Management of Oropharyngeal Cancer in the HPV Era Definitive

cancercontroljournal.org Vol. 23, No. 3, July 2016

H. LEE MOFFITT CANCER CENTER & RESEARCH INSTITUTE, AN NCI COMPREHENSIVE CANCER CENTER

Management of Oropharyngeal Cancer in the HPV Era Arash O. Naghavi, MD, Tobin J. Strom, MD, Kamran A. Ahmed, MD, et al

Definitive Radiotherapy for Squamous Cell Carcinoma of the Glottic Larynx William M. Mendenhall, MD, Roi Dagan, MD, Curtis M. Bryant, MD, et al

Clinical Benefits of Proton Beam Therapy for Tumors of the Skull Base Kamran A. Ahmed, MD, Stephanie K. Demetriou, Mark McDonald, MD, et al

Management of BCC and SCC of the Head and Neck Tobin J. Strom, MD, Jimmy J. Caudell, MD, PhD, and Louis B. Harrison, MD

Treatment of Head and Neck Paragangliomas Kenneth Hu, MD, and Mark S. Persky, MD

Multidisciplinary Management of Salivary Gland Cancers Matthew J. Mifsud, MD, Jon N. Burton, MD, Andy M. Trotti, MD, et al

Cancer Control is included in Index Medicus/MEDLINE

Editorial Board Members

Editor: Conor C. Lynch, PhD Production Team: Assistant Member Lodovico Balducci, MD Veronica Nemeth Tumor Biology Senior Member Editorial Coordinator Program Leader, Senior Adult Oncology Program Kristen J. Otto, MD Sherri Damlo Moffitt Cancer Center Associate Member Medical Copy Editor Head and Neck and Endocrine Oncology Diane McEnaney Deputy Editor: Graphic Design Consultant Julio M. Pow-Sang, MD Michael A. Poch, MD Senior Member Assistant Member Associate Editor of Chair, Department of Genitourinary Oncology Genitourinary Oncology Educational Projects Director of Moffitt Robotics Program Jeffery S. Russell, MD, PhD John M. York, PharmD Moffitt Cancer Center Assistant Member Akita Biomedical Consulting Endocrine Tumor Oncology 1111 Bailey Drive Editor Emeritus: Elizabeth M. Sagatys, MD Paso Robles, CA 93446 John Horton, MB, ChB Assistant Member Phone: 805-238-2485 Professor Emeritus of Medicine & Oncology Pathology - Clinical Fax: 949-203-6115 Saïd M. Sebti, PhD E-mail: [email protected] Moffitt Cancer Center Senior Member Journal Advisory Committee: Drug Discovery For Consumer and Daniel A. Anaya, MD Bijal D. Shah, MD General Advertising Associate Member Assistant Member Information: Gastrointestinal Oncology Malignant Hematology Veronica Nemeth Aliyah Baluch, MD Lubomir Sokol, MD, PhD Editorial Coordinator Assistant Member Senior Member Cancer Control: Journal of the Moffitt Cancer Center Infectious Diseases Hematology/Oncology Dung-Tsa Chen, PhD 12902 Magnolia Drive – MBC-JRNL Hatem H. Soliman, MD Associate Member Tampa, FL 33612 Assistant Member Biostatistics Phone: 813-745-1348 Breast Oncology Fax: 813-449-8680 Hey Sook Chon, MD E-mail: [email protected] Assistant Member Jonathan R. Strosberg, MD Gynecological Oncology Assistant Member Gastrointestinal Oncology Jasreman Dhillon, MD Assistant Member Jennifer Swank, PharmD Pathology - Anatomic Clinical Pharmacist - Medical Oncology/Internal Medicine Jennifer S. Drukteinis, MD Associate Member Sarah W. Thirlwell, MSc, MSc(A), RN Diagnostic Radiology Nurse Director Clement K. Gwede, PhD Supportive Care Medicine Program Associate Member Eric M. Toloza, MD, PhD Health Outcomes and Behavior Assistant Member Sarah E. Hoffe, MD Thoracic Oncology Associate Member Nam D. Tran, MD Radiation Oncology Assistant Member John V. Kiluk, MD Neuro-Oncology Associate Member Jonathan S. Zager, MD Breast Oncology Senior Member Cancer Control is a member of Richard D. Kim, MD Cutaneous Oncology the Medscape Publishers’ Circle®, Associate Member an alliance of leading medical Gastrointestinal Oncology publishers whose content is Bela Kis, MD, PhD featured on Medscape Assistant Member (www.medscape.com). Diagnostic Radiology Rami Komrokji, MD Most issues and supplements of Associate Member Cancer Control are available at Malignant Hematology cancercontroljournal.org

Cancer Control: Journal of the Moffitt Cancer Center (ISSN 1073-2748) is published by H. Lee Moffitt Cancer Center & Research Institute, 12902 Magnolia Drive, Tampa, FL 33612. Telephone: 813-745-1348. Fax: 813-449-8680. E-mail: [email protected]. Internet address: cancercontroljournal.org. Cancer Control is included in Index Medicus ®/MEDLINE® and EMBASE®/ Excerpta Medica, Thomson Reuters Science Citation Index Expanded (SciSearch®) and Journal Citation Reports/Science Edition. Copyright 2016 by H. Lee Moffitt Cancer Center & Research Institute. All rights reserved. Cancer Control: Journal of the Moffitt Cancer Center is a peer-reviewed journal that is published to enhance the knowledge needed by professionals in oncology to help them minimize the impact of human malignancy. Each issue emphasizes a specific theme relating to the detection or management of cancer. The objectives of Cancer Control are to define the current state of cancer care, to integrate recently generated information with historical practice patterns, and to enlighten readers through critical reviews, commentaries, and analyses of recent research studies. Disclaimer: All articles published in this journal, including editorials and letters, represent the opinions of the author(s) and do not necessarily reflect the opinions of the editorial board, the H. Lee Moffitt Cancer Center & Research Institute, Inc, or the institutions with which the authors are affiliated unless clearly specified. The reader is advised to independently verify the effectiveness of all methods of treatment and the accuracy of all drug names, dosages, and schedules. Dosages and methods of administration of pharmaceutical products may not be those listed in the package insert and solely reflect the experience of the author(s) and/or clinical investigator(s).

July 2016, Vol. 23, No. 3 Cancer Control 189 Table of Contents

Editorial

Maximizing Cures and Preserving Function in Head and Neck Cancers 192 Louis B. Harrison, MD, and Jimmy J. Caudell, MD, PhD

Articles

Management of Oropharyngeal Cancer in the HPV Era 197 Arash O. Naghavi, MD, Tobin J. Strom, MD, Kamran A. Ahmed, MD, Michelle I. Echevarria, MD, Yazan A. Abuodeh, MD, Puja S. Venkat, MD, Jessica M. Frakes, MD, Louis B. Harrison, MD, Andy M. Trotti, MD, and Jimmy J. Caudell, MD, PhD

Definitive Radiotherapy for Squamous Cell Carcinoma of the Glottic Larynx 208 William M. Mendenhall, MD, Roi Dagan, MD, Curtis M. Bryant, MD, Robert J. Amdur, MD, and Anthony A. Mancuso, MD

Clinical Benefits of Proton Beam Therapy for Tumors of the Skull Base 213 Kamran A. Ahmed, MD, Stephanie K. Demetriou, Mark McDonald, MD, and Peter A.S. Johnstone, MD

Management of BCC and SCC of the Head and Neck 220 Tobin J. Strom, MD, Jimmy J. Caudell, MD, PhD, and Louis B. Harrison, MD

Treatment of Head and Neck Paragangliomas 228 Kenneth Hu, MD, and Mark S. Persky, MD

Multidisciplinary Management of Salivary Gland Cancers 242 Matthew J. Mifsud, MD, Jon N. Burton, MD, Andy M. Trotti, MD, and Tapan A. Padhya, MD

Departments

Pathology Report: Practical Issues for Retroperitoneal Sarcoma 249 Vicky Pham, MS, Evita Henderson-Jackson, MD, Matthew P. Doepker, MD, Jamie T. Caracciolo, MD, Ricardo J. Gonzalez, MD, Mihaela Druta, MD, Yi Ding, MD, and Marilyn M. Bui, MD, PhD

190 Cancer Control July 2016, Vol. 23, No. 3 Table of Contents

Departments, continued

Infectious Diseases Report: Primary Gangrenous Cutaneous Mold Infections in a 265 Patient With Cancer and Neutropenia Abraham Yacoub, MD, Kiran K. Soni, Lysenia Mojica, MD, Jane Mai, MD, Jamie Morano, MD, C. Wayne Cruse, MD, Ramon L. Sandin, MD, Sowmya Nanjappa, MBBS, MD, Chandrashekar Bohra, MBBS, Ganesh Gajanan, MBBS, and John N. Greene, MD Case Series: Diffuse Alveolar Hemorrhage in Acute Myeloid Leukemia 272 Sowmya Nanjappa, MBBS, MD, Daniel K. Jeong, MD, Manjunath Muddaraju, MD, Katherine Jeong, MD, Eboné D. Hill, MD, and John N. Greene, MD Case Series: Marijuana Smoking in Patients With Leukemia 278 Sara Khwaja, MD, Abraham Yacoub, MD, Asima Cheema, MD, Nancy Rihana, MD, Robin Russo, MD, Ana Paula Velez, MD, Sowmya Nanjappa, MBBS, MD, Ramon L. Sandin, MD, Chandrashekar Bohra, MBBS, Ganesh Gajanan, MBBS, and John N. Greene, MD

Special Report: Platelet-to-Lymphocyte Ratio and Use of NSAIDs During the Perioperative Period 284 as Prognostic Indicators in Patients With NSCLC Undergoing Surgery Brenda M. Lee, MD, Andrea Rodríguez, MD, Gabriel Mena, MD, Vijaya Gottumukkala, MB, MD, Reza J. Mehran, MD, David C. Rice, MB, Lei Feng, MS, Jun Yu, MS, and Juan P. Cata, MD

Special Report: Resection of the First Rib With Preservation of the T1 Nerve Root in 295 Pancoast Tumors of the Lung Andreas K. Filis, MD, Lary A. Robinson, MD, and Frank D. Vrionis, MD, PhD

Special Report: Feasibility, Preliminary Efficacy, and Lessons Learned From a Garden-Based 302 Lifestyle Intervention for Cancer Survivors Colleen K. Spees, PhD, RDN, Emily B. Hill, Elizabeth M. Grainger, PhD, RDN, Jackie L. Buell, PhD, RDN, Susan E. White, PhD, Matthew D. Kleinhenz, PhD, and Steven K. Clinton, MD, PhD

Special Report: Hypofractionated Stereotactic Radiotherapy for Auditory Canal or Middle Ear Cancer 311 Taro Murai, MD, Shin-Etsu Kamata, MD, Kengo Sato, MD, Kouki Miura, MD, Mitsuhiro Inoue, Naoki Yokota, MD, Seiji Ohta, MD, Michio Iwabuchi, MD, Hiromitsu Iwata, MD, and Yuta Shibamoto, MD

Ten Best Readings Relating to Head and Neck Cancers 317

About the art in this issue: Born in 1943, Arthur Spencer Pina de Alba has always been passionately involved in the arts. After completing his studies at Shaker Heights High School, Arthur was given a National Arts Award and 3 Gold Key Awards. He earned a scholarship to study at Chounard Academy, and then went on to earn a bachelor of fine arts at the California Institute of the Arts. After college he established an all-encompassing fine-arts production company. His multimedia enterprise worked with private parties to design and produce large-scale murals, sculptures, and trompe l’oeil, as well as illustrations and storyboards for the film industry. Arthur's endeavors have also included art direction for album cover artwork on such documentaries as Gravity is My Enemy and Conquistadores, which was narrated by Orson Welles. Some of his clients have included Desi Arnez Jr and fashion designer Richard Blackwell. During his time in Los Angeles, Arthur held solo exhibitions at the Crest and Ryder Galleries. After living in Los Angeles, he moved to France, where he immediately made his mark at the Academy Royal de Paris. In Paris, Arthur became the art director for Click, a magazine devoted to fashion and photography. He also produced video clips and worked at Studio Puteaux. By appointment of the director of the Ballet de Opera in Paris, he was the only artist to paint and sculpt Rudolf Nureyev and Patrick Dupont and the ballerinas and belle tomes of the opera house. Arthur currently resides in the French Riviera, where he continues to paint and sculpt in bronze and other materials. He now paints on an electronic tablet and prints his work on special paper and canvas, and he travels across Europe to create his art. For more information or to contact Arthur, visit www.arthurpinadealba.com.

Cover and Articles: Artwork courtesy of Arthur Pina de Alba. www.arthurpinadealba.com.

July 2016, Vol. 23, No. 3 Cancer Control 191 Editorial

Maximizing Cures and Preserving Function in Head and Neck Cancers

Head and neck cancers are a fascinating collection of nificant morbidity. The authors underscore that an ex- tumors that challenge the multidisciplinary oncology perienced multidisciplinary team should manage these team to maximize rates of cure while preserving some tumors because these specialists can present patients of the most vital functions in a patient’s day-to-day life. with all the available care options, including observa- Advances in tumor biology and medical technology tion alone. Drs Hu and Persky also provide algorithmic have contributed to our understanding of these diseas- plans of care for radiotherapy or surgery in optimally es, helping us develop better treatments. selected patients. This issue of Cancer Control highlights the role Dr Mifsud and colleagues provide a review on the of radiotherapy with or without chemotherapy and broad and complex topic of salivary gland cancers, an surgery for the management of selected neoplasms interesting group of tumors typically managed with of the head and neck. We have included several top- surgery. The role of adjuvant radiotherapy and chemo- ics that highlight areas of our evolving understanding therapy has evolved in this setting. and goals of organ/function preservation in this oncol- We hope you find this issue of Cancer Control to ogy setting. be helpful in your practice of head and neck oncology. Dr Naghavi and colleagues provide an overview of the management of oropharyngeal cancer in the era Louis B. Harrison, MD of human papillomavirus (HPV). The observation that Department Chair, Radiation Oncology HPV-associated oropharyngeal cancer confers a better H. Lee Moffitt Cancer Center & Research Institute prognosis has increased our understanding into new Tampa, Florida areas of management. A key idea is the focus on treat- [email protected] ment deintensification — a rare opportunity in oncology to maximize the rates of cure with less therapy and Jimmy J. Caudell, MD, PhD fewer rates of toxicity — to maximize quality of life. Section Head, Head and Neck Cancer Another example of organ/function preservation Department of Radiation Oncology is the management of squamous cell carcinoma (SCC) H. Lee Moffitt Cancer Center & Research Institute of the glottic larynx. Dr Mendenhall and colleagues Tampa, Florida highlight treatment outcomes for SCC in a large popu- [email protected] lation of patients from their institution as well as from published reports from the medical literature. References Dr Ahmed and coauthors describe proton beam 1. Centers for Disease Control and Prevention. Skin cancer statistics. https://www.cdc.gov/cancer/skin/statistics/. Accessed August 8, 2016. radiation as a highly conformal approach for the man- 2. American Cancer Society. Basal and squamous cell skin cancers. agement of skull-base tumors. In this setting, proton Published April 1, 2016. Revised May 10, 2016. Accessed August 8, 2016. 3. Iannacone MR, Wang W, Stockwell HG, et al. Patterns and timing therapy can be used to exploit its physical dose-distri- of sunlight exposure and risk of basal cell and squamous cell carcinomas bution characteristics in an advantageous way in nu- of the skin--a case-control study. BMC Cancer. 2012;12:417. merous critical structures, including the skull base, cranial nerves, visual apparatus, and brainstem. Dr Strom and colleagues review treatment options for cutaneous basal cell carcinoma and SCC. Skin cancers make up the most common form of human malignancy, and most cases occur on the sun-exposed skin of the head and neck.1-3 Although patients with skin cancer commonly undergo surgery, radiotherapy is also a curative option and offers better cosmetic outcomes in select patients. Thus, the role of radiotherapy in cutaneous oncology is growing. Drs Hu and Persky present a balanced approach to paragangliomas of the head and neck. Although they are generally benign, these tumors can also cause sig-

192 Cancer Control July 2016, Vol. 23, No. 3 Radiation Oncology Program H. Lee Moffitt Cancer Center & Research Institute

The Radiation Oncology Program provides comprehensive, multimodality, patient-oriented care for all types and stages of cancer. Special attention is given to technological advancement, quality of life, and education. The goals of the program are to provide interdisciplinary care for new and established patients; develop and conduct clinical research trials that focus on radiotherapy as the modality of treatment; establish collaborative links with appropriate science programs at Moffitt Cancer Center; utilize state-of-the-art equipment in the most efficient and effective manner; and provide educational programs and training for medical students, house staff, physicians, fellows, and community physicians. Our specialty-trained radiation oncologists are knowledgeable in both standard and unique methods of radiotherapy. The Radiation Oncology Program holds weekly conferences to discuss cases where physician input is obtained and shared. These conferences also function as an educational forum for hospital staff involved in the care of patients receiving radiotherapy. The physicians provide direction to the staff in all technical aspects of radiotherapy. Quality of care of all patients is the program’s first priority and goal. This leads to highly expert, professional, quality patient care and very high rates of patient satisfaction.

CHAIRMAN Louis B. Harrison, MD

Vice Chairman Peter A. Johnstone, MD

Chief of Director of DIVISION ADMINISTRATOR AND DIRECTOR OF Medical Physics Clinical Research CLINICAL PHYSICS Eduardo G. Moros, PhD Javier F. Torres-Roca, MD Stuart Wasserman, MS

Faculty Jimmy J. Caudell, MD, PhDa Jessica Frakes, MD Sungjune Kim, MD, PhD Javier F. Torres-Roca, MD Roberto Diaz, MD, PhDb Louis B. Harrison, MD Michael E. Montejo, MD Andy M. Trotti, MD Thomas J. Dilling, MDc Sarah E. Hoffe, MDe Amber G. Orman, MD Kosj Yamoah, MD, PhD Vladimir Feygelman, PhD Peter A. Johnstone, MDf Brad A. Perez, MD Michael Yu, MD, ScMg Daniel C. Fernandez, MD, PhDd Kujtim Latifi, PhD Jacob G. Scott, MD Geoffrey Zhang, PhD

Technology Administrative/ Education AND Research/ Support Clinical Support Training Translation Physicists Physician assistants Therapists Clinical researchers Dosimetrists Advanced registered Medical students Trial coordinators Therapists nurse practitioners Residents Data managers Registered nurses/ Fellows Basic scientists licensed practical nurses Postgraduates Doctoral physicists Patient representatives International students Dietitians Social workers aSection Head for Head and Neck Cancer, bSection Head for Breast Oncology, cSection Head for Thoracic Oncology, dSection Head for Brachytherapy, eSection Head for GI Oncology, fSection Head for GU Oncology, gSection Head for Neuro-Oncology and Palliative Care. GI = gastrointestinal, GU = genitourinary.

July 2016, Vol. 23, No. 3 Cancer Control 193 Research

The Radiation Oncology Program participates as a full-member Radiation Therapy Oncology Group (RTOG) institution and has many national protocols open for enrollment at Moffitt Cancer Center. Faculty members are principal investigators or co-investigators on several RTOG trials and have investigator-initiated trials that cover the spectrum of human cancer. The program performs cancer control research in the areas of screening, quality of life, survivorship/late effects, toxicity assessment and intervention, patient decision-making, older adult/comorbidity research, and under-served populations. The department is also pioneering the Future of Radiation Therapy project, which focuses on developing personalized medicine programs specific to radiotherapy and cancer care.

To schedule a patient appointment with a physician in the Radiation Oncology Program, call the New Patient Appointment Center at 813-745-3980 or 888-860-2778 (during normal business hours). For information about clinical trials, call 813-745-4106.

www.moffitt.org

194 Cancer Control July 2016, Vol. 23, No. 3 Department of Head and Neck and Endocrinology H. Lee Moffitt Cancer Center & Research Institute

Moffitt Cancer Center’s Department of Head and Neck and Endocrinology features a multispecialty team of physicians, surgeons, medical oncologists, radiation oncologists, and other professionals who have world-class expertise in the diagnosis and treatment of malignant and benign tumors of the head and neck. The comprehensive approach of the department is evidence based, and it offers advanced screening, diagnosis, and treatment options, including a robust clinical trials program through which patients have access to promising new treatments before they are made widely available in other settings. The department is also a high-volume surgery center, and its head and neck surgeons are regarded among the best in the nation.

CHAIRMAN Christine Chung, MD

Endocrinology Medical Oncology Surgery Bryan McIver, MD, PhD, Christine Chung, MD Thomas V. McCaffrey, MD, PhD Program Leader, Endocrinology Jeffery S. Russell, MD, PhD Kristen Otto, MD Howard Lilienfeld, MD Tapan A. Padhya, MD John Tourtelot, MD Ashley Hodes, MD

Program Support Clinical Support Community and Education and Department Administrator Clinical Operations Manager Patient Programs Training Clifton R. Scott, MHA Yasmin Sterling, RN Physician Marketing Medical students Strategy Manager Residents Registered nurses Jason D. Bever Fellows Medical assistants International scholars Patient services specialists

Associated Clinical Faculty

Radiation Oncology Radiology Pathology Jimmy J. Caudell, MD, PhD John Arrington, MD Marino E. Leon, MD Louis B. Harrison, MD Gregory Carney, MD Barbara A. Centeno, MD Andy M. Trotti, MD F. Reed Murtagh, MD Laila Khazai, MD

Senior Adult Oncology Thoracic Otorhinolaryngology Julie Kish, MD Tawee Tanvetyanon, MD Kestutis P. Boyev, MD

July 2016, Vol. 23, No. 3 Cancer Control 195 Research

As a recognized leader in performing groundbreaking cancer research, Moffitt Cancer Center has a nation- ally acclaimed clinical trials program with active protocols for various types of head and neck cancers. For instance, our researchers are studying: • New surgical techniques • New radiation delivery methods • New chemotherapy drugs, drug combinations, and dosages • Emerging treatments, such as gene therapy • New ways to prevent cancer recurrence

Examples of active clinical studies at Moffitt include:

Clinical Trial No. Study Title

17716 A Biomarker-Related Approach to Screening for HPV-Related Oropharyngeal Cancer

17799 A Phase I Dose Escalation Trial of Stereotactic Body Radiotherapy and Concurrent Cisplatin for Re-Irradiation of Unresectable, Recurrent Squamous Cell Carcinomas of the Head and Neck

17902 Phase 1 Study of the Safety and Tolerability of ATR-101 in Adrenocortical Carcinoma

17985 A Phase 1/2 Study Exploring the Safety, Tolerability, and Efficacy of INCB024360 in Combination With MEDI4736 in Subjects With Selected Advanced Solid Tumors

17986 Phase 1 Study of Mogamulizumab (KW-0761) in Combination With MEDI4736 and Mogamulizumab in Combination With Tremelimumab in Subjects With Advanced Solid Tumors

18227 Randomized Phase II and Phase III Studies of Individualized Treatment for Nasopharyngeal Carcinoma Based on Biomarker Epstein Barr Virus (EBV) Deoxyribonucleic Acid (DNA)

18502 A First-in-Human Study of Repeat Dosing With REGN2810, a Monoclonal, Fully Human Antibody to Pro- grammed Death - 1 (PD-1), as Single Therapy and in Combination With Other Anti-Cancer Therapies, in Patients With Advanced Malignancies

18513 Multicenter Phase II Study of Nivolumab in Previously Treated Patients With Recurrent and Metastatic Nasopharyngeal Carcinoma

18527 Non-Comparative, Two-Cohort, Single-Arm, Open-Label, Phase 1/2 Study of Nivolumab (BMS-936558) in Subjects With Virus-Positive and Virus-Negative Solid Tumors

18572 A Phase III Randomized, Open-Label, Multi-Center, Global Study of MEDI4736 Alone or in Combination With Tremelimumab vs Standard of Care in the Treatment of First-Line Recurrent or Metastatic Squamous Cell Head and Neck Cancer Patients (KESTREL)

NCT01220583 Radiation Therapy With or Without Chemotherapy in Treating Patients With High-Risk Malignant Salivary Gland Tumors That Have Been Removed by Surgery

NRG-HN002 Reduced-Dose Intensity-Modulated Radiation Therapy With or Without Cisplatin in Treating Patients With Advanced Oropharyngeal Cancer

To schedule a patient appointment with a physician in the Department of Head and Neck and Endocrinology, call the New Patient Appointment Center at 813-745-7360 or 1-888-860-2778 (during normal business hours).

www.MOFFITT.org

196 Cancer Control July 2016, Vol. 23, No. 3 The prevalence of HPV-positive OPC

continues to increase at an epidemic rate.

Artwork courtesy of Arthur Pina de Alba. www.arthurpinadealba.com.

Management of Oropharyngeal Cancer in the HPV Era Arash O. Naghavi, MD, Tobin J. Strom, MD, Kamran A. Ahmed, MD, Michelle I. Echevarria, MD, Yazan A. Abuodeh, MD, Puja S. Venkat, MD, Jessica M. Frakes, MD, Louis B. Harrison, MD, Andy M. Trotti, MD, and Jimmy J. Caudell, MD, PhD

Background: Historically, oropharyngeal cancer (OPC) has been attributed to risk factors such as smoking and alcohol use. The increased incidence of OPC has been driven by human papillomavirus (HPV) infection. Methods: A search of the literature involving HPV infection and OPC was performed, along with a search of ongoing clinical trials regarding HPV-positive OPC. Results: This review summarizes the differences in epidemiology and prognosis of HPV-positive OPC compared with non–HPV-related OPC. It will also discuss use of de-escalating treatment to minimize toxicity while maintaining excellent outcomes. Disease management is also addressed, including prevention and follow-up recommendations for this cohort of patients. Conclusions: HPV-positive OPC is a distinct disease, and efforts should be made to personalize its management. Preventive measures and vaccinations, along with de-escalation of treatment, may help optimize outcomes in this population. Introduction carcinogenesis driven by infection with human papil- In 2016, an estimated 48,330 new cases of head and lomavirus (HPV).4,5 The incidence of HPV-related OPC neck cancers will be diagnosed in the United States has increased at an epidemic rate.4,5 HPV status is now and 9,570 deaths are likely to be reported.1 Worldwide, recognized as the single most prognostic factor for out- head and neck cancers are the eighth leading cause of come in OPC, although it has not yet been included in cancer-related deaths.2 Historically, head and neck can- staging or treatment considerations.6-12 cers were largely attributed to the use of tobacco and Patients with HPV-positive OPC are diagnosed at alcohol.3 However, despite decreasing tobacco use, the a younger age and have significantly better overall out- incidence of oropharyngeal cancer (OPC) continues to comes, making the decision to undergo de-escalation rise. Today, up to 72% of persons with OPC lack the of radiotherapy (RT), chemotherapy, or surgery particu- traditional risk factors of tobacco and alcohol use, with larly important in this population.13-15 Alternatively, the addition or substitution of other novel therapies such as immunotherapy or targeted agents in combination with From the H. Lee Moffitt Cancer Center & Research Institute, Department of Radiation Oncology, Tampa, Florida. RT may be ideal in this virally driven cancer. Consider- Address correspondence to Arash O. Naghavi, MD, Department of ation of preventive measures, such as vaccination to de- Radiation Oncology, Moffitt Cancer Center, 12902 Magnolia Drive, crease HPV incidence, is an important measure.16 Tampa, FL 33612. E-mail: [email protected] Submitted February 7, 2016; accepted April 5, 2016. Epidemiology No significant relationships exist between the authors and the companies/organizations whose products or services may be referenced HPV infection has been historically associated with in this article. some anogenital carcinomas, but it has also become

July 2016, Vol. 23, No. 3 Cancer Control 197 recognized as a risk factor for head and neck cancers.17 spective trials. HPV-positive OPC is prognostic for im- The overall worldwide prevalence of HPV infection in proved outcomes compared with HPV-negative OPC in persons with head and neck cancers is approximately the setting of postoperative RT,7 concurrent chemora- 26%, with the highest percentage among those with diation,10,29,30 RT alone,6,31 and induction chemothera- OPC (35.6%), and OPC is most commonly attributed py followed by chemoradiation.32,33 Most studies show to the HPV-16 strain (86.7%).18 Other strains, such as that HPV infection is independently prognostic for im- HPV-18, -31, -33, -35, -39, -45, -51, -52, -56, -58, and -59, proved disease-free survival (DFS) and overall survival have also been implicated in head and neck carcino- (OS), but some also show a benefit in locoregional fail- genesis.19 The prevalence of HPV in head and neck ure (LRF) and distant metastasis (Table 1).10,14,29,30-34 cancers differs based on geographical site, with North Radiation Alone: In a phase 3 trial conducted by America having a higher prevalence of HPV-associated the Radiation Therapy Oncology Group (RTOG), pa- OPC than Europe and Asia.18 From 1988 to 2004, the tients with head and neck squamous cell carcinoma United States saw a 50% decrease in the incidence of (SCC) were randomized to 4 different fractionation HPV-negative head and neck cancers and a 225% in- regimens: standard fractionation (70 Gy), concomitant crease in HPV-positive OPC.20 If this trend continues boost (72 Gy), split-regimen hyperfractionation (67.2 in the United States, the rate of HPV-positive OPC is Gy), or hyperfractionation (81.6 Gy).31 From this tri- expected to surpass cervical cancer by 2020.20 Patients al, Gillison et al31 reviewed 190 evaluable patients by with HPV-positive OPC are more likely to be younger p16 expression, of which 75 (39.5%) were p16 posi- (by 4–10 years),10,21 male (5:1 male to female),20,22 and tive. Positive p16 expression predicted improvements white.20,23 in rates of 5-year LRF (28.9% vs 54.9%; P < .001), DFS The rising incidence of HPV-associated OPC is (43.6% vs 19.0%), and OS (49.0% vs 19.6%), but p16 ex- likely a consequence of changing sexual practices. pression showed no difference in distant metastasis HPV is a known sexually transmitted infection: Oral (11.0% vs 13.0%; P = .71) or second primary tumors HPV prevalence appears during the ages of sexual ac- (13.8% vs 11.4%; P = .4).31 tivity initiation, with 1.5% at 12 to 15 years of age, 3.3% Lassen et al14 evaluated the relationship of p16 sta- at 16 to 20 years, and 4.5% at 21 years or older.24 As part tus from 2 previous phase 3 trials (Danish Head and of the National Health and Nutrition Examination Sur- Neck Cancer Group [DAHANCA] 5 and 7). DAHANCA vey, a study found a prevalence of oral HPV infection 5 enrolled 414 patients who were randomized to con- as high as 6.9% in participants aged 14 to 69 years.22 ventionally fractionated RT (5 fractions/week) with or Oral HPV infection has been associated with an in- without the hypoxic radiosensitizer nimorazole.35 Be- creasing number of lifetime oral and vaginal sex part- cause this trial showed improved outcomes with the ners and, in men aged 18 to 23 years, was also associ- addition of nimorazole, the agent was added to both ated with open-mouth kissing.25 In a case-control study arms of DAHANCA 7, which randomized 786 patients of HPV infection and OPC, D’souza et al26 showed that to conventional RT (5 fractions/week) vs acceler- OPC was associated with an increasing lifetime num- ated RT (6 fractions/week) to a dose of 66 to 68 Gy.14 ber of oral sex partners (≥ 6 partners; odds ratio [OR] The retrospective analysis of the 2 trials consisted 3.4; 95% confidence interval [CI], 1.3–8.8; P = .009) of 336 patients with OPC evaluable by p16 assays, of and vaginal sex partners (≥ 26 partners; OR 3.1; 95% which 120 (36%) were positive for p16.14 In this study, CI, 1.5–6.5; P = .002), as well as oral infection with any p16-positive tumors were smaller in size (P < .0002), type of HPV (OR 12.3; 95% CI, 5.4–26.4). more likely to have nodal spread (P < .02), and had an After HPV is transmitted via sexual contact, the improved rate of 5-year locoregional tumor control virus infects the epithelial lining of the oropharynx. (72% vs 38%; P < .001).14 On multivariate analysis, the The epithelial crypts that cover the base of the tongue presence of p16-positive OPC independently pre- and tonsils are more susceptible to infection be- dicted a benefit in LRF (hazard ratio [HR] 0.29; 95% cause of easier access to the basal epithelial layer and CI, 0.19–0.44), tumor site failure (HR 0.24; 95% CI, they serve as a viral reservoir.27 In a healthy adult, 0.14–0.42), disease-specific death (HR 0.28; 95% CI, 65% to 80% of oral HPV infections will clear within 0.17–0.41), and OS (HR 0.31; 95% CI, 0.23–0.41; 12 months, but a significant decline in clearance rates P < .0001).14 In a larger subset that included data from is seen in immunocompromised patients.28 all 794 patients with head and neck SCC analyzed from the DAHANCA 6 and 7 trials, accelerated frac- Outcomes in Oropharyngeal Cancer in tionation was observed to have an LRF benefit in both Association With Human Papillomavirus Status the p16-positive and p16-negative groups, along with a Retrospective Analysis disease-specific survival benefit in p16-positive tumors.6 The initial retrospective data showing the prognostic Concurrent Chemotherapy: In a phase 3 tri- value of HPV infection in OPC led multiple groups to al, the RTOG randomized patients with stage III/IV investigate HPV status in previously completed pro- head and neck SCC to standard fractionation RT

198 Cancer Control July 2016, Vol. 23, No. 3 with 3 cycles of cisplatin or accelerated RT with 2 cy- may have been misclassified as being negative for cles of cisplatin.10 No difference was seen in outcomes HPV, thus accounting for HPV-16-negative study par- between the 2 arms; therefore, the arms were pooled ticipants with elevated p16 expression.10 This study in this retrospective analysis. Of the 721 patients en- was one of the first to use recursive partitioning rolled, 323 had evaluable OPC specimens: 206 (63.8%) analysis to stratify study participants with OPC into were positive for HPV infection (via in situ hybrid- risk groups based on stage, smoking status, and p16 ization).10 HPV-positive tumors had improved rates of status.10 3-year LRF (13.6% vs 35.1%; P < .001), DFS (73.7% vs In another phase 3 trial conducted by the RTOG, 43.4%; P < .001), and OS (82.4% vs 57.1%; P < .001).10 patients with stage III/IV head and neck SCC were ran- HPV status was not associated with distant metastasis domized to accelerated RT (mainly 70 Gy in 6 weeks at 3 years (8.7% vs 14.6%; P = .23).10 In 214 study par- or 72 Gy in 6 weeks via concomitant boost) and con- ticipants (66%), this study also showed that elevated current cisplatin (100 mg/m2 on days 1 and 22) with p16 expression (via immunohistochemistry) was more or without cetuximab (400 mg/m2 loading dose the prognostic for both DFS and OS than HPV status.10 The week before RT, then 250 mg/m2 per week during authors suggest that HPV-16 in situ hybridization has RT).29 No difference was seen in outcome between a high sensitivity rate for HPV-16, but non–HPV-16 the 2 arms. Of the 891 study participants analyzed,

Table 1. — Retrospective Analysis of Select Cooperative Trials of OPC Study RCT Location No. of Method HPV+ Treatment, Systemic RT Follow- OS HPV+ 5-y HPV+ LRF Phase Patients of HPV OPC, Type (%)b Therapy Dose, Up, Rate HR DFS HR (Oro- Detec- % Gy mo (95% pharynx) tion CI) Ang10 3 United 323 ISH 63.8 Cisplatin + Concurrent: (A) 70 57.6 3-y: 0.42 3-y: 0.49 3-y: States p16 RT (SFX) Cisplatin (SFX) 82.4% (0.27– 73.7% (0.33– 13.6% (50) (B) (d 1 and 22) (B) 72 vs 0.66) vs 0.74) vs Cisplatin + Cisplatin (AFX-C) 57.1% P < .001 43.4% P < .001 35.1% RT (AFX-C) (d 1 and 22) P < .001 P < .001 P < .001 (50) Ang29 3 United 321 p16 73.2 Cisplatin + Concurrent: 70–72 28.8 3-y: 0.32 3-y: 0.49 3-y: States RT (49) Cisplatin (AFX-C) 85.6% (0.20– 72.8% (0.33– 17.3% Cisplatin + (d 1 and 22) vs 0.51) vs 0.71) vs cetuximab + Cisplatin 60.1% P < .001 49.2% P < .001 32.5% RT (51) (d 1 and 22) + P < .001 P < .001 P < .001 cetuximab ( loading dose followed by wkly dose) Fakhry32 2 United 62 ISH 61.0 Induction Induction: 70 (SFX) 39.1 — 0.39 — 0.38 — States p16 chemo- Carboplatin (0.15– (0.12– therapy + (AUC 6) and 1.05) 1.15) chemoRT paclitaxel P = .06 P = .09 (intravenous) for 2 cycles (d 1 and 21) Concurrent: paclitaxel (wkly) Gillison31 3 United 190 p16 39.5 SFX (22) None 70 (SFX) 111.6 5-y: 0.61 5-y: 0.65 5-y: States HFX (29) 81.6 49% (0.42– 43.6% (0.45– 28.9 vs 0.89) vs 0.93) vs AFX-S (26) (HFX) 19.6% P = .02 19% P = .02 54.9% AFX-C (23) 67.2 (AFX-S) P = .02 P < .001 72 (AFX-C) Lassen14 3 Denmark 336 p16 36.0 RT (SFX) 78% 66–68 30.4 — 0.31 — — 5-y: (65) nimorazole 5 frac- (0.23– 28% RT (AFX) tions/wk 0.41) vs 62% (35) 6 frac- P < .0001 tions/wk P < .001

continued on next page

July 2016, Vol. 23, No. 3 Cancer Control 199 Table 1. — Retrospective Analysis of Select Cooperative Trials of OPC, continued Study RCT Location No. of Method HPV+ Treatment, Systemic RT Follow- OS HPV+ 5-y HPV+ LRF Phase Patients of HPV OPC, Type Therapy Dose, Up, mo Rate HR DFS HR (Oro- Detec- % (%)b Gy (95% pharynx) tion CI) Lohaus34,a — Germany 126 PCR 48.0 Surgery Concurrent 64 47.0 — 0.36 — — HR 0.09 + PORT + cisplatin (SFX) (0.17– (0.01– cisplatin 0.73) 0.74) P < .01 P = .03 Posner33 3 United 111 PCR 50.5 Induction TPF Induction: 70–74 83.0 5-y: 0.2 5-y: — 5-y: States + chemoRT 3 cycles (SFX) 82% (0.1– 78% 13% (49) of PF vs 35% 0.38) vs 28% vs Induction PF (daily for P < .0001 P < .0001 P < .001 42% + chemoRT 4–5 d) ± P < .0006 (51) docetaxel Concurrent: carboplatin (1.5 AUC wkly for 7 wk) Rischin30 3 United 185 PCR 57.0 Concurrent Concurrent: 70 29.0 2-y: 0.43 2-y: 0.39 States p16 standard Cisplatin (SFX) 91% (0.02– 87% (0.20– Canada cisplatin (d 1 of wk 1, vs 0.93) vs 72% 0.74) + RT 4, and 7) 74% Australia P = .031 P = .003 P = .003 Concurrent Cisplatin P = .004 New lower-dose (d 1 of wk 1, Zealand cisplatin + 4, and 7) + Europe tirapazamine tirapazamine + RT (d 1 of wk 1, 4, and 7; 3x/wk for wk 2 and 3) aMulticentered retrospective trial. bPercentage in each subgroup. AFX-C = accelerated fractionation with concomitant boost, AFX-S = accelerated fractionation with split radiotherapy, AFX = accelerated fractionation, AUC = area under the curve, CI = confidence interval, DFS = disease-free survival, HPV = human papillomavirus, HR = hazard ratio, HFX = hyperfractionation, ISH = in situ hybridization, LRF = locoregional failure, OPC = oropharyngeal cancer, OS = overall survival, PCR = polymerase chain reaction, PF = cisplatin/ fluorouracil, PORT = postoperative radiotherapy, RCT = randomized controlled trial, RT = radiotherapy, SFX = standard fractionation, TPF = docetaxel/ cisplatin/fluorouracil.

321 (36%) had OPC assessable by p16 assays, and, of Overall, 2-year DFS (87% vs 72%; P = .003) and OS these, 235 (73.2%) were p16 positive.29 Study patients (91% vs 74%; P = .004) benefits were seen in p16-pos- with p16-positive OPC had improved rates of 3-year itive study patients.30 When the 2 arms were individu- LRF (17.3% vs 32.5%; P < .001), DFS (72.8% vs 49.2%; ally compared, the OS benefit of p16-positive tumors P < .001), and OS (85.6% vs 60.1%; P < .001).29 This was significant in the cisplatin-alone arm (89% vs 68%; was one of the first studies to show a significant as- P = .021) but no longer significant in the group receiv- sociation between p16 and distant metastasis rate ing tirapazamine (94% vs 80%; P = .36).30 A trend for (3 years: 6.5% vs 17.0%; P = .005).29 improved locoregional control was seen in the tira- The Trans Tasman Radiation Oncology Group con- pazamine arm among study participants negative for ducted a phase 3 trial that randomized patients with p16 (HR 0.33; 95% CI, 0.90–1.24; P = .13).30 stage III/IV head and neck SCC to standard RT (70 Gy Induction Chemotherapy: The Eastern Coop- in 35 fractions in 7 weeks) concurrently with either erative Oncology Group (ECOG) conducted a phase 2 standard bolus cisplatin (100 mg/m2 on day 1 of weeks trial to determine outcomes with induction chemora- 1, 4, and 7) or decreased dosage of cisplatin (75 mg/m2 diotherapy in patients with resectable stage III/IV SCC on day 1 of weeks 1, 4, and 7) in addition to tirapaza- of the larynx and oropharynx.32 Of the 111 patients en- mine (290 mg/m2 on day 1 of weeks 1, 4, and 7 and rolled, Fakhry et al32 was able to evaluate 96 patients, 160 mg/m2 3 times a week for weeks 2 and 3).30 A total 62 (64.5%) of whom had OPC. Study patients under- of 465 study patients with OPC received treatment, and went 2 cycles of induction chemotherapy (carboplatin/ 185 (40%) specimens were evaluable and received at paclitaxel) followed by concurrent RT (70 Gy in 35 frac- least 60 Gy of radiation; of these, 106 (57%) were posi- tions in 7 weeks) and weekly paclitaxel.32 HPV status, tive for p16 expression (via immunohistochemistry).30 defined by in situ hybridization and p16 status, was

200 Cancer Control July 2016, Vol. 23, No. 3 positive in 38 (61%) of the study patients with OPC.32 ception of acneiform rash and infusion reactions, ce- Study patients positive for HPV infection had a higher tuximab was well tolerated with little additional tox- response rate to induction chemotherapy (82% vs 55%; icity.40 This study did not evaluate outcome in terms P = .01) and rate of 2-year OS (95% vs 62%; P = .005).32 of HPV status but did show an enhanced benefit in Posner et al33 conducted a phase 3 trial that ran- younger age (< 65 years) and those with OPC, a find- domized patients with stage III/IV head and neck ing that suggests an agent could be used to effec- SCC into 2 arms of induction chemotherapy docetaxel tively treat patients with HPV-positive OPC.40 A ret- 75 mg/m2/cisplatin 100 mg/m2/fluorouracil 1000 mg/m2 rospective analysis of this trial showed a benefit with for 4 days vs cisplatin 100 mg/m2/fluorouracil 1000 mg/m2 cetuximab in both p16-positive and p16-negative study for 5 days every 3 weeks for 3 cycles, followed by concur- patients with respect to rates of locoregional control, rent RT (70–74 Gy in 7–7.5 weeks) and weekly carbopla- DFS, and OS.41 Previous studies have shown that epi- tin (area under the curve = 1.5).33 Of the 264 study pa- dermal growth factor receptor has a higher expres- tients with OPC, 111 of those specimens were evaluable; sion in those who smoke42 and is inversely related to of those, 56 (50%) were positive for HPV-16 on reverse HPV status,43,44 which may explain the limited efficacy transcriptase–polymerase chain reaction.33 Study pa- of cetuximab in HPV-positive cervical cancers.45,46 In tients positive for HPV infection had improved rates of a trial conducted by the RTOG, the addition of cetux- 5-year LRF (13% vs 42%; P = .0006), DFS (78% vs 28%; imab to chemoradiotherapy showed no benefit in the P < .001), and OS (82% vs 35%; P < .001).33 Although setting of HPV-positive OPC.29 study patients with HPV-positive OPC had a lower rate Several ongoing phase 3 trials are investigating of distant metastasis (5% vs 11%), this result was not outcomes in patients with stage III/IV HPV-positive significant.33 OPC and are comparing standard concurrent cisplatin with cetuximab (NCT01855451, NCT01302834, Meta-Analysis NCT01874171). Table 2 details several ongoing trials Two meta-analyses examined the relationship between (as of publication) studying the effect of de-escalation HPV and head and neck SCC outcomes. Ragin et al36 in the setting of HPV-positive OPC (NCT01302834, evaluated 23 studies and showed that HPV-positive NCT01874171, NCT01855451, NCT01530997, head and neck SCC was associated with an improved NCT01088802, NCT01891695, NCT02254278, outcome when compared with HPV-negative head and NCT01706939, NCT01084083, NCT01898494, neck SCC with regard to DFS (HR 0.62; 95% CI, 0.5–0.8) NCT02215265, NCT01687413). One study, RTOG 1016, and OS (HR 0.85; 95% CI, 0.7–1.0). A more recent meta- enrolled patients with p16-positive stage III/IV OPC- analysis that included 42 studies found a more pro- SCC (excluding T1–2 N1 or M1 disease) treated with ac- nounced DFS (HR 0.41; 95% CI, 0.27–0.64) and OS celerated intensity-modulated radiotherapy (IMRT; (HR 0.43; 95% CI, 0.35–0.52) benefit in patients with 70 Gy in 35 fractions in 6 weeks) and then randomized HPV-positive head and neck SCC.37 In this meta-analy- them to 2 cycles of bolus cisplatin (100 mg/m2 on days sis, little difference was seen when comparing the sur- 1 and 22) or weekly cetuximab (400 mg/m2 loading vival rates of the studies that adjusted for risk factors dose then 250 mg/m2 weekly; NCT01302834). This was (age, sex, smoking, alcohol, stage) to those that did not, a noninferiority trial with a primary end point of OS suggesting that the benefit seen in patients positive for and the goal of determining whether substituting ce- HPV infection may be largely attributable to the HPV tuximab for cisplatin allows for a decrease in morbidity infection itself.37 without sacrificing survival. This is similar to the trial being performed in the United Kingdom, which Current De-Escalation Trials for Human has enrolled patients with stage III/IV p16-positive Papillomavirus–Positive Oropharyngeal Cancer OPC-SCC treated with accelerated IMRT and then ran- Substituting Cisplatin With Cetuximab domized to either 3 cycles of bolus cisplatin vs weekly The addition of cisplatin to RT has been shown to im- cetuximab (NCT01874171). Higher-risk patients, de- prove rates of survival in head and neck cancers by fined as having M1 disease or a smoking history longer 5% to 8% but at the cost of significant adverse events, than 10 pack-years or those with greater than N2a stag- including ototoxicity (23%–50%), peripheral neuropa- ing, were excluded from this study. The primary end thy (30%–86%), nephrotoxicity (20%–41%), and my- point is severe acute and late toxicity (grades 3–5), with elosuppression (< 5%).38,39 These toxicities have led to secondary end points of survival and disease control. the investigation of alternative systemic agents, such A similar trial located in Australia will study patients as cetuximab, which is a monoclonal antibody that with standard fractionated RT (70 Gy in 35 fractions in targets epidermal growth factor receptor. A phase 7 weeks) who are randomized to either weekly cisplatin 3 study of locally advanced head and neck cancers (40 mg/m 2) or weekly cetuximab, with the primary end found that adding cetuximab to RT prolonged rates point of toxicity and secondary end points of survival of locoregional control, DFS, and OS.40 With the ex- and locoregional control (NCT01855451).

July 2016, Vol. 23, No. 3 Cancer Control 201 Table 2. — Ongoing De-Escalation Trials for HPV-Positive OPC-SCC Trial (NCT Phase No. of Outcome Study Aim Study Design Deintensi- Medication Radiation Inclusion Method No.) Study Measure fication Dose/ Criteria of HPV Patients Fraction Detection Substitution of Cisplatin With Cetuximab

De-ESCALaTE: 3 304 Cost Compare RT + Cisplatin to Cisplatin 70 Gy in Stages III– p16 Cet vs Cis LRC toxicity and cetuximab cetuximab (3 wk for 35 fractions IVa (T3N0– Toxicity outcomes 3 cycles) in 6/wk T4N0, T1N1– QOL RT + (NCT01874171) between cisplatin Cetuximab T4N3) OS cisplatin and (loading dose If smoking Toxicity cetuximab before RT, history > 10 (≥ grade 3) with acceler- then wkly py: N0–N2a ated RT with RT)

RTOG1016 3 706 DFS Compare RT + Cisplatin to Cisplatin (3 wk 70 Gy in Stage III p16 (NCT01302834) Distant toxicity and cetuximab cetuximab for 2 cycles) 35 fractions (excluding metastasis outcomes RT + Cetuximab in 6/wk T1–2N1) between or M1 LRC cisplatin (loading dose cisplatin and before RT, OS cetuximab then wkly Toxicity with acceler- with RT) ated RT

TROG 12.01/ 3 200 DFS Compare RT + Cisplatin to Intravenous 70 Gy in Smoking p16 ACTRN1261300- LRF toxicity and cetuximab cetuximab cisplatin 35 fractions history ≤ 10 py, 0279729 outcomes (wkly) in 7 wk stage III QOL RT + (NCT01855451) between cisplatin Cetuximab (excluding OS cisplatin and (intravenous T1–2N1) Toxicity cetuximab loading dose or stage IV with conven- before RT, (excluding tional RT then wkly T4, N3, M1) during RT) Smoking history > 10 py: N0–N2a

Deintensification of Chemoradiotherapy

LCCC1120 2 43 Complete Determine RT + RT dose Intravenous 54–60 Gy in T0-3 HPV DNA (NCT01530997) pathological response cisplatin cisplatin 30 fractions N0-2c p16 response and toxic- (wkly) rate ity with M0 HPV DNA + p16 DSS de-escalated Smoking RT history LRC ≤ 10 py or OS > 5 y of QOL abstinence

J0988 2 60 Adverse Determine RT RT dose Cisplatin 63 and T1-3 HPV DNA (NCT01088802) events response and (de-escalated) (wk 1–3 and 50.75 Gy in Any resect- p16 toxicity with + cisplatin 5–7) 35 fractions Late toxicity able node HPV DNA + (≥ grade 3) de-escalated in 7 wk RT p16 QOL

NRG-HN002 2 296 Distant Determine SFX RT + RT dose ± Intravenous 60 Gy/ Stages p16 (NCT02254278) metastasis if cisplatin cisplatin cisplatin cisplatin (wkly) 30 fractions III–IVa HPV bio- is needed in AFX RT None in 6 wk T1–2N1–2b markers lowered RT 60 Gy/ dose T3N0–2b LRC 30 fractions in 5 wk Smoking OS history PFS < 10 py Toxicity

UVA-16766 1 45 N0 nodal Determine RT Elective Unspecified 39.6 Gy Stages I–IVb p16 (NCT01891695) control rate response and nodal RT PFS toxicity with de-escalated Toxicity elective nodal RT

continued on next page

202 Cancer Control July 2016, Vol. 23, No. 3 Table 2. — Ongoing De-Escalation Trials for HPV-Positive OPC-SCC, continued Trial (NCT Phase No. of Outcome Study Aim Study Design Deintensi- Medication Radiation Inclusion Method No.) Study Measure fication Dose/ Criteria of HPV Patients Fraction Detection Induction Chemotherapy Followed by Low-Dose RT ECOG1308 2 83 2-y PFS Determine 2 cohorts strati- RT dose Induction: CR: 54 Gy/27 Stages HPV DNA (NCT01084083) Biomarkers if response fied by response 3 cycles of fractions III–IVb, p16 to induction to induction paclitaxel PR/stable: resectable OS chemotherapy chemotherapy: (d 1, 8, 15), HPV DNA + 66 Gy/33 p16 QOL allows for Induction cisplatin fractions Toxicity safe RT chemotherapy (d 1), cetux- de-escalation then concurrent imab (loading cetuximab with dose then response-de- wkly) pendent RT dose Concurrent: (CR) low-dose Cetuximab RT or (wkly) (PR/stable) standard-dose RT Quarterback 3 365 Biomarkers Determine 2 arms for CR/PR: RT dose Induction: TPF 56 Gy/28 Stages III–IV, HPV DNA Trial LRC if response Induction for 3 cycles fractions M0, without p16 (NCT01706939) to induction chemotherapy active OS Concurrent: 70 Gy/35 chemotherapy then carboplatin Carboplatin fractions smoking his- PFS allows for with either (wkly) tory (< 20 py Toxicity safe RT low-dose RT in last 20 y de-escalation or standard- or smoked dose RT within last 2 y Upfront Surgery ADEPT 3 496 CSS Determine TORS followed Cisplatin Cisplatin 60 Gy/30 T1–4a, SM–, p16 + + (NCT01687413) DFS if cisplatin by RT (wkly) fractions N with ECE added to in 6 wk Distant TORS followed low-dose RT by RT + cisplatin metastasis has benefit LRC in ECE+ Toxicity

ECOG3311 2 377 OS De-escalation 3 cohorts RT dose Low/inter- Low risk: Stages p16 (NCT01898494) PFS of adjuvant stratified by risk: mediate: None None III–IVa treatment QOL Surgery High: Cisplatin Intermediate based on his- followed by: (wkly) risk: Surgical topathological bleeding assessment Low risk: (A) 50 Gy/25 Observation fractions SM+ rate Intermediate risk: (B) 60 Gy/30 Toxicity (A) Low-dose RT fractions (B) Standard RT High risk: 60 Gy/30 High risk: fractions Standard RT + cisplatin PATHOS 2/3 242 DFS De-escalation 3 cohorts RT dose and Low/inter- Low risk: Stages p16 (NCT02215265) OS of adjuvant stratified by cisplatin mediate: None None T1–3, treatment risk: N0–2b QOL High: None Intermediate based on Surgery Exclude Swallowing histopatho- High: Cisplatin risk: followed by: (wkly) current function logical (A) 60 Gy/30 smokers assessment Low risk: fractions Toxicity Observation (< 2 y of (B) 50 Gy/25 diagnosis) Intermediate fractions with N2b risk: disease (A) Low-dose RT High risk: (B) Standard RT (A)/(B) 60 Gy/30 High risk: fractions (A) Standard RT (B) Standard RT + cisplatin

AFX = accelerated fractionation, CSS = cancer-specific survival, CR = complete response, DSS = disease-specific survival, ECE = extracapsular extension, HFX = hyperfractionation, HPV = human papillomavirus, LRC = locoregional control, LRF = locoregional failure, OPC = oropharyngeal cancer, OS = overall survival, PFS = progression-free survival, PR = partial response, py = pack year, QOL = quality of life, RT = radiotherapy, SCC = squamous cell carcinoma, SFX = standard fractionation, SM = surgical margins, TOR = transoral robotic-assisted surgery, TPF = docetaxel/cisplatin/fluorouracil.

July 2016, Vol. 23, No. 3 Cancer Control 203 Deintensification of Chemoradiotherapy standard RT to 70 Gy with cetuximab. Preliminary data Several trials are also attempting to deintensify the ra- show CRs in 62 of the 80 study patients (78%) and PRs/ diation dose in the setting of HPV-positive OPC-SCC stable disease in 15 (18%) of study patients after induc- by maintaining weekly cisplatin and de-escalating tion chemotherapy and then subsequently treated with the dose to the primary tumor (LCCC1120, J0988) or de-escalated and standard RT, respectively. The de-es- elective nodes (J0988, UVA-16766). A phase 2 trial en- calated RT group had a similar rate of 1-year DFS com- rolled patients with HPV-positive OPC-SCC (excluding pared with the standard RT group (91% vs 87%), and T4, N3, M1 disease) who had a smoking history of no the rate was even higher in study patients with a smok- more than 10 pack-years or those who used to smoke ing history shorter than 10 pack-years (97%). Patients but stopped more than 5 years ago, and then treated with high-risk features, including T4 disease, N2c dis- them with weekly cisplatin (30 mg/m2) and concur- ease, and a longer smoking history (>10 pack-years), rent IMRT (60 Gy in 30 fractions; NCT01530997). The had suboptimal DFS rates between 84% and 88%. Be- primary end point was complete pathological response cause this trial altered many variables from standard rate based on required biopsy of the primary site and chemoradiotherapy, it is difficult to determine which dissection of pretreatment regions of lymph-node in- variable could be adjusted to improve outcomes in pa- volvement 4 to 14 weeks following chemoradiation. tients with HPV-positive disease and high-risk features. The 43 patients enrolled had an 86% (37 of 43) path- Following the study data from Posner et al,33 a fol- ological response rate and all 6 non–pathological-re- low-up phase 3 trial was initiated and enrolled patients sponse rate cases had microscopic residual disease with stages III/IV HPV-positive OPC-SCC who were (1 primary site and 5 nodal sites). At a median follow- treated with 3 cycles of induction docetaxel/cisplatin/ up of 15 months, all study patients were alive and had fluorouracil followed by clinical/radiological evalua- no evidence of disease.47 This trial is similar to an Aus- tion (NCT01706939). Patients with a PR or CR to induc- tralian study, in which the researchers are de-escalat- tion chemotherapy will undergo a 2:1 randomization ing study patients to 63 Gy in 35 fractions with concur- of reduced-dose RT (56 Gy/28 fractions) vs standard- rent weekly cisplatin; severe late toxicity (grade ≥ 3) is dose RT (70 Gy/35 fractions) with concurrent weekly the primary end point (NCT01088802). Another study is carboplatin. Patients with M1 disease and active smok- also looking at de-escalating elective nodal volumes from ers (> 20 pack-years during the last 20 years or smok- 50.0 to 39.6 Gy in stage I/IVb HPV-positive OPC-SCC ing < 2 years ago) were excluded from the study. The (NCT01891695). In that study, nodal control is being primary end point will be the difference in DFS rate evaluated as the primary end point. with reduced-dose RT. Favorable risk is being investigated in a phase 2 randomized trial using a de-escalation rate of radia- Upfront Surgery tion from 70 to 60 Gy (NCT02254278). In this study, Minimally invasive surgery, such as transoral robotic- participants with stage III/IVa HPV-positive OPC-SCC assisted surgery, was approved in 2009 by the US Food (T12N12b or T3N02b) and a smoking history of no and Drug Administration for the treatment of patients longer than 10 pack-years will be randomized to ei- with T1- to T2-staged OPC. However, up to 80% of pa- ther accelerated-fractionation RT alone (60 Gy in 30 tients may also receive adjuvant RT.48 As of publica- fractions in 5 weeks) or standard-fractionation RT tion, multiple trials are in progress that will employ (60 Gy in 30 fractions in 6 weeks) with concurrent upfront surgery for OPC and randomize adjuvant treat- weekly cisplatin (40 mg/m2). The primary objective is ment strategies based on histopathological features to select the arm with a 2-year DFS rate of at least 85% (NCT01898494, NCT01687413, NCT02215265). without unacceptable swallowing toxicity at 1 year. A phase 2 trial conducted by the ECOG is stratifying patients with stage III/IVa HPV-positive OPC based Induction Chemotherapy Followed by Lower- on pathological findings after surgery (NCT01898494). Dose Radiation The study participants will undergo upfront transoral A phase 2 trial conducted by the ECOG has enrolled robotic-assisted surgery and neck dissection and are patients with resectable stage III/IVb HPV-positive to be stratified into 3 risk groups: low, intermediate, OPC initially treated with 3 cycles of induction pacli- and high. The low-risk group (those with T12N01 dis- taxel (90 mg/m2 on days 1, 8, and 15), cisplatin (75 mg/ ease with negative margins) will receive no further m2 on day 1), and cetuximab (loading dose 400 mg/m2; treatment. Those defined as having intermediate weekly dose 250 mg/m2) followed by clinical/radio- risk (< 1 mm positive extracapsular extension [ECE], logical evaluation (NCT01084083). Patients with a clini- 2–4 positive lymph nodes, perineural/lymphovas- cal complete response (CR) to induction chemotherapy cular invasion) will be randomized to either low-dose underwent de-escalated RT (54 Gy in 27 fractions) and IMRT (50 Gy/25 fractions) or standard-dose IMRT concurrent weekly cetuximab, whereas patients with a (60 Gy/30 fractions). Those in the high-risk group clinical partial response (PR) or stable disease received (positive margins, ECE+ > 1 mm or ≤ 5 positive lymph

204 Cancer Control July 2016, Vol. 23, No. 3 nodes) will receive IMRT (66 Gy/33 fractions) with ness to develop targeted treatments to provide alterna- weekly cisplatin (40 mg/m2). In addition to evaluating tive methods of de-escalation RT. the DFS rate, they will investigate surgical effective- Two T-cell vaccination studies that include all ness (margin status) and complication rates such as HPV malignancies (cervical, oropharyngeal, vagi- bleeding. nal, anal, and penile cancers) are currently underway A UK phase 2/3 trial is randomizing intermediate- (NCT02280811, NCT01585428). One phase 1/2 trial risk patients (T3 or T1–2 with additional risk factors, will grow white blood cells from study participants N2a or N2b, close margins [1–5 mm], or perineural/ and infect them with a retrovirus directed at the tu- lymphovascular invasion) to postoperative low- mor (anti-HPV E6) prior to injecting those cells back (50 Gy/25 fractions) or standard-dose RT (60 Gy/ into the study participants to determine the tumor 30 fractions; NCT02215265). High-risk patients (positive response rate (NCT02280811). In another study, margins [< 1 mm]) will be randomized to IMRT tumor-infiltrating lymphocytes will be grown ex vivo (60 Gy/30 fractions) with or without weekly cisplatin and then reinfused into study participants to evaluate (40 mg/m 2). The exclusion criteria for this study are response (NCT01585428). those with T4, N2c–3, or M1 disease and those with Programmed death ligand 1 (PDL-1) is a check- N2b disease who currently smoke. The primary end point inhibitor that binds to T cells and can cause de- point will be the effect of RT de-escalation on swal- creased T-cell function. Anti-PDL-1 can induce an im- lowing function and toxicity. munogenic response in various cancers by improving Among patients with HPV-positive OPC who T-cell activity and function.53 PDL-1 expression is com- choose to undergo surgery, retrospective data suggest mon in HPV-positive OPC, with expression levels high- no advantage to adjuvant concurrent chemotherapy er than HPV-negative OPC (49%–70% vs 29%–34%).54-56 in those with extracapsular extension.49 An ongoing Preliminary analysis in a phase 1b multicohort study phase 3 trial is investigating the optimal adjuvant treat- involving pembrolizumab showed that 99 study pa- ment in persons who are ECE positive (NCT01687413). tients with recurrent/metastatic head and neck cancers In this study, those with p16-positive OPC-SCC who had a response rate of 18.2%, regardless of HPV status underwent transoral robotic-assisted surgery with neg- (NCT01848834). Future studies may focus on the ef- ative margins and who were positive for ECE will be fectiveness of anti-PDL-1 in patients with HPV-positive randomized to IMRT (60 Gy/30 fractions in 6 weeks) OPC, especially in the upfront adjuvant setting. with or without weekly cisplatin (40 mg/m2). DFS will be the primary end point. Management of Human Papillomavirus– Positive Oropharyngeal Cancer Future Therapeutic Approaches Prevention Evidence suggests that improved outcomes in patients HPV-associated OPC has a different set of risk factors with HPV-positive OPC can be attributed to the than conventional head and neck cancers. It is impor- T-cell–mediated immune response.50 The presence of tant to educate patients on the risk factors associated infiltrating T cells may also be prognostic for improved with this disease and recommend safe sexual practic- outcomes in patients positive for HPV infection.51 es, as well as help them to understand how to mini- These data suggest that immunomodulators directed mize their exposure to tobacco and alcohol products. at activating T cells against HPV may provide an al- HPV-induced malignancies have been studied in other ternative treatment approach for these patients, pre- disease sites for decades. Preventive measures, such sumably with less toxicity than conventional treat- as HPV vaccines, can reduce the risk of HPV infection ment.51 Approaches may include therapeutic vaccines and theoretically decrease the incidence of HPV-asso- specific to HPV, injecting tumor-specific T cells, or by ciated OPC.24 inducing T-cell immunity in such patients. The most common vaccines for HPV are bivalent For example, therapeutic vaccinations targeting (HPV-16, -18), tetravalent (HPV-6, -11, -16, -18), and E6/E7 may induce CR in HPV-infected tumors.52 In head 9-valent (HPV-6, -11, -16, -18, -31, -33, -45, -52, -58).57 and neck cancers, select phase 1 studies are examining Initially, recommendations for HPV vaccines were ad- therapeutic vaccines specific to HPV-infected malig- vocated for adolescent females for the prevention of nancies, including targets against HPV-16 antigens cervical cancer; however, the US Food and Drug Ad- (NCT01598792) or p16INK4a peptides (NCT01462838). ministration has approved HPV vaccines for use in A phase 2 study is looking at whether a Listeria- both sexes to prevent the spread of infection and sub- based vaccine for HPV can induce an antigen-specific sequent malignancy.24,58 The Advisory Committee on T-cell response in patients with HPV-positive OPC un- Immunization Practices recommends HPV vaccination dergoing upfront transoral robotic-assisted surgery be initiated at the age of 11 or 12 years, but the vac- (NCT02002182). These 2 vaccination trials are evaluat- cine can be started as early as 9 years of age and as late ing patient safety and rate of immunogenic effective- as 26 years of age.57 Healthy males are recommended

July 2016, Vol. 23, No. 3 Cancer Control 205 to undergo vaccination up to 21 years of age, whereas excellent response to treatment, clinical trials are avail- females and high-risk males (those who are immuno- able to de-escalate treatment to minimize toxicity with- compromised, men who have sex with men) are rec- out sacrificing disease control or survival rates. New ommended up to 26 years of age.57 Contraindications therapeutic approaches are also being explored, such to the vaccine include pregnancy and immediate hy- as immunomodulators, which may improve outcomes persensitivity reaction.57 in this patient population. A randomized controlled trial in Costa Rica showed that the bivalent vaccination of women had a References significantly decreased prevalence of oral HPV infec- 1. American Cancer Society. Cancer Facts & Figures 2016. Atlanta, GA: tion and a vaccine effectiveness rate of 93.3%.16 Be- American Cancer Society; 2016. http://www.cancer.org/acs/groups/content/@ research/documents/document/acspc-047079.pdf. Accessed May 24, 2016. cause oral HPV infection can increase the risk of OPC 2. Shibuya K, Mathers CD, Boschi-Pinto C, et al. Global and regional by 1200%, HPV vaccination may aid in the prevention estimates of cancer mortality and incidence by site: II. Results for the global burden of disease 2000. BMC Cancer. 2002;2:37. of oropharyngeal malignancy in addition to anogenital 3. Ragin CC, Modugno F, Gollin SM. The epidemiology and risk factors malignancies.26 Overall, HPV vaccination is a relatively of head and neck cancer: a focus on human papillomavirus. J Dent Res. 2007;86(2):104-114. safe preventive measure and should be considered in 4. Chenevert J, Chiosea S. Incidence of human papillomavirus in all adolescents.58 oropharyngeal squamous cell carcinomas: now and 50 years ago. Hum Pathol. 2012;43(1):17-22. 5. Panwar A, Batra R, Lydiatt WM, et al. Human papilloma virus posi- Follow-Up tive oropharyngeal squamous cell carcinoma: a growing epidemic. Cancer Treat Rev. 2014;40(2):215-219. Guidelines from the National Comprehensive Cancer 6. Lassen P, Eriksen JG, Krogdahl A, et al. The influence of HPV- Center do not differentiate between HPV-related and associated p16-expression on accelerated fractionated radiotherapy in 59 head and neck cancer: evaluation of the randomised DAHANCA 6&7 trial. unrelated head and neck cancers. They recommend Radiother Oncol. 2011;100(1):49-55. that a history should be taken and a physical examina- 7. Lassen P, Primdahl H, Johansen J, et al. Impact of HPV-associated p16-expression on radiotherapy outcome in advanced oropharynx and tion should be performed every 1 to 3 months for year non-oropharynx cancer. Radiother Oncol. 2014;113(3):310-316. 1, 2 to 6 months for year 2, 4 to 8 months for years 3 to 5, 8. O’Sullivan B, Huang SH, Perez-Ordonez B, et al. Outcomes of 59 HPV-related oropharyngeal cancer patients treated by radiotherapy alone and then annually thereafter. Imaging is recom- using altered fractionation. Radiother Oncol. 2012;103(1):49-56. mended within 6 months and chest imaging is recom- 9. Overgaard J, Hansen HS, Specht L, et al. Five compared with six fractions per week of conventional radiotherapy of squamous-cell carci- mended if clinically indicated in patients with a histo- noma of head and neck: DAHANCA 6 and 7 randomised controlled trial. ry of smoking.59 However, patients with HPV-positive Lancet. 2003;362(9388):933-940. 10. Ang KK, Harris J, Wheeler R, et al. Human papillomavirus and survival OPC typically present at a younger age, have fewer of patients with oropharyngeal cancer. N Engl J Med. 2010;363(1):24-35. smoking-related comorbidities, have improved outcomes, 11. O’Sullivan B, Huang SH, Siu LL, et al. Deintensification candidate 10,11,21 subgroups in human papillomavirus-related oropharyngeal cancer accord- and experience significantly less rates of toxicity. ing to minimal risk of distant metastasis. J Clin Oncol. 2013;31(5):543-550. A retrospective study evaluated severe (grade ≥ 3) 12. Huang SH, Xu W, Waldron J, et al. Refining American Joint Com- mittee on Cancer/Union for International Cancer Control TNM stage and toxicity and LRF rates in patients with HPV-positive prognostic groups for human papillomavirus-related oropharyngeal carci- OPC to help determine the optimal clinical follow-up.60 nomas. J Clin Oncol. 2015;33(8):836-845. 13. Al-Swiahb JN, Huang CC, Fang FM, et al. Prognostic impact of p16, Of the 232 study patients with HPV-positive OPC, p53, epidermal growth factor receptor, and human papillomavirus in oro- 21 (9%) had severe late toxicity and a 3-year locore- pharyngeal cancer in a betel nut-chewing area. Arch Otolaryngol Head 60 Neck Surg. 2010;136(5):502-508. gional control rate of 94%. The study showed that 14. Lassen P, Overgaard J, Eriksen JG. Expression of EGFR and HPV- 64% of events (toxicity or failure) occurred within associated p16 in oropharyngeal carcinoma: correlation and influence on prognosis after radiotherapy in the randomized DAHANCA 5 and 7 trials. the first 6 months of follow-up, with an incidence Radiother Oncol. 2013;108(3):489-494. event rate lower than 2% on subsequent follow-up.60 15. Sheu T, Fuller CD, Mendoza TR, et al. Nomogram for predicting symptom severity during radiation therapy for head and neck cancer. Otolaryngol The authors recommend reducing follow-up in pa- Head Neck Surg. 2014;151(4):619-626. tients with HPV-positive OPC to every 3 months for 16. Herrero R, Quint W, Hildesheim A, et al. Reduced prevalence of oral human papillomavirus (HPV) 4 years after bivalent HPV vaccination in the first 6 months, then every 6 months for the first a randomized clinical trial in Costa Rica. PLoS One. 2013;8(7):e68329. 2 years, and then annually thereafter.60 17. Tota JE, Chevarie-Davis M, Richardson LA, et al. Epidemiology and burden of HPV infection and related diseases: implications for prevention strategies. Prev Med. 2011;53(suppl 1):S12-S21. Conclusions 18. Kreimer AR, Clifford GM, Boyle P, et al. Human papillomavirus types in head and neck squamous cell carcinomas worldwide: a system- The incidence rate of oropharyngeal cancer (OPC) as- atic review. Cancer Epidemiol Biomarkers Prev. 2005;14(2):467-475. sociated with human papillomavirus (HPV) infection 19. St Guily JL, Jacquard AC, Pretet JL, et al. Human papillomavirus 4,5 genotype distribution in oropharynx and oral cavity cancer in France--the is growing at an epidemic rate. HPV-driven oncogen- EDiTH VI study. J Clin Virol. 2011;51(2):100-104. esis makes persons with HPV-associated OPC biologi- 20. Chaturvedi AK, Engels EA, Pfeiffer RM, et al. Human papillomavirus and rising oropharyngeal cancer incidence in the United States. J Clin cally distinct from persons with OPC not associated Oncol. 2011;29(32):4294-4301. with HPV infection, and efforts are being made to per- 21. Gillison ML. Human papillomavirus-related diseases: oropharynx cancers and potential implications for adolescent HPV vaccination. J Adolesc sonalize the management of this type of OPC. Preven- Health. 2008;43(4 suppl):S52-S60. tive measures, such as education about sexual practic- 22. Gillison ML, Broutian T, Pickard RK, et al. Prevalence of oral HPV infection in the United States, 2009-2010. JAMA. 2012;307(7):693-703. es and other risk factors, along with HPV vaccinations, 23. Settle K, Posner MR, Schumaker LM, et al. Racial survival disparity should be considered.57 Because this population has an in head and neck cancer results from low prevalence of human papilloma-

206 Cancer Control July 2016, Vol. 23, No. 3 virus infection in black oropharyngeal cancer patients. Cancer Prev Res the literature. ISRN Oncol. 2012;2012:945162. (Phila). 2009;2(9):776-781. 49. Sinha P, Lewis JS Jr, Piccirillo JF, et al. Extracapsular spread and 24. D’Souza G, Dempsey A. The role of HPV in head and neck cancer adjuvant therapy in human papillomavirus-related, p16-positive oropha- and review of the HPV vaccine. Prev Med. 2011;53(suppl 1):S5-S11. ryngeal carcinoma. Cancer. 2012;118(14):3519-3530. 25. D’Souza G, Agrawal Y, Halpern J, et al. Oral sexual behaviors as- 50. Williams R, Lee DW, Elzey BD, et al. Preclinical models of HPV+ sociated with prevalent oral human papillomavirus infection. J Infect Dis. and HPV- HNSCC in mice: an immune clearance of HPV+ HNSCC. Head 2009;199(9):1263-1269. Neck. 2009;31(7):911-918. 26. D’Souza G, Kreimer AR, Viscidi R, et al. Case-control study 51. Ward MJ, Thirdborough SM, Mellows T, et al. Tumour-infiltrating of human papillomavirus and oropharyngeal cancer. N Engl J Med. lymphocytes predict for outcome in HPV-positive oropharyngeal cancer. 2007;356(19):1944-1956. Br J Cancer. 2014;110(2):489-500. 27. Begum S, Cao D, Gillison M, et al. Tissue distribution of human 52. Kenter GG, Welters MJ, Valentijn AR, et al. Vaccination against papillomavirus 16 DNA integration in patients with tonsillar carcinoma. Clin HPV-16 oncoproteins for vulvar intraepithelial neoplasia. N Engl J Med. Cancer Res. 2005;11(16):5694-5699. 2009;361(19):1838-1847. 28. Chung CH, Bagheri A, D’Souza G. Epidemiology of oral human 53. Formenti SC, Demaria S. Combining radiotherapy and cancer im- papillomavirus infection. Oral Oncol. 2014;50(5):364-369. munotherapy: a paradigm shift. J Natl Cancer Inst. 2013;105(4):256-265. 29. Ang KK, Zhang Q, Rosenthal DI, et al. Randomized phase III trial of 54. Yang W, Song Y, Lu YL, et al. Increased expression of programmed concurrent accelerated radiation plus cisplatin with or without cetuximab death (PD)-1 and its ligand PD-L1 correlates with impaired cell-mediated for stage III to IV head and neck carcinoma: RTOG 0522. J Clin Oncol. immunity in high-risk human papillomavirus-related cervical intraepithelial 2014;32(27):2940-2950. neoplasia. Immunology. 2013;139(4):513-522. 30. Rischin D, Young RJ, Fisher R, et al. Prognostic significance of 55. Lyford-Pike S, Peng S, Young GD, et al. Evidence for a role of the p16INK4A and human papillomavirus in patients with oropharyngeal cancer PD-1:PD-L1 pathway in immune resistance of HPV-associated head and treated on TROG 02.02 phase III trial. J Clin Oncol. 2010;28(27):4142-4148. neck squamous cell carcinoma. Cancer Res. 2013;73(6):1733-1741. 31. Gillison ML, Zhang Q, Jordan R, et al. Tobacco smoking and in- 56. Ukpo OC, Thorstad WL, Lewis JS, Jr. B7-H1 expression model for creased risk of death and progression for patients with p16-positive and immune evasion in human papillomavirus-related oropharyngeal squa- p16-negative oropharyngeal cancer. J Clin Oncol. 2012;30(17):2102-2111. mous cell carcinoma. Head Neck Pathol. 2013;7(2):113-121. 32. Fakhry C, Westra WH, Li S, et al. Improved survival of patients with 57. Petrosky E, Bocchini JA Jr, Hariri S, et al. Use of 9-valent human human papillomavirus-positive head and neck squamous cell carcinoma in papillomavirus (HPV) vaccine: updated HPV vaccination recommenda- a prospective clinical trial. J Natl Cancer Inst. 2008;100(4):261-269. tions of the advisory committee on immunization practices. MMWR Morb 33. Posner MR, Lorch JH, Goloubeva O, et al. Survival and human pap- Mortal Wkly Rep. 2015;64(11):300-304. illomavirus in oropharynx cancer in TAX 324: a subset analysis from an 58. Soares GR, Vieira Rda R, Pellizzer EP, et al. Indications for the HPV international phase III trial. Ann Oncol. 2011;22(5):1071-1077. vaccine in adolescents: a review of the literature. J Infect Public Health. 34. Lohaus F, Linge A, Tinhofer I, et al. HPV16 DNA status is a strong 2015;8(2):105-116. prognosticator of loco-regional control after postoperative radiochemo- 59. National Comprehensive Cancer Network. Head and Neck Cancer therapy of locally advanced oropharyngeal carcinoma: results from a mul- Cancers. v1.2016. https://www.nccn.org/professionals/physician_gls/pdf ticentre explorative study of the German Cancer Consortium Radiation /head-and-neck.pdf. Accessed May 24, 2016. Oncology Group (DKTK-ROG). Radiother Oncol. 2014;113(3):317-323. 60. Frakes JM, Naghavi AO, Demetriou SK, et al. Determining optimal 35. Overgaard J, Hansen HS, Overgaard M, et al. A randomized dou- follow-up in the management of human papillomavirus-positive oropharyn- ble-blind phase III study of nimorazole as a hypoxic radiosensitizer of pri- geal cancer. Cancer. 2016;122(4):634-641. mary radiotherapy in supraglottic larynx and pharynx carcinoma. Results of the Danish Head and Neck Cancer Study (DAHANCA) Protocol 5-85. Radiother Oncol. 1998;46(2):135-146. 36. Ragin CC, Taioli E. Survival of squamous cell carcinoma of the head and neck in relation to human papillomavirus infection: review and meta-analysis. Int J Cancer. 2007;121(8):1813-1820. 37. O’Rorke MA, Ellison MV, Murray LJ, et al. Human papillomavirus related head and neck cancer survival: a systematic review and meta- analysis. Oral Oncol. 2012;48(12):1191-1201. 38. Pignon JP, le Maitre A, Maillard E, et al. Meta-analysis of chemotherapy in head and neck cancer (MACH-NC): an update on 93 randomised trials and 17,346 patients. Radiother Oncol. 2009;92(1):4-14. 39. Chen AM, Zahra T, Daly ME, et al. Definitive radiation therapy without chemotherapy for human papillomavirus-positive head and neck cancer. Head Neck. 2013;35(11):1652-1656. 40. Bonner JA, Harari PM, Giralt J, et al. Radiotherapy plus cetuximab for squamous-cell carcinoma of the head and neck. N Engl J Med. 2006;354(6):567-578. 41. Rosenthal DI, Harari PM, Giralt J, et al. Impact of p16 status on the results of the phase III cetuximab (cet)/radiotherapy (RT). J Clin Oncol. 2014;32(5 suppl): 6001. 42. Baumeister P, Marte M, Reiter M, et al. EGFR and P-GP expression in oropharyngeal mucosa in relation to smoking. Anticancer Res. 2012;32(5):1639-1648. 43. Kumar B, Cordell KG, Lee JS, et al. EGFR, p16, HPV Titer, Bcl-xL and p53, sex, and smoking as indicators of response to therapy and survival in oropharyngeal cancer. J Clin Oncol. 2008;26(19):3128-3137. 44. Hong A, Dobbins T, Lee CS, et al. Relationships between epidermal growth factor receptor expression and human papillomavirus status as markers of prognosis in oropharyngeal cancer. Eur J Cancer. 2010;46(11):2088-2096. 45. Farley J, Sill MW, Birrer M, et al. Phase II study of cisplatin plus cetuximab in advanced, recurrent, and previously treated cancers of the cervix and evaluation of epidermal growth factor receptor immunohistochemical expression: a Gynecologic Oncology Group study. Gynecol Oncol. 2011;121(2):303-308. 46. Santin AD, Sill MW, McMeekin DS, et al. Phase II trial of cetuximab in the treatment of persistent or recurrent squamous or non-squamous cell carcinoma of the cervix: a Gynecologic Oncology Group study. Gynecol Oncol. 2011;122(3):495-500. 47. Chera BS, Amdur RJ, Tepper J, et al. Phase 2 trial of de-intensified chemoradiation therapy for favorable-risk human papillomavirus-associated oropharyngeal squamous cell carcinoma. Int J Radiat Oncol Biol Phys. 2015;93(5):976-985. 48. Dowthwaite SA, Franklin JH, Palma DA, et al. The role of transoral robotic surgery in the management of oropharyngeal cancer: a review of

July 2016, Vol. 23, No. 3 Cancer Control 207 Definitive radiotherapy

plays a role in the treatment of

patients with glottic SCC.

Artwork courtesy of Arthur Pina de Alba. www.arthurpinadealba.com.

Definitive Radiotherapy for Squamous Cell Carcinoma of the Glottic Larynx William M. Mendenhall, MD, Roi Dagan, MD, Curtis M. Bryant, MD, Robert J. Amdur, MD, and Anthony A. Mancuso, MD

Background: Depending on the extent of disease, squamous cell carcinoma (SCC) of the glottis is managed with surgery, radiotherapy (RT), or a combination of these modalities. Patients with advanced disease may receive concomitant chemotherapy in conjunction with definitive or postoperative RT. Methods: The treatment policies of the University of Florida and patient outcomes are reviewed. Results: The likelihood of cure after RT for carcinoma in situ (Tis) to T2 glottic SCC varies from 70% to 94% depending on tumor stage. Consideration should be given to adding weekly cisplatin for patients with T2b SCC because of the high local recurrence rate after RT alone. The probability of cure is about 65% to 80% for select low-volume (≤ 3.5 cc) T3 to T4 glottic SCC after RT. These patients should be considered for concomitant weekly cisplatin. Higher-volume tumors, particularly those with airway compromise, should be treated with laryngectomy and postoperative RT. Conclusion: Definitive RT is an excellent treatment for select patients with laryngeal cancer.

Introduction Unless otherwise specified, outcomes data are Depending on the extent of disease, squamous cell from the University of Florida. Patients were staged carcinoma (SCC) of the glottis is managed with sur- according to the prior edition of the American gery, radiotherapy (RT), or a combination of the 2 Joint Committee on Cancer (AJCC) staging system.1 modalities. Patients with advanced disease may re- Previously, vocal cord fixation was required to up- ceive concomitant chemotherapy in conjunction stage a patient to T3; however, the AJCC staging with definitive or postoperative RT. Herein we re- system was subsequently modified (7th edition) view the role of definitive RT in the treatment of paso that paraglottic space invasion without vo- tients with SCC of the glottis. cal cord fixation upstages T2 to T3.2 Dagan et al3 subsequently demonstrated that paraglottic space From the Departments of Radiation Oncology (WMM, RJA) and invasion did not adversely impact outcomes for Diagnostic Radiology (AAM), University of Florida College of patients with T2 SCC of the glottis treated with Medicine, Gainesville, Florida, and Proton Therapy Institute (RD, CMB), Jacksonville, Florida. definitive RT. Address correspondence to William M. Mendenhall, MD, 2000 Southwest Archer Road, Post Office Box 100385, Gainesville, FL 32610-0385. Carcinoma In Situ E-mail: [email protected] Cases of carcinoma in situ (Tis) are managed with vocal Submitted October 21, 2015; accepted December 11, 2015. cord stripping. If the lesion recurs and remains in situ, No significant relationships exist between the authors and the companies/organizations whose products or services may be referenced then the procedure may be repeated. However, after in this article. multiple vocal cord strippings, voice quality may dete-

208 Cancer Control July 2016, Vol. 23, No. 3 riorate so that definitive RT becomes an alternative op- A tion for patients with recurrent Tis after prior surgery. Patients are treated with parallel-opposed 4 to 6 MV photon fields limited to the glottis. For lesions limited to the true vocal cords, the top of the field is the midthyroid notch, the bottom is the inferior aspect of the cricoid cartilage, the posterior aspect is 1 cm behind the thyroid ala, and the anterior border is 1 to 1.5 cm anterior to the skin. If the lesion extends off of the true vocal cords, then the fields are accordingly enlarged. Depending on the extent of the lesion, the fields are weighted 1:1 or, if the lesion involves mostly 1 vocal cord, then 3:2 to the side of the tumor. A third anterior field may be used to give 5% to 10% of the dose to reduce the dose laterally (Fig 1). The dose-fractionation schedule employed is the same as that used for patients with T1 invasive SCC: 63 Gy at 2.25 Gy per once- daily fraction, and 5 fractions per week for 5.5 weeks. Sengupta et al4 reported on 37 study patients B with Tis of the glottic larynx treated with definitive RT between 1967 and 2005. The median dose given was 60 Gy (range, 56.25–66.50) at 2.25 Gy per fraction, and the mean follow-up time was 9.5 years.4 The 5-year local control rate was 91%, the local control rate with larynx preservation was 91%, the ultimate local control rate was 91%, and the cause-specific survival rate was 100%.4 No patient experienced a severe late complication.4

T1 to T2 Disease All patients with invasive SCC of the glottis undergo Fig 1A–B. Patient with T1N0 squamous cell carcinoma confined to the pretreatment computed tomography as part of their anterior two-thirds of the vocal cord. The patient is treated in the supine diagnostic evaluation. Although the yield is low, select position. (A) Schematic diagram of the field extending from the middle of patients with anterior commissure invasion will have the thyroid notch to the bottom of the cricoid cartilage, 1 cm posterior to the cartilages and with 1 cm fall-off anteriorly. (B) Computer-generated anterior extension along the Broyles ligament to the dosimetry; the dose is specified at the 95% isodose line. thyroid cartilage. It is also an excellent way to detect Reprinted from Int J Radiat Oncol Biol Phys, Volume 15, Mendenhall WM the presence and extent of subglottic invasion. et al, T1–T2 squamous cell carcinoma of the glottic larynx treated with radiation therapy: relationship of dose fractionation factors to local control Patients with T1 to T2 SCC may be treated with sur- 5 and complications, 1267–1273, 1988, with permission from Elsevier. gery or RT. We typically offer transoral CO2 laser excision to those with well-defined T1 SCC limited to the midthird of 1 vocal cord and then treat the remainder cord. Fig 2 shows the typical appearance of T1a SCC in with RT. Open partial laryngectomy is reserved for pa- the vocal cord before and after RT. tients with local recurrence following RT. Chera et al6 reported on 585 study patients with

The risk of subclinical disease in the regional T1N0 to T2N0 SCC of the glottis treated with definitive nodes is low so that RT fields are limited to the larynx. RT between the years 1964 and 2006. This is a rela- The dose-fractionation schedules are 63 Gy in 28 frac- tively unselected series of patients because the poli- tions for T1 to T2a disease (normal vocal cord mobility) cy during that time period was to treat nearly all pa- and 65.25 Gy in 29 fractions for T2b disease (decreased tients with T1 to T2 cancer with RT alone. T1 disease vocal cord mobility but not fixed). Patients with T2 was stratified as T1a (limited to 1 vocal cord with or disease — particularly T2b — may be considered for without invasion of the anterior commissure) or T1b treatment with hyperfractionated RT to 74.4 Gy at 1.2 (involvement of both true vocal cords), and T2 disease Gy twice a day.6 It is also reasonable to consider the was stratified as T2a (supraglottic extension, subglot- addition of concomitant weekly cisplatin 30 mg/m2 for tic extension, or both with normal vocal cord mobility) 7 those with unfavorable T2 lesions. Fig 1 illustrates the or T2b disease (vocal cord mobility diminished but not treatment field and dosimetry of a patient with T1N0 fixed).6 Paraglottic space invasion did not upstage a le- 6 SCC confined to the anterior two-thirds of the vocal sion to T3 in the absence of vocal cord fixation. No pa-

July 2016, Vol. 23, No. 3 Cancer Control 209 A B on tumor stage; however, the risk of nodal failure increases to 13% to 25% if the tumor recurs at the primary site. Therefore, we suggest that salvage surgery should include elective neck dissection. The 5-year distant metastasis-free rates were 99% for T1a, 99% for T1b, 99% for T2a, and 94% for T2b.6 The 5-year cause-specific and overall survival rates, respectively, were 97% and 82% for T1a, 99% and 83% for T1b, 94% and 76% for T2a, and 90% and 78% for 6 Fig 2A–B. Typical appearance of the larynx several years following radiotherapy. (A) T1a squamous cell car- T2b. cinoma involving most of the right true vocal cord. (B) By 1 year after radiotherapy with 63 Gy at 2.25 Gy, the Ten study patients larynx looks almost normal, except for a contour defect in the portion of the true cord previously destroyed (1.7%) experienced severe, by the tumor. Voice quality is usually good and dysphagia is rare. acute, late, or postoperative complications. One study tient received adjuvant chemotherapy or elective nodal patient with T2N0 SCC who was treated twice daily was irradiation.6 hospitalized for dehydration. Two others underwent sal- The distribution according to tumor stage appears vage laryngectomy for suspected local recurrence and in the Table.6 The anterior commissure was invaded had no tumor in either specimen. Three study patients in 369 study patients (63%). A total of 81 (14%) were with T2N0 SCC required permanent tracheostomy for treated twice daily at 1.2 to 1.25 Gy per fraction; the laryngeal edema. One study patient who underwent sal- remainder was treated once daily. The median follow- vage total laryngectomy for local recurrence had a pha- up period was 12.3 years. Parameters included in the ryngocutaneous fistula. Two received permanent per- multivariate analysis of local control were tumor stage cutaneous gastrostomy tubes. One patient died due to

(T1a vs T1b vs T2a vs T2b), anterior commissure inva- radiation-induced carotid artery angiosarcoma. sion, histological differentiation (well, moderate, or not otherwise specified vs poor), total dose (≤ 63 vs > 63 Gy), T3 to T4 Disease fractionation (≥ 2.25 daily vs ≥ 1.2 Gy twice daily vs Patients with fixed-cord T3 and T4 SCC may be treated < 2.25 Gy once daily), overall treatment time (continu- with definitive RT, surgery, or a combination of the ous variable), and beam energy (60Co vs other).6 2 modalities.8 The probability of local control with a The 5-year local and ultimate local control rates, functional larynx varies with the volume of the prima- which included successful salvage after local failure, appear in the Table.6 The 5-year rates of ultimate lo- Table. — Actuarial Local Control Rates by Tumor Stage for cal control with larynx preservation were 95% for T1a, Squamous Cell Carcinoma of the Glottic Larynx 94% for T1b, 81% for T2a, and 74% for T2b. Multivari- ate analysis revealed that the rate of local control was Tumor No. of 5-y Local 5-y Ultimate Stage Patients Control Rate, % Local Control adversely impacted by tumor stages of T2a and T2b Rate, % (P = .0001), overall treatment time of more than 41 days Tis 37 91 91 (P = .0016), poorly differentiated histology (P = .0126), T1a 253 94 98 and once-daily fractionation (P = .0494).6 The 5-year overall rates of neck control and neck T1b 72 93 97 control with continuous local control of the primary T2a 165 80 96 site, respectively, were 98% and 100% for T1a, 99% and T2b 95 70 93 100% for T1b, 96% and 98% for T2a, and 88% and 94% T3a 87 63 89b for T2b.6 The 5-year rates of neck control without con- a 22 81 b tinuous local control were 75% for T1a, 80% for T1b, T4 86 87% for T2a, and 75% for T2b.6 Thus, the risk of iso- aSelected low-volume T3 to T4 disease. bCrude ultimate local control lated nodal failure ranges from 0% to 6% depending rates. Data from reference 6.

210 Cancer Control July 2016, Vol. 23, No. 3 ry tumor calculated on a pretreatment scan.9 The pri- concomitant carboplatin and taxol, and 3 patients mary tumoral volume may vary based on the observer underwent planned neck dissection.8 The median and whether contrast-enhanced diagnostic computed follow-up period was 5.7 years for all patients and tomography is employed vs planning computed to- 10.6 years for living patients.8 mography; the former is preferred. The 5-year local control rates appear in the Table.6 Tumors with a primary tumor volume of 3.5 to The crude local control and ultimate local control 4 cm3 or smaller have a high probability of local con- rates for T3 disease were 67% (n = 58/87) and 89% trol with a functional larynx, and patients with such tu- (n = 77/87); for T4 disease, they were 82% (n = 18/22) mors are good candidates for definitive RT. Those with and 86% (n = 19/22); and the overall rates were 70% larger tumors have a high likelihood of local failure (n = 76/109) and 88% (n = 96/109).8 Multivariate analy- after definitive RT and, if controlled, a higher risk of a sis of local control revealed cord fixation P( = .5706), poorly functioning larynx and may require permanent once- vs twice-daily fractionation (P = .4936), exten- placement of a percutaneous endoscopic gastrostomy sion into soft tissue (P = .9324), subglottic extension tube, tracheostomy, or both. In addition, the risk of a of at least 1.0 cm (P = .2106), dose of no more than major complication after salvage laryngectomy is ap- 70 vs more than 70 Gy (P = .4655), histological dif- proximately 30% to 35%.8 These patients’ cancer is best ferentiation (P = .8181), and tumor stage (P = .1484). managed with total laryngectomy and neck dissection, Although all of the T3-stage cancers had cord fixa- typically followed by postoperative RT. Patients who re- tion, some of the T4-staged lesions had at least par- quire tracheostomy prior to treatment are likely to have tially mobile vocal cords. The 5-year outcomes by high-volume tumors that should be surgically managed. AJCC stage for stage III vs IVa, respectively, were Patients with T4 disease who are suitable for defin- 62% and 78% for locoregional control (P = .1010), itive RT are highly selected and have tumors less than 97% and 100% for distant metastasis-free survival 3.5 to 4.0 cm3 in size.8 Their disease is often staged as (P = .4323), 83% and 87% for cause-specific survival

T4a based on limited exolaryngeal spread through the (P = .5075), and 52% and 67% for overall survival thyrohyoid membrane, and they commonly have a (P = .5259). Severe complications were observed in more favorable prognosis than fixed-cord T3 SCC. The 13 patients (12%) and included cartilage necrosis more common, high-volume T4 SCC with cartilage de- (n = 1; fatal case), permanent gastrostomy (n = 2), struction has a poor prognosis after definitive RT and dysphagia/stricture (n = 1), edema requiring tracheos- is best treated with total laryngectomy, neck dissec- tomy (n = 2), and fatal aspiration (n = 1). tion, and postoperative RT. Pameijer et al11 reported on 42 study patients with Patients treated with definitive RT are irradi- fixed-cord T3 SCC of the glottis who were treated with ated with an altered fractionation schedule, prefer- definitive RT between the years 1980 and 1993, corre- ably 74.4 Gy in 62 twice-daily fractions in 31 treat- lating local control rates with primary tumor volume ment days for 6.5 weeks.7 If this is not feasible for and presence of cartilage sclerosis. All patients were fol- logistical reasons, then the Danish Head and Neck lowed for at least 2 years.11 Patients were stratified into Cancer Group schedule can be employed, which low risk (volume < 3.5 cm3 and no cartilage sclerosis consists of 70 Gy in 35 fractions for 30 treatment [n = 13] or volume < 3.5 cm3 and 1 cartilage sclerosis days with 2 fractions per day, 1 day per week, [n = 8]), moderate risk (volume < 3.5 cm3 and ≥ 2 carti- during weeks 2 through 6.7 The simultaneous inte- lage sclerosis [n = 5] or volume > 3.5 cm3 and no cartilage grated boost technique is employed if this schedule sclerosis [n = 3] or volume > 3.5 cm3 and single cartilage is chosen. The high-risk planning target volume re- sclerosis [n = 6]), and high risk (volume > 3.5 cm3 and ceives 70 Gy at 2 Gy per fraction, the intermediate- ≥ 2 cartilage sclerosis [n = 7]). Rates of local control after risk planning target volume receives 63 Gy at 1.8 Gy RT were 90% (19 of 21) for the low-risk group, 43% (6 of per fraction, and the low-risk planning target vol- 14) for the moderate-risk group, and 14% (1 of 7) for the ume receives 56 Gy at 1.6 Gy per fraction.7,10 Patients high-risk group.11 receive concomitant cisplatin 30 mg/m2 per week.7,10 Hinerman et al8 reported on 109 previously un- Conclusions treated patients with fixed-cord T3 (n = 87) and T4 Radiotherapy can produce a high rate of local control (n = 22) SCC of the glottic larynx treated between the for patients with squamous cell carcinoma of the glot- years 1966 and 2002. The stage distribution was T3N0 tic larynx. It also can be used to help maintain a func- in 68 patients (62%), T3N1 in 14 patients (13%), T3N2b tional larynx in patients with carcinoma in situ, T2 dis- in 5 patients (5%), T4N0 in 17 patients (16%), T4N1 in ease, and low-volume T3 to T4 disease. 4 patients (4%), and T4N2b in 1 patient.8 Five patients received induction chemotherapy with cisplatin and References fluorouracil, 4 patients received concomitant cisplatin 1. Greene FL, Page DL, Fleming ID, et al. AJCC Cancer Staging Manual. alone or combined with fluorouracil, 1 patient received 6th ed. New York: Springer; 2002.

July 2016, Vol. 23, No. 3 Cancer Control 211 2. Edge S, Byrd DR, Compton CC, et al, eds. AJCC Cancer Staging Manual. 7th ed. New York: Springer; 2010. 3. Dagan R, Morris CG, Bennett JA, et al. Prognostic significance of paraglottic space invasion in T2N0 glottic carcinoma. Am J Clin Oncol. 2007;30(2):186-190. 4. Sengupta N, Morris CG, Kirwan J, et al. Definitive radiotherapy for carcinoma in situ of the true vocal cords. Am J Clin Oncol. 2010;33(1):94-95. 5. Mendenhall WM, Werning JW, Hinerman RW, et al. Management of T1-T2 glottic carcinomas. Cancer. 2004;100(9):1786-1792. 6. Chera BS, Amdur RJ, Morris CG, et al. T1N0 to T2N0 squamous cell carcinoma of the glottic larynx treated with definitive radiotherapy. Int J Radiat Oncol Biol Phys. 2010;78(2):461-466. 7. Mendenhall WM, Riggs CE, Vaysberg M, et al. Altered fractionation and adjuvant chemotherapy for head and neck squamous cell carcinoma. Head Neck. 2010;32(7):939-945. 8. Hinerman RW, Mendenhall WM, Morris CG, et al. T3 and T4 true vocal cord squamous carcinomas treated with external beam irradiation: a single institution’s 35-year experience. Am J Clin Oncol. 2007;30(2):181-185. 9. Mendenhall WM, Morris CG, Amdur RJ, et al. Parameters that predict local control after definitive radiotherapy for squamous cell carcinoma of the head and neck. Head Neck. 2003;25(7):535-542. 10. Newlin HE, Amdur RJ, Riggs CE, et al. Concomitant weekly cisplatin and altered fractionation radiotherapy in locally advanced head and neck cancer. Cancer. 2010;116(19):4533-4540. 11. Pameijer FA, Mancuso AA, Mendenhall WM, et al. Can pretreatment computed tomography predict local control in T3 squamous cell carcinoma of the glottic larynx treated with definitive radiotherapy? Int J Radiat Oncol Biol Phys. 1997;37(5):1011-1021.

212 Cancer Control July 2016, Vol. 23, No. 3 Proton beam therapy may

be beneficial in reducing toxicity

and improving outcomes

in select tumors of the skull base.

Artwork courtesy of Arthur Pina de Alba. www.arthurpinadealba.com.

Clinical Benefits of Proton Beam Therapy for Tumors of the Skull Base Kamran A. Ahmed, MD, Stephanie K. Demetriou, Mark McDonald, MD, and Peter A.S. Johnstone, MD

Background: The unique radiobiological properties of protons have been understood for many years. In addition, many of the clinical benefits of radiotherapy were first noted in tumors involving the skull base. More public attention has been given to proton beam therapy due to the increasing number of centers now in operation or in the planning stages for offering this treatment option. Methods: We reviewed the physical properties of protons and the clinical studies performed to justify their use in the management of skull-base tumors and determine the benefits of proton beam therapy. Results: Published reports suggest a benefit to proton beam therapy for use in tumors of the skull base, including craniopharyngiomas, chordomas, skull-base sarcomas, and unresectable meningiomas. Conclusions: Use of proton beam therapy may be beneficial in select patients. Surgical and medical oncologists should have a general understanding of such cases to facilitate their appropriate referral.

Introduction ton beam therapy is being investigated in many other Proton beam therapy has risen in popularity in recent disease sites, including tumors of the lung, pancreas, years as a treatment modality in the field of radiation esophagus, prostate, and cervix.6-10 oncology. More than 10 centers are in operation with In this review, we summarize the unique proper- many more proton centers under construction or in ties of protons that make them particularly beneficial the planning stages.1 There are proven benefits to pro- for the treatment of skull-base lesions and review the ton treatment in select patients, including those with results of clinical studies supporting their use. chordomas, skull-base tumors, and in the setting of childhood malignancies where doses initiated in prox- Basic Proton Dosimetry imity to adjacent normal structures are of critical im- Although proton beam therapy has been receiving portance.2-5 The potential therapeutic benefit of pro- more attention in recent years, the unique physical properties of protons as charged particles have been From the Department of Radiation Oncology (KAA, SKD, PASJ), understood for many years.1 Both photons and elec- H. Lee Moffitt Cancer Center & Research Institute, Tampa, Florida, trons deposit the majority of their energy at the surface Health Proton Therapy Center (MM), Indiana University, Bloomington, Indiana. close to their entrance at the body. After transferring Submitted February 24, 2015; accepted June 4, 2015. this energy, the particle continues to deposit energy Address correspondence to Peter A.S. Johnstone, MD, Mof- along its path but at a decreasing rate. Thus, the major- fitt Cancer Center, 12902 Magnolia Drive, Tampa, FL 33612. ity of the dose of radiation prescribed to a patient is of- E-mail: [email protected] ten deposited outside of the target location. Therefore, No significant relationships exist between the authors and the companies/organizations whose products or services may be ref- the key goal of radiotherapy is to deliver high ablative erenced in this article. doses to gross or microscopic disease while limiting the

July 2016, Vol. 23, No. 3 Cancer Control 213 dose of radiation to normal tissues and to avoid exceed- Range Modulation ing normal tissue tolerance, which can lead to toxicity. Even though the fact that protons deposit the majority Due to this concern, numerous strategies have of their dose at a certain distance from their entrance been developed to decrease the dose of radiation de- seems appealing, few tumors are covered by the nar- livered to normal tissues and include multiple fields, row range provided by a pristine Bragg peak. The so- 3-dimensional (3D) conformal therapy, advanced inten- lution to this has been to create a so-called spread-out sity-modulated radiotherapy (IMRT)/computed tomog- Bragg peak (SOBP), which allows for radiation to be raphy, stereotactic radiosurgery, and arc therapy. Treat- deposited throughout the extent of the tumor (Fig 2).15 ment planning for IMRT and inverse radiotherapy has The SOBP is created using a method known as range advanced the field of radiation oncology, reducing the modulation. The rate of penetration is directly related doses received to normal structures and allowing the es- to beam energy: A higher energy beam deposits its calation of doses to target areas. Studies have shown sig- dose further than a lower energy beam. nificant improvement in tumor targeting and normal tissue sparing, thus improving local control and reducing toxicity rates.11-14 IMRT has now become the stan- 100 dard of care in select adult malignancies, including head and 80 neck cancers,11 gastrointestinal malignancies (eg, esophageal,12 60 anal cancer13), and prostate cancer.14 Although these techniques can significantly reduce the dose 40 delivered to normal tissues, both Relative Dose (%) acute and late toxicities remain a 20 concern.11-14 The pattern of dose deposi- 0 tion in proton particles differs 0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 from that of photons and elec- Depth (cm) trons. As the proton beam travels through tissue and maintains Fig 1. — Proton Bragg peak demonstrating peak energy deposition at the termination of the pro- velocity and energy, it begins to ton path. From McDonald MW, Fitzek MM. Proton therapy. Curr Probl Cancer. 2010;34(4):257-296. slow near the end of its path. At Reprinted with permission by Elsevier. this point, it abruptly deposits the majority of its energy. No further energy deposition occurs at the end of this path. This phenomenon is known as the Bragg 100 peak (Fig 1).15 This makes protons different from electrons and 80 photons, which do not exhibit a similar Bragg peak phenomenon. All 3 particles — photons, 60 electrons, and protons — have a place in modern radiotherapy, 40 each with unique benefits and Relative Dose (%) drawbacks. In tumors likely to 20 benefit from proton beam therapy, the decision to use proton vs 0 photon therapy either with IMRT 10 12 14 16 18 20 22 24 26 or 3D conformal therapy should Depth (cm) be made on a case-by-case basis, with the clinician evaluating the Fig 2. — Summation of individual Bragg peaks to create the spread-out Bragg peak. From McDonald necessary target coverage and MW, Fitzek MM. Proton therapy. Curr Probl Cancer. 2010;34(4):257-296. Reprinted with permission sparing of critical structures. by Elsevier.

214 Cancer Control July 2016, Vol. 23, No. 3 A range modulator places varying thicknesses of solete with the introduction of proton beam therapy. absorbing material in front of the beam to modify an No dose exists downstream of the Bragg peak for otherwise monoenergetic beam. This yields a spec- any individual beam; however, with compensation trum of proton beam energies, each with a different and range modulation, different depths of individual point of maximum dose deposition. When these in- beams may be included in a plan. For instance, match- dividual beamlets are summed together, the SOBP ing “through” and “patch” fields in a single plan allows is created. A consequence of employing such a strat- for noncoplanar sparing of the brainstem, spinal cord, egy is that a once relatively low entrance dose of a or optic chiasm (Fig 4). single proton beam is repeated by each of the beam The concepts of “through” and “patch” fields are energies, such that the cumulative dose is increased not valid in photon radiotherapy and are infrequent- when delivering the SOBP. However, the unique dose ly used in proton radiotherapy. No conceivable bene- fall-off rate following dose deposition after the final fit exists to their use in treating prostate cancer or in Bragg peak is preserved. standard craniospinal proton beam therapy; however, these concepts are critical when treating extensive le- Through and Patch Fields Historically, the radiotherapy concepts of “dose” and “target” were, at best, 2 dimensional. The delivered dose was visualized as the intersection of few irregularly shaped cylinders within the patient to form an irregular polyhedron — and was almost always an oddly shaped box.16 Sometimes, with arc treatments, a partial plane of distributed dose was delivered, with increasing intensity to a high-dose apex at the target depth, then fanning out again at exit. Because software for noncoplanar calculation did not exist in the 1980s, noncoplanar beams were rare. Beam-eye view involved drawing block outlines on the skin, and clinicians were most comfortable with relatively few gantry angles. In the 1990s, combined computed tomography/3D dose-calculation algorithms were introduced.17 En- hanced graphics added color and patient-specific or- Fig 3. — Radiation model revealing target dose entry and exit in 3-dimensional space. From Moore EM, Magrino TJ, Johnstone PA. Rectal bleed- gan anatomy to the 1980s framework, and clinicians ing after radiation therapy for prostate cancer: endoscopic evaluation. could now visualize the patient’s body from any angle Radiology. 2000;217(1):215-218. Reprinted with permission by the (Fig 3).16 Doing so allowed more frequent use of nonco- Radiological Society of North America. planar beams, off-axis calculations, multiple beam ar- rangements, and, ultimately, 3D conformal radiotherapy.17 Conceptually, however, the radiotherapy dose was A C generally delivered via a planar dose bath with more sites of beam entry/exit, although exceptions did exist (eg, use of oblique or tangential fields could minimize the axial volume treated for appropriate lesions). In the 2000s, intensity modulation, dynamic multi- leaf collimation, and more sophisticated beam-eye view software and image-fusion capacity became available.18 In this 4-dimensional environment, the perception of “target” changed. The target volume was anatomically defined and, thus, usually irregular. It could contain different volumes receiving different delivery doses within the same duration of therapy. However, despite the so- phistication of the display and the complexity of beams and fluences, the mental picture from the 1990s of a planar dose bath remained. The degree of “planarism” B D is based on the dose used; lower doses still outline a slice through the patient, with few arcs remaining dose Fig 4A–D. — (A) Right and left lateral “through” fields, (B) bilateral posterior oblique “patch” fields abutting the brainstem, (C) 4 proton beams free because they may provide no benefit. targeting a midline skull-base lesion (red), and (D) final dose distribution The construct of the planar dose bath became ob- showing brainstem sparing.

July 2016, Vol. 23, No. 3 Cancer Control 215 sions of the brainstem and optic apparatus to high cords of 52 children treated with proton beam therapy doses. A proton “through” field does not traverse the (n = 21) vs photon IMRT (n = 31). No difference was patient, but it does traverse the target. A proton “patch” detected between the treatment modalities in rates of field is added when the anatomical or spatial complex- survival, solid and cystic disease control, and late tox- ity of a lesion abuts a dose-limited normal structure, icities when measured for a median of 33 months for and the through field must sacrifice lesion coverage to those receiving proton beam therapy and 106 months avoid doses to normal tissue. for those receiving IMRT.22 However, another study reported on significant reductions in doses to normal Skull-Base Sarcomas structures such as the hippocampus, dentate gyrus, The unique physical properties distinguishing proton subventricular zone, and vascular structure while pre- beam therapy from other radiation modalities have serving target coverage in a comparison of 3D proton been employed to treat select malignancies.3,4,19 Due and intensity-modulated proton beam therapy with to their proximity to critical structures in the brain, photon IMRT.23 However, prospective, quality-of-life such as the brainstem, optic nerves, and optic chiasm, and formal neurocognitive testing are needed for more skull-base sarcomas can be challenging to manage. sensitive measures of benefit to determine how long-term Complete resection of these malignancies is often not toxicity profiles differ between the management of possible. Use of proton therapy for these malignan- these rare malignancies and when comparing proton cies offers clinicians the opportunity to deliver po- beam therapy with photon treatment. tentially ablative doses to areas of gross disease while avoiding toxicity in the brainstem and optic struc- Chordomas tures. Proton therapy also has the potential to pre- Chordomas are rare primary bone tumors thought to serve excellent local control while also limiting the arise from remnants of the embryonal notochord, al- dose to normal structures. though they can arise anywhere along the spine as One group of researchers reported dosimetric well as the base of the skull. When affecting the base differences between proton beam plans with photon of skull, chordomas can invade the clivus, petrous IMRT for the treatment of parameningeal rhabdomyo- bone, and extend to the cavernous sinus with direct sarcomas in pediatric patients.19 Acceptable target vol- involvement of cranial nerves, potentially leading to ume coverage was achieved using both modalities, cranial nerve palsies. These tumors are challenging to and the proton plans resulted in significant sparing of treat when located in the skull base because surgical all examined normal tissues, except for the ipsilateral resection may lead to poor outcomes, with gross resid- cochlea and mastoid.19 In a series involving 16 chil- ual disease requiring postoperative radiotherapy. How- dren with rhabdomyosarcoma, similar results were ever, dose escalation with radiation doses greater than found, including rates of good local control and ac- 60 Gy has been proven to lead to improved outcomes.24 ceptable toxicity profiles.3 Furthermore, when com- Due to the proximity to critical structures, the delivery paring 3D proton planning to 3D photon planning, of acceptable doses of radiation to achieve local control another study showed that protons were able to spare can represent a challenge for many clinicians. Many radiation to normal tissues in the brain and contralat- studies have been conducted on the benefit of proton eral orbital structures while also maintaining tumor beam therapy for the management of chordomas, mak- coverage in 7 children with orbital rhabdomyosarco- ing it the current standard of care for managing these ma who received treatment.4 rare tumors.25-28 Protons have also been instrumental in the dose Craniopharyngiomas escalation of chordomas (> 70 Gy); a steep dose-gradi- Craniopharyngiomas are histologically benign but agent limits toxicity to critical normal tissues.5 Long-term gressive tumors originating from the remnants of the follow-up of patients with chordomas whose condi- Rathke pouch, and they typically affect children and tions were managed with proton beam therapy has adolescents. Because of the location of these tumors in also been reported.29 With a median dose of 72.1 Gy the suprasellar region, the temporal lobes, hippocam- for combined proton/photon beam therapy, a 5-year pus, hypothalamus, optic nerves, and chiasm may be local control rate of 73% has been reported.29 Similar at risk for acute and late effects of radiotherapy. findings have been demonstrated in a report of 33 pa- Several institutions have reported their results in tients with chordomas and 25 with chondrosarcomas the treatment of these rare tumors.20-22 A study from who received radiotherapy.25 The mean target dose a proton facility reported 5- and 10-year local control was 70.7 Co Gy equivalent (range, 64.8–79.2), and lo- rates of 93% and 85%, respectively, in a group of 5 chil- cal control rates of 76% were achieved for those with dren and 10 adults.20 Similar outcomes were reported chordomas and 92% for those with chondrosarcomas.25 by Luu et al21 in a group of 16 patients with accept- Similar local control results were reported by Ares et al26 able toxicity profiles. Bishop et al22 compared the rein a cohort of 64 patients who had skull-base chor-

216 Cancer Control July 2016, Vol. 23, No. 3 domas (n = 42) or chondrosarcomas (n = 22). They were and 2 cases of grade 3 toxicity. Improvements were treated with protons using a spot-scanning technique, seen in eye-related symptoms for the majority of the and the median total dose was 73.5 Gy for those with study patients.34 chordomas.26 The 5-year disease-specific survival rate Another study reported on the management of was 81% with a 5-year rate of freedom from high-grade atypical and malignant meningiomas.35 In this series, toxicity of 94%.26 24 study patients were treated with a mean proton to- In the setting of reirradiation for 16 study par- tal dose of 34.05 Co Gy equivalent and a mean photon ticipants with progressive or recurrent chordomas, total dose of 30.96 Co Gy equivalent.35 The authors of McDonald et al30 demonstrated that definitive pro- that study suggested that a proton boost may improve ton beam therapy, with or without salvage surgery, outcomes and rates of survival with dose escalation.35 was encouraging for initial disease control and over- No acute cases of morbidity related to radiotherapy all survival. The median prior dose of radiation was were reported. 75.2 Gy, and the median dose of reirradiation was 75.6 Gy (relative biological equivalent).30 The 2-year Midsagittal Lesions local control rate was 85%, the overall survival rate Midline central nervous system lesions are optimally was 80%, and the rate of developing distant metasta- treated with proton beam therapy using modified mid- sis was 20%.30 Despite high cumulative, lifetime doses sagittal proton schemas that allow normal tissue to of radiation, acceptable toxicities were reported in be spared in most of the uninvolved brain while also 3 patients who experienced grade 3/4 events but no limiting doses to at-risk organs. Estabrook et al36 dem- grade 5 events; however, the researchers noted that onstrated the dosimetric advantage of using protons additional toxicities could potentially develop with a in 9 study patients (n = 5 craniopharyngiomas; n = 4 longer follow-up period.30 meningiomas). The Bragg peak characteristic of pro- Thus, these reports show that proton beam therapy tons allowed for the use of vertex and anterior and/or has some benefit for the management of chordomas. posterior superior oblique beams along the midsagittal plane to achieve a mean dose of 52.2 Gy (relative bio- Meningiomas logical equivalent) to the target.36 The decreased num- In general, meningiomas are benign tumors that arise ber of treatment beams and unique angles achieved from the meninges. Surgical resection is the standard with protons resulted in an improved plan distinct of care, and radiotherapy is typically reserved for in- from what is currently possible with IMRT photons.36 complete resections, adverse pathology, or for those The reported midline central nervous system proton tumors that are not resectable.31,32 For tumors that in- plans resulted in minimal doses to the normal tissue of volve the base of the skull, cavernous sinus, or optic the brain, as demonstrated by average mean doses of chiasm, radiotherapy alone may represent a manage- 18 Gy (relative biological equivalent) to the brainstem ment option. Although intensity modulation and daily (range, 0.0–40.1) and 17.1 Gy (relative biological equiv- image guidance allows these lesions to be treated with alent) to the hippocampi (range, 0.0–45.9).36 photons and a relatively low relative dose to the normal Fig 5 depicts a proton beam arrangement for para- brain, the volume of normal brain in the treatment field falcine meningioma.37 is often extensive. Although using vertex or angled, superior photon fields may be a viable option to decrease Conclusions brain hemispheric coverage, doing so often leads to a The radiobiological benefits of proton beam therapy high exit dose to the brainstem. Due to the benign na- have been understood for many years. As proton beam ture of these tumors, acute and late sequelae of treat- therapy becomes available for more and more patients, ment represent a primary concern. it will become increasingly important that the benefits Proton therapy offers benefits to select menin- of this treatment be based on solid clinical evidence. giomas located proximal to critical structures or in The majority of the studies reviewed were retrospec- the setting of recurrence.33-35 One such experience tive in nature. As more institutions become equipped was reported in 46 study patients receiving treat- to deliver this therapeutic option, more opportunities ment, 9 of whom received treatment following in- will exist for the development of prospective trials to complete tumor resection, 8 following tumor bi- study proton beam therapy. opsy, and 29 following tumor recurrence after gross Due to the rare nature of these tumors, multi-in- total resection and combined photon/proton beam stitutional and collaborative efforts will be important treatment plans.33 A 10-year recurrence-free survival to increase our knowledge. At this time, many tumors rate of 88% was achieved, as was a 10-year toxicity-free encompassing regions of the skull base have demon- survival rate of 80%.33 Noël et al34 reported on 51 study strated a proven benefit from proton treatment based patients treated with combined proton beam/photon on retrospective results. Studies into dose escalation radiotherapy, noting excellent rates of local control and conformal treatment plans with proton beam

July 2016, Vol. 23, No. 3 Cancer Control 217 A B C

Fig 5A–C. — Midsagittal proton beam array for parafalcine meningioma with (A) sagittal, (B) coronal, and (C) axial views. From McMullen KP, Ker- stiens J, Johnstone PA. Practical aspects of pediatric proton radiation therapy. Cancer J. 2014;20:393-396. Reprinted with permission by Wolters Kluwer Health.

therapy may further improve outcomes in these dis- 15. McDonald MW, Fitzek MM. Proton therapy. Curr Probl Cancer. 2010;34(4):257-296. ease sites without compromising the risk of toxicity to 16. Moore EM, Magrino TJ, Johnstone PA. Rectal bleeding after normal structures. In addition, future research might radiation therapy for prostate cancer: endoscopic evaluation. Radiology. 2000;217(1):215-218. also reveal that the benefits of proton beam therapy to 17. Bucci MK, Bevan A, Roach M III. Advances in radiation therapy: skull-base lesions can translate to treatment options for conventional to 3D, to IMRT, to 4D, and beyond. CA Cancer J Clin. 2005;55(2):117-134. other malignancies. 18. Teh BS, Woo SY, Butler EB. Intensity modulated radiation therapy (IMRT): a new promising technology in radiation oncology. Oncologist. 1999;4(6):433-442. References 19. Kozak KR, Adams J, Krejcarek SJ, et al. A dosimetric comparison of proton and intensity-modulated photon radiotherapy for pediat- 1. Foote RL, Stafford SL, Petersen IA, et al. The clinical case for proton ric parameningeal rhabdomyosarcomas. Int J Radiat Oncol Biol Phys. beam therapy. Radiat Oncol. 2012;7:174. 2009;74(1):179-186. 2. MacDonald SM, Safai S, Trofimov A, et al. Proton radiotherapy for 20. Fitzek MM, Linggood RM, Adams J, et al. Combined proton and childhood ependymoma: initial clinical outcomes and dose comparisons. photon irradiation for craniopharyngioma: long-term results of the early Int J Radiat Oncol Biol Phys. 2008;71(4):979-86. cohort of patients treated at Harvard Cyclotron Laboratory and Massachusetts 3. Timmermann B, Schuck A, Niggli F, et al. Spot-scanning proton General Hospital. Int J Radiat Oncol Biol Phys. 2006;64(5):1348-1354. therapy for malignant soft tissue tumors in childhood: first experiences at 21. Luu QT, Loredo LN, Archambeau JO, et al. Fractionated proton the Paul Scherrer Institute. Int J Radiat Oncol Biol Phys. 2007;67:497-504. radiation treatment for pediatric craniopharyngioma: preliminary report. 4. Yock T, Schneider R, Friedmann A, et al. Proton radiotherapy for Cancer J. 2006;12(2):155-159. orbital rhabdomyosarcoma: clinical outcome and a dosimetric comparison 22. Bishop AJ, Greenfield B, Mahajan A, et al. Proton beam therapy with photons. Int J Radiat Oncol Biol Phys. 2005;63(4):1161-1168. versus conformal photon radiation therapy for childhood craniopharyngi- 5. Austin-Seymour M, Munzenrider J, Goitein M, et al. Fractionated oma: multi-institutional analysis of outcomes, cyst dynamics, and toxicity. proton radiation therapy of chordoma and low-grade chondrosarcoma of Int J Radiat Oncol Biol Phys. 2014;90(2):354-361. the base of the skull. J Neurosurg. 1989;70(1):13-17. 23. Boehling NS, Grosshans DR, Bluett JB, et al. Dosimetric com- 6. Hong TS, Ryan DP, Blaszkowsky LS, et al. Phase i study of pre- parison of three-dimensional conformal proton radiotherapy, intensity- operative short-course chemoradiation with proton beam therapy and modulated proton therapy, and intensity-modulated radiotherapy for capecitabine for resectable pancreatic ductal adenocarcinoma of the treatment of pediatric craniopharyngiomas. Int J Radiat Oncol Biol Phys. head. Int J Radiat Oncol Biol Phys. 2011;79(1):151-157. 2012;82(2):643-652. 7. Mizumoto M, Okumura T, Hashimoto T, et al. Proton beam therapy 24. Amichetti M, Cianchetti M, Amelio D, et al. Proton therapy in for hepatocellular carcinoma: a comparison of three treatment protocols. chordoma of the base of the skull: a systematic review. Neurosurg Rev. Int J Radiat Oncol Biol Phys. 2011;81(4):1039-1045. 2009;32(4):403-416. 8. Mendenhall NP, Li Z, Hoppe BS, et al. Early outcomes from three 25. Hug EB, Loredo LN, Slater JD, et al. Proton radiation therapy prospective trials of image-guided proton therapy for prostate cancer. Int J for chordomas and chondrosarcomas of the skull base. J Neurosurg. Radiat Oncol Biol Phys. 2012;82(1):213-221. 1999;91(3):432-439. 9. Kagei K, Tokuuye K, Okumura T, et al. Long-term results of proton 26. Ares C, Hug EB, Lomax AJ, et al. Effectiveness and safety of spot beam therapy for carcinoma of the uterine cervix. Int J Radiat Oncol Biol scanning proton radiation therapy for chordomas and chondrosarcomas Phys. 2003;55(5):1265-1271. of the skull base: first long-term report. Int J Radiat Oncol Biol Phys. 10. Chang JY, Komaki R, Lu C, et al. Phase 2 study of high-dose proton 2009;75(4):1111-1118. therapy with concurrent chemotherapy for unresectable stage III nonsmall 27. Noël G, Feuvret L, Calugaru V, et al. Chordomas of the base of the cell lung cancer. Cancer. 2011;117(20):4707-4713. skull and upper cervical spine. One hundred patients irradiated by a 3D 11. Lee N, Puri DR, Blanco AI, et al. Intensity-modulated radiation thera- conformal technique combining photon and proton beams. Acta Oncol. py in head and neck cancers: an update. Head Neck. 2007;29(4):387-400. 2005;44(7):700-708. 12. Lin SH, Wang L, Myles B, et al. Propensity score-based compari- 28. Igaki H, Tokuuye K, Okumura T, et al. Clinical results of proton son of long-term outcomes with 3-dimensional conformal radiotherapy vs beam therapy for skull base chordoma. Int J Radiat Oncol Biol Phys.

intensity-modulated radiotherapy for esophageal cancer. Int J Radiat Oncol 2004;60(4):1120-1126. Biol Phys. 2012;84(5):1078-1085. 29. Munzenrider JE, Liebsch NJ. Proton therapy for tumors of the skull 13. Kachnic LA, Winter K, Myerson RJ, et al. RTOG 0529: a phase 2 base. Strahlenther Onkol. 1999;175(suppl 2):57-63. evaluation of dose-painted intensity modulated radiation therapy in com- 30. McDonald MW, Linton OR, Shah MV. Proton therapy for reirradia- bination with 5-fluorouracil and mitomycin-c for the reduction of acute tion of progressive or recurrent chordoma. Int J Radiat Oncol Biol Phys. morbidity in carcinoma of the anal canal. Int J Radiat Oncol Biol Phys. 2013;87(5):1107-1114. 2013;86(1):27-33. 31. Korah MP, Nowlan AW, Johnstone PA, et al. Radiation therapy 14. Sheets NC, Goldin GH, Meyer AM, et al. Intensity-modulated radia- alone for imaging-defined meningiomas. Int J Radiat Oncol Biol Phys. tion therapy, proton therapy, or conformal radiation therapy and morbidity and 2010;76(1):181-186. disease control in localized prostate cancer. JAMA. 2012;307(15):1611-1620.

218 Cancer Control July 2016, Vol. 23, No. 3 32. Debus J, Wuendrich M, Pirzkall A, et al. High efficacy of fractionated stereotactic radiotherapy of large base-of-skull meningiomas: long-term results. J Clin Oncol. 2001;19(15):3547-3553. 33. Wenkel E, Thornton AF, Finkelstein D, et al. Benign meningioma: partially resected, biopsied, and recurrent intracranial tumors treated with combined proton and photon radiotherapy. Int J Radiat Oncol Biol Phys. 2000;48(5):1363-1370. 34. Noël G, Bollet MA, Calugaru V, et al. Functional outcome of patients with benign meningioma treated by 3d conformal irradiation with a combination of photons and protons. Int J Radiat Oncol Biol Phys. 2005;62(5):1412-1422. 35. Boskos C, Feuvret L, Noël G, et al. Combined proton and photon conformal radiotherapy for intracranial atypical and malignant meningioma. Int J Radiat Oncol Biol Phys. 2009;75(2):399-406. 36. Estabrook NC, McDonald MW, Hoene TA, et al. Proton radiotherapy for midline central nervous system lesions: a class solution. Oncology. 2015;89(2):111-117. 37. McMullen KP, Kerstiens J, Johnstone PA. Practical aspects of pediatric proton radiation therapy. Cancer J. 2014;20(6):393-396.

July 2016, Vol. 23, No. 3 Cancer Control 219 RT plays a role in the treatment of BCC

and SCC of the head and neck.

Artwork courtesy of Arthur Pina de Alba. www.arthurpinadealba.com.

Management of BCC and SCC of the Head and Neck Tobin J. Strom, MD, Jimmy J. Caudell, MD, PhD, and Louis B. Harrison, MD

Background: For decades radiotherapy (RT) has been shown to treat skin cancers; however, the indications, delivery methods, and techniques for RT continue to evolve. Methods: Relevant prospective and retrospective reports were reviewed that addressed outcomes with, indications for, and delivery techniques used with RT for the management of cutaneous basal cell carcinoma (BCC) and squamous cell carcinoma (SCC) of the head and neck. Results: Rates of local control higher than 90% are typically achievable for early-stage BCC and SCC of the head and neck. RT is often recommended for tumors located in cosmetically or functionally sensitive areas of the face, for patients who cannot tolerate anesthesia, for those taking anticoagulants, or for patients who prefer RT to other treatment options. A wide range of radiation doses, daily fractionation schedules, and radiation techniques have been shown to be effective for management. In general, postoperative local radiation is recommended following excision for patients with high-risk factors, including those whose tumors have close or positive margins, perineural invasion, invasion of the bone or nerves, or those with recurrent disease. Conclusions: RT plays an integral role in the treatment of primary and postoperative cutaneous BCC and SCC of the head and neck. Prospective trials are in progress to address the roles of concurrent systemic therapy and RT for both cutaneous BCC and SCC.

Introduction to ultraviolet light or sunlight, changing clothing styles, An estimated 5.4 million cases of cutaneous basal cell and increasing longevity.2-4 The socioeconomic burden carcinoma (BCC) and squamous cell carcinoma (SCC) of these cancers is often underestimated because BCC are diagnosed every year in the United States — a num- and SCC are frequently excluded from cancer regis- ber that is more than all other types of cancer com- tries.1,3,5 Cutaneous BCC and SCC typically present lo- bined.1 Many factors play a role in the development of cally — rather than regionally or distant — and may be cutaneous SCC and BCC, and the incidence of these curable with local management. cancers is also rising as a result of increased exposure Cutaneous BCC arises from pluripotent cells in the basal layer of the epidermis and tends to develop following exposure to the sun during childhood.2,6 BCC From the Department of Radiation Oncology, H. Lee Moffitt Cancer Center & Research Institute, Tampa, Florida. often presents as slow-growing, translucent papules Address correspondence to Louis B. Harrison, MD, Department of with raised, telangiectatic borders on the head and Radiation Oncology, Moffitt Cancer Center, 12902 Magnolia Drive, neck. Rarely, the tumors metastasize to the regional Tampa, FL 33612. E-mail: [email protected] lymph nodes (0.01%–0.50% of cases), although lo- Submitted February 26, 2015; accepted April 14, 2016. cally advanced lesions can invade nearby structures No significant relationships exist between the authors and the companies/organizations whose products or services may be ref- and cause symptoms such as numbness, pain, and erenced in this article. weakness.7-9 In general, BCC has a slow rate of annual

220 Cancer Control July 2016, Vol. 23, No. 3 growth (≤ 5 m) and, compared with SCC, has a higher The American Joint Commission on Cancer has cure rate following local management.7,10 taken many of these risk factors into account in its SCC arises from epithelial keratinocytes and tends combined staging system of BCC and SCC, defining to progress over decades of exposure to the sun from factors such as Breslow thickness (> 2 mm), Clark level precursor lesions known as actinic keratoses, which (≥ IV), perineural invasion, primary tumor locations of are typically small, red, scaly, 1- to 3-mm papules with the ear or hair-bearing lips, and poorly differentiated a rough texture.7,11 Compared with BCC, SCC has a or undifferentiated tumors as being high risk.17 Given higher risk of local recurrence (8%–15% of cases) and that prospective validation of these risk factors is lack- distant metastasis (0.5%–16.0% of cases).9,12-14 The risk ing, factors considered to be high risk by the Ameri- of recurrence or metastasis with BCC and SCC depends can Joint Commission on Cancer are primarily based on multiple risk factors. on consensus opinion; thus, in response, new staging schema have been proposed.15,20,28 Risk Factors for Recurrence High-risk factors for locoregional recurrence and dis- Radiotherapy tant metastases following definitive therapy of BCC Radiotherapy (RT) plays an integral role in the treat- and SCC are important prognostic information and ment of cutaneous BCC and SCC of the head and neck central for appropriate management recommenda- and can be used in the definitive setting, adjuvant set- tions. Most risk factors have been identified based ting, or both. Patient selection for RT is important and on retrospective studies and, thus, lack prospective best determined in the setting of multidisciplinary validation.15 Risk factors identified for BCC recur- care. Lesions in cosmetically sensitive areas (eyelids, rence include tumor size (> 2 cm), tumors located in nasal skin, central face, contour of the ear) are often the “H zone” of the face (representing embryonic fu- best managed with RT. The Fig shows 2 examples of sion planes and includes portions of the nose, scalp, large-sized SCCs on the face before and after treat- ears, and lips), recurrent tumors, perineural invasion, ment with definitive RT. Similarly, RT allows for func- poorly defined borders, and more aggressive histolo- tional preservation of organs, eyelids, nasal folds, gies (infiltrative, morpheaform, or basosquamous ears, nose, and lips, whereas surgery could lead to histology).7,16,17 functional impairment. Because of the compact and complex neurovas- Many retrospective studies have addressed the use cular anatomy of the head and neck, as well as the of definitive RT for epithelial skin cancers of the head functional and cosmetic importance of the head and and neck (Table).9,10,29-49 Local control rates with defini- neck region, wide surgical margins are often difficult tive radiation are generally considered to be excellent, to attain and therefore represent a higher risk factor but they vary based on the proportion of BCC and SCC for recurrence than other sites of the body. and recurrent tumors.9,10,29-49 Similarly, high-risk factors for cutaneous SCC include tumor lo- A B cations at specific sites of the head and neck, including the lip or ear, tumors arising from a scar, recurrent tumors, large tumor size (> 2 cm), tumor depth of more than 4 mm or Clark level of at least IV, invasion beyond the subcutaneous tissue, rapidly growing lesions, perineural invasion, desmoplasia, poor differentiation, or infiltrative margins.18-22 C D Among patients who underwent resection, positive margins have also been reported as prognostic.23,24 Of note, clinically evident perineural invasion with BCC or SCC has been shown to carry a higher risk of recurrence than incidental perineural invasion found on pathology.25,26 SCC with perineural invasion also Fig A–D. — A multifocal recurrence of SCC of the skin of the forehead (A) before and (B) after carries a higher incidence of lymph- definitive radiotherapy. A separate patient with ulcerative SCC involving the nose and right medial node metastasis.27 canthus (C) before and (D) after definitive radiotherapy. SCC = squamous cell carcinoma.

July 2016, Vol. 23, No. 3 Cancer Control 221 Table. — Select Studies of Definitive Radiotherapy for Cutaneous BCC and SCC of the Head and Neck

Study No. of Lesions, BCC SCC T4 Recurrent Follow- Overall BCC SCC Cosmesis Grade ≥ 4 Study n (%) Lesions, Lesions, Lesions, Lesions, Up, y Local Local Local (Good/ Late Patients n (%) n (%) n (%) n (%) Control Control Control Excellent), Toxicity, Rate, % Rate, % Rate, % % % Head and Neck Abbatucci39 675 675 (100) Unknown Unknown Unknown Unknown 2.0 > 96.0 (crude) Unknown Unknown 48.0 Unknown Al-Othman38 85 88 (100) 41 (46.6) 37 (42.0) 88 (100) 45 (51.1) 4.0 53 (5-y Unknown Unknown Unknown Unknown actuarial) Avril29,a 173 173 (100) 173 (100) 0 (0) 0 (0) 0 (0) 4.0 92.5 (4-yr 92.5 (4-y — 69.0 Unknown actuarial) actuarial) Cognetta36 854 854 (100) 712 (83.4) 133 (15.6) 0 (0) Unknown 2.6 97.0 (crude) 95.8 (5-y 93.3 (5-yr Unknown Unknown actuarial) actuarial) Olschewski37 85 104 (100) 104 (100) 0 (0) Unknown Unknown 3.1 100 (crude) 100 (crude) — 94.0 Unknown Petrovich32 646 646 (100) 465 (72.0) 116 (18.0) Unknown Unknown 5.0 90.7 (crude) Unknown Unknown Unknown Unknown Nasal Skin Caccialanza40 620 671 (100) 656 (97.8) 15 (2.2) 2 (0.3) 89 (13.3) 3.2 94.3 (crude) 93.3 (crude) 94.4 (crude) 74.5 Unknown Childers41 26 26 (100) 26 (100) 0 (0) Unknown 5.0 (19.0) 9.3 96.0 (crude) 96.0 (crude) — 81.0 0 Tsao33 94 94 (100) 0 (0) 94 (100) 7 (7.4) Unknown 2.9 85.0 (5-y — 85.0 (5-y Unknown 0 actuarial) actuarial) Eyelid and Medial Canthus Fitzpatrick45 1,166 1,166 (100) 1,062 (91.1) 104 (8.9) 14 (1.2) 418 (35.8) 5.0 94.9 (5-y 95.0 (5-y 93.0 (5-y Unknown Unknown (eyelid) crude) crude) crude) Krema42 90 90 (100) 90 (100) 0 (0) Unknown 26 (28.0) 6.7 94.0 (10-y 94.0 (10-y — Unknown Unknown (medial actuarial) actuarial) canthus) Rodriguez- 631 631 (100) 631 (100) 0 (0) Unknown ~ 315 6.1 95.8 (crude) 95.8 (crude) — Unknown Unknown Sains44 (50.0) (eyelid) Swanson43 23 23 (100) 23 (100) 0 (0) 0 (0) 6 (26.0) 9.9 87.0 (crude) 87.0 — Unknown 0 (medial (crude) canthus) Pinna Caccialanza46 108 115 (100) 99 (86.1) 16 (13.9) Unknown Unknown 2.4 89.6 (crude) 87.9 100 (crude) 74.8 Unknown (actuarial) Silva30 313 334 (100) 201 (60.2) 122 (36.5) 44 (13.2) 50 (15.0) 3.3 86.6 (2-y 93.0 (2-y 82.0 (2-y Unknown 5.7 actuarial) actuarial) actuarial) Multiple Sites Barysch47 179 156 (86.7) 0 (0) 180 (100) Unknown 3 (1.7) 4.9 86.7 (crude) — 86.7 (crude) Unknown Unknown Griep34 389 370 295 (75.8) 94 (24.2) Unknown 75 (19.3) 2.0 95.1 (crude) 95.9 (crude) 92.5 (crude) 73.1 Unknown (95.0) Locke31 468 491 (92.5) 389 (73.3) 142 (26.7) 31 (5.8) 167 (31.5) 5.8 88.7 (crude) 92.0 (crude) 79.6 (crude) 92.0 Unknown Lovett9 339 319 (94.1) 242 (71.4) 92 (27.1) Unknown 137 (37.5) 2.0 86.1 (crude) 90.9 (crude) 75.3 (crude) 91.6 Unknown (minimum; median not reported) Kwan10 182 163 (89.6) 61 (33.5) 121 (66.5) 29 (15.9) 98 (53.8) 3.5 70.3 (crude) 86.0 (4-y 58 (4-y Unknown Unknown actuarial) actuarial) van 333 355 (81.8) 332 (76.5) 102 (23.5) 0 (0) 0 (0) 3.6 97.5 (crude) 97.0 95.1 (crude) Unknown Unknown Hezewijk48 (crude) Zagrodnik49 148 135 (77.1) 175 (100) 0 (0) Unknown 0 (0) 4.0 84.2 (5-y 84.2 (5-y — Unknown Unknown actuarial) actuarial) aProspective trial. BCC = basal cell carcinoma, SCC = squamous cell carcinoma.

222 Cancer Control July 2016, Vol. 23, No. 3 A dermatological surgeon and radiation oncolo- by 2 weeks (33% of study patients).53 This high-dose, gist should closely collaborate with each other as well hypofractionated radiation schedule is associated with as with the patient to devise an appropriate treatment worse cosmetic outcomes.30,31,54 By contrast, cosmetic plan. Balancing time, cosmetic and functional out- outcomes are generally good or excellent and rates of comes, and cost should all factor into the treatment de- toxicities related to radiation are rare with modern ra- cision. RT is typically delivered during several weeks, diation techniques and more protracted fractionation possibly creating logistical issues with patients, where- schema when properly performed (see Table9,10,29-49). as extensive surgery with reconstruction can extend treatment and healing time. For patients taking anti- Adjuvant Therapy coagulant medication, RT offers the safest option, be- Following the wide excision of BCC and SCC, the goal cause it does not require patients to interrupt their an- of postoperative RT is to further minimize the risk of ticoagulants prior to surgery. Other issues to consider local or regional recurrence. In general, adjuvant RT is when discussing surgery and RT include patient age, appropriate when either the risk of recurrence is high treatment preference, and treatment availability.50 Clin- or the probability of successful salvage surgery is rela- ical practice guidelines from the National Comprehen- tively low.50 Adjuvant RT is often recommended after sive Cancer Network recommend that definitive RT be wide excision with close or positive margins or when offered to nonsurgical candidates; they also state that other high-risk factors are present, including perineu- radiation is often reserved for patients 60 years or older ral invasion, invasion of the bone or nerves, or recur- because of concern for long-term sequellae.51,52 rent disease.50 A trial that began enrolling patients in 1982 is the Management of regional lymphatics, including only trial ever completed that compared surgery with clinically or radiographically positive lymph nodes, RT for the definitive treatment of BCC.29 Avril et al29 typically consists of nodal dissection and adjuvant RT. randomized 347 patients with BCC of the face (tumor No prospective trials have evaluated adjuvant regional size < 4 cm) to surgery (n = 174 [≤ 2-mm margin]) or RT for cutaneous BCC or SCC, although some retro- RT (n = 173; 55% interstitial brachytherapy, 33% contact spective series have reported improved rates of locore- therapy, 12% conventional therapy). The study report- gional control with use of adjuvant RT.55-58 ed an actuarial 4-year local control rate of 99.3% with For patients with mucosal SCC of the head and surgery compared with 92.5% for RT (P = .003).29 At neck, 2 large prospective trials have assessed the util- 3 months, patient-reported cosmesis was equivalent ity of chemoradiotherapy compared with RT alone.59,60 in the 2 groups; by 4 years, 87% of participants in the A combined analysis of both trials found a significant surgical arm reported having a good cosmetic outcome improvement in rate of survival for study patients with compared with 69% of those in the RT arm (P = .05).29 positive surgical margins or extracapsular extension However, the trial took place prior to the availability who were treated with adjuvant chemoradiotherapy.61 of 3-dimensional treatment planning and, as such, ad- Prospective trials with adjuvant locoregional chemora- equate tumor volume coverage and radiation dosim- diotherapy for cutaneous BCC and SCC are also lack- etry to prevent extreme radiation hot-spots would ing. However, the results of adjuvant chemoradiothera- have been difficult, particularly for deeper tumors. py for mucosal SCC are often extrapolated and applied This could have been what led to worse cosmetic out- to cutaneous SCC, and chemoradiotherapy is often rec- comes. The recurrence rates and cosmetic outcomes ommended in patients with positive margins or extra- were also not reported by the respective method of ra- capsular extension. diation delivery — information that could have been A retrospective analysis of 61 patients with Amer- useful to clarify the outcomes and cosmetic results.29 ican Joint Commission on Cancer stage III/IV SCC of Interstitial iridium-192 brachytherapy was delivered to the skin and at least 1 risk factor, including at least 55% of study volunteers, thus causing a heterogeneous 2 lymph nodes, close (< 1 mm) or positive surgical radiation dose distribution: high radiation doses near margins, or extracapsular nodal extension, found a the catheters and, theoretically, a higher risk for skin significantly improved rate of median recurrence- toxicity and fibrosis.29 free survival with adjuvant chemoradiotherapy com- A French retrospective series reported acceptable pared with adjuvant RT alone (40.3 vs 15.4 months, rates of toxicity with interstitial brachytherapy for the respectively).62 An ongoing prospective phase 3 trial treatment of cutaneous carcinomas, but the research- is also evaluating the potential benefit of concur- ers also reported a lower rate of cosmetic toxicity when rent carboplatin with RT compared with RT alone in using a dose rate of less than 98 cGy/hour53; the dose individuals with high-risk postoperative cutaneous rate in the trial by Avril et al29 was not reported. The SCC of the neck (NCT00193895). The trial is expected trial also allowed superficial, orthovoltage contact to close at the end of 2016, and its results may help therapy (50 kV) with a hypofractionated radiation guide treatment decisions with concurrent adjuvant dose schedule of 2 sessions of 18 to 20 Gy separated systemic therapy.

July 2016, Vol. 23, No. 3 Cancer Control 223 Dosing Schedule limited to use with superficial tumors and, as a result, Various radiation dose and fractionation schedules are not frequently used.7 have been used for both the definitive and adjuvant For the treatment of larger and more complex tu- treatment of BCC and SCC. Few retrospective series mors and postoperative tumor beds, treatment with have shown improved rates of local tumor control for megavoltage photons using either 3-dimensional con- epithelial skin cancers with increasing the biologi- formal or intensity-modulated RT may be appropriate. cally effective dose, increasing the daily fraction size, Intensity-modulated RT can be advantageous for larger and/or increasing the total dose of radiation.31,32,54 cutaneous head and neck tumors because of its ability For definitive treatment, radiation doses higher than to limit radiation to nearby critical structures, including 5 Gy, typically delivered every other day or twice per the parotid and submandibular glands, and it has been week, have been shown to offer equivalent rates of lo- shown to improve quality of life.63 cal control at the cost of increased toxicity and, often, Select retrospective, single-institution compari- poor cosmesis.31,32,54 sons of electrons and superficial photons have found Because of these increased rates of toxicity, dai- no difference in outcomes. Locke et al31 reviewed ly radiation doses higher than 5 Gy are typically re- and compared outcomes of 531 patients with BCC served for patients who are frail and unable to make and SCC who were treated with superficial electrons daily office visits. By contrast, radiation schedules us- (19%), superficial kV photons (60%), or a combination ing lower daily radiation doses between 2 and 5 Gy of both (20%); they found no difference in rates of lo- (ie, 3 Gy daily fractions for a total dose of 51–54 Gy cal recurrence. Similarly, Griep et al34 assessed out- or 2.5 Gy daily fractions for a total dose of 50–60 Gy) comes of 389 patients with BCC and SCC who were are commonly used due to their favorable toxicity and treated with superficial photons or electrons, and cosmetic profiles.30-34 they too reported no difference in local recurrence For lesions involving bone or cartilaginous tissues by treatment technique; however, they did report im- or for very large tumors, conventional daily radiation proved cosmetic outcomes with smaller daily doses of doses of 2 Gy are often used, with total doses rang- radiation (3 Gy daily fractions) compared with larger ing from 66 to 76 Gy.34 In general, fraction sizes of less daily doses (6–10 Gy daily fractions). Because recur- than 4 Gy offer optimal cosmesis and fraction sizes of rence is similar between techniques, the appropriate less than 3 Gy should be considered if bone or carti- technique should instead be based on factors such as lage is present in the treatment field.7 In our experi- tumor size, depth, location, and field size. ence, abbreviated schedules, such as 51 Gy in 17 fractions of 3 Gy per fraction, produce an excellent balance Cutaneous Brachytherapy of cosmetic outcomes and time sensitivity. Brachytherapy is frequently used for superficial (< 1 cm depth) skin cancers. High-dose-rate brachy- Technique therapy can incorporate 3-dimensional dosimetric Many RT techniques have been used for the treatment planning, and it can be used with surface applicators of epithelial skin cancers. The appropriate radiation or molds that conform to the contours of a patient’s technique depends on multiple factors, including pri- skin. For smaller tumors of the skin (size < 3 cm, depth mary tumor location, tumor size, scar length (in the ≤ 5-mm), a Leipzig applicator (Nucletron, Veenendaal, postoperative setting), neighboring anatomy, and The Netherlands) or Valencia applicator (Nucletron) whether the regional lymph nodes are included. can be used. For more extensive skin lesions (surface Many options are also available for the treatment depth < 1 cm), a Freiburg Flap applicator (Elekta, of cutaneous cancers. Superficial megavoltage elec- Stockholm, Sweden) or HAM applicator (Mick) is of- trons are typically used for primary lesions or tumor ten used.64 Guix et al65 reported on 102 facial cases beds 5 mm or deeper. Electrons are produced from of BCC and 34 cases of SCC definitively treated us- linear accelerators and have a dose distribution, with ing high-dose-rate brachytherapy with iridium-192 the peak dose near the skin surface and a rapid dose and surface molds. Gross tumors were treated with a fall-off beyond the target.7 Superficial orthovoltage minimum dose of 6000 to 6500 cGy in 180 cGy daily x-rays, or photons, have also traditionally been used fractions calculated at a 5-mm depth.65 The group re- to treat cutaneous BCC and SCC and provide less of a ported a 5-year actuarial local control rate of 98% for penumbra (the distance from the area receiving 20% all tumors and local control rates of 99% for primary and 90% of the dose) surrounding the radiation target tumors and 87% for recurrent tumors.65 No severe ear- than electrons. Superficial kV x-rays also have a peak ly or late complications were reported. dose near the skin surface, do not require a bolus, and work well on irregular surfaces without having to cre- Systemic Approaches to Aggressive ate a smooth, tissue-equivalent bolus, as is necessary Cutaneous Malignancies with electrons. Despite this, superficial kV x-rays are In an attempt to improve rates of locoregional control

224 Cancer Control July 2016, Vol. 23, No. 3 for high-risk skin carcinomas, systemic therapy with recurrent and metastatic mucosal SCC of the head or without concurrent RT has been investigated. The and neck when cetuximab is combined with fluoro- literature addressing systemic chemotherapy for cuta- uracil/platinum compared with fluorouracil/platinum neous BCC and SCC is generally limited to small ret- alone.77 Adding cetuximab to RT has also been shown rospective series and case reports.56-68 Substantial im- to improve rates of survival for patients with locally ad- provements have been made in our understanding of vanced mucosal SCC of the head and neck compared the molecular changes present in BCC and SCC. For ex- with RT alone.78 Despite this, the data addressing ce- ample, the sonic hedgehog signaling pathway has been tuximab use in patients with cutaneous SCC are in shown to play a key role in the development of most their infancy and largely consist of cases reports.79,80 A cases of BCC, so it is being investigated as a therapeu- pilot study reported on the use of cetuximab in 20 pa- tic target.3,69,70 The efficacy and safety of vismodegib, a tients with inoperable cutaneous SCC.81 Overall, 47% sonic hedgehog inhibitor, was first investigated by Ra- of study patients had a response; 33% with cetuximab leigh et al,71 and then was tested in patients with met- alone (n = 6), 80% with cetuximab and RT (n = 5), and astatic (n = 33) or locally advanced (n = 71) BCC that 38% with carboplatin (n = 9).81 recurred following surgery and in those who were not Two additional epidermal growth factor recep- candidates for surgery or RT.72 Results of the trial were tor tyrosine kinase inhibitors, erlotinib and gefitinib, promising. Among the study patients with metastatic have been assessed in phase 1 and 2 studies, respec- and locally advanced tumors, objective response rates tively, with concurrent RT for cutaneous SCC.35,82 In of 33.3% and 47.6%, respectively, were seen by inde- the phase 1 trial, erlotinib was concurrently delivered pendent review, as were median duration of responses with RT in 15 patients with locally advanced cutane- of 7.6 and 9.5 months, respectively.72 ous SCC (93% with T4 disease).82 The combination A second larger trial addressed rates of response therapy resulted in a grade 2/3 skin reaction in 100% of and toxicity of vismodegib among a similar group of the study volunteers, mucositis in 87%, and diarrhea in patients with metastatic (n = 31) or locally advanced 20%; the researchers concluded that the combination (n = 468) BCC who were not surgical candidates.73 therapy had an acceptable toxicity profile.82 Gefitinib The results were promising, with overall and com- was studied in a phase 2 trial in which it was first de- plete response rates by investigator review of 66.7% livered neoadjuvantly and study participants were as- and 33.8%, respectively, among patients with locally sessed for response.35 Those who responded were then advanced disease.73 The rate of median duration of treated with definitive local therapy followed by main- response was 22.7 months among study patients with tenance gefitinib. Among 22 study patients evaluable locally advanced and metastatic disease.73 Nearly all for response, 18% achieved a complete response and study participants experienced toxicities (98.0%), al- 27% had a partial response; 59% experienced adverse though grade 5 toxicities were relatively rare, occur- effects from the neoadjuvant therapy.35 Following in- ring in 21 (5.0%).73 duction, 12% were treated with surgery, 18% with RT Another sonic hedgehog inhibitor, sonidegib, was alone, 12% with concurrent gefitinib and RT, and 47% tested in a randomized phase 2 trial at 2 different dose with surgery and postoperative RT and concurrent ge- levels (200 and 800 mg) in a similar setting with both fitinib.35 The outcomes were promising: The 2-year dis- locally advanced and metastatic BCC.60 At a median ease-specific and progression-free survival rates were follow-up of nearly 14 months, 43% and 15% of study 72% and 64%, respectively.35 patients with locally advanced and metastatic BCC, However, further studies addressing the use of respectively, had an objective response at the 200-mg epidermal growth factor receptor inhibitors alone and dose level; similar response rates were seen in the concurrently with RT for cutaneous SCC are needed. higher dose group but at the expense of increased adverse events.60 Conclusions Preliminary case reports of concurrent treatment Radiotherapy (RT) plays an integral role in the treat- with RT and vismodegib have been reported with ment of primary and postoperative cutaneous basal promising results.75 A phase 2 trial is also underway to cell and squamous cell carcinomas of the head and test the safety and tolerability of vismodegib combined neck. In general, definitive RT is recommended for with RT for locally advanced BCC (NCT01835626). The tumors located in cosmetically or functionally sensi- results of this trial will help guide future trials and tive areas of the face, for patients who cannot toler- treatment decisions with respect to combined modality ate anesthesia and, often, for those who are taking treatment for BCC. anticoagulants, or for patients who prefer RT to other Similar to mucosal SCC of the head and neck, treatment modalities. A wide range of radiation dos- cutaneous SCC also frequently overexpresses epider- es, daily fractionation schedules, and radiation tech- mal growth factor receptor.76 Cetuximab has been niques are effective for basal and squamous skin can- shown to prolong survival in patients with untreated cers, and recommendations should take into account

July 2016, Vol. 23, No. 3 Cancer Control 225 World J Surg Oncol. 2012;10:117. the size, depth, location, field size, and neighboring 24. O’Brien CJ, McNeil EB, McMahon JD, et al. Significance of clinical anatomy of the tumor. stage, extent of surgery, and pathologic findings in metastatic cutaneous squamous carcinoma of the parotid gland. Head Neck. 2002;24(5):417-422. In general, adjuvant local RT is recommended fol- 25. Mendenhall WM, Amdur RJ, Hinerman RW, et al. Skin cancer of the lowing excision for patients with high-risk factors, in- head and neck with perineural invasion. Am J Clin Oncol. 2007;30(1):93-96. 26. Jackson JE, Dickie GJ, Wiltshire KL, et al. Radiotherapy for perineural cluding those with close or positive tumoral margins, invasion in cutaneous head and neck carcinomas: toward a risk-adapted perineural invasion, invasion of the bone or nerves, or treatment approach. Head Neck. 2009;31(5):604-610. 27. Moore BA, Weber RS, Prieto V, et al. Lymph node metastases from recurrent disease. Adding concurrent chemotherapy cutaneous squamous cell carcinoma of the head and neck. Laryngoscope. to adjuvant RT can be considered for patients at higher 2005;115(9):1561-1567. 28. Andruchow JL, Veness MJ, Morgan GJ, et al. Implications for risk, including those with positive tumoral margins or clinical staging of metastatic cutaneous squamous carcinoma of the head nodal extracapsular extension. Prospective trials are and neck based on a multicenter study of treatment outcomes. Cancer. 2006;106(5):1078-1083. currently underway to address the role of concurrent 29. Avril MF, Auperin A, Margulis A, et al. Basal cell carcinoma of the systemic therapy and RT for both cutaneous basal cell face: surgery or radiotherapy? Results of a randomized study. Br J Cancer. 1997;76(1):100-106. and squamous cell carcinomas. 30. Silva JJ, Tsang RW, Panzarella T, et al. Results of radiotherapy for epithelial skin cancer of the pinna: the Princess Margaret Hospital experience, 1982-1993. Int J Radiat Oncol Biol Phys. 2000;47(2):451-459. References 31. Locke J, Karimpour S, Young G, et al. Radiotherapy for epithelial 1. American Cancer Society. Cancer Facts & Figures 2016. Atlanta, GA: skin cancer. Int J Radiat Oncol Biol Phys. 2001;51(3):748-755. American Cancer Society; 2016. http://www.cancer.org/acs/groups/content/@ 32. Petrovich Z, Kuisk H, Langholz B, et al. Treatment results and pat- research/documents/document/acspc-047079.pdf. Accessed June 1, 2016. terns of failure in 646 patients with carcinoma of the eyelids, pinna, and 2. Leiter U, Eigentler T, Garbe C. Epidemiology of skin cancer. Adv Exp nose. Am J Surg. 1987;154(4):447-450. Med Biol. 2014;810:120-140. 33. Tsao MN, Tsang RW, Liu FF, et al. Radiotherapy management for 3. Goppner D, Leverkus M. Basal cell carcinoma: from the molecular squamous cell carcinoma of the nasal skin: the Princess Margaret Hospital understanding of the pathogenesis to targeted therapy of progressive experience. Int J Radiat Oncol Biol Phys. 2002;52(4):973-979. disease. J Skin Cancer. 2011;2011:650258. 34. Griep C, Davelaar J, Scholten AN, et al. Electron beam therapy is 4. Preston DS, Stern RS. Nonmelanoma cancers of the skin. N Engl J Med. not inferior to superficial x-ray therapy in the treatment of skin carcinoma. 1992;327(23):1649-1662. Int J Radiat Oncol Biol Phys. 1995;32(5):1347-1350. 5. Diepgen TL, Mahler V. The epidemiology of skin cancer. Br J Dermatol. 35. Lewis CM, Glisson BS, Feng L, et al. A phase II study of gefitinib 2002;(146 suppl 61):1-6. for aggressive cutaneous squamous cell carcinoma of the head and neck. 6. Lear JT, Tan BB, Smith AG, et al. Risk factors for basal cell car- Clin Cancer Res. 2012;18(5):1435-1446. cinoma in the UK: case-control study in 806 patients. J R Soc Med. 36. Cognetta AB, Howard BM, Heaton HP, et al. Superficial x-ray in the 1997;90(7):371-374. treatment of basal and squamous cell carcinomas: a viable option in select 7. Harrison LB, Sessions RB, Kies MS. Head and Neck Cancer: A patients. J Am Acad Dermatol. 2012;67(6):1235-1241. Multidisciplinary Approach. 4th ed. Philadelphia, PA: Lippincott Williams 37. Olschewski T, Bajor K, Lang B, et al. Radiotherapy of basal cell & Wilkins; 2014. carcinoma of the face and head: Importance of low dose per fraction on 8. Rubin AI, Chen EH, Ratner D. Basal-cell carcinoma. N Engl J Med. long-term outcome [in German]. J Dtsch Dermatol Ges. 2006;4(2):124-130. 2005;353(21):2262-2269. 38. Al-Othman MO, Mendenhall WM, Amdur RJ. Radiotherapy alone 9. Lovett RD, Perez CA, Shapiro SJ, et al. External irradiation of epi- for clinical T4 skin carcinoma of the head and neck with surgery reserved thelial skin cancer. Int J Radiat Oncol Biol Phys. 1990;19(2):235-242. for salvage. Am J Otolaryngol. 2001;22(6):387-390. 10. Kwan W, Wilson D, Moravan V. Radiotherapy for locally advanced 39. Abbatucci JS, Boulier N, Laforge T, et al. Radiation therapy of skin basal cell and squamous cell carcinomas of the skin. Int J Radiat Oncol carcinomas: results of a hypofractionated irradiation schedule in 675 cas- Biol Phys. 2004;60(2):406-411. es followed more than 2 years. Radiother Oncol. 1989;14(2):113-119. 11. Jorizzo JL, Carney PS, Ko WT, et al. Treatment options in the man- 40. Caccialanza M, Piccinno R, Percivalle S, et al. Radiotherapy of car- agement of actinic keratosis. Cutis. 2004;74(6 suppl):9 -17. cinomas of the skin overlying the cartilage of the nose: our experience in 12. Jackson A. Prevention, early detection and team management of 671 lesions. J Eur Acad Dermatol Venereol. 2009;23(9):1044-1049. skin cancer in primary care: contribution to the health of the nation objec- 41. Childers BJ, Goldwyn RM, Ramos D, et al. Long-term results of irra- tives. Br J Gen Pract. 1995;45(391):97-101. diation for basal cell carcinoma of the skin of the nose. Plast Reconstr Surg. 13. Czarnecki D, Staples M, Mar A, et al. Metastases from squamous cell 1994;93(6):1169-1173. carcinoma of the skin in southern Australia. Dermatology. 1994;189(1):52-54. 42. Krema H, Herrmann E, Albert-Green A, et al. Orthovoltage radio- 14. Quaedvlieg PJ, Creytens DH, Epping GG, et al. Histopathological therapy in the management of medial canthal basal cell carcinoma. Br J characteristics of metastasizing squamous cell carcinoma of the skin and Ophthalmol. 2013;97(6):730-734. lips. Histopathology. 2006;49(3):256-264. 43. Swanson EL, Amdur RJ, Mendenhall WM, et al. Radiotherapy 15. Farasat S, Yu SS, Neel VA, et al. A new American Joint Commit- for basal cell carcinoma of the medial canthus region. Laryngoscope. tee on Cancer staging system for cutaneous squamous cell carcinoma: 2009;119(12):2366-2368. creation and rationale for inclusion of tumor (T) characteristics. J Am Acad 44. Rodriguez-Sains RS, Robins P, Smith B, et al. Radiotherapy of peri- Dermatol. 2011;64(6):1051-1059. ocular basal cell carcinomas: recurrence rates and treatment with special 16. Sexton M, Jones DB, Maloney ME. Histologic pattern analysis of attention to the medical canthus. Br J Ophthalmol. 1988;72(2):134-138. basal cell carcinoma. Study of a series of 1039 consecutive neoplasms. 45. Fitzpatrick PJ, Thompson GA, Easterbrook WM, et al. Basal and J Am Acad Dermatol. 1990;23(6 pt 1):1118-1126. squamous cell carcinoma of the eyelids and their treatment by radiotherapy. 17. Edge S, Byrd D, Compton C, et al. AJCC Cancer Staging Manual. Int J Radiat Oncol Biol Phys. 1984;10(4):449-454. 7th ed. New York: Springer; 2010. 46. Caccialanza M, Piccinno R, Kolesnikova L, et al. Radiotherapy of 18. Clayman GL, Lee JJ, Holsinger FC, et al. Mortality risk from squamous skin carcinomas of the pinna: a study of 115 lesions in 108 patients. Int J cell skin cancer. J Clin Oncol. 2005;23(4):759-765. Dermatol. 2005;44(6):513-517. 19. Brantsch KD, Meisner C, Schonfisch B, et al. Analysis of risk factors 47. Barysch MJ, Eggmann N, Beyeler M, et al. Long-term recurrence determining prognosis of cutaneous squamous-cell carcinoma: a prospective rate of large and difficult to treat cutaneous squamous cell carcinomas study. Lancet Oncol. 2008;9(8):713-720. after superficial radiotherapy. Dermatology. 2012;224(1):59-65. 20. Jambusaria-Pahlajani A, Kanetsky PA, Karia PS, et al. Evaluation of 48. van Hezewijk M, Creutzberg CL, Putter H, et al. Efficacy of a hypofrac- AJCC tumor staging for cutaneous squamous cell carcinoma and a proposed tionated schedule in electron beam radiotherapy for epithelial skin cancer: alternative tumor staging system. JAMA Dermatol. 2013;149(4):402-410. Analysis of 434 cases. Radiother Oncol. 2010;95(2):245-249. 21. Eroglu A, Berberoglu U, Berreroglu S. Risk factors related to locore- 49. Zagrodnik B, Kempf W, Seifert B, et al. Superficial radiotherapy for gional recurrence in squamous cell carcinoma of the skin. J Surg Oncol. patients with basal cell carcinoma: recurrence rates, histologic subtypes, 1996;61(2):124-130. and expression of p53 and Bcl-2. Cancer. 2003;98(12):2708-2714. 22. Rowe DE, Carroll RJ, Day CL, Jr. Prognostic factors for local recur- 50. Mendenhall WM, Amdur RJ, Hinerman RW, et al. Radiotherapy for rence, metastasis, and survival rates in squamous cell carcinoma of the cutaneous squamous and basal cell carcinomas of the head and neck. skin, ear, and lip. Implications for treatment modality selection. J Am Acad Laryngoscope. 2009;119(10):1994-1999. Dermatol. 1992;26(6):976-990. 51. National Comprehensive Cancer Network. Basal Cell Skin Cancer. 23. Goh RY, Bova R, Fogarty GB. Cutaneous squamous cell carcinoma v2.2016. https://www.nccn.org/professionals/physician_gls/pdf/nmsc.pdf. metastatic to parotid - analysis of prognostic factors and treatment outcome. Accessed June 1, 2016.

226 Cancer Control July 2016, Vol. 23, No. 3 52. National Comprehensive Cancer Network. Squamous Cell Skin imab for squamous-cell carcinoma of the head and neck. N Engl J Med. Cancer. v1.2016. https://www.nccn.org/professionals/physician_gls/pdf 2006;354(6):567-578. /squamous.pdf. Accessed June 1, 2016. 79. Wollina U. Cetuximab in non-melanoma skin cancer. Expert Opin 53. Ducassou A, David I, Filleron T, et al. Retrospective analysis of lo- Biol Ther. 2012;12(7):949-956. cal control and cosmetic outcome of 147 periorificial carcinomas of the 80. Wollina U. Update of cetuximab for non-melanoma skin cancer. Expert face treated with low-dose rate interstitial brachytherapy. Int J Radiat On- Opin Biol Ther. 2014;14(2):271-276. col Biol Phys. 2011;81(3):726-731. 81. Preneau S, Rio E, Brocard A, et al. Efficacy of cetuximab in the treat- 54. Lim JT. Irradiation of the pinna with superficial kilovoltage radio- ment of squamous cell carcinoma. J Dermatolog Treat. 2014;25(5):424-427. therapy. Clin Oncol (R Coll Radiol). 1992;4(4):236-239. 82. Heath CH, Deep NL, Nabell L, et al. Phase 1 study of erlotinib plus 55. Veness MJ, Morgan GJ, Palme CE, et al. Surgery and adjuvant radiation therapy in patients with advanced cutaneous squamous cell radiotherapy in patients with cutaneous head and neck squamous cell carcinoma. Int J Radiat Oncol Biol Phys. 2013;85(5):1275-1281. carcinoma metastatic to lymph nodes: combined treatment should be considered best practice. Laryngoscope. 2005;115(5):870-875. 56. Audet N, Palme CE, Gullane PJ, et al. Cutaneous metastatic squamous cell carcinoma to the parotid gland: analysis and outcome. Head Neck. 2004;26(8):727-732. 57. delCharco JO, Mendenhall WM, Parsons JT, et al. Carcinoma of the skin metastatic to the parotid area lymph nodes. Head Neck. 1998;20(5):369-373. 58. Wang JT, Palme CE, Morgan GJ, et al. Predictors of outcome in patients with metastatic cutaneous head and neck squamous cell carcinoma involving cervical lymph nodes: Improved survival with the addition of adjuvant radiotherapy. Head Neck. 2012;34(11):1524-1528. 59. Bernier J, Domenge C, Ozsahin M, et al. Postoperative irradiation with or without concomitant chemotherapy for locally advanced head and neck cancer. N Engl J Med. 2004;350(19):1945-1952. 60. Cooper JS, Pajak TF, Forastiere AA, et al. Postoperative concurrent radiotherapy and chemotherapy for high-risk squamous-cell carcinoma of the head and neck. N Engl J Med. 2004;350(19):1937-1944. 61. Bernier J, Cooper JS, Pajak TF, et al. Defining risk levels in locally advanced head and neck cancers: a comparative analysis of concurrent postoperative radiation plus chemotherapy trials of the EORTC (#22931) and RTOG (#9501). Head Neck. 2005;27(10):843-850. 62. Tanvetyanon T, Padhya T, McCaffrey J, et al. Postoperative concurrent chemotherapy and radiotherapy for high-risk cutaneous squamous cell carcinoma of the head and neck. Head Neck. 2015;37(6):840-845. 63. Nutting CM, Morden JP, Harrington KJ, et al. Parotid-sparing intensity modulated versus conventional radiotherapy in head and neck cancer (PARSPORT): a phase 3 multicentre randomised controlled trial. Lancet Oncol. 2011;12(2):127-136. 64. Kowalik L, Lyczek J, Sawicki M, et al. Individual applicator for brachytherapy for various sites of superficial malignant lesions. J Contemp Brachytherapy. 2013;5(1):45-49. 65. Guix B, Finestres F, Tello J, et al. Treatment of skin carcinomas of the face by high-dose-rate brachytherapy and custom-made surface molds. Int J Radiat Oncol Biol Phys. 2000;47(1):95-102. 66. Ganti AK, Kessinger A. Systemic therapy for disseminated basal cell carcinoma: an uncommon manifestation of a common cancer. Cancer Treat Rev. 2011;37(6):440-443. 67. Carneiro BA, Watkin WG, Mehta UK, et al. Metastatic basal cell carcinoma: complete response to chemotherapy and associated pure red cell aplasia. Cancer Invest. 2006;24(4):396-400. 68. Jefford M, Kiffer JD, Somers G, et al. Metastatic basal cell carcinoma: rapid symptomatic response to cisplatin and paclitaxel. ANZ J Surg. 2004;74(8):704-705. 69. Hutchin ME, Kariapper MS, Grachtchouk M, et al. Sustained Hedge- hog signaling is required for basal cell carcinoma proliferation and survival: conditional skin tumorigenesis recapitulates the hair growth cycle. Genes Dev. 2005;19(2):214-223. 70. Taipale J, Beachy PA. The Hedgehog and Wnt signalling pathways in cancer. Nature. 2001;411(6835):349-354. 71. Raleigh DR, Algazi A, Arron ST, et al. Induction Hedgehog pathway inhibition followed by combined-modality radiotherapy for basal cell carcinoma. Br J Dermatol. 2015;173(2):544-546. 72. Sekulic A, Migden MR, Lewis K, et al. Pivotal ERIVANCE basal cell carcinoma (BCC) study: 12-month update of efficacy and safety of vismodegib in advanced BCC. J Am Acad Dermatol. 2015;72(6):1021-1026.e8. 73. Basset-Seguin N, Hauschild A, Grob JJ, et al. Vismodegib in patients with advanced basal cell carcinoma (STEVIE): a pre-planned interim analysis of an international, open-label trial. Lancet Oncol. 2015;16(6):729-736. 74. Migden MR, Guminski A, Gutzmer R, et al. Treatment with two different doses of sonidegib in patients with locally advanced or metastatic basal cell carcinoma (BOLT): a multicentre, randomised, double-blind phase 2 trial. Lancet Oncol. 2015;16(6):716-728. 75. Pollom EL, Bui TT, Chang AL, et al. Concurrent vismodegib and radiotherapy for recurrent, advanced basal cell carcinoma. JAMA Dermatol. 2015;151(9):998-1001. 76. Alter M, Satzger I, Mattern A, et al. Treatment of advanced cutaneous squamous cell carcinomas with epidermal growth factor receptor inhibitors. Dermatology. 2013;227(4):289-294. 77. Vermorken JB, Mesia R, Rivera F, et al. Platinum-based chemotherapy plus cetuximab in head and neck cancer. N Engl J Med. 2008;359(11):1116-1127. 78. Bonner JA, Harari PM, Giralt J, et al. Radiotherapy plus cetux-

July 2016, Vol. 23, No. 3 Cancer Control 227 When treatment for paragangliomas

of the head and neck is indicated,

an algorithm-based approach can help

optimize outcomes.

Artwork courtesy of Arthur Pina de Alba. www.arthurpinadealba.com.

Treatment of Head and Neck Paragangliomas Kenneth Hu, MD, and Mark S. Persky, MD

Background: Commonly occurring in the head and neck, paragangliomas are typically benign, highly vascular neoplasms embryologically originating from the extra-adrenal paraganglia of the neural crest. Frequently, these tumors are associated with the vagus, tympanic plexus nerve, the carotid artery, or jugular bulb. Their clinical presentation can vary across a wide spectrum of signs and symptoms. Methods: We reviewed and compared standard treatment approaches for paragangliomas of the head and neck. Results: In general, surgery is the first-line choice of therapy for carotid body tumors, whereas radiotherapy is the first-line option for jugular and vagal paragangliomas. Conclusions: Because of the complexity of clinical scenarios and treatment options for paragangliomas, a multidisciplinary algorithmic approach should be used for treating paragangliomas. The approach should emphasize single-modality treatment that yields excellent rates of tumor control, low rates of severe, iatrogenic morbidity, and the preservation of long-term function in this patient population.

Introduction In the era of safe embolization protocols and so- Commonly occurring in the head and neck, paragan- phisticated surgical approaches to the skull base, sur- gliomas are typically benign, highly vascular neoplasms gery has become the preferred treatment method for embryologically originating from the extra-adrenal paragangliomas.1-5 Historically, health care profession- paraganglia of the neural crest, and they are typically als relied on radiotherapy to treat debilitated or elderly associated with the vagus, tympanic plexus nerve, the patients with unresectable, extensive tumors or para- carotid artery, or jugular bulb.1-3 Clinically, patients with gangliomas, but long-term experience and advances in paragangliomas present with diverse signs and symp- the field of radiation oncology have demonstrated that toms. They may have a family history of paragangliomas first-line radiotherapy for the treatment of paraganglio- and can present with multicentric tumors regardless of mas has low rates of long-term complications and high sporadic or familial origin. 1-3 rates of tumor control.1-3 Thus, radiotherapy now plays an important role in treatment algorithms.1-3 In addition, individualized treatment strategies may also in- From the Departments of Radiation Oncology (KH) and Otolaryn- clude observation. gology-Head and Neck Surgery (MSP), New York University Langone Medical Center, New York, New York. Address correspondence to Kenneth Hu, MD, Department of Radiation Multicentricity Oncology, NYU Langone Medical Center, 160 East 34th Street, Rarely, sporadic cases (10%) will present with a concur- New York City, NY 10016. E-mail: [email protected] rent second paraganglioma, and multicentricity may be Submitted March 21, 2016; accepted April 28, 2016. present in up to 85% of those with a genetic predisposi- No significant relationships exist between the authors and the 6-8 companies/organizations whose products or services may be refer- tion. A second carotid body tumor is by far the most enced in this article. common pattern of a synchronous secondary paragan-

228 Cancer Control July 2016, Vol. 23, No. 3 glioma (20% of carotid body tumors).6-8 An additional the carotid bifurcation, the carotid body is a discrete, contralateral or ipsilateral paraganglioma, as well as bi- oval structure that directly receives its blood supply lateral carotid body tumors, poses significant and chal- from the carotid bifurcation via the glomic arteries lenging treatment problems for the clinician, because (Fig 1).3 The afferent reflex is mediated by a glosso- patients undergoing resection of bilateral carotid body pharyngeal nerve branch called the nerve of Hering. tumors may experience baroreceptor function loss and Expanding tumors that originate from here can im- deficits in the cranial nerves resulting in labile hyper- pair sympathetic nerve chain function as well as cra- tension.1,2,9 These tumors may be metachronous, indi- nial nerves X and XII.1-3 cating that surveillance may be appropriate for select Distributed within the temporal bone in close as- paraganglioma cases. Routine follow-up imaging with sociation with the Jacobsen nerve, which is the tym- magnetic resonance imaging (MRI) or fluorodeoxy- panic branch of the glossopharyngeal nerve (see Fig glucose/fluorodopa or indium pentetreotide positron 1), are the jugulotympanic paraganglia.3 Typically, emission tomography (PET) is warranted in multicen- temporal bone paraganglia are located in the jugu- tric, metachronous tumors.1-3 lar fossa, and symptoms may involve early functional impairment of cranial nerves IX, X, XI within the Malignancy jugular foramen, and XII as it exits the hypoglossal Paragangliomas are rarely malignant, representing a canal.1-3 small subset of extra-adrenal paragangliomas with a Vagal paraganglia are distinctly separate from propensity for regional lymph-node and distant meta- jugulotympanic paraganglia because they do not form static disease, primarily to bone, the lungs, and liver.1-3 discrete bodies; in addition, they may be interspersed A higher rate of malignancy has been observed in within the vagal nerve fibers in the pars nervosa of the sporadic paragangliomas compared with the familial jugular foramen (which transmits lower cranial nerves type, with the exception of the familial syndrome IX, X, and XI) or located within the vagus nerve be- associated with SDHB mutation (rate of malignancy, neath the perineurium.1-3 The superior vagal ganglion 30%–70%).10,11 The rate of malignancy depends on is visible at the level of the jugular foramen. The ori- the site of origin: Although rare, orbital and laryngeal paragangliomas have a 25% rate of malignancy, which is the highest rate of any paraganglioma.1-3 By comparison, the rate of malignancy for vagal paragangliomas is between 16% and 19%, between 5% and 6% for jugulotympanic paragangliomas, and between 3% and 4% for carotid body tumors.10 No histological criteria exist to diagnose malignancy in primary tumors. The health care professional can make the diagnosis only if malignancy is confirmed by the presence of a tumor in the lymph nodes or the disease has metastatically spread to distant sites.1-3 It is worth emphasizing that the implications for treatment are significant when malignant disease is present.11-13 Locoregional control is best achieved with primary resection followed by adjuvant radiotherapy.1-3 The disease is known to recur for up to 20 years.1-3 The 5-year survival rates are 50% to 80% for those with nodal disease and up to 11% for those with distant spread.14

Anatomy and Physiology Diffusely distributed throughout the upper part of the body, paraganglia facilitate the chemoreceptive reflexes of the cardiovascular system. They contain chief cells with the capacity to secrete neuropeptides Fig 1. — Schematic representation of common sites of head and neck (eg, norepinephrine), which, in turn, influence vascu- paragangliomas and their relationship to the lower cranial nerves and major lar reflexes.1-3 vessels. The carotid body, which is a type of paragan- From Persky MS, Hu KS. Paragangliomas of the head and neck. In: Harrison LB, Sessions RB, Hong WK, eds. Head and Neck Cancer: A Multidisciplinary glion, can sense changes in pH, arterial oxygen pres- Approach. 3rd ed. Lippincott Williams and Wilkins, Philadelphia, PA; 2009. sure, and level of carbon dioxide.1-3 Located within Reprinted with permission by Wolters Kluwer.

July 2016, Vol. 23, No. 3 Cancer Control 229 gin of most vagal paragangliomas is the nodose vagal health care professional can obtain computed tomog- ganglion, which is located approximately 1 to 2 cm raphy (CT) with contrast for delineating paraganglio- below the jugular foramen (see Fig 1).3 Both the su- mas because the contrast enhances these highly vas- perior and nodose vagal ganglia are proximal to the cular tumors.1-3 In addition, CT is frequently diagnostic pars venosa of the jugular foramen, cranial nerves IX given the characteristic patterns tumors displace to to XII, and the ascending portion of the petrous inter- the internal and external carotid arteries.3 CT can be nal carotid artery.15 Therefore, vagal paragangliomas utilized to define the erosion and any possible skull- have distinct therapeutic sequela based on their close base involvement, including in the temporal bone. A anatomical association with the superior portion well-circumscribed mass occupying the carotid bifur- of the vagal nerve and other adjacent neurovascular cation that has splayed the external and internal ca- structures.1-3 rotid arteries is characteristic of carotid body tumors, Extra-adrenal paraganglia lack methyltransferase, which posterolaterally displace the internal carotid a requirement for converting norepinephrine to epi- artery. Typically, when vagal paragangliomas are pres- nephrine; the metabolic breakdown of catecholamines ent, the internal carotid artery is anteriorly/medially to metanephrine (from epinephrine) and normeta- displaced, occupying the superior parapharyngeal nephrine (from norepinephrine) as well as vanillyl- space (with or without skull-base involvement).1-3 In mandelic acid can be detected in urine. Thus, appro- the early disease stage, jugulotympanic paragan- priate urine and serum analysis can be used to detect gliomas can be distinguished — in particular when actively secreting paragangliomas.1-3 a paraganglioma is located in the tympanic cavity alone — and bone destruction patterns typical for this Workup type of paraganglioma are observed.1-3 History and Physical Examination MRI with gadolinium contrast is a complementary The health care professional should obtain a thor- imaging modality that the health care professional can ough history and perform a physical examination for also employ for evaluating head and neck paragangli- any patient presenting with a paraganglioma, eval- omas.1-3 Gadolinium should be used, if available, be- uating for cranial nerve dysfunction, possible signs cause paragangliomas show intense signal enhance- and symptoms related to catecholamine secretion ment in this medium.19 Similar to CT, use of contrast for as well as evidence of malignant transformation.1-3 MRI aids the health care professional in the delineation Typically, carotid body tumors present as an enlarg- of the tumor, and it also helps to survey and detect syn- ing asymptomatic mass at the level of the carotid bi- chronous paragangliomas of the head and neck as well furcation. Dysfunction of the cranial nerve may be as confirm the diagnosis. On T2-weighted MRI, the indicative of the presence of large tumors extending characteristic “salt and pepper” appearance observed into the jugular foramen.1-3 Patients with jugulotym- is essentially pathognomonic and is related to the panic tumors may present with pulsatile tinnitus high-flow vascular voids within the vascular tumor. or conductive hearing loss. Otoscopic examination Findings on magnetic resonance angiography may reveal the Brown sign involving a red-blue le- that demonstrate the patency of the circle of Willis sion in the middle ear that blanches with positive provide the surgeon with critical information about pressure.1-3 For patients with more advanced disease, intracranial circulation. Angiography can be used lower cranial nerve deficits may become apparent. to detail the features of tumor flow dynamics (the Frequently (< 50% of cases), vagal paragangliomas venous drainage and blood supply of the tumor ve- present with multiple cranial neuropathies involv- nous drainage) as well as the vascular anatomy of ing the hypoglossal, vagus, and spinal accessory the intracranial space and head and neck (Fig 2).3 If nerves.3,13,16,17 carotid artery sacrifice is anticipated, then cerebral Attention should be given to a possible familial angiography with ipsilateral internal carotid balloon history of paragangliomas, von Hippel–Lindau syn- occlusion can be obtained to define the intracranial drome, and type 2 multiple endocrine neoplasia.1-3 cerebral circulation. However, because techniques in Sweating, hypertension, tachycardia, and nervousness high-resolution MRI and CT have evolved, the role of may be symptoms of secreting tumors. Laboratory magnetic resonance angiography is limited in diag- studies, including 24-hour urinalysis and serum cat- nosis, although it does remain important for use in echolamine screening (norepinephrine, epinephrine, preoperative planning for tumor resection. Obtaining and metanephrine), should be ordered for all patients superselective angiography and performing emboli- with suspected paraganglioma.1-3 zation of the arterial supply can decrease the risk for intraoperative blood loss. Diagnostic Imaging Radionuclide imaging can be used to target the In general, fine needle biopsy is not indicated because biochemical pathways of catecholamine synthesis. radiographic studies are pathognomonic.1-3,16,18 The Such imaging techniques include fluorodeoxyglucose

230 Cancer Control July 2016, Vol. 23, No. 3 raphy because it can be used for arterial mapping to identify tumor blood supply, tumor flow dynamics, and — perhaps the most important — the displacement of major vessels. Angiography is particularly useful for patients whose tumors are large and are sup- plied by the external and internal carotid arteries and anastomoses are present between the internal and external carotid systems.22 The health care professional should evaluate the internal carotid artery for areas of irregularity or constriction as well as the structural in- tegrity of the vessels, because either may indicate that removal could be necessary. Equally important is the venous phase of angiography because it can help the health care professional identify draining vessels, and evaluate the amount of invasion and occlusion of the lumen of the jugular bulb, jugular vein, and sigmoid sinus — which are characteristic findings of jugulotympanic paragangliomas. If the surgeon anticipates that disruption or removal of the internal carotid artery might be necessary during surgery, then he or she must evaluate for adequate contralateral cerebral circulation; several methods can be used for this purpose. Although it is not available in all settings, xenon CT in conjunction with ipsilateral balloon occlusion can be useful to measure cerebral flow.23,24 While monitoring the clinical neurological status of the patient, temporary balloon occlusion of the internal carotid artery if a patent Fig 2. — Carotid body tumor splaying the carotid bifurcation. circle of Willis is present may be easier to perform. From Persky MS, Hu KS. Paragangliomas of the head and neck. In: Harrison LB, Sessions RB, Hong WK, eds. Head and Neck Cancer: A Multidisciplinary In addition, this method is reliable: Approximately Approach. 3rd ed. Lippincott Williams and Wilkins, Philadelphia, PA; 2009. 93% of patients can tolerate removal of their ipsilat- Reprinted with permission by Wolters Kluwer. eral internal carotid artery based on findings from the angiographic evaluation.9 It is worth noting that con- with or without metaiodobenzylguanidine, dihydroxy- ditions in the operating room and the angiographic phenylalanine, fluorodopamine, fluorodopa, or indium suite may differ in terms of the cerebral delivery of octreotide for use with PET.20,21 oxygen.9 A patient’s tolerance for the temporary occlusion of the internal carotid does not preclude the Surgery possibility that a delayed cerebrovascular ischemic Prior to surgery, the health care professional should event may occur following surgery. Particularly typically perform an angiographic evaluation. To eval- among patients with jugulotympanic paraganglio- uate for use of anesthesia, the presence of catechol- mas, venous outflow of the contralateral and ipsilat- amine-secreting tumors should be considered because eral transverse sinus and jugular systems should be they require an alpha- and beta-adrenergic blockade. evaluated. Because anatomical variations exist, an ab- Continuous arterial pressure monitoring is ideal, and sent or contralateral hypoplastic jugular system may preparing for possible blood transfusions should be be a contraindication for surgery because of an in- expected. creased risk of venous stroke after surgery.25-27 Based on our experience with paragangliomas, any proposed treatment must be individualized to each Embolization patient. Issues to be considered include patient age, Embolization is important adjunct therapy, and this is medical comorbidities, location and size of the tumor, especially true for surgery to treat large paraganglio- possible presence of synchronous tumors, and a his- mas. Experienced interventional radiology teams are tory of progressive neurological dysfunction. necessary to avoid surgical complications. Emboliza- tion carries the risk of the potential migration of em- Angiographic Evaluation bolization particles into the cerebral circulation, thus An important preoperative adjunct for the surgical ap- resulting in stroke, and it is known to occur in the pres- proach to paragangliomas is superselective angiog- ence of anastomoses between the internal and exter-

July 2016, Vol. 23, No. 3 Cancer Control 231 nal carotid systems or flow reversal from the arterial involved and their dissection may cause dysfunc- blood supply of the tumor. Advantages to embolization of the hypoglossal, vagus, and glossopharyngeal tion include tumor shrinkage, decreased blood flow, nerves. Ligation of the external carotid artery has and additional surgical benefits. A decrease in the been previously reported as a way to control blood rate of intraoperative bleeding after embolization re- flow to the tumor, but this maneuver should be avoid- sults in fewer transfusions, making for more effective ed because it does not affect blood flow to the tumor tumor dissection with more well-defined tissue planes and collateral circulation may be profuse.1-3 and more effective identification and preservation of normal anatomical structures, including the cranial Jugulotympanic Paragangliomas nerves. Larger paragangliomas usually demonstrate The surgeon can perform tympanotomy to approach multiple arterial feeding vessels that should be individ- small-sized tympanic paragangliomas through the ually addressed through superselective angiography. external auditory canal. Embolization of these small With progressive embolization of these vessels, addi- tumors is not required. Larger-sized tympanic para- tional compartments of the tumor are devascularized gangliomas that are confined to the mastoid, middle until a tumor “blush” is absent. Surgery should be per- ear, or both places without breaching the bone across formed within 48 hours of embolization to avoid collat- the jugular bulb or the jugulo carotid spine can be eral or intracranial arterial supply, and steroids should exposed through a combined postauricular/endaural be administered to reduce the inflammatory response approach. following embolization.28,29 The surgeon is required to combine the temporal and cervical approaches when the jugular bulb is in- Observation volved. Because intraluminal vascular invasion will be Judicious observation may be appropriate for select present, the surgeon is required to pack or ligate the patients prior to performing surgery or providing ra- internal jugular ligation located inferior and the sig- diotherapy. Small- or moderate-sized, unilateral carotid moid sinus located superior to tumor involvement. The body tumors can be easily and safely resected; howev- surgeon must also proximally and distally trace and er, enlarging tumors or progressive neurological dys- identify cranial nerves IX, X, and XI. More advanced function warrants intervention. tumors with the potential to involve various branches of the petrous carotid artery that may also extend into Carotid Body Tumors the intracranial area might require a postauricular in- The surgeon can make an oblique vertical incision fratemporal fossa approach. along the anterior border of the sternocleidomastoid The modes of spread for jugular paragangliomas muscle to visualize large-sized tumors; by contrast, a are shown in Fig 3.3 transverse neck incision across the sternocleidomastoid muscle should be made to approach small-sized Vagal Paragangliomas carotid body tumors. Dissection distal and proximal Located below the jugular foramen (~ 2 cm), vagal to the tumor helps the surgeon identify the common, paragangliomas typically begin in the inferior (no- internal, and external carotid arteries, and vessel loop dose) vagal ganglion. They have also been known to control must be applied to anticipate possible injury to originate in the middle and superior ganglia, situated the carotid. Precautions related to carotid bypass must within the jugular foramen; this location results in be available. early skull-base invasion and, in some cases, extends The surgeon should expose the most superior to the intracranial space. Growth of tumors bidirec- and inferior extent of the internal carotid artery; typi- tionally along the vagus nerve will also inferiorly and cally, the vessel is posterolaterally displaced. Because superiorly involve the poststyloid parapharyngeal hemostasis is generally feasible with paragangliomas, space and jugular foramen, respectively. In general, tumor embolization prior to surgery is unwarrant- the internal carotid artery is medially and anteriorly ed.1-3 The surgeon should use caution when dissect- displaced.7 The surgical approach for jugulotympanic ing the tumor in the subadventitial plane. Although paragangliomas involving the skull base is similar to it should be avoided in most cases, external carotid vagal paragangliomas. artery sacrifice may be required if the artery is en- The modes of spread for vagal paragangliomas are cased and infiltrated by the tumor. The last surgical shown in Fig 4.3 step when dissecting the tumor (at the bifurcation of the common carotid artery) occurs when the artery Complications is at its most vulnerable point for damage.1-3 The sur- Vascular Injury rounding cranial nerves in carotid body tumors may The incidence rate for stroke following surgery for have marked hyperemia of the vasa nervosum of their paragangliomas has been reported to be as high as nerve sheath. In larger tumors, these nerves may be 20% and as low as 2% or less.1-3 With the evolution of

232 Cancer Control July 2016, Vol. 23, No. 3 preoperative planning, surgical techniques, and diagnostic evaluations, the need to sacrifice the internal carotid and the risk of injury are minimal. The risk of injury to the carotid artery or need for sacrifice following treatment of carotid body tumors is size specific: tumors larger than 5 cm are likely to require carotid reconstruction. Unlike jugulotympanic paragangliomas and carotid body tumors, vagal paragangliomas are not closely associated with the carotid artery, although the internal carotid artery may be involved in its petrous portion in advanced disease. Rarely, injury may occur, even with adequate surgical exposure and mi- crosurgical technique. If the patient is at high risk for vessel injury within the petrous carotid portion and balloon occlusion testing has been safely and satisfac- torily performed, then the surgeon may consider permanent preoperative occlusion of the carotid distal to the tumor.1-3

Baroreflex Failure Following the resection of bilateral carotid body tumors, the baroreceptor reflex is lost and bilateral de- Fig 3. — Modes of spread for jugular paragangliomas. nervation of the carotid sinus is unavoidable, resulting From Persky MS, Hu KS. Paragangliomas of the head and neck. In: Harrison in postoperative labile refractory hypertension, tachy- LB, Sessions RB, Hong WK, eds. Head and Neck Cancer: A Multidisciplinary cardia, diaphoresis, and headache.30 The long-term Approach. 3rd ed. Lippincott Williams and Wilkins, Philadelphia, PA; 2009. 31 Reprinted with permission by Wolters Kluwer. treatment of choice is clonidine. Algorithms have been created for the management of bilateral carotid body tumors to avoid long-term postoperative hyper- tensive issues.1-3

Cranial Nerve Injury The surgical risk to the lower cranial nerves for the treatment of paragangliomas is site specific and directly related to tumor size. Vagal, jugulotympanic, and carotid body paragangliomas represent, in decreasing order, risk of injury to the cranial nerves. The size of the tumor is especially important in vagal and jugulotympanic paragangliomas. Tumors presenting with extensive skull-base involvement are likely to have extensive lower cranial nerve involvement and often preoperatively present with multiple cranial nerve deficits. Facial nerve involvement in conjunction with preoperative paralysis is a sign of such extensive involvement.16,30,32-34 Following surgery, deficits in the cranial nerve may involve impaired aspiration, deglutition, tongue motion, and phonation. Patients with more than 1 cranial nerve deficit do not respond well to treatment because these deficits have additive effects. Those who are older may experience difficulty in recovery because vocal cord medialization may be necessary if a risk of aspira- Fig 4. — Modes of spread for vagal paragangliomas. tion is present. In such patients, tracheostomy or gas- From Persky MS, Hu KS. Paragangliomas of the head and neck. In: Harrison trostomy may be necessary — this is especially true for LB, Sessions RB, Hong WK, eds. Head and Neck Cancer: A Multidisciplinary Approach. 3rd ed. Lippincott Williams and Wilkins, Philadelphia, PA; 2009. patients with injuries to the high vagus nerve. Reprinted with permission by Wolters Kluwer. Injury to the accessory nerve results in func-

July 2016, Vol. 23, No. 3 Cancer Control 233 tional loss of the sternocleidomastoid and trape- Histopathological Changes zius muscles. Patients with such injury should be Multiple reports have described the histopathological referred to physical therapy to help avoid shoulder impact of radiation on paragangliomas.35-37 Gardner pain secondary to shoulder “drop” and limited range et al36 studied 6 irradiated tumor specimens resected of motion. If the nerve is injured below the jugular 4 to 6 weeks following radiotherapy and found evi- foramen, then primary repair or nerve grafting may dence of vascular endarteritis with mural thrombi as be performed.17 well as necrotic infarct and pyknotic cellular death. Hypoglossal nerve injury results in paralysis of Fibrosis around nests of the pathognomonic chief the ipsilateral tongue. Long-term hypoglossal nerve cells has been reported at 6 months post-treatment paralysis results in hemiatrophy of the tongue with- with diminished vascularity.35,38-40 Chief cells may in a few months. If this occurs in combination with show evidence of senescence such as nuclear pleo- other lower cranial nerve injuries, then swallowing morphism, irregular nuclear outlines, and chroma- therapy will be required to prevent aspiration. Swal- tin clumping.36,37,40 Thus, radiation appears to cre- lowing therapy is usually focused on educating the ate sclerosing endarteritis with subsequent fibrosis, patient to direct the bolus to the functioning side.17 which in turn prevents tumor growth and involution More significant, persistent swallowing and aspira- while also causing a loss of reproductive capacity of tion issues may require feeding via tracheostomy the chief cells.22,35,41 and gastrostomy tubes. External Beam Radiotherapy Radiotherapy Ever since the first major review was published more Traditionally, radiotherapy was offered to those with than 50 years ago of 106 cases of paragangliomas that postoperative recurrence or for those patients who demonstrated similar rates of efficacy between radio- were not surgical candidates due to technical unresect- therapy and surgery, numerous retrospective stud- ability issues, patient refusal, age, or illness. Unlike ies have confirmed the effectiveness of radiation for surgery, concern exists with radiotherapy because le- managing head and neck paragangliomas.35,42-68 Re- sions may remain dormant as they rarely regress com- flecting a wide variety of delivery techniques, beam pletely after this treatment modality. Yet, evidence energies, and dosing schedules, cumulative rates of supports the long-term efficacy of moderate doses local control average 90% (range, 65%–100%) and are of radiation to prevent tumor progression while also based on more than 1,000 cases (median follow-up time, preserving cranial nerve function.1-3 Thus, radiother- 10 years).35,42-68 apy has become a first-line treatment for these large- Primary radiotherapy may be delivered with con- ly benign tumors, and it is commonly considered for ventional fractionation, the results of which have the skull-base tumors of a jugulotympanic origin as well as largest and longest experience. Typically, a dose of vagal tumors.1-3 Given the demonstrable and favorable 45 Gy in 5 weeks is given with conventional frac- long-term outcomes, successful treatment is typically tionation (Fig 5).3 The effectiveness of radiotherapy is defined as lack of tumor progression on serial radio- not affected by site of origin, whether it be the jugulo- graphic follow-up and includes stability of tumor size temporal, carotid body, or vagal space.43,48,62,69 Gilbo or partial regression.1-3 et al63 reported on a 45-year experience with conven- Primary radiotherapy may be delivered with con- tional fractionation and prescribed a dose of 45 Gy. ventional external beam radiotherapy (EBRT), stereotac- A total of 131 patients were enrolled and 156 paragan- tic radiosurgery, or hypofractionated stereotactic radio- gliomas were studied. The 5- and 10-year rates of local therapy — all of these approaches have excellent rates of control were 99% and 96%, respectively, at a median local control and outcomes.1-3 Typically, doses of 45 Gy follow-up of 8.7 years.63 Five tumors recurred between in 5 weeks are given with conventional EBRT, 12 to approximately 1 and 8 years following treatment.63 15 Gy with stereotactic radiosurgery, and 21 Gy for Compared With Surgery: The effectiveness of 3 fractions or 25 Gy for 5 fractions with hypofractionat- surgery compared with EBRT is difficult to ascertain ed stereotactic radiotherapy.1-3 Patients whose intracra- because of the retrospective nature of data from au- nial tumors measure less than 3 cm are the best candi- thors at single institutions, who report small patient dates for stereotactic approaches, whereas those whose numbers confounded with selection bias and other tumors are larger or have a component of extracranial conflicting outcomes.66,70-73 To extract meaningful com- spread are best suited for EBRT.1-3 The ablative nature of parative outcomes based on a common staging system, stereotactic radiosurgery can cause a small increase or an analysis of 5 studies was performed to report on exacerbation in the rate of cranial neuropathy, and, thus, the outcomes of temporal bone tumors using McCabe/ a more fractionated approach with stereotactic radiother- Fletcher staging.43,47,51,60,68 That analysis demonstrated apy or conventional EBRT may be considered in those that study patients treated with radiotherapy, surgery, whose baseline cranial nerve function is excellent.1-3 or a combination of both had average rates of local

234 Cancer Control July 2016, Vol. 23, No. 3 Relative Dose, % 0 22.222 44.444 66.666 88.888 100

20 Ratio of Total Structure Volume, %

0 0 100 200 300 400 Dose, cGy

Fig 5. — Jugular paraganglioma treated with external beam radiotherapy. A man aged 27 years presented with diplopia and tinnitus and was found to have paraganglioma of the jugular area that extended extracranially and to the cavernous sinus. He was treated with definitive intensity-modulated radiotherapy (dose of 45 Gy in 25 fractions). Both symptoms resolved by the end of radiotherapy. control of 93%, 78%, and 85%, respectively, for a medi- ment, 3% developed permanent stroke, and 1.3% died an follow-up period of 11 to 16 years.43,47,51,60,68 Despite because of postoperative complications.73 The rates a larger number of advanced tumors in the radiother- of iatrogenic cranial neuropathy and vascular compli- apy group, the rate of local control was similar or bet- cations were 2.3% for Shamblin class 1/2 tumors and ter than those undergoing surgery alone or receiving 35.7% for Shamblin class 3 tumors (P < .001).74 With combination treatment.43,47,51,60,68 Others have shown use of EBRT, a potential increased risk of ischemic debulking surgery does not improve outcomes when stroke of approximately 12% has been observed with patients are treated with radiotherapy.60 long-term follow-up of 15 years.75,76 In a systematic literature review, Suarez et al10 re- Suarez et al73 also studied patients with jugu- ported on the role of surgery and EBRT in the treat- lar or vagal paragangliomas treated with surgery ment of carotid body paragangliomas. The mean fol- (n = 1,310), EBRT (n = 461), or stereotactic radiosur- low-up times were 80.6 months for surgery (n = 2,175) gery (n = 261). The median follow-up times for sur- and 99.9 months for those receiving EBRT (n = 127).10 gery, EBRT, and stereotactic radiosurgery were No difference was seen in local control in the surgery 66 months, 113 months, and 41 months, respective- and EBRT arms (93.8% vs 94.5%, respectively).10 ly.73 Among patients with jugular paragangliomas, Reduction in tumor size was reported in 25.2% of better rates of local control were observed with ra- study patients treated with EBRT. All received conven- diotherapy compared with surgery (91.5% vs 78.1%; tional doses of radiotherapy (40–65 Gy); 44% received P = .002), and fewer major complications were seen doses between 40 and 50 Gy.10 Iatrogenic cranial with radiotherapy compared with surgery (11% vs neuropathy — primarily occurring in cranial nerves 26%; P = .02) — in particular, lower iatrogenic cranial X and XII — occurred in 22.2% of patients treated with neuropathy (0.08 vs 1.0/per patient; P < .001).73 Peri- surgery vs 0% in those treating with EBRT (P = .004).10 operative complications of major importance were, Iatrogenic Horner syndrome occurred in 2.5% of pain order of decreasing incidence, cerebrospinal fluid tients treated with surgery.10 The carotid artery was leak, aspiration/pneumonia, wound infection, men- resected in 12.5% because of injury or tumor encase- ingitis, and stroke.73 A rate of perioperative mortal-

July 2016, Vol. 23, No. 3 Cancer Control 235 ity was reported to be 1.6%.73 Severe complications mas in 276 patients to evaluate the proportion of pa- from radiotherapy included, in order of decreasing tients whose tumors had regressed after stereotactic incidence, deafness, osteonecrosis, death, and brain radiosurgery or conventional EBRT. All studies had to necrosis.73 No significant difference was observed in have a minimum of 12 months of follow-up with ad- rates of local control between EBRT and stereotactic equate radiologic evaluation.79 The percentages of radiosurgery.73 patients demonstrating some regression after stereotactic techniques and treated with definitive intent, Impact on Neurological Function combined modality, and salvage treatment were 21%, In the majority of patients with jugulotympanic tumors 33%, and 52%, respectively; for those receiving con- who present with tinnitus, EBRT can be used to reduce ventional EBRT, the corresponding outcomes were or resolve it.48,55 Cummings et al55 demonstrated com- 4%, 0%, and 64%.79 No differences in local control were plete resolution of tinnitus in 79% and stable or par- noted between the 2 treatment techniques nor in those tial relief in 21% of the cases they studied. With regard with tumor regression vs those without.79 to sensorineural hearing loss, they found that 5% of patients reported a return to normal hearing, 30% re- Fractionated Radiotherapy ported some improvement, and 62% noted no change Rates of morbidity following radiotherapy vary accord- after radiotherapy.55 Improvement of other cranial ing to the radiation technique and treatment site.1-3 neuropathies after radiotherapy has been reported in Common toxicities related to radiotherapy include mu- approximately one-third of patients.44,47,48,50,51,53,55,62,69 cositis, fatigue, otitis, dermatitis, nausea, xerostomia, Complete restoration of cranial nerve function epilation, skin dryness, fibrosis, and cerumen build is less common, occurring in about 10% (range, up.1-3 Severe complications after radiotherapy have 8%–20%).44,47,48,50,51,53,55,62,69 The probability that cranial been reported in 30 series and occur in approximately nerve function will improve following radiotherapy is 6% of patients; the rate of treatment-related mortality most likely inversely related to the duration of cranial has been observed to be 0.6%.35,42-68 Severe morbidity neuropathy.48,73 Suarez et al73 reported improvement of primarily consists of osteoradionecrosis, chronic otitis, cranial nerve function in 8.8% of patients after stereo- brain necrosis, radiation-related cranial neuropathy, tactic radiosurgery compared with 4.1% in those treat- radiation-induced sarcoma, external auditory canal ste- ed with EBRT. Hearing loss occurred in 6.5% patients nosis, and trismus.1 treated with stereotactic radiosurgery.73 Most of these severe complications are related to Radiation-induced cranial neuropathy follow- radiotherapy that exceed the current recommended ing EBRT is rare and has been associated with doses dose, use outdated treatment techniques, or are due to above what are recommended.48,53,62 Four such cases toxicity from reirradiation.1 For example, of the 4 re- have been reported: 2 cases of cranial nerve VII palsy, ported cranial neuropathies related to radiotherapy, 1 case of cranial nerve VIII dysfunction occurring after 3 occurred after receiving 64 to 66 Gy of radiation.48 a high dose of radiotherapy (64–66 Gy), and 1 case of Cole et al45 reported that all of the severe complica- cranial nerve VI palsy that the authors stated had an tions seen in their series occurred in patients treated “unclear etiology.”48,53,62 with orthovoltage; none occurred in those treated with megavoltage. It is worth noting that the reported inci- Impact on Radiographic Tumor Regression dence of radiation-induced secondary malignancies Results of the radiographic follow-up of patients af- was low (0.4%).45 These included 2 fibrosarcomas oc- ter EBRT demonstrate stability in tumor size or mod- curring 15 and 25 years after treatment and 1 osteosar- est tumor regression.51,60,77 Mukherji et al78 reported on coma occurring 5 years post-treatment.45 In 2015, Gilbo 17 patients with 18 paragangliomas treated with defini- et al63 reported on 131 patients (156 benign paragangli- tive radiotherapy who underwent pre-treatment and omas) in a 45-year report. The patients were treated for post-treatment imaging using CT or MRI. A total of 61% paragangliomas of the jugular bulb, vagal area, tempo- showed a decrease in tumor size, with an average re- ral bone, and carotid body. At nearly 12 years of median duction of 23% (range, 8%–45%) at a median follow-up follow-up, they observed no severe (grade 4/5) com- of 2.5 years. Postradiotherapy findings on MRI includ- plications, and none of their study patients developed ed reduction in flow voids, decreased heterogeneous iatrogenic cranial neuropathy/malignancy.63 Thus, with enhancement, and a reduced T2 signal.78 Other studies current recommended dosing guidelines and modern have demonstrated tumor regression in 57% to 73% of treatment techniques, definitive radiotherapy can be patients followed by CT.47,49 Thus, paragangliomas will well tolerated in patients with paragangliomas. show modest radiographic change or stable tumor in the majority of cases. Stereotactic Radiosurgery van Hulsteijn et al79 performed a meta-analysis of Stereotactic radiosurgery is a successful first-line 15 studies involving 283 jugulotympanic paraganglio- treatment for paragangliomas and as salvage therapy

236 Cancer Control July 2016, Vol. 23, No. 3 for treatment failure. Stereotactic radiosurgery uses 374 patients.87 The mean age for patients undergoing a highly focused, single ablative dose of radiation to surgery was 47 vs 57 years of age for those treated with a small target with a steep dose gradient to spare as stereotactic radiosurgery.87 A total of 48% of patients much surrounding normal tissue as possible. For most underwent first-line radiosurgery and 52% underwent patients, the plan for treatment and the treatment itself salvage or adjuvant radiosurgery.87 Tumor shrinkage are performed in a single session. Radiation is deliv- was reported in 36% of patients treated with radiosur- ered using non-coplanar beams, rigid immobilization, gery while 61% remained unchanged.87 With regard and MRI-based treatment planning. to neurological symptoms, 39% showed improve- Multiple series have reported on the success of this ment, 58% showed no change, and 3% had worsening approach to treat jugular paragangliomas and have symptoms; moreover, most study patients experienc- observed generally high rates of local control.77,80-84 A ing neurological improvement did so within the first wide range of median doses has also been reported be- 12 months.87 At a follow-up period of 49 and 39 months tween 15 and 32 Gy, with 15 Gy being the most com- (range, 20.0–86.4 months) for the surgical and radiosur- mon.77,80-84 Chen et al85 reported on a 15-patient series gical groups, respectively, the rates of local control in and found that 13 Gy was associated with higher rates the surgical group were 92% and 98% for the radiosur- of failure than 15 Gy (P = .08). Treatments appeared to gical group.87 The neurological complication rate after be well tolerated, with low incidence rates of transient stereotactic radiosurgery was 8.5% (6.4% transient, facial neuropathy, hearing impairment, and vertigo.85 2.1% permanent).87 Transient complications included Stereotactic radiosurgery does have limitations, which exacerbation of pre-existing cranial nerve deficits, tin- include the location of the intracranial tumor and tu- nitus, or vertigo; permanent complications were report- mor size.80 Tumors are typically no larger than 3 cm to ed as worsening facial nerve function (n = 3), worsen- achieve tight dose conformality. ing vertigo (n = 1), and progressive hearing loss to A population-based meta-analysis of 19 studies deafness (n = 1).87 No patients experienced new deficits comprising 335 glomus jugulare cases treated with ra- in the lower cranial nerves.87 Meningitis, wound infec- diosurgery was reported by Guss et al.86 For all of the tion, pneumonia, ischemia, cerebrospinal fluid leak, 19 studies, the rate of tumor control was 97%; among and aspiration were the major surgical complications 8 reports studied whose median follow-up times were reported.87 Iatrogenic cranial nerve deficits were re- longer than 36 months, the rate of control achieved ported in several surgical series and involved cranial was 95%.86 No difference in outcome was reported by nerves VII and IX to XII.87 The rate of perioperative radiosurgical technique used.86 mortality was 1.3%.87 Compared With Surgery: Gottfried et al87 performed a meta-analysis of 109 studies comprising Hypofractionated Stereotactic Radiotherapy 869 glomus jugulare cases treated either by gross total Hypofractionated stereotactic radiotherapy combines resection, subtotal resection, stereotactic radiosurgery the high precision of stereotactic planning, the function alone, or subtotal resection combined with stereotac- preservation advantages of fractionation, and the con- tic radiosurgery. The median follow-up times for gross venience of short-course treatment. Some authors have total resection, subtotal resection, subtotal resection reported excellent outcomes with low rates of severe in combination with stereotactic radiosurgery, and toxicity. For example, Wegner et al88 reported an early stereotactic radiosurgery were 88 months, 72 months, experience of 18 glomus jugulare cases treated with lin- 96 months, and 71 months, respectively.87 The rates of ear accelerator–based stereotactic radiotherapy; 8 had tumor control for gross total resection, subtotal resec- persistent tumors after prior surgery, 1 had recurrence tion, subtotal resection in combination with stereotac- after EBRT, and 2 were treated with combined extracra- tic radiosurgery, and stereotactic radiosurgery were nial stereotactic radiotherapy with intracranial Gamma 86%, 69%, 71%, and 95%, respectively.87 The study pa- Knife (Elekta, Crawley, UK) radiotherapy.88 The me- tients undergoing stereotactic radiosurgery alone had dian radiation dose was 21 Gy delivered in 3 fractions the best tumor control rates (P < .001).87 Those under- (16–25 Gy in 1–5 fractions) to the 80% isodose line.88 At going gross total resection had worse deficits in cra- a median follow-up of 22 months, the observed rate of nial nerves IX to XI compared with those assigned to local control was 100%. No cases of new or worsening stereotactic radiosurgery, although deficits in cranial pre-existing neurological deficits were observed.88 nerve XII were comparable in the 2 groups.87 Chun et al89 reported on 31 patients with jugular Gottfried et al87 also performed a comprehensive lit- paragangliomas (n = 30) or carotid body tumors (n = 1) erature review comparing stereotactic radiosurgery and treated with CyberKnife (Accuray, Sunnyvale, Califor- conventional surgery for the treatment of jugular para- nia) fractionated stereotactic radiotherapy at a dose of gangliomas in 576 patients. They reviewed 8 radiosur- 25 Gy in 5 fractions.89 Mean tumor size was reported as gery series reporting outcomes on 142 patients and 10.7 mL, which is twice the size of the typical volumes 7 conventional surgical series that detailed outcomes on used in Gamma Knife series.88,89 At a median follow-

July 2016, Vol. 23, No. 3 Cancer Control 237 up time of 24 months, the rate of local control was and more sensitive imaging techniques can detect and reported as 100%.89 Tumor volume was reduced for accurately follow the progress of these tumors.92 In an the entire study population by 37%; at 2 years, that observational study of 47 tumors followed for 5 years, rate was 49%.89 A total of 60% of patients reported the researchers found that 38% progressed at an an- improvement in tinnitus.89 Grade 1/2 adverse events nual growth of 2 mm while 42% were stable and 20% (primarily headache) were reported in 19% of the pa- decreased in size.92 Others have reported a higher rate tients studied.89 of progression (≤ 60%) and an annual average growth Lieberson et al90 reported on 36 patients (41 para- rate of 1 mm, with variable doubling times between gangliomas): 17 of whom were treated with fraction- 6 months and 21 years.93 Whether observation for all ated stereotactic radiotherapy, and 19 were treated asymptomatic patients is the best initial strategy re- with surgery. They reported that, on average, smaller mains to be determined. Moreover, additional research lesions (< 8 mL) were treated with 18 Gy in 1 fraction, is needed to identify the following: moderately sized lesions (8–16 mL) were treated with • Optimal surveillance strategies across a long time 20 Gy in 2 fractions, and larger-sized lesions (≥ 16 mL) period were treated with 22 Gy in 3 fractions (median lesion • Thresholds for active treatment with regard to size = 1.64 mL).90 One large lesion measuring 69 mL symptom onset or changes in tumor size that im- received 25 Gy in 5 fractions, and another measuring pact morbidity 42 mL received 24 Gy in 3 fractions.90 Five of the 19 pa- • Reversibility of symptoms that develop from active tients treated with single-dose radiation had worsening surveillance with subsequent treatment of their pretreatment deficit, 2 patients had transient If treatment is to be pursued, then we propose worsening of cranial nerve deficits, and 7 patients had multidisciplinary algorithms that take into account an improvement of their pretreatment deficit, including multiple factors, such as type of paraganglioma, ma- 4 treated with stereotactic radiosurgery and 3 treated lignant status, whether the tumor is catecholamine-se- with stereotactic radiotherapy.90 creting, presence of synchronous tumors, and extent of pre-treatment cranial neuropathy (Figs 6 and 7).1-3 Malignant and Catecholamine-Secreting Patient factors include age, presence of comorbidities, Paragangliomas potential for rehabilitation, lifestyle, and wishes. The The diagnosis of malignant tumors is radiographically focus of the treatment algorithms is unimodality treat- determined, because no distinguishing histological fea- ment to control the tumor, minimize iatrogenic mor- tures of transformation exist. The standard treatment for bidity, and preserve cranial nerve function as optimally malignant paragangliomas is resection of the primary tu- as possible. mor and neck dissection. 91 Data regarding the outcomes First-line surgery is preferred for patients with carot- of malignant paragangliomas after definitive radiother- id body tumors, especially those whose tumors are less apy are rare.37,43 Hinerman et al43 reported on 3 patients than 5 cm in size and also lack carotid artery encase- with malignant carotid body tumors and neck metasta- ment, because these tumors are usually excised with a ses who underwent definitive radiotherapy (64.8–70 Gy) low associated risk of carotid artery sacrifice. In cases of who were without evidence of disease at 15 months, bilateral synchronous carotid body tumors, bilateral re- 4 years, and 6 years. Catecholamine-secreting tumors section is contraindicated due to baroreflex failure syn- should be treated with resection as results from case re- drome. Rather, resection of the smaller tumor and radio- ports indicate poor functional control in patients treated therapy to the larger tumor are recommended. with radiotherapy despite lack of tumor progression.37 For patients with catecholamine-secreting or malignant tumors, first-line resection is recommended Multidisciplinary Treatment Algorithms with adjuvant radiotherapy as needed based on patho- The management of paragangliomas is challenging logical factors. given the complexity of possible presentations, the Radiotherapy is the preferred modality for the treat- availability of multiple treatment options, and the na- ment of jugular and vagal paragangliomas because ture of these tumors to indolently grow but with the this therapeutic option offers the opportunity for high potential to severely impact functioning of the head rates of local control combined with likely resolution of and neck. pre-existing tinnitus and possibly improvement in cra- Observation without treatment is an appropri- nial neuropathy. First-line surgery may be considered ate initial option in patients who are asymptomatic, for small tympanic tumors and in those with advanced particularly among older individuals with comorbidi- jugular/vagal paragangliomas in which resection would ties who are willing to undergo watchful surveillance. not worsen any pre-existing cranial nerve palsy. Some authors indicate that paragangliomas can re- First-line radiotherapy should be considered in main stable for long periods of time and, if growth is patients who are elderly, those with multiple comor- detected, then the rate of expansion is generally slow bidities, and in those whose tumors have advanced

238 Cancer Control July 2016, Vol. 23, No. 3 Presence of CBT

Jugular or Single CBT Bilateral CBT vagal tumor + CBT

Magnetic resonance RT for jugular/ Surgery for RT for Surgery for CBT anglography vagal tumor smaller-sized CBT larger-sized CBT

Salvage surgery Salvage reirradiation Surgery if regrowth or observation

Salvage RT if growth

Fig 6. — Algorithm for the treatment of CBTs. CBT = carotid body tumor, RT = radiotherapy.

Jugular or vagal

Small-volume Large, complex tumors tumors ( > 3 cm)

Conservative function- Stereotactic RT IMRT IMRT preserving surgery

No growth Growth

Radiographic Salvage surgery, RT, surveillance or reirradiation

Fig 7. — Algorithm for the treatment of vagal and jugular paragangliomas. IMRT = intensity-modulated radiotherapy, RT = radiotherapy.

July 2016, Vol. 23, No. 3 Cancer Control 239 relationships with the facial nerve. J Neurosurg. 2010;112(1):88-98. into the skull base or extended into the intracranial 19. Valavanis A, Schubiger O, Oguz M. High-resolution CT investiga- space. tion of nonchromaffin paragangliomas of the temporal bone. AJNR Am J Neuroradiol. 1983;4(3):516-519. 20. Barr JD, Lemley TJ, McCann RM. Carotid artery balloon test occlu- Conclusions sion: combined clinical evaluation and xenon-enhanced computed tomo- graphic cerebral blood flow evaluation without patient transfer or balloon Management of head and neck paragangliomas has reinflation: technical note. Neurosurgery. 1998;43(3):634-638. evolved with advances in the understanding of the 21. Chrisoulidou A, Kaltsas G, Ilias I, et al. The diagnosis and management of malignant phaeochromocytoma and paraganglioma. Endocr Relat natural history of the disease, genetic testing, and Cancer. 2007;14(3):569-585. the improvements of multiple treatment options. Be- 22. Poznanovic SA, Cass SP, Kavanagh BD. Short-term tumor control and acute toxicity after stereotactic radiosurgery for glomus jugulare tumors. cause paragangliomas are benign in the vast majority Otolaryngol Head Neck Surg. 2006;134(3):437-442. of cases, survival is not a common end point reported 23. Jackson CG, Cueva RA, Thedinger BA, et al. Cranial nerve preservation in lesions of the jugular fossa. Otolaryngol Head Neck Surg. in studies. Rather, the primary focus of study is pre- 1991;105(5):687-693. serving function and the prevention of morbidity from 24. Carlson A, Yonas H. Xenon techniques in predicting patients at risk for stroke after balloon test occlusion. Neurosurgery. 2009;64(6):E1206. progressive disease. Watchful waiting represents an 25. Young NM, Wiet RJ, Russell EJ, et al. Superselective embolization of initial management strategy that may be appropriate glomus jugulare tumors. Ann Otol Rhinol Laryngol. 1988;97(6 pt 1):613-620. 26. Piazza P, Di Lella F, Menozzi R, et al. Absence of the contralateral for select patients whose tumors are incidentally found internal carotid artery: a challenge for management of ipsilateral glomus and remain asymptomatic. For those who require treat- jugulare and glomus vagale tumors. Laryngoscope. 2007;117(8):1333-1337. 27. Forshaw MA, Higgins N, Hardy DG, et al. Rupture of an internal ment, a multidisciplinary approach that emphasizes carotid artery aneurysm in the petrous temporal bone. Br J Neurosurg. single-modality therapy offers optimal outcomes. 2000;14(5):479-482. 28. Brink I, Hoegerle S, Klisch J, et al. Imaging of pheochromocytoma Acknowledgment: We gratefully acknowledge and paraganglioma. Fam Cancer. 2005;4(1):61-68. the editorial assistance of Sherri Damlo of Damlo 29. Andrews JC, Valavanis A, Fisch U. Management of the internal carotid artery in surgery of the skull base. Laryngoscope. 1989;99(12):1224-1229. Does, LLC, for her substantive editing and extensive 30. Maturo S, Brennan J. Baroreflex failure: a rare complication of carotid collaboration with us on this manuscript. paraganglioma surgery. Laryngoscope. 2006;116(5):829-830. 31. Sanna M, De Donato G, Piazza P, et al. Revision glomus tumor surgery. Otolaryngol Clin North Am. 2006;39(4):763-782, vii. References 32. Jackson CG, McGrew BM, Forest JA, et al. Lateral skull base surgery for glomus tumors: long-term control. Otol Neurotol. 2001;22(3):377-382. 1. Hu K, Persky MS. The multidisciplinary management of para- 33. Manolidis S, Jackson CG, Von Doersten PG, et al. Lateral skull base gangliomas of the head and neck, part 2. Oncology (Williston Park). surgery: the otology group experience. Skull Base Surg. 1997;7(3):129 -137. 2003;17(8):1143-1161. 34. Glasscock ME III. The history of glomus tumors: a personal per- 2. Hu K, Persky MS. Multidisciplinary management of paragangliomas spective. Laryngoscope. 1993;103(11 pt 2 suppl 60):3-6. of the head and neck, Part 1. Oncology (Williston Park). 2003;17(7):983-993. 35. Spector GJ, Compagno J, Perez CA, et al. Glomus jugulare tumors: 3. Persky MS, Hu KS. Paragangliomas of the head and neck. In: Harri- effects of radiotherapy. Cancer. 1975;35(5):1316-1321. son LB, Hong WK, Sessions RB, ed. Head and Neck Cancer: A Multi- 36. Gardner G, Cocke EW Jr, Robertson JT, et al. Combined ap- disciplinary Approach. 3rd ed. Philadephia: Lippincott Williams & Wilkins; proach surgery for removal of glomus jugulare tumors. Laryngoscope. 2009: 655 - 687. 1977;87(5 pt 1):665-688. 4. Sanna M, Jain Y, De Donato G, et al. Management of jugular paragan- 37. Glasscock ME, Jackson CG. Glomus tumors: diagnosis and sur- gliomas: the Gruppo Otologico experience. Otol Neurotol. 2004;25(5):797-804. gery. Rev Laryngol Otol Rhinol (Bord). 1979;100(1-2):131-136. 5. Jackson CG. Neurotologic skull base surgery for glomus tumors. 38. Rosenwasser H. Glomus jugulare tumors. I. Historical background. Diagnosis for treatment planning and treatment options. Laryngoscope. Arch Otolaryngol. 1968;88(1):1-40. 1993;103(11 pt 2, suppl 60):17-22. 39. Rosenwasser H. Current management glomus jugulare tumors. 6. Grufferman S, Gillman MW, Pasternak LR, et al. Familial carotid body Ann Otol Rhinol Laryngol. 1967;76(3):603-610. tumors: case report and epidemiologic review. Cancer. 1980;46(9):2116-2122. 40. Brackmann DE, House WF, Terry R, et al. Glomus jugulare tu- 7. Netterville JL, Jackson CG, Miller FR, et al. Vagal paraganglioma: mors: effect of irradiation. Trans Am Acad Ophthalmol Otolaryngol. a review of 46 patients treated during a 20-year period. Arch Otolaryngol 1972;76(6):1423-1431. Head Neck Surg. 1998;124(10):1133-1140. 41. Vanmiert PJ. The treatment of chemodectomas by radiotherapy. 8. Netterville JL, Reilly KM, Robertson D, et al. Carotid body tumors: a Proc R Soc Med. 1964;57:946-951. review of 30 patients with 46 tumors. Laryngoscope. 1995;105(2):115-126. 42. Alford BR, Guilford FR. A comprehensive study of tumors of the 9. Persky MS, Setton A, Niimi Y, et al. Combined endovascular and glomus jugulare. Laryngoscope. 1962;72:765-805. surgical treatment of head and neck paragangliomas--a team approach. 43. Hinerman RW, Mendenhall WM, Amdur RJ, et al. Definitive radio- Head Neck. 2002;24(5):423-431. therapy in the management of chemodectomas arising in the temporal 10. Suarez C, Rodrigo JP, Mendenhall WM, et al. Carotid body para- bone, carotid body, and glomus vagale. Head Neck. 2001;23(5):363-371. gangliomas: a systematic study on management with surgery and radio- 44. Mumber MP, Greven KM. Control of advanced chemodectomas of therapy. Eur Arch Otorhinolaryngol. 2014;271(1):23-34. the head and neck with irradiation. Am J Clin Oncol. 1995;18(5):389-391. 11. Timmers HJ, Gimenez-Roqueplo AP, Mannelli M, et al. Clinical 45. Cole JM, Beiler D. Long-term results of treatment for glomus aspects of SDHx-related pheochromocytoma and paraganglioma. Endocr jugulare and glomus vagale tumors with radiotherapy. Laryngoscope. Relat Cancer. 2009;16(2):391-400. 1994;104(12):1461-1465. 12. Muller U. Pathological mechanisms and parent-of-origin effects in 46. Larner JM, Hahn SS, Spaulding CA, et al. Glomus jugulare tumors. hereditary paraganglioma/pheochromocytoma (PGL/PCC). Neurogenetics. Long-term control by radiation therapy. Cancer. 1992;69(7):1813-1817. 2011;12(3):175-181. 47. Schild SE, Foote RL, Buskirk SJ, et al. Results of radiotherapy for 13. Baysal BE. Genomic imprinting and environment in hereditary para- chemodectomas. Mayo Clin Proc. 1992;67(6):537-540. ganglioma. Am J Med Genet C Semin Med Genet. 2004;129C(1):85-90. 48. Powell S, Peters N, Harmer C. Chemodectoma of the head and 14. Pacheco-Ojeda L. Malignant carotid body tumors: report of three neck: results of treatment in 84 patients. Int J Radiat Oncol Biol Phys. cases. Ann Otol Rhinol Laryngol. 2001;110(1):36-40. 1992;22(5):919-924. 15. Plenat F, Leroux P, Floquet J, et al. Intra and juxtavagal paraganglia: 49. Verniers DA, Keus RB, Schouwenburg PF, et al. Radiation therapy, a topographical, histochemical, and ultrastructural study in the human. Anat an important mode of treatment for head and neck chemodectomas. Eur J Rec. 1988;221(3):743-753. Cancer. 1992;28A(6-7):1028-1033. 16. Jackson CG, Glasscock ME III, Nissen AJ, et al. Glomus tumor 50. Boyle JO, Shimm DS, Coulthard SW. Radiation therapy for para- surgery: the approach, results, and problems. Otolaryngol Clin North Am. gangliomas of the temporal bone. Laryngoscope. 1990;100(8):896-901. 1982;15(4):897-916. 51. Pryzant RM, Chou JL, Easley JD. Twenty year experience with ra- 17. Netterville JL, Civantos FJ. Defect reconstruction following neuroto- diation therapy for temporal bone chemodectomas. Int J Radiat Oncol Biol logic skull base surgery. Laryngoscope. 1993;103(11 pt 2 suppl 60):55-63. Phys. 1989;17(6):1303-1307. 18. Borba LA, Araujo JC, de Oliveira JG, et al. Surgical management of 52. Konefal JB, Pilepich MV, Spector GJ, et al. Radiation therapy in the glomus jugulare tumors: a proposal for approach selection based on tumor treatment of chemodectomas. Laryngoscope. 1987;97(11):1331-1335.

240 Cancer Control July 2016, Vol. 23, No. 3 53. Dawes PJ, Filippou M, Welch AR, et al. The management of glomus and conventional surgery for the treatment of glomus jugulare tumors. jugulare tumours. Clin Otolaryngol Allied Sci. 1987;12(1):15-24. Neurosurg Focus. 2004;17(2):E4. 54. Sharma PD, Johnson AP, Whitton AC. Radiotherapy for jugulo- 88. Wegner RE, Rodriguez KD, Heron DE, et al. Linac-based stereotactic tympanic paragangliomas (glomus jugulare tumours). J Laryngol Otol. body radiation therapy for treatment of glomus jugulare tumors. Radiother 1984;98(6):621-629. Oncol. 2010;97(3):395-398. 55. Cummings BJ, Beale FA, Garrett PG, et al. The treatment of glo- 89. Chun SG, Nedzi LA, Choe KS, et al. A retrospective analysis of mus tumors in the temporal bone by megavoltage radiation. Cancer. tumor volumetric responses to five-fraction stereotactic radiotherapy for 1984;53(12):2635-2640. paragangliomas of the head and neck (glomus tumors). Stereotact Funct 56. Reddy EK, Mansfield CM, Hartman GV. Chemodectoma of glomus Neurosurg. 2014;92(3):153-159. jugulare. Cancer. 1983;52(2):337-340. 90. Lieberson RE, Adler JR, Soltys SG, et al. Stereotactic radiosurgery as 57. Dickens WJ, Million RR, Cassisi NJ, et al. Chemodectomas arising the primary treatment for new and recurrent paragangliomas: is open surgical in temporal bone structures. Laryngoscope. 1982;92(2):188-191. resection still the treatment of choice? World Neurosurg. 2012;77(5-6):745-761. 58. Tidwell TJ, Montague ED. Chemodectomas involving the temporal 91. Manolidis S, Shohet JA, Jackson CG, et al. Malignant glomus tumors. bone. Radiology. 1975;116(1):147-149. Laryngoscope. 1999;109(1):30-34. 59. Simko TG, Griffin TW, Gerdes AJ, et al. The role of radiation thera- 92. Langerman A, Athavale SM, Rangarajan SV, et al. Natural history py in the treatment of glomus jugulare tumors. Cancer. 1978;42(1):104-106. of cervical paragangliomas: outcomes of observation of 43 patients. Arch 60. Kim JA, Elkon D, Lim ML, et al. Optimum dose of radiotherapy Otolaryngol Head Neck Surg. 2012;138(4):341-345. for chemodectomas of the middle ear. Int J Radiat Oncol Biol Phys. 93. Jansen JC, van den Berg R, Kuiper A, et al. Estimation of growth 1980;6(7):815-819. rate in patients with head and neck paragangliomas influences the treat- 61. Gibbin KP, Henk JM. Glomus jugulare tumours in South Wales--a ment proposal. Cancer. 2000;88(12):2811-2816. twenty-year review. Clin Radiol. 1978;29(6):607-609. 62. Lybeert ML, van Andel JG, Eijkenboom WM, et al. Radiotherapy of paragangliomas. Clin Otolaryngol Allied Sci. 1984;9(2):105-109. 63. Gilbo P, Tariq A, Morris CG, et al. External-beam radiation therapy for malignant paraganglioma of the head and neck. Am J Otolaryngol. 2015;36(5):692-696. 64. Maruyama Y. Radiotherapy of tympanojugular chemodectomas. Radiology. 1972;105(3):659-663. 65. Fuller AM, Brown HA, Harrison EG, Jr., et al. Chemodectomas of the glomus jugulare tumors. Laryngoscope. 1967;77(2):218-238. 66. Moore G, Yarington CT Jr, Mangham CA Jr. Vagal body tumors: diagnosis and treatment. Laryngoscope. 1986;96(5):533-536. 67. Grubb WB, Jr., Lampe I. The role of radiation therapy in the treatment of chemodectomas of the glomus jugulare. Laryngoscope. 1965;75(12):1861-1871. 68. Wang ML, Hussey DH, Doornbos JF, et al. Chemodectoma of the temporal bone: a comparison of surgical and radiotherapeutic results. Int J Radiat Oncol Biol Phys. 1988;14(4):643-648. 69. Mitchell DC, Clyne CA. Chemodectomas of the neck: the response to radiotherapy. Br J Surg. 1985;72(11):903-905. 70. Glasscock ME III, Harris PF, Newsome G. Glomus tumors: diagnosis and treatment. Laryngoscope. 1974;84(11):2006-2032. 71. Ogura JH, Spector GJ, Gado M. Glomus jugulare and vagale. Ann Otol Rhinol Laryngol. 1978;87(5 pt 1):622-629. 72. Carrasco V, Rosenman J. Radiation therapy of glomus jugulare tumors. Laryngoscope. 1993;103(11 pt 2; suppl 60):23-27. 73. Suarez C, Rodrigo JP, Bodeker CC, et al. Jugular and vagal paragangliomas: systematic study of management with surgery and radiotherapy. Head Neck. 2013;35(8):1195-1204. 74. Makeieff M, Raingeard I, Alric P, et al. Surgical management of carotid body tumors. Ann Surg Oncol. 2008;15(8):2180-2186. 75. Smith GL, Smith BD, Buchholz TA, et al. Cerebrovascular disease risk in older head and neck cancer patients after radiotherapy. J Clin Oncol. 2008;26(31):5119-5125. 76. Scott AS, Parr LA, Johnstone PA. Risk of cerebrovascular events after neck and supraclavicular radiotherapy: a systematic review. Radiother Oncol. 2009;90(2):163-165. 77. Jordan JA, Roland PS, McManus C, et al. Stereotastic radiosurgery for glomus jugulare tumors. Laryngoscope. 2000;110(1):35-38. 78. Mukherji SK, Kasper ME, Tart RP, et al. Irradiated paragangliomas of the head and neck: CT and MR appearance. AJNR Am J Neuroradiol. 1994;15(2):357-363. 79. van Hulsteijn LT, Corssmit EP, Coremans IE, et al. Regression and local control rates after radiotherapy for jugulotympanic paragangliomas: systematic review and meta-analysis. Radiother Oncol. 2013;106(2):161-168. 80. Feigenberg SJ, Mendenhall WM, Hinerman RW, et al. Radiosurgery for paraganglioma of the temporal bone. Head Neck. 2002;24(4):384-389. 81. Liscak R, Vladyka V, Simonova G, et al. Leksell gamma knife radiosurgery of the tumor glomus jugulare and tympanicum. Stereotact Funct Neurosurg. 1998;70(suppl 1):152-160. 82. Eustacchio S, Leber K, Trummer M, et al. Gamma knife radiosurgery for glomus jugulare tumours. Acta Neurochir (Wien). 1999;141(8):811-818. 83. Foote RL, Coffey RJ, Gorman DA, et al. Stereotactic radiosurgery for glomus jugulare tumors: a preliminary report. Int J Radiat Oncol Biol Phys. 1997;38(3):491-495. 84. Foote RL, Pollock BE, Gorman DA, et al. Glomus jugulare tumor: tumor control and complications after stereotactic radiosurgery. Head Neck. 2002;24(4):332-338; discussion 338-339. 85. Chen PG, Nguyen JH, Payne SC, et al. Treatment of glomus jugulare tumors with gamma knife radiosurgery. Laryngoscope. 2010;120(9):1856-1862. 86. Guss ZD, Batra S, Limb CJ, et al. Radiosurgery of glomus jugulare tumors: a meta-analysis. Int J Radiat Oncol Biol Phys. 2011;81(4):e497-502. 87. Gottfried ON, Liu JK, Couldwell WT. Comparison of radiosurgery

July 2016, Vol. 23, No. 3 Cancer Control 241 Outcomes for advanced salivary gland

disease are unsatisfactory, so novel,

systemic, therapeutic strategies should be

determined through future research.

Artwork courtesy of Arthur Pina de Alba. www.arthurpinadealba.com.

Multidisciplinary Management of Salivary Gland Cancers Matthew J. Mifsud, MD, Jon N. Burton, MD, Andy M. Trotti, MD, and Tapan A. Padhya, MD

Background: Salivary carcinomas are a rare group of biologically diverse neoplasms affecting the head and neck. The wide array of different histological entities and clinical presentations has historically limited attempts to establish well-defined treatment algorithms. In general, low-risk lesions can be managed with a single treatment modality, whereas advanced lesions require a more complex, multidisciplinary approach. Methods: The relevant literature was reviewed, focusing on diagnostic and treatment algorithms for salivary malignancies. Results: Salivary carcinomas with high-risk features require an aggressive treatment approach with complete surgical resection, neck dissection to appropriate cervical lymph-node basins, and postoperative radiotherapy. Conclusions: The heterogeneity of salivary neoplasms represents a unique clinical challenge. Despite the multidisciplinary management paradigm detailed in this review, outcomes for advanced disease are unsatisfactory. Future progress will likely require the addition of novel systemic therapeutic strategies.

Introduction Health Organization identified 24 well-defined malig- Salivary carcinomas are a rare collection of neoplasms nant histologies that can arise within major or minor that account for roughly 3% to 5% of all head and neck salivary tissue.4,6 The spectrum of biological behavior malignancies.1-3 In the United States, 2,500 new cases expressed is diverse, ranging from indolent, low-grade are diagnosed each year, and salivary carcinoma has tumors that almost mimic a benign process to aggres- an estimated incidence rate of 2.2 to 3.0 cases per sive, high-risk lesions with a propensity for early dis- 100,000 people per year.1,4,5 A broad array of unique seminated metastasis.3 clinical entities comprises this group of salivary car- Approximately 70% of these lesions originate cinomas. From a pathological perspective, the World within the parotid glands, often presenting as a painless but slowly enlarging preauricular mass.1 A distinct From the Department of Otolaryngology - Head & Neck Surgery (MJM), feature of these carcinomas is their wide anatomical University of Toronto, Toronto, Ontario, Canada, Department of distribution.7,8 The remaining 30% of tumors can arise Otolaryngology - Head & Neck Surgery (JNB, TAP), University of South Florida Morsani College of Medicine, Tampa, Florida, Depart- throughout the upper aerodigestive tract, particularly ments of Radiation Oncology (AMT), Head and Neck and Endo- affecting the oral cavity, pharynx, larynx, nose/para- crine Oncology (TAP), H. Lee Moffitt Cancer Center & Research nasal sinuses, and middle ear space.9 The initial pre- Institute, Tampa, Florida. sentation and potential morbidity depend on the site Address correspondence to Matthew J. Mifsud, MD, Department of Otolaryngology-Head & Neck Surgery, Mount Sinai Hospital, of involvement, which necessitates a nuanced and indi- 600 University Avenue, Room 401, Toronto, ON M5G1X5. E-mail: vidualized treatment strategy.1,6,9 [email protected] A well-defined treatment algorithm has long been Submitted December 15, 2015; accepted March 30, 2016. lacking for salivary carcinomas due to their clinical No significant relationships exist between the authors and the companies/organizations whose products or services may be ref- and biological heterogeneity. Research is further con- erenced in this article. founded by a known propensity for late treatment fail-

242 Cancer Control July 2016, Vol. 23, No. 3 ure, with 25% to 50% of all disease recurring more than over FNAB, but it can cause patient discomfort, comes 5 years after initial management.10 In general, tumors with risk of facial nerve injury, and the theoretical can be stratified into low- or high-risk categories on the (though clinically unproven) risk of tumor seeding.21 basis of clinicopathological risk features. Some key fea- Therefore, the initial approach to salivary neoplasms tures include patient age, tumor size, cervical lymph- must combine data obtained from the clinical exami- node metastasis, grade/histology, American Joint nation as well as imaging, biopsy, and potentially intra- Committee on Cancer (AJCC) tumor stage, and gross operative findings to ensure patients receive optimal perineural invasion.11-13 first-line treatment. For low-risk tumors, a single-modality treatment approach is likely to yield a high rate of locoregional Surgical Considerations disease control (88%–95%) with a correspondingly Regardless of tumor location, histology, or stage, the high 5- to 10-year survival rate.14-16 By contrast, neo- first-line treatment approach for salivary malignan- plasms with high-risk features tend to have an aggres- cies should be curative surgical resection, when fea- sive clinical course and 5-year survival rates ranging sible.1,6,7 Other first-line treatment modalities have from 30% to 40%.1,15,17,18 been studied, but those modalities have generally been reserved for patients with either unresectable Diagnostic Challenges disease or medical comorbidities preventing a surgi- The difficulty distinguishing benign tumors from ma- cal approach.25 Preoperative decision-making is chal- lignancies is a key management challenge for salivary lenging in these cases because the operative strategy gland carcinomas, particularly those within the parot- must take into account the anatomical tumor localiza- id glands. They commonly present as a painless and tion, the extent of locoregional disease, the cosmetic/ slowly evolving neck mass. “Worrisome” clinical symp- functional morbidity of tumor extirpation, and need toms (eg, pain, hard consistency, rapid tumor growth, for reconstruction of the postoperative defect.26,27 An- cranial nerve deficits, cervical adenopathy) may be other important consideration is a tendency of cer- utilized to identify a malignancy in approximately tain tumor histologies (eg, adenoid cystic carcinomas) 30% of cases.7 Imaging studies (eg, ultrasonography, for perineural invasion.28,29 In this scenario, the op- computed tomography, magnetic resonance imaging erative strategy may require extensive nerve dissec- [MRI]) are often utilized in this setting to better define tion (eg, trigeminal or facial nerve branches) toward the neoplastic lesion. The soft-tissue definition afford- and through the skull-base foramina in an attempt to ed by MRI makes this imaging modality the most use- achieve tumor clearance.28,30 ful for characterizing salivary pathology.19,20 MRI can Because the parotid location is the most common also be utilized to discriminate between benign and site, many of these tumors will be addressed with malignant tumors because certain findings on MRI, some form of parotid surgery. Conservative resection including T2 hypointensity, ill-defined tumor margins, (superficial parotidectomy) with facial nerve dissec- diffuse growth pattern, and invasion of subcutaneous tion and preservation is considered to be appropriate tissue, are strongly supportive of a malignant process.19 in most cases.7,27,31,32 Nerve sacrifice is considered nec- However, a certain subset of salivary cancers will have essary only when gross nerve adherence/invasion is a relatively benign appearance on imaging, potentially intraoperatively observed. When possible, nerve leading to misdiagnosis in select patients if pathologi- grafting should be considered at the time of resection cal analysis is not performed.20 to theoretically optimize long-term functional out- The role of routine pretreatment biopsy for sali- comes. 7,27,31,32 Rarely, lateral temporal bone resection vary masses remains a controversial topic.21 Fine nee- or parapharyngeal space dissection/mandibulotomy dle aspiration biopsy (FNAB) is considered by some may also be required for complete excision of gross to be an important diagnostic tool to assist with ini- disease in extensive tumors. For lesions outside of the tial treatment planning and patient counseling.7,22 The parotid tissue, a broad array of surgical strategies may sensitivity rates of FNAB for distinguishing malignant be applicable, ranging from traditional open ablative from benign lesions range from between 80% and 98%, approaches to transoral, endoscopic, or robotic-as- and this technique is particularly accurate at differen- sisted techniques.9,33,34 Regardless of the approach, tiating high-grade from less-aggressive lesions.7,21-23 the operative management of salivary carcino- However, FNAB is poor at clearly defining tumor pa- mas should be to attempt complete tumor eradica- thologies, and it has a relatively low specificity rate and tion, with the goal being microscopically negative a potentially high false-negative rate.22,24 Cytological surgical margins.6,7,9,32 assessment of salivary pathology by FNAB can also be challenging to interpret, potentially leading to inaccu- Role of Adjuvant Radiotherapy rate interpretations at low-volume centers. Ultrasono- Despite an adequate surgical resection, locoregion- graphic-guided core biopsy may be an improvement al treatment failure is common for advanced tumors.

July 2016, Vol. 23, No. 3 Cancer Control 243 Chen et al26 observed a 10-year locoregional disease proach to node-positive disease, with a neck dissec- control rate of 80% for T1 to T2 lesions vs 63% for T3 tion advocated at the time of primary surgery followed to T4 lesions. Postoperative radiotherapy (RT) has by adjuvant RT.7,9,16 In particular, for major salivary tu- been utilized, in particular, to enhance locoregional mors, when node-positive disease is present, multiple control in this setting. Armstrong et al14 presented a ipsilateral neck levels are often involved. In 31 parot- matched-pair analysis of salivary malignancies treated id malignancies with node-positive disease, Ali et al40 with surgery alone (n = 46) vs surgery plus adjuvant RT identified the risk of nodal involvement to be greater (n = 46). Although no difference was seen in locore- than 70% for levels 2 and 3, greater than 50% for levels gional disease control rates in low-stage disease, the 1 and 4, and 40% for level 5. Lim et al41 further con- 5-year rate of locoregional disease control improved firmed these findings in their review of 39 primary from 17% to 51% when RT was added for treatment of parotid malignancies with node-positive necks at pre- advanced lesions (stages 3/4).14 A review by Chung et al35 sentation. In these cases, pathologically positive nodal of multiple retrospective series published between metastases were uniformly found across all neck lev- 1989 and 2012 further supports the benefit of adjuvant els when dissected.41 Thus, when a therapeutic neck RT. In that series for which surgery and adjuvant RT dissection is undertaken, comprehensive clearance of were combined, 5-year locoregional control rates of RT all major nodal basins in the involved side of the neck ranging from 73% to 96% were achieved vs 66% to 84% is advocated.40-42 for surgery alone.35 Elective treatment of the N0 neck is a contro- The decision as to when postoperative RT should versial topic. Some authors propose node sam- be recommended for salivary carcinomas generally pling (neck dissection) in all patients with salivary depends on the presence/absence of “high-risk” clini- carcinoma, because current methods of predict- cal/pathological features. This timing is derived from ing the presence of lymphatic metastasis (in N0 study authors who have identified the most predic- necks) can be imprecise.43 However, the litera- tive variables of locoregional treatment failure.26,36,37 ture is also suggestive of a more selective manage- Of these characteristics, tumor size (> 4 cm), cervi- ment algorithm on the basis of clinicopathological cal lymph-node metastases, high-grade histopathol- risk features. A retrospective series of 363 cases by ogy, and positive surgical margins are considered the Yoo et al42 helps to elaborate on patterns of cervi- most significant.26,36,37 Conversely, a small series of cal lymph-node metastasis in major/minor salivary 58 study patients conducted by Zbären et al38 sug- tumors. The total rate of nodal involvement was gested a potential benefit of adjuvant RT, even in the 20.1%, including 51 study patients with clinically setting of low-risk disease (T1 and T2 and N0). How- node-positive disease at presentation.42 Of the ever, this series is considered by some authors to be an 312 N0 cases, 110 study patients underwent elective outlier and, consequently, adjuvant RT may be of little neck dissection and had a 15.5% rate of occult metas- benefit for small (< 4 cm), low-grade lesions without tasis.42 The other 202 study patients underwent sur- additional adverse pathological features.6,14,16,26 gical resection of the primary lesion without neck Terhaard et al16 conducted the largest institution- dissection (with or without RT) and experienced al series to address the use of postoperative RT for a 2.5% rate treatment failure within the cervical salivary malignancies. Of 498 study patients with ma- lymph-node basins.42 jor/minor salivary carcinoma, 386 received a combined Of the literature series reviewed, most indicated treatment approach (surgery in combination with that high tumor grade, adverse histology, and poten- RT).16 With a mean follow-up period of 76 months, tially tumor size (> 4 cm or T3/T4 tumors) are the most multivariate analysis indicated a locoregional control important predictors for cervical lymph-node metasta- benefit of adjuvant RT only in the settings of T3 to T4 sis.16,42,44-46 When positive nodes are identified in these tumors, node-positive disease, positive/close margins, cases after elective neck dissection, they are clustered bone invasion, or perineural invasion.16 The most re- within levels 1 to 3 in both major and minor salivary cent guidelines from the National Comprehensive malignancies.42,46 Consequently, elective neck man- Cancer Network reflect these findings, advocating use agement is generally advocated only in these high-risk of RT for patients with major or minor salivary malig- scenarios to the most likely affected nodal basin.16,42,44 nancies and these particular adverse features.39 Opti- A paucity of evidence exists to distinguish the optimal results may be achieved with a radiation dose of mal treatment modality in this setting. Because these 60 Gy, increased to between 66 and 70 Gy for those high-risk patients meet the criteria for adjuvant RT, with gross residual disease.16,25,36 elective neck irradiation would seem appropriate. An elective neck irradiation approach, generally with a Treatment of the Neck dose of 46 to 50 Gy for N0 disease, has also been em- Management of the cervical lymph nodes warrants ployed in select patients with a high degree of regional particular discussion. There is little debate on the ap- disease control (> 90%).16,45 However, in most cases, a

244 Cancer Control July 2016, Vol. 23, No. 3 combined approach is offered, given both the prognos- rates of severe late toxicity (in particular, cranial neu- tic benefit of pathological nodal staging and the poten- ropathy and soft-tissue/brain necrosis) have been re- tial therapeutic benefit of identifying these occult nod- ported with neutron beam RT when compared with al metastases.16,42,46 conventional RT.7,16,48 Neutron beam RT has also failed An alternative approach for the N0 neck in pa- to improve survival rates (despite its improved rates tients at high risk is selective level 2 nodal sampling.7 of locoregional control) due to the likelihood of dis- The approach is performed at the time of primary sur- ease dissemination.48,50 gery with frozen section analysis: Patients with positive nodal disease undergo comprehensive lymph- Prognostication and Survival Outcomes adenectomy plus adjuvant RT, whereas those with Given the varying range of tumor histologies and negative findings on biopsy receive elective neck ir- disease presentation, surrogate clinicopathological radiation alone. Such an approach still requires vali- factors are used as a means of stratifying the pre- dation, but it could offer a more selective approach sumed biological aggressiveness of each individual with diminished morbidity.7 However, we continue to case.10,37,51-53 Bhatacharyya et al54 reviewed 903 study recommend elective neck dissection (generally incor- patients with parotid gland carcinomas identified by porating levels 1–3) followed by adjuvant RT for pa- the Surveillance, Epidemiology and End Results da- tients with high-grade or locally advanced (T3 or T4) tabase. Patient age, tumor histology, extraglandular salivary carcinoma. spread, tumor grade, and cervical lymph-node status (N stage) were all found to have an impact on rates Management of Inoperable Disease of survival.54 For patients who present with metastatic disease, ex- Other authors have attempted to synthesize these tensive locoregional disease deemed to be unresect- data into clinical nomograms to theoretically pre- able, or with medical comorbidities that make them dict the likelihood of disease recurrence and patient inappropriate surgical candidates, nonsurgical treat- death.12,13 Carillo et al55 describe a 3-tiered risk-strat- ment strategies can be considered.25 In these settings, ification scheme utilizing patient and tumor-specific high-dose RT (> 66 Gy) may be reasonable to promote features, including age and tumor stage, grade, and durable disease control.7,16,17,26 For example, a series of margin status. In their single institution experience, 64 cases attained a 10-year locoregional disease con- disease recurred in 71.4% of high-risk and 8.8% of trol rate of 40% rate with RT alone.26 In a series of low-risk cases.55 Therefore, a multidisciplinary treat- 40 study patients, Terhaard et al16 further elaborated ment approach is advocated for those with interme- on the dose-response relationship between radia- diate- or high-risk disease, whereas single-modality tion and treatment effect. They found that a 5-year (surgical) management is considered acceptable for locoregional disease control rate of 50% could be low-risk cases.55 Although nomograms may be of ben- achieved at a radiation dose of at least 66 Gy vs 0% efit for patient counseling, none have been widely ad- for lower than 66 Gy.16 opted or validated for routine clinical use.56 High linear energy transfer neutron RT has been A risk-stratification scheme that incorporates explored in this patient population (particularly for current AJCC tumor, node, and metastasis guidelines adenoid cystic carcinomas) due to presumed radio- and tumor histological grading (for appropriate enti- resistance.47,48 A randomized trial completed during ties such as mucoepidermoid or salivary duct carci- the 1980s compared neutron beam RT vs photon nomas) can be used to guide treatment decisions and beam therapy in unresectable disease.49 At 10 years, for reasonably accurate prognostication.57 A multimo- a significant benefit in locoregional control was seen dality treatment approach with primary tumor resec- with neutron RT vs photo beam therapy (56% vs 17%; tion — likely elective neck dissection — and adju- P = .009).7,48,49 Some retrospective series have con- vant RT is advocated for high-stage (AJCC stage 3/4), firmed this finding, with 5- to 6-year locoregional high-grade salivary carcinomas, or both. This ap- control rates as high as 75% after particle beam thera- proach has been proven successful in a number py.48 This has led some authors to conclude that neu- of series because durable rates of locoregional tron beam RT could be considered the treatment of disease control within the head and neck have been choice for unresectable salivary malignancies or in achieved.7,9,16,17,35 those with gross residual disease following attempted However, this aggressive multimodality treat- surgical extirpation.47,49,50 ment approach has not had a robust impact on Despite the potential benefits afforded by neu- survival rates among patients with advanced sali- tron beam RT (compared with conventional RT) in vary malignancies.26,35 Bjørndal et al17 reported a this population, few barriers have prevented its wide 5-year crude survival rate of 30% for stage 4 tumors utilization. It is an expensive technology and avail- (n = 251), despite 60% of study patients receiving ad- able in limited centers worldwide.48 Relatively high juvant RT. A meta-analysis of parotid adenocarcino-

July 2016, Vol. 23, No. 3 Cancer Control 245 ma, which included 19 series (n = 4,631), did identify apy for these carcinomas, but chemoradiotherapy is a modest improvement in survival with use of adju- being explored for the treatment of advanced sali- vant RT; however, the 5-year rate of overall survival vary tumors. among those with high-grade lesions was 35%, often The failure of most conventional chemothera- despite a multimodality approach.18 A multivariate peutic agents has led to the use of targeted agents analysis of Surveillance, Epidemiology and End Re- to treat salivary neoplasms.61 Preliminary studies of sults data reported a statistically significant improve- agents, including tyrosine kinase inhibitors (eg, ce- ment in overall survival with adjuvant RT in cases of tuximab, gefitinib, imatinib) and appropriate mono- salivary malignancies classified as high-risk, locally clonal antibodies (eg, trastuzumab, lapatinib), had advanced (T3/T4 or node-positive), or both (hazard mixed results, with no agent showing clear antitumor ratio = 0.76; P < .001).58 However, squamous cell carci- activity.58,60,65,66 However, other drug targets continue nomas (likely metastatic cutaneous neoplasms) were to be identified, ranging from key cell-surface recep- included, making these data somewhat less applicators (eg, epidermal growth factor, C-kit, human epi- ble to tumors of true salivary origin.58 Overall, a poor dermal growth factor receptor 2, androgen receptors) prognosis is generally attributed to advanced salivary to histologically specific oncogenic gene fusions.6,60,67 malignancies. The presumed mechanism for this is How this exploratory phase into the molecular ge- a biological propensity for distant metastasis, which nomics of salivary malignancies will evolve into new thus far has been unchanged by locoregional man- systemic treatment paradigms remains uncertain.63 agement strategies.1,11,17,59 Conclusions Systemic Therapy The optimal management of salivary malignancies Applying effective systemic treatment approaches requires combining accurate diagnostic pathology presumably will be necessary to improve outcomes with an understanding of key patient- and tumor- for patients with high-risk salivary malignancies due specific clinicopathological features to facilitate to their likelihood of distant disease dissemination.26 risk stratification. Although low-risk tumors can be The appropriate use of chemotherapy has not been controlled in most situations with surgery alone, confirmed and is generally studied only in the setting a more complex multidisciplinary approach is neces- of palliation of metastatic disease. Some barriers ex- sary in cases of intermediate- to high-risk disease. ist that prevent the appropriate use of systemic treat- In this setting, aggressive surgical extirpation, neck ment strategies for salivary malignancies, including dissection, and postoperative radiotherapy (to re- historically low response rates to standard chemo- section bed and at-risk cervical lymph-node basins) therapeutic agents, biological tumor diversity (which are advocated. Treatment escalation with concur- theoretically necessitates use of a histologically rent chemoradiotherapy is also under investigation specific drug regimen), and a paucity of literature as a potential means of improving management out- mostly comprised of small, nonrandomized phase 2 comes. However, in the absence of a more robust studies.60 evidence base, this approach cannot be routinely In the palliative setting, a wide array of conven- recommended. tional chemotherapeutic agents can be used (both It is our belief that future progress in this field singe- and poly-drug regimens); partial and complete will require the innovative use of systemic therapeu- response rates range from 25% to 60%.60-63 These tic agents. This may include the use of newer, targeted responses are also generally of a short duration, as therapies or the novel use of conventional agents in ei- demonstrated by a median 7-month partial response ther the adjuvant or neoadjuvant setting. Regardless, rate of 27% achieved in a single phase 2 trial of cis- the biological diversity of salivary gland malignancies platin, doxorubicin, and cyclophosphamide.62,64 Be- will likely continue to pose both unique and fascinat- cause these study results are not promising for the ing clinical challenges. palliative setting, conventional chemotherapeutic agents have not generally been considered for the References adjuvant management of advanced salivary malig- 1. Guzzo M, Locati LD, Prott FJ, et al. Major and minor salivary gland tumors. Crit Rev Oncol Hematol. 2010;74(2):134-148. nancies; however, the exception has been a group 2. Gillespie MB, Albergotti WG, Eisele DW. Recurrent salivary gland of exploratory studies that have employed concur- cancer. Curr Treat Options Oncol. 2012;13(1):58-70. 65 3. Seethala RR. An update on grading of salivary gland carcinomas. rent chemoradiotherapy. Rationale exists for use of Head Neck Pathol. 2009;3(1):69-77. chemotherapy as a radiosensitizer against salivary 4. Barnes L, Eveson JW, Reichart P, et al. Pathology and Genetics of Head and Neck Tumours. In: World Health Organization. World Health malignancies, which could theoretically impact the Organization Classification of Tumours. Lyon, France: IARC Press; 2005: 210. risk of distant disease dissemination.66,67 At pres- 5. National Cancer Institute. Salivary gland cancer treatment. Revised September 25, 2015. http://www.cancer.gov/types/head-and-neck/hp ent, the research base is too small to advocate for a /salivary-gland-treatment-pdq#section/all. Accessed May 18, 2016. standard approach involving use of chemoradiother- 6. Adelstein DJ, Koyfman SA, El-Naggar AK, et al. Biology and manage-

246 Cancer Control July 2016, Vol. 23, No. 3 ment of salivary gland cancers. Semin Radiat Oncol. 2012;22(3):245-253. tients with high-risk locally advanced salivary gland cancers treated with 7. Vander Poorten V, Bradley PJ, Takes RP, et al. Diagnosis and man- surgery and postoperative radiotherapy. Head Neck. 2011;33(3):318-323. agement of parotid carcinoma with a special focus on recent advances in 37. Hocwald E, Korkmaz H, Yoo GH, et al. Prognostic factors in major molecular biology. Head Neck. 2012;34(3):429-440. salivary gland cancer. Laryngoscope. 2001;111(8):1434-1439. 8. Wong DS. Signs and symptoms of malignant parotid tumours: an 38. Zbären P, Nuyens M, Caversaccio M, et al. Postoperative radiation objective assessment. J R Coll Surg Edinb. 2001;46(2):91-95. therapy for T1 and T2 primary parotid carcinoma: is it useful? Otolaryngol 9. Vander Poorten V, Hunt J, Bradley PJ, et al. Recent trends Head Neck Surg. 2006;145(1):140-143. in the management of minor salivary gland carcinoma. Head Neck. 39. National Comprehensive Cancer Network. Head and Neck Cancer 2014;36(3):444-455. Cancers. v1.2016. https://www.nccn.org/professionals/physician_gls/pdf 10. Chen AM, Garcia J, Granchi PJ, et al. Late recurrence from salivary /head-and-neck.pdf. Accessed May 17, 2016. gland cancer: when does “cure” mean cure? Cancer. 2008;112(2):340-344. 40. Ali S, Palmer FL, DiLorenzo M, et al. Treatment of the neck in carci- 11. Speight PM, Barrett AW. Prognostic factors in malignant tumours of noma of the parotid gland. Ann Surg Oncol. 2014;21(9):3042-3048. the salivary glands. Br J Oral Maxillofac Surg. 2009;47(8):587-593. 41. Lim CM, Gilbert M, Johnson JT, et al. Is level V neck dissection 12. Poorten V, Hart A, Vauterin T, et al. Prognostic index for patients necessary in primary parotid cancer? Laryngoscope. 2015;125(1):118-121. with parotid carcinoma: International external validation in a Belgian-Ger- 42. Yoo S-H, Roh J-L, Kim S-O, et al. Patterns and treatment of man database. Cancer. 2009;115(3):540-550. neck metastases in patients with salivary gland cancers. J Surg Oncol. 13. Ali S, Palmer FL, Yu C, et al. A predictive nomogram for recurrence 2015;111(8):1000-1006. of carcinoma of the major salivary glands. JAMA Otolaryngol Head Neck 43. Nobis C-P, Rohleder NH, Wolff K-D, et al. Head and neck salivary Surg. 2013;139(7):698-705. gland carcinomas--elective neck dissection, yes or no? J Oral Maxillofac 14. Armstrong JG, Harrison LB, Spiro RH, et al. Malignant tumors of Surg. 2014;72(1):205-210. major salivary gland origin. A matched-pair analysis of the role of com- 44. Chen AM, Garcia J, Lee NY, et al. Patterns of nodal relapse after bined surgery and postoperative radiotherapy. Arch Otolaryngol Head surgery and postoperative radiation therapy for carcinomas of the major Neck Surg. 1990;116(3):290-293. and minor salivary glands: what is the role of elective neck irradiation? 15. Spiro RH. Changing trends in the management of salivary tumors. Int J Radiat Oncol Biol Phys. 2007;67(4):988-994. Semin Surg Oncol. 1995;11(3):240-245. 45. Herman MP, Werning JW, Morris CG, et al. Elective neck management 16. Terhaard CHJ, Lubsen H, Rasch CRN, et al. The role of radiothera- for high-grade salivary gland carcinoma. Am J Otolaryngol - Head Neck py in the treatment of malignant salivary gland tumors. Int J Radiat Oncol Med Surg. 2013;34(3):205-208. Biol Phys. 2005;61(1):103-111. 46. Lau VH, Aouad R, Farwell DG, et al. Patterns of nodal involvement 17. Bjørndal K, Krogdahl A, Therkildsen MH, et al. Salivary gland car- for clinically N0 salivary gland carcinoma: refining the role of elective neck cinoma in Denmark 1990-2005: Outcome and prognostic factors: results irradiation. Head Neck. 2014;36(10):1435-1439. of the Danish Head and Neck Cancer Group (DAHANCA). Oral Oncol. 47. Douglas JG, Koh W, Austin-Seymour M, et al. Treatment of salivary 2012;48(2):179-185. gland neoplasms with fast neutron radiotherapy. Arch Otolaryngol Head 18. Jeannon J-P, Calman F, Gleeson M, et al. Management of advanced Neck Surg. 2003;129(9):944-948. parotid cancer. A systematic review. Eur J Surg Oncol. 2009;35(9):908-915. 48. Jereczek-Fossa BA, Krengli M, Orecchia R. Particle beam radio- 19. Christe A, Waldherr C, Hallett R, et al. MR imaging of parotid tumors: therapy for head and neck tumors: radiobiological basis and clinical expe- typical lesion characteristics in MR imaging improve discrimination between rience. Head Neck. 2006;28(8):750-760. benign and malignant disease. Am J Neuroradiol. 2011;32(7):1202-1207. 49. Laramore GE, Krall JM, Griffin TW, et al. Neutron versus photon irra- 20. de Ru JA, van Leeuwen MS, van Benthem PPG, et al. Do magnetic diation for unresectable salivary gland tumors: final report of an RTOG-MRC resonance imaging and ultrasound add anything to the preoperative work- randomized clinical trial. Int J Radiat Oncol Biol Phys. 1993;27(2):235-240. up of parotid gland tumors? J Oral Maxillofac Surg. 2007;65(5):945-952. 50. Huber PE, Debus J, Latz D, et al. Neutron beam radiation therapy: 21. Douville NJ, Bradford CR. Comparison of ultrasound-guided core an overview of treatment and oral complications when treating salivary biopsy versus fine-needle aspiration biopsy in the evaluation of salivary gland malignancies. J Oral Maxillofac Surg. 2003;3016(2):377-382. gland lesions. Head Neck. 2013;35(11):1657-1661. 51. Chen AM, Lau VH, Farwell DG, et al. Mucoepidermoid carci- 22. Alphs HH, Eisele DW, Westra WH. The role of fine needle aspira- noma of the parotid gland treated by surgery and postoperative ra- tion in the evaluation of parotid masses. Curr Opin Otolaryngol Head Neck diation therapy: clinicopathologic correlates of outcome. Laryngoscope. Surg. 2006;14(2):62-66. 2013;123(12):3049-3055. 23. Kim BY, Hyeon J, Ryu G, et al. Diagnostic accuracy of fine needle 52. Stodulski D, Mikaszewski B, Stankiewicz C. Are all prognostic fac- aspiration cytology for high-grade salivary gland tumors. Ann Surg Oncol. tors in parotid gland carcinoma well recognized? Eur Arch Oto-Rhino-Lar- 2013;20(7):2380-2387. yngology. 2012;269(3):1019-1025. 24. Tryggvason G, Gailey MP, Hulstein SL, et al. Accuracy of fine-nee- 53. Renehan AG, Gleave EN, Slevin NJ, et al. Clinico-pathological and dle aspiration and imaging in the preoperative workup of salivary gland treatment-related factors influencing survival in parotid cancer. Br J mass lesions treated surgically. Laryngoscope. 2013;123(1):158-163. Cancer. 1999;80(8):1296-1300. 25. Mendenhall WM, Morris CG, Amdur RJ, et al. Radiotherapy 54. Bhattacharyya N, Fried MP. Determinants of survival in parotid alone or combined with surgery for salivary gland carcinoma. Cancer. gland carcinoma: a population-based study. Am J Otolaryngol Head Neck 2005;103(12):2544-2550. Med Surg. 2005;26(1):39-44. 26. Chen AM, Granchi PJ, Garcia J, et al. Local-regional recurrence 55. Carrillo JF, Vázquez R, Ramírez-Ortega MC, et al. Multivariate pre- after surgery without postoperative irradiation for carcinomas of the major diction of the probability of recurrence in patients with carcinoma of the salivary glands: Implications for adjuvant therapy. Int J Radiat Oncol Biol parotid gland. Cancer. 2007;109(10):2043-2051. Phys. 2007;67(4):982-987. 56. Takahama A, Sanabria Á, Benevides GM, et al. Comparison of 27. Spiro JD, Spiro RH. Cancer of the parotid gland: role of 7th nerve two prognostic scores for patients with parotid carcinoma. Head Neck. preservation. World J Surg. 2003;27(7):863-867. 2009;31(9):1188-1195. 28. Amit M, Binenbaum Y, Trejo-Leider L, et al. International collabora- 57. Van Der Schroeff MP, Terhaard CHJ, Wieringa MH, et al. Cytology tive validation of intraneural invasion as a prognostic marker in adenoid and histology have limited added value in prognostic models for salivary cystic carcinoma of the head and neck. Head Neck. 2014;55(7):691-696. gland carcinomas. Oral Oncol. 2010;46(9):662-666. 29. Coca-Pelaz A, Rodrigo JP, Bradley PJ, et al. Adenoid cystic carci- 58. Mahmood U, Koshy M, Goloubeva O, et al. Adjuvant radiation ther- noma of the head and neck - an update. Oral Oncol. 2015;51(7):652-661. apy for high-grade and/or locally advanced major salivary gland tumors. 30. Tarsitano A, Pizzigallo A, Gessaroli M, et al. Intraoperative biopsy Arch Otolaryngol Head Neck Surg. 2011;137(10):1025-1030. of the major cranial nerves in the surgical strategy for adenoid cystic carci- 59. Ali S, Sarhan M, Palmer FL, et al. Cause-specific mortality in pa- noma close to the skull base. Oral Surg Oral Med Oral Pathol Oral Radiol. tients with mucoepidermoid carcinoma of the major salivary glands. Ann 2012;113(2):214-221. Surg Oncol. 2013;20(7):2396-2404. 31. Chang Y, Young S, Kim K, et al. Conservative parotidectomy for the 60. Rizk S, Robert A, Vandenhooft A, et al. Activity of chemotherapy treatment of parotid cancers. Oral Oncol. 2005;41(10):1021-1027. in the palliative treatment of salivary gland tumors: review of the literature. 32. Bell RB, Dierks EJ, Homer L, et al. Management and outcome of Eur Arch Oto-Rhino-Laryngology. 2007;264(6):587-594. patients with malignant salivary gland tumors. J Oral Maxillofac Surg. 61. Lagha A, Chraiet N, Ayadi M, et al. Systemic therapy in the man- 2005;63(7):917-928. agement of metastatic or advanced salivary gland cancers. Head Neck 33. Villanueva NL, de Almeida JR, Sikora AG, et al. Transoral robotic Oncol. 2012;4(10):19. surgery for the management of oropharyngeal minor salivary gland tu- 62. Debaere D, Vander Poorten V, Nuyts S, et al. Cyclophosphamide, mors. Head Neck. 2014;36(1):28-33. doxorubicin, and cisplatin in advanced salivary gland cancer. B-ENT. 34. Moore BA, Burkey BB, Netterville JL, et al. Surgical management of 2011;7(1):1-6. minor salivary gland neoplasms of the palate. Ochsner J. 2008;8(4):172-180. 63. Vander Poorten V, Meulemans J, Delaere P, et al. Molecular markers 35. Chung MP, Tang C, Chan C, et al. Radiotherapy for nonadenoid and chemotherapy for advanced salivary cancer. Curr Otorhinolaryngol cystic carcinomas of major salivary glands. Am J Otolaryngol Neck Med Rep. 2014:85-96. Surg. 2013;34(5):425-430. 64. Licitra L, Cavina R, Grandi C. Cisplatin, doxorubicin and cyclophospha- 36. Feinstein TM, Lai SY, Lenzner D, et al. Prognostic factors in pa- mide in advanced salivary gland carcinoma. Ann Oncol. 1996;7(6):640-642.

July 2016, Vol. 23, No. 3 Cancer Control 247 65. Tanvetyanon T, Qin D, Padhya T, et al. Outcomes of postoperative concurrent chemoradiotherapy for locally advanced major salivary gland carcinoma. Arch Otolaryngol Head Neck Surg. 2009;135(7):687-692. 66. Cerda T, Shan X, Vignot S, et al; REFCOR. A rationale for chemoradiation (vs radiotherapy) in salivary gland cancers ? Crit Rev Oncol / Hematol. 2014;91(2):142-158. 67. Pederson AW, Salama JK, Haraf DJ, et al. Adjuvant chemoradiotherapy for locoregionally advanced and high-risk salivary gland malignancies. Head Neck Oncol. 2011;3(1):31.

248 Cancer Control July 2016, Vol. 23, No. 3 Pathology Report

Practical Issues for Retroperitoneal Sarcoma Vicky Pham, MS, Evita Henderson-Jackson, MD, Matthew P. Doepker, MD, Jamie T. Caracciolo, MD, Ricardo J. Gonzalez, MD, Mihaela Druta, MD, Yi Ding, MD, and Marilyn M. Bui, MD, PhD

Background: Retroperitoneal sarcoma is rare. Using initial specimens on biopsy, a definitive diagnosis of histological subtypes is ideal but not always achievable. Methods: A retrospective institutional review was performed for all cases of adult retroperitoneal sarcoma from 1996 to 2015. A review of the literature was also performed related to the distribution of retroperitoneal sarcoma subtypes. A meta-analysis was performed. Results: Liposarcoma is the most common subtype (45%), followed by leiomyosarcoma (21%), not otherwise specified (8%), and undifferentiated pleomorphic sarcoma (6%) by literature review. Data from Moffitt Cancer Center demonstrate the same general distribution for subtypes of retroperitoneal sarcoma. A pathology-based algorithm for the diagnosis of retroperitoneal sarcoma is illustrated, and common pitfalls in the pathology of retroperitoneal sarcoma are discussed. Conclusions: An informative diagnosis of retroperitoneal sarcoma via specimens on biopsy is achievable and meaningful to guide effective therapy. A practical and multidisciplinary algorithm focused on the histopathology is helpful for the management of retroperitoneal sarcoma.

Introduction tic, and predictive information based on a relatively Soft-tissue sarcomas are mesenchymal neoplasms small amount of tissue obtained on biopsy. that account for up to 1% of all newly diagnosed ma- In this study, we compare our experience, espe- lignancies at a rate of 3.6 per 100,000 per year.1-3 Com- cially the distribution of histological subtypes, with pared with bone or visceral sarcomas, they make up the literature to develop a better understanding of 58% of all sarcomas.3 Although the extremities rep- common and rare tumor types, pertinent ancil- resent the most common location of soft-tissue sarco- lary testing, diagnostic pitfalls, and refine a practi- ma, retroperitoneal sarcoma accounts for 9% to 15% cal, multidisciplinary, and algorithmic approach to of all adult soft-tissue sarcomas.3,4 For surgical pathol- achieve an informative diagnosis to help guide the ogists, retroperitoneal tumors will be encountered in therapeutic plan. daily practice regardless of the practice setting. Ret- roperitoneal sarcoma presents diagnostic challenges Methods due to its rarity, variety of tumor types, a general level After obtaining Institutional Review Board approval, of unfamiliarity among surgical pathologists, and lack the database Transmed (H. Lee Moffitt Cancer Center of generally accepted guidelines in its diagnostic ap- & Research Institute, Tampa, Florida) was retrospec- proach. In the era of personalized medicine, pathol- tively reviewed for all histological subtypes of retro- ogists play a critical and central role in patient care. peritoneal sarcomas from 1996 to 2015. Transmed in- Demand has increased to obtain diagnostic, prognos- tegrates data from all patients seen at Moffitt Cancer Center since 1996 (approximately 395,000 patients, From the University of South Florida Morsani College of Medicine regardless of diagnosis) as well as other patients not (VP), the Departments of Anatomic Pathology (EH-J, MMB), Sar- treated at Moffitt Cancer Center but who consented to coma (EH-J, JTC, RJG, MD, MMB), Surgical Oncology (MPD), and the Total Cancer Care protocol at one of 17 consor- Diagnostic Imaging (JTC), H. Lee Moffitt Cancer Center & Research Institute, Tampa, Florida, and the Department of Pathology (YD), tium sites (approximately 36,000 patients). This result JiShuiTan Hospital, Beijing, China. was cross-checked by sarcoma pathologists as well as Address correspondence to Marilyn M. Bui, MD, PhD, Department the deidentified result of another retroperitoneal sar- of Anatomic Pathology, Moffitt Cancer Center, 12902 Magnolia coma retrospective institutional review. Drive, Tampa, FL 33612. E-mail: [email protected] Submitted August 17, 2015; accepted December 29, 2015. A systematic review of the English literature This study received financial assistance from Information Shared was conducted for works published between 2000 Services Department/Collaborative Data Services Core at the H. Lee and 2015. This yielded 85 search results. Another Moffitt Cancer Center & Research Institute supported under NIH literature search was conducted yielding an ad- grant P30-CA76292. This research was also supported by an award from the Scholarly Concentrations Program at USF Health Morsani ditional 444 results. After excluding case reports College of Medicine. and studies with locations of retroperitoneal sarco-

July 2016, Vol. 23, No. 3 Cancer Control 249 ma mixed with other locations, our search yielded 54 studies from which we extracted information Table 1. — Distribution of Histological Subtypes of Retroperitoneal Soft-Tissue Sarcoma about the distribution of histological subtypes within each study patient population. Subtype Distribution Distribution From the From The meta-analysis was conducted using Stats- Literature, Moffitt, % Direct (Cheshire, United Kingdom). % Liposarcoma (all subtypes) 45.1 54.6 Results Well differentiated 23.9 15.3 Literature Review A literature search revealed the following list of most Dedifferentiated 20.9 15.6 to least common histological subtypes of tumors. Per- Myxoid/round cell 0.9 6.8 centages shown are out of all retroperitoneal sarcoma Pleomorphic 0.4 1 subtypes (Table 1). Our review and data analysis of Leiomyosarcoma 21.3 26.1 retroperitoneal sarcoma studies concluded that the Other 8.2 0 most common subtype of retroperitoneal sarcoma Malignant fibrous histiocytoma/ 6.4 1 is liposarcoma, which constitutes 45.1% of all retro- undifferentiated pleomorphic peritoneal sarcoma (Fig 1). The next most common sarcoma subtypes are leiomyosarcoma (21.3%), other (8.2%), Sarcoma not otherwise specified 1.6 4.9 and malignant fibrous histiocytoma/undifferentiated Malignant peripheral nerve sheath 1.3 1 pleomorphic sarcoma (6.4%; see Table 1). Well-differ- tumor Well differentiated entiated liposarcoma and dedifferentiated liposarco- Solitary fibrous tumor/ 0.8 0 Dedifferentiated hemangiopericytoma ma constituted 45.8% and 44.8% of all retroperitoneal Myxoid/round cell liposarcomas, respectively (Fig 2). The less common Fibrosarcoma (nondesmoid) 0.5 1 Pleomorphic subtypes were myxoid/round cell liposarcoma and Rhabdomyosarcoma 0.4 1 pleomorphic liposarcoma. Schwannoma 0.3 0 Angiosarcoma 0.1 0 Institutional Results Anaplastic sarcoma < 0.1 0 A review of Transmed, which represents single-institution data, is summarized in Table 1. Our data for Adenosarcoma < 0.1 0 retroperitoneal sarcoma subtypes are consistent with Chondrosarcoma < 0.1 0 our literature search. Liposarcomas of all subtypes at Desmoplastic small-round cell tumor < 0.1 0 Moffitt Cancer Center make up 54.6% (168 cases) of all Epithelioid sarcoma < 0.1 0 retroperitoneal sarcoma (307 cases), followed by leio- Extraskeletal osteosarcoma < 0.1 1 myosarcomas (80 cases), sarcoma not otherwise spec- Fibromyxosarcoma < 0.1 1 ified (15 cases), malignant peripheral nerve tumor Giant cell sarcoma < 0.1 3 (3 cases), and other less common subtypes (1–2 cases; see Table 1). The trend for most to least common tu- Mesenchymoma < 0.1 0 mor types appears to be similar (Fig 3). In Fig 4, data Primitive neuroectodermal tumor/ < 0.1 1 extraskeletal Ewing sarcoma from Moffitt Cancer Center on the histological subtypes of retroperitoneal liposarcomas are compared Small cell/embryonal/synovial/ < 0.1 0 undifferentiated sarcoma with the data found during our literature search. Spindle cell sarcoma < 0.1 1 The median age of the patients in all 54 studies was 58 years. The percentages of those who were men Undifferentiated round-cell < 0.1 1 sarcoma and women were 53.2% and 46.8%, respectively.

Gap Between Practice and Literature tochemical hallmarks and characteristics that can be Most of the literature we reviewed describes sarcoma diagnosed with a small amount of tissue acquired on in categories based on histological type and morpho- biopsy; however, commonly, sarcoma does not have logical pattern.5 Morphological features are evaluated such a signature and morphologically overlaps with to determine the differential diagnosis, ranking pos- other tumors. Ancillary tests can be performed to sible diagnoses from most likely to least likely in con- narrow the diagnosis, with the understanding that it junction with clinical and radiological clues, and then may not be possible to achieve a definitive diagno- applying ancillary testing to narrow down the diagnosis. However, it is important to realize that, when a sis. Thus, it is important to understand the common definitive diagnosis is not achievable, an alternative and rare types of retroperitoneal sarcoma. option is to provide useful, diagnostic information to Some sarcomas have molecular and immunohis- guide the clinical team in the next appropriate step

250 Cancer Control July 2016, Vol. 23, No. 3 Liposarcoma 45.1 Leiomyosarcoma 21.3

Other 8.2 Malignant fibrous histiocytoma/ 6.4 undifferentiated pleomorphic sarcoma Sarcoma not otherwise specified 1.6 Malignant peripheral nerve sheath tumor 1.3

Histological Subtype Solitary fibrous tumor/hemangiopericytoma 0.8 Rhabdomyosarcoma 0.4

0 5 10 15 20 25 30 35 40 45 50 Percentage

Fig 1. — Distribution of histological subtypes of retroperitoneal sarcoma.

Well differentiated 45.8 Dedifferentiated 44.8 Myxoid/round cell 4.7 Pleomorphic 0.8 Histological Subtype 0 5 10 15 20 25 30 35 40 45 50 Percentage

Fig 2. — Distribution of histological subtypes of retroperitoneal liposarcoma. for the patient’s care. In such situations, being aware and higher-grade tumors being more appropriate for of the management plan for various diagnoses is criti- treatment.1 Future research may enable more histolo- cal. Some diagnoses are managed according to estab- gy-specific decisions. We speculate that new data may lished institutional pathways. The literature often fo- suggest no role exists for radiotherapy in the treatment cuses on achieving a definitive diagnosis, rather than of leiomyosarcoma. realizing an informative diagnosis. However, histological subtypes play a role in chemotherapy. Some subtypes may respond better to Multidisciplinary Approach to certain chemotherapeutic agents or regimens and, Pathology-Focused Management conversely, several different sarcoma subtypes will The interdisciplinary management of retroperitoneal similarly respond to identical chemotherapy treat- sarcoma involves chemotherapy, radiotherapy, and ments (Fig 5; Table 2).7-18 The role of chemotherapy surgery, along with providing prognostic, predictive, in the management of retroperitoneal sarcoma is not and diagnostic information to the clinical team and the well defined. The 2 most important determinants of patient. The mainstay of treatment for retroperitoneal overall survival (OS) are tumor grade and extent of re- sarcoma is surgical resection. Complete resection with section, with subtypes having a less important role in a microscopically negative margin (R0) remains the po- determining prognosis.6,19 Therefore, when consider- tential likelihood for cure.6 Surgical decision-making ing whether to pursue a definitive diagnosis for a tu- in retroperitoneal sarcoma is not solely based on his- mor subtype, the relative costs and turnaround time tological subtypes, but rather on factors such as per- of ancillary tests vs the importance or relevance of the formance status, patient comorbidities, and extent of test results in the larger context of the interdisciplin- tumor involvement into adjacent organs and vascular ary management of retroperitoneal sarcoma must be structures.1 Adjunctive therapies such as perioperative considered; therefore, pathologists must make judi- radiotherapy and chemotherapy are also selectively cious use of their tissue samples, time, and institutional used within the context of interdisciplinary review on resources. a case-by-case basis. At this time, use of radiotherapy does not depend on histological subtypes but rather on Neoadjuvant Chemotherapy tissue tolerance and tumor grades, with intermediate- The role of neoadjuvant chemotherapy is not well de-

July 2016, Vol. 23, No. 3 Cancer Control 251 All liposarcoma subtypes

All leiomyosarcoma subtypes

Sarcoma not otherwise specified Malignant peripheral nerve

Histological Subtype sheath tumor

0 10 20 30 40 50 60 Percentage

Data Source Malignant Peripheral Sarcoma Not Otherwise All Subtypes of All Subtypes of Nerve Sheath Tumor Specified Leiomyosarcoma Liposarcoma Moffitt 1.0 4.9 26.1 54.7 Literature 1.3 1.6 21.3 45.1 Fig 3. — Comparison of retroperitoneal sarcoma subtypes demonstrates a consistent distribution at Moffitt Cancer Center vs literature search.

Dedifferentiated

Well differentiated

Myxoid/round cell

Pleomorphic Histological Subtype

0 5 10 15 20 25 30 Percentage

Data Source Pleomorphic Myxoid/Round Cell Well Differentiated Dedifferentiated Moffitt 1.0 6.8 15.3 15.6 Literature 0.4 0.9 23.9 20.9 Fig 4. — Comparison of liposarcoma subtypes distribution at Moffitt Cancer Center vs literature search. Percentages take into account all retroperitoneal sarcoma subtypes. fined in retroperitoneal sarcoma due to the rarity of the The administration of systemic therapy in the neo- disease; therefore, its role is often extrapolated from adjuvant setting is often combination therapy with studies that include extremity sarcoma. Several case doxorubicin and ifosfamide, a regimen with potential series have been reported, but no definitive prospec- for renal toxicity but with a higher response rate (31% tive trials have examined OS differences in those re- vs 14% in the single-arm doxorubicin alone).24 As a con- ceiving preoperative vs postoperative chemotherapy.1 sequence of resection, nephrectomy is often required, Similarly, to date, no randomized trials exist of neoad- so chemotherapy in the neoadjuvant setting allows for juvant chemotherapy vs resection alone for retroperi- use of ifosfamide, which, in combination with doxo- toneal sarcoma.7 Although data are limited, preopera- rubicin, may have a larger effect on tumor response tive chemotherapy appears to be safe and occasionally when compared with doxorubicin alone.24 induces a modest radiographic response, which may impact surgical outcomes in select patients.20-23 The Localized Treatment (Curative Intent) theoretical advantage of preoperative chemothera- Retroperitoneal sarcoma is a heterogeneous group of py focuses on the potential to reduce the complex- tumors with multiple histological subtypes and grades ity of potential surgery for the tumor subtypes that re- that vary in chemosensitivity. For localized treatment, spond to systemic therapy, especially intermediate- or doxorubicin and ifosfamide are commonly used.25 high-grade tumors.1,8 Opportunity also exists to deter- Among the liposarcomas, well-differentiated liposar- mine the response to chemotherapy in the neoadjuvant coma does not respond (response rate = 0%) to che- setting as a determinant of therapy continuation in the motherapy and dedifferentiated liposarcoma responds adjuvant setting. poorly (response rate = 25%).9 Myxoid liposarcoma has

252 Cancer Control July 2016, Vol. 23, No. 3 the highest response rate at 48%.9 Round cell liposarco- Chemotherapy for Metastatic or Advanced ma, a high-grade spectrum of myxoid liposarcoma, has Soft-Tissue Sarcoma (Palliative Intent) a response rate of 17%.9 Pleomorphic liposarcoma has Chemotherapy has an established role in the palliative a response rate of 33%.1 The response rate of leiomyo- management of metastatic or advanced soft-tissue sar- sarcoma to chemotherapy is 25%.10-12 For the remain- coma.6,14 In the metastatic setting, use of single-agent ing retroperitoneal sarcoma subtypes, response rates doxorubicin or combination doxorubicin/ifosfamide has range from 21% to 31% (see Fig 5).7,9,11-14,16,26 shown a consistent response.24 However, some subtypes

Table 2. — Use of Chemotherapy for the Management of Clinical and radiological information (including contrast-enhanced Retroperitoneal Soft-Tissue Sarcomas abdominopelvic CT or MRI to evaluate local extent of disease and CT of the chest to evaluate for distant metastases) a Subtype Chemotherapy Type of assessed by interdisciplinary team Response Chemotherapy Patients with Li-Fraumeni syndrome should be referred for Rate,a % further genetic assessment11,16 Myxoid/round cell Myxoid: 48 Neoadjuvant Setting liposarcoma Round cell: 179 Doxorubicin/ifosfamide for localized treatment

Well-Differentiated Dedifferentiated 21–3310-12 Metastatic Setting Liposarcoma (24%) liposarcoma Single-agent doxorubicin Further biopsy not mandatory Leiomyosarcoma Adjuvant Setting Adjuvant chemotherapy depending on cross-sectional Pleomorphic imaging determined on Image-Guided Core Biopsy liposarcoma Complete resection is case-by-case basis; with mainstay of treatment Preferred to open surgical Undifferentiated large high-grade tumors, biopsy pleomorphic sarcoma adjuvant chemotherapy Unresponsive to anthracycline- Other subtypes increases metastasis-free based chemotherapy or 8 radiotherapy9 survival rate Most locoregionally occur rather than metastasize9,14 Angiosarcomab Histological specific Solitary fibrous treatment (not tumorb anthracycline-based) recommended16 Hemangiopericytomab Other Retroperitoneal Differential Diagnoses Ewing/primitive — Soft-Tissue Sarcoma Subtypes GIST neuroectodermal Desmoid tumor tumorb (aggressive fibromatosis) Nonpleomorphic Metastatic lesion Surgery = Mainstay rhabdomyosarcomab,c Treatment Lipoma Lymphoma Synovial sarcoma/ Synovial Synovial Sarcoma If tumor is partially resect- MPNST sarcoma: Responds best to Primary neoplasm arising able, then neoadjuvant 25–3110,12 ifosfamide-based radiotherapy/chemotherapy from pancreas, adrenal 13 first-line therapy becomes more important glands, kidneys, duodenum MPNST: 21 to treat periphery of lesion as (doxorubicin/ Adenopathy from testicles ifosfamide)10,12; it abuts critical structures in a young male with intent of consolidating also sensitive to peritumoral reactive zone radiotherapy15 and rendering the close MPSNT margin sterile Responds best to Neoadjuvant chemotherapy doxorubicin/ifosfamide May Not Require Surgery also important for in metastatic setting13 chemosensitive histologies7,11 as First Approach New evidence may suggest Examples: lymphoma, GIST, Preoperative Radiotherapy: External beam radiotherapy used for no role for radiotherapy in Ewing sarcoma intermediate- to high-grade tumors, especially in more radiosensitive treating leiomyosarcoma tumors (eg, extraosseous Ewing sarcoma, primitive neuroectodermal tumor). The more common subtypes, such as well-differentiated liposarcoma and leiomyosarcoma, are generally unresponsive to radiotherapy.18 Refer to Table 2 Postoperative Radiotherapy: Not commonly used due to toxic effects to adjacent organs with no apparent benefit.17 aIn general, response rates extracted from studies of soft-tissue sarcomas Fig 5. — Proposed algorithm for the pathology-focused management of (not restricted to retroperitoneum). retroperitoneal soft-tissue sarcoma. bComprises < 0.8% of all subtypes. aTeam is made up of a surgical oncologist, radiologist, oncologist, and cMost studies reviewed list rhabdomyosarcoma as a general subtype pathologist. without denoting more specific subtypes. CT = computed tomography, GIST = gastrointestinal stromal tumor, MPNST = malignant peripheral nerve sheath tumor. MRI = magnetic resonance imaging.

July 2016, Vol. 23, No. 3 Cancer Control 253 (eg, solitary fibrous tumor, well-differentiated liposarco- nal damage, secondary cancers, cardiomyopathy, and ma) are resistant to anthracycline-based therapy, while overall impairment of quality of life.16 others show a differential response to cytotoxic systemic therapy.9,27 Thus, the concept of histology-driven treat- Radiotherapy ment has arisen, rather than a one-size-fits-all or all-in- The overall benefit of radiotherapy for use with retro- clusive approaches to therapy in patients with metastatic peritoneal sarcoma has yet to be established, with most soft-tissue sarcoma. Angiosarcomas have been shown of the data being extrapolated from studies of soft-tissue to respond to paclitaxel and pegylated liposomal doxo- sarcoma of the extremities.33 However, concern remains rubicin, leiomyosarcomas to combination gemcitabine/ about the increased risk of treatment-related toxicity to docetaxel, desmoid tumors to liposomal doxorubicin, highly radiosensitive visceral structures due to their rap- and synovial sarcoma and malignant peripheral nerve idly proliferating mucosa and rich blood supply.34 The sheath tumor (MPNST) respond best to doxorubicin/ relatively low rate of radiation tolerance for surrounding ifosfamide.10-15,26-28 Pazopanib, a multitargeted tyrosine normal tissues (liver, kidney, gastrointestinal tract, spi- kinase inhibitor, has demonstrated single-agent activity nal cord) predisposes patients to risks of intestinal per- in patients with advanced soft-tissue sarcoma subtypes, foration, peritonitis, and peripheral neuropathy.35 except for liposarcoma.16 Consensus-based recommen- Use of preoperative radiotherapy is currently being dations for the treatment of metastatic soft-tissue sarco- investigated in an accruing, prospective, randomized, ma from the National Comprehensive Cancer Network multicenter trial (NCT01344018). This type of trial is in- provide specific therapy regimen recommendations vestigating the potential for external beam radiotherapy for the following retroperitoneal sarcoma subtypes: (EBRT) to reduce local regional failure.36 Proponents of soft-tissue sarcoma with nonspecific histologies, angio- preoperative radiation cite the potential benefits of posarcoma, solitary fibrous tumor/hemangiopericytoma, tentially using lower doses, while the tumor displaces Ewing sarcoma/primitive neuroectodermal tumor, non- radiosensitive viscera outside the field of radiation.37 pleomorphic rhabdomyosarcoma, and desmoid tumors Proponents also claim gross tumor volume can be more (see Fig 5).7,9,11,14,16 Although select patients with sarcoma adequately defined, which would allow for more accu- do derive substantial clinical benefit from chemothera- rate preoperative treatment planning.37 py, most patients develop metastatic disease that is in- Use of postoperative EBRT has been studied but curable.29 In a study of 488 participants with advanced largely abandoned due to its toxic effects of the remain- soft-tissue sarcoma who were treated with first-line che- ing organs within the tissue bed after resection, with no motherapy at a single institution, 45% of them derived apparent improvement in survival.17 Another concern clinical benefit from treatment in terms of partial re- with postoperative radiation suggests the difficulty in sponse (PR) or prolonged disease stabilization; the me- defining a precise area of the tumor bed to apply EBRT.17 dian rate of OS was 12 months.30 Our experience at Moffitt Cancer Center favors use Furthermore, in the past decade, greater empha- of preoperative EBRT for intermediate- to high-grade sis has been placed on identifying the underlying mo- tumors, especially in more radiosensitive tumors, such lecular drivers of sarcomas.31 Several potential novel, as extraosseous Ewing sarcoma/primitive neuroecto- systemic therapies for soft-tissue sarcoma have been dermal tumor. The more common subtypes, such as identified, including MDM2 targets, cyclin-dependent well-differentiated liposarcoma and leiomyosarcoma, kinase 4 inhibitors, peroxisome proliferator-activated are generally unresponsive to radiation.18 Thus, concern receptors (critical regulators of normal adipocyte dif- remains about the increased risk of treatment-related ferentiation), and tyrosine kinase receptors.32 Other toxicity to visceral structures. targets reported but not yet tested include YEATS4, c-Jun, and JNK.32 Pathological Prognostic Factors The most important prognostic factors for survival are Adjuvant Chemotherapy extent of tumor resection and histological grade, al- Several prospective, randomized trials of study pa- though histological subtype is also emphasized as an tients who received adjuvant regimens following sur- important factor.38 Other factors influencing prognosis gical resection have demonstrated decreased local re- include tumor stage, patient age, tumor size, and mul- currence rates, but the effect on OS is less clear.1 The tifocality. Nomograms have been developed and vali- benefits of adjuvant chemotherapy must be addressed dated to more accurately predict postoperative survival based on the individual while simultaneously taking based on these and other factors.39 Well-differentiated into consideration performance status, disease loca- liposarcomas have the most favorable outcome, where- tion, tumor size, comorbid factors (including age), and as leiomyosarcomas, pleomorphic sarcoma/malignant histological subtype. The potential for benefit must be fibrous histiocytoma, MPNST, and dedifferentiated li- discussed in the context of expected treatment-related posarcomas exhibit the least favorable outcomes.38 toxicities, including sterility in younger individuals, re- Other prognostic concerns include risks for locore-

254 Cancer Control July 2016, Vol. 23, No. 3 gional recurrence and metastasis. Local regional relapse Other symptoms such as hematuria, uncontrollable is the main cause of disease-related death and, in con- hypertension, or early satiety may raise concern for a junction with retroperitoneal sarcoma, the risk for de- primary tumor of a retroperitoneal origin rather than veloping abdominal sarcomatosis also results in death, primary retroperitoneal sarcoma.1 even in the absence of systemic dissemination.40,41 Tu- Findings on cross-sectional imaging may also help mor grade and histological subtypes are major prognos- assess the internal composition of the tumor, allowing tic factors related to metastatic occurrence.19,42 Toule- for the preoperative prediction of tumor histopathol- monde et al42 reviewed the data of 586 study patients ogy and assessment of local extent of disease; these with retroperitoneal sarcoma during a multicenter anal- may have an impact on neoadjuvant treatment, surgical ysis, looking at patterns of care at diagnosis and prog- planning, or both. Tumor staging should include imag- nostic factors, with a focus on main histological sub- ing of the chest to evaluate for pulmonary metastatic types. Those findings are summarized in Table 3.8,14,42 disease and is typically performed with CT. Of all the retroperitoneal sarcoma subtypes, well- Proposed Algorithm for Management differentiated liposarcomas make up 24% and are com- A patient’s clinical and radiological information is first prised of simple-appearing fat that has a characteristic assessed by an interdisciplinary team that generally in- appearance on cross-sectional imaging (Fig 6).1 For cludes a surgical oncologist, radiologist, medical and lesions demonstrating typical imaging findings of radiation oncologists, and a pathologist — preferably well-differentiated retroperitoneal liposarcoma, all of whom have expertise in soft-tissue sarcoma (see biopsy is typically not indicated given the high rate Fig 5). Cross-sectional imaging, including computed to- of diagnostic sensitivity of CT for low-grade adipo- mography (CT) or magnetic resonance imaging (MRI), cytic lesions and no role for neoadjuvant therapy preferably with intravenous contrast, can demonstrate with these tumors.1 predictable displacement of retroperitoneal structures, If imaging findings do not suggest well-differen- allowing for localization of a tumor to the retroperito- tiated liposarcoma, then image-guided core needle bi- neum rather than the peritoneal cavity. When a mass opsy or fine needle aspiration (FNA) is required and is detected in the retroperitoneum, the differential is preferred to open surgical biopsy.16 However, core diagnoses may include primary retroperitoneal soft- needle biopsy is preferred to FNA due to the difficulty tissue sarcomas, metastatic disease, including lymph- that exists in discerning histological subtype by FNA. adenopathy from primary sites of disease elsewhere, If results on biopsy suggest retroperitoneal sar- lymphoma, primary neoplasms arising from retroperi- coma, then surgery is the mainstay treatment for all toneal viscera such as the pancreas, kidneys, ureters, subtypes (see Fig 5). If the tumor is partially resectable, adrenal glands, duodenum, or ascending/descending then neoadjuvant radiotherapy and chemotherapy be- colon such as adenocarcinoma or gastrointestinal stro- come more important for treating the periphery of the mal tumor (GIST), paraspinal neurogenic tumors, and lesion because it will abut critical structures; both types retroperitoneal fibrosis. Clinical symptomatology, such of therapy have the intent of consolidating the peritu- as B symptoms in cases of lymphoma or findings on moral reactive zone and rendering close margins sterile. physical examination such as a palpable testicular ab- Neoadjuvant chemotherapy is also important for tumors normality, which may be associated with retroperito- with chemosensitive histologies.7,11 For the palliative neal lymphadenopathy, will direct diagnostic consid- management of metastatic or advanced soft-tissue sar- eration away from primary retroperitoneal sarcoma. coma, chemotherapy has an established role.6

Table 3. — Risk by Subtype for Local Regional Relapse, Metastasis, and OS Rates Subtype Risk for Locoregional Relapse42 OS Metastasis8 Leiomyosarcoma Piecemeal resection predictive Adjacent organ involvement associated Exceptions to typical pattern of of relapse with worst OS metastatic disease Associated with high risk of competitive Commonly metastasizes to liver, lung 42 systemic relapse High risk of metastatic disease14 Well-differentiated Piecemeal resection predictive Adjacent organ involvement associated Nearly all recur locoregionally, not via liposarcoma of relapse with worst OS metastasis14 Multifocality significantly associated with worse survival42 Dedifferentiated Surgeon specialization and periopera- Grade 3 associated with worst OS42 Nearly all recur locoregionally, not via liposarcoma tive radiotherapy associated with lower metastasis14 risk of relapse OS = overall survival.

July 2016, Vol. 23, No. 3 Cancer Control 255 Fig 5 summarizes relevant information regarding subtypes and perioperative therapy considerations.7,9,11,14,16 Myxoid/round cell liposarcoma as a primary malignancy is rare in the retroperitoneum (< 0.9% of all retroperitoneal sarcoma subtypes), and physical examination and imaging of the extremities to rule out a primary site in an extremity with distant retroperitoneal metastasis are indicated.6,43 Primitive neuroectodermal tumors, Ewing sarcoma, and alveolar/embryonal rhabdomyosarcoma are radiosensitive and chemosensitive, so they should not be considered for first-line surgery. Because primary chemotherapy is always given, the strategy for these tumors significantly differs from that undertaken for adult-type soft-tissue sarcoma. Subsequent surgery is then considered for responsive tumors to achieve complete resection of all visible tumors; however, adjacent viscera is only included if true invasion is evident.14 Our experience with GIST is that neoadjuvant tyrosine ki- Fig 6. — Axial contrast-enhanced computed tomography demonstrates nase inhibitor therapy for large lesions and abutting a large, round, adipocytic left retroperitoneal mass (dashed circle) that displaces the descending colon and compresses the left kidney with critical structures should be used to reduce tumor size minimal fat stranding and few thin septations consistent with well-dif- and limit the complexity of the procedure. Reduction ferentiated liposarcoma. of tumor size due to effective preoperative imatinib occurs in the majority of patients.44,45 In such a sce- and CDK4 are present in most cases and are in keeping nario, proper therapy is warranted for larger tumors with gene amplification.49 Molecular, cytogenetic, and that may require multivisceral resection. immunohistochemistry applications help distinguish well-differentiated liposarcoma from benign adipose Determination of Tumor Types With tumors. Well-differentiated liposarcoma is locally ag- Radiological and Histological Correlation gressive and lacks metastatic potential.47 Unless it be- Well-Differentiated and Dedifferentiated comes dedifferentiated (dedifferentiated liposarcoma), Liposarcoma well-differentiated liposarcoma is potentially curable CT or MRI can be used to recognize well-differentiated with complete excision when located in the extremities; liposarcoma (see Fig 6). Histomorphological findings however, sites like the retroperitoneum and mediasti- that support a diagnosis of well-differentiated liposar- num pose difficulty in obtaining a microscopically neg- coma include proliferation of mature adipocytes with ative margin (R0).46,51 The differential diagnosis of well- marked variation in cell size and a focal nuclear atypia differentiated liposarcoma from benign adipose tumors in both adipocytes and stromal cells. Identification of in a limited specimen, such as specimens from core lipoblasts does not make (or is required for) a diagno- needle biopsies, can be difficult. MDM2 fluorescence sis of liposarcoma.46 Lipoblasts may be seen in benign in situ hybridization is sensitive and specific to iden- lesions such as lipoblastoma, pleomorphic lipoma, and tify MDM2 amplification in morphologically atypical chondroid lipoma. and typical lesional cells.51 The immunohistochemistry Based on morphology, 4 main subtypes exist: adi- performance of MDM2 decreases when working with pocytic (lipoma-like), sclerosing, inflammatory, and limited specimens on biopsy, a rate typically attributed spindle cell. The sclerosing subtype of well-differen- to the focal staining seen within the majority of well- tiated liposarcoma tends to occur in the retroperito- differentiated liposarcoma on whole-tissue sections.51 neum and microscopically demonstrate scattered, MDM2 has been useful in identifying dedifferen- hyperchromatic, bizarre stromal cells set in extensive tiated liposarcomas in addition to well-differentiated fibrillary collagenous stroma. With the exception of the liposarcoma. Several studies have reported that most spindle-cell subtype, well-differentiated liposarcoma poorly differentiated sarcomas arising in the retroperi- has been found by molecular and cytogenetic studies toneum are, in fact, dedifferentiated liposarcoma and to harbor a characteristic ring and giant marker chro- can be diagnosed on the basis of MDM2 amplification mosomes containing amplification of the 12q13-15 re- even without an atypical adipocytic component.52-54 gion, including MDM2.47-50 Several other genes located Like well-differentiated liposarcoma, dedifferentiated in the 12q14-15 region, including CDK4 and HMGA2, liposarcoma harbors the ring or giant marker chromo- are frequently coamplified with MDM2.46 somes, whereas MDM2 (12q15) is consistently ampli- Immunohistochemically, expressions of MDM2 fied and overexpressed.46,49,55-58

256 Cancer Control July 2016, Vol. 23, No. 3 Histologically, dedifferentiated liposarcoma con- sarcoma constitute 0.9% and 0.4% of all retroperitoneal tains a well-differentiated liposarcoma component jux- soft-tissue sarcomas, respectively (see Table 1). Myxoid taposed to areas of either low- or high-grade nonlipo- liposarcoma demonstrates areas of myxomatous tis- genic sarcoma (dedifferentiation). The dedifferentiated sue that appear similar to fluid (low density on CT; component may exhibit a wide variety of morphology high signal intensity on T2-weighted MRI) prior to from high-grade sarcoma resembling undifferentiated pleomorphic sarcoma or high-grade myxofibrosar- A coma to low-grade dedifferentiation resembling fibromatosis or fibrosarcoma.59-62 Historically, many dedifferentiated liposarcomas were diagnosed as malignant fibrous histiocytoma, particularly de novo dedifferentiated liposarcomas.54 Dedifferentiated liposarcoma commonly occurs in the retroperitoneum, with more than 90% arising de novo, whereas less than 10% represent recurrences.46,58 Radiologically, the presence of a nonlipogenic soft-tissue mass within an otherwise fatty tumor would be consistent with dedifferentiation (Fig 7). Given the internal heterogeneity, it is important to correlate the histopathological findings with radiological imaging because results on tissue sampling may yield low-grade lipomatous tissue or high-grade sarcoma depending on the target of the biopsy. Surgically, the B dedifferentiated part of the tumor is regarded as the only part needed to be excised, whereas surrounding well-differentiated liposarcoma is regarded as normal tissue. Therefore, dedifferentiated liposarcoma is rarely excised with margins free of well-differentiated liposarcoma. Dedifferentiation is associated with a 15% to 20% metastatic rate; however, the mortality rate is more often related to uncontrolled local recurrences than to metastatic spread.50 A case illustration of dedifferentiated liposarcoma appears in Fig 8.

Other Subtypes of Liposarcoma Myxoid/round cell liposarcoma and pleomorphic lipo-

Fig 7. — Axial contrast-enhanced computed tomography demonstrates Fig 8A–C. — Dedifferentiated liposarcoma. (A) Nonlipogenic malignancy is a large infiltrative retroperitoneal mass encasing the right kidney (K), revealed on core biopsy. Hematoxylin and eosin stain, × 400. (B) Nuclear- displacing the ascending colon (C) with lipogenic elements (large arrow), positive MDM2. Immunohistochemistry, × 400. (C) Amplification of MDM2 thick septations (small arrow), and nonadipocytic soft-tissue masses (*) (red signals). Green signals are chromosome 17 and confirm dedifferenti- anterior to the inferior vena cava and aorta (A). ated liposarcoma. Fluorescent in situ hybridization, × 1000.

July 2016, Vol. 23, No. 3 Cancer Control 257 administering intravenous contrast. However, myx- Retroperitoneal leiomyosarcoma shows a high pro- oid elements will be enhanced as opposed to non- pensity of occurrence in women. When arising from enhanced fluid in areas of cystic degeneration or large blood vessels, it commonly occurs from the in- tumor necrosis. Occasionally, liposarcoma may dem- ferior vena cava and its major tributaries.65 Grossly, onstrate areas of mineralization or ossification. More leiomyosarcomas may exhibit either a grey-to-white than 95% of cases of myxoid liposarcoma demon- whorled appearance or form fleshy, tan-white masses strate a classical t(12;16)(q13;p11) or t(12;22)(q13;q12) with hemorrhage, necrosis, and/or cystic change in- translocation, which results in fusion of FUS-CHOP distinguishable from other sarcomas.66 Typical his- or EWSR1-CHOP, respectively.46 A diagnosis of pri- tomorphology consists of intersecting fascicles of mary retroperitoneal myxoid liposarcoma should spindle cells with elongated and blunt-ended nuclei made with caution, because such cases represent and eosinophilic cytoplasm. Nuclear hyperchroma- either metastatic myxoid liposarcoma or well-differ- sia and pleomorphism (ranging from mild to severe) entiated/dedifferentiated liposarcoma with myxoid can be observed. Pleomorphism in leiomyosarcoma stromal change.43,63 Round cell liposarcoma is not an may resemble undifferentiated pleomorphic sarco- independent subtype but rather a high-grade spec- ma.66 Mitoses, including atypical mitotic figures, are trum of myxoid liposarcoma. Round cell liposarcoma usually present. Occasionally, retroperitoneal leio- shares the same molecular change as myxoid liposar- myosarcoma may have areas exhibiting epithelioid coma but is morphologically composed of blue round cytomorphology, multinucleated giant cells, promi- tumor cells. nent inflammatory infiltrate, or exuberant myxoid Pleomorphic liposarcoma often presents as a non- change.67-70 Rarely, leiomyosarcoma contains granular specific soft-tissue mass without obvious fat evident by cytoplasmic change.71 Leiomyosarcoma, specifically results on imaging mimicking other high-grade retro- of soft-tissue origin, is considered malignant when peritoneal sarcomas. Pleomorphic liposarcoma exhib- some degree of nuclear atypia and mitotic activity, its pleomorphic lipoblasts without well differentiation which may be very low (< 1/10 hpf), are present.66 An- or other types of liposarcoma.46 This is a rare type of tibodies to smooth-muscle actin, desmin, and caldes- high-grade liposarcoma that can be differentiated from mon are positive in most types of leiomyosarcoma.72-74 dedifferentiated liposarcoma by the absence of MDM2 Smooth-muscle, myosin heavy chain is less sensitive amplification. Pleomorphic liposarcoma also has a dif- and may be expressed in myoepithelial cells.75-77 Ab- ferent clinical metastatic rate than other pleomorphic errant cytokeratin is frequent and epithelial mem- or high-grade sarcoma types (Table 4).46 brane antigen expression may also be seen.78-81 S100, CD34, estrogen, and progesterone receptors may also Retroperitoneal Leiomyosarcoma be positive in leiomyosarcoma.46,82 Hormone-receptor Retroperitoneal leiomyosarcomas typically arise expression can be seen in leiomyosarcoma of a uter- from smooth muscle within the soft tissues of the ret- ine origin but is not specific for leiomyosarcoma of a roperitoneum itself or within the walls of large ret- gynecological origin.66 roperitoneal vessels such as the inferior vena cava, aorta, or gonadal veins.64 Radiographically, the most Undifferentiated Pleomorphic Sarcoma helpful diagnostic clue may be identification of a Undifferentiated pleomorphic sarcoma was previ- mass associated with one of these vascular struc- ously referred to as malignant fibrous histiocytoma. tures. Typically, retroperitoneal leiomyosarcomas are It has no identifiable line of differentiation when hypervascular solid tumors with minimal necrosis analyzed by current technologies and represents a and a notable absence of internal fat compared with diagnosis of exclusion.66 Undifferentiated pleomor- liposarcomas. Mineralization is rarely seen. Com- phic sarcoma typically presents as a large, nonspe- pression of the underlying vessel may result in ve- cific, heterogeneous, enhancing soft-tissue mass with nous thrombosis, and intraluminal extension of dis- internal necrosis and intratumoral hemorrhage, of- ease is not uncommon. ten with lobulated morphology and well-defined margins due to the presence of a pseudocapsule. Table 4. — Sarcoma Types and Clinical Metastatic Rates Occasionally, internal fibrous tissue may be identified as areas of low signal intensity on T1- and Type Incidence, % T2-weighted MRI (Fig 9).64 Myxoid elements may be Dedifferentiated liposarcoma 15–20 seen as regions of increased T2-weighted signal with High-grade myxofibrosarcoma 25–30 enhanced intravenous contrast. Tumor heterogene- Pleomorphic liposarcoma 30–50 ity, including hemorrhage, is typical of undifferenti- Pleomorphic leiomyosarcoma 70 ated pleomorphic sarcoma. Given the propensity for Pleomorphic rhabdomyosarcoma 90 hemorrhage, underlying neoplasms, including undifferentiated pleomorphic sarcoma, must be consid-

258 Cancer Control July 2016, Vol. 23, No. 3 of undifferentiated pleomorphic sarcoma with prominent myxoid stroma and features such as multinodular growth pattern and prominent curvilinear vasculature, even focally, should be diagnosed as high-grade myxofibrosarcoma.66 Immunohistochemistry serves as a method to exclude other pleomorphic tumors. Undifferentiated pleomorphic sarcoma often shows patchy to rare cells positive for cytokeratin, actin, desmin, or epithelial membrane antigen. Vimentin and CD34 may be positive but are of no diagnostic value.46 Over the years, complex cytogenetic aberrations have been identified in undifferentiated pleomorphic sarcoma, but they are nonspecific.84,85 Similarities between undifferentiated pleomorphic sarcoma and leiomyosarcoma have been reported from comparative genomic hybridization Fig 9. — Axial contrast-enhanced computed tomography demonstrates analyses, suggesting a possible shared lineage.86,87 Fur- a heterogeneous nonadipocytic, largely necrotic upper retroperitoneal 87 mass along the aorta (stent graft in place) and pulmonary metastases. thermore, Carneiro et al reported that losses of 4q13 Although nonspecific, imaging characteristics would favor undiffer- (encompassing SMAD1) and 18q22 were independent entiated pleomorphic sarcoma rather than liposarcoma or leiomyo- predictors of metastasis. sarcoma. Neurogenic Tumors ered in any patient presenting with spontaneous he- Neurogenic tumors, including nerve sheath tumors matoma. When arising in the retroperitoneum, these (both benign and malignant), are more common in tumors tend to be larger than when they occur in the the extremities, but they have been known to occur extremities. Adjacent bone invasion is more common in the retroperitoneum, often in paraspinal or pre- than in liposarcoma or leiomyosarcoma and may help sacral locations. Neurogenic tumors tend to demon- suggest the diagnosis. Undifferentiated pleomorphic strate a fusiform shape, increased T2-weighted sig- sarcoma represents the most common soft-tissue nal intensity, and intravenous contrast enhancement. sarcoma following prior radiotherapy. Bony erosions and scalloping, including widening of When encountering a retroperitoneal soft-tissue the neural foramina, are commonly associated osse- tumor with the appearance of undifferentiated pleo- ous findings. Differentiation of benign and malignant morphic sarcoma, pathologists must exclude possible neoplasms can be difficult, but increased size, rapid pleomorphic sarcoma of a specific type (eg, pleomor- growth, internal necrosis, and increased vascularity phic leiomyosarcoma, pleomorphic liposarcoma), a de- all favor malignancy. Malignant neurogenic tumors differentiated component of another type of sarcoma also tend to demonstrate greater fluorodeoxyglucose (eg, dedifferentiated liposarcoma), and other sarcoma- uptake than benign tumors on positron emission to- toid carcinomas. Rarely, sarcomatoid mesothelioma, mography/CT. Malignant tumors are often seen in the dedifferentiated melanoma, and anaplastic lymphoma setting of type 1 neurofibromatosis.64 can also occur.66 Ganglioneuroma is a rare, benign, differentiat- The histology of undifferentiated pleomorphic sar- ed neoplasm of the sympathetic nervous system that coma is variable and may reveal several morphologi- contains no immature neuroblastic elements.66,88 Gan- cal patterns, from storiform areas composed of spindle glioneuromas predominantly arise within the poste- cells to pleomorphic areas composed of large, round- rior mediastinum and retroperitoneum.66,89 They are ed, fibroblastic-like cells with marked nuclear atypia well-circumscribed tumors with a fibrous capsule, and and bizarre, multinucleated-tumor giant cells.46,66 Mi- cut sections are gray to yellow with a whorled-like pat- totic activity is prominent with atypical mitotic fig- tern similar to leiomyoma.66 Histologically, this tumor ures. The background stroma is usually collagenous, consists of Schwann cells with scattered deposits of gan- but, rarely, the stroma may have metaplastic osteoid glion cells, isolated, or small clusters. Surgical excision is or chondroid material. Some cases of undifferentiated appropriate. Ganglioneuromas rarely recur. Malignant pleomorphic sarcoma have prominent background transformation into MPNST has been reported.90,91 xanthomatous and neutrophilic infiltrates.83 Schwannoma, a peripheral nerve sheath tumor Prominent stromal myxoid change may be pres- consisting of well-differentiated Schwann cells, is usu- ent in undifferentiated pleomorphic sarcoma, either ally encapsulated, and its cut surfaces have a pink, focally within the tumor or as large areas abutting cel- white, or yellow appearance.66 Retroperitoneal tumors lular zones. Tumors with cytomorphology indicative are large and may have areas of cystic degeneration

July 2016, Vol. 23, No. 3 Cancer Control 259 and calcification. Classic histology shows a pattern ated with myxoid matrix. They can resemble melano- of alternating Antoni A (cellular areas of spindle cells ma or carcinoma. Immunohistochemically, epitheli- with occasional palisading) and Antoni B (loose myxoid MPNSTs show strong and diffuse S100 positivity, oid areas with scattered spindle cells and thick-walled, one-half of cases lack SMARCB1 staining, and rare hyalinized vessels) areas. cases show keratin positivity.110,111 MPNST may also Retroperitoneal schwannomas may be exclusive- have rhabdomyoblastic differentiation (malignant ly or predominantly composed of Antoni A tissue.92,93 triton tumor), which has a worse prognosis than con- Schwannomas with increased cellularity and occasion- ventional MPNST.112 al mitoses are referred to as cellular schwannomas, a Cellular schwannoma may be misdiagnosed as variant of schwannoma.92-96 Degenerative changes such MPNST due to it hypercellularity. Significant institu- as cyst formation, calcification, hemorrhage, and hya- tional data do not exist to address how to differentiate linization may be present in retroperitoneal schwanno- MPNST from cellular schwannoma in the retroperito- mas, especially if the tumor has been present for a long neum. One large study suggested that certain features time. Marked nuclear atypia characterized by Schwann distinguish cellular schwannoma from MPNST, but cells with large, hyperchromatic nuclei are usually as- only in general and not specifically to the retroperito- sociated with schwannomas with degenerative change neum (Table 5).113 (ancient schwannoma). They behave similar to conventional schwannomas. S100 is strongly and diffuse- Solitary Fibrous Tumor/Hemangiopericytoma ly expressed in schwannomas.97 SOX10, a marker of Solitary fibrous tumor is a mesenchymal tumor of fi- neural crest differentiation, exhibits nuclear staining broblastic origin that occurs in deep soft tissue such in schwannomas.98 Of note, retroperitoneal schwanno- as the thigh, pelvis, retroperitoneum, and serosal mas may express cytokeratin AE1/3 due to cross-reac- surfaces.114 Solitary fibrous tumor typically demon- tivity with GFAP.46 strates marked enhancement when viewed with in- MPNST is an aggressive sarcoma arising from a travenous contrast and presents as a solid, enhancing peripheral nerve (eg, sciatic nerve, brachial plexus, soft-tissue mass with prominent tortuous and ser- sacral plexus) or preexisting benign nerve sheath tu- pentine vessels extending to and seen within the pe- mor (eg, neurofibroma). Nearly 50% of MPNSTs occur riphery of the mass. Central necrosis is common. The in patients with type 1 neurofibromatosis and the re- cut surface of the tumor is gray-white to red-brown mainder sporadically occurs.46,66 Retroperitoneal in- in color and hemorrhage or cystic degeneration may volvement is rare.99,100 Grossly, MPNSTs arising from a be seen.66 Histologically, solitary fibrous tumors nerve form a large fusiform mass and often measure more than 5 cm and have a tan-white, fleshy cut sur- 46 face with areas of hemorrhage and necrosis. Tumor Table 5. — Distinguishing Features Between Cellular histomorphology is diverse. Classic cases of MPNST Schwannoma and Malignant Peripheral Nerve Sheath Tumora exhibit spindle cells arranged in densely cellular fas- Cellular Schwannoma Malignant Peripheral Nerve cicles alternating with less cellular, myxoid areas, creat- Sheath Tumor 66 ing a marble-like effect. The cells can have a whorling Benign Malignant or, rarely, palisading architecture with large areas of Free of metastasis necrosis. The spindle cells have hyperchromatic nuclei Disease-specific–related deaths and pale cytoplasm. Mitoses are readily seen. Occa- Schwannian whorls Perivascular hypercellularity sionally, MPNSTs demonstrate marked pleomorphism, Peritumoral capsule Tumor herniation into vascular simulating high-grade, undifferentiated pleomorphic Subcapsular lymphocytes lumens sarcoma. Skeletal muscle differentiation, glandular dif- Macrophage-rich infiltrates Necrosis ferentiation, and heterologous elements have been re- Fascicles absent ported in MPNSTs.101,102 Between 50% and 90% of cases Expression of p75NTR Expression of p75NTR observed in of MPNST are positive for S100, which demonstrates observed in 31% of cellular 80% of tumors* focal staining, and 2% to 15% show weak expression schwannomas* 97,103-108 of TLE1. SOX10 has recently been reported to Complete loss of SOX10* show better sensitivity and specificity for the diagno- Neurofibromin or p16 expression* 109 sis of MPNST than S100. A rare variant, epithelioid Presence of EGFR immunoreactivity MPNST, is not associated with type 1 neurofibromato- is specific* sis and commonly arises from preexisting schwanno- Ki-67 labeling indices > 20% ma.46 Histologically, epithelioid MPNST is composed highly predictive (87% sensitivity, of short cords of large epithelioid cells arranged in a 96% specificity) vague, nodular pattern.66 The cells have large, round aIn general, not specific to the retroperitoneum. *P < .001. nuclei with prominent nucleoli. They may be associ-

260 Cancer Control July 2016, Vol. 23, No. 3 can demonstrate a variable appearance, from being well-differentiated liposarcoma on imaging and highly cellular to densely hyalinized and hypocellu- findings on biopsies are often inconclusive. Molecu- lar tumors. Cellular, solitary fibrous tumors show a lar testing is recommended to support a diagnosis patternless architecture composed of tightly packed, of retroperitoneal lipoma confirming the absence bland-appearing, spindle- to fusiform-shaped cells of MDM2 amplification; however, a negative result with indistinct cytoplasmic borders arranged around does not exclude the possibility of well-differentiat- prominent dilated, branched vessels. Myxoid change ed liposarcoma. Amplification varies in individual is common, as is stromal and perivascular hyaliniza- tumors and among different cells in the same tumor. tion. Mitoses are scarce. Additional research is necessary to understand the A solitary fibrous tumor variant, the fat-forming etiology and genetic mechanisms of retroperitoneal solitary fibrous tumor, often affects the thigh and lipomas. Close and regular follow-up is recommend- retroperitoneum.46 It contains a variable amount ed for such cases. of mature adipocytes that should not be confused with well-differentiated liposarcoma. Solitary fi- Conclusions brous tumors can be malignant, and these cases When a definitive diagnosis of retroperitoneal sarco- demonstrate dense cellularity, increased mitoses ma, including its histological tumor type and grade, (> 4 mitoses/10 hpf), variable cytological atypia, tu- is achievable, then such a diagnosis can help pro- mor necrosis, and/or infiltrative margins.115,116 Tu- vide useful prognostic and predictive information to mor cells are typically positive for CD34 (80%–90%), help guide effective therapy. However, not all tumors CD99 (70%), BCL2 (30%), epithelial membrane an- of the retroperitoneum must be biopsied, subjected tigen (30%), and actin (20%).117-123 Desmin, cyto- to work up by ancillary testing, or both methods to keratin, and S100 are usually absent.117 Recently, achieve a definitive diagnosis. Certain tumor types NAB2/STAT6 was identified in solitary fibrous tu- may call for specific neoadjuvant therapeutic regi- mors, strongly suggesting that this type of tumor is mens that might produce excellent responses, thus a translocation-associated neoplasm and that STAT6 warranting a definitive diagnosis with judicious ancil- immunostain is a sensitive and specific marker for lary testing. By contrast, other tumor types may share solitary fibrous tumor.124,125 similar therapeutic regimens, but neoadjuvant therapy may not provide any benefit to the patient. In such Other Rare Considerations cases, an informative diagnosis should be rendered GIST in the retroperitoneum is rare: One case report as an alternative, which includes meaningful and use- has been documented in the literature.126 Characteris- ful information to guide the next best step in man- tic immunostain pattern of CD117, CD34, and DOG1 agement, especially when there is a small amount of immunoreactivity is helpful in confirming the diag- tissue on biopsy available for analysis. A definitive nosis.46 Ewing sarcoma in the retroperitoneum is also diagnosis can be deferred and then be rendered at rare. Hallmark translation (EWSR1/FLI1 or EWSR1/ resection following a full examination of the entire ERG) detection is helpful in confirming Ewing sarco- tumor. This type of multidisciplinary, pathology-fo- ma in the majority of cases.46 cused approach is practical and, in our experience, Retroperitoneal desmoid tumor has been de- works well to serve the needs of patients with retro- scribed in 11 articles since 1991.127-137 These tumors peritoneal sarcoma. are characterized by proliferation of spindle (fibro- Acknowledgment: Total Cancer Care was enabled blast) cells, with a moderate amount of collagen fi- in part by the support of the DeBartolo Family, and we bers, bland cellular appearance, scant mitosis, and thank the many patients who provided data and tissue lack of metastasis.130,135 Expression of beta catenin in for Total Cancer Care. We also thank Ambuj Kumar, tumor cells is helpful for confirming the diagnosis.46 MD, of USF Health for his assistance with statistical Most sporadic cases of aggressive fibromatosis con- analysis and the continuous commitment to medical tain a somatic mutation in either APC or CTNNB1.128 education by the staff of the Anatomic Pathology Ed- Lazar et al138 reported 3 discrete mutations (ACC- ucation Program. We also appreciate Julia A. Bridge, 41GCC, TCT45TTT, and TCT45CCT) in 2 codons of MD, from the Nebraska Medical Center for performing CTNNB1 exon 3. Targeted therapy for desmoid tumor/ the molecular testing. fibromatosis may be a potential treatment option in the future.135 References Retroperitoneal lipoma is a rare benign tumor of 1. Mullinax JE, Zager JS, Gonzalez RJ. Current diagnosis and management of retroperitoneal sarcoma. Cancer Control. 2011;18(3):177-187. mature adipocytes occurring in the retroperitoneum. 2. Mastrangelo G, Coindre JM, Ducimetiere F, et al. Incidence of soft- Eighteen cases of retroperitoneal lipoma have been tissue sarcoma and beyond: a population-based prospective study in 3 European 139-155 regions. Cancer. 2012;118(21):5339-5348. described in the literature since 1980. Retro- 3. Ducimetiere F, Lurkin A, Ranchere-Vince D, et al. Incidence of sarcoma peritoneal lipoma may not be distinguishable from histotypes and molecular subtypes in a prospective epidemiological study with

July 2016, Vol. 23, No. 3 Cancer Control 261 central pathology review and molecular testing. PLoS One. 2011;6(8):e20294. 29. Hoekstra HJ, Thijssens K, van Ginkel RJ. Role of surgery as pri- 4. Liles JS, Tzeng CW, Short JJ, et al. Retroperitoneal and intra-abdominal mary treatment and as intervention in the multidisciplinary treatment of soft sarcoma. Curr Probl Surg. 2009;46(6):445-503. tissue sarcoma. Ann Oncol. 2004;15(suppl 4):iv181-iv186. 5. Dodd LG, Bui MM. Atlas of Soft Tissue and Bone Pathology: With Histo- 30. Karavasilis V, Seddon BM, Ashley S, et al. Significant clinical ben- logic, Cytologic, and Radiologic Correlations. New York: Demos Medical; 2015. efit of first-line palliative chemotherapy in advanced soft-tissue sarcoma: 6. Matthyssens LE, Creytens D, Ceelen WP. Retroperitoneal liposarcoma: retrospective analysis and identification of prognostic factors in 488 pa- current insights in diagnosis and treatment. Front Surg. 2015;2:4. tients. Cancer. 2008;112(7):1585-1591. 7. Management of primary retroperitoneal sarcoma (RPS) in the adult: 31. Neuville A, Chibon F, Coindre JM. Grading of soft-tissue sarcomas: a consensus approach from the Trans-Atlantic RPS Working Group. Ann from histological to molecular assessment. Pathology. 2014;46(2):113-120. Surg Oncol. 2015;22(1):256-263. 32. Hoffman A, Lazar AJ, Pollock RE, et al. New frontiers in the treat- 8. Italiano A, Delva F, Mathoulin-Pelissier S, et al. Effect of adjuvant ment of liposarcoma, a therapeutically resistant malignant cohort. Drug chemotherapy on survival in FNCLCC grade 3 soft-tissue sarcomas: a Resist Updat. 2011;14(1):52-66. multivariate analysis of the French Sarcoma Group Database. Ann Oncol. 33. Windham TC, Pisters PW. Retroperitoneal sarcomas. Cancer Control. 2010;21(12):2436-2441. 2005;12(1):36-43. 9. Jones RL, Fisher C, Al-Muderis O, et al. Differential sensitivity of lipo- 34. Payne WG, Naidu DK, Wheeler CK, et al. Wound healing in patients sarcoma subtypes to chemotherapy. Eur J Cancer. 2005;41(18):2853-2860. with cancer. Eplasty. 2008;8:e9. 10. Sleijfer S, Ouali M, van Glabbeke M, et al. Prognostic and predic- 35. Thomas TO, Hasan S, Small W, et al. The tolerance of gastrointesti- tive factors for outcome to first-line ifosfamide-containing chemotherapy nal organs to stereotactic body radiation therapy: what do we know so far? for adult patients with advanced soft-tissue sarcomas: an exploratory, ret- J Gastrointest Oncol. 2014;5(3):236-246. rospective analysis on large series from the European Organization for 36. Gronchi A, Colombo C, Raut CP. Surgical management of localized Research and Treatment of Cancer-Soft Tissue and Bone Sarcoma Group soft tissue tumors. Cancer. 2014;120(17):2638-2648. (EORTC-STBSG). Eur J Cancer. 2010;46(1):72-83. 37. Cormier JN, Pollock RE. Soft tissue sarcomas. CA Cancer J Clin. 11. Seddon B, Scurr M, Jones RL, et al. A phase II trial to assess the 2004;54(2):94-109. activity of gemcitabine and docetaxel as first line chemotherapy treat- 38. Perez EA, Gutierrez JC, Moffat FL Jr, et al. Retroperitoneal and ment in patients with unresectable leiomyosarcoma. Clin Sarcoma Res. truncal sarcomas: prognosis depends upon type not location. Ann Surg 2015;5:13. Oncol. 2007;14(3):1114-1122. 12. Van Glabbeke M, van Oosterom AT, Oosterhuis JW, et al. Prognos- 39. Gronchi A, Miceli R, Shurell E, et al. Outcome prediction in primary tic factors for the outcome of chemotherapy in advanced soft-tissue sar- resected retroperitoneal soft-tissue sarcoma: histology-specific overall coma: an analysis of 2,185 patients treated with anthracycline-containing survival and disease-free survival nomograms built on major sarcoma first-line regimens--a European Organization for Research and Treatment center data sets. J Clin Oncol. 2013;31(13):1649-1655. of Cancer Soft Tissue and Bone Sarcoma Group Study. J Clin Oncol. 40. Stojadinovic A, Yeh A, Brennan MF. Completely resected recurrent 1999;17(1):150-157. soft-tissue sarcoma: primary anatomic site governs outcomes. J Am Coll 13. Kroep JR, Ouali M, Gelderblom H, et al. First-line chemotherapy for Surg. 2002;194(4):436-447. malignant peripheral nerve sheath tumor (MPNST) versus other histologi- 41. Anaya DA, Lev DC, Pollock RE. The role of surgical margin status cal soft-tissue sarcoma subtypes and as a prognostic factor for MPNST: in retroperitoneal sarcoma. J Surg Oncol. 2008;98(8):607-610. an EORTC Soft Tissue and Bone Sarcoma Group study. Ann Oncol. 42. Toulmonde M, Bonvalot S, Meeus P, et al. Retroperitoneal sarco- 2011;22(1):207-214. mas: patterns of care at diagnosis, prognostic factors and focus on main 14. Thomas DM, O’Sullivan B, Gronchi A. Current concepts and future histological subtypes: a multicenter analysis of the French Sarcoma perspectives in retroperitoneal soft-tissue sarcoma management. Expert Group. Ann Oncol. 2014;25(3):735-742. Rev Anticancer Ther. 2009;9(8):1145-1157. 43. de Vreeze RS, de Jong D, Tielen IH, et al. Primary retroperitoneal myx- 15. Saynak M, Bayir-Angin G, Kocak Z, et al. Recurrent solitary fibrous oid/round cell liposarcoma is a nonexisting disease: an immunohistochemical tumor of the pleura: significant response to radiotherapy. Med Oncol. and molecular biological analysis. Mod Pathol. 2009;22(2):223-231. 2010;27(1):45-48. 44. Zaydfudim V, Okuno SH, Que FG, et al. Role of operative therapy 16. National Comprehensive Cancer Network. NCCN Clinical Practice in treatment of metastatic gastrointestinal stromal tumors. J Surg Res. Guidelines in Oncology: Soft Tissue Sarcoma. V2.2016. http://www.nccn. 2012;177(2):248-254. org/professionals/physician_gls/pdf/sarcoma.pdf. Accessed April 20, 2016. 45. Fiore M, Palassini E, Fumagalli E, et al. Preoperative imatinib me- 17. Kneisl JS, Coleman MM, Raut CP. Outcomes in the management of sylate for unresectable or locally advanced primary gastrointestinal stromal adult soft-tissue sarcomas. J Surg Oncol. 2014;110(5):527-538. tumors (GIST). Eur J Surg Oncol. 2009;35(7):739-745. 18. Tuan J, Vitolo V, Vischioni B, et al. Radiation therapy for retroperito- 46. Fletcher CD; World Health Organization; International Agency for neal sarcoma. Radiol Med. 2014;119(10):790-802. Research on Cancer. WHO Classification of Tumours of Soft Tissue and 19. Stoeckle E, Coindre JM, Bonvalot S, et al. Prognostic factors in Bone. 4th ed. Lyon: IARC Press; 2013. retroperitoneal sarcoma: a multivariate analysis of a series of 165 pa- 47. Dal Cin P, Kools P, Sciot R, et al. Cytogenetic and fluorescence in situ tients of the French Cancer Center Federation Sarcoma Group. Cancer. hybridization investigation of ring chromosomes characterizing a specific 2001;92(2):359-368. pathologic subgroup of adipose tissue tumors. Cancer Genet Cytogenet. 20. Meric F, Hess KR, Varma DG, et al. Radiographic response to neo- 1993;68(2):85-90. adjuvant chemotherapy is a predictor of local control and survival in soft- 48. Karakousis CP, Dal Cin P, Turc-Carel C, et al. Chromosomal tissue sarcomas. Cancer. 2002;95(5):1120-1126. changes in soft-tissue sarcomas. A new diagnostic parameter. Arch Surg. 21. Wendtner CM, Abdel-Rahman S, Krych M, et al. Response to neo- 1987;122(11):1257-1260. adjuvant chemotherapy combined with regional hyperthermia predicts 49. Pedeutour F, Suijkerbuijk RF, Forus A, et al. Complex composition long-term survival for adult patients with retroperitoneal and visceral high- and co-amplification of SAS and MDM2 in ring and giant rod marker chro- risk soft-tissue sarcomas. J Clin Oncol. 2002;20(14):3156-3164. mosomes in well-differentiated liposarcoma. Genes Chromosomes Cancer. 22. Meric F, Milas M, Hunt KK, et al. Impact of neoadjuvant chemo- 1994;10(2):85-94. therapy on postoperative morbidity in soft-tissue sarcomas. J Clin Oncol. 50. Ikoma N, Torres KE, Somaiah N, et al. Accuracy of preoperative 2000;18(19):3378-3383. percutaneous biopsy for the diagnosis of retroperitoneal liposarcoma sub- 23. Donahue TR, Kattan MW, Nelson SD, et al. Evaluation of neo- types. Ann Surg Oncol. 2015;22(4):1068-1072. adjuvant therapy and histopathologic response in primary, high-grade 51. Weaver J, Rao P, Goldblum JR, et al. Can MDM2 analytical tests retroperitoneal sarcomas using the sarcoma nomogram. Cancer. performed on core needle biopsy be relied upon to diagnose well-differen- 2010;116(16):3883-3891. tiated liposarcoma? Mod Pathol. 2010;23(10):1301-1306. 24. Judson I, Verweij J, Gelderblom H, et al. Doxorubicin alone versus 52. Chibon F, Mariani O, Derre J, et al. A subgroup of malignant fibrous intensified doxorubicin plus ifosfamide for first-line treatment of advanced histiocytomas is associated with genetic changes similar to those of well- or metastatic soft-tissue sarcoma: a randomised controlled phase 3 trial. differentiated liposarcomas. Cancer Genet Cytogenet. 2002;139(1):24-29. Lancet Oncol. 2014;15(4):415-423. 53. Coindre JM, Hostein I, Maire G, et al. Inflammatory malignant 25. Krikelis D, Judson I. Role of chemotherapy in the management of fibrous histiocytomas and dedifferentiated liposarcomas: histological review, soft-tissue sarcomas. Expert Rev Anticancer Ther. 2010;10(2):249-260. genomic profile, and MDM2 and CDK4 status favour a single entity. J Pathol. 26. Young RJ, Natukunda A, Litiere S, et al. First-line anthracycline- 2004;203(3):822-830. based chemotherapy for angiosarcoma and other soft-tissue sarcoma 54. Coindre JM, Mariani O, Chibon F, et al. Most malignant fibrous subtypes: pooled analysis of eleven European Organisation for Research histiocytomas developed in the retroperitoneum are dedifferentiated lipo- and Treatment of Cancer Soft Tissue and Bone Sarcoma Group trials. Eur J sarcomas: a review of 25 cases initially diagnosed as malignant fibrous Cancer. 2014;50(18):3178-3186. histiocytoma. Mod Pathol. 2003;16(3):256-262. 27. Constantinidou A, Jones RL, Olmos D, et al. Conventional anthra- 55. Dei Tos AP, Doglioni C, Piccinin S, et al. Coordinated expression cycline-based chemotherapy has limited efficacy in solitary fibrous tumour. and amplification of the MDM2, CDK4, and HMGI-C genes in atypical lipo- Acta Oncol. 2012;51(4):550-554. matous tumours. J Pathol. 2000;190(5):531-536. 28. Constantinidou A, Pollack S, Loggers E, et al. The evolution of sys- 56. Gisselsson D, Hoglund M, Mertens F, et al. Chromosomal organiza- temic therapy in sarcoma. Expert Rev Anticancer Ther. 2013;13(2):211-223. tion of amplified chromosome 12 sequences in mesenchymal tumors de-

262 Cancer Control July 2016, Vol. 23, No. 3 tected by fluorescence in situ hybridization. Genes Chromosomes Cancer. 84. Gazziola C, Cordani N, Wasserman B, et al. Malignant fibrous his- 1998;23(3):203-212. tiocytoma: a proposed cellular origin and identification of its characterizing 57. Pedeutour F, Forus A, Coindre JM, et al. Structure of the supernu- gene transcripts. Int J Oncol. 2003;23(2):343-351. merary ring and giant rod chromosomes in adipose tissue tumors. Genes 85. Al-Agha OM, Igbokwe AA. Malignant fibrous histiocytoma: between Chromosomes Cancer. 1999;24(1):30-41. the past and the present. Arch Pathol Lab Med. 2008;132(6):1030-1035. 58. Coindre JM, Pedeutour F, Aurias A. Well-differentiated and dedif- 86. Larramendy ML, Gentile M, Soloneski S, et al. Does compara- ferentiated liposarcomas. Virchows Arch. 2010;456(2):167-179. tive genomic hybridization reveal distinct differences in DNA copy number 59. McCormick D, Mentzel T, Beham A, et al. Dedifferentiated lipo- sequence patterns between leiomyosarcoma and malignant fibrous histiocy- sarcoma. Clinicopathologic analysis of 32 cases suggesting a better toma? Cancer Genet Cytogenet. 2008;187(1):1-11. prognostic subgroup among pleomorphic sarcomas. Am J Surg Pathol. 87. Carneiro A, Francis P, Bendahl PO, et al. Indistinguishable genomic 1994;18(12):1213-1223. profiles and shared prognostic markers in undifferentiated pleomorphic 60. Weiss SW, Rao VK. Well-differentiated liposarcoma (atypical li- sarcoma and leiomyosarcoma: different sides of a single coin? Lab Invest. poma) of deep soft tissue of the extremities, retroperitoneum, and miscel- 2009;89(6):668-675. laneous sites. A follow-up study of 92 cases with analysis of the incidence 88. Zare S, Parvin M, Ghohestani SM. Retroperitoneal ganglioneuroma. of “dedifferentiation.” Am J Surg Pathol. 1992;16(11):1051-1058. Urol J. 2008;5(4):232. 61. Elgar F, Goldblum JR. Well-differentiated liposarcoma of the ret- 89. Jain M, Shubha BS, Sethi S, et al. Retroperitoneal ganglioneuroma: roperitoneum: a clinicopathologic analysis of 20 cases, with particu- report of a case diagnosed by fine-needle aspiration cytology, with review lar attention to the extent of low-grade dedifferentiation. Mod Pathol. of the literature. Diagn Cytopathol. 1999;21(3):194-196. 1997;10(2):113-120. 90. Drago G, Pasquier B, Pasquier D, et al. Malignant peripheral nerve 62. Henricks WH, Chu YC, Goldblum JR, et al. Dedifferentiated liposar- sheath tumor arising in a “de novo” ganglioneuroma: a case report and coma: a clinicopathological analysis of 155 cases with a proposal for an ex- review of the literature. Med Pediatr Oncol. 1997;28(3):216-222. panded definition of dedifferentiation. Am J Surg Pathol. 1997;21(3):271-281. 91. de Chadarevian JP, MaePascasio J, Halligan GE, et al. Malignant 63. Sioletic S, Dal Cin P, Fletcher CD, et al. Well-differentiated and dedif- peripheral nerve sheath tumor arising from an adrenal ganglioneuroma in ferentiated liposarcomas with prominent myxoid stroma: analysis of 56 cases. a 6-year-old boy. Pediatr Dev Pathol. 2004;7(3):277-284. Histopathology. 2013;62(2):287-293. 92. Fletcher CD, Davies SE, McKee PH. Cellular schwannoma: a dis- 64. Kransdorf MJ, Murphey MD. Imaging of Soft Tissue Tumors. 2nd ed. tinct pseudosarcomatous entity. Histopathology. 1987;11(1):21-35. Philadelphia, PA: Lippincott Williams & Wilkins; 2006. 93. Woodruff JM, Godwin TA, Erlandson RA, et al. Cellular schwanno- 65. Weiss SW. Smooth muscle tumors of soft tissue. Adv Anat Pathol. ma: a variety of schwannoma sometimes mistaken for a malignant tumor. 2002;9(6):351-359. Am J Surg Pathol. 1981;5(8):733-744. 66. Goldblum JR, Folpe AL, Weiss SW, et al. Enzinger and Weiss’s Soft 94. Casadei GP, Scheithauer BW, Hirose T, et al. Cellular schwan- Tissue Tumors. 6th ed. Philadelphia, PA: Saunders/Elsevier; 2014. noma. A clinicopathologic, DNA flow cytometric, and proliferation marker 67. Mentzel T, Calonje E, Fletcher CD. Leiomyosarcoma with promi- study of 70 patients. Cancer. 1995;75(5):1109-1119. nent osteoclast-like giant cells. Analysis of eight cases closely mimicking 95. Lodding P, Kindblom LG, Angervall L, et al. Cellular schwannoma. the so-called giant cell variant of malignant fibrous histiocytoma. Am J A clinicopathologic study of 29 cases. Virchows Arch A Pathol Anat Histopathol. Surg Pathol. 1994;18(3):258-265. 1990;416(3):237-248. 68. Merchant W, Calonje E, Fletcher CD. Inflammatory leiomyosar- 96. White W, Shiu MH, Rosenblum MK, et al. Cellular schwan- coma: a morphological subgroup within the heterogeneous family of noma. A clinicopathologic study of 57 patients and 58 tumors. Cancer. so-called inflammatory malignant fibrous histiocytoma. Histopathology. 1990;66(6):1266-1275. 1995;27(6):525-532. 97. Weiss SW, Langloss JM, Enzinger FM. Value of S-100 protein in 69. King ME, Dickersin GR, Scully RE. Myxoid leiomyosarcoma of the the diagnosis of soft tissue tumors with particular reference to benign and uterus. A report of six cases. Am J Surg Pathol. 1982;6(7):589-598. malignant Schwann cell tumors. Lab Invest. 1983;49(3):299-308. 70. Salm R, Evans DJ. Myxoid leiomyosarcoma. Histopathology. 98. Karamchandani JR, Nielsen TO, van de Rijn M, et al. Sox10 and 1985;9(2):159-169. S100 in the diagnosis of soft-tissue neoplasms. Appl Immunohistochem 71. Nistal M, Paniagua R, Picazo ML, et al. Granular changes in vascular Mol Morphol. 2012;20(5):445-450. leiomyosarcoma. Virchows Arch A Pathol Anat Histol. 1980;386(2):239-248. 99. Niwa T, Aida N, Fujita K, et al. Diffusion-weighted imaging of retro- 72. Oda Y, Miyajima K, Kawaguchi K, et al. Pleomorphic leiomyosar- peritoneal malignant peripheral nerve sheath tumor in a patient with neurofi- coma: clinicopathologic and immunohistochemical study with special bromatosis type 1. Magn Reson Med Sci. 2008;7(1):49-53. emphasis on its distinction from ordinary leiomyosarcoma and malignant 100. Yan B, Meng X, Shi B, et al. A retroperitoneal NF1-independent fibrous histiocytoma. Am J Surg Pathol. 2001;25(8):1030-1038. malignant peripheral nerve sheath tumor with elevated serum CA125: case 73. Rangdaeng S, Truong LD. Comparative immunohistochemical report and discussion. J Neurooncol. 2012;109(1):205-211. staining for desmin and muscle-specific actin. A study of 576 cases. Am J 101. deCou JM, Rao BN, Parham DM, et al. Malignant peripheral nerve Clin Pathol. 1991;96(1):32-45. sheath tumors: the St. Jude Children’s Research Hospital experience. Ann 74. Swanson PE, Stanley MW, Scheithauer BW, et al. Primary cutane- Surg Oncol. 1995;2(6):524-529. ous leiomyosarcoma. A histological and immunohistochemical study of 102. Daimaru Y, Hashimoto H, Enjoji M. Malignant “triton” tumors: a clini- 9 cases, with ultrastructural correlation. J Cutan Pathol. 1988;15(3):129-141. copathologic and immunohistochemical study of nine cases. Hum Pathol. 75. Hisaoka M, Wei-Qi S, Jian W, et al. Specific but variable expression 1984;15(8):768-778. of h-caldesmon in leiomyosarcomas: an immunohistochemical reassess- 103. Zou C, Smith KD, Liu J, et al. Clinical, pathological, and molecular ment of a novel myogenic marker. Appl Immunohistochem Mol Morphol. variables predictive of malignant peripheral nerve sheath tumor outcome. 2001;9(4):302-308. Ann Surg. 2009;249(6):1014-1022. 76. Watanabe K, Kusakabe T, Hoshi N, et al. h-Caldesmon in leio- 104. Matsunou H, Shimoda T, Kakimoto S, et al. Histopathologic and im- myosarcoma and tumors with smooth muscle cell-like differentiation: munohistochemical study of malignant tumors of peripheral nerve sheath its specific expression in the smooth muscle cell tumor. Hum Pathol. (malignant schwannoma). Cancer. 1985;56(9):2269-2279. 1999;30(4):392-396. 105. Nakajima T, Watanabe S, Sato Y, et al. An immunoperoxidase study 77. Watanabe K, Tajino T, Sekiguchi M, et al. h-Caldesmon as a spe- of S-100 protein distribution in normal and neoplastic tissues. Am J Surg cific marker for smooth muscle tumors. Comparison with other smooth Pathol. 1982;6(8):715-727. muscle markers in bone tumors. Am J Clin Pathol. 2000;113(5):663-668. 106. Wick MR, Swanson PE, Scheithauer BW, et al. Malignant peripheral 78. Brown DC, Theaker JM, Banks PM, et al. Cytokeratin expres- nerve sheath tumor. An immunohistochemical study of 62 cases. Am J Clin sion in smooth muscle and smooth muscle tumours. Histopathology. Pathol. 1987;87(4):425-433. 1987;11(5):477-486. 107. Terry J, Saito T, Subramanian S, et al. TLE1 as a diagnostic immu- 79. Miettinen M. Immunoreactivity for cytokeratin and epithelial membrane nohistochemical marker for synovial sarcoma emerging from gene expres- antigen in leiomyosarcoma. Arch Pathol Lab Med. 1988;112(6):637-640. sion profiling studies. Am J Surg Pathol. 2007;31(2):240-246. 80. Miettinen M. Keratin subsets in spindle cell sarcomas. Keratins are 108. Kosemehmetoglu K, Vrana JA, Folpe AL. TLE1 expression is not widespread but synovial sarcoma contains a distinctive keratin polypep- specific for synovial sarcoma: a whole section study of 163 soft tissue and tide pattern and desmoplakins. Am J Pathol. 1991;138(2):505-513. bone neoplasms. Mod Pathol. 2009;22(7):872-878. 81. Iwata J, Fletcher CD. Immunohistochemical detection of cytokera- 109. Nonaka D, Chiriboga L, Rubin BP. Sox10: a pan-schwannian and tin and epithelial membrane antigen in leiomyosarcoma: a systematic melanocytic marker. Am J Surg Pathol. 2008;32(9):1291-1298. study of 100 cases. Pathol Int. 2000;50(1):7-14. 110. Laskin WB, Weiss SW, Bratthauer GL. Epithelioid variant of malig- 82. Kelley TW, Borden EC, Goldblum JR. Estrogen and progester- nant peripheral nerve sheath tumor (malignant epithelioid schwannoma). one receptor expression in uterine and extrauterine leiomyosarcomas: Am J Surg Pathol. 1991;15(12):1136-1145. an immunohistochemical study. Appl Immunohistochem Mol Morphol. 111. Hollmann TJ, Hornick JL. INI1-deficient tumors: diagnostic features 2004;12(4):338-341. and molecular genetics. Am J Surg Pathol. 2011;35(10):e47-63. 83. Khalidi HS, Singleton TP, Weiss SW. Inflammatory malignant fibrous 112. Thway K, Fisher C. Malignant peripheral nerve sheath tumor: pathology histiocytoma: distinction from Hodgkin’s disease and non-Hodgkin’s lym- and genetics. Ann Diagn Pathol. 2014;18(2):109-116. phoma by a panel of leukocyte markers. Mod Pathol. 1997;10(5):438-442. 113. Pekmezci M, Reuss DE, Hirbe AC, et al. Morphologic and immuno-

July 2016, Vol. 23, No. 3 Cancer Control 263 histochemical features of malignant peripheral nerve sheath tumors and toneal lipoma: a soft tissue pathology heresy?: report of a case with clas- cellular schwannomas. Mod Pathol. 2015;28(2):187-200. sic histologic, cytogenetics, and molecular genetic features. Am J Surg 114. McMaster MJ, Soule EH, Ivins JC. Hemangiopericytoma. A Pathol. 2008;32(6):951-954. clinicopathologic study and long-term followup of 60 patients. Cancer. 146. Drop A, Czekajska-Chehab E, Maciejewski R. Giant retroperitoneal 1975;36(6):2232-2244. lipomas--radiological case report. Ann Univ Mariae Curie Sklodowska 115. Nielsen GP, O’Connell JX, Dickersin GR, et al. Solitary fibrous tu- Med. 2003;58(2):142-146. mor of soft tissue: a report of 15 cases, including 5 malignant examples 147. Martinez CA, Palma RT, Waisberg J. Giant retroperitoneal lipoma: with light microscopic, immunohistochemical, and ultrastructural data. a case report. Arq Gastroenterol. 2003;40(4):251-255. Mod Pathol. 1997;10(10):1028-1037. 148. Raftopoulos I, Lee T, Byrne MP. Image of the month: retroperitoneal 116. Gold JS, Antonescu CR, Hajdu C, et al. Clinicopathologic corre- lipoma. Arch Surg. 2002;137(7):865-866. lates of solitary fibrous tumors. Cancer. 2002;94(4):1057-1068. 149. Foa C, Mainguene C, Dupre F, et al. Rearrangement involving chro- 117. Gengler C, Guillou L. Solitary fibrous tumour and haemangiopericy- mosomes 1 and 8 in a retroperitoneal lipoma. Cancer Genet Cytogenet. toma: evolution of a concept. Histopathology. 2006;48(1):63-74. 2002;133(2):156-159. 118. Vallat-Decouvelaere AV, Dry SM, Fletcher CD. Atypical and malig- 150. Forte F, Maturo G, Catania A, et al. Retroperitoneal lipoma. Unusual nant solitary fibrous tumors in extrathoracic locations: evidence of their com- presentation with detrusor instability. Minerva Urol Nefrol. 2002;54(2):131-133. parability to intra-thoracic tumors. Am J Surg Pathol. 1998;22(12):1501-1511. 151. Marshall MT, Rosen P, Berlin R, et al. Appendicitis masquerading 119. Suster S, Nascimento AG, Miettinen M, et al. Solitary fibrous tu- as tumor: a case of two diagnoses. J Emerg Med. 2001;21(4):397-399. mors of soft tissue. A clinicopathologic and immunohistochemical study of 152. Matsubara N, Yoshitaka T, Matsuno T, et al. Multiple tumors and a 12 cases. Am J Surg Pathol. 1995;19(11):1257-1266. novel E2F-4 mutation. A case report. Digestion. 2000;62(2-3):213-216. 120. van de Rijn M, Lombard CM, Rouse RV. Expression of CD34 by 153. Acheson A, McIlrath E, Barros D’Sa AA. Pelvic lipoma causing ve- solitary fibrous tumors of the pleura, mediastinum, and lung. Am J Surg nous obstruction syndrome. Eur J Vasc Endovasc Surg. 1997;14(2):149-150. Pathol. 1994;18(8):814-820. 154. Zhang SZ, Yue XH, Liu XM, et al. Giant retroperitoneal pleomorphic 121. Chilosi M, Facchettti F, Dei Tos AP, et al. Bcl-2 expression in pleural lipoma. Am J Surg Pathol. 1987;11(7):557-562. and extrapleural solitary fibrous tumours. J Pathol. 1997;181(4):362-367. 155. Deppe G, Malviya VK, Hercule J, et al. Retroperitoneal pelvic lipoma. 122. Hanau CA, Miettinen M. Solitary fibrous tumor: histological and im- J Natl Med Assoc. 1985;77(7):574-576. munohistochemical spectrum of benign and malignant variants presenting at different sites. Hum Pathol. 1995;26(4):440-449. 123. Hasegawa T, Hirose T, Seki K, et al. Solitary fibrous tumor of the soft tissue. An immunohistochemical and ultrastructural study. Am J Clin Pathol. 1996;106(3):325-331. 124. Mohajeri A, Tayebwa J, Collin A, et al. Comprehensive genetic analysis identifies a pathognomonic NAB2/STAT6 fusion gene, non- random secondary genomic imbalances, and a characteristic gene expression profile in solitary fibrous tumor. Genes Chromosomes Cancer. 2013;52(10):873-886. 125. Yoshida A, Tsuta K, Ohno M, et al. STAT6 immunohistochemistry is helpful in the diagnosis of solitary fibrous tumors. Am J Surg Pathol. 2014;38(4):552-559. 126. Hwang KT CJ, Jung IM, Heo SC, et al. Primary retroperitoneal malignant gastrointestinal stromal tumor mimicking adrenal mass. J Korean Surg Soc. 2009;77(5):357-361. 127. Budzynski A, Wysocki A. Retroperitoneal desmoid tumor. Przegl Lek. 1996;53(6):506-507. 128. Ferenc T, Sygut J, Kopczynski J, et al. Aggressive fibromatosis (desmoid tumors): definition, occurrence, pathology, diagnostic problems, clinical behavior, genetic background. Pol J Pathol. 2006;57(1):5-15. 129. Ghidirim G, Mishin I, Gagauz I, et al. Sporadic retroperitoneal aggressive fibromatosis: report of a case. Zentralbl Chir. 2010;135(1):79-82. 130. Kikkawa A, Kido A, Kumai T, et al. Extraabdominal fibromatosis in retroperitoneal space. World J Surg Oncol. 2004;2:33. 131. Kim YW, Choi SJ, Jeon UB, et al. Retroperitoneal fibromatosis presenting as a presacral mass. Acta Radiol Short Rep. 2014;3(2):2047981614523760. 132. Li L, Lang J, Fan Q, et al. Primary pelvic retroperitoneal tumors: clinical analysis of 36 cases. Zhonghua Fu Chan Ke Za Zhi. 2014;49(11):838-841. 133. Mohos E, Kovacs T, Brittig F, et al. Desmoid tumors in three patients. Magy Seb. 2001;54(6):387-392. 134. Popowicz W, Drzewiecki L, Murawa D, et al. Desmoids: report of cases. Pol Merkur Lekarski. 2003;15(90):554-555. 135. Shih LY, Wei CK, Lin CW, et al. Postoperative retroperitoneal desmoid tumor mimics recurrent gastrointestinal stromal tumor: a case report. World J Gastroenterol. 2012;18(42):6172-6176. 136. Umemoto S, Makuuchi H, Amemiya T, et al. Intra-abdominal desmoid tumors in familial polyposis coli: a case report of tumor regression by prednisolone therapy. Dis Colon Rectum. 1991;34(1):89-93. 137. Whittle PC, Sanchez RL, Albarracin GJ, et al. Retroperitoneal fibromatosis: report of a case. Rev Med Chil. 2006;134(1):85-89. 138. Lazar AJ, Tuvin D, Hajibashi S, et al. Specific mutations in the beta- catenin gene (CTNNB1) correlate with local recurrence in sporadic desmoid tumors. Am J Pathol. 2008;173(5):1518-1527. 139. Weniger M, D’Haese JG, Kunz W, et al. En-bloc resection of a giant retroperitoneal lipoma: a case report and review of the literature. BMC Res Notes. 2015;8:75. 140. Saito S. Retroperitoneal lipoma presenting with nutcracker-like phenomenon. Case Rep Urol. 2013;2013:893242. 141. Wei D, Shen L, Yang K, et al. Giant retroperitoneal lipoma in a pregnant patient. J Obstet Gynaecol. 2013;33(5):522. 142. Chander B, Krishna M, Thakur S, et al. Extremely rare giant retroperitoneal fibrolipoma: a case report. J Cancer Res Ther. 2012;8(2):314-316. 143. Singh G, Bharadwaj RN, Purandare SN, et al. Giant retroperitoneal lipoma presenting as inguinal hernia. Indian J Surg. 2011;73(3):187-189. 144. Ukita S, Koshiyama M, Ohnaka M, et al. Retroperitoneal lipoma arising from the urinary bladder. Rare Tumors. 2009;1(1):e13. 145. Ida CM, Wang X, Erickson-Johnson MR, et al. Primary retroperi-

264 Cancer Control July 2016, Vol. 23, No. 3 Infectious Diseases Report

Primary Gangrenous Cutaneous Mold Infections in a Patient With Cancer and Neutropenia Abraham Yacoub, MD, Kiran K. Soni, Lysenia Mojica, MD, Jane Mai, MD, Jamie Morano, MD, C. Wayne Cruse, MD, Ramon L. Sandin, MD, Sowmya Nanjappa, MBBS, MD, Chandrashekar Bohra, MBBS, Ganesh Gajanan, MBBS, and John N. Greene, MD

Background: Opportunistic fungal infections caused by Aspergillus and Candida followed by infections with Fusarium, Rhizopus, Mucor, and Alternaria species are an important cause of morbidity and mortality in patients with hematological malignancies. Cutaneous mucormycosis infections are rare, and the incidence, outcomes, and factors associated with survival in the setting of hematological malignancies are not clear. Methods: A literature search was conducted for all cases of primary cutaneous mold infections in patients with hematological malignancy, of which 50 cases were found. Our case of a patient with a hematological malignancy who sustained a cat bite that in turn caused a primary cutaneous mold infection is also included. Results: In the 51 cases identified, 66.7% were neutropenic upon presentation, and 54.9% were male with an average age of 32 years. Aspergillus species (33.3%) was the most cited followed by Rhizopus species (19.6%). Overall mortality rate was 29.4% and was observed more frequently in patients with neutropenia (60.0%) and without surgical intervention (73.3%). Survival rate was higher (35.3%) for cases utilizing both antifungal and surgical intervention. The antifungal agent with the highest survival rate was amphotericin B and its formulations (58.8%). Conclusions: Neutropenia within hematological malignancies demonstrate a risk for developing severe cutaneous fungal infections, of which primary cutaneous mucormycosis can carry significant mortality. Combination antifungal therapy and surgical debridement appears to be associated with higher survival outcomes and warrants further investigation.

Introduction aplastic anemia who developed primary cutaneous Fungi from the Mucorales order are abundant, ubiqui- Rhizopus fungal infection on her right forearm follow- tous, and saprophytic, with rapidly growing mycelium ing a bite from a household cat. The cutaneous fungal that are typically unseptate or irregularly septate.1 Al- infection expanded and caused extensive necrotizing though important in the ecosystem,2 they are lethal fasciitis of the entire extensor compartment of the pa- opportunistic fungal infections when encountered in tient’s right forearm. persons who are immunocompromised.3 Two main types of fungal infections of the skin may be present in Case Report immunocompromised patients with neutropenia: pri- A 72-year-old woman diagnosed with aplastic anemia mary cutaneous fungal infections and cutaneous mani- received her first chemotherapy cycle of antithymocyte festations of fungemia.4 We present a case of a 72-year- globulin, methylprednisolone, and cyclosporine in July old, immunocompromised, neutropenic woman with 2013. She remained neutropenic for 3 weeks prior to being admitted to the hospital in August 2013. Following From the Division of Infectious Diseases and Tropical Medicine chemotherapy, she continued on prophylactic acyclovir (AY, LM, JNG), Departments of Internal Medicine (SN) and Oncologic and voriconazole while she concurrently received cyclo- Sciences (SN), Clinical Microbiology and Virology Laboratories sporine and weekly intravenous hydrocortisone 100 mg. (RLS), H. Lee Moffitt Cancer Center & Research Institute (CB, GG), and the Morsani College of Medicine (AY, KKS, SN), Division of She presented with a flat, homogenous, circular, Infectious Diseases and International Medicine (JMa, JMo), and nonerythematous lesion with a large black eschar Departments of Surgery (CWC), Oncologic Sciences (CWC, RLS), (Fig 1). She sustained a traumatic cat bite to the area ap- and Pathology (RLS), University of South Florida, Tampa, Florida. proximately 5 days prior to her cycle of chemotherapy. Address correspondence to John N. Greene, MD, Department of Infectious Diseases and Tropical Medicine, Moffitt Cancer The patient denied any fever, chills, or sweats. She was Center, 12902 Magnolia Drive, Tampa, FL 33612. E-mail: not dependent on packed red blood cell transfusions [email protected] but did receive biweekly platelets. She had no history Submitted July 23, 2015; accepted January 14, 2016. of hypoglycemic or iron-chelating agents. She was No significant relationships exist between the authors and the companies/organizations whose products or services may be empirically started on cefepime, metronidazole, and referenced in this article. azithromycin.

July 2016, Vol. 23, No. 3 Cancer Control 265 Fig 1. — A prominent 12 × 14 cm flat, homogenous, circular, and nonerythematous lesion with an eschar lesion.

Her vital signs were within normal limits. Physi- cal examination revealed a prominent 12 × 14 cm lesion with a large central eschar on her right forearm (see Fig 1). Recent laboratory studies included a white blood cell count of 0.77/mm3, an absolute neutrophil count (ANC) of 70 cells/mm3, a platelet level of 63,000/mm3, and a creatinine level of 0.9 mg/dL. The results of liver function tests were within normal limits. Punch biopsy from the histological section of the lesion on her forearm revealed multiple broad, hyposeptated hyphae in cross and longitudinal sections with abundant vascular invasion (Fig 2). This was consistent with Zygomycetes fungi (Mucorales) infection. Treatment was modified to liposomal amphotericin B 5 mg/kg intravenously once daily and posaconazole 200 mg suspension 4 times a day via a gastrostomy tube. This antifungal regimen was started immediately, and the patient underwent prompt surgical evaluation. The wound was extensively debrided (Fig 3). The cutaneous fungal infection had invaded the entire forearm extensor compartment, including the skin, fat, fascia and underlying musculature, demonstrating significant necrosis. Thus, we determined that the wound would best be resolved with amputation of the extremity, which the patient and daughter both declined. The area was closed and a temporary skin allograft was applied (Fig 4). The patient and her family subsequently declined any further surgical intervention or treatment, and she was discharged home with hospice care. She died 2 days later. Debrided, nonviable tissue was positive for Rhizopus by both fungal stain and culture. Angio- Fig 2. — Hematoxylin and eosin stains of specimens obtained from punch biopsy showing multiple broad, ribbon-like hyposeptated hyphae lymphatic invasion by Rhizopus was also seen on the in cross and longitudinal sections with abundant vascular invasion. This pathology specimen. is consistent with Zygomycetes fungi (Rhizopus species) infection.

266 Cancer Control July 2016, Vol. 23, No. 3 Table 1. — Clinical Characteristics of the Case Series Variable No. of No. of No. of Patients Patients Patients (N = 51) Who Who Died Survived (n = 15) (n = 36) Age, y Mean 32 Median 26 Sex Male 28 21 7 Female 23 15 8 Neutropenia, cells/mm3 ANC < 500 34 25 9

Fig 3. — The wound was extensively debrided surgically, including the ANC > 500 5 3 2 skin, fat, fascia, and muscle. Not stated 12 8 4 Malignancy AML/MDS/AMML 18 15 3 ALL 18 12 6 CML 4 3 1 CLL 1 0 1 Hodgkin lymphoma 3 1 2 NHL/mixed cutaneous 4 4 0 lymphoma Aplastic anemia 2 0 2 Multiple myeloma 1 1 0 ALL = acute lymphocytic leukemia, AML = acute myeloid leukemia, AMML = acute myelomonocytic leukemia, ANC = absolute neutrophil count, CLL = chronic lymphocytic leukemia, CML = chronic myelogenous leukemia, MDS = myelodysplastic syndrome, NHL = non-Hodgkin lymphoma. From references 5 to 23.

Fig 4. — The wound was temporarily closed with skin allograft. Acremonium 4% Fusarium Literature Review 6% In the 51 cases identified, 66.7% were neutropenic on Absidia presentation and 54.9% were male (Table 1).5-23 All pri- 6% Aspergillus mary cutaneous mold infections in patients with he- 33% matological malignancy were caused by molds from Other 1 of 2 phylum, Zygomycota or Ascomycota (Fig 5).5-23 10% Within the phylum Zygomycota, the following 5 genera/classes caused infections: Absidia, Mucorales, Rhizopus, Rhizomucor (“Other”), and Cunningha- Mucorales mella (“Other”; see Fig 5).5-23 Within the phylum As- 10% comycota, the following 5 genera caused infections: Aspergillus, Alternaria, Fusarium, Acremonium, 5-23 and Curvularia (“Other”; see Fig 5). Alternaria Rhizopus In patients with hematological malignancy and 12% 20% primary cutaneous mold infection, outcomes were determined on whether or not the infection was lo- 5-23 calized vs disseminated. In cases in which the in- Fig 5. — Distribution of reported primary cutaneous mold infections. fection remained localized, 100% of patients recov- Percentages do not add up to 100 due to rounding. From references 5 to 23.

July 2016, Vol. 23, No. 3 Cancer Control 267 ered; by contrast, in cases of disseminated infection, taneous mucormycosis. It has been postulated that in- the mortality rate was 29.4%.5-23 dividuals who lack phagocytes or who have impaired phagocytic function are at higher risk of mucormyco- Discussion sis than their normal counterparts.29 By contrast, pa- Mucormycosis, also known as zygomycosis, phyco- tients with HIV/AIDS do not appear to be at increased mycosis, or hyphomycosis, is a broad term for a mul- risk for developing mucormycosis.29 Therefore, neutro- titude of diseases caused by fungi in the Mucorales phils — but not necessarily T-cell lymphocytes — are order, with the most common causative organism be- critical for inhibiting fungal-spore proliferation.29 ing of the Rhizopus species.24 Other genera include, Cutaneous mucormycosis is an increasingly preva- in descending order of prevalence, Rhizomucor, Cun- lent form of mucormycosis that has been predominant- ninghamella, Apophysomyces, Saksenaea, Absidia, ly described in immunocompromised patients who are Mucor, Syncephalastrum, Cokeromyces, and Mortier- recipients of transplants, including both hematopoi- ella.24,25 Six clinical forms of mucormycosis have been etic stem cell and solid organ transplants after minor described: rhinocerebral, cutaneous, pulmonary, dis- trauma.30,31 Some studies suggest that the incidence of seminated, gastrointestinal, and miscellaneous.26 The zygomycosis appears to be increasing in oncology cen- most common clinical presentation of mucormycosis ters and specifically in patient populations receiving is rhino-orbital-cerebral infection.27 hematopoietic stem cell transplants, possibly related to Severe neutropenia with an ANC of less than the use of voriconazole prophylaxis.32 500 cells/mm3 increases susceptibility to bacte- Penetrating trauma remains the major cause of rial or fungal infections.28 Our patient’s ANC was primary cutaneous zygomycosis (Table 2).5-23,33-54 70 cells/mm3, which predisposed her to developing cu- Other causes reported include elasticized bandages

Table 2. — Primary Cutaneous Zygomycetes/Mucorales Infections in Patients With Cancer Study Age, Sex Malignancy Neutro- Fungus Cause of Treatment Outcome y penia Infection Arnaiz-Garcia38 71 M CML NA Mucor Dog scratch to Liposomal amphotericin B Survived species back of hand Hand amputation 72 M MDS NA Mucor NA Excision Survived Prostate species cancer Becker42 13 M ALL Yes Absidia Fall into an Liposomal amphotericin B Survived corymbifera emptied pond Surgical debridement Bethge39 26 F AML Yes Absidia NA Amphotericin B Survived species Surgical debridement Cantatore- 10 F AML Yes Mucorales Possible Liposomal amphotericin B Survived 34 Francis insect bite Surgical debridement 15 F Hodgkin Yes Mucorales Intravenous Liposomal amphotericin B Deceased lymphoma site Bedside excision Dennis50 6 F ALL Yes Rhizopus Under a Amphotericin B Survived oryzae pressure-gauge Surgical debridement bandage 7 M ALL Yes Rhizopus Under a Amphotericin B Survived oryzae pressure-gauge Surgical debridement bandage Durand47 53 M AML Yes Rhizopus Site of Foley Liposomal amphotericin B Deceased microspores catheter Surgical debridement Hadithi43 42 M AML Yes Absidia NA Amphotericin B Survived species Wide surgical excision, including periosteum of mandible and skin flap Hansen44 4 M Neuroecto- Yes Rhizopus Scraped knees Liposomal amphotericin B Survived dermal tumor species after fall Surgical debridement with skin graft

continued on next page

268 Cancer Control July 2016, Vol. 23, No. 3 Table 2. — Primary Cutaneous Zygomycetes/Mucorales Infections in Patients With Cancer, continued Study Age, Sex Malignancy Neutro- Fungus Cause of Treatment Outcome y penia Infection Hocker53 45 M AML Yes Rhizomucor NA Liposomal amphotericin B Survived species Posaconazole Surgical debridement Johnson51 79 M Prostate No Rhizopus Possible Amphotericin B Deceased cancer rhizopodiformis contamination from endotracheal tape Karam40 36 M AML Yes Absidia NA Amphotericin B Survived corymbifera Hyperbaric oxygen therapy Surgical debridement Khardori45 52 F AML Yes Rhizopus Site of Amphotericin B Survived species intravenous catheter Miyamoto49 21 M ALL Yes Rhizopus Adhesive Amphotericin B Deceased oryzae tape at site of Surgical debridement intravenous catheter Amputation Motohashi54 60 F CML No Cunninghamella Site of Liposomal amphotericin B Deceased bertholletiae elastic tape after puncture for pleural effusion Mueller37 46 M AML No Mucorales Site of Liposomal amphotericin B Survived catheter insertion for blood transfusion Roux41 13 F Osteosarcoma Yes Absidia Potential Amphotericin B Survived corymbifera skin trauma Posaconazole caspofungin under nonsterile plaster Rubin46 70 F CLL NA Rhizopus Site of arterial Amphotericin B Deceased arrhizus line Voriconazole Salati36 23 M ALL Yes Mucorales Adhesive tape Amphotericin B Survived at site of lum- Surgical debridement with skin graft bar puncture Sankar35 3 M ALL Yes Mucorales Adhesive tape Amphotericin B Survived at site of lum- Rifampicin bar puncture Surgical debridement Sundararajan48 7 F ALL Yes Rhizopus NA Amphotericin B Survived oryzae Surgical debridement

Wirth52 3 M ALL Yes Rhizopus Under Amphotericin B Survived species adhesive tape on venipuncture site Our case 72 F Aplastic Yes Rhizopus Cat bite Amphotericin B Deceased anemia species Surgical debridement Skin grafting ALL = acute lymphocytic leukemia, AML = acute myeloid leukemia, CLL = chronic lymphocytic leukemia, CML = chronic myelogenous leukemia, F = female, M = male, MDS = myelodysplastic syndrome, NA = not applicable, NHL = non-Hodgkin lymphoma. and wound dressings,55 burns,56 motor vehicle colli- bite is a rare finding. Although it was not confirmed, sion,57 intravenous access,58 and dog bite.59 Our case we suspect that this patient’s fungal infection may have report of primary cutaneous mucormycosis after a cat resulted from the cat’s mouth being contaminated with

July 2016, Vol. 23, No. 3 Cancer Control 269 2005;44(5):521-524. soil that contained Mucorales, because the patient did 6. Knapp CF, Tucci VT, Keeler JA, et al. Primary cutaneous Acremo- not recall having any other penetrating trauma. nium infection in a neutropenic patient after trauma. Infect Dis Clin Pract. 59 2010;18(4):277-278. Zachary et al suggest that, in canine flora, Mucor 7. Fincher RM, Fisher JF, Lovell RD, et al. Infection due to the fungus species is an atypical finding. Therefore, this could be Acremonium (cephalosporium). Medicine (Baltimore). 1991;70(6):398-409. 8. Vasiloudes P, Morelli JG, Weston WL. Painful skin papules caused true for a cat’s oral flora. This case is noteworthy be- by concomitant Acremonium and Fusarium infection in a neutropenic cause we found no existing case reports of cutaneous child. J Am Acad Dermatol. 1997;37(6):1006-1008. 9. Cooke NS, Feighery C, Armstrong DK, et al. Cutaneous Fusarium so- mucormycosis following a cat bite. lani infection in childhood acute lymphoblastic leukaemia. Clin Exp Dermatol. The diagnosis may be difficult to make because 2009;34(5):e117-e119. 10. Dai W, Dharamsi JW, Soliman S, et al. Cutaneous fusariosis devel- these infectious lesions can be confused with ischemic oping in a post-irradiation site. Dermatol Online J. 2011;17(5):5. pathology.60 Mucormycosis relies on analyses of clini- 11. Hsu CK, Hsu MM, Lee JY. Fusariosis occurring in an ulcerated cutaneous CD8+ T cell lymphoma tumor. Eur J Dermatol. 2006;16(3):297-301. cal samples by microscopy and culture. Recovery of 12. Perez-Perez L, Pereiro M Jr, Sanchez-Aguilar D, et al. Ulcerous Zygomycetes from tissue can be problematic, and fun- lesions disclosing cutaneous infection with Fusarium solani. Acta Derm Venereol. 2007;87(5):422-424. gal detection is often hindered by the absence of fun- 13. Iwen PC, Tarantolo SR, Sutton DA, et al. Cutaneous infection gal growth in culture; therefore, the diagnosis is often caused by Cylindrocarpon lichenicola in a patient with acute myelogenous 61 leukemia. J Clin Microbiol. 2000;38(9):3375-3378. delayed or made postmortem. Furthermore, Larché et al 14. Torrelo A, Hernandez-Martín A, Scaglione C, et al. Primary cutaneous and Nosari et al62 proposed use of a molecular test to aspergillosis in a leukemic child. Actas Dermosifiliogr. 2007;98(4):276-8. 15. Galimberti R, Kowalczuk A, Hidalgo PI, et al. Cutaneous aspergillosis: enable rapid identification of Mucorales at the genus a report of six cases. Br J Dermatol. 1998;139(3):522-526. and species levels. 16. Grossman ME, Fithian EC, Behrens C, et al. Primary cutaneous aspergillosis in six leukemic children. J Am Acad Dermatol. 1985;12(2 pt 1):313-318. Treatment consists of high doses of amphotericin 17. Estes SA, Hendricks AA, Merz WG, et al. Primary cutaneous asper- B combined with aggressive surgery and correction of gillosis. J Am Acad Dermatol. 1980;3(4):397-400. 18. D’Antonio D, Pagano L, Girmenia C, et al. Cutaneous aspergillosis in the immune deficits. Using lipid formulations of am- patients with haematological malignancies. Eur J Clin Microbiol Infect Dis. photericin B helps reduce nephrotoxicity and permits 2000;19(5):362-365. 19. Lucas GM, Tucker P, Merz WG. Primary cutaneous Aspergillus the administration of higher doses that may increase nidulans infection associated with a Hickman catheter in a patient with efficacy.63-65 In spite of appropriate management, the neutropenia. Clin Infect Dis. 1999;29(6):1594-1596. 20. Gene J, Azon-Masoliver A, Guarro J, et al. Cutaneous infection mortality rate for cutaneous mucormycosis has been caused by Aspergillus ustus, an emerging opportunistic fungus in immu- reported to be up to 16%.66 nosuppressed patients. J Clin Microbiol. 2001;39(3):1134-1136. 21. Knapp CF, Tucci VT, Sandin RL, et al. Cutaneous Curvularia infection The higher overall mortality rate of 29.4% seen in a neutropenic patient. Asian Biomed. 2008;2:393-395. in the review of the literature was observed more fre- 22. Ioannidou D, Maraki S, Kruger Krasagakis S, et al. Cutaneous alter- nariosis revealing acute myeloid leukaemia in an adult patient. Mycoses. quently in patients with neutropenia (60.0%) and with- 2004;47(5-6):227-230. out surgical intervention (73.3%), and the survival rate 23. Srinivasan A, Wickes BL, Romanelli AM, et al. Cutaneous infection caused by Macrophomina phaseolina in a child with acute myeloid leuke- was higher (35.3%) for cases utilizing both antifungal mia. J Clin Microbiol. 2009;47(6):1969-1972. and surgical interventions.5-23 The antifungal agent 24. Waness A, Dawsari GA, Al Jahdali H. The rise of an opportunistic infection called “invasive zygomycosis.” J Glob Infect Dis. 2009;1(2):131-138. with the highest survival rate was amphotericin B and 25. Morace G, Borghi E. Invasive mold infections: virulence and patho- its formulations (58.8%).5-23 genesis of mucorales. Int J Microbiol. 2012;2012:349278. 26. Tapia EO, Chahín AC, Concha FC. Primary cutaneous mucormycosis: two case reports and review of the literature. Rev Chilena Infectol. Conclusions 2011;28(3):269-273. 27. Elsiesy H, Saad M, Shorman M, et al. Invasive mucormycosis in a Neutropenia within the setting of hematological ma- patient with liver cirrhosis: case report and review of the literature. Hepat lignancies increases the risk for developing severe Mon. 2013;13(8):e10858. 28. Boxer LA. How to approach neutropenia. Hematol Am Soc Hematol cutaneous fungal infections, of which primary cuta- Educ Program. 2012;2012:174-182. neous mucormycosis can carry significant mortality. 29. Ibrahim AS, Spellberg B, Walsh TJ, et al. Pathogenesis of mucormycosis. Clin Infect Dis. 2012;54(suppl 1):S16-S22. Combination antifungal therapy and surgical debride- 30. Baraia J, Munoz P, Bernaldo de Quiros JC, et al. Cutaneous mucor- ment appear to be associated with higher survival mycosis in a heart transplant patient associated with a peripheral catheter. Eur J Clin Microbiol Infect Dis. 1995;14(9):813-815. outcomes and, thus, this treatment plan warrants fur- 31. Petrikkos G, Drogari-Apiranthitou M. Zygomycosis in immunocom- ther investigation. promised non-haematological patients. Mediterr J Hematol Infect Dis. 2011;3(1):e2011012. 32. Person AK, Kontoyiannis DP, Alexander BD. Fungal infections in trans- References plant and oncology patients. Infect Dis Clin North Am. 2010;24(2):439-459. 33. Skiada A, Petrikkos G. Cutaneous zygomycosis. Clin Microbiol Infect. 1. Hoffmann K, Pawłowska J, Walther G, et al. The family structure of 2009;15(suppl 5):41-45. the Mucorales: a synoptic revision based on comprehensive multigene- 34. Cantatore-Francis JL, Shin HT, Heilman E, et al. Primary cutane- genealogies. Persoonia. 2013;30:57-76. ous zygomycosis in two immunocompromised children. Pediatr Dermatol. 2. Tian B-Y, Huang Q-G, Xu Y, et al. Microbial community structure 2007;24(3):257-262. and diversity in a native forest wood-decomposed hollow-stump ecosys- 35. Sankar J, Arun S, Sankar MJ, et al. Primary cutaneous mucormy- tem. World J Microbiol Biotechnol. 2010;26(2):233-240. cosis during induction chemotherapy in a child with acute lymphoblastic 3. Koklu E, Akcakus M, Torun YA, et al. Primary gangrenous cutane- leukemia. Indian J Pediatr. 2009;76(11):1161-1163. ous mucormycosis of the scalp in a child: a case report. Pediatr Emerg 36. Salati SA, Rabah SM. Cutaneous mucormycosis in a leukemic pa- Care. 2008 Feb;24(2):102-104. tient. J Coll Physicians Surg Pak. 2011;21(2):109-110. 4. Mays SR, Bogle MA, Bodey GP. Cutaneous fungal infections in 37. Mueller BU, Pabst T. Cutaneous zygomycosis at catheter insertion the oncology patient: recognition and management. Am J Clin Dermatol. site in AML-M4Eo. Ann Hematol. 2006;85(3):194-195. 2006;7(1):31-43. 38. Arnaiz-Garcia ME, Alonso-Pena D, Gonzalez-Vela Mdel C, et al. 5. Chang YH, Huang LM, Hsueh PR, et al. Acremonium pyomyo- Cutaneous mucormycosis: report of five cases and review of the literature. sitis in a pediatric patient with acute leukemia. Pediatr Blood Cancer. J Plast Reconstr Aesthet Surg. 2009;62(11):e434-e441.

270 Cancer Control July 2016, Vol. 23, No. 3 39. Bethge WA, Schmalzing M, Stuhler G, et al. Mucormycoses in patients with hematologic malignancies: an emerging fungal infection. Haema- tologica. 2005;90(suppl):ECR22. 40. Karam A, Ianotto JC, Metges JP, et al. Ulceration of the penis due to Absidia corymbifera. Br J Dermatol. 2003;148(6):1286-1287. 41. Roux BG, Mechinaud F, Gay-Andrieu F, et al. Successful triple combination therapy of disseminated Absidia corymbifera infection in an adolescent with osteosarcoma. J Pediatr Hematol Oncol. 2010;32(2):131-133. 42. Becker BC, Schuster FR, Ganster B, et al. Cutaneous mucormycosis in an immunocompromised patient. Lancet Infect Dis. 2006;6(8):536. 43. Hadithi M, Leemans CR, Jonkhoff AR. Cutaneous mucormycosis in acute myeloid leukemia; hit hard and quickly. Haematologica. 2001;86(12):1320. 44. Hansen JP, Kristjansson AK, Stone MS. A neutropenic child with a purple lesion on the thigh. Pediatr Dermatol. 2007;24(5):560-561. 45. Khardori N, Hayat S, Rolston K, et al. Cutaneous Rhizopus and Aspergillus infections in five patients with cancer. Arch Dermatol. 1989;125(7):952-956. 46. Rubin AI, Grossman ME. Bull’s-eye cutaneous infarct of zygomycosis: a bedside diagnosis confirmed by touch preparation. J Am Acad Dermatol. 2004;51(6):996-1001. 47. Durand CM, Alonso CD, Subhawong AP, et al. Rapidly progressive cutaneous Rhizopus microsporus infection presenting as Fournier’s gan- grene in a patient with acute myelogenous leukemia. Transpl Infect Dis. 2011;13(4):392-396. 48. Sundararajan T, Kumar CP, Menon T, et al. Cutaneous zygomycosis due to Rhizopus oryzae in a patient with acute lymphoblastic leukemia. Mycoses. 2004;47(11-12):521-523. 49. Miyamoto H, Hayashi H, Nakajima H. Cutaneous mucormycosis in a patient with acute lymphocytic leukemia. J Dermatol. 2005;32(4):273-277. 50. Dennis JE, Rhodes KH, Cooney DR, et al. Nosocomical Rhizopus infection (zygomycosis) in children. J Pediatr. 1980;96(5):824-828. 51. Johnson KE, Leahy K, Owens C, et al. An atypical case of fatal zygomycosis: simultaneous cutaneous and laryngeal infection in a patient with a non-neutropenic solid prostatic tumor. Ear Nose Throat J. 2008;87(3):152-155. 52. Wirth F, Perry R, Eskenazi A, et al. Cutaneous mucormycosis with subsequent visceral dissemination in a child with neutropenia: a case report and review of the pediatric literature. J Am Acad Dermatol. 1997;36(2 pt 2):336-341. 53. Hocker TL, Wada DA, Bridges A, et al. Disseminated zygomycosis heralded by a subtle cutaneous finding. Dermatol Online J. 2010;16(9):3. 54. Motohashi K, Ito S, Hagihara M, et al. Cutaneous zygomycosis caused by Cunninghamella bertholletiae in a patient with chronic myelogenous leukemia in blast crisis. Am J Hematol. 2009;84(7):447-448. 55. Patterson JE, Barden GE, Bia FJ. Hospital-acquired gangrenous mucormycosis. Yale J Biol Med. 1986;59(4):453-459. 56. Constantinides J, Misra A, Nassab R, et al. Absidia corymbifera fungal infection in burns: a case report and review of the literature. J Burn Care Res. 2008;29(2):416-419. 57. Ryan-Poirier K, Eiseman RM, Beaty JH, et al. Post-traumatic cutaneous mucormycosis in diabetes mellitus. Short-term antifungal therapy. Clin Pediatr (Phila). 1988;27(12):609-612. 58. Wollstein R, Palekar A. Mucormycosis infection following intravenous access in the forearm. Can J Plast Surg. 2010;18(2):e30-e32. 59. Zachary D, Chapin K, Binns L, et al. Cutaneous mucormycosis complicating a polymicrobial wound infection following a dog bite. Case Rep Infect Dis. 2011;2011:348046. 60. Raizman NM, Parisien M, Grafe MW, et al. Mucormycosis of the upper extremity in a patient with alcoholic encephalopathy. J Hand Surg Am. 2007;32(3):384-388. 61. Larché J, Machouart M, Burton K, et al. Diagnosis of cutaneous mucormycosis due to Rhizopus microsporus by an innovative PCR-restriction fragment-length polymorphism method. Clin Infect Dis. 2005;41(9):1362-1365. 62. Nosari A, Oreste P, Montillo M, et al. Mucormycosis in hematologic malignancies: an emerging fungal infection. Haematologica. 2000;85(10):1068-1071. 63. Losee JE, Selber J, Vega S, et al. Primary cutaneous mucormycosis: guide to surgical management. Ann Plast Surg. 2002;49(4):385-390. 64. Kontoyiannis DP, Wessel VC, Bodey GP, et al. Zygomycosis in the 1990s in a tertiary-care cancer center. Clin Infect Dis. 2000;30(6):851-856. 65. Sharma A, Lokeshwar N. Febrile neutropenia in haematological malignancies. J Postgrad Med. 2005;51(suppl 1):S42-S48. 66. Jones NF, Shin EK, Eo S, et al. Successful salvage of mucormycosis infection of the forearm and osteomyelitis of the ulna. Hand (N Y). 2008;3(4):332-336.

July 2016, Vol. 23, No. 3 Cancer Control 271 Case Series

Diffuse Alveolar Hemorrhage in Acute Myeloid Leukemia Sowmya Nanjappa, MBBS, MD, Daniel K. Jeong, MD, Manjunath Muddaraju, MD, Katherine Jeong, MD, Eboné D. Hill, MD, and John N. Greene, MD

Summary: Diffuse alveolar hemorrhage is a potentially fatal pulmonary disease syndrome that affects individuals with hematological and nonhematological malignancies. The range of inciting factors is wide for this syndrome and includes thrombocytopenia, underlying infection, coagulopathy, and the frequent use of anticoagulants, given the high incidence of venous thrombosis in this population. Dyspnea, fever, and cough are commonly presenting symptoms. However, clinical manifestations can be variable. Obvious bleeding (hemoptysis) is not always present and can pose a potential diagnostic challenge. Without prompt treatment, hypoxia that rapidly progresses to respiratory failure can occur. Diagnosis is primarily based on radiological and bronchoscopic findings. This syndrome is especially common in patients with hematological malignancies, given an even greater propensity for thrombocytopenia as a result of bone marrow suppression as well as the often prolonged immunosuppression in this patient population. The syndrome also has an increased incidence in individuals with hematological malignancies who have received a bone marrow transplant. We present a case series of 5 patients with acute myeloid leukemia presenting with diffuse alveolar hemorrhage at our institution. A comparison of clinical manifestations, radiographic findings, treatment course, and outcomes are described. A review of the literature and general overview of the diagnostic evaluation, differential diagnoses, pathophysiology, and treatment of this syndrome are discussed.

Introduction 5 patients. The median patient age was 57 years (range, Diffuse alveolar hemorrhage (DAH) is a potentially fa- 51–70 years). Three (60%) patients were women. No tal pulmonary disease that affects individuals with can- congruity was observed on the AML subtype diag- cer and hematological/nonhematological malignan- nosed by pathology. Two (40%) patients had received cies. Clinicians caring for patients with acute myeloid allogeneic hematopoietic stem cell transplantation leukemia (AML) must maintain a high index of sus- (HSCT) with matched unrelated donors in the year picion for DAH, because these patients are immuno- prior to developing DAH, whereas the other 3 patients compromised by their underlying disease process and were receiving induction chemotherapy (7 days stan- subsequently at increased risk because they typically dard-dose cytarabine plus 3 days of daunorubicin). receive treatment with immunosuppressive agents and The 2 patients who received HSCT had prior pulmo- stem cell transplantation. Here we present a case series nary function tests available, with results revealing de- of 5 patients with AML who developed DAH. We also creased diffusing lung capacity for carbon monoxide review the clinical presentation, diagnostic evaluation, of below 80%. Three patients had a history of a prior differential diagnosis, and therapeutic management of or underlying pulmonary disorder, including 1 patient this potentially life-threatening clinical syndrome. with a history of left lower lobe resection, 1 patient with a history of chronic obstructive pulmonary dis- Case Series ease, and 2 patients with a history of cryptogenic-or- Table 1 summarizes the clinical characteristics for the ganizing pneumonia. One patient had a history of an unknown autoimmune disorder affecting his ear. Each patient’s clinical presentation was different. From the Departments of Internal Medicine (SN, KJ, EDH), Radi- One patient presented with fever at the time of diag- ology (DKJ), and Infectious Diseases (JNG), H. Lee Moffitt Cancer Center & Research Institute, Department of Oncologic Sciences nosis of DAH (by bronchoscopy). Four patients devel- (SN, DKJ, EDH), University of South Florida Morsani College of oped hypoxia with increasing oxygen requirements Medicine, Tampa, Florida, and the Department of Critical Care and dyspnea. Three patients complained of cough, and Medicine (MM), SSM St Mary’s Health Center, St Louis, Missouri. 2 patients presented with bleeding at other sites in the Address correspondence to Sowmya Nanjappa, MBBS, MD, and John N. Greene, MD, Moffitt Cancer Center, 12902 Magnolia Drive, form of epistaxis. One patient had hemoptysis and that Tampa, FL 33612. E-mails: [email protected] and patient was neither hypoxic nor dyspneic at the time. [email protected] All patients were thrombocytopenic and anemic. Submitted February 17, 2016; accepted March 4, 2016. On radiological imaging, all 5 patients had bilateral, No significant relationships exist between the authors and the companies/organizations whose products or services may be referenced ground-glass pulmonary opacities on computed tomog- in this article. raphy (CT). In 3 patients, the findings of bronchoalveo-

272 Cancer Control July 2016, Vol. 23, No. 3 lar lavage (BAL) on bronchoscopy confirmed the diag- tor VII. Three patients died due to hypoxic respiratory nosis of DAH by retrieving progressively bloody effluent failure within 30 days of confirming the DAH diagno- on BAL. DAH was suspected in 1 patient by her clinical sis by bronchoscopy. The other 2 patients did not re- syndrome and pinkish effluent on BAL, and 1 patient quire intubation and were weaned off supplemental had suspected DAH based on documentation of bloody oxygen and discharged home in stable condition. secretions noted from bronchoscopy performed at a different institution. Although none of the cultures taken Discussion during bronchoscopy in the 5 patients revealed an un- DAH is a clinical syndrome defined by a disturbance of derlying infectious process, all patients were being em- the alveolar-capillary basement membrane that causes pirically given broad-spectrum antibiotics. bleeding into the pulmonary alveoli.1 It is a final, com- All 5 patients were treated with high-dose steroids mon pathway for a variety of underlying issues affect- and 1 patient also received aminocaproic acid and fac- ing the lungs, including autoimmune vasculitides, drug

Table 1. — Select Clinical Characteristics of Study Patients Patient Sex Age, Malignancy Oncological Pulmonary Exposure Symptoms Bleeding at Available Vitals WBC, Hemo- Platelet y Therapy Disease to Antico- Other Sites on Presentation 103/ globin, Count, at DAH History agulants mcL g/dL 103/mcL Diagnosis 1 M 63 AML 7 + 3 induction COP No Cough Epistaxis BP: 124/78 mm Hg 3.74 7.4 14 (FAB, M5a) LLL resection Dyspnea HR: 102 beats/ Unknown Fever minute autoimmune Hypoxia RR: 20 breaths/ disorder minute Temp: 98.8 °F 2 F 51 AML MUD COP Prophy- Cough Epistaxis BP: 121/80 mm Hg 0.02 8.1 35 (FAB, M1 Allogenic COPD lactic Fever Hemoptysis HR: 109 beats/ in relapse) HSCT dalteparin minute Rhinovirus (5000 U) Hemoptysis RR: 20 breaths/ minute

O2 saturation: 96% on room air Temp: 98.7 °F

3 M 70 ALL Reinduction Nodular No Dyspnea No BP: 99/51 mm Hg 0.03 7.4 5 (WHO, MEC pneumonia Hypoxia HR: 76 beats/ therapy- Suspected minute related sleep apnea myeloid RR: 17 breaths/ neoplasms) minute Temp: 97.7 °F 4 F 57 AML MUD None No Dyspnea No BP: 127/71 mm Hg 0.30 9.8 15 (WHO, with Allogenic Hypoxia HR: 78 beats/ myelodys- HSCT minute plasia-related Left chest changes) wall pain RR: 30 breaths/ minute

O2 saturation: 97% on room air Temp: 97.8 °F 5 F 56 AML 7 + 3 induction None No Cough No BP: 139/88 mm Hg 95.10 7.1 20 (FAB, M4) Dyspnea HR: 109 beats/ Hypoxia minute RR: 22 breaths/ minute

O2 saturation: 95% on room air Temp: 99.3 °F continued on the next page

July 2016, Vol. 23, No. 3 Cancer Control 273 Table 1. — Select Clinical Characteristics of Study Patients, continued Patient Pro- BAL CMV Cultures Medical Antibiotic Treatment Intubation Mortality Length calcitonin, Appearance PCR Treatment of ng/mL Respiratory Sputum Blood for DAH Given on Duration Hospital Admission Stay, d 1 NR Bloody Negative Negative No Negative Steroids None NA Yes Deceased 15 results 2 NR Bloody Negative Negative No VRE Aminocaproic Yes 41 d Yes Deceased 41 results Candida acid Received for glabrata Factor VII prolonged Steroids neutropenia, treatment of bacteremia 3 NR Bloody No Negative No Strepto- Steroids Yes 32 d No Alive 32 results results coccus Empiric mitis antibiotics for Strepto- neutropenia, coccus treatment of oralis bacteremia 4 0.23 Bloody No Negative No Negative Steroids Yes 10 d No Deceased 12 results results Empiric antibiotics for neutropenia 5 1.21 Pinkish No Negative No Negative Steroids Yes 22 d No Alive 22 results results Initially started for treatment of pneumonia Continued for prophylaxis after prolonged neutropenia 7 + 3 = 7 days standard-dose cytarabine plus 3 days of daunorubicin, ALL = acute lymphocytic leukemia, AML = acute myeloid leukemia, BAL = bronchoal- veolar lavage, BP = blood pressure, CMV = cytomegalovirus, COP = cryptogenic organizing pneumonia, COPD = chronic obstructive pulmonary disease, DAH = diffuse alveolar hemorrhage, F = female, FAB = French American British classification, HR = heart rate, HSCT = hematopoietic stem cell transplantation, LLL = left lower lobe, M = male, MEC = mitoxantrone/etoposide/intermediate-dose cytarabine, MUD = matched unrelated donor, NR = not reported, PCR = polymerase chain reaction, RR = respiratory rate, Temp = temperature, VRE = vancomycin-resistant enterococci, WBC = white blood count, WHO = World Health Organization classification. toxicities, infections, organ transplantation, and radio- anticoagulants.2,7 Diffuse alveolar damage with edema therapy.1,2 Among patients with leukemia, DAH is one of of the alveolar septa without signs of lung inflammation the most common, noninfectious complications related or direct extravasation of red blood cells characterizes a to thrombocytopenia.2 In particular, DAH has been as- third pattern of DAH, with hyaline membranes forming sociated with patients following HSCT, with a reported along the alveolar spaces.7 Infections, transplantation, incidence of 3% to 20% in patients following HSCT and a including HSCT, and radiotherapy have all been associ- mortality rate that ranges from 50% to 80%.3 Early recog- ated with this pattern of DAH.7 nition and management is crucial given the potential for rapid progression and the high rate of mortality. Clinical Presentation DAH can be divided into 3 characteristic patterns In accordance with the patients in our case series, the on histological examination, thus reflecting different clinical signs and symptoms of DAH may vary. The pathophysiological mechanisms. The most common onset of symptoms is typically acute and nonspecific; presentation described is associated with pulmonary fever, cough, dyspnea, and chest pain are among the vasculitis or capillaritis and involves neutrophilic infil- most commonly presenting symptoms.1 tration and destruction of the pulmonary interstitium.2,7 Although hemoptysis may occur in up to This pattern of DAH is often caused by systemic auto- two-thirds of individuals with DAH, our case series immune vasculitides. Another characteristic pattern in- demonstrates that hemoptysis is often absent.1 A de- volves extravasation of red blood cells into the alveoli crease in hematocrit level should alert the clinician to without signs of vasculitis or inflammation, and it may possible DAH. Another report has underscored the ab- be seen with infections and a variety of drugs, including sence of hemoptysis in patients with DAH, so hemopty-

274 Cancer Control July 2016, Vol. 23, No. 3 sis alone is not a reliable indicator for diagnosing DAH.4 However, bleeding from other sites may support clinical suspicion for DAH.5 Our patients exhibited epistaxis, but other signs or symptoms of coagulopathy such as purpura may be present. Acute respiratory failure with hypoxia and an increasing oxygen requirement is also often seen as a presenting symptom.2 Findings on a pulmonary examination are typically nonspecific and may reveal tachypnea, crackles, or bronchial breath sounds. Although the underlying pathophysiology that leads to DAH following HSCT is not well understood, patients who have received stem cell transplants are often at greatest risk for DAH within the first 30 days following transplantation.1,3 DAH occurs in approximately 5% of allogenic and autologous recipients and has a mortality rate that ranges from 50% to 100%.12 Other risk factors for DAH following HSCT include advanced age, myeloablative chemotherapy prior to transplantation, radiotherapy, and severe acute graft-vs-host 2 disease. Our 2 patients who received stem cell trans- Fig 1. — Posteroanterior radiograph of the chest demonstrates bilateral, plants did not have DAH in the early post-transplant multifocal opacities. period, but it is worth noting that patients with leukemia remain at risk given their underlying immunosuppression and hematological disorder. A

Diagnostic Evaluation Abnormal blood cell counts, including those resulting in anemia and leukocytosis, and elevated inflammatory markers are typical in the setting of DAH.1 However, in patients with AML, these laboratory abnormalities are often ubiquitous and, thus, are limited in offering diagnostic clues. Furthermore, prior research in patients receiving HSCTs has found that the initial degree of thrombocytopenia does not correlate with the development of DAH.2,6 Findings on pulmonary function tests in the setting of DAH are typically characterized by an increased diffusing lung capacity for carbon monoxide and decreased levels of exhaled nitric oxide.1 The pa- B tients in our case series had normal findings on pulmonary function testing, but these tests were performed prior to the onset of clinical symptoms of DAH. With the rapid progression and severity of DAH, patients are often unable to complete pulmonary function testing at the time of symptoms and diagnosis.6 Findings on chest radiography may appear normal or may show nonspecific, diffuse opacities reflec- tive of diffuse lung disease (Fig 1).2 CT of the chest will characteristically show diffuse or patchy, bilateral ground-glass opacities, as were seen in all 5 of our case patients (Fig 2).1 These opacities are often more centrally located and at the bases; they typically spare the periphery.3 In addition, interlobular septal thicken- ing may develop.2 Findings on CT are likely to rapidly Fig 2A–B. — (A) Axial CT of the chest demonstrates bilateral, multifocal 2 ground-glass opacities with a geographical distribution compatible with progress in conjunction with disease progression. Al- alveolar hemorrhage. (B) These opacities were acutely increased from though DAH is a clinical syndrome drawing together findings on CT obtained 5 days prior. clinical symptoms, laboratory abnormalities, and find- CT = computed tomography.

July 2016, Vol. 23, No. 3 Cancer Control 275 ings on imaging, bronchoscopy with BAL showing pro- It is important that clinicians remember that DAH gressively bloody effluent is usually diagnostic.7 When can disguise itself in areas of infection as well as in- it is performed, Prussian blue staining of the BAL fluid farction.2 As reflected in our case series, it seems rea- will characteristically reveal hemosiderin-laden mac- sonable to consider empirical antibiotic coverage while rophages or siderophages.3 Prussian blue staining was establishing a diagnosis. Furthermore, pulmonary ede- not performed in our case series. However, previous ma and prior injury from radiotherapy can complicate reports have attempted to define alveolar hemorrhage or obscure the diagnosis and should be worked-up and by the presence of at least 20% of siderophages on ex- treated if necessary.2 amination of BAL fluid samples to aid in the diagnosis In patients who have received HSCT, idiopathic when BAL findings are equivocal.6 pneumonia syndrome and engraftment syndrome should also be considered in the differential diagno- Differential Diagnosis sis.2 Engraftment syndrome is usually characterized Although it is essential to maintain a high index of sus- by noninfectious fever following HSCT, skin rash or picion for DAH, clinicians must also consider a broad erythroderma, and increased capillary permeability differential of alternative diagnoses. In addition, con- leading to pulmonary edema with similar radiologi- sideration should be given to the potential etiologies cal appearances to DAH.2 Engraftment syndrome oc- instigating DAH. Patients with leukemia may develop curs in 30% to 40% of patients following bone marrow DAH related to thrombocytopenia or another coagu- transplant, and it is thought to be related to cytokine lopathy triggered by the malignancy. These patients production during engraftment.3 Idiopathic pneumo- also often require treatment with anticoagulants given nia syndrome has been described in patients follow- their propensity for venous thrombosis. Other treating HSCT, typically within the first 6 months after ments can place patients at risk for developing DAH transplantation.2,8 This syndrome is associated with through bone marrow suppression. Use of cytarabine symptoms and signs of pneumonia and widespread al- and all-transretinoic acid are among several immuno- veolar injury despite the absence of infection, cardiac suppressive agents associated with DAH.2 dysfunction, renal failure, or iatrogenic fluid.2,8 Indeed, When considering alternative diagnoses to DAH, this expansive definition encompasses a variety of localized pulmonary hemorrhage caused by bronchi- disorders, and DAH following HCST and engraftment ectasis, tumor, or infection should be distinguished syndrome have both been considered subset classifica- from DAH.1 Pulmonary infections, acute eosinophilic tions of idiopathic pneumonia syndrome.8 pneumonia, aspiration pneumonitis, and acute respiratory distress syndrome can all present with similar Treatment clinical and radiographic findings. Diffuse leukemic Standard treatment strategies for DAH are typically di- infiltration of the lungs may be a manifestation of ma- rected toward the underlying etiology. Supportive-care lignancy and can be confused with DAH.5 Viral pneu- measures, including hemodynamic support and inva- monias, Pneumocystis jiroveci infection, Mycoplasma sive or noninvasive oxygen supplementation, the dis- pneumoniae infection, and toxoplasmosis, among oth- continuation of any offending agents, and reversal of er atypical infections, can often present with diffuse coagulopathy are basic first-line therapies. In cases of pulmonary infiltrates comparable with CT findings for DAH with an etiology suspected to be attributable to DAH.2 Table 2 lists the infectious causes of DAH.9-20 autoimmune vasculitides, immunosuppressive therapies and plasmapheresis have been used. Systemic high-dose steroids are typically recommended for non- Table 2. — Infectious Causes of Diffuse Alveolar Hemorrhage infectious cases of DAH to combat the inflammation Category Agent/Disease precipitated by the underlying disease process.9 How- Bacterial Legionella13 ever, research has not shown significant differences in Leptospirosis10 rates of mortality between patients with DAH receiving 9 Mycobacterium tuberculosis11 low-, medium-, or high-dose steroids. Mycoplasma pneumoniae12 Use of platelet transfusions to treat thrombocyto- Stenotrophomonas maltophilia9 penia in the setting of hemorrhage is widely accepted, but further research is necessary for the use of other Fungal Invasive aspergillosis18 Mucormycosis19 treatments such as aminocaproic acid and activated factor VII. Although initial study results have suggest- 20 Parasitic Strongyloides stercoralis ed a benefit to using aminocaproic acid in addition to Viral Cytomegalovirus15 steroids to treat DAH in patients who had received al- 14 Dengue logeneic HSCT, subsequent research has showed no Influenza H1N116 benefit.9 Prior case reports and research have shown Hantavirus17 that use of factor VII can control bleeding in the set-

276 Cancer Control July 2016, Vol. 23, No. 3 ting of DAH, but the cost can be restrictive in patients 17. Hong YM, Moon JC, Yang HC, et al. Hemorrhagic fever with renal 10 syndrome and coexisting hantavirus pulmonary syndrome. Kidney Res whose underlying disease remains grim. Use of factor Clin Pract. 2012;31(2):118-120. VII must also be weighed against the risk of thrombo- 18. Wah TM, Moss HA, Robertson RJ, et al. Pulmonary complications following bone marrow transplantation. Br J Radiol. 2003;76(906):373-379. sis, and clinicians should use it with caution or avoid it 19. Petrikkos G, Skiada A, Lortholary O, et al. Epidemiology and clinical altogether in patients who have had recent surgery or manifestations of mucormycosis. Clin Infect Dis. 2012;54(suppl 1):S23-S34. 20. Steinhaus D, Gainor J, Vernovsky I, et al. Survival in a case of dif- have a history of thrombosis, liver disease, myocardial fuse alveolar hemorrhage due to Strongyloides stercoralis hyperinfection. infarction, or stroke.10 The patients in our case series Respir Med Case Rep. 2012;5:4-5. 21. Panoskaltsis-Mortari A, Griese M, Madtes DK, et al; American all received high-dose steroids in accordance with cur- Thoracic Society Committee on Idiopathic Pneumonia Syndrome. An rent recommendations; 1 patient received aminocapro- official American Thoracic Society research statement: noninfectious lung injury after hematopoietic stem cell transplantation: idiopathic pneumonia ic acid and factor VII without success. syndrome. Am J Respir Crit Care Med. 2011;183(9):1262-1279. 22. Rathi NK, Tanner AR, Dinh A, et al. Low-, medium- and high-dose steroids with or without aminocaproic acid in adult hematopoietic SCT Conclusions patients with diffuse alveolar hemorrhage. Bone Marrow Transplant. Early recognition and prompt diagnosis of diffuse alve- 2015;50(3):420-426. 23. Pathak V, Kuhn J, Gabriel D, et al. Use of activated factor VII in olar hemorrhage (DAH) in patients with acute myeloid patients with diffuse alveolar hemorrhage: a 10 years institutional experience. leukemia can be a challenging but decisive step toward Lung. 2015;193(3):375-379. decreasing the high mortality rate associated with this life-threatening syndrome. Our case series highlights the combination of clinical signs and diagnostic evidence that can help distinguish DAH from alternative diagnoses. Further research may help elucidate the underlying pathophysiological mechanisms of DAH, particularly following hematopoietic stem cell transplantation, and unlock prospects for improved diagnosis and treatment.

References

1. Lichtenberger JP III, Digumarthy SR, Abbott GF, et al. Diffuse pulmonary hemorrhage: clues to the diagnosis. Curr Probl Diagn Radiol. 2014;43(3):128-139. 2. Escuissato DL, Warszawiak D, Marchiori E. Differential diagnosis of diffuse alveolar haemorrhage in immunocompromised patients. Curr Opin Infect Dis. 2015;28(4):337-342. 3. Sharma SK, Kumar S, Singh AK, et al. Diffuse alveolar hemorrhage following allogeneic peripheral blood stem cell transplantation: a case report and a short review. Indian J Hematol Blood Transfus. 2014;30(1):41-44. 4. Lara AR, Schwarz MI. Diffuse alveolar hemorrhage. Chest. 2010;137(5):1164-1171. 5. Rabe C, Appenrodt B, Hoff C, et al. Severe respiratory failure due to diffuse alveolar hemorrhage: clinical characteristics and outcome of intensive care. J Crit Care. 2010;25(2):230-235. 6. Schwarz MI, Whitcomb ME, Goldman AL. The spectrum of diffuse pulmonary infiltration in malignant disease. Chest. 1973;64(1):88-93. 7. Lewis ID, DeFor T, Weisdorf DJ. Increasing incidence of diffuse alveolar hemorrhage following allogenic bone marrow transplantation. Bone Marrow Transplant. 2000;26(5):539-543. 8. Lassence AD, Fleury-feith J, Escudier E, et al. Alveolar hemorrhage: diagnostic criteria and results in 194 immunocompromised hosts. Am J Respir Crit Care Med. 1995;151(1):157-163. 9. Tada K, Kurosawa S, Hiramoto N, et al. Stenotrophomonas maltophilia infection in hematopoietic SCT recipients: high mortality due to pulmonary hemorrhage. Bone Marrow Transplant. 2013;48(1):74-79. 10. Gulati S, Gulati A. Pulmonary manifestations of leptospirosis. Lung India. 2012;29(4):347-353. 11. Marruchella A, Corpolongo A, Tommasi C, et al. A case of pulmonary tuberculosis presenting as diffuse alveolar haemorrhage: is there a role for anticardiolipin antibodies? BMC Infect Dis. 2010;10:33. 12. Kong MX, Newman K, Goldenberg R, et al. Fatal mycoplasma pneumoniae infection: case report and review of the literature. N A J Med Sci. 2012;5(2):126-130. 13. Sundar KM, Pearce MJ. Diffuse alveolar hemorrhage due to Legio- nella pneumonia. Sarcoidosis Vasc Diffuse Lung Dis. 2004;21(2):158-159. 14. Gulati S, Maheshwari A. Atypical manifestations of dengue. Trop Med Int Health. 2007;12(9):1087-1095. 15. Gasparetto EL, Ono SE, Escuissato D, et al. Cytomegalovirus pneumonia after bone marrow transplantation: high resolution CT findings. Br J Radiol. 2004;77(921):724-727. 16. Kennedy ED, Roy M, Norris J, et al; 2009 Pandemic H1N1 Influen- za-Associated Lower Respiratory Tract Hemorrhage Working Group. Low- er respiratory tract hemorrhage associated with 2009 pandemic influenza A (HINI) virus infection. Influenza Other Respir Viruses. 2013;7(5):761-765.

July 2016, Vol. 23, No. 3 Cancer Control 277 Case Series

Marijuana Smoking in Patients With Leukemia Sara Khwaja, MD, Abraham Yacoub, MD, Asima Cheema, MD, Nancy Rihana, MD, Robin Russo, MD, Ana Paula Velez, MD, Sowmya Nanjappa, MBBS, MD, Ramon L. Sandin, MD, Chandrashekar Bohra, MBBS, Ganesh Gajanan, MBBS, and John N. Greene, MD

Summary: Worldwide, marijuana (cannabis) is a widely used drug. The incidence of marijuana smoking is increasing and is second only to tobacco as the most widely smoked substance in the general population. It is also the second most commonly used recreational drug after alcohol. Some adverse effects of marijuana smoking have been documented; however, the number of studies on the pulmonary effects of marijuana in individuals with leukemia is limited. In our case series, we report on 2 men with acute myeloid leukemia with miliary nodular lung patterns on computed tomography of the chest due to heavy marijuana use. We also report on 2 patients with acute lymphocytic leukemia who had a history of smoking marijuana and then developed lung opacities consistent with mold infection.

Introduction in the 2000s.8 The most common method of marijua- Use of marijuana (cannabis) is common by many peo- na use is smoking, either as a rolled cigarette (joint), ple worldwide.1 The incidence of marijuana smok- through a water-filled pipe (bong), or the hookah, with ing is increasing, second only to tobacco as the most coal used to vaporize water and then mixed with to- widely smoked substance in the general population.2 bacco. The bong is considered the most harmful meth- It is also the second most commonly used recreation- od.9 Marijuana can be inhaled up to the peak inspira- al drug after alcohol.3 As of publication, 23 states plus tion, holding the hot fumes for as long as possible prior the District of Columbia allow marijuana to be used as to slow exhalation. Such smoking technique results a medicinal drug and 4 states have passed measures in a greater deposition of toxic substances, such as tar to make its recreational use legal.4,5 However, the US and carbon monoxide, as well as damage to the lung Drug Enforcement Administration still considers mari- parenchyma typically seen in those who smoke tobac- juana to be a schedule 1 substance.6 Marijuana is a co cigarettes.10-12 greenish-gray mixture of the dried, shredded leaves, The gaseous and particulate composition of mari- stems, seeds, and flowers of Cannabis sativa. Prepara- juana is similar to tobacco cigarettes, except that the tions of C sativa have been used for their euphoric ef- active component in tobacco is nicotine (compared fects.7 The most psychoactive constituent compound with THC in marijuana).13 An accurate study of ad- in marijuana is tetrahydrocannabinol (THC), which is verse effects of marijuana is difficult to perform due to rapidly absorbed from the lungs and binds to endog- its illegal status in many states, various smoking tech- enous cannabinoid receptors in the central nervous niques, and its shorter duration of use compared with system. tobacco.14 In addition, tobacco smoking is frequently The potency of THC in marijuana has risen from a confounding factor in marijuana studies, although approximately 4% in the early 1980s to as much as 30% some pulmonary adverse effects of marijuana smoking have been documented and include cough, dyspnea, bronchitis, pneumomediastinum, spontaneous pneu- From the Departments of Infectious Diseases (AY, JNG) and Internal mothorax, and apical lung bullae.15-18 Medicine (SN), Clinical Microbiology and Virology Laboratories Marijuana mold contamination has also been re- (RLS), H. Lee Moffitt Cancer Center & Research Institute (SK, AC, 19 CB, GG), Departments of Internal Medicine (NR, RR, APV, JNG) and ported. The association between nodular pneumonia Oncologic Sciences (SN, RLS, JNG), Division of Infectious Diseases and bronchiolitis, with its associated miliary micronod- and International Medicine (NR, APV), University of South Florida ular pattern, and smoking marijuana has been de- Morsani College of Medicine (SK, AC), Tampa, Florida. scribed in the literature.20 However, the effect of mari- Address correspondence to John N. Greene, MD, Department of Infectious Diseases and Tropical Medicine, Moffitt Can- juana smoking in immunocompromised populations cer Center, 12902 Magnolia Drive, Tampa, FL 33612. has not been studied, except for a few case reports.20 E-mail: [email protected] Submitted May 16, 2015; accepted March 15, 2016. Case Series No significant relationships exist between the authors and the companies/organizations whose products or services may be We describe 4 cases of nodular lung lesions seen on referenced in this article. computed tomography (CT) of the chest in patients

278 Cancer Control July 2016, Vol. 23, No. 3 with leukemia who smoked marijuana. Two patients His last use of marijuana was 1 day prior to his had acute lymphocytic leukemia (ALL) and 2 had acute hospital admission. myelogenous leukemia (AML). The patients with ALL Findings on CT of the chest showed a nodular responded well to treatment and their nodular lung le- consolidation pattern consistent with fungal infection sions resolved with antifungal therapy. Both patients (Fig 2). He was treated with oral voriconazole 200 mg with AML died due to disseminated fungal infection. twice a day for 3 months. Following treatment com- The Table summarizes the clinical presentation, pletion, the nodular lesion resolved, as evidenced on treatment, and outcomes of each of the 4 case patients. follow-up CT.

Figure 1. Acute Lymphocytic Leukemia Case 1: A 27-year-old man with a history of ALL and marijuana and cigarette smoking developed neutropenia while on chemotherapy and was started on antibiotic prophylaxis (levofloxacin) and fluconazole. He was discharged home with neutropenia. He continued to smoke cigarettes and marijuana. On a second visit, he was admitted to the hospital with neutropenic fever. CT of the chest was performed and demonstrated a new 1.08-cm cavitary nodule (Fig 1). Findings on bronchoscopy, Gram stain, fungal stain, acid-fast bacilli smear, viral polymerase chain re- Fig 1. — Computed tomography of the chest reveals a 1.08-cm cavitary action, and the respective cultures were all negative. He nodule that responded to voriconazole treatment. was treated with voriconazole 300 mg twice a day for

3 months and micafungin 150 mg every day for 6 weeks. CT findings obtained 2 months later showed the nodule dramatically decreased in size from 1.08 to 0.1 cm. Case 2: A 20-year-old man presented with B-cell ALL. One week after receiving chemotherapy, he presented with bilateral eye pain, photophobia, sore throat, nonproductive cough, rhinorrhea, and low-grade fever for 5 days. He had been in contact with his niece who had upper respiratory tract infection with cough. He missed his dose of prophylactic levofloxacin the previous night. Except for bilateral erythema of his eyes, nasal congestion, fever (100.1 °F), and sore throat, findings on the remainder of his physical examination were normal. Laboratory studies revealed a white blood cell (WBC) count of 80 cells/mm3; all other values were normal. A respiratory viral panel from nasopharyngeal sampling was positive for human metapneumovirus. Chest x-ray revealed no evidence of pneumonia. HIV Fig 2. — Computed tomography of the chest showed a nodular pneumo- testing was negative. nia infiltrate consistent with mold infection.

Table. — Summary of the Case Series Case No. Malignancy Clinical Signs Antifungal Antifungal Treatment Outcome Prophylaxis 1 ALL Neutropenic fever Fluconazole /levofloxacin Voriconazole/micafungin Favorable 2 ALL None Fluconazole /levofloxacin Voriconazole Favorable 3 AML None Posaconazole Liposomal amphotericin B Death Voriconazole 4 AML Neutropenic fever Posaconazole Liposomal amphotericin B Death Micafungin Voriconazole ALL = acute lymphocytic leukemia, AML = acute myelogenous leukemia.

July 2016, Vol. 23, No. 3 Cancer Control 279 Acute Myelogenous Leukemia Case 3: A 36-year-old man with a history of AML presented with declining neutrophil levels. He was in re- mission after several cycles of chemotherapy, and he had received a matched unrelated donor allogeneic stem cell transplant. The prior chemotherapy course was complicated by Streptococcus mitis bacteremia with acute respiratory distress syndrome and septic shock. His history was significant for smoking 1 pack of cigarettes per day for 22 years and smoking marijuana up until his hospital admission. He also had a history of alcohol abuse (quit 11 years ago) and a history of heroin, cocaine, and methamphetamine abuse (stopped 12 years ago). CT of the chest demonstrated ground-glass pneumonia that resolved within a few weeks with cortico- steroid therapy. Findings on CT at a previous follow-up visit were normal. Results from respiratory viral and Fig 3. — Computed tomography of the chest demonstrates a miliary reticulonodular pattern presumed to be due to bronchiolitis related to atypical pneumonia panels of a nasopharyngeal speci- smoking marijuana. men were negative, as was a sputum culture. A miliary nodular pattern was noted on repeat CT and was most likely due to his heavy marijuana use (Fig 3). A review of systems was unremarkable and findings on physical examination were normal. His white blood cell count was 5280 cells/mm3 and a lymphocyte count was 560 cells/mm3. Results on serum galactomannan testing were positive, but repeat testing was negative. Findings on pulmonary function testing were normal. Cardiac testing revealed an ejection fraction of 57%. Despite taking posaconazole as prophylaxis, he died 2 months later from disseminated fungal infection. The infection was demonstrated on a histopathology examination from a skin lesion but was not able to grow on culture. Fig 4. — Computed tomography of the chest showed ground-glass Case 4: A 53-year-old man with relapsed AML was pneumonia and a miliary pattern due to bronchiolitis related to smoking marijuana. admitted to the hospital for reinduction chemotherapy. A review of systems was unremarkable, and findings on physical examination were normal. This patient Discussion had a history of smoking 1 pack of cigarettes a day for Marijuana, hash, and hash oil are the 3 main forms of 22 years and smoking marijuana twice a week for the cannabis. Cannabis contains more than 500 chemi- same duration. Laboratory studies were ordered and cals.21 Both natural and synthetic cannabinoids are revealed a white blood cell count of 3340 cells/mm3 thought to possess neuroprotective properties. The im- and a platelet count of 73,000 cells/mm3. pact of marijuana smoking on respiratory health has Findings on CT of the chest revealed ground-glass some similarities to that of tobacco smoking14; however, pneumonia and a miliary pattern due to bronchiolitis studies conducted on marijuana smoking are subject (negative on serum galactomannan testing; Fig 4). He to confounding use of tobacco as well as other social received chemotherapy and posaconazole prophylax- factors.3,14,22,23 is, but he later developed neutropenic fever. He also Marijuana is commonly used by adolescents and developed altered sensorium initially thought to be young adults.21 The incidence of marijuana smoking caused by use of pain medication or due to dissemi- has increased in the western world, possibly because nated Fusarium infection from an injury to his right of the euphoric effects of cannabis and the loosening great toe and his right second toe, revealing eschar-like legal restrictions.4,21 Rarely, marijuana use can trigger lesions. His medication was switched to amphotericin acute myocardial infarction by inducing coronary ar- B and later to voriconazole and high-dose micafungin. tery vasospasm24; marijuana use may also be associ- However, several weeks later the patient succumbed to ated with atrial fibrillation.25 Marijuana smoking has disseminated Fusarium infection. been implicated as a causative factor in traditionally

280 Cancer Control July 2016, Vol. 23, No. 3 tobacco-related tumors of head, neck,26 and lung,27 as pattern is comparable with respiratory bronchiolitis well as in cases of transitional cell carcinoma.28 commonly seen among those who smoke cigarettes. Tobacco smoking is also a recognized behavior- Respiratory bronchiolitis is characterized by accumu- al risk factor for periodontal disease, and marijuana lation of tan- or yellow-pigmented, “smoker” macro- smoking likely contributes to this risk in a similar fash- phages in the bronchiolar lumens, with associated ion.29 Daily use of marijuana is associated with moder- chronic inflammation and fibrosis extending from re- ate to severe liver fibrosis, and its misuse may be a risk spiratory bronchioles to the alveolar wall.44,45 This dis- factor for cerebrovascular disease — this is particularly tinct finding of miliary micronodular imaging should true among young people.30,31 Marijuana use may ad- alert physicians to include synthetic marijuana abuse versely affect the mucous membranes and the skin, in the differential diagnosis.46 as well as cause Raynaud phenomenon and arteritis.32 Regular smoking of marijuana may cause alveo- Long-term, heavy cannabis use has also been associat- lar macrophage damage, and the immune systems ed with cognitive dysfunction, leading to attention and of individuals who smoke marijuana on a regular ba- memory impairment.33,34 sis may be unable to fight against inhaled pathogens Marijuana smoke contains up to twice as many such as bacteria, fungi, or cease the growth of tumor polyaromatic hydrocarbons as tobacco smoke.27 These cells.47 Marijuana use is also known to increase the ac- hydrocarbons, along with other carcinogenic com- cumulation of alveolar macrophages, which trigger pounds found in marijuana smoke and the technique an inflammatory response.45 Cannabis may also cause of marijuana smoking, suggest that an associated risk squamous metaplasia in the bronchial tree, thus in- of malignancy may be present. Numerous case reports, creasing an individual’s susceptibility to infection and cohort, and case-control studies have examined the ef- cancer. In addition, cannabis may inhibit the produc- fect of marijuana smoking on the risk of malignancies tion or function of cytokines, possibly leading to the but with variable results.35-38 inability of the immune system to fight infections.43 One cohort study looked at such a link for 40 years.39 Because THC is a general “immunosuppressant” The authors examined approximately 50,000 Swedish affecting macrophages, natural killer cells, and T cells, military recruits during the years 1969 and 1970 who its use as a medicinal agent in patients with preexisting were between 18 and 20 years of age and who also immune deficits might have serious consequences.48 used marijuana.39 The cohort was followed until 2009. Several infectious complications from marijuana smok- After taking into account the baseline use of marijuana ing have been reported.49 In 1 case, a 46-year-old man and alcohol, as well as various respiratory and socio- who was immunocompetent presented with a 1-week economic conditions, the study authors concluded that history of fever, headache, cough, dyspnea, and with the study participants who were heavy smokers had evidence of a healed ulcer on his left upper lip.50 Find- more than a twofold risk of developing lung cancer.39 ings on transbronchial biopsy supported a diagnosis of Use of marijuana as a medicinal agent is also gain- herpes simplex virus–related bronchiolitis and pneu- ing in popularity among individuals with cancer to treat monitis due to long-term marijuana use.50 This study il- neuropathic pain and to stimulate appetite due to the lustrated marijuana-induced squamous metaplasia that potent, antiemetic properties of cannabis.40,41 One his- promoted replication of the herpes simplex virus in the toric report from 1997 concluded that nearly one-third lower respiratory tract. of clinical oncologists surveyed supported the use of Results of an Australian study suggest that tubercu- marijuana as an antiemetic agent.42 Another survey from losis could be spread after sharing a bong with an indi- Colorado, where recreational marijuana use has become vidual with pulmonary tuberculosis.51 Of the 45 people legal and more prevalent, concluded that approximate- who shared the same bong, 29 tested positive for latent ly one-fifth of physicians support use of medical mari- tuberculosis; several cases of active tuberculosis also juana, and more than 60% believe that it poses serious occurred.51 On typing of Mycobacterium tuberculosis, mental and physical health problems.43 all of the isolates were identical.51 Regarding the pulmonary effects of marijuana, the Marijuana can also become contaminated with vari- relationship between marijuana smoking and respira- ous molds such as Aspergillus, Mucorales, and Fusar- tory complications is poorly understood. Short-term ium.52 In a single study, marijuana was cultured and its exposure has been associated with bronchodilation, mold load was compared with that of tobacco.53 The re- and adverse effects of long-term marijuana smoking insult was significant: 100,000 colony-forming units of mold clude cough, dyspnea, chronic bronchitis, spontaneous on marijuana vs 200 colony-forming units on tobacco.53 pneumothorax, and apical lung bullae.10-12 Marijuana smoking has been linked to the devel- A miliary nodular pattern, similar to what is seen opment of invasive pulmonary aspergillosis in immu- in cases of miliary tuberculosis and disseminated fun- nocompromised individuals, including those with can- gal infection, has also been seen in individuals who cer and neutropenia. One of the earliest reports found use marijuana44; this was true for 2 of our cases. The of pulmonary aspergillosis due to inhaling marijuana

July 2016, Vol. 23, No. 3 Cancer Control 281 was in a patient with chronic granulomatous disease; nodeficiency, although other risk factors and lifestyle it was published in 1975.54 Following this report, the practices may contribute to this pulmonary pathology. development of bronchopulmonary aspergillosis due Marijuana smoking is associated with fungal pneumo- to smoking marijuana was highlighted again in 1978.55 nia due to the possible inhalation of mold spores on Subsequently, case reports of pulmonary aspergillosis the leaves of the marijuana plant. This is because mari- in individuals with cancer or those needing transplan- juana cigarettes typically do not have a filter to prevent tation have been reported and include small-cell lung spore inhalation, which is generally in contrast to ciga- cancer,56 colorectal cancer,57 leukemia,58 and bone mar- rettes. Marijuana can also damage the lungs and impair row59 and renal transplants.60 local lung immune responses, thus possibly predispos- Other fungal infections directly or indirectly ing those who smoke marijuana to invasive pulmonary linked to smoking include cryptococcosis, coccidi- fungal infection. Bronchiolitis may also occur in those oidomycosis, paracoccidioidomycosis, and penicil- who smoke marijuana because the small airways of liosis. These infections have been primarily related the lungs may become irritated when hot smoke is to smoking tobacco, but marijuana may also play inhaled — this can also occur with smoking tobacco, a role. Although no reports have been published, in- breathing air pollution and other types of smoke, and gestion of fungally contaminated marijuana is a cause being exposed to chemical irritants. for concern among persons who are immunocompro- The findings from this case series highlight the mised. However, other relevant examples have been need to consider antifungal prophylaxis and cessa- published: Following a bone marrow transplant, 1 pa- tion of marijuana use among patients with prolonged tient developed hepatic Mucor infection after ingesting neutropenia and concomitant marijuana use. In addi- naturopathic medications, and another individual with tion to miliary tuberculosis and endemic mycosis, inva- neutropenia developed disseminated fusariosis after sive fungal infections possibly related to cannabis use eating cereal contaminated with the pathogen.49 should also be included in the differential diagnosis of Confirming the diagnosis of invasive fungal in- those who demonstrate a nodular pattern and repre- fection in severely immunocompromised patients resent a susceptible host. mains a challenge for infectious diseases specialists. Coagulation abnormalities and coexisting thrombocy- References topenia may preclude invasive procedures for tissue 1. World Health Organization. Management of substance abuse: cannabis. http://www.who.int/substance_abuse/facts/cannabis/en/. Accessed harvesting, and sputum cultures are neither sensitive June 16, 2016. nor specific. In a setting where cultures may be diffi- 2. Lange P. Cannabis and the lung. Thorax. 2007;62(12):1036-1037. 3. Moore BA, Augustson EM, Moser RP, et al. Respiratory effects cult to obtain, a non–culture-based diagnostic modal- of marijuana and tobacco use in a U.S. sample. J Gen Intern Med. ity, such as CT, can be used as an alternative approach 2005;20(1):33-37. 61,62 4. Motel S; Pew Research Center. 6 facts about marijuana. Published based on revised definitions and guidelines. Imag- April 14, 2015. http://www.pewresearch.org/fact-tank/2015/04/14/6-facts ing data support using CT of the chest for the early de- -about-marijuana/. Accessed June 16, 2016. 5. Office of National Drug Control Policy. Marijuana resource center: tection of invasive pulmonary aspergillosis in severely state laws related to marijuana. https://www.whitehouse.gov/ondcp/state immunocompromised patients.61,62 Most patients (94%) -laws-related-to-marijuana. Accessed June 16, 2016. 6. US Drug and Enforcement Agency. Drug scheduling. presented with at least 1 macronodule (nodule diam- https://www.dea.gov/druginfo/ds.shtml. Accessed June 16, 2016. eter ≥ 1 cm) on CT of the chest.63 In a study conducted 7. National Institute on Drug Abuse. Marijuana. Revised February 2016. 64 https://www.drugabuse.gov/drugs-abuse/marijuana. Accessed June 16, 2016. by Marchiori et al, findings on CT helped to differen- 8. Mehmedic Z, Chandra S, Slade D, et al. Potency trends of Δ9-THC tiate between invasive fungal infection and organizing and other cannabinoids in confiscated cannabis preparations from 1993 to 2008. J Forensic Sci. 2010;55(5):1209-1217. pneumonia. 9. Gill A. Bong lung: regular smokers of cannabis show relatively dis- Prophylaxis, empirical therapy, and treatment for tinctive histologic changes that predispose to pneumothorax. Am J Surg Pathol. 2005;29(7):980-982. probable or proven fungal infection are 3 strategies for 10. Aldington S, Williams M, Nowitz M, et al. Effects of cannabis on pul- the prevention and treatment of individuals at high risk monary structure, function and symptoms. Thorax. 2007;62(12):1058-1063. 65 11. Johnson MK, Smith RP, Morrison D, et al. Large lung bullae in mari- for fungal infection. Invasive fungal infection pro- juana smokers. Thorax. 2000;55(4):340-342. duced by yeasts and molds is the main infectious cause 12. Hii SW, Tam JD, Thompson BR, et al. Bullous lung disease due to marijuana. Respirology. 2008;13(1):122-127. of death in patients with hematological malignancies; 13. Tashkin D P. Airway effects of marijuana, cocaine, and other inhaled therefore, antifungal prophylaxis is recommended.66,67 illicit agents. Curr Opin Pulm Med. 2001;7(2):43-61. 14. Howden ML, Naughton MT. Pulmonary effects of marijuana inhalation. Clinical practice guidelines from the Infectious Diseas- Expert Rev Respir Med. 2011;5(1):87-92. es Society of America62 recommend intravenous or oral 15. Miller WE, Spiekerman RE, Hepper NG. Pneumomediastinum resulting from performing Valsalva maneuvers during marihuana smoking. voriconazole as first-line therapy for invasive pulmo- Chest. 1972;62(2):233-234. nary aspergillosis. 16. Phan TD, Lau KK, Li X. Lung bullae and pulmonary fibrosis associated with marijuana smoking. Australas Radiol. 2005;49(5):411-414. 17. Tan WC, Lo C, Jong A, et al; Vancouver Burden of Obstructive Lung Conclusions Disease (BOLD) Research Group. Marijuana and chronic obstructive lung disease: a population-based study. CMAJ. 2009;180(8):814-820. Marijuana smoking can be dangerous and life threat- 18. Feldman AL, Sullivan JT, Passero MA, et al. Pneumothorax in poly- ening in the setting of neutropenia and T-cell immu- substance-abusing marijuana and tobacco smokers: three cases. J Subst

282 Cancer Control July 2016, Vol. 23, No. 3 Abuse. 1993;5(2):183-186. 2011;3(1):e2011005. 19. Kagen SL. Aspergillosis: an inhalable contaminant of marijuana. 53. Verweij PE, Kerremans JJ, Voss A, et al. Fungal contamination of N Engl J Med. 1981;304(8):483-484. tobacco and marijuana. JAMA. 2000;284(22):2875. 20. Ruchlemer R, Amit-Kohn M, Raveh D, et al. Inhaled medicinal 54. Chusid MJ, Gelfand JA, Nutter C, et al. Pulmonary aspergillosis, cannabis and the immunocompromised patient. Support Care Cancer. inhalation of contaminated marijuana smoke, chronic granulomatous dis- 2015;23(3):819-822. ease. Ann Intern Med. 1975;82(5):682-683. 21. American Society of Addiction Medicine. Marijuana use fact sheet. 55. Llamas R, Hart DR, Schneider NS. Allergic bronchopulmonary asper- http://www.asam.org/docs/default-source/default-document-library/marijuana gillosis associated with smoking moldy marihuana. Chest. 1978;73(6):871-872. -use-fact-sheet.pdf?sfvrsn=0. Accessed June 16, 2016. 56. Sutton S, Lum BL, Torti FM. Possible risk of invasive aspergillosis 22. Versteeg PA, Slot DE, van der Velden U, et al. Effect of cannabis with marijuana use during chemotherapy for small cell lung cancer. Drug usage on the oral environment: a review. Int J Dent Hyg. 2008;6(4):315-320. Intell Clin Pharm. 1986;20(4):289-291. 23. Coffman KL. The debate about marijuana usage in transplant can- 57. Cescon DW, Page AV, Richardson S, et al. Invasive pulmonary didates: recent medical evidence on marijuana health effects. Curr Opin aspergillosis with marijuana use in a man with colorectal cancer. J Clin Organ Transplant. 2008;13(2):189-195. Oncol. 2008;26(13):2214-2215. 24. Fisher BA, Ghuran A, Vadamalai V, et al. Cardiovascular complica- 58. Szyper-Kravitz M, Lang R, Manor Y, et al. Early invasive pulmonary tions induced by cannabis smoking: a case report and review of the literature. aspergillosis in a leukemia patient linked to aspergillus contaminated mari- Emerg Med J. 2005;22(9):679-680. juana smoking. Leuk Lymphoma. 2001;42(6):1433-1437. 25. Korantzopoulos P, Liu T, Papaioannides D, et al. Atrial fibrillation 59. Hamadeh R, Ardehali A, Locksley RM, et al. Fatal aspergillosis as- and marijuana smoking. Int J Clin Pract. 2008;62(2):308-313. sociated with smoking contaminated marijuana, in a marrow transplant 26. Donald PJ. Marijuana smoking--possible cause of head and neck car- recipient. Chest. 1988;94(2):432-433. cinoma in young patients. Otolaryngol Head Neck Surg. 1986;94(4):517-521. 60. Marks WH, Florence L, Lieberman J, et al. Successfully treated in- 27. Reid PT, Macleod J, Robertson JR. Cannabis and the lung. J R Coll vasive pulmonary aspergillosis associated with smoking marijuana in a Physicians Edinb. 2010;40(4):328-323. renal transplant recipient. Transplantation. 1996;61(12):1771-1774. 28. Chacko JA, Heiner JG, Siu W, et al. Association between marijuana 61. De Pauw B, Walsh TJ, Donnelly JP, et al. Revised definitions of use and transitional cell carcinoma. Urology. 2006;67(1):100-104. invasive fungal disease from the European Organization for Research and 29. Thomson WM, Poulton R, Broadbent JM, et al. Cannabis smoking Treatment of Cancer/Invasive Fungal Infections Cooperative Group and the and periodontal disease among young adults. JAMA. 2008;299(5):525-531. National Institute of Allergy and Infectious Diseases Mycoses Study Group 30. Ishida JH, Peters MG, Jin C,, et al. Influence of cannabis use on se- (EORTC/MSG) Consensus Group. Clin Infect Dis. 2008;46(12):1813-1821. verity of hepatitis C disease. Clin Gastroenterol Hepatol. 2008;6(1):69-75. 62. Walsh TJ, Anaissie EJ, Denning DW, et al. Treatment of aspergil- 31. Mateo I, Pinedo A, Gomez-Beldarrain M, et al. Recurrent losis: clinical practice guidelines of the Infectious Diseases Society of stroke associated with cannabis use. J Neurol Neurosurg Psychiatry. America. Clin Infect Dis. 2008;46(3):327-360. 2005;76(3):435-437. 63. Greene RE, Schlamm HT, Oestmann JW, et al. Imaging findings 32. Tennstedt D, Saint-Remy A. Cannabis and skin diseases. Eur J in acute invasive pulmonary aspergillosis: clinical significance of the halo Dermatol. 2011;21(1):5-11. sign. Clin Infect Dis. 2007;44(3): 373-379. 33. Solowij N, Stephens R, Roffman RA, et al. Does marijuana use 64. Marchiori E, Marom EM, Zanetti G, et al. Reversed halo sign in inva- cause long-term cognitive deficits? JAMA. 2002;287:2653-2654. sive fungal infections: criteria for differentiation from organizing pneumonia. 34. McHale S, Hunt N. Executive function deficits in short-term abstinent Chest. 2012;142(6):1469-1473. cannabis users. Hum Psychopharmacol. 2008;23(5):409-415. 65. Cumbo TA, Segal BH. Prevention, diagnosis, and treatment of in- 35. Aldington S, Harwood M, Cox B, et al. Cannabis use and risk of lung vasive fungal infections in patients with cancer and neutropenia. J Natl cancer: a case-control study. Eur Respir J. 2008;31(2):280-286. Compr Canc Netw. 2004;2(5):455-69. 36. Mehra R, Moore B A, Crothers K, et al. The association between 66. Villafuerte-Gutierrez P, Villalon L, Losa JE, et al. Treatment of fe- marijuana smoking and lung cancer: a systematic review. Arch Intern Med. brile neutropenia and prophylaxis in hematologic malignancies: a critical 2006;166(13):1359-1367. review and update. Adv Hematol. 2014;2014:986938. 37. Berthiller J, Straif K, Boniol M, et al. Cannabis smoking and risk of 67. Glockner A. Treatment and prophylaxis of invasive candidiasis with lung cancer in men: a pooled analysis of three studies in Maghreb. J Thorac anidulafungin, caspofungin and micafungin: review of the literature. Eur J Oncol. 2008;3(12):1398-1403. Med Res. 2011;16(4):167-179. 38. Brambilla C, Colonna M. Cannabis: the next villain on the lung cancer battlefield? Eur Respir J. 2008;31(2):227-228. 39. Callaghan RC, Allebeck P, Sidorchuk A. Marijuana use and risk of lung cancer: a 40-year cohort study. Cancer Causes Control. 2013;24(10):1811-1820. 40. Grant I, Atkinson J H, Gouaux B, et al. Medical marijuana: clearing away the smoke. Open Neurol J. 2012;6:18-25. 41. Tramer MR, Carroll D, Campbell FA, et al. Cannabinoids for control of chemotherapy induced nausea and vomiting: quantitative systematic review. BMJ. 2001;323(7303):16-21. 42. Schwartz RH, Voth EA, Sheridan MJ. Marijuana to prevent nausea and vomiting in cancer patients: a survey of clinical oncologists. South Med J. 1997;90(2):167-172. 43. Kondrad E, Reid A. Colorado family physicians’ attitudes toward medical marijuana. J Am Board Fam Med. 2013;26(1):52-60. 44. Heyneman LE, Ward S, Lynch DA et al. Respiratory bronchiolitis, respiratory bronchiolitis-associated interstitial lung disease, and desqua- mative interstitial pneumonia: different entities or part of the spectrum of the same disease process? Am J Roentgenol. 1999;173(6):1617. 45. Baldwin G C, Tashkin D P, Buckley D M, et al. Marijuana and cocaine impair alveolar macrophage function and cytokine production. Am J Respir Crit Care Med.1997;156(5):1606-1613. 46. Berkowitz EA, Henry TS, Veeraraghavan S, et al. Pulmonary effects of synthetic marijuana: chest radiography and CT findings. AJR Am J Roentgenol. 2015;204(4):750-757 47. Tetrault JM, Crothers K, Moore B A, et al. Effects of marijuana smoking on pulmonary function and respiratory complications: a systematic review. Arch Intern Med. 2007;167(3):221-228. 48. Tashkin DP, Baldwin GC, Sarafian T, et al. Respiratory and Immuno- logic consequences of marijuana smoking. J Clin Pharmacol. 2002; 42(11 suppl):71S-81S. 49. Sipsas NV, Kontoyiannis DP. Occupation, lifestyle, diet, and invasive fungal infections. Infection. 2008:36(6):515-525. 50. Libraty DH, Bocelli L, Fraire A. Herpes simplex virus bronchiolitis in a cannabis user. IDCases. 2013;1(1):2-4. 51. Thu K, Hayes M, Miles S, et al. Marijuana ‘bong’ smoking and tuberculosis. Intern Med J. 2013;43(4):456-458. 52. Gargani Y, Bishop P, Denning DW. Too many mouldy joints: marijuana and chronic pulmonary aspergillosis. Mediterr J Hematol Infect Dis.

July 2016, Vol. 23, No. 3 Cancer Control 283 Special Report

Platelet-to-Lymphocyte Ratio and Use of NSAIDs During the Perioperative Period as Prognostic Indicators in Patients With NSCLC Undergoing Surgery Brenda M. Lee, MD, Andrea Rodríguez, MD, Gabriel Mena, MD, Vijaya Gottumukkala, MB, MD, Reza J. Mehran, MD, David C. Rice, MB, Lei Feng, MS, Jun Yu, MS, and Juan P. Cata, MD

Background: Hematological biomarkers of inflammation such as the neutrophil-to-lymphocytic rate have been reported as predictors of survival in a variety of cancers. The aim of the present study was to investigate the prognostic value of the perioperative platelet-to-lymphocyte ratio in patients with non–small-cell lung cancer (NSCLC) and to elucidate the effects of the perioperative use of nonsteroidal anti-inflammatory drugs (NSAIDs) on tumor recurrence and survival in patients undergoing surgical resection for NSCLC. Methods: This retrospective study included data from 1,637 patients who underwent surgical resection for stage I, II, or III NSCLC. Perioperative data and tumor-related variables were included. Univariate and multivariable Cox proportional hazard ratio (HR) models were used to evaluate the association between perioperative platelet-to-lymphocyte ratio and NSAID use on recurrence-free survival (RFS) and overall survival (OS). Results: Multivariate analysis showed that a preoperative platelet-to-lymphocyte ratio of at least 180 was associated with reduced rates of RFS (HR = 1.22; 95% confidence interval [CI], 1.03–1.45; P = .019) and OS (HR = 1.33; 95% CI, 1.10–1.62; P = .004). Perioperative use of NSAIDs showed no statistically significant changes in RFS and OS rates (P = .72 and P = .44, respectively). Conclusions: A higher preoperative inflammatory status is associated with decreased rates of RFS and OS in patients with NSCLC undergoing curative surgery. Perioperative use of NSAIDs was not found to be an independent predictor of survival.

Introduction Current risk-stratification characteristics used in Lung cancer is the leading cause of cancer-related patients with NSCLC are associated with surgical stag- death in both men and women in the United States.1 ing and include tumor size, invasion to adjacent or- Non–small cell lung cancer (NSCLC) makes up 83% of gans, nodal status, and presence of metastasis. Some lung cancer cases,2 and it has a 5-year overall survival clinically available inflammatory biomarkers have (OS) rate of 21% when considering all stages together.1 been identified to improve and guide therapy in pa- Surgery remains the first-line treatment for the major- tients with cancer, particularly those with NSCLC.5 We ity of patients with stage I or II NSCLC, and most pa- previously reported the importance of the preopera- tients with stage III or IV NSCLC will typically receive tive neutrophil-to-lymphocyte ratio in the context of chemotherapy with or without radiotherapy.2 There- NSCLC, indicating that those with a high preopera- fore, great effort has been placed on investigating peri- tive inflammatory status have an elevated mortality operative factors that could predict and/or improve the risk; however, other investigated markers, including poor survival rates of this patient population.3,4 the platelet-to-lymphocyte ratio, have been shown to correlate with poor prognosis in various types of cancer.6-9 The interaction between platelets and cancer From the Department of Surgery (BML), Johns Hopkins Hospital, cells, as well as between platelets and immune cells, Baltimore, Maryland, Departments of Anesthesiology and Peri- operative Medicine (AR, GM, VG, JPC), Thoracic & Cardiovascular has previously been described, and platelets may me- Surgery (RJM, DCR), and Biostatistics (LF, JY), University of diate tumor growth by stimulating angiogenesis and Texas MD Anderson Cancer Center, Houston, Texas, and Anesthesi- inducing immunosuppression through the release of ology and Surgical Oncology Research Group (JPC), Texas. growth factors.10,11 Address correspondence to Juan P. Cata, MD, Department of Anesthesiology and Perioperative Medicine, MD Anderson The potential link between inflammation and can- Cancer Center, 1515 Holcombe Boulevard, Houston, TX 77030. cer has been previously studied, and findings suggest E-mail: [email protected] that an inflammatory microenvironment promotes Submitted September 30, 2015; accepted February 19, 2016. tumor growth, progression, and immunosuppres- No significant relationships exist between the authors and the com- 12,13 panies/organizations whose products or services may be referenced sion. Understanding cancer-associated inflamma- in this article. tion is crucial because this information can be used to This work is supported by NIH/NCI award No. P30CA016672. predict survival in most patient with cancer, including

284 Cancer Control July 2016, Vol. 23, No. 3 those with NSCLC.5 Similarly, use of nonsteroidal anti- as the platelet count divided by the lymphocyte count. inflammatory drugs (NSAIDs) has been implicated in Preoperative platelet-to-lymphocyte ratios and postop- the modulation of cancer-related inflammation; more- erative platelet-to-lymphocyte ratios on days 1, 2, and over, use of NSAIDs could reduce proliferation, inva- 3 were calculated from routine complete blood counts. sion, and micrometastasis of malignant cells by reduc- The primary end points of this study were RFS and ing the inflammatory state associated with the tumor OS. RFS was defined as the time in months from the microenvironment and by promoting apoptosis in can- date of surgery to the date of recurrence or the date cer cells.14-16 In the context of oncological surgery, a ret- of death, whichever occurred first.5 Patients were cen- rospective study demonstrated an association between sored at the last recurrence-free date if neither recur- intraoperative use of NSAIDs and improved oncologi- rence nor death occurred. OS was defined as the time cal outcomes in breast cancer; this finding may be the in months from surgery to death from any cause.5 result of a reduction in the inflammatory response as- Descriptive statistics for demographics and base- sociated with surgery.5,17 line patient characteristics, including mean, standard The predictive value of the platelet-to-lymphocyte deviation, median, and interquartile range, were calcu- ratio has not been determined in patients with NSCLC. lated for continuous variables. Frequency counts and Therefore, the aim of this retrospective study is to test percentages were calculated for categorical variables. the hypothesis that a low perioperative platelet-to-lym- A Fisher exact test or chi-square test was used to evalu- phocyte ratio is associated with improved recurrence- ate the association between 2 categorical variables. A free survival (RFS) and OS rates in patients with NSCLC Kruskal–Wallis test was used to evaluate the difference undergoing surgery. We also investigated the effects of in a continuous variable between patient groups. Two NSAID use on survival rates in patients with high and different cutoff values were used to discriminate pa- low platelet-to-lymphocyte ratios. tients with high inflammatory status from those with low status. First, the median platelet-to-lymphocyte ra- Methods tio at each time point (preoperative and postoperative Patient Selection days 1, 2, and 3) was used to divide patients with high This study was approved by the Institutional Review inflammatory status from those with low status. We Board of the University of Texas MD Anderson Cancer also used a cutoff value of 180 to discriminate between Center (Houston, TX). Demographical, perioperative, patients with high preoperative and postoperative and survival data were retrieved from the MD Anderson platelet-to-lymphocyte ratios (≥ 180) from those with Research Electronic Data Capture database for patients low ratios (< 180).18 The Kaplan–Meier method was with NSCLC. We obtained information for 1,637 patients used for time-to-event analysis, including RFS and OS with stage I, II, or III NSCLC who underwent surgery analyses. The median time to event in months and sur- between January 1, 2004, and July 31, 2011. Patients vival rates at 3 and 5 years with 95% confidence inter- 18 years of age or older who underwent thoracic sur- vals (CIs) were calculated. The log-rank test was used gery for NSCLC with intention to cure were included in to evaluate the differences in time-to-event end points the analysis. We excluded those who had undergone between patient groups. Univariate Cox proportional palliative surgery or had secondary malignancies. hazard ratio (HR) models were fitted to evaluate the ef- All patients had surgery under balanced, gen- fects of continuous variables on RFS and OS. Multivari- eral anesthesia. Postoperative pain management able Cox proportional HR models were used for multi- consisted of patient-controlled epidural analgesia or variate analyses to include covariates associated with patient-controlled intravenous opioid analgesia with prognosis in previous studies or had a P value of less or without the addition of NSAIDs that started within than .2 in univariate analyses.5,19 A P value less than .05 72 hours after surgery. Those patients treated with was considered statistically significant. SAS 9.1.3 (SAS, NSAIDs received ketorolac (30–60 mg/day), ibuprofen Cary, NC) and S-Plus 8.0 (TIBCO Software, Palo Alto, (200–800 mg/day), rofecoxib (50 mg/day), or celecoxib CA) were used for the analyses. (200–300 mg/day) alone or in combination. Results Statistical Analysis Patient Characteristics The following variables were included in our analy- A total of 1,637 patients with NSCLC were included in sis: age, sex, body mass index, American Society of the analysis. Approximately one-half (51.9%) had stage Anesthesiologists (ASA) physical status, tumor histol- I cancer. In general, patients with stage I or II disease ogy, stage of disease, type of surgery, types and doses were older than those with stage III disease (Table 1). of NSAIDs, and neoadjuvant chemotherapy with or More women had stage I tumors than did women with without adjuvant chemotherapy with or without ra- stage II/III tumors (see Table 1). Overall, more patients diotherapy. had adenocarcinoma (903; 55.2%) than other histolo- The platelet-to-lymphocyte ratio was calculated gies of NSCLC. When comparing the percentage of

July 2016, Vol. 23, No. 3 Cancer Control 285 Table 1. — Characteristics of Patients (N = 1,637) Covariate Stage P Covariate Stage P Value Value I II III I II III Age, median 66.62 66.08 64.67 .0005 Postoperative .0237 (IQR; y) (19.29– (18.14– (19.02– day 1 platelet-to- 89.96) 88.96) 88.61) lymphocyte ratio BMI, median (IQR; 26.6 26.37 26.21 .5031 Median (IQR) 213.33 219 230.23 .2478 2 kg/m ) (16.10– (16.10– (17.13– (2.96– (22.09– (19.38– 54.91) 54.91) 57.13) 3,022.22) 1071.74) 1,476.92) Sex, n (%) .0001 High: ≥ median, 702 (82.7) 320 (79.6) 324 (83.9) Female 454 (53.5) 166 (41.3) 172 (44.6) n (%) Male 395 (46.5) 236 (58.7) 214 (55.4) Low: < median, 147 (17.3) 82 (20.4) 62 (16.1) n (%) ASA physical .4436 status, n (%) Postoperative .0049 1–2 91 (10.7) 44 (10.9) 33 (8.5) day 2 platelet-to- lymphocyte ratio 3–4 758 (89.3) 358 (89.1) 353 (91.5) Median (IQR) 150.76 170.18 171.43 .0821 Adenocarcinoma, .0004 (32.31– (39.77– (24.16– n (%) 1,900.00) 821.65) 1,269.23) Yes 497 (58.5) 188 (46.8) 218 (56.5) High: ≥ median, 187 (61.3) 102 (66.2) 133 (71.1) No 352 (41.5) 214 (53.2) 168 (43.5) n (%) Type of surgery, < .0001 Low: < median, 118 (38.7) 52 (33.8) 54 (28.9) n (%) n (%) Thoracotomy 545 (64.2) 323 (80.3) 328 (85) Postoperative .088 day 3 platelet-to- Thoracoscopy 304 (35.8) 79 (19.7) 58 (15) lymphocyte ratio Neoadjuvant < .0001 chemotherapy, Median (IQR) 163.41 185.63 180.92 .5415 n (%) (16.28– (48.32– (27.31– 890.00) 792.00) 1,250.00) Yes 64 (7.5) 77 (19.2) 164 (42.5) High: ≥ median, 184 (70.5) 110 (74.8) 107 (69.5) No 785 (92.5) 325 (80.8) 222 (57.5) n (%) Neoadjuvant ra- .0056 Low: < median, 77 (29.5) 37 (25.2) 47 (30.5) diotherapy, n (%) n (%) Yes 4 (0.5) 5 (1.2) 10 (2.6) Preoperative .0001 No 845 (99.5) 397 (98.8) 76 (97.4) platelet-to- lymphocyte Adjuvant chemo- < .0001 ratio, n (%) therapy, n (%) High: ≥ 180 169 (19.9) 122 (30.3) 107 (27.7) Yes 91 (10.7) 148 (36.8) 150 (38.9) Low: < 180 680 (80.1) 280 (69.7) 279 (72.3) No 758 (89.3) 254 (63.2) 236 (61.1) Postoperative .3128 Adjuvant radio- < .0001 therapy, n (%) day 1 platelet-to- lymphocyte ratio, Yes 24 (2.8) 43 (10.7) 154 (39.9) n (%) No 825 (97.2) 59 (89.3) 232 (60.1) High: ≥ 180 532 (62.7) 260 (64.7) 259 (67.1) Use of ketorolac, .0065 Low: < 180 317 (37.3) 142 (35.3) 127 (32.9) n (%) Postoperative .0099 Yes 375 (44.2) 188 (46.8) 208 (53.9) day 2 platelet-to- No 474 (55.8) 214 (53.2) 178 (46.1) lymphocyte ratio, n (%) Preoperative < .0001 platelet-to-lym- High: ≥ 180 109 (35.7) 72 (46.8) 90 (48.1) phocyte ratio Low: < 180 196 (64.3) 82 (53.2) 97 (51.9) Median (IQR) 128.31 146.93 137.75 .0001 (3.05– (21.75– (13.19– Postoperative .1816 1141.3) 1,203.57) 1,028.21) day 3 platelet- to-lymphocyte High: ≥ median, 386 (45.5) 233 (58) 200 (51.8) ratio, n (%) n (%) High: ≥ 180 115 (44.1) 77 (52.4) 79 (51.3) Low: < median, 463 (54.5) 169 (42) 186 (48.2) n (%) Low: < 180 146 (55.9) 70 (47.6) 75 (48.7)

ASA = American Society of Anesthesiologists, BMI = body mass index, IQD = interquartile deviation, SD = standard deviation.

286 Cancer Control July 2016, Vol. 23, No. 3 patients who had adenocarcinoma with the percent- P = .0148). This difference was also statistically signifi- age of those with other forms of NSCLC within each cant in the platelet-to-lymphocyte ratio values of post- disease stage, the percentages were higher among operative day 1 (ketorolac: 212.64 [range, 2.96–1,611.76] those with stage I disease (58.5%) and among those vs no ketorolac: 225.67 [range, 22.09–3,022.22]; P = .0292), with stage III disease (56.5%) than among those with day 2 (ketorolac: 150.22 [range, 24.16–840.91] vs no stage II disease (46.8%). Thoracotomy was the most com- ketorolac: 171.12 [range, 40.69–1,900.00]; P = .0080), mon surgical procedure across the 3 groups of patients, and day 3 (ketorolac: 167.44 [range, 16.28–733.33] vs although more than one-third (35.8%) of patients no ketorolac: 181.98 [range, 31.07–1,250.00]; P = .0429). with stage I disease underwent video-assisted thoracoscopic surgery (see Table 1). Most patients with Recurrence-Free Survival stage I or II tumors did not receive neoadjuvant or adju- The median RFS rate of the overall population was vant chemotherapy or radiotherapy; by contrast, patients 76 months (95% confidence interval [CI], 68.33–88.53), with stage III disease were more likely to receive adju- and the RFS rates at 3 and 5 years were 64% (95% CI, vant chemotherapy (38.9%) or radiotherapy (39.9%; see 62–66) and 55% (95% CI, 53–58), respectively. The uni- Table 1). variate analysis showed that age, male sex, ASA physical A statistically significant difference was found in status 3/4, neoadjuvant chemotherapy, adjuvant radio- the median preoperative platelet-to-lymphocyte ra- therapy, and open thoracotomy were all associated with tio values among patients with various tumor stages. decreased rates of RFS at 3 and 5 years (Table 2). The Patients with stage I disease had the lowest preopera- analysis also demonstrated that patients with stage I dis- tive platelet-to-lymphocyte ratios compared with those ease had the highest RFS rates at 3 and 5 years (76% and with stage II/III disease (P < .0001; see Table 1). Simi- 68%, respectively) compared with patients with stage II larly, when the analysis was conducted using the me- (58% and 46%, respectively) or III disease (44% and 38%, dian values of the entire population or a cutoff plate- respectively; P < .0001). The univariate analysis also let-to-lymphocyte ratio of 180, a larger percentage of demonstrated that a preoperative platelet-to-lympho- patients in stage I had lower platelet-to-lymphocyte ra- cyte ratio greater than the median (133) or higher than tios than these cutoff values (P = .0001 for both cutoff 180 was a predictor of poor RFS rate (P = .0002; see levels). Despite an increase in postoperative platelet-to- Table 2). Similarly, a platelet-to-lymphocyte ratio above lymphocyte ratios across all groups, the statistical anal- 180 but not greater than the median value on postop- ysis demonstrated that patients’ median values on days erative day 3 was associated with significant changes in 1 and 2 significantly differed among the various tumor RFS (P = .0414), but not postoperative day 1 (P = .3937) stages. As shown in Table 1, patients with stage III tu- or 2 (P = .2206; Fig 1). Use of NSAIDs perioperatively mors had the highest postoperative median platelet-to- was not associated with longer rates of RFS (P = .719). lymphocyte ratios. When we analyzed the proportion In a multivariate model analysis, a high preop- of patients with platelet-to-lymphocyte ratios greater erative platelet-to-lymphocyte ratio (≥ 180), older age, than or less than the median or those that were high- and male sex were associated with 22% (P = .019), 2% er or lower than 180, no statistically significant differ- (P < .0001), and 19% (P = .021) increases the risk of ences were observed, except on postoperative day 2. recurrence, respectively (Table 3). Tumor staging was Patients with a platelet-to-lymphocyte ratio of at least also an independent prognostic factor of poor RFS. 180 on postoperative day 2 had a higher tumor stage Thus, patients with stage II/III disease had an in- (stage I = 35.7% vs stage II = 46.8% vs stage III = 48.1%; creased recurrence risk of 77% (P < .0001) and 116% P = .009). A similar trend was found when the median (P < .0001), respectively, compared with the risk for was considered as the cutoff value (P = .082). patients with stage I disease. Other variables associ- In regard to the perioperative use of NSAIDs, ated with shorter rates of RFS were administration 931 patients (56.87%) received this type of analgesic; of neoadjuvant chemotherapy (HR = 1.339; 95% CI, 771 (47.1%) were treated with ketorolac alone or in 1.107–1.621; P = .0026) and adjuvant radiotherapy combination with ibuprofen, celecoxib, or both. A (HR = 1.340; 95% CI = 1.083–1.657; P = .0070). The sin- statistically significant difference was found between gle factor that was significantly associated with longer use and nonuse of ketorolac among the various tumor rates of survival was minimally invasive surgery; pa- stages. Ketorolac was more commonly administered to tients who underwent this surgery had a 19% reduced patients with higher tumor stages than to those with risk of recurrence (P = .049; see Table 3). stage I disease (P = .006; see Table 1). The statistical Because the preoperative and postoperative analysis also demonstrated a significant difference be- platelet-to-lymphocyte ratios of patients treated with tween the median (interquartile range) preoperative ketorolac were statistically different, we investigated platelet-to-lymphocyte ratio of patients treated with ke- the association between platelet-to-lymphocyte ra- torolac (129.91 [range, 3.05–1203.70]) and patients not tios and RFS rates in these 2 populations of patients. treated with ketorolac (138.88 [range, 21.58–1028.21]; The log-rank analysis demonstrated that a preopera-

July 2016, Vol. 23, No. 3 Cancer Control 287 tive (P = .104) or postoperative platelet-to-lymphocyte 108 months. OS rates at 3 and 5 years were 76% (95% ratio higher than 180 or less than 180 was not as- CI, 74–79) and 67% (95% CI, 64–69), respectively. The sociated with a change in survival (postoperative univariate analysis demonstrated that a high preop- day 1: P = .104; postoperative day 2: P = .952; and post- erative platelet-to-lymphocyte ratio (≥ 180), older age, operative day 3: P = .307) in patients taking ketorolac. male, ASA physical status 3/4, ASA physical status 2/3, By contrast, a platelet-to-lymphocyte ratio higher than use of neoadjuvant chemotherapy, use of adjuvant ra- 180 was associated with poor survival in patients not diotherapy, and thoracotomy were associated with treated with ketorolac (preoperative: P = .0006; post- a decreased OS rate (Table 4; Fig 2). The OS rates at operative day 1: P = .272; postoperative day 2: P = .008; 3 and 5 years were not statistically significant different and postoperative day 3: P = .014). between patients who did or did not receive perioperative NSAIDs (P = .443), or between those with a ratio Overall Survival higher than or less than 180, and a median higher or The median OS rate for the entire population was lower platelet-to-lymphocyte ratio on postoperative

Table 2. — Univariate Analysis for RFS Variable No. of Study No. of RFS HR Median RFS, mo 3-y RFS Rate 5-y RFS Rate P Patients Events (95% CI) (95% CI) (95%CI) (95% CI) Value Age (continuous) 1,637 683 1.02 < .0001 (1.02–1.03) BMI (continuous) 1,637 683 0.99 .2749 (0.98–1.01) Sex Female 792 296 87.58 (75.33–108.38) 0.68 (0.65–0.72) 0.61 (0.57–0.65) .0002 Male 845 387 63.21 (50.56–77.63) 0.6 (0.57–0.64) 0.5 (0.47–0.54) ASA physical status 1–2 168 57 98.52 (76.02–NA) 0.76 (0.69–0.83) 0.65 (0.57–0.74) .0147 3–4 1,469 626 71.75 (65.44–83.51) 0.63 (0.60–0.65) 0.54 (0.51–0.57) Adenocarcinoma No 734 301 81.80 (65.90–108.38) 0.63 (0.60–0.67) 0.55 (0.51–0.60) .8137 Yes 903 382 71.35 (65.44–88.53) 0.65 (0.61–0.68) 0.55 (0.52–0.59) Stage I 849 258 NA (99.77–NA) 0.76 (0.73–0.79) 0.68 (0.64–0.71) < .0001 II 402 193 51.08 (41.92–71.09) 0.58 (0.53–0.63) 0.46 (0.41–0.53) III 386 232 24.11 (20.07–34.89) 0.44 (0.39–0.50) 0.38 (0.33–0.44) Type of surgery Thoracotomy 1,196 553 67.54 (57.62–77.63) 0.61 (0.59–0.64) 0.52 (0.49–0.56) < .0001 Thoracoscopy 441 130 NA (95.24–NA) 0.72 (0.67–0.76) 0.65 (0.60–0.71) Neoadjuvant chemotherapy No 1,332 506 87.58 (76.84–99.77) 0.68 (0.66–0.71) 0.59 (0.56–0.62) < .0001 Yes 305 177 28.48 (19.05–44.42) 0.46 (0.41–0.52) 0.39 (0.34–0.46) Neoadjuvant radiotherapy No 1,618 673 76.38 (68.33–88.53) 0.64 (0.62–0.67) 0.55 (0.53–0.58) .3285 Yes 19 10 18.73 (14.06–NA) 0.47 (0.29–0.76) 0.47 (0.29–0.76) Adjuvant chemotherapy No 1,248 499 82.19 (68.66–96.81) 0.65 (0.62–0.68) 0.56 (0.53–0.60) .2608 Yes 389 184 68.33 (52.4–87.58) 0.62 (0.57–0.67) 0.53 (0.48–0.58) Adjuvant radiotherapy No 1,416 539 89.55 0.68 (0.65–0.70) 0.59 (0.56–0.62) < .0001 (78.52–108.28) Yes 221 144 24.84 (18.43–34.92) 0.42 (0.36–0.5) 0.34 (0.27–0.41)

continued on next page

288 Cancer Control July 2016, Vol. 23, No. 3 Table 2. — Univariate Analysis for RFS, continued Variable No. of Study No. of RFS HR Median RFS, mo 3-y RFS Rate 5-y RFS Rate P Patients Events (95% CI) (95% CI) (95%CI) (95% CI) Value Preoperative platelet-to- lymphocyte ratio High: ≥ median 819 370 71.16 (57.39 – 87.58) 0.61 (0.58–0.65) 0.53 (0.49–0.57) .0342 Low: < median 818 313 81.80 (68.66–95.24) 0.67 (0.64–0.71) 0.58 (0.54–0.62) Postoperative platelet-to- lymphocyte ratio day 1 High: ≥ median 1,346 570 75.33 (67.05–88.53) 0.63 (0.61–0.66) 0.55 (0.52–0.58) .3789 Low: < median 291 113 83.51 (63.24–NA) 0.66 (0.61–0.72) 0.57 (0.51–0.64) Postoperative platelet-to- lymphocyte ratio day 2 High: ≥ median 422 218 51.08 (37.52–72.34) 0.56 (0.51–0.61) 0.48 (0.43–0.54) .0598 Low: < median 224 96 68.66 (48.65–NA) 0.66 (0.59–0.73) 0.54 (0.47–0.62) Postoperative platelet-to- lymphocyte ratio day 3 High: ≥ median 401 202 56.67 (40.28–82.92) 0.58 (0.53–0.63) 0.49 (0.44–0.54) .8775 Low: < median 161 74 51.25 (36.79–NA) 0.58 (0.50–0.67) 0.49 (0.41–0.59) Preoperative platelet-to- lymphocyte ratio High: ≥ 180 398 202 55.22 (38.34–80.32) 0.56 (0.51–0.62) 0.48 (0.43–0.54) .0002 Low: < 180 1,239 481 81.80 (71.16–96.81) 0.67 (0.64–0.69) 0.58 (0.55–0.61) Postoperative platelet-to- lymphocyte ratio day 1 High: ≥ 180 1,051 450 76.02 (65.54–88.53) 0.63 (0.60–0.66) 0.55 (0.52–0.58) .3937 Low: < 180 586 233 78.52 (64.85–99.77) 0.66 (0.62–0.71) 0.56 (0.52–0.61) Postoperative platelet-to- lymphocyte ratio day 2 High: ≥ 180 271 143 49.24 (33.57–82.92) 0.55 (0.49–0.61) 0.49 (0.43–0.56) .2206 Low: < 180 375 171 66.52 (44.51–83.51) 0.62 (0.57–0.68) 0.51 (0.45–0.57) Postoperative platelet-to- lymphocyte ratio day 3 High: ≥ 180 271 149 40.11 (25.26–71.09) 0.53 (0.47–0.60) 0.44 (0.38–0.51) .0414 Low: < 180 291 127 67.54 (49.24–108.28) 0.62 (0.57–0.68) 0.53 (0.47–0.60)

ASA = American Society of Anesthesiologists, BMI = body mass index, CI = confidence interval, HR = hazard ratio, NA = not applicable, RFS = recurrence-free survival. days 1 (P = .907), 2 (P = .103), and 3 (P = .066). RFS, preoperative (P = .197) and postoperative (day 1: The multivariate analysis showed that age P = .302; day 2: P = .678; and day 3: P = .653) plate- (HR = 1.034; 95% CI, 1.024–1.044; P < .0001), male sex let-to-lymphocyte ratios higher than or less than 180 (HR = 1.328; 95% CI, 1.109–1.590; P = .002), postop- were not associated with a change in survival among erative radiotherapy (HR = 1.337; 95% CI, 1.052–1.698; patients treated with ketorolac. However, a platelet-to- P = .0176), and higher tumor stage (II vs I: HR = 1.864, lymphocyte ratio higher than 180 was associated with 95% CI, 1.488–2.335, P < .0001; III vs I: HR = 2.451, 95% poor survival in patients not treated with ketorolac CI, 1.922–3.125, P < .0001) were independent predictor (preoperative: P = .0002; postoperative day 1: P = .278; factors of poor OS (Table 5). The analysis also demon- postoperative day 2: P = .013; and postoperative day 3: strated that a preoperative platelet-to-lymphocyte ratio P = .039). of at least 180 was associated with a 33% increase in mortality after surgery (P = .0036). Discussion We also investigated the association between The platelet-to-lymphocyte ratio has emerged as a platelet-to-lymphocyte ratios and OS in patients treat- potential inflammatory biomarker and a predictor of ed with or without ketorolac. Similar to the findings in toxicity- and cancer-related survival and OS rates in pa-

July 2016, Vol. 23, No. 3 Cancer Control 289 tients with NSCLC.20-22 Similar to the results of previ- Table 3. — Multivariate Cox Proportional Hazard Model for ously published studies, our work shows that the pre- Recurrence-Free Survival Rates operative platelet-to-lymphocyte ratio — but not the Covariate HR (95% CI) P Value early postoperative platelet-to-lymphocyte ratio — is a Age 1.023 (1.015–1.031) < .0001 predictor of recurrence and mortality in patients with BMI 0.991 (0.976–1.007) .2749 NSCLC undergoing surgery. Sex (male vs female) 1.198 (1.027–1.396) .0212 Three other studies failed to demonstrate the ASA physical status 3–4 vs 1–2 1.123 (0.851–1.482) .4129 prognostic value of the platelet-to-lymphocyte ratio; Stage however, several differences can be seen between those studies and the present study.20,23,24 The sample II vs I 1.775 (1.466–2.150) < .0001 size in the other studies was significantly smaller than III vs I 2.165 (1.753–2.673) < .0001 that in the present work; therefore, these previous re- Neoadjuvant chemotherapy 1.339 (1.107–1.621) .0026 (yes vs no) ports might have been statistically underpowered to detect a difference in survival between patients with Adjuvant radiotherapy 1.340 (1.083–1.657) .0070 20,23,24 (yes vs no) a high or low inflammatory status. By contrast, Type of surgery 0.818 (0.670–0.999) .0491 to the best of our knowledge, the sample size of the (thoracoscopy vs thoracotomy) present study is the largest of all studies that have ad- Preoperative platelet-to- 1.224 (1.033–1.450) 0.0193 dressed the prognostic value of platelet-to-lymphocyte lymphocyte ratio (≥ 180 vs < 180) ratios in patients with NSCLC. The studied population ASA = American Society of Anesthesiologists, BMI = body mass index, of patients in the previously published 3 reports was CI = confidence interval, HR = hazard ratio. different from ours.20,23,24 For example, Miyazaki et al23

A C

100 High (E/N = 202/398) 100 High (E/N = 143/271) Low (E/N = 481/1,239) Low (E/N = 171/375) 80 80 P = .0002 P = .2206

60 60

40 40 Probability, % Probability, %

20 20

0 0 0 12 24 36 48 60 72 84 96 108 120 132 0 12 24 36 48 60 72 84 96 108 120 132 Months Months B D

100 High (E/N = 450/1051) 100 High (E/N = 149/271) Low (E/N = 233/586) Low (E/N = 127/291) 80 80 P =.3937 P = .0414

60 60

40 40 Probability, % Probability, %

20 20

0 0 0 12 24 36 48 60 72 84 96 108 120 132 0 12 24 36 48 60 72 84 96 108 120 132 Months Months

Fig 1A–D. Kaplan-Meier curves show recurrence-free survival in patients with platelet-to-lymphocyte ratios greater than (high) or less than (low) 180. Each panel represents platelet-to-lymphocyte ratios at various time points: (A) preoperative, (B) postoperative day 1, (C) postoperative day 2, and (D) postoperative day 3. E = event.

290 Cancer Control July 2016, Vol. 23, No. 3 included elderly patients with stage I disease, and of malignancy, including tumor escape, cell survival Unal et al20 had a larger proportion of patients with and proliferation, epithelial–mesenchymal transfor- squamous cell histology. Our study included patients mation, and metastasis.25,26 Activation of platelets in who were younger, and adenocarcinoma was the most the context of inflammation can favor tumor progres- common histology. Although the cutoff values (deter- sion by several mechanisms, including immunosup- mined as the median platelet-to-lymphocyte or plate- pression and angiogenesis. It has even been suggested let-to-lymphocyte ratio = 180) used in our study were that platelets can “cover” cancer cells, resulting in the arbitrary, they were different from those used in the escape of cancer cells from immune cytotoxic cells other prior reports.20,23,24 such as natural killer cells.27 Platelets can also release The findings of this study highlight the impor- growth factors, such as prostaglandin E, vascular en- tance of inflammatory status on the survival of patients dothelial growth factor, and transforming growth fac- with NSCLC. Inflammation is involved in many aspects tor that disrupt the function of natural killer cells and

Table 4. — Univariate Analysis for Overall Survival Rates Variable No. of Study No. of RFS HR Median RFS, mo 3-y RFS Rate 5-y RFS Rate P Patients Events (95% CI) (95% CI) (95%CI) (95% CI) Value Age (continuous) 1,637 505 1.03 (1.02–1.03) < .0001 BMI (continuous) 1,637 505 0.99 (0.97–1.01) .2490 Sex Female 792 206 NA (102.14–NA) 0.82 (0.79–0.85) 0.72 (0.68–0.76) < .0001 Male 845 299 97.86 (77.63– NA) 0.71 (0.68–0.74) 0.61 (0.58–0.65) ASA physical status 1–2 168 41 NA (98.52–NA) 0.87 (0.81–0.92) 0.77 (0.69–0.85) .0211 3–4 1,469 464 108.38 (95.2–NA) 0.75 (0.73–0.77) 0.65 (0.63–0.68) Adenocarcinoma No 734 230 NA (94.74–NA) 0.73 (0.70–0.77) 0.65 (0.61–0.69) .3212 Yes 903 275 99.67 (89.55–NA) 0.79 (0.76–0.82) 0.68 (0.64–0.72) Stage I 849 180 NA 0.86 (0.84–0.89) 0.78 (0.74–0.81) < .0001 II 402 144 77.63 0.70 (0.66–0.76) 0.59 (0.54–0.66) (66.52–102.14) III 386 181 58.44 0.61 (0.56–0.66) 0.49 (0.44–0.55) (46.91–80.32) Type of surgery Thoracotomy 1,196 418 99.67 (89.39–NA) 0.74 (0.71–0.77) 0.64 (0.61–0.67) .0002 Thoracoscopy 441 87 NA 0.83 (0.79–0.87) 0.75 (0.70–0.81) Neoadjuvant chemotherapy No 1,332 376 109.66 (99.67–NA) 0.78 (0.76–0.81) 0.70 (0.67–0.73) < .0001 Yes 305 129 70.83 (52.83–NA) 0.67 (0.62–0.73) 0.53 (0.46–0.6) Neoadjuvant radiotherapy No 1,618 497 108.28 (95.93–NA) 0.76 (0.74–0.79) 0.67 (0.64–0.69) .2800 Yes 19 8 NA (16.59–NA) 0.63 (0.45–0.89) 0.54 (0.34–0.86) Adjuvant chemotherapy No 1,248 371 NA (99.77–NA) 0.77 (0.74–0.79) 0.67 (0.64–0.70) .8335 Yes 389 134 88.53 (81.80–NA) 0.75 (0.71–0.80) 0.67 (0.62–0.72) Adjuvant radiotherapy No 1,416 393 NA (102.14–NA) 0.79 (0.77–0.81) 0.71 (0.68–0.73) < .0001 52.04 0.61 (0.55–0.68) 0.44 (0.37–0.52) Yes 221 112 (45.86–77.63) continued on next page

July 2016, Vol. 23, No. 3 Cancer Control 291 Table 4. — Univariate Analysis for Overall Survival Rates, continued Variable No. of Study No. of RFS HR Median RFS, mo 3-y RFS Rate 5-y RFS Rate P Patients Events (95% CI) (95% CI) (95%CI) (95% CI) Value Preoperative platelet-to- lymphocyte ratio High: ≥ median 819 279 99.67 (87.58 – NA) 0.74 (0.71–0.77) 0.65 (0.61–0.69) .0739 Low: < median 818 226 NA (95.24–NA) 0.78 (0.75–0.82) 0.68 (0.65–0.72) Postoperative platelet-to- lymphocyte ratio day 1 High: ≥ median 1,346 420 108.28 (91.49–NA) 0.76 (0.74–0.79) 0.66 (0.63–0.69) .7146 Low: < median 291 85 108.38 (95.93–NA) 0.75 (0.70–0.81) 0.67 (0.61–0.74) Postoperative platelet-to- lymphocyte ratio day 2 High: ≥ median 422 167 94.74 (76.51–NA) 0.69 (0.65–0.74) 0.59 (0.54–0.65) .1457 Low: < median 224 72 97.86 (75.33–NA) 0.78 (0.72–0.84) 0.65 (0.58–0.73) Postoperative platelet-to- lymphocyte ratio day 3 High: ≥ median 401 160 88.53 (78.52–NA) 0.70 (0.65–0.74) 0.59 (0.53–0.64) .6054 Low: < median 161 60 73.03 (52.53–NA) 0.66 (0.58–0.74) 0.58 (0.49–0.67) Preoperative platelet-to- lymphocyte ratio High: ≥ 180 398 160 87.58 .0002 0.70 (0.65–0.75) 0.60 (0.55–0.66) (77.63–108.38) Low: < 180 1,239 345 NA (102.14–NA) 0.78 (0.76–0.81) 0.69 (0.66–0.72) Postoperative platelet-to- lymphocyte ratio day 1 High: ≥ 180 1,051 329 109.66 (91.49–NA) 0.76 (0.73–0.79) 0.66 (0.63–0.70) .9073 Low: < 180 586 176 102.89 (95.24–NA) 0.77 (0.73–0.8) 0.67 (0.62–0.71) Postoperative platelet-to- lymphocyte ratio day 2 High: ≥ 180 271 114 90.54 (67.05 – NA) 0.69 (0.63–0.75) 0.59 (0.53–0.66) .1060 Low: < 180 375 125 99.77 (75.49–NA) 0.75 (0.7–0.79) 0.63 (0.57–0.69) Postoperative platelet-to- lymphocyte ratio day 3 78.52 High: ≥ 180 271 122 (58.31–108.38) 0.65 (0.59–0.71) 0.54 (0.48–0.61) .0668 Low: < 180 291 98 99.77 (75.49–NA) 0.72 (0.67–0.78) 0.62 (0.56–0.69) ASA = American Society of Anesthesiologists, BMI = body mass index, CI = confidence interval, HR = hazard ratio, RFS = recurrence-free survival. stimulate the growth of cancer cells.27,28 Our study also showed that patients with more Absolute or relative lymphopenia is another im- advanced disease had a higher platelet-to-lymphocyte portant component of a high platelet-to-lymphocyte ra- ratio, suggesting a link between the amount of tumor tio in patients with NSCLC. The results from a previous burden and the magnitude of systemic inflammatory study indicated that patients with low lymphocyte na- response. This finding is in agreement with previ- dirs undergoing radiotherapy have lower rates of sur- ous studies, including a smaller cohort of patients, in- vival than patients whose lymphocyte levels are high- dicating an association between stage of disease and er.29 Likewise, Campian et al30 observed an association systemic inflammation.5,32 For example, patients with between severe lymphopenia and reduced survival in gynecological, breast, or colorectal cancer who have el- patients with NSCLC. The results of these publications evated platelet-to-lymphocyte ratios have a higher risk are in agreement with a recent study demonstrating of positive lymph nodes and larger tumors than those that patients with NSCLC who have a high inflamma- whose ratios are lower.32-35 tory status and a number of low CD8-positive T cells at In terms of the postoperative inflammatory re- the tumor level have a higher risk of mortality.31 sponse, our analysis demonstrated that patients with

292 Cancer Control July 2016, Vol. 23, No. 3 A C

100 High (E/N = 160/398) 100 High (E/N = 114/271) Low (E/N = 345/1,239) Low (E/N = 125/375) 80 80 P = .0002 P = .1060

60 60

40 40 Probability, % Probability, %

20 20

0 0 0 12 24 36 48 60 72 84 96 108 120 132 0 12 24 36 48 60 72 84 96 108 120 132 Months Months B D

100 High (E/N = 329/1,051) 100 High (E/N = 122/271) Low (E/N = 176/586) Low (E/N = 98/291) 80 80 P =.9073 P = .0688

60 60

40 40 Probability, % Probability, %

20 20

0 0 0 12 24 36 48 60 72 84 96 108 120 132 0 12 24 36 48 60 72 84 96 108 120 132 Months Months

Fig 2A–D. Kaplan-Meier curves show overall survival in patients with platelet-to-lymphocyte ratio greater than (high) or less than (low) 180. Each figure panel represents platelet-to-lymphocyte ratios at various time points: (A) preoperative, (B) postoperative day 1, (C) postoperative day 2, and (D) postoperative day 3. E = event.

stage II/III disease had significantly higher platelet-to- Table 5. — Multivariable Cox Proportional Hazard Model for Overall Survival lymphocyte ratios than did patients with stage I disease. This phenomenon might be related to the magni- Covariate HR (95% CI) P Value tude of the surgical insult because a larger proportion Age 1.034 (1.024–1.044) < .0001 of patients with stage II/III disease underwent thora- BMI 0.996 (0.977–1.014) .6371 cotomies than thoracoscopies. These findings are sup- 36 Sex (male vs female) 1.328 (1.109–1.590) .0020 ported by Whitson et al, who showed that markers of inflammatory response such as interleukin 6 corre- ASA physical status 3–4 vs 1–2 1.083 (0.781–1.502) .6328 late with the magnitude of surgical trauma. Although Stage we failed to demonstrate an association between high II vs I 1.864 (1.488–2.335) < .0001 postoperative platelet-to-lymphocyte ratios and poor III vs I 2.451 (1.922–3.125) < .0001 rates of survival, it is possible that using higher cutoff Neoadjuvant chemotherapy (yes 1.152 (0.922–1.438) .2138 values than those chosen for this study may have re- vs no) sulted in better predictive ability. Adjuvant radiotherapy (yes vs no) 1.337 (1.052–1.698) .0176 Type of surgery (thoracoscopy vs 0.790 (0.620–1.006) .0563 Limitations thoracotomy) Our study has the inherent limitations of any retro- Preoperative platelet-to-lympho- 1.334 (1.099–1.620) .0036 spective study; therefore, the results must be interpret- cyte ratio (≥ 180 vs < 180) ed with caution. Missing and/or confounding variables ASA = American Society of Anesthesiologists, BMI = body mass index, not included in the statistical analysis could have in- CI = confidence interval, HR = hazard ratio. fluenced the survival findings of the studied popula-

July 2016, Vol. 23, No. 3 Cancer Control 293 tion of patients. For example, we did not account for cancer patients. Br J Cancer. 2014;110(10):2524-2530. 10. Nurden AT. Platelets, inflammation and tissue regeneration. the presence of mutations, for the type of biologically Thromb Haemost. 2011;105(suppl 1):S13-S33. targeted therapeutic agents, or for conventional che- 11. Goubran HA, Burnouf T, Radosevic M, et al. The platelet-cancer loop. Eur J Intern Med. 2013;24(5):393-400. motherapy given postoperatively. We used 2 different 12. Coussens LM, Werb Z. Inflammation and cancer. Nature. cutoff values to assess the impact of the platelet-to-lym- 2002;420(6917):860-867. 13. Balkwill F, Mantovani A. Inflammation and cancer: back to Virchow? phocyte ratio on rates of RFS and OS. The chosen val- Lancet. 2001;357(9255):539-545. ues were arbitrarily determined, so they are not based 14. Benish M, Bartal I, Goldfarb Y, et al. Perioperative use of beta- blockers and COX-2 inhibitors may improve immune competence and re- on a receiving operative-curve analysis. The platelet- duce the risk of tumor metastasis. Ann Surg Oncol. 2008;15(7):2042-2052. to-lymphocyte ratio calculation was based on the ab- 15. Zhang S, Da L, Yang X, et al. Celecoxib potentially inhibits metastasis of lung cancer promoted by surgery in mice, via suppression of the solute lymphocyte count obtained from the complete PGE2-modulated β-catenin pathway. Toxicol Lett. 2014;225(2):201-207. blood cell count; hence, the statistical analysis did not 16. Singh B, Cook KR, Vincent L, et al. Role of COX-2 in tumorospheres derived from a breast cancer cell line. J Surg Res. 2011;168(1):e39-49. account for subpopulations of lymphocytic cells. Use 17. Forget P, Bentin C, Machiels JP, et al. Intraoperative use of ke- of NSAIDs after surgery might have been influenced torolac or diclofenac is associated with improved disease-free survival and overall survival in conservative breast cancer surgery. Br J Anaesth. by factors such as intraoperative bleeding and comor- 2014;113(suppl 1):i82-i87. bidities that might be considered sources of bias. Our 18. Zhou X, Du Y, Huang Z, et al. Prognostic value of PLR in various cancers: a meta-analysis. PLoS One. 2014;9(6):e101119. analysis also did not include variables such as intraop- 19. Cata JP, Villarreal J, Keerty D, et al. Perioperative beta-blocker use erative vs postoperative use of NSAIDs, class of NSAID, and survival in lung cancer patients. J Clin Anesth. 2014;26(2):106 -117. 20. Unal D, Eroglu C, Kurtul N, et al. ABO blood groups are not associ- or the total amount of each analgesic. Therefore, we ated with treatment response and prognosis in patients with local advanced cannot comment on whether the class of NSAID, time, non- small cell lung cancer. Asian Pac J Cancer Prev. 2013;14(6):3945-3948. 21. Cannon NA, Meyer J, Iyengar P, et al. Neutrophil-lymphocyte and and amount of administration of each NSAID might platelet-lymphocyte ratios as prognostic factors after stereotactic radia- have had an impact on the survival of our population tion therapy for early-stage non-small-cell lung cancer. J Thorac Oncol. 2015;10(2):280-285. of patients. 22. Arrieta O, Michel Ortega RM, Villanueva-Rodriguez G, et al. Asso- ciation of nutritional status and serum albumin levels with development of toxicity in patients with advanced non-small cell lung cancer treated Conclusions with paclitaxel-cisplatin chemotherapy: a prospective study. BMC Cancer. Our results show that a high preoperative inflam- 2010;10:50. 23. Miyazaki T, Yamasaki N, Tsuchiya T, et al. Inflammation-based matory status is an independent predictor of recur- scoring is a useful prognostic predictor of pulmonary resection for elderly rence-free and overall survival rates in patients with patients with clinical stage I non-small-cell lung cancer. Eur J Cardiothorac Surg. 2015;47(4):e140-e145. non–small-cell lung cancer. The preoperative platelet- 24. Pinato DJ, Shiner RJ, Seckl MJ, et al. Prognostic performance of to-lymphocyte ratio is a simple and readily available inflammation-based prognostic indices in primary operable non-small cell lung cancer. Br J Cancer. 2014;110(8):1930-1935. laboratory parameter that can be used as an indepen- 25. Lazennec G, Richmond A. Chemokines and chemokine recep- dent prognostic factor of recurrence-free survival and tors: new insights into cancer-related inflammation. Trends Mol Med. 2010;16(3):133-144. overall survival in patients with nonmetastatic, oper- 26. Fridman WH, Pages F, Sautes-Fridman C, et al. The immune con- able non–small-cell lung cancer. Further research is texture in human tumours: impact on clinical outcome. Nat Rev Cancer. 2012;12(4):298-306. needed to determine the association between a high 27. Bambace NM, Holmes CE. The platelet contribution to cancer pro- platelet-to-lymphocyte ratio and type of recurrence gression. J Thromb Haemost. 2011;9(2):237-249. 28. Mao Y, Sarhan D, Steven A, et al. Inhibition of tumor-derived pros- (local vs distant). Other independent factors of sur- taglandin-e2 blocks the induction of myeloid-derived suppressor cells and vival include age, male sex, tumor stage, and type of recovers natural killer cell activity. Clin Cancer Res. 2014;20(15):4096-4106. 29. Tang C, Liao Z, Gomez D, et al. Lymphopenia association with surgery. gross tumor volume and lung V5 and its effects on non-small cell lung cancer patient outcomes. Int J Radiat Oncol Biol Phys. 2014;89(5):1084-1091. 30. Campian JL, Ye X, Brock M, et al. Treatment-related lymphope- References nia in patients with stage III non-small-cell lung cancer. Cancer Invest. 1. American Cancer Society. Cancer Facts & Figures 2016. Atlanta, GA: 2013;31(3):183-188. American Cancer Society; 2016. http://www.cancer.org/acs/groups/content/@ 31. Alifano M, Mansuet-Lupo A, Lococo F, et al. Systemic inflammation, research/documents/document/acspc-047079.pdf. Accessed June 22, 2016. nutritional status and tumor immune microenvironment determine outcome 2. Miller KD, Siegel R, Lin CC, et al. Cancer treatment and survivor- of resected non-small cell lung cancer. PLoS One. 2014;9(9):e106914. ship statistics, 2016. CA Cancer J Clin. 2016;66(4):271-289. 32. Wang D, Wu M, Feng FZ, et al. Pretreatment neutrophil-to-lympho- 3. Green JS, Tsui BC. Impact of anesthesia for cancer surgery: Con- cyte and platelet-to-lymphocyte ratios do not predict survival in patients tinuing professional development. Can J Anaesth. 2013;60(12):1248-1269. with cervical cancer treated with neoadjuvant chemotherapy and radical 4. Snyder GL, Greenberg S. Effect of anaesthetic technique and other hysterectomy. Chin Med J (Engl). 2013;126(8):1464-1468. perioperative factors on cancer recurrence. Br J Anaesth. 2010;105(2):106-115. 33. Kwon HC, Kim SH, Oh SY, et al. Clinical significance of preopera- 5. Choi JE, Villarreal J, Lasala J, et al. Perioperative neutrophil:lymphocyte tive neutrophil-lymphocyte versus platelet-lymphocyte ratio in patients ratio and postoperative NSAID use as predictors of survival after lung cancer with operable colorectal cancer. Biomarkers. 2012;17(3):216-222. surgery: a retrospective study. Cancer Med. 2015;4(6):825-833. 34. Raungkaewmanee S, Tangjitgamol S, Manusirivithaya S, et al. 6. Asher V, Lee J, Innamaa A, et al. Preoperative platelet lymphocyte ratio Platelet to lymphocyte ratio as a prognostic factor for epithelial ovarian as an independent prognostic marker in ovarian cancer. Clin Transl Oncol. cancer. J Gynecol Oncol. 2012;23(4):265-273. 2011;13(7):499-503. 35. Azab B, Shah N, Radbel J, et al. Pretreatment neutrophil/lympho- 7. Kos M, Hocazade C, Kos FT, et al. Prognostic role of pretreatment cyte ratio is superior to platelet/lymphocyte ratio as a predictor of long- platelet/lymphocyte ratio in patients with non-small cell lung cancer. Wien term mortality in breast cancer patients. Med Oncol. 2013;30(1):432. Klin Wochenschr. 2015. Epub ahead of print. 36. Whitson BA, D’Cunha J, Andrade RS, et al. Thoracoscopic versus 8. Chen Q, Dai Z, Yin D, et al. Negative impact of preoperative platelet- thoracotomy approaches to lobectomy: differential impairment of cellular lymphocyte ratio on outcome after hepatic resection for intrahepatic cholan- immunity. Ann Thorac Surg. 2008;86(6):1735-1744. giocarcinoma. Medicine (Baltimore). 2015;94(13):e574. 9. Krenn-Pilko S, Langsenlehner U, Thurner EM, et al. The elevated preoperative platelet-to-lymphocyte ratio predicts poor prognosis in breast

294 Cancer Control July 2016, Vol. 23, No. 3 Special Report

Resection of the First Rib With Preservation of the T1 Nerve Root in Pancoast Tumors of the Lung Andreas K. Filis, MD, Lary A. Robinson, MD, and Frank D. Vrionis, MD, PhD

Background: Surgical outcomes for Pancoast (superior sulcus) tumors of the lung have significantly changed during the last few decades and have improved with use of curative-intent surgery by utilizing en bloc complete resections. Methods: A retrospective analysis was conducted of 11 selected patients treated at Moffitt Cancer Center from 2007 to 2016. Data from patient records were collected and analyzed. Results: All 11 patients with a Pancoast tumor involving the first rib had their T1 root preserved at surgery. In 10 patients (90.9%), the tumor was removed en bloc. Clear margins of resection were documented in 4 cases (36.0%). No patient developed postoperative hand weakness, but 3 patients (27.3%) had minor postoperative complications, including air leak, chylothorax, and pericardial effusion. One iatrogenic injury to the subclavian artery was reported during surgery; the injury was subsequently repaired. No operative mortality was reported. Conclusions: Radical resection of Pancoast tumors is considered to be safe, and preserving the T1 nerve root provides more favorable, functional outcomes.

Introduction root), its involvement and consequent sacrifice by vir- Pancoast tumor of the lung, also known as superior tue of involvement of the first rib contributes to some sulcus tumor, constitutes a special variant of lung car- degree of functional impairment from paresis of the ip- cinoma localized at the apex of either lung. These carci- silateral hand. nomas comprise 3% to 5% of all cases of non–small-cell In this study, we report our experience with en lung cancer cases, most commonly squamous cell car- bloc resection of the tumor along with the first rib, cinoma.1 Up to 5% of Pancoast tumors are small cell with concomitant preservation of the T1 nerve root. carcinomas.2 Originally described by Pancoast3 in the early 20th century, these tumors were considered to Methods have a uniformly fatal prognosis. By employing com- Study Protocol and Patient Population bined triple-modality treatment using chemoradio- A retrospective cohort analysis was performed of pa- therapy with surgical resection, as described by the tients undergoing surgical resection for Pancoast tu- Southwest Oncology Group (SWOG),4 outcomes have mors with concomitant preservation of T1 nerve root changed and 5-year survival rates have been markedly between 2007 and 2014 at the H. Lee Moffitt Cancer improved (≤ 55% with complete resection).5 Center & Research Institute (Tampa, Florida). Although radical resection is often considered the key element to increased survival following induction Data Collection chemotherapy or chemoradiotherapy, it may be associ- Eleven patients with Pancoast tumors who underwent ated with some degree of functional compromise.1 Al- surgical treatment and T1 nerve root preservation were though the T1 nerve root is not the main contributor included in the study. Electronic medical records for of hand muscle function (by contrast to the C8 nerve these patients were reviewed. Demographical information, preoperative clinical characteristics, postop- From the Departments of Neuro-Oncology (AKF, FDV) and Thoracic erative assessments, and long-term follow-up data (if Oncology (LAR), H. Lee Moffitt Cancer Center and Research Institute, available) were extracted. University of South Florida Morsani College of Medicine (AKF, FDV), Tampa, Florida. Primary End Point Dr Filis is now affiliated with the Department of Neurosurgery, Imland Klinik, Rendsburg, Germany. The primary purpose of this study was to assess the Address correspondence to Andreas K. Filis, MD, Imland oncological and functional outcomes of patients un- Klinik, Lilienstrasse 20-28, 24768 Rendsburg, Germany. dergoing surgical resection for Pancoast tumors with E-mail: [email protected] T1 nerve root preservation. Submitted January 5, 2016; accepted March 31, 2016. No significant relationships exist between the authors and the companies/organizations whose products or services may be Presurgical Evaluation referenced in this article. All patients were initially evaluated by a thoracic onco-

July 2016, Vol. 23, No. 3 Cancer Control 295 logical surgeon and subsequently by a neurosurgeon. based on imaging and clinical findings, spinal instru- Pretreatment computed tomography (CT)–guided nee- mentation was made available during surgery. The dle biopsy was performed to confirm the diagnosis in SWOG protocol for preoperative chemotherapy, con- all cases. For operative planning, patients were evalu- sisting of 2 cycles of cisplatin and etoposide, with con- ated with contrast-enhanced CT of the chest and mag- current radiotherapy (45 Gy) was used in the earlier netic resonance imaging (MRI) of the brachial plexus. cases, and preoperative chemotherapy alone was given Positron emission tomography (PET) was obtained in the most recent 5 cases.4 If the pathology specimen for staging to rule out any local or systemic metasta- margins were microscopically positive for the tumor, ses. MRI of the brain for staging was also routinely ob- an additional radiation dose of 20 Gy was postopera- tained prior to treatment. Fig 1 shows the MRI results tively administered. Alternatively, patients who did of a patient with a Pancoast tumor. not receive preoperative radiotherapy were given a At the time of the neurosurgical consultation, dose of 60 Gy in cases of R1 resection (microscopi-

A B

C D

Fig 1A–D. — (A) Sagittal, (B) coronal, and (C) axial views of magnetic resonance imaging from a patient with a Pancoast tumor involving the T1 vertebral body and the (D) lower brachial plexus.

296 Cancer Control July 2016, Vol. 23, No. 3 cally positive margins) or 70 Gy in cases of R2 resection (grossly Suspicion CT of chest with contrast Thoracic surgery positive margins); radiotherapy of MRI of brachial plexus Neurosurgery was always postoperatively ad- Pancoast PET Consultation ministered to the tumor bed. tumor A unique multidisciplinary approach for Pancoast tumors Surgical Work-Up: has been put into place at Moffitt PFTs Cancer Center. Fig 2 illustrates the ABGs algorithm of the steps to be fol- Cardiac evaluation lowed, starting with the diagnosis (if indicated) and going through the selected treatment plan for Pancoast tumors. In our more than 20 years Moffitt Tumor Board review for of experience, no patient has had resection status R2 resection or N2 (ipsilateral me- diastinal) positive nodes. CT-guided CT-guided Statistical Analysis Chemotherapy biopsy for Yes Resectable? No biopsy for histology A descriptive analysis was per- histology formed; qualitative variables were reported as frequencies and pro- Contrast CT Concurrent portions, whereas quantitative of the chest and chemo- variables were reported as mean adrenals radiotherapy plus or minus standard deviation. Joint Surgery: Differences across the mean pre- Upper lobectomy/apical operative and postoperative sub- Partial segmentectomy jective pain scores were obtained Response? response and chest wall resection or stable ± vertebrectomy, by an independent sample t test. disease instrumentation, A type 1 error at 5% was set as lymph-node dissection being statistically significant. All Progressive disease analyses were 2-tailed and performed using SPSS version 22.0 Resectable (IBM, Armonk, New York). without nodal Yes involvement Surgical Technique At Moffitt Cancer Center, the No Positive margins Negative margins surgical resection is jointly performed by a thoracic oncological 2 Concurrent 60 Gy surgeon and a neurosurgeon. 50 Gy chemoradiotherapy if radiotherapy The operative plan consists of radiotherapy margins posterolateral thoracotomy with microscopically upper lobectomy or segmental positive (R1) resection of the tumor by the tho- 70 Gy radiotherapy if racic oncological surgeon, with margins grossly the neurosurgeon performing en positive (R2) bloc resection of the tumor along the posteromedial, superior chest wall, the brachial plexus, and Fig 2. — Flowchart illustrating the multidisciplinary approach to Pancoast tumors at Moffitt Cancer Center. spine (if involved). ABG = arterial blood gas, CT = computed tomography, MRI = magnetic resonance imaging, The patient is placed in the PET = positron emission tomography, PFT = pulmonary function test. lateral decubitus position with the lung deflated using a double- lumen endotracheal tube. The initial thoracotomy inci- is inferiorly transected along the incision and the me- sion is posteriorly and superiorly extended, parallel to dial attachment of the trapezius and rhomboid muscles the medial border of the scapula. The trapezius muscle to the spine are detached, allowing mobilization of

July 2016, Vol. 23, No. 3 Cancer Control 297 the scapula and exposure of the upper chest wall. Ini- placement is then used, if necessary (Fig 4). tially, the lateral chest wall cuts are made to the first Intraoperative somatosensory-evoked poten- and typically the second and third ribs, as required. tial monitoring is used with all surgical procedures A window of approximately 1 cm in the chest wall al- directly involving vertebrae. If concerns exist about lows visualization of the upper chest cavity during the close soft-tissue surgical margins, then an intraoper- surgery. Caution should be exercised while resecting ative frozen section evaluation of the surgical speci- the anterior segment of the first rib to avoid injuries men is performed. to the overlying subclavian vessels by periosteal dissection and bone removal. The underlying intercostal Follow-Up Care nerves and vessels are distally cut, together with the At Moffitt Cancer Center, the initial inpatient postop- ribs. Subsequently, ipsilateral paraspinal muscles are erative care is primarily handled by the thoracic onco- partially dissected off the transverse processes, expos- logical surgeon in conjunction with the neurosurgeon; ing the lateral aspect of the facet joints and pars as well this is because any potential complications or prob- as the transverse processes. Using a small osteotome, lems are likely to be related to the lung resection. Pa- the transverse processes are removed. The costover- tients are then evaluated in the outpatient clinic 2 to tebral ligaments are cut and a Cobb elevator is placed 3 weeks following surgery by the thoracic oncological between the head of the rib and the lateral aspect of surgeon and the radiation oncologist. The neurosurgi- the vertebral body. By elevating the proximal rib, the cal evaluation occurs within 8 to 12 weeks following exiting nerve roof is stretched, clipped, and cut (for T2 surgery, after which most of the neurological outcome and below). is apparent. At 6 weeks, adjuvant radiotherapy at a For the T1 nerve root, special attention is given, dose of 65 Gy is delivered to the tumor bed in patients both at the foramen and distally at the plexus, supe- who only received preoperative induction chemothera- rior to the first rib. The attachments of the scalenus py. Long-term follow-up care is usually coordinated by muscles to the first rib must be carefully dissected un- the medical oncologist and radiation oncologist. til the inferior part of the brachial plexus is exposed. The C8 nerve root above the first rib and the junction Results of the T1 nerve root with the former superiorly and lat- The mean age of the 11 patients who underwent erally to the first rib is visualized. The T1 nerve root is surgical management for Pancoast tumors was subsequently traced proximally along its course posterior to the proximal segment of the first rib. If the surgeon has difficulty in identifying the proximal T1 nerve root, then removal of the lateral aspect of the First rib pars-facet complex may be required at the T1 to T2 level. Fig 3 shows key anatomical landmarks and their relation to the Pancoast tumor during surgical dissection. Following identification of the T1 nerve proxi- C8 mally and distally, mobilization of the tumor is initiated. The portion of chest wall for resection is inferi- Lung orly pushed to identify the subclavian artery and the stellate ganglion. Tumor extension to involve the latter structure requires cutting the ganglion as low as possible to avoid Horner syndrome. The proximal first rib must be carefully rotated between the C8 and T1 nerve roots and inferiorly pushed together with the rest of the specimen, which is then marked for orientation, T1 and the entire specimen with the en bloc attached tumor and lung tissue is sent to pathology. This description of the resection is applicable to the majority of Pancoast tumors involving the chest Pancoast wall. In the event that the tumor invades the lateral tumor aspect of the vertebral body, the paraspinal muscle is severed and retracted, following which unilateral lam- inectomy is performed. Following the clipping of the involved nerve roots proximal to the ganglion, osteot- omy medial to the pedicle is performed. Pedicle screw Fig 3. — Illustration of the anatomical considerations for resecting a instrumentation or single-rod vertebral body screw Pancoast tumor while preserving the T1 nerve.

298 Cancer Control July 2016, Vol. 23, No. 3 A most recent 5 patients (45%); these patients received preoperative chemotherapy alone. In 2 patients (18%), chemoradiotherapy or radiotherapy did not precede resection; rather, full-dose radiotherapy was postoperatively administered. Subclavian artery involvement was seen in 2 patients (18%), 1 of whom had an iatrogenic injury that was intraoperatively repaired. Two patients (18%) required intraoperative spinal instrumentation. The rate of mean blood loss was 436 ± 408 mL. Four patients (36%) had margins free of disease based on the results of the final pathological evaluation. In all cases, postoperative function of the hand was normal. Although preservation of the T1 nerve root in its continuity could be achieved for B all 11 patients, 1 patient (9%) developed hypothenar atrophy with altered sensation along the C8 and T1 distribution but without significant hand weakness. Three patients (27.3%) developed postoperative complications, including late pericardial effusion 1 month after surgery, persistent air leak, and chylothorax (n = 2) that required repair with re-entry thoracotomy at 1 week. No operative mortality was reported. Of the 11 patients, 5 died within a mean of 21.9 months following surgery. Causes of death included distant metastasis to the brain (n = 1) and adrenal metastases (n = 1); these patients developed metastases 14 and 15 months after lung surgery, respectively. Three additional patients died a mean of 15 months after surgery, but the causes of these deaths were unknown because the patients were lost to follow-up.

Discussion The management of Pancoast tumors has significantly changed during the last few decades. Locally advanced disease with invasion into the vertebral bodies was historically considered a relative contraindication for surgery. Komaki et al6 reported a significantly improved survival rate in these patients who Fig 4. — Postoperative (A) lateral and (B) anteroposterior radiographs underwent surgical resection compared with those show the instrumentation used in a patient after a more extensive resec- who did not. Gandhi et al7 reported outcomes in pa- tion of the tumor. tients with Pancoast tumors with vertebral involvement who were preoperatively treated with radio- 60.8 ± 11.8 years (range, 43–81 years; median, 59 years), therapy, and they reported an increase in the 2-year and 82% (n = 9) were men. Their mean height, survival rate among those treated with combined mo- weight, and body mass index were 172.1 ± 7.3 cm, dality therapy compared with those given monothera- 69.4 ± 10.7 kg, and 23.5 ± 3.9 kg/m2, respectively. py. SWOG researchers4 reported a 5-year survival rate Their mean subjective preoperative pain score was of 54% in patients with T3-4N0 disease who under- 4.4 ± 2.9; this score improved early on following sur- went complete R0 surgical resection following preop- gery to 4.1 ± 2.2, although this was not statistically erative, concurrent chemotherapy and radiotherapy. significant (P = .598). Concurrent induction chemoradiotherapy has All patients presented with radicular thoracic pain, been the standard of care since 2001, which is when and none had any evidence of metastasis on pretreat- the Intergroup Pancoast tumor trial results were pub- ment evaluation. All patients were evaluated by the lished.4 However, in this study population (perfor- same thoracic and neurosurgical team. The SWOG mance status 0–1), the induction chemoradiation regi- regimen4 was followed in most patients, except in the men was so toxic that a treatment-related mortality rate

July 2016, Vol. 23, No. 3 Cancer Control 299 of 2.7% was observed, and only 76.0% of all study pa- compared with standard preoperative chemoradio- tients underwent potentially curative surgery.4 therapy are being analyzed in a large series at Moffitt Despite improving options for chemotherapy and Cancer Center (LAR, June 8, 2016). radiotherapy, gross complete surgical resection re- Similar to a prior series,14 a relative high rate of mains fundamentally important for better rates of sur- cerebral metastasis (27%; n = 3) and local recurrence vival, although some neurological compromise may (9%; n = 11) was observed in our series after long-term be seen. Such compromise is generally related to the follow-up. Based on our experience, free surgical mar- anatomical nature of tumor involvement in the neu- gins and maximal multimodality therapy did not pre- rovascular structures, including the subclavian artery clude the occurrence of late cerebral metastases. MRI and vein and the lower brachial plexus trunks. In 2007, of the brain is not a routine part of the treatment plan Davis et al8 described how Paulson routinely dissected for long-term follow-up, and it is only preoperatively the lower plexus to obtain radical resection. In 1975 performed. Paulson9 reported a 5-year survival rate of 34% with The results of T1 nerve root functional preserva- preoperative radiotherapy and surgical resection. Ac- tion in our series are similar to those of Davis et al,8 cording to Davis et al,8 Shahian et al10 were the first to with approximately 91% of patients having good func- report on postoperative functional outcomes with radi- tional, postoperative outcomes. One patient (9%) had cal nerve root resection. In other published series, the ipsilateral hypothenar atrophy and altered sensation T1 nerve root was either resected or not reported.11-13 along the distribution of the C8 and T1 nerve roots, The report by Davis et al8 was the initial series that thereby correlating with the anatomical preservation reported on T1 preservation as a key element of sur- of T1 nerve root intraoperatively. Preoperative chemo- gical treatment in this patient population. Although it therapy alone facilitated the optimal dissection, owing was a small series (n = 7), complete tumor resection to tumor shrinkage, maintenance of more visible dis- with sparing of the T1 nerve root was achieved in section planes, and satisfactory clinical outcomes. 5 patients (71%).8 Preoperative chemoradiotherapy was performed and the resection was jointly carried out by Limitations a thoracic surgeon and a neurosurgeon, which is simi- The primary limitations of this study are those related lar to our patient cohort. Following this paradigm, the to its retrospective nature, including inherent selection group achieved a 2-year survival rate of 80% and nor- bias and a smaller sample size. This study focused on mal hand function.8 describing the functional outcomes of a select group of In our series of 11 patients for whom the T1 nerve patients with Pancoast tumor who underwent T1 nerve root was preserved during surgery, 5 patients (45%) re- root preservation. Inferential statistics were not feasi- ceived preoperative chemotherapy, 4 (36%) received ble to explore other clinical parameters. preoperative chemoradiotherapy, and 2 patients (18%) received postoperative radiotherapy alone. The Conclusions long-term survival rate with these varying approaches With technical advancements, surgical resection for is still pending evaluation. Pancoast tumors appears to be a feasible therapeu- During the last 5 years, we have adopted an ap- tic option that should be pursued using a multidisci- proach of 3 cycles of preoperative chemotherapy alone plinary team approach. Induction chemoradiotherapy followed by surgical resection at 3 to 5 weeks, then fol- or chemotherapy alone preceding surgical resection lowed by full-dose radiotherapy (65 Gy) at the tumor provides overall improvements in resectability and sur- bed. After induction chemotherapy alone, typically a vival rates. Meticulous intraoperative dissection with clear dissection plane can be seen for surgery; this is complete resection by a multidisciplinary surgical team critical for radical resection and helps to minimize in- will generally lead to improved functional outcomes by traoperative complications. This chemotherapy-alone preserving the T1 nerve root and without sacrificing approach has lower rates of morbidity and helps to fa- oncological principles. cilitate surgical resection; full-dose radiotherapy postoperatively is also possible, which is especially impor- References tant for local control if close or microscopically positive 1. Foroulis CN, Zarogoulidis P, Darwiche K, et al. Superior sulcus (Pancoast) tumors: current evidence on diagnosis and radical treatment. margins are present. J Thorac Dis. 2013;5(suppl 4):S342-S358. In an unpublished series, nearly 100% of study 2. Setzer M, Robinson LA, Vrionis FD. Management of locally advanced Pancoast (superior sulcus) tumors with spine involvement. Cancer patients completed therapy, including surgery, and Control. 2014;21(2):158-167. toxicity related to the induction regimen was not a 3. Pancoast HK. Superior pulmonary sulcus tumor. JAMA. 1932;99:16 -17. 4. Rusch VW, Giroux DJ, Kraut MJ, et al. Induction chemoradiation limiting factor (compared with concurrent chemora- and surgical resection for superior sulcus non-small-cell lung carcinomas: diotherapy; Lary A. Robinson, MD, personal commu- long-term results of Southwest Oncology Group Trial 9416 (Intergroup Trial 0160). J Clin Oncol. 2007;25(3):313-318. nication, June 8, 2016). The long-term results of this 5. Glassman LR, Hyman K. Pancoast tumor: a modern perspective on modified, preoperative chemotherapy-alone approach an old problem. Curr Opin Pulm Med. 2013;19(4):340-343.

300 Cancer Control July 2016, Vol. 23, No. 3 6. Komaki R, Mountain CF, Holbert JM, et al. Superior sulcus tumors: treatment selection and results for 85 patients without metastasis (Mo) at presentation. Int J Radiat Oncol Biol Phys. 1990;19(1):31-36. 7. Gandhi S, Walsh GL, Komaki R, et al. A multidisciplinary surgical approach to superior sulcus tumors with vertebral invasion. Ann Thorac Surg. 1999;68(5):1778-1784. 8. Davis GA, Knight S. Pancoast tumor resection with preservation of brachial plexus and hand function. Neurosurg Focus. 2007;22(6):E15. 9. Paulson DL. Carcinomas in the superior pulmonary sulcus. J Thorac Cardiovasc Surg. 1975;70(6):1095-1104. 10. Shahian DM, Neptune WB, Ellis FH Jr. Pancoast tumors: improved survival with preoperative and postoperative radiotherapy. Ann Thorac Surg. 1987;43(1):32-38. 11. Martinod E, D’Audiffret A, Thomas P, et al. Management of superior sulcus tumors: experience with 139 cases treated by surgical resection. Ann Thorac Surg. 2002;73(5):1534-1540. 12. Urschel HC Jr. Superior pulmonary sulcus carcinoma. Surg Clin North Am. 1988;68(3):497-509. 13. Kraut MJ, Vallières E, Thomas CR Jr. Pancoast (superior sulcus) neoplasms. Curr Probl Cancer. 2003;27(2):81-104. 14. Shah H, Anker CJ, Bogart J, et al. Brain: the common site of relapse in patients with Pancoast or superior sulcus tumors. J Thorac Oncol. 2006;1(9):1020-1022.

July 2016, Vol. 23, No. 3 Cancer Control 301 Special Report

Feasibility, Preliminary Efficacy, and Lessons Learned From a Garden-Based Lifestyle Intervention for Cancer Survivors Colleen K. Spees, PhD, RDN, Emily B. Hill, Elizabeth M. Grainger, PhD, RDN, Jackie L. Buell, PhD, RDN, Susan E. White, PhD, Matthew D. Kleinhenz, PhD, and Steven K. Clinton, MD, PhD

Background: Cancer survivors remain at increased risk for secondary malignancies, comorbidities, and all-cause mortality. Lifestyle behaviors, such as diet and physical activity, are strongly linked to a decreased risk of chronic disease and improved health outcomes, yet a paucity of research has been conducted in this vulnerable population. Methods: Adult cancer survivors were recruited to participate in Growing Hope, an experimental single-group study designed to assess the feasibility and efficacy of a theory-driven and evidence-based intervention. For 4 months, 22 participants received group and individual education and had access to harvesting fresh produce at an urban garden. Data on program satisfaction, compliance, diet, and physical activity were collected via surveys; anthropometrics, blood values, and skin carotenoids were objectively measured. Results: The intervention resulted in significant improvements in consumption of fruits and vegetables (P = .003), decreased consumption of red and processed meats (P = .030) and sugar-sweetened beverages (P = .020). Levels of skin carotenoids, fasting blood glucose, and non–high density lipoprotein cholesterol were also significantly improved (P = .011, P = .043, and P = .05, respectively). Conclusions: The results of this study support the feasibility and efficacy of a multifaceted, garden-based intervention for cancer survivors. In addition, these preliminary results demonstrate a positive impact aligning with the current lifestyle recommendations for cancer survivorship. Larger randomized controlled trials are warranted to define impact on sustained health outcomes.

Introduction cardiovascular disease, and type 2 diabetes mellitus.7-9 Advances in early cancer detection and treatment have Long-term secondary effects of treatment often affect increased the number of Americans who survive cancer.1 this population, resulting in higher rates of all-cause In 1971, cancer survivors were numbered at 3 million; mortality.10,11 by 2022, the number of survivors is expected to exceed Cancer survivors often lack consistent follow-up 18 million.2 Yet the completion of active cancer treat- care and management as they transition from active ment frequently does not signal an end to the disease patient with cancer to a cancer survivor.3,4,12 For many, or its ill effects. Cancer survivors often face ongoing the encounter with cancer provides a renewed appre- medical issues as well as emotional and financial chal- ciation for health and an opportunity for empowerment. lenges.3-6 In addition, survivors remain at increased Survivors demonstrate a desire to take “ownership” of risk for recurring primary cancers and secondary ma- their own health and make changes that enhance or con- lignancies as well as comorbidities such as osteoporosis, tribute to the success of cancer therapy, the amelioration of treatment-related toxicity, and the overall quality of survivorship. From the Medical Dietetics & Health Sciences (CKS, EBH), Health Information Management and Systems (SEW), School of Health and Modifiable lifestyle behaviors, including consum- Rehabilitation Sciences (JLB), Comprehensive Cancer Center (CKS, ing a plant-based diet and engaging in physical activ- EMG, SKC), Department of Internal Medicine – Division of Hema- ity, have been shown to increase the quality of life in tology and Oncology (EMG, SKC), College of Medicine, and the De- 13,14 partment of Horticulture and Crop Science (MDK), The Ohio State cancer survivors. More than 35% of cancer-related University, Columbus, Ohio. deaths are estimated to have been prevented by diet, Address correspondence to Colleen K. Spees, RDN, PhD, Medical Dietet- physical activity, and other lifestyle modifications.13,15,16 ics and Health Sciences, OSU College of Medicine, 453 West Tenth Ave- However, the dietary and physical activity patterns of nue, 306E Atwell Hall, Columbus, OH 43210. E-mail: [email protected] Submitted February 18, 2016; accepted April 4, 2016. the majority of cancer survivors mirror that of the US 17 This project was supported by the OSU Food Innovation Center, the population and remain suboptimal. Most cancer sur- National Center for Advancing Translational Sciences, National vivors consume diets high in saturated fat and low in Institutes of Health, through Grant UL1TR001070, JamesCare for fruits, vegetables, and whole grains, and they often fail Life, and Waterman Farm. The content is solely the responsibility of 7,18 the authors and does not necessarily represent the official views of to meet physical activity guidelines. the National Institutes of Health. Impactful programs regarding diet, nutrition,

302 Cancer Control July 2016, Vol. 23, No. 3 and lifestyle are critical for survivorship. Modifi- stable angina), taking medications that did not allow able lifestyle behaviors, including engaging in regu- for increased intake of fruits and vegetables (eg, warfa- lar physical activity, consuming a plant-based diet, rin), planning to begin medications during the course and maintaining a healthy body weight, have been of the intervention, use of select nonprescription sub- shown to improve outcomes, overall functioning, stances (eg, herbs, botanical products), active metabol- and quality of life among cancer survivors.17,19-21 Ex- ic or digestive illnesses (Crohn disease, celiac disease, pert committees, such as the American Institute irritable bowel syndrome), renal or hepatic insufficien- for Cancer Research/World Cancer Research Fund cy, cachexia, or women who were pregnant. All study and the American Cancer Society, have defined procedures were approved by the Institutional Review evidence-based recommendations for physical ac- Board and informed consent was obtained from all in- tivity and dietary patterns relative to cancer sur- dividual participants included in the study. vivors.13,14,22,23 Results of the largest prospective study on diet and lifestyle behaviors of more than Study Design 520,000 individuals revealed that participants with This single-arm study was designed to assess feasi- the greatest adherence to the American Institute for bility and compliance with a novel and multifaceted Cancer Research/World Cancer Research Fund guide- intervention. Drawing on successful theoretical prin- lines had a 34% lower rate of death from all causes, ciples specific to cancer survivorship, the interven- including cancer-related death, than those with the tion was based on social cognitive theory framework lowest concordance with the recommendations.24 Be- and supplemented by motivational interviewing.27-29 cause lifestyle behaviors are influenced by a myriad After completion of baseline clinical, anthropomet- of factors, the Academy of Nutrition and Dietetics rec- ric, and dietary assessments, study participants were ommends interventions that address several key com- involved in a multifaceted intervention focusing on ponents, rather than targeting any single factor, when cancer survivor–specific nutrition, physical activity, creating programs to facilitate behavioral change.25,26 and behavioral modifications delivered within a garden setting. The 4-month intervention components Purpose included13,14,22,23,30: The purpose of this study was to determine the fea- • Harvesting produce (fruit, vegetables, herbs) sibility, acceptability, and preliminary efficacy of a ≥ 3 times/week at an urban garden aligned with multifaceted, evidence-based intervention for cancer a comprehensive cancer center supported by the survivors transitioning out of active treatment and or- National Institutes of Health chestrated around a season of herb, fruit, and vegeta- • Biweekly group education classes structured ble harvesting in an urban garden. around evidence-based cancer survivor guidelines • Access to remote motivational interviewing coach- Methods ing by a trained registered dietitian nutritionist Participants and Recruitment • Access to a secure online web portal to provide Adult cancer survivors were recruited from the James lifestyle behavior recommendations and survivor- Cancer Hospital and Solove Research Institute (Co- specific health and wellness tips, recipes, and lumbus, OH) at The Ohio State University (OSU) other resources Comprehensive Cancer Center, its associated on- At the end of the 4-month intervention, the clinical, cology clinics, and its affiliated JamesCare for Life anthropometric, and dietary assessments were repeated. community-based cancer survivor outreach program. Garden Harvesting and Tracking: JamesCare Study brochures were distributed, and interested indi- for Life, an extension of the OSU Comprehensive Can- viduals were instructed to e-mail the study coordinator cer Center, provides resources and services to assist and complete an 11-item eligibility screening. cancer survivors and their families during their can- Participation inclusion criteria included: adults cer journey. Among the offerings at JamesCare for Life, 18 years of age or older, an English speaker, access to programs to educate cancer survivors on the critical the Internet, basic computer skills, and cancer survi- role of nutrition, physical activity, and other modifiable vors who had completed active cancer treatment (che- lifestyle behaviors for cancer prevention and survivor- motherapy, radiotherapy, and/or surgery) within the ship are available. One such program is the Waterman previous 24 months. Use of adjuvant hormone therapy Farm’s Garden of Hope, a community garden estab- was acceptable. Exclusion criteria were: cognitively un- lished in 2012, which is supported by the OSU College able to provide informed consent, had physical or men- of Food, Agriculture, and Environmental Sciences. The tal limitations that would prevent full participation in garden is a 2.5-acre plot with herbs, fruits, and vege- the program, receiving active cancer treatment, preex- tables. Cancer survivors enrolled in the JamesCare for isting medical conditions that precluded unsupervised Life program were permitted to harvest up to 3 times physical activity (eg, severe orthopedic conditions, un- per week (in 2-hour blocks) during the growing season

July 2016, Vol. 23, No. 3 Cancer Control 303 (typically May to October) free of charge. week, the motivational interviewing coach sent weekly Prior to harvesting at the Garden of Hope, par- e-mails, text messages, or both to encourage adher- ticipants attended a mandatory garden orientation ence to the recommendations and to emphasize the to complete registration forms, receive harvesting in- key points of the group education sessions. structions, become familiar with the policies and pro- Web Portal: Participants were provided unlimited cedures of Waterman Farm, and meet the support staff access to a secure web portal where they could access at JamesCare for Life. At this orientation, participants additional evidence-based lifestyle recommendations, were given a red harvest bag and a registration card current research, cancer survivor–specific health and with a unique bar code to electronically track the num- wellness tips, recipes utilizing produce from the gar- ber of harvests attended throughout the season. den, links to external websites with evidence-based At the harvest check-in station, participants pre- content, nutritional information, and updated lists of sented their Garden of Hope registration cards to staff the produce ready to harvest each week at the garden. at JamesCare for Life. Using a mobile scanning device, Participants also had access to the group education cal- the staff members at JamesCare for Life scanned each endar that listed expert presenters, class topics, and participant’s unique bar code to track harvest visits. objectives. The web portal also housed electronic cop- During each 2-hour harvest block, Garden of Hope ies of all class handouts, slides, and recipes from each registrants were encouraged to fill up their red bag session. with fresh garden produce and interact with dietetic in- terns and dietitians that attended each harvest. These Data Collection and Analysis trained volunteers were accessible to participants and Medical, dietary, and lifestyle questionnaires were provided safe food-handling tips, recipe suggestions, collected at baseline (month 0) and immediately pos- nutrition education, harvesting assistance, and addi- tintervention (month 4). Similarly, objective anthropo- tional motivation and support.30 metric and clinical biomarkers were collected at the Group Education and Training: In addi- baseline and postintervention periods during clinical tion to harvesting, study participants attended 10 ex- laboratory visits. To obtain additional programmatic pert-led classes in a group setting focusing on cancer feedback, participants completed a class evaluation survivor–specific guidelines. These 1-hour classes, after each group education session throughout the presented in a conference room adjacent to the gar- program. Study data were collected and managed us- den, were held biweekly on Monday evenings, and ing the Research Electronic Data Capture data capture they offered in-depth discussions of evidence-based tools hosted at OSU.32 survivorship guidelines. Each class focused on a different recommendation and detailed evidence behind Feasibility the recommendation, application of the guidelines To determine feasibility, the reach, adoption, adher- into practice, and behavioral regulation guidance ence, and acceptability of the program were mea- (goal setting, overcoming barriers, coping strategies, sured.33 Each component of the intervention was also and self-management). In addition, each class ended evaluated for translation of research into practice.34,35 with a cooking demonstration that utilized produce Reach of the program was measured via recruitment harvested from the garden that day to demonstrate and enrollment logs. Adoption of the program was safe food-handling and preparation methods, to share evaluated through participant retention and question- healthy recipes, and to serve as an example of translat- naires. Tracking of participation at all harvests and ing guidelines into practice. The chef and dietitians at group education sessions were tallied throughout the the medical center led these cooking demonstrations. intervention to further define compliance and use of Motivational Interviewing: Motivational in- the program. To assess acceptability of the interven- terviewing is a clinical practice that grew out of theo- tion, individual group education evaluations and over- retical and practical considerations. Motivational in- all programmatic satisfaction surveys were completed. terviewing relies on both relational aspects between interventionist and client as well as behavioral rein- Dietary Intake forcement.31 Our motivational interviewing coach was The validated, 26-item Dietary Screener Questionnaire a registered dietitian nutritionist trained and compe- developed by the National Cancer Institute was used tent in motivational interviewing techniques as well to evaluate dietary intake.36 This screening tool assess- as medical nutrition therapy, theoretical behavioral es the frequency of consumption of a wide variety of constructs, oncology, and sports nutrition.29,31 Partici- foods and beverages. Participants were asked to report pants communicated with the motivational interview- their frequency of consumption of food items during ing coach via e-mail, text, voice-over internet proto- 1 month prior, and these data were used to estimate col methods, or phone throughout the intervention. intakes of fruits, vegetables, red and processed meats, In addition to monitoring participant progress each dairy, and added sugars.

304 Cancer Control July 2016, Vol. 23, No. 3 Clinical Measurements tial cancer diagnosis was 59 years (Table 1). Breast and Participants completed laboratory visits at baseline and prostate cancers were the most prevalent primary can- postintervention to assess clinical and biological indi- cers reported in this cohort (see Table 1). ces of health risk. Prior to these laboratory visits, par- A total of 88% of participants completed preinter- ticipants were instructed to adhere to a 12-hour fast- vention and postintervention questionnaires, 80% at- ing period, hydrate, and abstain from vigorous exercise tended both laboratory visits, and 73% completed the and alcohol consumption 72 hours prior to the visit. postintervention programmatic evaluation. Overall Fasting samples were collected between 7:00 and study retention was high at 88% (n = 22/25), and no ad- 10:00 am to control for diurnal variation. Anthropometric data (height, weight) were collect- 37 ed to calculate body mass index. Participants were Table 1. — Baseline Characteristics of Study Participants weighed wearing light clothing and without shoes on (N = 22) a standard and calibrated digital scale, standing height Demographics No. of was measured, and blood pressure was obtained. Participants, Capillary blood samples were taken by trained n (%) personnel using standardized sterile techniques. Lipid Age, y Start of study 61.7 (mean) profile and blood glucose screening with a single fin- Primary cancer diagnosis 59.3 (mean) ger stick were performed. Skin carotenoid levels were Sex Male 5 (22.7) measured with a scanner that uses Raman spectroscopy to estimate skin carotenoids. The scanner has been Female 17 (77.3) demonstrated to be sensitive, reproducible, and correl- Race/Ethnicity Black/African American 1 (4.5) ative with both plasma carotenoid levels and fruit and White 21 (95.5) vegetable intake.38-44 Othera 0 (0) Marital Status Married 13 (59.1) Statistical Analysis Descriptive statistics were generated for all demo- Divorced 6 (27.3) graphical and outcome measurements. All analyses Never married 1 (4.5) were performed using SPSS version 21 (IBM SPSS Sta- Widowed 2 (9.1) tistics, Armonk, NY). Compliance was assessed using Otherb 0 (0) attendance records from harvests and the group educa- Education Completed < grade 12 0 (0) tion classes. Participants were considered compliant if Completed ≥ grade 12 or 3 (13.6) they attended 75% of the offered sessions. General Education Statistical analysis for the effect of the interven- Development equivalent tion on anthropometric, dietary, and clinical mea- Completed 1–3 y college 2 (9.1) sures were conducted by comparing the prestudy Completed ≥ 4 y college 7 (31.8) and poststudy scores. The hypothesis of no change in Earned professional or 10 (45.5) these variables was tested with paired t tests to deter- graduate degree mine which were significantly different. The P values Employment Employed or self-employed 10 (45.5) of the individual t tests were adjusted via a Bonferroni Retired 9 (40.9) correction. Statistical analysis for efficacy of the inter- Homemaker 3 (13.6) vention was performed using descriptive statistics for the outcome variables. A paired t test was performed Annual Household ≤ 19,999 0 (0) Income, $ to assess the significance of the shift in the relevant 20,000–24,999 2 (9.1) outcomes variables. 25,000–34,999 2 (9.1) 35,000–49,999 2 (9.1) Results 50,000–74,999 3 (13.6) The recruitment goals for this intervention were met ≥ 75,000 9 (40.9) within 4 weeks. A total of 35 adult cancer survivors Prefer not to answer 4 (18.2) were screened for eligibility and 25 (n = 20 women, n = 5 men) were deemed eligible and enrolled. During Primary Cancer Breast 12 (54.5) the course of the intervention, 3 female participants Prostate 4 (18.2) withdrew from the study secondary to cancer recurrence Otherc 6 (27.3) or related health concerns (n = 2) or personal issues aAmerican Indian, Asian, Hispanic, not specified, and Pacific Islander. that interfered with full participation (n = 1). Of the final bSeparated or domestic partner. cohort (N = 22), the majority was white and women; cColorectal, endometrial, leukemia, lung, lymphoma, melanoma, not the mean age was 62 years and the mean age of ini- specified, ovarian, skin, thyroid, and tonsillar.

July 2016, Vol. 23, No. 3 Cancer Control 305 verse events were reported as part of this study. The results of the postintervention program- Table 2. — Comprehensive Postintervention Programmatic Evaluation matic evaluation showed that more than 90% of the participants reported the dates and times for both Question Response No. of Responses, harvesting (Monday evening, Thursday morning, % Saturday morning) and biweekly group classes (Mon- Were the Garden of Hope harvesting Yes 15 (93.8) day evenings) were convenient (Table 2). All par- times (Monday evening and Thursday ticipants rated the overall program as either “very and Saturday mornings) scheduled good” or “excellent” and would recommend the pro- at a convenient time for you? No 1 (6.3) gram to other survivors, and they also reported that Were the group educational classes Yes 16 (100) the program helped them achieve better adherence (Monday evenings at 6 pm) scheduled at a convenient time for you? No 0 (0) to the evidence-based dietary recommendations and had a positive impact on their health (see Table 2). How would you rate the group educa- Poor 0 (0) tional sessions? Open-ended participant feedback from the postinter- Fair 0 (0) vention evaluation appears in Table 3. Good 1 (6.3) On average, participants harvested 1.5 times per Very good 8 (50.0) week during the 4-month growing season. The mean Excellent 7 (43.8) attendance rate for the group education was 73% Did the program help you to achieve Yes 16 (100) (n = 16). Objectively, participants rated each separate better dietary patterns that more education class positively utilizing a Likert-type scale closely aligned with the evidence- (0–5). More than 90% of participants rated the edu- based recommendations (primarily plant-based, rich in whole grains, No 0 (0) cational sessions as “very good” or “excellent” (see fruits, and vegetables and low in Table 2). Overall, 95% of session attendees “agreed” or sodium, simple sugars, and red/ “strongly agreed” that expert presenters met the class processed meats)? objectives. Subjective data were also positive on all Did the program impact your overall Yes 16 (100) health in a positive way? postclass evaluations, with recurring statements that No 0 (0) included words like “educational,” “informative,” and Did the program provide you with a Yes 15 (93.8) sense of community and support? “entertaining” (see Table 3). No 1 (6.3) Fifteen of the participants (68%) utilized the mo- Which program activity was the most Group 9 (56.3) tivational interviewing coach and all of them indicat- effective for you? classes ed they would utilize the motivational interviewing Harvesting 4 (25.0) coach if offered again in the future. Furthermore, 93% Remote 1 (6.3) of those participating in the motivational interview- motivational ing coaching “agreed” or “strongly agreed” that indi- interviewing vidualized motivational interviewing reinforced the coaching group education sessions. When asked to describe Web portal 2 (12.5) the most beneficial part of the motivational interview- Would you recommend this program Yes 16 (100) to other survivors? ing coach, the majority of comments indicated satis- No 0 (0) faction: Recurring phrases praised the accessibility, How would you rate the Harvesting Poor 0 (0) knowledge, and support of the motivational inter- Health Program as a whole? viewing coach (see Table 3). Fair 0 (0) Sixteen participants (73%) utilized the web portal Good 0 (0) with mixed results related to ease of use. Fifteen (94%) Very good 2 (12.5) of the portal users “somewhat” or “strongly” agreed the Excellent 14 (87.5) web portal reinforced the group education objectives and key topics. Of those who did not utilize the web portal, qualitative feedback included difficulty with decreased red and processed meat intake (P = .030; portal access (log in), frustration navigating through Fig). No participants reported smoking at baseline or the portal efficiently, and a preference for in-person or postintervention, and all participants reported alcohol telephonic correspondence (see Table 3). consumption within the recommended goal range at Based on the dietary intake data, participant ad- both baseline and postintervention. herence to the evidence-based guidelines for cancer Positive trends were noted in multiple clinical in- survivorship improved during the course of the inter- dices of health (Table 4). Blood pressure was the single vention. Significant improvements included increased variable that did not trend positively, yet the difference consumption of fruits and vegetables (P = .003), de- was clinically insignificant. Significant changes were creased consumption of added sugars (P = .020), and noted in values of fasting blood glucose, non–high

306 Cancer Control July 2016, Vol. 23, No. 3 Table 3. — Qualitative Postintervention Programmatic Evaluation Intervention Component Participant Feedback Garden harvesting The volunteers (students) in and out of the garden were extremely helpful. I liked the wonderful produce from the garden, the friendly and helpful volunteers … the study was a gift! This study was one of the highlights of my life. I can’t believe how much I have learned about “real” food! Group education Until now, I did not know all the health benefits associated with my dietary changes. Thank you for a wonderful opportunity. I feel very privileged. This has forever changed my life and way of eating. The classes were excellent and well presented. The Monday night classes and cooking demonstrations were very interesting. I can’t say enough good things about the health coaching and the website. It has made a lasting impression on my health. I thoroughly enjoyed attending the Monday night classes and following the weekly tips. … It has been a fantastic opportunity and I will miss all of you. Remote coaching I appreciated her knowledge base and sincere willingness to help. She had good ideas and made them relevant to me. The goals she helped me set were realistic. I loved having someone to be accountable to. I am definitely more informed about my food choices and have started making better choices more often. I’m trying to approach one issue at a time … and will continue to improve. “I got materials from [the remote motivational interviewing coach] to look at my own diet and track and develop my own dietary patterns. I think those were helpful for me and would be for others, too. Loved having a health coach and [the study principal investigator] in my back pocket. With their support and energy, I felt like I could do it. Being able to discuss my own personal nutrition issues was the most beneficial part of the health coaching. [The remote motivational interviewing coach] always provided helpful information and was supportive of any positive steps (no matter how small) that I made. Every communication appears to be online. It might have made a difference in my decision if the coach had made a telephone call. Some of us need to hear a voice until we get used to technology. Web portal The portal was very easy to use. I liked the website, the recipes, the tips, the help from [the remote motivational interviewing coach] … everything was valuable! The discussion boards took a little while to get used to navigating them and following the discussions. If there are other formats to use, that might help. Group support I did make many changes to my eating habits, and my husband has made the changes right along beside me. I only wish this kind of support were available while I was going through cancer treatment. When I first started in the study, I thought I knew a lot about nutrition. I discovered I did not. For the first time since my cancer diagnosis I felt like I had the know-how and support to make life-long changes. I enjoyed the camaraderie with other participants. I very much enjoyed the spirit of community. All of the study team was fabulous to work with, very knowledgeable, helpful, and enthusiastic. I was blessed to be a participant in this study.

density lipoprotein choles- Table 4. — Clinical Laboratory Results terol, and skin carotenoids (P = .043, P = .050, P = .011, Variable No. of Baseline Postintervention Difference, Participants (mean) (mean) P value respectively; see Table 4 and Fig). Weight, pounds 20 170.7 169.4 –1.2 (.603) Body mass index, kg/m2 20 28.1 27.9 –0.2 (.586) Discussion Systolic blood pressure, mm Hg 20 132.9 133.5 +0.7 (.815) This study demonstrated Diastolic blood pressure, mm Hg 20 82.5 83.2 +0.7 (.704) the feasibility, acceptability, Total cholesterol, mg/dL 19 211.3 201.7 –9.5 (.150) and preliminary efficacy of HDL cholesterol, mg/dL 18 61.4 65.2 +3.7 (.176) a comprehensive interven- Non–HDL cholesterol, mg/dL 18 152.3 137.9 –14.4 (.050)* tion delivered at an urban garden targeted toward Fasting blood glucose, mg/dL 19 97.5 93.8 –3.7 (.043)* cancer survivors. Feasibil- Skin carotenoid score (RRS) 20 38,225.5 43,251.4 +5,025.9 (.011)* ity was evaluated from par- *Statistically significant (P ≤ .05). ticipant attendance records HDL = high density lipoprotein, RRS = resonance Raman spectroscopy.

July 2016, Vol. 23, No. 3 Cancer Control 307 and follow-up questionnaires. The rates of rapid re- was rated as the second most important intervention cruitment, high enrollment, and low attrition suggest component. that the intervention was acceptable and well received. Nearly 70% of participants took advantage of the Previous behavioral interventions report approxi- remote motivational interviewing coach. Consistent us- mately two-thirds of participant attendance at educa- ers reported that the remote motivational interviewing tion classes.45-47 The ability to meet recruitment goals reinforced the classes and was critical in providing an within 30 days suggests demand for this type of inter- additional level of support often overlooked in other vention and appropriateness for the target population. behavioral interventions.48 Those who declined help The accrued sample was relatively modest in size and from the motivational interviewing coach felt that they homogenous in terms of primary cancer site (mainly did not have enough time or did not need coaching to breast and prostate), primarily white, older, well edu- achieve their goals. cated, and upper middle class. These factors are likely The secure web portal was utilized with mixed re- to have contributed to higher attendance and overall sults in almost three-quarters of the study cohort. The adherence rates. majority of participants enjoyed the website and com- The group education classes were led by an expe- mented that they appreciated the print-ready recipes, rienced expert aligned with the evidence-based topic copies of class handouts, safe food-handling tips, web presented. For example, the Food Safety for Survivors links, and continued updates. However, low comput- session was led by the state extension specialist in food er literacy was evident in a subset of the older partici- safety. Other classes were taught by registered dieti- pants throughout the intervention. The research team tians, medical oncologists, and academic researchers. and motivational interviewing coach attempted to re- This level of expertise allowed for additional support solve these issues throughout the study but remained and discussions of issues pertaining to advanced top- largely unsuccessful. These participants appeared to ics such as adjuvant therapy, post-treatment consider- prefer face-to-face or telephonic contact and had little ations, and ongoing survivor care plans. Overall, the interest in learning to navigate the website. Similar majority of participants reported that the group classes findings have been found in other interventions target- were the most important component of the interven- ing older individuals.49,50 tion leading to their success. Harvesting free produce Using evidence-based guidelines for cancer pre-

Pre Post A B C

6 6 • 1.5

5 • 4 1.0 4

Cup • Unit 3 2 0.5

2 0 0.0 Added Sugar, teaspoon Sugar, Added n = 22 n = 21 n = 22 Fruit and Vegetable Consumption Sugar-Sweetened Beverage Consumption Red and Processed Meat Consumption

D E F

80 120 250

110 60 200 100 40 150 90

100 20 Glucose Level, mg/dL 80 Cholesterol Level, mg/dL n = 20 n = 19 n = 18 Raman Intensity Count, in thousands Skin Carotenoids Fasting Blood Glucose Non–HDL Cholesterol

Fig A–F. — Preintervention to postintervention changes in dietary intake and indices of health. Box plots depicting significant differences in (A) fruit and vegetables (P = .003), (B) sugar-sweetened beverages (P = .020), (C) red and processed meats (P = .030), (D) skin carotenoids (P = .011), (E) fasting blood glucose (P = .043), and (F) non–HDL cholesterol (P = .050). HDL = high density lipoprotein.

308 Cancer Control July 2016, Vol. 23, No. 3 AICR guidelines for cancer prevention is associated with lower mortality vention and survivorship, significant improvements among older female cancer survivors. Cancer Epidemiol Biomarkers Prev. aligning with current recommendations were demon- 2013;22(5):792-802. 8. Berrino F. Lifestyle prevention of cancer recurrence: the yin and the strated for dietary intake (increased consumption of yang. In: Zappia V, Panico S, Russo GL, et al, eds. Advances in Nutrition and fruits and vegetables and decreased consumption of Cancer. Cancer Treatment and Research. Berlin: Springer; 2014:341-351. 9. Demark-Wahnefried W, Platz EA, Ligibel JA, et al. The role of obe- red, processed meats and sugar-sweetened beverages) sity in cancer survival and recurrence. Cancer Epidemiol Biomarkers Prev. and clinical indices of health (improved fasting blood 2012;21(8):1244-1259. 10. McCullough ML, Patel AV, Kushi LH, et al. Following cancer pre- glucose, systolic blood pressure, non–high density li- vention guidelines reduces risk of cancer, cardiovascular disease, and all- poprotein cholesterol, and skin carotenoids). cause mortality. Am Soc Prev Oncol. 2011;20(6):1089-1097. 11. Early Breast Cancer Trialists’ Collaborative Group. Favourable and un- favourable effects on long-term survival of radiotherapy for early breast can- Limitations cer: an overview of the randomised trials. Lancet. 2000;355(9217):1757-1770. 12. McCorkle R, Ercolano E, Lazenby M, et al. Self-management: A limitation to this study was the small sample size enabling and empowering patients living with cancer as a chronic illness. and lack of control group, although the size was ap- CA Cancer J Clin. 2011;61(1):50-62. 13. Wiseman M. The second World Cancer Research Fund/American propriate for a feasibility pilot study. A Hawthorne ef- Institute for Cancer Research expert report. Food, nutrition, physical ac- fect — whereby the mere presence of the intervention, tivity, and the prevention of cancer: a global perspective. Proc Nutr Soc. 2008;67(3):253-256. not the intervention itself, is associated with favorable 14. Kushi LH, Doyle C, McCullough M, et al. American Cancer Society changes in outcome measures — must be considered. guidelines on nutrition and physical activity for cancer prevention: reducing the risk of cancer with healthy food choices and physical activity. CA Cancer Biases associated with drop-out rates, although mini- J Clin. 2012;62(1):30-67. mal, may have resulted in an overestimation of the 15. Anand P, Kunnumakkara AB, Kunnumakara AB, et al. Cancer is a preventable disease that requires major lifestyle changes. Pharm Res. effect of the intervention. It is possible that observed 2008;25(9):2097-2116. changes were caused by a “healthy participant effect,” 16. Calle EE, Rodriguez C, Walker-Thurmond K, et al. Overweight, obesity, and mortality from cancer in a prospectively studied cohort of U.S. adults. in which participants may have self-reported healthi- N Engl J Med. 2003;348(17):1625-1638. er lifestyles than nonparticipants and, thus, may have 17. Spark LC, Reeves MM, Fjeldsoe BS, et al. Physical activity and/or dietary interventions in breast cancer survivors: a systematic review of the been more likely to comply with the study protocol. maintenance of outcomes. J Cancer Surviv Res Pract. 2013;7(1):74-82. However, improvements seen in biological indices of 18. Blanchard CM, Courneya KS, Stein K. Cancer survivors’ adherence to lifestyle behavior recommendations and associations with health-related health were aligned with self-reported gains (eg, fruit quality of life: results from the American Cancer Society’s SCS-II. J Am and vegetable intake, skin carotenoid levels), thus pro- Soc Clin Oncol. 2008;26(13):2198-2204. 19. Ballard-Barbash R, Friedenreich CM, Courneya KS, et al. Physical viding more information than either presented in iso- activity, biomarkers, and disease outcomes in cancer survivors: a systematic lation. In addition, we did not have a diverse or large review. J Natl Cancer Inst. 2012;104(11):815-840. 20. Hawkes AL, Pakenham KI, Chambers SK, et al. Effects of a multiple enough sample to permit adequate comparisons be- health behavior change intervention for colorectal cancer survivors on psy- tween variables of interest. chosocial outcomes and quality of life: a randomized controlled trial. Ann Behav Med. 2014:1-12. 21. Nathan PC, Ford JS, Henderson TO, et al. Health behaviors, medi- Conclusions cal care, and interventions to promote healthy living in the Childhood Cancer Survivor Study cohort. J Clin Oncol. 2009;27(14):2363-2373. The results of this study support the feasibility and 22. Ligibel JA, Denlinger CS. New NCCN guidelines for survivorship efficacy of a multifaceted, garden-based intervention care. J Natl Compr Cancer Netw. 2013;11(5 suppl):640-644. 23. McCabe MS, Bhatia S, Oeffinger KC, et al. American Society of for cancer survivors. In this pilot study, the interven- Clinical Oncology statement: achieving high-quality cancer survivorship tion resulted in significant improvements in dietary care. J Clin Oncol. 2013;31(5):631-640. 24. Vergnaud A-C, Romaguera D, Peeters PH, et al. Adherence to the intake and key biomarkers of health. The intervention World Cancer Research Fund/American Institute for Cancer Research was staged within the season of an urban garden. This guidelines and risk of death in Europe: results from the European Prospec- tive Investigation into Nutrition and Cancer cohort study. Am J Clin Nutr. innovative clinical intervention might be a model for 2013;97(5):1107-1120. health care professionals to use to address the gap in 25. Raynor HA, Champagne CM. Position of the Academy of Nutrition and Dietetics: interventions for the treatment of overweight and obesity in programs for patients transitioning from active cancer adults. J Acad Nutr Diet. 2016;116(1):129-147. treatment to becoming healthy cancer survivors. 26. Green AC, Hayman LL, Cooley ME. Multiple health behavior change in adults with or at risk for cancer: a systematic review. Am J Health Behav. 2015;39(3):380-394. References 27. Bandura A. Social Foundations of Thought and Action: A Social Cognitive Theory. Rockville, MD: Prentice-Hall; 1986. 1. American Cancer Society. Cancer Facts & Figures 2016. Atlanta, GA: 28. Miller WR, Rollnick S. Motivational Interviewing: Helping People American Cancer Society; 2016. http://www.cancer.org/acs/groups/content/@ Change. New York: Guilford Press; 2012. research/documents/document/acspc-047079.pdf. Accessed June 23, 2016. 29. Stacey FG, James EL, Chapman K, et al. A systematic review and 2. de Moor JS, Mariotto AB, Parry C, et al. Cancer survivors in the meta-analysis of social cognitive theory-based physical activity and/or United States: prevalence across the survivorship trajectory and implica- nutrition behavior change interventions for cancer survivors. J Cancer Surviv tions for care. Am Soc Prev Oncol. 2013;22(4):561-570. Res Pract. 2015;9(2):305-338. 3. Hewitt M, Greenfield S, Stovall E. From Cancer Patient to Cancer 30. Spees CK, Joseph A, Darragh A, et al. Health behaviors and per- Survivor: Lost in Transition. Washington, DC: National Academies Press; 2005. ceptions of cancer survivors harvesting at an urban garden. Am J Health 4. Campbell MK, Tessaro I, Gellin M, et al. Adult cancer survivorship Behav. 2015;39(2):257-266. care: experiences from the LIVESTRONG centers of excellence network. 31. Simmons LA, Wolever RQ. Integrative health coaching and motiva- J Cancer Surviv Res Pract. 2011;5(3):271-282. tional interviewing: synergistic approaches to behavior change in health- 5. Hawkins NA, Smith T, Zhao L, et al. Health-related behavior change care. Glob Adv Health Med. 2013;2(4):28-35. after cancer: results of the American Cancer Society’s studies of cancer 32. Harris PA, Taylor R, Thielke R, et al. Research electronic data cap- survivors (SCS). J Cancer Surviv Res Pract. 2010;4(1):20-32. ture (REDCap)—a metadata-driven methodology and workflow process 6. DeSantis CE, Lin CC, Mariotto AB, et al. Cancer treatment and sur- for providing translational research informatics support. J Biomed Inform. vivorship statistics, 2014. CA Cancer J Clin. 2014;64(4):252-271. 2009;42(2):377-381. 7. Inoue-Choi M, Robien K, Lazovich D. Adherence to the WCRF/ 33. Bowen DJ, Kreuter M, Spring B, et al. How we design feasibility

July 2016, Vol. 23, No. 3 Cancer Control 309 studies. Am J Prev Med. 2009;36(5):452-457. 34. Glasgow RE, Vogt TM, Boles SM. Evaluating the public health impact of health promotion interventions: the RE-AIM framework. Am J Public Health. 1999;89(9):1322-1327. 35. Glasgow RE, Klesges LM, Dzewaltowski DA, et al. Evaluating the impact of health promotion programs: using the RE-AIM framework to form summary measures for decision making involving complex issues. Health Educ Res. 2006;21(5):688-694. 36. National Cancer Institute. Dietary Screener Questionnaire in the NHANES 2009-10: background. Revised October 2015. http://appliedre- search.cancer.gov/studies/nhanes/dietscreen. Accessed June 23, 2016. 37. Deurenberg P, Weststrate JA, Seidell JC. Body mass index as a measure of body fatness: age- and sex-specific prediction formulas. Br J Nutr. 1991;65(02):105-114. 38. Aguilar SS, Wengreen HJ, Lefevre M, et al. Skin carotenoids: a biomarker of fruit and vegetable intake in children. J Acad Nutr Diet. 2014;114(8):1174-1180. 39. Ermakov IV, Sharifzadeh M, Ermakova M, et al. Resonance Raman detection of carotenoid antioxidants in living human tissue. J Biomed Opt. 2005;10(6):064028. 40. Mayne ST, Cartmel B, Scarmo S, et al. Noninvasive assessment of dermal carotenoids as a biomarker of fruit and vegetable intake. Am J Clin Nutr. 2010;92(4):794-800. 41. Scarmo S, Henebery K, Peracchio H, et al. Skin carotenoid status measured by resonance Raman spectroscopy as a biomarker of fruit and vegetable intake in preschool children. Eur J Clin Nutr. 2012;66(5):555-560. 42. Zidichouski J, Mastaloudis A, Poole S, et al. Clinical validation of a noninvasive, Raman Spectroscopic method to assess carotenoid nutritional status in humans. J Am Coll Nutr. 2009;28(6):687-693. 43. Mayne ST, Cartmel B, Scarmo S, et al. Resonance Raman spectroscopic evaluation of skin carotenoids as a biomarker of carotenoid status for human studies. Arch Biochem Biophys. 2013;539(2):163-170. 44. Jahns L, Johnson LK, Mayne ST, et al. Skin and plasma carotenoid response to a provided intervention diet high in vegetables and fruit: uptake and depletion kinetics. Am J Clin Nutr. 2014;100(3):930-937. 45. Courneya KS, Segal RJ, Mackey JR, et al. Effects of aerobic and resistance exercise in breast cancer patients receiving adjuvant chemotherapy: a multicenter randomized controlled trial. J Clin Oncol. 2007;25(28):4396-4404. 46. Djuric Z, DiLaura NM, Jenkins I, et al. Combining weight-loss counseling with the Weight Watchers plan for obese breast cancer survivors. Obes Res. 2002;10(7):657-665. 47. Campbell KL, Van Patten CL, Neil SE, et al. Feasibility of a lifestyle intervention on body weight and serum biomarkers in breast cancer survivors with overweight and obesity. J Acad Nutr Diet. 2012;112(4):559-567. 48. Wagland R, Fenlon D, Tarrant R, et al. Rebuilding self-confidence after cancer: a feasibility study of life-coaching. Support Care Cancer. 2015;23(3):651-659. 49. Heart T, Kalderon E. Older adults: Are they ready to adopt health- related ICT? Int J Med Inf. 2013;82(11):e209-e231. 50. Czaja SJ, Charness N, Fisk AD, et al. Factors predicting the use of technology: Findings from the center for research and education on aging and technology enhancement (create). Psychol Aging. 2006;21(2):333-352.

310 Cancer Control July 2016, Vol. 23, No. 3 Special Report

Hypofractionated Stereotactic Radiotherapy for Auditory Canal or Middle Ear Cancer Taro Murai, MD, Shin-Etsu Kamata, MD, Kengo Sato, MD, Kouki Miura, MD, Mitsuhiro Inoue, Naoki Yokota, MD, Seiji Ohta, MD, Michio Iwabuchi, MD, Hiromitsu Iwata, MD, and Yuta Shibamoto, MD

Background: Stereotactic radiotherapy (SRT) may represent a new treatment option for individuals with auditory canal or middle ear cancer. Methods: Study participants with pathologically proven ear cancer were treated with SRT (35 Gy for 3 fractions or 40 Gy for 5 fractions) as first-line therapy. When local tumor recurrence developed following SRT, subtotal temporal bone resection and postoperative chemoradiotherapy were performed as salvage treatment. Boluses were used for the initial 14 study patients. Results: Twenty-nine study participants were enrolled and staged with T1 (n = 3), T2 (n = 7), T3 (n = 14), or T4 disease (n = 5). Three-year overall survival rates were 69% for T1/2 disease, 79% for T3 disease, and 0% for T4 disease. Three-year local control rates were 70% for T1/2 disease, 50% for T3 disease, and 20% for T4 disease. Grade 2 or higher dermatitis or soft-tissue necrosis occurred more frequently in study patients treated with boluses (8/14 vs 2/15; P = .02). Salvage treatment was safely performed for 12 recurrent cases. Conclusions: These results suggest that SRT outcomes are promising for patients with ear cancer (≤ T3 disease). The rate of toxicity was acceptable in the study patients treated without boluses. Outcomes of salvage surgery and postoperative radiotherapy following SRT were also encouraging.

Introduction treatment of ear cancer. For example, reported out- Because of the rarity of external auditory canal and comes of radiotherapy alone or chemoradiotherapy middle ear cancers, large-scale prospective studies are not satisfactory in patients with advanced-stage have not been performed, and optimal treatment re- ear cancer.6 Local recurrence after radiotherapy may mains controversial.1,2 Nevertheless, surgery followed require surgical intervention. Due to the difficulties of by conventional radiotherapy has been recommend- en bloc resection and the increased risk of complica- ed.3-5 Subtotal temporal bone resection or total tem- tions, surgeons may be reluctant to approach recur- poral bone resection are often performed.3,4 However, rent ear cancer previously treated with radiotherapy.7,8 these procedures often cause severe complications. Furthermore, postoperative radiotherapy is difficult for Deafness and facial palsy are generally unavoidable. patients with a prior history of radiotherapy due to the A need exists for the development of less-invasive possible toxicities of reirradiation. treatments, although some hurdles remain for the Data regarding the outcomes of radiotherapy for ear cancer are primarily from series of patients who underwent conventional 2- or 3-dimensional radio- From the Department of Radiology (TM, YS), Nagoya City Uni- 5,6,9-11 versity Graduate School of Medical Sciences, Nagoya, Japan, therapy. However, developments in conformal Yokohama CyberKnife Center (TM, KS, MIn, NY, SO, MIw, HI), radiotherapy may provide a new treatment option for Yokohama, Japan, Head and Neck Tumor Center (S-EK), Chiba ear cancer. Stereotactic radiotherapy (SRT) has the do- Tokushukai Hospital, Funabashi, Japan, Department of Head and Neck Oncology and Surgery (S-EK, KM), International Uni- simetric advantages of improved target coverage with versity of Health and Welfare Mita Hospital, Tokyo, Japan, De- a steep dose gradient and spares surrounding normal partment of Neurosurgery (KS), Japanese Red Cross Medical Cen- tissues.12,13 Thus, salvage surgery and postoperative ra- ter, Tokyo, Japan, Radiation Oncology Center (NY), Suzukake Central Hospital, Hamamatsu, Shizuoka, Japan, Department of diotherapy can be performed following SRT for intra- Radiation Oncology (HI), Nagoya Proton Therapy Center, Na- cranial tumors.14,15 goya City West Medical Center, Nagoya, Japan. Address correspondence to Taro Murai, MD, Department of Ra- Materials and Methods diology, Nagoya City University Graduate School of Medical Sci- ences, 1 Kawasumi, Mizuho-cho, Mizuho-ku, Nagoya 467-8601, Study Design and Participants Japan. E-mail: [email protected] Initiated in 2008, this prospective study included pa- Submitted April 7, 2016; accepted May 16, 2016. tients with ear cancer who fulfilled the following in- No significant relationships exist between the authors and the clusion criteria: pathologically proven external au- companies/organizations whose products or services may be referenced in this article. All authors have no conflict of interest ditory canal or middle ear cancer and a Karnofsky with the vender mentioned in this article. Performance Status scale of at least 70. Exclusion crite-

July 2016, Vol. 23, No. 3 Cancer Control 311 ria were multiple malignant tumors, pregnancy, breast- side and 5-mm margins to the soft-tissue structure side. feeding, previous diagnosis of any psychological dis- Dose distribution was calculated with the ray-tracing orders, contraindications to iodine or gadolinium, and algorithm.17 Boluses with high CT attenuations previous radical surgery or irradiation to the skull or (250–900 HU) were used for the initial 14 study partici- brain. Once enrolled, the participants were classified pants to avoid dose reduction of the skin. Bolus usage according to the modified Pittsburgh grading system.2,3 was ceased for the latter 15 study participants. Those with T1 to T4 or N1-staged ear cancer were treat- The prescribed dose covered 95% of the PTV. De- ed with this protocol. Each study patient’s staging clas- pending on tumor volume and risk-organ doses, 37.5 Gy sification was determined prior to treatment with SRT in 3 fractions or 40.0 Gy in 5 fractions was delivered on based on findings from magnetic resonance imaging, consecutive days. The maximum doses to risk organs computed tomography (CT), and a physical examina- were limited to 21 to 25 Gy for 5 fractions and 15 to tion. Study participants were enrolled at a single insti- 18 Gy for 3 fractions.12,13 The 5-fraction regimen was ap- tution (International University of Health and Welfare plied to large tumors (size > 30 cc) or tumors close to the Mita Hospital [Tokyo, Japan]). One head and neck sur- internal carotid artery. When local tumor recurrence de- geon performed all of the physical examinations. Writ- veloped following SRT, subtotal resection of the tempo- ten informed consent was obtained from all patients. ral bone and postoperative 3-dimentional chemoradio- This study was approved by our Institutional Research therapy (60 Gy/30 fractions) were performed using 4 or and Ethics committee. 10 MV photons generated by a conventional linear accelerator.3,4 Intravenous cisplatin 80 mg/m2 was admin- Treatment Strategy and Details istered twice on a 3-week interval. Radical neck dissec- The treatment strategy is described in Fig 1. All study tion was performed for lymph-node metastases.3,4 participants underwent SRT as first-line treatment. SRT was carried out at a single institution (Yokohama End Points CyberKnife Center [Yokohama, Japan]). When local We evaluated rates of nonhematological toxicities, recurrence occurred, salvage surgery followed by local tumor control, and overall survival (OS) as end conventional 3-dimentional radiotherapy was per- points. Toxicities were recorded according to the formed at the International University of Health and National Cancer Institute Common Terminology Cri- Welfare Mita Hospital. SRT was performed using teria for Adverse Events, version 4. Grade 3 or higher CyberKnife (Accuray, Sunnyvale, California).13,16 This toxicity was considered unacceptable. All toxicities machine was chosen due to the high conformity in were evaluated by an otolaryngologist. The OS and lo- SRT for head and neck tumors. cal control rates were calculated by the Kaplan-Meier The study participants were placed in a supine method. The cumulative incidence of local control fail- position and a thermoplastic mask was molded to ure was calculated by accounting for death as a com- the head and attached to the head support. Axial peting risk. All local control failures were pathological- contrast-enhanced CT with a slice thickness of 2 mm ly confirmed. The log-rank test and the Gray test were was conducted for treatment planning. The clinical used in the univariate analyses. These analyses were target volume was defined as an abnormal contrast- implemented in the R package survival (Therneau, enhanced lesion seen on CT and magnetic resonance Rochester, Minnesota) or package cmprsk (Gray, Cam- imaging. To create the planning target volume (PTV), bridge, Massachusetts). All analyses were performed 2-mm margins were added to the bone structure in R version 3.0.0 (Hornik, Wien, Austria) for Windows (Microsoft, Redmond, Washington).

Results Progression free Between 2008 and 2011, a total of 29 patients with ear Observation cancer were enrolled. Details of the study participants and treatment prior to SRT are summarized in Table 1. SRT Three study participants had T1 tumors. One study pa- 37.5 Gy/3 fractions Operable recurrent tient had a Karnofsky Performance Status score of 70 40.0 Gy/5 fractions Salvage surgery + CRT and 22 study participants had scores of 100. Histology was squamous cell carcinoma in all study participants. Inoperable recurrent The median observation period was 40 months. Best supportive care Overall Survival The three-year OS rates were 69% for T1/2 (95% confi- Fig 1. — Protocol design. CRT = conventional radiotherapy, dence interval [CI], 31–89), 79% for T3 (95% CI, 47–93), SRT = stereotactic radiotherapy. and 0% for T4 disease (95 CI, 0–52; P = 0.0003; Fig 2).

312 Cancer Control July 2016, Vol. 23, No. 3 In 3 study participants with T1 tumors, recurrence was served. All lymph-node metastases developed in the not present at 3 years, and their hearing had been pre- ipsilateral internal jugular nodes, which are usually served. Operability was also a significant prognostic not irradiated in conventional radiotherapy for ear factor for median OS (44.0 vs 23.5 months; P = .01). No cancer.9 In univariate analyses, no statistically signifi- statistically significant differences were observed with cant differences were observed with respect to PTV respect to age (< 65 vs ≥ 65 years), PTV (< 25 vs ≥ 25 cc), (< 25 vs ≥ 25 cc), fractionation schedule, operability, bolus usage, fractionation schedule, or sex. or sex. Similar local control rates were obtained after ceasing bolus usage (P = .99). Local Control Three-year local control rates were 70% (95% CI, Toxicity 41–94) for T1/2, 50% for T3 (95 CI%, 27–79), and 20% Toxicities after SRT are summarized in Table 2. Grade for T4 disease (95 CI%, 2–95; Fig 3). Patterns of re- 3 vertigo was observed as an early toxicity. These currence were local failure in 9 study participants and symptoms ceased a few weeks following SRT. For late mixture of local failure and lymph-node metastases in toxicities, grade 2 vertigo and grade 2 facial nerve 5 study participants. No marginal recurrence was ob- disorder were observed (1 patient each). Grade 2 or higher dermatitis or soft-tissue necrosis occurred in 8 of the initial 14 study participants treated with bo- Table 1. — Characteristics of Study Participants, Tumors, luses; among them, 1 patient had grade 3 dermatitis, and Therapy which developed around the bolus. Therefore, use of Variable Stage boluses was stopped. After boluses were stopped, no T1/2 (n = 10) T3 (n = 14) T4 (n = 5) grade 2 or higher skin toxicity was observed. Asymp- tomatic auditory canal cartilage exposure was evalu- Median age, y 65 (37–82) 61 (40–89) 62 (55–70) ated as grade 2 soft-tissue necrosis because intraear Ratio of age 5:5 11:3 3:2 washing was regularly performed. Necrosis of at least (< 65:≥ 65) grade 3 occurred in 1 patient. Dermatitis or soft-tissue Male: female 5:5 10:4 1 : 4 necrosis of at least grade 2 occurred more frequent- Planning target 15.0 ± 9.7 42.1 ± 28.4 53.2 ± 19.6 ly in study participants treated with boluses (8/14 vs volume,a cc 2/15; P = .02). No other statistically significant risk fac- Ratio of 10:0 11:3 0:5 operable:inoperable tors were identified. Ratio of bolus usage 5:5 7:7 2:3 (+:–) Salvage Surgery and Postoperative Radiotherapy Ratio of No. of As salvage treatment, subtotal temporal bone resection 9:1 10:4 1:4 fractions (3:5) and reconstruction were carried out for 3 study par- Ratio of lymph-node ticipants with T1/2 disease, 7 study participants with 10:0 13:1 4:1 metastasis (–:+) T3 disease, and 2 study participants with T4 disease aMean ± standard deviation. T = tumor. (Table 3). Among them, 3 study participants (T3 = 1,

100 100

80 80

60 60

40 T1/2 40 T1/2

Rate of OS, % T3 T3 20 T4 20 T4 Rate of Local Control, % P = .0003 P = .053 0 0 0 12 24 36 48 60 0 12 24 36 48 60 No. at No. at Risk Months Risk Months

T1/2 10 9 6 6 6 1 T1/2 10 7 7 7 6 3 T3 14 12 12 11 7 3 T3 14 7 7 6 4 2 T4 5 4 1 0 0 0 T4 5 0 0 0 0 0

Fig 2. — OS curves after stereotactic radiotherapy. OS = overall survival. Fig 3. — Local control curves after stereotactic radiotherapy.

July 2016, Vol. 23, No. 3 Cancer Control 313 27 days. Postoperative chemoradiotherapy was deliv- Table 2. — Toxicity Rates of Stereotactic Radiotherapy ered to 7 study participants (T3 = 5, T4 = 2; see Table 3). Toxicity Stage The median follow-up period after salvage irradiation T1/2 (n = 10) T3 (n = 14) T4 (n = 5) was 40 months (range, 4–60 months). Grade 2 contra- Early lateral side otitis medina occurred 5 years after reirradiation. No grade 2 or higher brain necrosis or soft-tis- Ear pain sue necrosis was observed. Grade 0 8 12 3 Grade 1 1 Discussion Grade 2 2 1 2 In our study, tumor stage was a prognostic factor of 3,6 Radiation dermatitis OS. The reported 5-year survival rates for ear cancer vary, ranging from 21% to 86%.4,6 Prognosis of early- Grade 0 9 10 5 stage ear cancer is generally satisfactory; thus, organ Grade 1 1 4 preservation should be considered as the treatment for External ear early stage ear cancer. By contrast, further improve- inflammation ment in rates of OS and local control also remains an Grade 0 10 14 3 issue to be resolved for advanced-stage ear cancer in Grade 1 2 addition to avoiding surgery. The present study was Late designed to achieve these outcomes using SRT. Ear pain T1/2-staged disease is confined to the auditory canal and receives full benefit from definitive thera- Grade 0 8 13 5 py. The 5-year OS rates after temporal bone resection Grade 1 have ranged from 28% to 100%.4 Zhang et al18 per- Grade 2 2 1 formed definitive surgery to conserve the facial nerve Radiation-related for early-stage ear cancer; however, positive surgical dermatitis margins were observed in 54% of patients, and the Grade 0 10 10 4 recurrence rate was 46%. In patients with early-stage Grade 1 ear cancer, it may be difficult to conserve the facial Grade 2 3 nerve at surgery. By contrast, conventional radiotherapy achieved sufficient outcomes. Hashi et al10 treated Grade 3 1 1 8 patients with T1 disease with radiotherapy, and the Soft-tissue necrosis disease control rate was 100%. Pemberton et al11 also Grade 0 7 12 5 treated 27 patients with early-stage ear cancer; the Grade 1 5-year cancer-specific survival rate was 85%. SRT also Grade 2 3 1 0 achieved comparable rates of local control and OS in Grade 3 1 the present study. Although a comparison of the treatment strategies is difficult, owing to the heterogeneity of staging classi- T4 = 2) had been evaluated as inoperable prior to SRT. fication and the types of treatment used, radiotherapy, Among study participants with recurrent T3 disease, including SRT, may be preferable to surgery for early- 5 study participants survived more than 3 years after stage ear cancer. salvage treatment. Of the 4 study participants with The prognosis of T3 disease was nearly equivalent T4 tumors, intrathecal dissemination was observed in to that of early-stage ear cancer in our study. Testa et al19 2 at the recurrence sites. En bloc resection was per- reported that the 5-year survival rate for patients who formed in 9 of the 12 study participants (see Table 3). At surgery, dissemination to the dura was observed in 2 study participants (T3 = 1, T4 = 1). Neck dissection Table 3. — Rates of Local Recurrence and Salvage Treatment was also performed in 5 study participants who devel- Stage oped lymph-node metastases (T1/2 = 1, T3 = 4). T1/2 (n = 10) T3 (n = 14) T4 (n = 5) As complications of salvage surgery, grade 1 cere- Local recurrence 3 7 4 brospinal fluid leak was observed in 1 patient. Grade 3 abscess, grade 2 vertigo, and grade 3 necrosis of the Salvage surgery 3 7 2 skin flap also occurred (n = 1 for all). Because of these En bloc resection 3 5 1 events, 2 study participants required hospital admis- Postoperative 5 2 sion for 71 and 110 days, respectively. The median hos- radiotherapy pitalization period of the 12 study participants was

314 Cancer Control July 2016, Vol. 23, No. 3 underwent radiotherapy was 29%, but that rate was toxicities in our study compared with the results from 63% for patients treated with a combination of surgery previous studies.6,9-11,23 Grade 3/4 necrosis or derma- and radiotherapy. Austin et al20 also reported that com- titis was previously reported to develop in 6% to 15% bination therapy involving surgery and radiotherapy of patients.6,10,11,23 However, 3 out of 4 of these studies provided a higher 5-year survival rate than either sur- were retrospective, and the toxicities were evaluated gery or radiotherapy alone. Thus, combination therapy by radiation oncologists.6,10,11,23 By contrast, ours was of has been considered optimal for the treatment of T3 tuan exploratory, prospective study design, and toxicities mors.3,4 The 3-year OS rates after surgery in combina- were evaluated by an otolaryngologist. Without bolus tion with postoperative conventional radiotherapy for usage, grade 3 necrosis occurred in 1 of our 15 study T3 disease has ranged from 21% to 86%.4 participants, suggesting that SRT had acceptable rates In our study, the OS rate was satisfactory, and 50% of toxicity and did not require bolus treatment. of our study participants avoided surgery. Although Conventional radiotherapy may complicate the 50% of study participants experienced local recur- care of patients who require salvage surgery. Surgical rence, these patients could undergo radical salvage complication rates in patients with a prior history of treatment, including subtotal temporal bone resection radiotherapy are generally higher than those with un- and postoperative radiotherapy. SRT reduces tumor treated disease.7,8 In our study, surgical complications bulk and allows salvage surgery, even if pathological of grade 2 or higher were observed in 2 study patients, complete response cannot be obtained; hence, our and the median admission period was acceptable, sug- treatment protocol may improve patient prognosis and gesting that salvage surgery had acceptable rates of quality of life among those with T3 disease. toxicity. In addition, after careful follow-up, no grade By contrast, the prognosis for T4 tumors remained 3 or higher radiation-associated late toxicities (eg, soft poor in the present study. Many investigators have also tissue necrosis, brain necrosis) occurred. With use of reported similar, unfavorable outcomes. Xie et al21 re- SRT, the irradiated volume was limited, and the irradi- ported a 2-year OS rate of 22% in 39 patients with ated tissues were excised and reconstructed with nor- T4 tumors undergoing temporal bone resection and mal tissues. In our protocol, doses to the cranial nerves postoperative radiotherapy. Leong et al22 performed and brain stem were restricted within the tolerance subtotal resection and postoperative chemoradio- level. These results indicated that SRT for ear cancer therapy in 23 patients with T4N0M0 disease. Among provides a sufficiently steep dose gradient to deliver re- them, 14 died of ear cancer within 22 months of sur- irradiation after surgery. gery.22 Koto et al23 reported a 3-year OS rate of 44% with carbon-ion radiotherapy in 13 patients with Future Directions T4 tumors. These outcomes are attributed to the tumor Results of the current protocol appear promising, but location and anatomical features. T4 tumors erode the further improvement in local control is desirable, es- deep parts of the auditory canal, including the cochlea, pecially because the local control rate for T3 tumors petrous apex, carotid canal, dura, and facial nerve. The should be increased from 50% of the present study. foramina, fissures, and channels for the nerves and A large tumor may contain many hypoxic cells re- vessels are thought to provide the routes for tumor dis- sistant to radiation.1,25 Thus, the phenomenon of re- semination. oxygenation and redistribution should be utilized In our study, dissemination to the dura was ob- in radiotherapy against these large tumors.25 Frac- served in 3 of the 4 patients with T4 tumors. T4 dis- tionation has radiobiological advantages because ease may spread beyond the definitively irradiated vol- reoxygenation of hypoxic tumor cells and the redis- ume of SRT. Therefore, patients with T4 tumors are not tribution of the cell cycle to a more sensitive phase good candidates for SRT, which is in contrast to those can be expected.12 Therefore, further increase in the with T1/2 or T3 disease. fraction number may improve rates of local control. Grade 2 or higher dermatitis occurred more fre- In addition, use of radiosensitizers may enhance ra- quently in study participants treated with boluses in diation effects and improve prognosis of individuals our study. Dermatitis was observed in areas around with advanced cancer.26 Conventional cytotoxic che- the bolus, suggesting that its usage is associated with motherapy agents like cisplatin and fluorouracil are these toxicities. The boluses had high CT values, and considered to have a radiosensitizing effect.26 An oral the PTV was in contact with air-density areas. In such formation of tegafur/gimeracil/oteracil has also been cases, the dose around the bolus was overestimated shown to enhance the radiation effect.27,28 Conven- or underestimated in the treatment planning system tional hypoxic cell sensitizers may also be effective due to electron scattering.17 Although dosimetry of ac- when combined with high-dose radiotherapy.29 Thus, tual skin doses is difficult, ceasing bolus usage may the increase the fraction number and use of these decrease actual doses by 5% to 10%.24 With respect to radiosensitizers should be considered for inclusion in soft-tissue necrosis, we may have overdiagnosed the a new protocol to improve local efficacy rates.

July 2016, Vol. 23, No. 3 Cancer Control 315 Conclusions of the depth dose in the build-up region, and surface dose for a 6-MV therapeutic photon beam: Monte Carlo simulation and measurements. To the best of our knowledge, this is the first study that J Radiat Res. 2013;54(2):374-382. demonstrates the efficacy of stereotactic radiotherapy 25. Murata R, Shibamoto Y, Sasai K, et al. Reoxygenation after single irradiation in rodent tumors of different types and sizes. Int J Radiat Oncol for ear cancer. The results of this protocol for individu- Biol Phys. 1996;34(4):859-865. als with T3 disease or higher are promising. In study 26. Budach W, Hehr T, Budach V, et al. A meta-analysis of hyperfractionated and accelerated radiotherapy and combined chemotherapy and participants treated without a bolus, toxicities were radiotherapy regimens in unresected locally advanced squamous cell car- acceptable and tolerable. Salvage surgery and postop- cinoma of the head and neck. BMC Cancer. 2006;6:28. 27. Fukushima M, Sakamoto K, Sakata M, et al. Gimeracil, a compo- erative radiotherapy after stereotactic radiotherapy are nent of S-1, may enhance the antitumor activity of X-ray irradiation in hu- also feasible, and these results are encouraging. man cancer xenograft models in vivo. Oncol Rep. 2010;24(5):1307-1313. 28. Tahara M, Kiyota N, Mizusawa J, et al. Phase II trial of chemoradiotherapy with S-1 plus cisplatin for unresectable locally advanced head and References neck cancer (JCOG0706). Cancer Sci. 2015;106(6):726-733. 29. Shibamoto Y, Nishimoto S, Mi F, et al. Evaluation of various types of 1. Kuhel WI, Hume CR, Selesnick SH. Cancer of the external auditory new hypoxic cell sensitizers using the EMT6 single cell - spheroid - solid tumour canal and temporal bone. Otolaryngol Clin North Am. 1996;29(5):827-852. system. Int J Radiat Biol Relat Stud Phys Chem Med. 1987;52(3):347-357. 2. Moody SA, Hirsch BE, Myers EN. Squamous cell carcinoma of the external auditory canal: an evaluation of a staging system. Am J Otol. 2000;21(4):582-588. 3. Bacciu A, Clemente IA, Piccirillo E, et al. Guidelines for treating temporal bone carcinoma based on long-term outcomes. Otol Neurotol. 2013;34(5):898-907. 4. Prasad SC, D’Orazio F, Medina M, et al. State of the art in temporal bone malignancies. Curr Opin Otolaryngol Head Neck Surg. 2014;22(2):154-165. 5. Yin M, Ishikawa K, Honda K, et al. Analysis of 95 cases of squamous cell carcinoma of the external and middle ear. Auris Nasus Larynx. 2006;33(3):251-257. 6. Ogawa K, Nakamura K, Hatano K, et al. Treatment and prognosis of squamous cell carcinoma of the external auditory canal and middle ear: a multi-institutional retrospective review of 87 patients. Int J Radiat Oncol Biol Phys. 2007;68(5):1326-1334. 7. Goodwin WJ Jr. Salvage surgery for patients with recurrent squamous cell carcinoma of the upper aerodigestive tract: when do the ends justify the means? Laryngoscope. 2000;110(3 pt 2 suppl 93):1-18. 8. Agra IM, Carvalho AL, Pontes E, et al. Postoperative complications after en bloc salvage surgery for head and neck cancer. Arch Otolaryngol Head Neck Surg. 2003;129(12):1317-1321. 9. Pfreundner L, Schwager K, Willner J, et al. Carcinoma of the external auditory canal and middle ear. Int J Radiat Oncol Biol Phys. 1999;44(4):777-788. 10. Hashi N, Shirato H, Omatsu T, et al. The role of radiotherapy in treating squamous cell carcinoma of the external auditory canal, especially in early stages of disease. Radiother Oncol. 2000;56(2):221-225. 11. Pemberton LS, Swindell R, Sykes AJ. Primary radical radiotherapy for squamous cell carcinoma of the middle ear and external auditory canal - an historical series. Clin Oncol (R Coll Radiol). 2006;18(5):390-394. 12. Murai T, Ogino H, Manabe Y, et al. Fractionated stereotactic radiotherapy using CyberKnife for the treatment of large brain metastases: a dose escalation study. Clin Oncol (R Coll Radiol). 2014;26(3):151-158. 13. Iwata H, Sato K, Tatewaki K, et al. Hypofractionated stereotactic radiotherapy with CyberKnife for nonfunctioning pituitary adenoma: high local control with low toxicity. Neuro Oncol. 2011;13(8):916-922. 14. Tsao M, Xu W, Sahgal A. A meta-analysis evaluating stereotactic radiosurgery, whole-brain radiotherapy, or both for patients presenting with a limited number of brain metastases. Cancer. 2012;118(9):2486-2493. 15. Shibamoto Y, Sugie C, Iwata H. Radiotherapy for metastatic brain tumors. Int J Clin Oncol. 2009;14(4):281-288. 16. Iwata H, Tatewaki K, Inoue M, et al. Salvage stereotactic reirradiation using the CyberKnife for the local recurrence of nasal or paranasal carcinoma. Radiother Oncol. 2012;104(3):355-360. 17. Sharma SC, Ott JT, Williams JB, et al. Clinical implications of adopt- ing Monte Carlo treatment planning for CyberKnife. J Appl Clin Med Phys. 2010;11(1):3142. 18. Zhang T, Li W, Dai C, et al. Evidence-based surgical management of T1 or T2 temporal bone malignancies. Laryngoscope. 2013;123(1):244-248. 19. Testa JR, Fukuda Y, Kowalski LP. Prognostic factors in carcinoma of the external auditory canal. Arch Otolaryngol Head Neck Surg. 1997;123(7):720-724. 20. Austin JR, Stewart KL, Fawzi N. Squamous cell carcinoma of the external auditory canal. Therapeutic prognosis based on a proposed staging system. Arch Otolaryngol Head Neck Surg. 1994;120(11):1228-1232. 21. Xie B, Zhang T, Dai C. Survival outcomes of patients with temporal bone squamous cell carcinoma with different invasion patterns. Head Neck. 2015;37(2):188-196. 22. Leong SC, Youssef A, Lesser TH. Squamous cell carcinoma of the temporal bone: outcomes of radical surgery and postoperative radiotherapy. Laryngoscope. 2013;123(10):2442-2448. 23. Koto M, Hasegawa A, Takagi R, et al. Carbon ion radiotherapy for locally advanced squamous cell carcinoma of the external auditory canal and middle ear. Head Neck. 2016;38(4):512-516. 24. Apipunyasopon L, Srisatit S, Phaisangittisakul N. An investigation

316 Cancer Control July 2016, Vol. 23, No. 3 Ten Best Readings Relating to Head and Neck Cancers

McDonald MW, Linton OR, Shah MV. Proton O’Sullivan B, Huang SH, Su J, et al. Development therapy for reirradiation of progressive or re- and validation of a staging system for HPV-recurrent chordoma. Int J Radiat Oncol Biol Phys. lated oropharyngeal cancer by the International 2013;87(5):1107-1114. Collaboration on Oropharyngeal Cancer Network This retrospective review of data from 16 patients for Staging (ICON-S): a multicentre cohort study. was undertaken to report on reirradiation with proton Lancet Oncol. 2016;17(4):440-451. therapy for progressive/recurrent chordoma in com- The International Collaboration on Oropharyn- bination with or without salvage surgery. The 2-year geal Cancer Network for Staging (ICON-S) aimed estimated rate of local control was 85%, the rate of de- to develop a tumor, node, metastasis (TNM) clas- veloping distant metastases was 20%, overall survival sification for human papillomavirus (HPV)-positive was 80%, and the rate of chordoma-specific survival oropharyngeal cancer. Patients with oropharyngeal was 88%. Toxicities were observed at longer follow-up cancer whose cancer had not spread were included. times. Full-dose proton reirradiation was associated Stage classifications were ranked according to sur- with initial disease control and good overall survival vival performance. Because the rate of 5-year overall for patients with progressive/recurrent chordoma. survival was similar for T4a and T4b disease, T4 classification was not subdivided in the retermed Forastiere AA, Zhang Q, Weber RS, et al. Long-term ICON-S T categories. The rates of 5-year overall sur- results of RTOG 91-11: a comparison of three non- vival were similar among N1, N2a, and N2b disease, surgical treatment strategies to preserve the larynx so the authors retermed the 7th edition N categories. in patients with locally advanced larynx cancer. They then proposed the following ICON-S classifi- J Clin Oncol. 2013;31(7):845-852. cation: stage I (T1–T2N0–N1), stage II (T1–T2N2 or These researchers sought to determine the long- T3N0–N2), and stage III (T4 or N3). They also classi- term results of treating 520 patients with stage III/IV fied metastatic disease (M1) as ICON-S stage IV. The glottic/supraglottic squamous cell cancer with either authors concluded that their ICON-S staging system concomitant cisplatin/radiotherapy (RT), induction for HPV-positive oropharyngeal cancer is now suit- cisplatin/fluorouracil followed by radiotherapy (RT), able for the 8th edition of the Union for International or RT alone. After a follow-up period of 10 years, the Cancer Control/American Joint Committee on Cancer data revealed that induction cisplatin/fluorouracil fol- TNM classification. lowed by RT and concomitant cisplatin/RT had similar effectiveness for laryngectomy-free survival. Rates of Keck MK, Zuo Z, Khattri A, et al. Integrative analy- both larynx preservation and locoregional control were sis of head and neck cancer identifies two biologi- significantly improved with concomitant cisplatin/RT cally distinct HPV and three non-HPV subtypes. compared with induction chemotherapy or RT alone. Clin Cancer Res. 2015;21(4):870-881. Gene expression–based consensus clustering, hu- Mifsud MJ, Tanvetyanon T, Mccaffrey JC, et al. Adju- man papillomavirus (HPV) status, and copy number vant radiotherapy versus concurrent chemoradiother- profiling were obtained for a clinically homogenous apy for the management of high-risk salivary gland patient cohort of 134 locoregionally advanced head carcinomas. Head Neck. 2016. Epub ahead of print. and neck squamous cell carcinomas to identify sub- The purpose of this retrospective study of 140 pa- types and translationally relevant characteristics. Five tients with salivary carcinomas was to explore the subtypes of head and neck squamous cell carcinomas use of adjuvant chemoradiotherapy for the treatment were identified, 2 of which were biologically distinct of advanced salivary malignancies because of their HPV subtypes. One HPV negative and 1 HPV posi- aggressive disease course. Patients were treated with tive subtype had a prominent immune and mesen- either postoperative chemoradiotherapy or radiother- chymal phenotype. Compared with other subtypes, apy (RT). High rates of locoregional control were ob- these 2 subtypes had low expression and no copy served in both groups: 91% for RT and 79% for chemo- number events for epidermal growth factor recep- radiotherapy. A difference in 3-year progression-free tor (EGFR)/human epidermal growth factor receptor survival was not seen with chemoradiotherapy vs RT. (HER) ligands. The basal subtype was characterized Until prospective evidence is available, chemoradio- by a prominent EGFR/HER signaling phenotype, therapy as standard treatment for advanced salivary HPV-negative status, and with strong hypoxic differen- malignancies cannot be advised. tiation. The authors concluded that their classification

July 2016, Vol. 23, No. 3 Cancer Control 317 subtypes carry translational implications for develop- Spencer S, Schulten J, Hossain A, et al. Cetuximab ing biomarkers and personalizing patient care. and radiotherapy in laryngeal preservation for cancers of the larynx and hypopharynx: a second- Graham DM, Isaranuwatchai W, Habbous S, et al. ary analysis of a randomized clinical trial. JAMA A cost-effectiveness analysis of human papilloma- Otolaryngol Head Neck Surg. 2016. Epub ahead virus vaccination of boys for the prevention of oro- of print. pharyngeal cancer. Cancer. 2015;121(11):1785-1792. An international, randomized, phase 3 study was A preliminary analysis and literature review were undertaken to assess the rates of laryngeal preserva- performed to assess the cost effectiveness of human tion and laryngectomy-free survival among 168 pa- papillomavirus (HPV) vaccination in Canadian boys tients with hypopharyngeal or laryngeal cancer after with oropharyngeal cancer. The researchers deter- being treated with cetuximab and radiotherapy (RT) mined that the quadrivalent HPV vaccine (HPV4) pro- or RT alone. The rates of laryngeal preservation at duced 0.05 more quality-adjusted life years (QALYs) 2 years were 87.9% and 85.7% for those receiving and saved $145 Canadian dollars (CAD) per indi- cetuximab/RT and RT alone, respectively. The hazard vidual compared with no vaccine. When the authors ratio for laryngectomy-free survival was 0.78 when adjusted for assumed rates of 50% vaccine efficacy comparing cetuximab/RT and RT alone. These results and 50% uptake, HPV4 resulted in 0.02 more QALYs indicate that a cetuximab-based regimen may provide and had a cost savings $42 CAD. Thus, the researchers some benefit for patients with cancer of the larynx/ concluded that use of HPV4 for boys living in Canada hypopharynx. aged 12 years might be a cost-effective method for preventing oropharyngeal cancer. Rosenthal DI, Harari PM, Giralt J, et al. Association of human papillomavirus and p16 status with out- Grossi Marconi D, da Costa Resende B, Rauber E, comes in the IMCL-9815 phase III registration trial et al. Head and neck non-melanoma skin cancer for patients with locoregionally advanced oropha- treated by superficial x-ray therapy: an analysis ryngeal squamous cell carcinoma of the head and of 1021 cases. PLoS One. 2016;11(7):e0156544. neck treated with radiotherapy with or without These authors retrospectively reviewed the re- cetuximab. J Clin Oncol. 2016;34(12):1300-1308. cords of 597 patients (n = 1,021 lesions) treated with A retrospective evaluation of the association of superficial X-ray therapy (SXRT) for basal cell carci- human papillomavirus (HPV) and p16 protein expres- noma, squamous cell carcinoma (SCC), and SCC in sion status was undertaken by researchers to study situ to compare treatment outcomes by fractionation outcomes in patients with oropharyngeal carcinoma schedules. No significant difference was observed receiving radiotherapy (RT) plus weekly cetuximab or among the 3 prescribed fractionation schemes in RT alone. When treated with RT alone or RT plus ce- terms of toxicity. Use of daily doses higher than tuximab, patients with p16-positive status had a higher 2 Gy was significantly associated with improved rates rate of overall survival than those whose p16 status of local control. The researchers concluded that SXRT was negative. Adding cetuximab to RT increased rates is well tolerated in this patient population, achieves ex- of progression-free survival, locoregional control, cellent rates of local control, and they recommend that and overall survival in those with p16-positive and it should be recommended for disease management. p16-negative oropharyngeal carcinoma. The authors observed that p16 status was also strongly prognostic Tanvetyanon T, Fisher K, Caudell J, et al. Adjuvant for patients with oropharyngeal carcinoma. chemoradiotherapy versus with radiotherapy alone for locally advanced salivary gland carcinoma among older patients. Head Neck. 2016;38(6):863-870. Researchers studied adjuvant chemoradiotherapy with adjuvant radiotherapy (RT) in an older patient population using Surveillance, Epidemiology, and End Results/Medicare data. A total of 13.5% received adjuvant chemoradiotherapy and 86.5% received adjuvant RT. The rates of median overall survival were 41 and 24 months for adjuvant RT and chemoradiotherapy, respectively. The authors observed that adjuvant chemoradiotherapy was associated with increased mortality and that rates of toxicity were significantly higher in those treated with chemoradiotherapy.

318 Cancer Control July 2016, Vol. 23, No. 3