Annalsplasticsurgery.Com VOLUME 78 | SUPPLEMENT 5 | JUNE 2017 of of Lastic Surgery Lastic Lastic Surgery Lastic P P Annals Annals

www.annalsplasticsurgery.com VOLUME 78 | SUPPLEMENT 5 | JUNE 2017 of of lastic Surgery lastic lastic Surgery lastic P P Annals Annals

June 2017 Annals of Plastic Surgery Volume 78 Supplement 5 (Pages S257–S350) Annals of Plastic Surgery

Volume 78 / Supplement 5 / June 2017 Editor in Chief Emeritus Editors William C. Lineaweaver, MD, Richard Stark, MD Lars Vistnes, MD William Morain, MD FACS (1978Y1981) (1982Y1992) (1992Y2007) Managing Editor Publishing Staff Associate Editor Jane Yongue Wood, MA Alexandra Manieri, Michelle Smith, Richard Goodwin, MD, PhD Deputy Editor Publisher Advertising Sales Bruce Mast, MD Representative Editorial Board

Aesthetic Surgery Fazhi Qi, MD, PhD Ming-Huei Chen, MD James Zins, MD, Associate Editor Maurice Nahabedian, MD Matthew Choi, MD Ahmed Afifi, MD Martin Newman, MD Roberto Flores, MD Mohammed Alghoul, MD Vu Nguyen, MD Johnny U. Franco, MD Lorelei Grunwaldt, MD Ozan Bitik, MD Burn Surgery and Research Christopher Chang, MD Janice Lalikos, MD Jorge de la Torre, MD Markus Kuentscher, MD, Prof, Joseph E. Losee, MD Michael Dobryansky, MD Associate Editor W. Scott McDonald, MD Antonio Jorge Forte, MD ScottC.Hultman,MD,AssociateEditor Parit Patel, MD Ahmed Hashem, MD Sigrid Blome-Eberwein, MD Aamir Sadik, MD Raymond Isakov, MD Bernd Hartmann, MD Dhruv Singhal, MD C.J. Langevin, DMD, MD Christoph Hirche, MD Jeffrey F. Topf, DDS George Varkarakis, MD Harry K. Moon, MD Marc Jeschke, MD Colin Morrison, MD Lars P. Kamolz, MD, Prof Richard Zeri, MD Farzad R. Nahai, MD Marcus Lehnhardt, MD Microsurgery Abel de la Pen˜ a, MD David Benjamin Lumenta, MD Gordon Lee, MD, Associate Editor Henry C. Vasconez, MD Andrea Pozez, MD Matthew Hanasono, MD, Joshua Waltzman, MD Paul Wurzer, MD Associate Editor Michael J. Yaremchuk, MD Christian J. Vercler, MD, MA, FACS Richard J. Zienowicz, MD Anuja K. Antony, MD Hand Surgery Joseph Carey, MD Breast Surgery Robert Weber, MD, Associate Editor Edward Chang, MD Steven Bernard, MD, Associate Editor Aaron Burger, MD, PhD Anthony Echo, MD Mustafa Akyurek, MD Christi Cavaliere, MD Yoon Chun, MD Victor Hassid, MD Umar Choudry, MD James Chang, MD Mihai Ionac, MD, PhD Amy Colwell, MD Alphonsus Chong, MD Nathan Menon, MD Stirling Craig, MD Loretta Coady-Fariborzian, MD Arash Momeni, MD Lynn Damitz, MD E. Gene Deune, MD Peter Neligan, MD Kent Hidgon, MD Doug Fornfeist, MD Alexandru Nistor, MD Eric Halvorson, MD Jennifer Giuffre, MD Dung H. Nguyen, MD, PharmD John Hijjawi, MD Scott Hansen, MD Hassan Noaman, MD Mark Leyngold, MD Ginard Henry, MD Leo Ryo Otake, MD, PhD Steve Henry, MD Daniel Luppens, MD Edmond Ritter, MD John Ingraham, MD Anne Peled, MD Roman Skoracki, MD Galen Perdikis, MD Kai Megerle, MD Ewa Timek, MD John Kim, MD Wendy Parker, MD Neil Salyapongse, MD Peripheral Nerve Surgery and Wei Z. Wang, MD Research Norman Weinzweig, MD Brian Rinker, MD, Associate Editor Listed in Index Medicus/MEDLINE, Head and Neck Surgery Gregory H. Borschel, MD EMBASE/Excerpta Medica, William A. Wooden, MD, Associate Paul Cederna, MD Biological Abstracts, Biosis, Editor Michael Chen, MD Current Contents/Clinical Medicine, Michael T. Friel, MD, Associate Editor Howard Clarke, MD Current Awarness in Biological David M. Adelman, MD, PhD A. Lee Dellon, MD, PhD Sciences, and International Center Ovunc Akdemir, MD Teoman Dogan, MD for Scientific Research Annals of Plastic Surgery

Taolin Fang, MD, PhD Peter Arnold, MD, PhD Jaimie Shores, MD Tessa Gordon, PhD Spencer Brown, PhD Thomas Tung, MD Tessa A. Hadlock, MD Wayne E. Carver, MD Fu-Chan Wei, MD Martijn Malessy, MD Yimin Chai, MD, PhD Shaun D. Mendenhall, MD Elliott Chen, MD Circumspectus Medicinae Wyndell Merritt, MD Koray Coskunfirat, MD Nash Naam, MD Kayvan Shokrollahi, LLM, FRCS(Plast), Kim Creek, PhD Associate Editor Gedge D. Rosson, MD David Drake, MD Iain S. Whitaker, FRCS, PhD, Shai M. Rozen, MD Holger Engel, MD Bauback Safa, MD Stephan A. Fann, MD Associate Editor Robert J. Spinner, MD Robert Gourdie, PhD Warren Rozen, MBBS, PhD Wesley P. Thayer, MD, PhD Geoffrey Hallock, MD Christopher G. Wallace, MBChB, MS Allen L. Van Beek, MD Deepak V. Kilpadi, PhD, MBA Huan Wang, MD, PhD Howard Levinson, MD Editors at Large Wendong Xu, MD Qingfeng Li, MD, PhD Francesco Boccardo, MD Hede Yan, MD Kacey Marra, PhD Reconstructive Surgery Irene Miko, MD Burt Brent, MD Corradino Campisi, MD Raffi Gurunluoglu, MD, PhD, Norbert Nemeth, MD, PhD Irfan Ozyazgan, MD Scott Corlew, MD Associate Editor Ivica Ducic Michael Matthew, MD, Jay Potts, PhD Y.J. Seol, PhD Jobe Fix, MD Associate Editor Leonard Furlow, MD Ramin Shayan, MBBS, PhD Patricio Andrades, MD Paul M. Glat, MD Merribeth Bruntz, MD Richard Simman, MD John Clayton, MD Sunil Tholpady, MD, PhD Yvette Godwin, FRCS(Plast) William Dzwierzynski, MD Krishna S. Vyas, MD, MHS Albert Losken, MD Brian Gastman, MD Michael Yost, PhD Cuneyt Ozek, MD George Huemer, MD, PhD Jian Zhang, MD, PhD Somprasong Songchaeroen, MD James Long, MD Li-wu Zheng, DDS, MD, PhD Peter Taub, MD Paul Montero, MD Takumi Yamamoto, MD Nurdan Ozturk, MD Transplantation Surgery and Chul Park, MD Research Corresponding Editors Lee Pu, MD, PhD Peter Ray, MD Chad Gordon, DO, Associate Editor Mutaz Habal, MD, Journal of Laurence Z. Rosenberg, MD Angelo A. Leto Barone, MD Craniofacial Surgery Graham Schwarz, MD Gerald Brandacher, MD Bernard Lee, MD, Journal of Eric Williams, MD Pedro Cavadas, MD, PhD Reconstructive Microsurgery Hadley Wood, MD Scott Hollenbeck, MD Hsian-Jenn Wang, MD, Journal of Research Zuhaib Ibrahim, MD Taiwan Society of Plastic Surgeons Harold Friedman, MD, PhD, Christina Kaufman, PhD Michael Wong, MD, California Society Associate Editor Luis Landin, MD of Plastic Surgery Supplement Editor Justine Lee, MD, PhD, Associate Editor Gerald Lipshutz, MD Feng Zhang, MD, PhD, Microsurgery Aron D. Wahrman, MD, MBA, Samir Mardini, MD MHCDS, FACS, Associate Editor Stefan Schneeberger, MD DISCLAIMER Advertisements in Annals of Plastic Surgery for products, services, courses, and symposia are published with a Bcaveat emptorYbuyer beware[ understanding. The Publisher, the Editor, and individual members of the Editorial Board do not imply warranty of products, nor quality, validity, or approval of the educational material offered by such advertisements.

Annals of Plastic Surgery (ISSN 0148-7043), is published monthly by Wolters Kluwer Health, Inc., at 14700 Citicorp Drive, Bldg. 3, Hagerstown, MD 21742. Business offices are located at Two Commerce Square, 2001 Market St., Philadelphia, PA, 19103. Periodical postage paid at Hagerstown, MD, and at additional mailing offices. Copyright * 2016 by Wolters Kluwer Health, Inc. Annual Subscription Rates: United States - $749 Individual, $1,708 Institution, $344 In-training. Rest of World - $968 Individual, $2,090 Institution, $333 In- training. Single copy rate $172. All prices include a handling charge. Subscriptions outside of North America must add $30 for airfreight delivery. United States residents of AL, CO, DC, FL, GA, HI, IA, ID, IN, KS, KY, LA, MD, MO, ND, NM, NV, PR, RI, SC, SD, UT, VT, WA, WV add state sales tax. The GST tax of 7% must be added to all orders shipped to Canada (Wolters Kluwer Health GST Identification #895524239, Publications Mail Agreement #1119672). Subscription prices outside the United States must be prepaid. Prices subject to change without notice. Visit us online at www.lww.com. Individual and in-training subscription rates include print and access to the online version. Institutional rates are for print only; online subscriptions are available via Ovid. Institutions can choose to purchase a print and online subscription together for a discounted rate. Institutions that wish to purchase a print subscription, please contact Wolters Kluwer Health, Inc., at 14700 Citicorp Drive, Bldg. 3, Hagerstown, MD 21742; phone 800-638-3030 (outside the United States 301-223-2300; fax 301-223-2400). Institutions that wish to purchase an online subscription or online with print, please contact the Ovid Regional Sales Office near you or visit www.ovid.com/site/index.jsp and select Contact and Locations. Address for non-member subscription information, orders, or change of address: Wolters Kluwer Health, Inc. phone 800-638-3030 (outside the United States 301-223-2300); fax 301-223-2400. In Japan, contact LWW Igaku-Shoin Ltd., 3-23-14 Hongo, Bunkyo-ku, Tokyo 113-0033; phone 81-3-5689-5400; fax 81-3-5689-5402. In Bangladesh, India, Nepal, Sri Lanka, and Pakistan, contact Globe Publications Pvt. B-13 3rd Floor, A Block, Shopping Complex, Naraina Vihar, Ring Road, New Delhi, 110028; phone 91-11-579-3211; fax 91-11-579-8876. Postmaster: Send address changes to Annals of Plastic Surgery, P.O. Box 1550, Hagerstown, MD 21740. Annals of Volume 78 Supplement 5 Plastic Surgery June 2017

Supplemental Digital Content is available in the article. Contents Editorial S257 President’s Introduction to the 2017 SESPRS Supplement Walter L. Erhardt, MD, FACS

Clinical Papers S258 Neoadjuvant Therapy Combined With Oncoplastic Reduction for High-Stage Breast Cancer Patients Justine S. Broecker, BA, Alexandra M. Hart, MD, Toncred M. Styblo, MD, and Albert Losken, MD S263 The Muscle-Sparing Latissimus Dorsi Flap for Breast Reconstruction: A Retrospective Review of 126 Consecutive Flaps Jonathan Cook, MD, Jessica Waughtel, DO, Christopher Brooks, MD, Dawn Hardin, RN, Yin Kan Hwee, MD, and Yoav Barnavon, MD S269 Complications and Recurrence in Implant-Sparing Oncologic Breast Surgery Joshua B. Elston, MD, Sangeetha Prabhakaran, MD, Amina Lleshi, MPH, Brianna Castillo, BS, Weihong Sun, MD, Ambuj Kumar, MD, Zhenjun Ma, MD, Paul D. Smith, MD, and Deniz Dayicioglu, MD S275 Unilateral Versus Bilateral Breast Reconstruction: Is Less Really More? Lauren V. Kuykendall, MD, Bugra Tugertimur, MS, Corin Agoris, MS, Sara Bijan, MS, Ambuj Kumar, MD, MPH, and Deniz Dayicioglu, MD S279 Numerical Correlation of Levator Advancement in Preoperative Planning Valeria Makeeva, BSc, Sherry S. Collawn, MD, PhD, Evelina N. Pierce, BA, Mina S. Mousa, MD, Jennifer H. Yang, MD, Peter N. Davison, BA, and Elodie C. Jospitre, BSc S282 Dermal Autograft Using Donor Breast as Alternative to Acellular Dermal Matrices in Tissue Expander Breast Reconstruction: A Comparative Review William Darden North, MD, Christopher S. Kubajak, BS, Brad St. Martin, MPH, and Brian Rinker, MD S286 The Horseshoe Labiaplasty: Problems and Pearls Adam J. Oppenheimer, MD, FACS S289 Single Institution Review of Patients With Prior Breast Augmentation Undergoing Breast Conservation Therapy for Breast Cancer Sangeetha Prabhakaran, MD, Joshua B. Elston, MD, Amina Lleshi, MPH, Ambuj Kumar, MD, Weihong Sun, MD, Nazanin Khakpour, MD, and Deniz Dayicioglu, MD S292 Use of Processed Nerve Allografts to Repair Nerve Injuries Greater Than 25 mm in the Hand Brian Rinker, MD, FACS, Jozef Zoldos, MD, Renata V. Weber, MD, Jason Ko, MD, Wesley Thayer, MD, PhD, Jeffrey Greenberg, MD, Fraser J. Leversedge, MD, Bauback Safa, MD, and Gregory Buncke, MD (continued next page) Annals of Plastic Surgery & Volume 78, Supplement 5, June 2017

Contents (continued)

S296 Fluorescein Isothiocyanate: A Novel Application for Lymphatic Surgery Lisa Spiguel, MD, Christiana Shaw, MD, Adam Katz, MD, Lifei Guo, MD, PhD, Hung-Chi Chen, MD, Bernard T. Lee, MD, and Dhruv Singhal, MD S299 A Comparison of 4 Analgesic Regimens for Acute Postoperative Pain Control in Breast Augmentation Patients Pamela Tan, MD, Morgan Sparks Martin, MD, Nina Shank, BA, Leann Myers, PhD, Emily Wolfe, MD, John Lindsey, MD, and Stephen Metzinger, MD

Research Papers S305 Clinical, Biomechanical, and Anatomic Investigation of Colles Fascia and Pubic Ramus Periosteum for Use During Medial Thighplasty Martin J. Carney, BS, Tim Matatov, MD, Matthew Freeman, MD, John Miller, BS, Rahul Vemula, MD, Jason Schuster, MSBME, Michael Dancisak, PhD, John Lindsey, MD, and Guenevere Rae, PhD S311 Plasticity of Auricular Cartilage in Response to Hormone Therapy Patrick S. Cottler, PhD, Matthew D. McLeod, MD, MPH, Jesse I. Payton, BS, Angela Pineros-Fernandez, MD, and Jonathan S. Black, MD S315 Abdominal Wall Allograft: Preclinical Biomechanical Investigation of a Novel Reconstructive Adjunct Brent R. DeGeorge, Jr, MD, PhD, Jonathan Bryce Olenczak, MD, Angela Pineros-Fernandez, MD, Raymond F. Morgan, MD, Patrick S. Cottler, PhD, and David B. Drake, MD S322 Identifying Risk Factors Leading to Unanticipated Postoperative Readmission Peter A. Felice, MD, David T. Kerekes, BS, and Bruce A. Mast, MD S325 Preserving the Legitimacy of Board Certification Michael S. Hanemann, Jr, MD, FACS, Holly C. Wall, MD, FACS, and John A. Dean, MD, FACS S328 Bridging the Gap: Engineered Porcine-derived Urinary Bladder Matrix Conduits as a Novel Scaffold for Peripheral Nerve Regeneration Lyly Nguyen, MD, Ashkan Afshari, MD, Nathaniel D. Kelm, BS, Alonda C. Pollins, MS, R. Bruce Shack, MD, Mark D. Does, PhD, and Wesley P. Thayer, MD, PhD S335 Effects of Collagenase Digestion and Stromal Vascular Fraction Supplementation on Volume Retention of Fat Grafts Jonathan B. Olenczak, MD, Scott A. Seaman, PhD, Kant Y. Lin, MD, Angela Pineros-Fernandez, MD, Catherine E. Davis, Lisa S. Salopek, LVT, RLATG, Shayn M. Peirce, PhD, and Patrick S. Cottler, PhD S343 A Formula for Planning and Predicting Postoperative Mammoplasty Results Mary G. Smithson, BS, Sherry S. Collawn, MD, PhD, Mina S. Mousa, MD, and Carly M. Bramel, PA-C S347 The Sternum-Nipple Distance is Double the Nipple-Inframammary Fold Distance in Macromastia Thomas N. Steele, BS, Julian J. Pribaz, MD, and Frank H. Lau, MD

The Instructions for Authors can be accessed on this journal’s Web site at www.annalsplasticsurgery.com. EDITORIAL

President's Introduction to the 2017 SESPRS Supplement Walter L. Erhardt, MD, FACS

t is with pleasure that I introduce this Annals of Plastic Surgery featuring articles presented at the 59th Annual Meet- I ing of the Southeastern Society of Plastic and Reconstructive Surgeons, which was held on June 11 to 15, 2016, at Disney World's Grand Floridian Resort and Spa. This supplement marks the latest of many years of collaboration through which the southeastern has had the honor of having many of the articles presented at our annual meeting accepted for publication in the Annals of Plastic Surgery. The stated mission of the SESPRS is “advancing professional excellence, quality education, and regional colle- giality,” and the Society's main educational endeavors, the Breast and Oculoplastic Symposia held annually in Atlanta in January and the Society's Annual Meeting held at a resort in the southeast each June, are 2 renowned plastic surgery educational events. Whereas the symposia focus on single anatomic regions, the annual meeting's content, developed in response to member input, covers the spectrum of plastic surgery. A combination of panels, lectureships (including MOC-approved presentations), member articles, and a robustly competitive resident article competition, the Annual Meeting format contains both aesthetic and reconstructive topics, as well as topics to aid our members in maintaining professional and efficient practices. True to the scope of last year's meeting, the 19 articles included in this Annals of Plastic Surgery issue range from research to clinical and aesthetic to reconstructive. All of the abstracts were peer reviewed by the SESPRS Scien- tific Program Committee chaired by Dr Braun Graham and formatted by the authors for inclusion in this issue. My thanks to the authors and to the Annals of Plastic Surgery editorial staff for their help in making this supplement possible. I hope you will find the following articles stimulating.

Received April 18, 2017, and accepted for publication, after revision April 18, 2017. Conflicts of interest and sources of funding: none declared. Copyright © 2017 Wolters Kluwer Health, Inc. All rights reserved. ISSN: 0148-7043/17/7806–S257 DOI: 10.1097/SAP.0000000000001154

Annals of Plastic Surgery • Volume 78, Supplement 5, June 2017 www.annalsplasticsurgery.com S257 CLINICAL PAPER

Neoadjuvant Therapy Combined With Oncoplastic Reduction for High-Stage Breast Cancer Patients Justine S. Broecker, BA,* Alexandra M. Hart, MD,† Toncred M. Styblo, MD,‡ and Albert Losken, MD†

the oncologic safety of neoadjuvant therapy (NT) combined with Objective: Oncoplastic surgery has been shown to be a good alternative to breast oncoplastic reduction has also been well established.10,11 conservation surgery (BCS) alone for patients with breast cancer. Its role in pa- There has been limited evaluation of the oncologic safety of NT tients with advanced disease is unclear. In this study, we evaluate the safety of followed by oncoplastic reduction for higher stage breast cancer pa- oncoplastic BCS (OBCS) in patients who received neoadjuvant therapy (NT) tients.12,13 The purpose of this report is to determine whether it is safe for high stage breast cancer. to perform NT followed by oncoplastic reduction in higher stage breast Methods: The oncologic outcomes of consecutive patients classified as high cancer patients. stage (>T2 or at least N1) who received NT followed by BCS at EUH by a single breast surgeon (T.M.S.) from September 2004 until June 2015 were compared METHODS with those who received BCS combined with an oncoplastic reduction. Patients This is a retrospective review of all patients with a diagnosis of were surveyed using the BREAST-Qto determine their satisfaction after surgery. high stage breast cancer (>T1 or at least N1) who received NT before Results: A total of 87 patients were included in this series. The mean initial tumor breast conservation therapy between September 2004 and June 2015. size (4.37 vs 2.56 cm), the weight of the surgical specimen, and the post-NT tumor Patients were excluded from this study if they had metastasis (M1) at size were all larger in the OBCS group as compared with BCS alone (1.54 vs the time of diagnosis or did not adhere to the recommended standard 1.29 cm). The mean follow-up was 44 months. The average percent reduction in tu- treatment for their disease. Study approval was obtained from the insti- mor size in response to NT was slightly greater in the OBCS group (61 vs 52%). tutional review board. Oncologic outcomes were similar for OBCS reduction and BCS groups, respec- Patients were stratified into 2 groups based on surgical proce- tively: positive margin rate, reexcision rate, completion mastectomy rate, local recur- dures received: those who underwent BCT followed by immediate rence rate, and 5-year DSS. Patient satisfaction was similar between the 2 groups. oncoplastic reduction (group 1) and those who underwent BCT alone Conclusions: The oncoplastic approach in high stage patients treated with neoad- without any immediate reconstruction (group 2). All BCTs were per- juvant systemic Powered by Editorial Manager and ProduXion Manager from formed by a single oncologic surgeon (T.M.S.) and all additional recon- Aries Systems Corporation therapy appears to be as safe and effective when com- struction was performed by a single plastic surgeon (AL). Clinical and pared to BCS alone. Oncoplastic BCS paired with NT broadens the indication for pathologic variables evaluated included patient demographics, preop- BCS for patients with larger tumor size. erative histology and clinical stage, NT received, details of BCTas well Key Words: oncoplastic surgery, oncoplastic reduction, breast surgery as oncoplastic reduction if received, postoperative histology, receptor (Ann Plast Surg 2017;78: S258–S262) status, adjuvant therapy and clinical course after surgery. Patients were defined as high stage based on clinical stage before initiating NT (>T1 ncoplastic surgery merges oncology and plastic surgery to simul- or at least N1). Intraoperative data evaluated included tumor size, spec- O taneously treat patients with breast cancer and address tissue de- imen weights, nodal status, and specimen radiography when appropri- fects, and has gained popularity as an alternative technique to breast ate. In all cases, the tumor was excised by the surgical oncologist before conservation therapy (BCT) alone.1,2 Oncoplastic breast conservation reduction and submitted for routine pathologic analysis. All patients surgery (OBCS) can be further subdivided into volume displacement having preoperative imaging guidance confirmed specimen adequacy and volume replacement forms of reconstruction. Oncoplastic reduc- with intraoperative specimen radiography. Both pre-NT and post-NT tion is defined as volume displacement form of reconstruction at the tumor size were determined by imaging (either ultrasound or magnetic time of lumpectomy.3 Oncoplastic techniques have evolved to broaden resonance imaging) and percent change in tumor size in response to NT the indications for breast conservation therapy and to improve aesthetic was then calculated between the 2 tumor sizes. results after BCT.4–7 Additional oncologic benefits of oncoplastic tech- Outcomes of interest included positive surgical margins, surgical niques compared with BCT alone have also been well established.8,9 reexcision, progression to completion mastectomy, local recurrence, metastasis, and death. Before the 2013 SSO/ASTRO guidelines some Oncoplastic techniques may broaden indications of BCT to include 14 larger tumor breast cancer (>4 cm) and locally advanced breast cancer patients received re-excision for margins 1 mm or less. by offering larger resection potential with wider free surgical margins, Patients from both groups were surveyed using the Breast-Q to lower incidence of positive margins, and potentially fewer reexcisions determine patient satisfaction with their breast surgery. Surveys were and completion mastectomies.8,9 mailed to patients or conducted over the telephone. The ability of the OBCS technique to preferentially treat larger tumors also makes it amenable to neoadjuvant chemotherapy, and Statistical Analysis General frequencies, chi-square and Kaplan-Meier statistical analysis were calculated using SPSS 23.0 (Armonk, NY: IBM Corp). Received December 20, 2016, and accepted for publication, after revision December Chi square analysis was performed to compare BCS and OBCS co- 31, 2016. ’ From the *Emory University School of Medicine, †Department of Plastic Surgery, and horts. Fischer s exact test was used to calculate P values. Survival anal- ‡Department of Surgical Oncology, Emory University School of Medicine, ysis was performed using Kaplan-Meier. Atlanta, GA. Conflicts of interest and sources of funding: none declared. Reprints: Albert Losken, MD, Division of Plastic and Reconstructive Surgery, RESULTS Department of Surgery, Emory University School of Medicine, 550 Peachtree St NE 7 Emory MOT, Atlanta, GA. E-mail: [email protected]. Demographic and Clinicopathologic Results Copyright © 2017 Wolters Kluwer Health, Inc. All rights reserved. ISSN: 0148-7043/17/7806–S258 We included a total of 87 patients in our review who received NT DOI: 10.1097/SAP.0000000000001062 followed by breast-conserving therapy (BCT) as shown in Table 1.

S258 www.annalsplasticsurgery.com Annals of Plastic Surgery • Volume 78, Supplement 5, June 2017 Annals of Plastic Surgery • Volume 78, Supplement 5, June 2017 Oncoplastic Reduction for High-Stage Breast Cancer

TABLE 1. Demographic, Clinicopathologic, and Oncologic Results

All Patients, 87 OBCS, 47 (53) BCS, 40 (47) P Age: mean (range), y 57 (22–86) 53 (22–73) 61 (35–86) 0.002 Race 0.41 White 43 (51) 21 (45) 22 (60) African-American 38 (45) 24 (51) 14 (38) Asian 2 (2) 1 (2) 1 (3) Hispanic 1 (1) 1 (2) 0 BMI, mean (range) 30.5 (18.5–53.7) 30.8 (22.0–44.7) 28.9 (19.5–43.1) 0.43 IDC 73 (84) 36 (90) 37 (79) 0.25 ILC 12 (14) 4 (10) 8 (17) ER+/PR+ 48 (55) 26 (55) 22 (55) 1.00 Luminal A 34 (39) 19 (30) 15 (38) 0.95 Luminal B 14 (16) 7 (15) 7 (18) 0.97 HER-2+ 29 (33) 17 (36) 12 (30) 0.70 Triple-negative 22 (25) 9 (19) 13 (33) 0.24 Pre-NT AJC stage: 0.94 2 66 (74) 35 (75) 31 (78) 3 21(24) 12(26) 9(23) Preoperative T stage: 0.05 T1 19 (21) 6 (13) 13 (33) T2 49 (55) 28 (60) 21 (53) T3 11 (12) 9 (19) 2 (5) T4 8 (9) 4 (9) 4 (10) Preoperative N stage: 0.31 N0 39 (44) 20 (43) 19 (48) N1 38 (43) 23 (49) 15 (38) N2 9 (10) 3 (6) 6 (15) N3 1 (1) 1 (2) 0 Pre-NT tumor size: mean (range), cm 3.58 (0.40–11.00) 4.37 (0.70–11.00) 2.65 (0.40–6.50) 0.001 Postoperative tumor size: mean (range), cm 1.43 (0–5.2) 1.54 (0 = 4.4) 1.29 (0–5.2) 0.28 % Reduction in tumor size, mean (range) 57 (−80 to 100) 61 (−80 to 100) 52 (−40 to 100) 0.28 Response to NT: 0.26 Complete 32 (27) 17 (36) 15 (38) Partial 44 (51) 26 (55) 18 (45) Zero 3 (4) 0 3 (8) Progression 8 (9) 4 (9) 4 (10) Positive margin 6 (7) 3 (6) 3 (8) 1.00 Reexcision 5 (6) 2 (4) 3 (8) 0.66 Reexcision positive margins 0 0 0 n/a Completion mastectomy 5 (6) 3 (6) 2 (5) 1.00 NT Chemo (Adria) 41 (47) 19 (40) 22 (55) 0.19 Chemo (Tax) 21 (24) 12 (24) 9 (23) 0.83 Herc/Per chem 19 (21) 12 (24) 7 (18) 0.42 Hormonal 6 (7) 4 (9) 2 (5) 0.36 Local recurrence 5 (6) 2 (5) 3 (6) 1.00 Metastasis 12 (14) 5 (11) 7 (18) 0.37 Time to local recurrence: mean (range), mo 23 (10–48) 19 (17–22) 29 (10–48) 0.69 Time to metastasis: mean (range), mo 34 (11–71) 29 (11–71) 38 (19–59) 0.43 Death 10 (11) 4 (9) 6 (15) 0.49 NED 69 (78) 40 (85) 29 (73) 0.50 Length of follow-up: mean (range), mo 44 (2–125) 44 (2–125) 43.4 (10–101) 0.94 n/a indicates not applicable; NED, no evidence of disease.

Forty-seven patients underwent an oncoplastic reduction at the same time as BCT (group 1) and 40 patients received BCT alone (group 2). Patients in group 1 were younger (mean age, 53 vs 61; P = .002). Patients in group 1 had significantly larger initial tumor sizes before NT (4.37 cm vs 2.65 cm; P < .001) and had a more advanced T stage (28% vs 15% T3/T4; P = 0.05). The majority of patients in both groups were hormone receptor positive (either ER or PR positive or both) (55% vs 55%; P = 1.00). There were similar rates of Her- 2–positive breast cancer (36% vs 25%, P = 0.37) but there was a slightly higher rate of triple negatives in the group 2 (23% vs 33%, P = 0.48). Most patients received neoadjuvant chemotherapy in both groups. Most triple negative breast cancer patients received Adriamycin-based chemotherapy (60%) followed by taxol-based chemotherapy (40%). All eligible patients (by date) with HER-2neu am- plification received herceptin or pertuzumab based induction, and a smaller percentage (7%) received hormonal NT. Response to NT was FIGURE 2. Molecular subtypes. greater in group 1 compared with group 2 (average % reduction in tumor size, 61% vs 52%; P = 0.28) (Figs. 1–4). The mean weight of excised tumor was larger in group 1 (66% vs 50%, P = 0.63) and how their breasts lined up with each other (152.3 vs 70.2 g; P = 0.012). The tumor size determined by imaging (53% vs 31%, P = 0.42). Please refer to Figure 5 for results from the pa- after NT and surgical excision was also larger in group 1 (1.29 cm vs tient survey regarding their satisfaction with breast surgery. We also sur- 1.04 cm, P = 0.45). veyed patients subjectively, and a common theme mentioned among The majority of patients received adjuvant hormonal therapy several patients was that radiation changed the cosmetic result of their (51% vs 60%) as well as radiation (96% vs 93%) after surgery (ex- breast after surgery. ceptions included completion mastectomy or patient and/or Medical Doctor choice). DISCUSSION Currently, there is a 12.4% lifetime risk of invasive breast cancer Oncologic Outcomes for women in the United States. Screening and management of breast Positive margin rate (6% vs 8%, P = 1.00), re-excision rate cancer has evolved over the past few decades, resulting in an increased (4% vs 8%, P = 0.66), completion mastectomy rate (6% vs 5%, survival rate.15 Therefore, the surgical management of breast cancer has P = 1.00), local recurrence rate (5% vs 6%, P = 1.00), metastasis rate also shifted focus to reflect these changes in survival to enhance patient (11% vs 18%, P = 0.37), and death rate (9% vs 15%, P =0.49)were quality of life after breast cancer. Breast-conserving surgery was intro- similar between the 2 groups. The mean time to local recurrence was duced in 198116 and has continued to demonstrate comparable oncolog- 23 months, and similar between the 2 groups (19 vs 29 months; ical outcomes to radical mastectomy with superior quality of life for the P = 0.69). Mean time to metastasis (29 months vs 38 months; treatment of some breast cancers.17–19 However, breast conservation P = 0.43) was longer in the BCS group. Average length of follow up surgery has limitations to the size of tumor that can be adequately re- – was 44 months for both groups combined (2 125 months) with similar moved without significant breast deformity. Oncoplastic breast con- rate of follow-up between the 2 groups. Five-year disease-specific sur- serving surgery (OBCS) was introduced into the literature in the vival (95% vs 100%, P = 0.64) were similar between groups 1 and 2, mid-1990s20 as a means to expand the benefit of breast conserving to respectively. At most recent follow-up, 85% of patients in group 1 include wider excisions while maintaining cosmesis.1 There are a vari- and 73% of patients in group 2 were disease-free (no evidence of ety of OBCS techniques available depending on various tumor and pa- disease, (P = 0.05). tient characteristics1 that are broadly divided into volume displacement, volume replacement and reduction.21 Oncoplastic reduction combines Patient Satisfaction Results lumpectomy with traditional breast reduction. Although OBCS includ- A total of 30 patients completed a subset of questions from the ing oncoplastic reduction has been developed with oncologic safety in postoperative BREAST-Q survey regarding satisfaction with their mind and has often shown similar oncologic outcomes to breast con- breast surgery. More patients from the OBCS reported to be very satis- serving surgery alone, little is known about the indications for and on- fied with the appearance (P = 0.03). Also, patients in the OBCS group cologic safety of oncoplastic reduction for higher-stage patients who were more often satisfied with the size of their breast postoperatively have a higher risk of mortality from breast cancer.

FIGURE 1. Change in tumor size in response to NT.

S260 www.annalsplasticsurgery.com © 2017 Wolters Kluwer Health, Inc. All rights reserved. Annals of Plastic Surgery • Volume 78, Supplement 5, June 2017 Oncoplastic Reduction for High-Stage Breast Cancer

FIGURE 3. Complete pathologic response by molecular subtype.

We reviewed 87 high-stage breast cancer patients who underwent breast conserving surgery after NT. Forty patients received breast conservation therapy alone and 47 received immediate oncoplastic reduction after their breast-conserving surgery. Our patient population reflects similar demographics and clinical characteristics to high-stage breast cancer patients diagnosed and treated in the United States and were relatively similar between the 2 groups. The mean age at time of surgery of our combined groups was 57 years, slightly younger than the median age of breast cancer diagnosis in the United States (61 years)22; and, the patients who had oncoplastic reduction were significantly younger than those who underwent breast conserving surgery alone (53 vs 61, P = 0.002). This reflects a trend nationally in that younger women are more likely to pursue breast reconstruction compared to older women.8 The average tumor size preoperatively was significantly larger for the OBCS group compared with the BCT group (4.36 cm vs 2.54 cm, P < 0.001), and OBCS patients had a significantly more advanced T stage (28% vs 15%, P = 0.05). The ability of oncoplastic surgery and immediate reconstruction to enable greater resections with breast conserving therapy as opposed to mastectomy has been well

FIGURE 4. Disease-specific survival OBCS versus BCS (Kaplan-Meier) (P = 0.64; OBCS median OS, 123.6; BCS FIGURE 5. Patient satisfaction results. 1, very dissatisfied; 2, median OS, 101.1). somewhat dissatisfied; 3, somewhat satisfied; 4, very satisfied.

established.8 The OBCS patients tended to have a greater response to 3. Losken A, Hamdi M. Partial breast reconstruction: current perspectives. Plast NT with a slightly greater average percent reduction in tumor size Reconstr Surg. 2009;124:722–736. (69.5% vs 48.5%, P = 0.26), although, oncoplastic patients had larger 4. Veiga DF, Veiga-Filho J, Ribeiro LM, et al. Quality-of-life and self-esteem out- resections (152.3 g vs 70.8g, P = 0.012). It is unclear why these patients comes after oncoplastic breast-conserving surgery. Plast Reconstr Surg.2010; 125:1051–1056. may have responded to a greater degree despite similar molecular 5. Losken A, Pinell XA, Eskenazi B. The benefits of partial versus total breast re- subtypes and treatment. As a result, tumor sizes were relatively construction for women with macromastia. Plast Reconstr Surg. 2010;125: similar between the 2 groups at the time of surgery (1.54 cm vs 1051–1056. 1.29 cm, P = 0.28). 6. Regano S, Hernanz F, Ortega E, et al. Oncoplastic techniques extend breast con- Our results demonstrate similar oncologic outcomes for high- serving surgery to patients with neoadjuvant chemotherapy response unfit for con- stage breast cancer patients treated with oncoplastic breast-conserving ventional techniques. World J Surg. 2009;33:2082–2086. surgery as compared to breast conserving surgery alone. Positive mar- 7. Kaviani A, Safavi A, Mohammadzadeh N, et al. Oncoplastic surgery in breast conservation: a prospective evaluation of the patients, techniques and oncologic gin rate, reexcision rate, completion mastectomy rate, local recurrence, outcomes. Am J Surg. 2014;208:727–734. metastasis and death rate were similar between the 2 groups. In addi- 8. Losken A, Dugal CS, Styblo TM, et al. A meta-analysis comparing breast conser- tion, 5-year disease-specific survival was similar between the 2 groups. vation therapy alone to the oncoplastic technique. Ann Plast Surg. 2014;72: The oncologic safety of oncoplastic techniques has been well estab- 145–149. lished for lower-stage breast cancers,23 but its safety among higher- 9. Losken A, Pinell-White X, Hart AM, et al. The oncoplastic reduction approach to stage breast cancer patients, particularly after NT, has been less well breast conservation therapy: benefits for margin control. Aesthet Surg J.2014;34: – studied.12,13 Although we have a small study size, our results along with 1185 1191. others suggest that oncoplastic reduction is a safe alternative after NT 10. Regaño S, Hernanz F, Ortega E, et al. Oncoplastic techniques extend breast- conserving surgery to patients with neoadjuvant chemotherapy response unfit for high-stage breast cancer patients and should be further studied with for conventional techniques. World J Surg. 2009;33:2082–2086. larger sample sizes and longer follow-up. 11. Mazouni C, Naveau A, Kane A, et al. The role of oncoplastic breast surgery in the One of the primary benefits of oncoplastic surgery is its demon- management of breast cancer treated with primary chemotherapy. Breast.2013; strated ability to improve patient cosmesis and satisfaction after surgical 22:1189–1193. resection for breast cancer.8,23 Oncoplastic reduction in particular has 12. Emiroglu M, Sert I, Karaali C, et al. The effectiveness of simultaneous oncoplastic the added benefit of reducing macromastia and because it is often per- breast surgery in patients with locally advanced breast cancer. Breast Cancer. 2016;23:463–470. formed bilaterally can retain breast symmetry after lumpectomy. Un- fortunately, as has been well documented, radiation after BCT can 13. Bogusevicius A, Cepuliene D, Sepetauskiene E. The integrated evaluation of the 24 results of oncoplastic surgery for locally advanced breast cancer. Breast J.2014; significantly alter cosmetic results of breast reconstruction, and many 20:53–60. of our patients noted that radiation similarly negatively impacted the 14. Moran MS, Schnitt SJ, Giuliano AE, et al. Society of Surgical Oncology-American cosmesis of their breast after either surgery through the development Society for Radiation Oncology consensus guideline on margins for breast-conserving surgery with whole-breast irradiation in stages I and II in- of skin fibrosis, arm lymphedema, and change in the size of the radiated – breast. However, in general, based on the answers to the BREAST-Q, vasive breast cancer. Ann Surg Oncol. 2014;21:704 716. 15. Park JH, Anderson WF, Gail MH. Improvements in US breast cancer survival and patients who underwent oncoplastic reduction appeared more likely proportion explained by tumor size and estrogen-receptor status. JClinOncol. to be satisfied with the comfort, appearance and size of their breast 2015;33:2870–2876. compared with the patients who underwent breast conservation 16. Veronesi U, Saccozzi R, Del Vecchio M, et al. Comparing radical mastectomy therapy alone. with quadrantectomy, axillary dissection, and radiotherapy in patients with small Our study is limited to a small, retrospective review of patients cancers of the breast. N Engl J Med. 1981;305:6–11. whose surgery was chosen based on Medical Doctor/patient preference 17. Veronesi U, Cascinelli N, Mariani L, et al. Twenty-year follow-up of a randomized rather than randomization. To further evaluate the impact of oncoplastic study comparing breast-conserving surgery with radical mastectomy for early breast cancer. N Engl J Med. 2002;347:1227–1232. surgery oncologic outcomes, larger prospective randomized studies with 18. Fisher B, Bauer M, Margolese R, et al. Five-year results of a randomized clinical longer follow-up are needed. trial comparing total mastectomy and segmental mastectomy with or without radiation in the treatment of breast cancer. N Engl J Med. 1985;312:665–673. CONCLUSIONS 19. Fisher B, Jeong J, Anderson S, et al. Twenty-five-year follow-up of a randomized trial comparing radical mastectomy, total mastectomy, and total mastectomy Our results suggest that oncoplastic reduction is an oncologically followed by irradiation. N Engl J Med. 2002;347:567–575. safe alternative to breast-conserving surgery alone with the potential for 20. Audretsch WP. Reconstruction of the partial mastectomy defect: classification and superior patient cosmesis and satisfaction for high stage breast cancer method. In: Surgery of the breast: principles and art. Philadelphia: Lippincott patients after NT. Furthermore, oncoplastic reduction paired with NT Raven; 1998. appears to be able to broaden the indication for breast-conserving sur- 21. Munhoz AM, Montag E, Gemperil R. Oncoplastic breast surgery: indications, gery for higher-stage patients with larger tumor sizes who would other- techniques and perspectives. Gland Surg. 2013;2:143–157. wise require mastectomy. 22. Howlader N, Noone AM, Krapcho M, et al SEER Cancer Statistics Review, 1975–2010. Bethesda, MD: National Cancer Institute; 2013. http://seer.cancer. gov/csr/1975_2010/, based on November 2012 SEER data submission. REFERENCES 23. De La Cruz L, Blankenship SA, Chatterjee A, et al. Outcomes after oncoplastic 1. Berry MG, Fitoussi AD, Curnier A, et al. Oncoplastic breast surgery: a review and breast-conserving surgery in breast cancer patients: a systematic literature review. systematic approach. J Plast Reconstr Aesthet Surg. 2010;63:1233–1243. AnnSurgOncol. 2016;23:3247–3258. Epub. 2. Losken A, Ghazi B. An update on oncoplastic surgery. Plast Reconstr Surg. 2012; 24. Meric F, Buchholz TA, Mirza NQ, et al. Long-term complications associated with 129:382e–383e. breast-conservation surgery and radiotherapy. Ann Surg Oncol. 2002;9:543.

The Muscle-Sparing Latissimus Dorsi Flap for Breast Reconstruction A Retrospective Review of 126 Consecutive Flaps Jonathan Cook, MD,* Jessica Waughtel, DO,† Christopher Brooks, MD,‡ Dawn Hardin, RN,‡ Yin Kan Hwee, MD,‡ and Yoav Barnavon, MD‡

was performed. Inclusion criteria were any patient who underwent Abstract: Our experience in the use of muscle-sparing latissimus dorsi (MSLD) MSLD flap reconstruction by our senior author during the defined flaps for breast reconstruction is presented. The procedure was performed on study period and had a minimum of 3 months of follow-up. 83 patients by the senior author over an 8-year period. Of the 83 patients re- Patient demographics, body mass index (BMI), and breast reviewed, a total of 126 MSLD flaps were done for immediate (26) or delayed constructive history were collected through careful review of the pa- (100) breast reconstructions. Preoperative and postoperative photographs were tients' medical records. Patients were divided into 3 groups based on taken of all patients, and complications as well as ancillary procedures were re- their indications for MSLD flap reconstruction: immediate reconstruc- corded. The MSLD flap is shown to be a versatile option for breast reconstruction tion, delayed primary reconstruction, and salvage of previously failed in a variety of clinical settings, with minimal complications and satisfactory breast reconstruction. aesthetic results. Out of 83 patients included in this study, there were a total of 126 Key Words: breast reconstruction, latissimus dorsi flap, myocutaneous flap, flaps. Mean age at time of MSLD flap reconstruction was 56 years muscle-sparing, MSLD, thoracodorsal artery, descending branch (range, 30–81 years). Mean BMI at the time of surgery was 30 kg/m2 (range, 17–47.2 kg/m2). All patients were followed in the office for a (Ann Plast Surg 2017;78: S263–S268) minimum of 3 months postoperatively (range, 3 months to 5 years). Complications were recorded and included the presence of infection (defined as the presence of erythema, wound drainage, or positive BACKGROUND wound cultures), clinical seroma, hematoma, and partial flap necrosis. The latissimus dorsi (LD) musculocutaneous flap has been a re- Ancillary procedures were also recorded, including placement of breast liable option for breast reconstruction since it was first described in implants and fat grafting. 1906.1–5 In 1995, Angrigiani et al2 were the first to describe the thoracodorsal artery perforator flap. Schwabegger et al3 in 2003 reported the advantages of a “muscle-sparing” approach, harvesting a larger skin paddle carried by a relatively small segment of LD muscle. Patient Selection In 2008, Hamdi4 noted that this skin paddle could be oriented obliquely Preoperative photographs were taken upon initial consultation. for improved contour. Saint-Cyr1 further refined this approach in 2009, The full range of operative choices was considered for all patients. emphasizing that a transversely oriented skin paddle with a pedicled When abdominally based reconstruction was not considered an option, descending-branch muscle-sparing latissimus dorsi (MSLD) flap re- or if patients were poor candidates for implant-based reconstruction, sulted in minimal functional deficits, a low incidence of donor-site they were offered MSLD flap reconstruction. Patients were considered seromas, and a cosmetically acceptable scar. appropriate candidates for MSLD flap reconstruction if they desired Considering the work done by previous authors, we aimed to de- autologous tissue reconstruction and if they were deemed unfit for termine whether these results are reproducible over a variety of clinical an abdominally based procedure. The ideal patient was one whose settings, having utility in an immediate and delayed fashion, and upper back adipose tissue provided an approximate match for their whether the MSLD flap could be utilized once other reconstructive pro- premastectomy breast tissue volume. cedures had failed. Whereas previous studies have reported on the efficacy, aesthetics, and complication rates of the MSLD flap, our study provides the largest retrospective review of the MSLD flap for breast re- Operative Technique construction. Our surgical technique, as well as its success in multiple The skin flap and the anterior border of the LD muscle are clinical settings, ancillary procedures, and associated complications, is marked preoperatively with the patient in the standing position. A pinch reported here. test of the transverse skin crease below the axillary fat pad is used to identify the widest portion of the transversely oriented flap. This crease PATIENTS AND METHODS is continuous with the inframammary fold (Fig. 1). The patient is posi- A retrospective chart review of patients who underwent breast re- tioned intraoperatively in the lateral decubitus position. An elliptical in- construction with MSLD flaps from September 2008 to January 2016 cision is made, beginning at the inframammary fold and extending towards the midline of the back, along the inferior margin of the skin Received October 31, 2016, and accepted for publication, after revision December 24, flap. The native back skin is elevated from the LD muscle inferiorly, 2016. and the anterior muscle border is identified. The latissimus muscle is di- From the *Department of Plastic Surgery, Cleveland Clinic Florida, Weston, FL; vided distal to the flap to create a 5-cm–wide muscle strip, and a retro- †Department of Plastic Surgery, Philadelphia College of Osteopathic Medicine, Philadelphia, PA; and ‡Department of Plastic Surgery, Memorial Regional grade dissection of the flap and overlying soft tissue paddle is Hospital, Hollywood, FL. performed. The anterior descending branch of the thoracodorsal artery Conflicts of interest and sources of funding: none declared. is identified with a Doppler probe, and muscle splitting is then further Reprints: Yoav Barnavon, MD, Plastic Surgical Specialists of South Florida, 1201 N performed. When available, indocyanine green imaging is used to con- 35th St, Suite 200, Hollywood, FL 33021. E-mail: [email protected]. Copyright © 2017 Wolters Kluwer Health, Inc. All rights reserved. firm adequate flap perfusion. The mastectomy defect is then prepared, if ISSN: 0148-7043/17/7806–S263 necessary, and the flap is transferred to the recipient bed. The back in- DOI: 10.1097/SAP.0000000000001036 cision is closed in layers over a closed-suction drain. The patient is then

Annals of Plastic Surgery • Volume 78, Supplement 5, June 2017 www.annalsplasticsurgery.com S263 Cook et al Annals of Plastic Surgery • Volume 78, Supplement 5, June 2017

TABLE 1. Ancillary Procedures

Ancillary Procedure No. Cases (%) Implants 22 (26.5) Fat grafting 52 (62.7) Total ancillary procedures 74 (89)

was then subdivided into delayed primary (33%) and salvage reconstruction (46%) (Fig. 2). The follow-up period ranged from 3 months to 5 years, and all patients were evaluated by the operating surgeon. Ancillary procedures were deemed necessary by assessing for symmetry, the presence of contour deformities, patient satisfaction, and ultimate goals (Table 1). Twenty-two patients (26.5%) underwent placement of breast implants as part of their reconstruction, and 52 patients (62.7%) had fat grafting. Regardless of indication, there were a total of 39 flap-related complications (summarized in Table 2). The most common complication in our study was minor infection, with a rate of 11.1%. This was followed by fat necrosis (7.9%), donor-site seroma (5.5%), partial flap necrosis (4%), and hematoma (1.6%). There were no cases of complete flap loss. Of the immediate primary MSLD flap group, there were 2 cases of minor infection (7.7%), including 1 case of an infected seroma diagnosed on culture of aspirated fluid and 1 case of seropurulent incisional drainage. Both cases were managed successfully with antibiotics alone. In addition, there was 1 case of donor-site seroma (3.8%), 2 cases of fat necrosis (7.7%), and 1 case of partial flap necrosis (3.8%). There were no hematomas or complete flap loss in the immediate primary group. The total complication rate for the immediate primary group was 23% (6 of a total of 26 flaps). In the delayed primary group, there were 2 cases of minor infection (4.8%), 5 cases of donor-site seroma (11.9%), 1 hematoma (2.4%), FIGURE 1. Preoperative marking of the skin flap and LD muscle. 4 cases of fat necrosis (9.5%), and 4 cases of partial flap necrosis (9.5%). One case of infection was simple cellulitis, and the other was repositioned into the supine position on the operating room table, and considered infected because of prolonged wound healing time, with the flap is partially deepithelialized before inset. subsequent resolution after antibiotics. The total complication rate in the delayed primary group was 38% (16 of 42 flaps). In the salvage MSLD flap group, there was a similar rate of complications at 28% (16 of 58 flaps). There were 10 cases of minor infec- RESULTS tion (17.2%), 1 case of seroma (1.7%), 1 case of hematoma (1.7%), and Eighty-three patients were included in this study, for a total of 4 cases of fat necrosis (6.9%). These results are summarized in Table 3. 126 flaps. Seventeen patients underwent immediate reconstruction, for a total of 26 flaps (21%). Sixty-six patients underwent delayed reconstruction, for a total of 100 flaps (79%). Delayed reconstruction CASE REPORTS Case 1 A 57-year-old patient presented status post left modified radical mastectomy and radiation therapy, with severe skin hyperpigmentation

TABLE 2. Flap-related Complications

Flap-related Complication No. Flaps (%) No complications 87 (69) Infection 14 (11.1) Seroma 7 (5.5) Hematoma 2 (1.6) Partial flap necrosis 5 (4) Fat necrosis 10 (7.9) FIGURE 2. Distribution of the clinical settings in which MSLD flap Total complications 38 (30) was used for reconstruction.

reconstruction. Her BMI was 29 kg/m2. She underwent delayed left TABLE 3. Flap-related Complications Based on Type of breast reconstruction with MSLD flap and subsequent fat grafting, nip- Reconstruction ple reconstruction, and nipple tattooing (Fig. 4).

Flap-related Complication Immediate (%) Delayed (%) Salvage (%) Case 3 Infection 2 (7.7) 2 (4.8) 10 (17.2) A 61-year-old female presented in consultation for possible im- Seroma 1 (3.8) 5 (11.9) 1 (1.7) mediate breast reconstruction. She was recently diagnosed with left Hematoma 0 (0) 1 (2.4) 1 (1.7) breast multifocal ductal carcinoma in-situ. Her BMI was 35 kg/m2. Fat necrosis 2 (7.7) 4 (9.5) 4 (6.9) Her surgical history included bilateral reduction mammoplasty. She Partial flap necrosis 1 (3.8) 4 (9.5) 0 (0) underwent immediate left breast reconstruction with MSLD flap with Total complications 6 (23) 16 (42) 16 (28) subsequent right mastopexy, fat grafting, and upper body lift (Fig. 5).

Case 4 and tissue contraction. She also had a past medical history of suction A 55-year-old female presented with a diagnosis of right breast lipectomy of the abdomen. She was not a candidate for implant-based invasive ductal carcinoma (Fig. 6). Genetic testing was also completed reconstruction. Her BMI was 31 kg/m2. She underwent delayed left given a strong family history of breast cancer. She was found to be pos- breast reconstruction with MSLD flap, subsequent fat grafting, and itive for the BRCA2 mutation. Her BMI was 27 kg/m2, and she had a right mastopexy for symmetry (Fig. 3). history of bilateral breast mastopexy and abdominoplasty. Initially, the patient desired implant-based reconstruction, and she underwent bilateral skin-sparing and nipple-sparing mastectomies with placement of Case 2 tissue expanders. Her postoperative course was complicated by partial A 53-year-old woman presented after having left modified flap necrosis. Tissue expanders were removed, and after 3 months, she radical mastectomy and radiation therapy. She desired breast recon- returned to the operating room for bilateral MSLD flaps reconstruction. struction with autologous tissue but was fearful of abdominally based Her second recovery was uneventful. We were able to attain reasonable

FIGURE 3. A 57-year-old female presented status post left modified radical mastectomy and radiation therapy. Preoperative imaging (top row). Fourteen months after delayed left breast reconstruction with MSLD, subsequent fat grafting and right mastopexy (bottom row).

FIGURE 4. A 53-year-old female previously underwent left modified radical mastectomy and radiation therapy. Preoperative photographs (top row). Twenty-eight months after left MSLD flap and subsequent fat grafting (bottom row).

FIGURE 5. A 61-year-old female diagnosed with left breast multifocal ductal carcinoma in-situ desired autologous breast reconstruction. Her surgical history included bilateral reduction mammoplasty. Preoperative photographs (top row). After immediate left breast reconstruction with MSLD flap with subsequent right mastopexy, fat grafting, and upper body lift (bottom row).

This 5-cm strip typically constitutes less than 20% of the muscle by width and, we believe, is the minimum amount of muscle that should be harvested to ensure perfusion of the overlying skin paddle. By limiting the width of this muscle cuff, the transverse branch of the thoracodorsal artery and the main thoracodorsal nerve may be left in situ. Like previous authors, we believe that the muscle strip plays an important role in protecting the neurovascular pedicle and decreasing seroma formation.1,3,6 Our study also highlights the versatility of the MSLD flap as a reconstructive tool in a variety of situations. One reason for the broad applicability of this flap is in the relatively forgiving relationship between the skin paddle and the underlying perforator.1,5 This enables the flap to be designed to accommodate a wide range of body habitus. Indeed, our patients' mean BMI at the time of surgery was 30 kg/m2 (range, 17–47.2 kg/m2). We achieved highly satisfactory results using the MSLD flap for immediate primary breast reconstruction. This was an attractive option for patients who preferred not to have a breast implant as part of their reconstruction and who desired to maintain an active lifestyle (such as tennis-players, golfers, or bowlers).6 In patients whose mastectomies included resection of the nipple-areola complex, we were often able to utilize the MSLD skin paddle as the foundation for areola reconstruction in subsequent procedures. We found that the difference in skin color, turgor, and texture of the MSLD skin paddle provided an ideal contrast to the surrounding skin of the native breast flap (Fig. 8). The MSLD flap was also particularly well-suited for the so- called salvage breast reconstruction (eg, after previous attempts at breast reconstruction surgery). In the irradiated or previously operated field, the MSLD flap provided the advantages of requiring a relatively simple dissection, with versatility in supporting a variety of flap orien- tations with respect to thoracodorsal artery perforator location.4,5 The anterior extent of the skin paddle dissection recruits additional volume to the flap and facilitates the creation of a conical shape for the breast mound. Coupled with a generous arc of rotation (up to 180 degrees), this flap can reach as far medially as the anterior midline. This expands the surgeon's options for flap placement and provides an opportunity to inset in a way that enables increased breast projection along with more natural ptosis. This study has several limitations. First, it is a retrospective FIGURE 6. Preoperative imaging of a 55-year-old female with a study, and although it does provide the largest series of patients reported history of bilateral breast mastopexy and abdominoplasty who in the literature to date, it is not possible to draw definitive conclusions was diagnosed with right breast intraductal carcinoma. about specific outcomes or complications. Indeed, the incidence of complications reported in our series is higher than noted in previous studies.4,6 This could be explained in part by the relatively broad criteria symmetry and a good aesthetic outcome after bilateral MSLD flaps we used in defining postoperative infection and seroma and could be with fat grafting to bilateral breasts (Fig. 7). corrected in future study through the use of standard definitions for postoperative complications. Further investigation in the form of a well-controlled cohort DISCUSSION study would be necessary to evaluate whether specific patient or opera- Techniques for breast reconstruction using thoracodorsal artery tive factors are associated with different complications, as outlined by based flaps have evolved over time, from the use of the entire LD mus- previous authors. Another area of future study would be in quantifying cle to the thoracodorsal artery perforator flap. Most recently, the pedi- the differences in operative times between the MSLD and other autolo- cled descending branch MSLD flap was described by Saint-Cyr et al,1 gous breast reconstructive procedures; we speculate that the relative who recognized that a transverse skin paddle may be oriented with release of positioning the patient (initially in the lateral decubitus position ative freedom in relation to the underlying perforator vessels. This en- and then supine for completion of flap insetting) provides for a particu- abled a more aesthetic outcome with minimal complications, when larly efficient operative approach. Finally, we plan to continue our study compared to earlier pedicle flap approaches.1,4,6 We report our experi- by quantifying patient satisfaction using a standardized instrument pro- ence using the MSLD flap for immediate primary, delayed primary, vided to patients during their postoperative reconstructive journey. and salvage breast reconstruction in a series of 83 consecutive patients (126 flaps). We achieved excellent subjective and objective outcomes, with follow-up ranging from 3 months to 5 years. CONCLUSIONS One key feature of our technique is the utilization of a 5-cm– This study reports our experience using the MSLD flap in a va- wide muscle strip of the anterior LD during flap harvest. Although this riety of settings for breast reconstruction in 83 patients (126 flaps). To differs slightly from previous recommendations,1,3–5,7 we found that our knowledge, this represents the largest series of MSLD flaps reported using a 5-cm muscle width allowed us to reliably include the anterior de- in the literature. Based on our experience, we believe that the MSLD flap scending branch of the thoracodorsal artery and its muscular perforators. can be considered a “workhorse” flap for breast reconstruction. We found

FIGURE 7. A 55 year-old-female after bilateral skin-sparing and nipple-sparing mastectomies and failed tissue expander reconstruction secondary to partial flap necrosis (top row). Fifteen months after bilateral MSLD flap reconstruction and fat grafting to bilateral breasts (bottom row).

excellent aesthetic and functional results from the use of this flap for immediate breast reconstruction, as well as in the delayed reconstructive setting (after mastectomy) and for salvage after prior failed breast reconstruction. The MSLD flap is a versatile alternative to breast reconstruction using other fasciocutaneous flaps or free tissue transfer techniques.

REFERENCES 1. Saint-Cyr M, Nagarkar P, Schaverien M, et al. The pedicled descending branch muscle-sparing latissimus dorsi flap for breast reconstruction. Plast Reconstr Surg. 2009;123:13–24. 2. Angrigiani C, Grilli D, Siebert J. Latissimus dorsi musculocutaneous flap without muscle. Plast Reconstr Surg. 1995;96:1608–1614. 3. Schwabegger AH, Harpf C, Rainer C. Muscle-sparing latissimus dorsi myocutaneous flap with maintenance of muscle innervation, function, and aesthetic appearance of the donor site. Plast Reconstr Surg. 2003;111:1407–1411. 4. Hamdi M, Van Landuyt K, Monstrey S, et al. Pedicled perforator flaps in breast reconstruction: a new concept. Br J Plast Surg. 2004;57:531–539. 5. Wong C, Saint-Cyr M. The pedicled descending branch muscle-sparing latissimus dorsi flap for trunk and upper extremity reconstruction. J Plast Reconstr Aesthet Surg. 2010;63:623–632. 6. Kim H, Wiraatmadja ES, Lim SY,et al. Comparison of morbidity of donor site following pedicled muscle-sparing latissimus dorsi flap versus extended latissimus dorsi flap breast reconstruction. J Plast Reconstr Aesthet Surg. 2013;66:640–646. 7. Zhu L, Mohan AT, Saint-Cyr M. A simple approach to harvest of the pedicled de- FIGURE 8. Contrast between the skin paddle and native breast scending branch muscle-sparing latissimus dorsi flap. J Plast Reconstr Aesthet flap skin. Surg. 2015;68:e179–e181.

Complications and Recurrence in Implant-Sparing Oncologic Breast Surgery Joshua B. Elston, MD,* Sangeetha Prabhakaran, MD,† Amina Lleshi, MPH,* Brianna Castillo, BS,* Weihong Sun, MD,† Ambuj Kumar,MD,* Zhenjun Ma, MD,† Paul D. Smith, MD,* and Deniz Dayicioglu, MD*

breast augmentation undergoing lumpectomy or mastectomy for Background: Patients with a history of prior breast augmentation and newly breast cancer and identify differences in complications between these diagnosed breast cancer represent a rapidly expanding and unique subset of 2 groups. patients. Prior studies have described changes in breast parenchyma and charac- Augmentation mammaplasty is one of the most common cos- teristic body habitus of previously augmented patients, as well as increased rates metic procedures performed in the United States and continues to rise of capsular contracture associated with breast conservation therapy. In our current in frequency. Since 2000, there has been a 35% increase in the num- study, we aimed to study the risk factors contributing to morbidity and whether ber of cosmetic augmentation performed annually with 286,000 aug- recurrence rates are higher in patients with prior breast augmentation undergoing mentations performed in 2014.1 As these numbers continue to rise, it lumpectomy or mastectomy for breast cancer and identify differences in compli- is not surprising that a unique population of patients with prior aug- cations between these 2 groups. mentation will develop breast cancer. There will be 231,840 pre- Methods: Retrospective analysis approved by institutional review board was per- dicted new diagnoses of breast cancer in females in 2015 occurring formed on patients with prior breast augmentation undergoing lumpectomy most frequently in those aged 55 to 64 years with a 5-year survival rate (N = 52) and mastectomy (N = 64) for breast cancer. of roughly 89%.2 Results: Patients with prior breast augmentation undergoing mastectomy had a Previously augmented patients represent a difficult popula- higher rate of complications compared with those undergoing lumpectomy tion for which to care. They have been previously documented as (20.3% vs 5.9% respectively, P = 0.031), after adjusting for patient-specific fac- being younger and have lower body mass index (BMI) than women tors including body mass index [odds ratio (OR), 0.242; 95% confidence interval 3 – – of similar age without augmentation. For these reasons, previously (CI), 0.063 0.922; P = 0.0376], tumor stage (OR, 0.257; 95% CI, 0.064 1.036; augmented patients discussing reconstructive options after a cancer P = 0.0562), smoking status (OR, 0.244; 95% CI, 0.065–0.918; P =0.0370),and – diagnosis may have higher cosmetic demands that need to be chemotherapy (OR, 0.242; 95% CI, 0.064 0.914; P = 0.0364). Four patients considered preoperatively. (7.7%) developed late complications in the lumpectomy group with 2 developing As more procedures, such as skin-sparing and nipple-sparing capsular contractures, 1 had fat necrosis and 1 needed complex reconstruction be- mastectomies, become more commonplace, patients may avoid a cause of flattening of the nipple-areolar complex. There was no difference in re- portion of the stigma associated with removing their breast by currence or tumor margins between lumpectomy and mastectomy groups. retaining anatomic units that are tied into the psychological aspect Conclusions: Patients with prior breast augmentation undergoing mastectomy of having a breast. Patients continue to opt for implant-based recon- have higher complication rates compared with lumpectomy even after adjusting struction with over 80% of reconstructions in 2014 involving some for tumor stage. There appears to be no increased oncologic risk associated with form of breast prosthetic.1 Those undergoing expander-implant re- either procedure given our current follow-up. Understanding these operative construction have the benefit of no donor sites but at the expense of risks may help in patients' decision-making process with regards to type of multiple office visits, more extensive surgery, and the associated oncologic surgery. stigma of an acquired breast absence.4 Patients with previously Key Words: implant-sparing, augmentation, lumpectomy, mastectomy, breast, placed submuscular implants present an opportunity for both the re- reconstruction, submuscular, complications, recurrence constructive surgeon and the patient: the cancer can be dealt with as oncologically feasible, and the patient retains the soft tissue envelope (Ann Plast Surg 2017;78: S269–S274) necessary for further procedures. The choice of oncologic surgery is decided between the patient and breast surgeon with well-documented efficacy and near equiva- atients with a history of prior breast augmentation and newly diag- lence in survival rates in mastectomy as well as breast conservation P nosed breast cancer represent a rapidly expanding and unique surgery.5 Although the literature is somewhat scarce in previously subset of patients. Prior studies have described changes in breast pa- augmented patients undergoing breast cancer removal, it appears that renchyma and characteristic body habitus of previously augmented sparing implants previously placed in a submuscular plane at the time patients, as well as increased rates of capsular contracture associated of lumpectomy or mastectomy is a reasonable option.3,6–8 The com- with breast conservation therapy. In our current study, we aimed to plications associated with this procedure have been described, al- study the risk factors contributing to morbidity in patients with prior though patient-specific factors that cause these complications are lacking at this time. A recent study has demonstrated an increased Received October 14, 2016, and accepted for publication, after revision December risk of complications in previously augmented patients undergoing 24, 2016. removal of implants with exchange to expanders versus those who From the *Division of Plastic Surgery Residency Program, USF Health, University of undergo implant-sparing mastectomy.9 This lends support to the idea South Florida; and †Division of Breast Oncology, H. Lee Moffitt Cancer Center and Research Institute, Tampa, FL. that reconstructive surgeons should consider leaving previously Conflicts of interest and sources of funding: none declared. placed submuscular implants in place for both aesthetic and morbidity Reprints: Joshua Elston, MD, Division of Plastic Surgery Residency Program, USF reduction purposes. Health, University of South Florida, 2 Tampa General Circle, 7th Floor Tampa, This study examines the largest cohort of patients with prior FL 33606. E-mail: [email protected]. Copyright © 2017 Wolters Kluwer Health, Inc. All rights reserved. submuscular breast augmentation undergoing lumpectomy versus ISSN: 0148-7043/17/7806–S269 mastectomy through an implant-sparing technique with the hopes DOI: 10.1097/SAP.0000000000001039 of identifying risk factors associated with increased complication

Annals of Plastic Surgery • Volume 78, Supplement 5, June 2017 www.annalsplasticsurgery.com S269 Elston et al Annals of Plastic Surgery • Volume 78, Supplement 5, June 2017

analyses were performed using SAS (version 9.4; SAS Institute, TABLE 1. Complications in Patients Undergoing Implant-Sparing Cary, NC). Mastectomy Versus Lumpectomy

Variable OR P RESULTS All complications 5.863 0.022 Complications Adjusted for BMI 6.013 0.025 Patients with prior submuscular augmentation mammaplasty Chemotherapy 5.840 0.026 undergoing implant-sparing mastectomy compared with implant- Smoking 5.857 0.025 sparing lumpectomy had a statistically significant increased odds of – When examining complication rates for patients undergoing implant-sparing complications (Table 1). The OR was 5.863 (95% CI, 1.255 27.378; lumpectomy versus mastectomy, we found, even after adjusting for comorbidities, P = 0.022). The results remained unchanged in favor of lumpectomy a higher rate of complications associated with implant-sparing mastectomy. compared with mastectomy when adjusted for BMI (OR, 6.013; 95% CI, 1.255–28.807; P = 0.025), receipt of chemotherapy (OR, 5.840; 95% CI, 1.231–27.712; P = 0.026), and smoking status (OR, – rates as well as differences in complications between the 2 types of 5.857; 95% CI 1.254 27.362; P = 0.025). Four patients (7.7%) devel- oncologic management. oped late complications in the lumpectomy group with 2 developing capsular contractures, 1 had fat necrosis and 1 needed complex reconstruction because of flattening of the nipple-areolar complex. There METHODS was no difference in recurrence or tumor margins between lumpectomy and mastectomy groups (Table 2). Patient Selection An institutional board review–approved retrospective analysis was performed on patients with prior breast augmentation undergoing CONCLUSIONS lumpectomy (N = 52) and mastectomy (N = 64) for breast cancer. Pa- Given the nature of breast implants remaining in place for a num- tients with previously placed subglandular implants undergoing mastec- ber of years in combination with the increasing number of cosmetic tomy or lumpectomy were excluded. augmentations performed annually, it is not surprising that the largest numbers of patients with prior augmentation will present with new Statistical Analysis breast cancer diagnoses desiring reconstruction (Figs. 1, 2). These pa- Patients were stratified into 2 groups for complication analysis: tients will be faced with a population of surgeons unfamiliar with those having any postoperative complication and those who did not. how to handle their oncologic surgery. Performing a lumpectomy or The difference in recurrence and complication rates according to pa- mastectomy while preserving a previously placed subpectoral breast tients receiving mastectomy versus lumpectomy was assessed using a implant will become a more commonly performed procedure. The core χ2 test and summarized as odds ratio (OR) along with 95% confidence desire is to provide a safe operation, but we should also endeavor to pro- intervals (CIs). Wilcoxon rank sum test with normal approximation was vide the least morbidity as well as superb cosmetic outcomes. used to test difference in BMI between lumpectomy and mastectomy. Implant-sparing lumpectomy or mastectomy is a safe procedure Fisher exact test was used to test the difference in complication, tumor to perform and has the benefits of maintaining the breast mound after margins, and other clinical variables between lumpectomy and mastec- mastectomy, fewer office visits that forego the painful weekly expan- tomy patients. Multiple logistic regression was used to obtain adjusted sion, avoiding the stigma associated with the mastectomy defect, and ORs accounting for smoking status, BMI, chemotherapy, or tumor less strenuous recovery.4,8 Although delaying the reconstruction limits stage. Bonferroni correction was applied to adjust for multiple compar- stress to mastectomy skin flaps, those patients desiring immediate isons. All P values are 2-sided, unless otherwise stated, and considered single-stage reconstruction with incremental size changes of the im- statistically significant for at 0.05 for all comparisons. All statistical plant can be considered. This works particularly well for a patient with

TABLE 2. Recurrence

Lumpectomy Mastectomy Variable Level N (%) N (%) P Smoking Current 10 (19.2) 7 (10.9) 0.455 Past 16 (30.8) 23 (35.9) Nonsmoker 26 (50.0) 34 (53.1) Tumor stage Tis 0 (0.0) 8 (13.1) <0.001 T0–T1 41 (87.2) 33 (54.1) T2–T4 6 (12.8) 20 (32.8) Recurrence No 47 (92.2) 61 (95.3) 0.7 Yes 4 (7.8) 3 (4.7) Tumor margins Close (<2 mm) 11 (22.9) 12 (20.0) 0.32 Negative (≥2 mm) 31 (64.6) 45 (75.0) Positive 6 (12.5) 3 (5.0) When examining recurrence rates or tumor margins between implant-sparing lumpectomy versus mastectomy groups, we found no statistical differences.

FIGURE 1. A, A 47-year-old female with 400 cm3 submuscular saline implants. B, After implant- and nipple-sparing bilateral mastectomies with resection weight of approximately 380 g bilaterally. C, She was ultimately exchanged to 650 cm3 silicone smooth round implants bilaterally.

implants placed in the remote past that desires immediate exchange be- in the setting of retained prosthesis as well as the potential for cos- cause of concerns of the age of the implant. If a patient wants a signif- metic deformity given the small remaining native breast volume. icant size difference, consideration can be given to exchange of the Women with larger breast volumes may find this resultant deformity submuscular implant to a standard 2-stage expander-implant–based re- acceptable; however, the option always exists to use the implant as a construction. However, we feel that the greatest strength of this tech- static expander of the skin envelope that can later be converted to an nique lies in the ability to perform a safe oncologic procedure with an autologous reconstruction. initial implant-sparing procedure, followed by exchange to a larger im- At this time, having an informed discussion with a patient about plant approximately 6 weeks later to achieve the desired size and shape. the risks, benefits, and possible complications is limited by the small Patients undergoing breast cancer surgery with previous aug- sample sizes for many studies examining patients undergoing implant- mentation have been shown to have less than 50% of their overall breast sparing mastectomy or lumpectomy. A new study reveals the difference volume corresponding to native breast tissue.6 Although survival is the in complications between previously augmented and non-augmented same, patients should be counseled on the adverse effects of radiation patients undergoing immediate reconstruction after mastectomy is

FIGURE 2. A, A 43 year-old female with history of 380 cm3 saline submuscular augmentation. B, She ultimately underwent unilateral right mastectomy with a 160-g resection specimen and was later exchanged to 480 cm3 saline implant.

higher in the previously augmented group, although it does not reach capsular contracture, which is in contrast with the published literature statistical significance.7 A prior study demonstrated the risk factors as- and likely is an effect of short follow up at this time. sociated with complications in non-augmented patients undergoing Limitations are present in any scientific study. As mentioned expander-implant–based reconstruction.10 Although it is helpful to ex- above, capsular contracture is a significant concern for any surgeon amine complications in the general population of expander-implant– using implants. This is certainly a shortfall of our article in that any true based reconstruction, this provides a myopic view as the previously long-term complication rates for capsular contracture are unable to be augmented population represents a unique subset that is typically youn- gleaned because of our relatively short follow-up as well as small co- ger with a lower BMI.3,11 horts owing to the relatively new nature of this procedure. Our study demonstrates a higher early complication rate associ- Future studies will aim to further elucidate whether long-term ated with implant-sparing mastectomies versus lumpectomies. This is complications bear out any differences from those described above or not surprising given that the higher complication rates associated stan- with the radiation therapy associated with implant-sparing lumpecto- dard and modified radical mastectomies and breast conservation ther- mies. We will also be considering the unique cohort of genetic carriers apy via lumpectomy. Future studies are warranted to determine if for increased breast cancer risk and the philosophy of performing these hold consistent for long-term complications. breast augmentation in anticipation of planned bilateral prophylactic Capsular contracture remains one of the most concerning com- mastectomies (Fig. 3). plications facing patients undergoing implant-sparing techniques be- Our protocol typically involves preoperative collaboration because it requires reoperation and can cause substantial psychosocial tween the breast and plastic surgeons to agree that the procedure is stress. Previous studies have reported on the risk of developing subse- oncologically feasible as well as for preoperative markings. Patients quent capsular contracture in the setting of retained prostheses with then underwent implant-sparing lumpectomy or mastectomy, and the rates approximately 60% quoted in each series.6,12–14 Two (3.8%) of tissues were allowed to heal for 3 months, at which time the volume our patients undergoing implant-sparing lumpectomy developed late of the removed specimen and previously submuscularly placed implant

FIGURE 3. A, A 27 year-old female BRCA2 carrier in whom we performed submuscular augmentation in anticipation of bilateral prophylactic mastectomies. B, Silicone implants (400 cm3) placed. C, After resection sizes of 490 g (left) and 370 g (right). volume were used to determine an estimated volume for the exchanged aesthetic outcomes, and the attempt of this study was to begin qualifying new implant. and quantifying those factors associated with complications in this pa- The pool of literature that exists on patients with prior augmen- tient population. Understanding these operative risks may help in pa- tation undergoing breast cancer surgery is steadily growing. Most stud- tients' decision-making process with regards to type of oncologic surgery. ies describe all patients with implants and a new diagnosis of breast cancer including those placed in a subglandular position. In our experience, these patients are likely not candidates for an implant-sparing pro- REFERENCES cedure. Our study is the largest to date describing a cohort of previous 1. American Society of Plastic Surgeons. 2015 Report of the 2014 Statistics. Na- submuscular breast augmentation patients undergoing implant-sparing tional Clearinghouse of Plastic Surgery Statistics. Available at: http://www. mastectomy or lumpectomy and the complication rates for them. Preop- plasticsurgery.org/Documents/news-resources/statistics/2014-statistics/plastic- erative discussion of outcomes is critical in patients that demand superb surgery-statsitics-full-report.pdf. Accessed November 1, 2015.

2. National Cancer Institute. SEER Cancer Statistics Factsheets: Female Breast 9. Parabkaharan S, Melody M, Trotta R, et al. Comparison of reconstructive Cancer. Available at: http://seer.cancer.gov/statfacts/html/breast.html. Accessed outcomes in breast cancer patients with preexisting subpectoral implants: November 1, 2015. Implant-sparing mastectomy with delayed implant exchange versus imme- 3. Singh KA, Saunders N, Carlson GW. Immediate breast reconstruction in the pre- diate tissue expander reconstruction. Ann Plast Surg. 2016;76(suppl 4): viously augmented patient. Ann Plast Surg. 2012;68:477–480. S332–S335. 4. Perrone M, Suber J, Smith P, et al. An advantage to submuscular breast augmen- 10. McCarthy CM, Mehrara BJ, Riedel E, et al. Predicting complications following tation: implant-sparing mastectomy in breast cancer patients with previous aug- expander/implant breast reconstruction: an outcomes analysis based on preopera- mentation. San Diego, CA: American Society of Plastic Surgeons Scientific tive clinical risk. Plast Reconstr Surg. 2008;121:1886–1892. Program; 2013. 11. Birdsell DC, Jenkins H, Berkel H. Breast cancer diagnosis and survival in 5. Fisher B, Anderson S, Bryant J, et al. Twenty-year follow-up of a randomized trial women with and without breast implants. Plast Reconstr Surg. 1993;92: comparing total mastectomy, lumpectomy, and lumpectomy plus irradiation for 795–800. the treatment of invasive breast cancer. NEnglJMed. 2002;347:1233–1241. 12. Handel N, Lewinsky B, Silverstein MJ, et al. Conservation therapy for breast 6. Karanas YL, Leong DS, Da Lio A, et al. Surgical treatment of breast cancer in pre- cancer following augmentation mammaplasty. Plast Reconstr Surg. 1991;87: viously augmented patients. Plast Reconstr Surg. 2003;111:1078–1083. 873–878. 7. Roostaeian J, Yoon AP, Rahgozar P, et al. Implant-based immediate breast recon- 13. Handel N, Lewinsky B, Jensen JA, et al. Breast conservation therapy after aug- struction in the previously augmented patient. J Plast Reconstr Aesthet Surg. mentation mammaplasty: is it appropriate? Plast Reconstr Surg.1996;98: 2015;68:e71–e79. 1216–1224. 8. Suber J, Malafa M, Smith P, et al. Prosthetic breast reconstruction after implant- 14. Mark RJ, Zimmerman RP, Greif JM. Capsular contracture after lumpectomy and sparing mastectomy in patients with submuscular implants. Ann Plast Surg. radiation therapy in patients who have undergone uncomplicated bilateral aug- 2011;66:546–550. mentation mammoplasty. Radiology. 1996;200:621–625.

Unilateral Versus Bilateral Breast Reconstruction Is Less Really More? Lauren V.Kuykendall, MD,* Bugra Tugertimur, MS,† Corin Agoris, MS,† Sara Bijan, MS,† Ambuj Kumar, MD, MPH,§ and Deniz Dayicioglu, MD*‡

with a concomitant increase in breast reconstruction.1–4 Since 2000, Objective: Over the recent years, there has been an increase in prophylactic there has been a 35% increase in breast reconstruction procedures with mastectomies with an associated increase in bilateral breast reconstruction. more than 60% of these being bilateral reconstruction.2 In women with We aimed to compare outcomes in terms of patient satisfaction with unilateral unilateral breast cancer, the annual incidence of contralateral breast versus bilateral breast reconstruction after deep inferior epigastric perforator cancer ranges from 0.5% to 0.75%, with a lifetime risk of 15%.4–6 (DIEP) flap and implant-based reconstruction. Factors associated with increased risk of developing contralateral Methods: Patients who underwent breast reconstruction by a single surgeon breast cancer include young age, family history of breast cancer, lobular between July 2011 and July 2015 were surveyed using the independently type histology, multicentric breast cancer, previous chest radiation, and validated BREAST-Q questionnaire. Mean satisfaction scores between patients BRCA1/2 mutations.3,4,6,7 Women choose contralateral prophylactic undergoing unilateral versus bilateral breast reconstruction were compared and mastectomy for many reasons, including fear of recurrence, anxiety stratified based on the type of reconstruction [eg, DIEP flap, tissue expander to about future risk of new cancer, and improvement in breast reconstruction implant (TE/I)]. Groups were further categorized by age (patients <55 years 4,5,8 ≥ outcomes. However, this rise in contralateral prophylactic mastec- and 55 years of age) and body mass index (<24.9 and >24.9). Complications tomy is most prevalent in women with early-stage disease for whom were recorded. long-term cancer survival outcomes are equivalent between breast Results: Of the 308 patients included, 118 (38%) had unilateral reconstruction conservation and mastectomy.8 With comparable oncologic outcomes (42 TE/I and 76 DIEP) and 190 (62%) had bilateral reconstruction (124 TE/I and survival rates, evaluations of patient-reported outcomes are helpful and 66 DIEP). A total of 95 patient surveys were included (31% response rate). in facilitating the decision-making process and optimizing long-term Overall, patients receiving unilateral reconstruction demonstrated increased health and satisfaction.9 Advocates of prophylactic mastectomy report satisfaction with outcome (P = 0.028), psychosocial well-being (P = 0.043), improved sense of well-being because of decreased disease-related and sexual well-being (P = 0.002). Complication rates were similar between anxiety and the potential for improved aesthetic outcomes, psychosocial unilateral and bilateral reconstruction. No significant differences for satisfaction well-being, and overall satisfaction.5,10–14 were found in the TE/I group (N = 58; unilateral, 10; bilateral, 48). Although studies have reported that patients undergoing post- In the DIEP group (N = 37; unilateral, 20; bilateral, 17), those receiving unilateral mastectomy breast reconstruction have improved satisfaction with reconstruction had higher satisfaction with outcome (P =0.013)andsexual outcome, there are limited data regarding satisfaction levels when well-being (P =0.014). unilateral and bilateral reconstruction are compared.10,12,14 The de- Additionally, younger patients (<55 years) were more likely to undergo bilateral mand for more rigorous outcome measures has led to increased studies reconstruction (P = 0.018). Body mass index did not have a significant associa- focused on validated patient-reported outcomes.10 Patient-reported tion with unilateral or bilateral reconstruction. outcome measures provide a means of quantifying the way patients Conclusions: Patients undergoing DIEP flap reconstruction showed higher perceive their health and the impact treatments have on their quality satisfaction with unilateral reconstruction, whereas patients receiving TE/I of life.15 The BREAST-Q questionnaire was developed to provide reconstruction, either unilateral or bilateral, were equally satisified. Additionally, reliable, valid, and clinically meaningful patient-reported outcomes younger women were more likely to undergo bilateral reconstruction, which is data.15,16 This questionnaire has been validated in more than 15,000 pa- consistent with current data trends. When considering surgical options, unilateral tients and has been proven to help study and measure the impact and DIEP flap reconstruction may provide improved outcomes in terms of patient effectiveness of breast surgery from the patient's perspective.15,16 satisfaction when compared with bilateral reconstruction in select patients. Using this tool, we aimed to compare patient satisfaction with unilat- Key Words: breast reconstruction, implant-based reconstruction, eral versus bilateral breast reconstruction after deep inferior epigastric DIEP flap reconstruction, patient satisfaction, perforator (DIEP) flap and implant-based reconstruction. We hypothe- unilateral versus bilateral reconstruction, BREAST-Q, patient reported outcomes sized that differences between unilateral and bilateral reconstruction would exist among the defined groups. (Ann Plast Surg 2017;78: S275–S278)

ver the past decade, there has been an increase in contralateral METHODS prophylactic mastectomies in women with unilateral breast cancer O Women who underwent breast reconstruction by a single surgeon between July 2011 and July 2015 were identified using operative Received October 14, 2016, and accepted for publication, after revision December 6, 2016. case logs. Data were gathered from a retrospective chart review includ- From the *Division of Plastic Surgery, Department of Surgery, Morsani College of ing patient demographics, type of reconstruction, and complications. Medicine, †Morsani College of Medicine, and §Department of Internal Medicine, All women were patients at H. Lee Moffitt Cancer Center and Research Morsani College of Medicine, University of South Florida; and ‡Comprehensive Breast Program, Department of Women's Oncology, H. Lee Moffitt Cancer Institute, a National Comprehensive Cancer Center, under the care of Center, Tampa, FL. the senior author. Patients were categorized based on unilateral versus Conflicts of interest and sources of funding: none declared. bilateral reconstruction. Patients underwent either DIEP free flap or Reprints: Lauren V. Kuykendall, MD, USF Health South Tampa Center, 2 Tampa tissue expander to implant (TE/I) breast reconstruction. In the TE/I General Circle, 7th Floor, Tampa, FL 33606. E-mail: [email protected]. Copyright © 2017 Wolters Kluwer Health, Inc. All rights reserved. group, patients who underwent nipple-sparing mastectomy, delayed ISSN: 0148-7043/17/7806–S275 reconstruction, radiation therapy, additional flap reconstruction, and DOI: 10.1097/SAP.0000000000001030 placement of acellular dermal matrix as well as those with incomplete

Annals of Plastic Surgery • Volume 78, Supplement 5, June 2017 www.annalsplasticsurgery.com S275 Kuykendall et al Annals of Plastic Surgery • Volume 78, Supplement 5, June 2017

data were excluded. In the DIEP flap group, those with incomplete Within the unilateral group, 58% were more than 55 years of age data were excluded. Complications were defined as dehiscence, infection, and 74% had a BMI of more than 24.9. Within the bilateral group, 57% hematoma, full-thickness skin loss, or flap loss requiring reoperation of patients were under 55 years of age and 68% had a BMI of more than within 30 days. 24.9 (Fig. 1). Younger patients were significantly more likely to A total of 308 patients were mailed postoperative satisfaction undergo bilateral reconstruction (OR, 1.812; P = 0.018), whereas surveys using the independently validated BREAST-Q questionnaire BMI had no significant association (OR, 1.359; P = 0.319) (Fig. 2). [BREAST-Q Reconstruction Module (Postoperative) 1.0], and satisfac- Complication rates were similar between the 2 groups (11% tion scores were recorded on a scale of 0 to 100. The mean satisfaction overall) with unilateral reconstruction having a 3% complication scores between patients undergoing unilateral versus bilateral breast rate and bilateral reconstruction having an 8% complication rate reconstruction were compared and stratified based on the type of recon- (P =0.229)(Fig.2). struction (eg, DIEP flap, TE/I). A total of 95 patients completed the BREAST-Q questionnaires Secondarily, we evaluated age and body mass index (BMI) in (31% response rate). Of those patients who responded, 30 were relation to the choice for unilateral versus bilateral reconstruction. unilateral and 65 were bilateral. Fifty-eight patients underwent Age groups were determined by median age of the patient population TE/I reconstruction (10 unilateral and 48 bilateral). Thirty-seven (patients <55 years and ≥55 years of age). Body mass index groups patients underwent DIEP free flap reconstruction (20 unilateral and were determined by national criteria for normal and obese weight 17 bilateral) (Fig. 1). (BMI <24.9 and ≥24.9). Overall, patients receiving unilateral reconstruction demonstrated All analyses were performed using SPSS statistical analysis increased satisfaction with outcome (P = 0.028), psychosocial well-being software (version 22; IBM SPSS Statistics, Inc). Difference in satisfac- (P = 0.043), and sexual well-being (P = 0.002). No significant differences tion scores was assessed using the independent sample t test and sum- for satisfaction were found in the TE/I group. In the DIEP group, those marized as mean difference along with 95% confidence intervals. receiving unilateral reconstruction had higher satisfaction with outcome Association between categorical variables was summarized as odds (P = 0.013) and sexual well-being (P = 0.014) (Fig. 3). ratio (OR) and 95% confidence interval, and difference was assessed using Fisher exact test. To adjust for multiple comparison, Bonferroni DISCUSSION correction was applied. Statistical significant level was set at P < 0.05. In recent years, there has been a growing trend toward contralateral prophylactic mastectomy given its ability to reduce the incidence of RESULTS contralateral breast cancer by more than 90%.7 This increase in bilateral Of the 308 patients initially included, 166 (54%) underwent TE/I mastectomy has led to more women electing for bilateral breast recon- reconstruction and 142 (46%) underwent DIEP flap reconstruction. A struction.12 Postmastectomy breast reconstruction can be a difficult total of 118 (38%) had unilateral reconstruction (42 TE/I and 76 DIEP), decision for women given the wide array of surgical options from im- and 190 (62%) had bilateral reconstruction (124 TE/I and 66 DIEP) (Fig. 1). plant to various autologous-based reconstructions. To better understand

FIGURE 1. Patient population stratified by age, BMI, laterality, and type of reconstruction. Overall, 31% patient survey response rate with comparable rates between unilateral and bilateral reconstruction groups.

FIGURE 2. Patient characteristics: age and BMI in unilateral versus bilateral reconstruction. Younger patients were more likely to undergo bilateral breast reconstruction (P = 0.018). Body mass index showed no significant association (P = 0.319). Complications were comparable (P =0.229). these decisions, there has been a growing body of literature focusing on reconstruction, specifically with surgical outcome and sexual well-being. patient-reported outcomes and satisfaction following mastectomy and These findings reinforce current outcomes-based literature and might be breast reconstruction. Recent data have demonstrated that, although attributable to better ability to recreate relative symmetry and shape of the cancer-related anxiety has improved, overall satisfaction following native breast using autologous tissue. Additionally, bilateral autologous bilateral mastectomy and reconstruction has been less favorable.10,14 reconstruction may not provide adequate volume to replicate missing In this study, we utilized the independently validated BREAST-Q ques- breasts in selected patients with the potential for increased donor site tionnaire to evaluate patient-reported outcomes in terms of satisfaction morbidity.10,14 A recent study found that patients who underwent with unilateral versus bilateral breast reconstruction after DIEP flap and unilateral mastectomy and reconstruction reported preservation of TE/I reconstruction. sensation as a reason for improved satisfaction.5 This may indicate We found that patients undergoing unilateral breast reconstruc- that maintaining the contralateral breast contributes to improved sexual tion compared with bilateral breast reconstruction demonstrated higher well-being and patient satisfaction. satisfaction in multiple categories including surgical outcome, psycho- Patients who underwent TE/I reconstruction demonstrated no social well-being, and sexual well-being. When patients were stratified significant difference in mean satisfaction scores between unilateral and into groups based on type of reconstruction, patients who underwent bilateral reconstruction. These results differ somewhat from previous DIEP flap reconstruction demonstrated higher satisfaction with unilateral reports which found improved aesthetic outcome and patient satisfaction

FIGURE 3. BREAST-Q mean satisfaction scores. Overall, patients receiving unilateral reconstruction demonstrated increased satisfaction with outcome (P = 0.028), psychosocial well-being (P =0.043),andsexualwell-being(P =0.002).

with bilateral implant reconstruction compared to unilateral. This was important in improving quality of care, and further studies will continue attributed to the fact that both breast contour and degree of ptosis are to highlight the factors that lead to improved satisfaction for our patients. symmetrical and the reconstructed breasts age in a similar fashion.17 Notably, many of these early studies lacked validated patient reported REFERENCES outcome measures. 1. Kummerow KL, Du L, Penson DF, et al. Nationwide trends in mastectomy for Of the patients in the TE/I group who responded to the question- early-stage breast cancer. JAMA Surg. 2015;150:9–16. naire, 82% underwent bilateral reconstruction. Patients undergoing bi- 2. American Society of Plastic Surgeons. Plastic Surgery Statistics Report 2015. lateral as compared to unilateral reconstruction were younger and had Available from: http://www.plasticsurgery.org/Documents/news-resources/ higher BMI, although the latter finding did not reach statistical signifi- statistics/2015-statistics/plastic-surgery-statsitics-full-report.pdf. Accessed June cance. A recent study looked closely at patients who opted for bilateral 2, 2016. reconstruction and found them more likely to undergo immediate 3. Jones NB, Wilson J, Kotur L, et al. Contralateral prophylactic mastectomy for unilateral breast cancer: an increasing trend at a single institution. Ann Surg Oncol. reconstruction, have reconstruction at a younger age, and have a 2009;16:2691–2696. prophylactic mastectomy. These patients were also less likely to undergo 4. Tuttle TM, Habermann EB, Grund EH, et al. Increasing use of contralateral pro- radiation and chemotherapy, which likely influences patient motivations phylactic mastectomy for breast cancer patients: a trend toward more aggressive and aesthetic standards.10,14 surgical treatment. JClinOncol. 2007;25:5203–5209. The overall complication rate was 11% and was comparable be- 5. Buchanan PJ, Abdulghani M, Waljee JF, et al. An analysis of the decisions made tween unilateral and bilateral reconstruction (3% vs 8%, P =0.229). for contralateral prophylactic mastectomy and breast reconstruction. Plast – These rates are slightly less than previous reports, which is likely owed Reconstr Surg. 2016;138:29 40. 6. Frost MH, Slezak JM, Tran NV,et al. Satisfaction after contralateral prophylactic to the fact that only major complications requiring return to the mastectomy: the significance of mastectomy type, reconstructive complications, operating room within 30 days were evaluated. and body appearance. JClinOncol. 2005;23:7849–7856. There are a number of limitations to our study. As all surveys 7. McDonnell SK, Schaid DJ, Myers JL, et al. Efficacy of contralateral prophylactic were sent at the same time point, patients responded to the surveys mastectomy in women with a personal and family history of breast cancer. JClin at varying stages of their postoperative course. The surveys were sent Oncol. 2001;19:3938–3943. postoperatively, and patients responded at intervals varying from 8. Hwang ES, Locklear TD, Rushing CN, et al. Patient-reported outcomes after – approximately 6 months to 4 years. Additionally, as patients only choice for contralateral prophylactic mastectomy. J Clin Oncol. 2016;34:1518 1527. 9. Atisha DM, Rushing CN, Samsa GP,et al. A national snapshot of satisfaction with responded once to the survey, there was no ability to assess how indi- breast cancer procedures. Ann Surg Oncol. 2015;22:361–369. vidual patient satisfaction evolved. Recent studies have demonstrated 10. Craft RO, Colakoglu S, Curtis MS, et al. Patient satisfaction in unilateral and bi- that satisfaction with breast reconstruction by procedure type may di- lateral breast reconstruction [outcomes article]. Plast Reconstr Surg. 2011;127: minish over time. Yueh et al14 reported that health-related quality of life 1417–1424. and length of time since surgery significantly affected patient satisfac- 11. Brandberg Y, Sandelin K, Erikson S, et al. Psychological reactions, quality of life, tion, suggesting that older reconstructions are less satisfied than newer and body image after bilateral prophylactic mastectomy in women at high risk rfor breast cancer: a prospective 1-year follow-up study. JClinOncol. 2008;26: reconstructions. As in all survey-based studies, responder bias is inher- 3943–3949. ent as patients at extremes of satisfaction or dissatisfaction are more 12. Cohen O, Small K, Lee C, et al. Is unilateral implant or autologous breast recon- likely to respond to the survey. Furthermore, the study was a retrospec- struction better in obtaining breast symmetry? Breast J. 2016;22:75–82. tive study and recall bias may exist as patients try to remember details of 13. Metcalfe KA, Semple JL, Narod SA. Satisfaction with breast reconstruction in their reconstruction. Finally, the response rate of the survey was 31%, women with bilateral prophylactic mastectomy: a descriptive study. Plast Reconstr which is low compared with other current outcomes based studies. Surg. 2004;114:360–366. Since the data collection, we have started implementing preoperative 14. Yueh JH, Slavin SA, Adesiyun T, et al. Patient satisfaction in postmastectomy breast reconstruction: a comparative evaluation of DIEP, TRAM, latissimus flap, and postoperative BREAST-Q questionnaires for all patients undergo- and implant techniques. Plast Reconstr Surg. 2010;125:1585–1595. ing breast reconstruction that will help with future prospective studies 15. Pusic AL, Lemaine V, Klassen AF, et al. Patient-reported outcome measures in where patients can be surveyed at predetermined times. plastic surgery: use and interpretation in evidence-based medicine. Plast Reconstr Despite the increasing trends in bilateral breast reconstruction, Surg. 2011;127:1361–1367. we found that patients undergoing unilateral breast reconstruction have 16. Pusic AL, Klassen AF, Scott AM, et al. Development of a new patient-reported higher satisfaction in several categories compared to bilateral recon- outcome measure for breast surgery: the BREAST-Q. Plast Reconstr Surg. – struction. In relation to type of reconstruction, we found that women 2009;124:345 353. 17. Cordeiro PG, McCarthy CM. A single surgeon's 12-year experience with tissue with DIEP flap reconstruction have higher satisfaction with unilateral expander/implant breast reconstruction: part II. An analysis of long-term compli- reconstruction whereas patients with TE/I reconstruction were equally cations, aesthetic outcomes, and patient satisfaction. Plast Reconstr Surg.2005; satisfied. Patient-reported outcome measures have become increasingly 118:832–838.

Numerical Correlation of Levator Advancement in Preoperative Planning Valeria Makeeva, BSc,* Sherry S. Collawn, MD, PhD,† Evelina N. Pierce, BA,* Mina S. Mousa, MD,* Jennifer H. Yang, MD,‡ Peter N. Davison, BA,* and Elodie C. Jospitre, BA§ However, swelling and weight of the anterior lamella of the eyelids may Background: Several procedures have been proposed for the treatment of eyelid make intraoperative MRD1 measurements difficult.7 We propose an al- ptosis, and both levator advancement and levator plication are widely used to ternate method correlating differing degrees of levator advancement to shorten the levator palpebrae superioris. The purpose of this study was to quantify change in MRD1 to help guide clinicians in preoperative planning. perioperative lid measurements in patients undergoing bilateral levator aponeurosis advancements to aid in preoperative planning. Methods: Between July 2014 and June 2016, the authors performed a retrospec- SURGICAL TECHNIQUE tive analysis of all bilateral upper eyelid levator advancement procedures for pto- Levator shortening techniques were popularized in the 1970s 20 17 sis performed by the senior surgeon. There are a total of 21 patients (6 men and and 1980s. Jones et al further used the technique of external levator 15 women) with a mean age of 63 years (range, 48–79 years). The average time at superioris aponeurosis repair using local anesthesia in 1975. follow-up was 5.3 months, with a range of 1 to 26 months. At our institution and in our report, all patients underwent bilat- Results: In this retrospective study, we collected data on presurgical mea- eral levator superioris aponeurosis advancement. After prep and drape surements including marginal reflex distance 1 (MRD1), surgical technique in sterile fashion, local anesthesia was injected using 1% xylocaine, used (symmetrical/asymmetrical levator advancement) with millimeters of 2 mL, with 1/100,000 epinephrine. Eye shields were used for globe pro- advancement used, and postsurgical measurements. We found that on average, tection and were applied with topical tetracaine drops. Anesthesia was an advancement of 4 mm led to an improvement in MRD1 of 2.26 mm (n = 14), with monitored anesthesia care. An ellipse of skin was removed approx- and advancement of 5 mm led to an improvement in MRD1 of 2.74 mm imately 7 to 8 mm above the lash line in men and 8 mm in women. A (n = 15). Patients also reported improvements in their quality of life. 1-mm strip of orbicularis was removed, and the levator aponeurosis Conclusions: Our results may be used to guide clinicians in preoperative planning. was then identified with deeper dissection. The tarsal plate was identified, and planned millimeters of advancement were marked on the leva- Key Words: levator advancement, levator plication, blepharoplasty, tor aponeurosis. The levator aponeurosis was then incised over the tarsal marginal reflex distance 1, eyelid ptosis plate, and the distal 1 mm of aponeurosis excised. Three horizontal 5–0 (Ann Plast Surg 2017;78: S279–S281) Nylon mattress sutures were used to attach the upper tarsal plate to the site marked on the aponeurosis. The amount of advancement was mea- arginal reflex distance 1 (MRD1) as shown in Figure 1 is defined sured and recorded for each eyelid. The supratarsal fold was recreated M as the distance between the upper lid margin and the corneal light with three 5–0 Vicryl sutures attaching the orbicularis down to the site reflection at forward gaze with eyebrows remaining fixated to eliminate of levator advancement. their contribution to lid elevation. It is influenced by the levator aponeurosis, sympathetic innervation to the Muller muscle, the skin METHODS 1 and orbicularis mass, and brow position. Marginal reflex distance 1 A retrospective review between July 2014 and June 2016 included 2 was first described by Putterman and Urist in 1975 to standardize up- 42 eyelids of 21 patients (6 men and 15 women) with blepharoptosis per eyelid ptosis measurements. Before the popularization of MRD1, corrected by levator advancement technique performed by the senior sur- clinicians measured the width of the palpebral fissure to determine eye- geon. Patients underwent either 4- or 5-mm advancement, and less com- lid elevation, a measure that was unreliable if lower eyelid level was ab- monly 3-, 6-, 7-, and 8-mm advancement. The advancement distance was 2 normal. Normal MRD1 is classified as greater than 2.5 mm, whereas based on the expected elevation of the eyelid over the pupil. University of 3 mild upper lid ptosis is classified as less than 2 mm. Alabama at Birmingham, Institutional Review Board for Human Use ap- Multiple reports document preoperative and postoperative proval was obtained for review of eyelid ptosis patient records. The av- MRD1 increases after different blepharoptosis procedures including le- erage age of the 21 patients included in this study was 63 years, ranging 4–14 vator advancement. However, postoperative eyelid height is still from47to79years. 8–10,15–19 cited as difficult to predict. Although previous reports address Evaluation of preoperative parameters included age, sex, and intertechnique differences in postoperative eyelid height, they rarely ad- preoperative eyelid measurements including MRD1. Postoperative data, dress the subject of predicting such intratechnique differences in post- such as eyelid measurements including MRD1 and length of follow-up, operative eyelid height. were recorded. Marginal reflex distance 1 measurements were collected Intraoperative MRD1 has been suggested as a tool to investigate using handheld ruler, a method cited as yielding reliable results in the 7 the relationship between preoperative and postoperative eyelid heights. hands of a skilled clinician.1 In both the 4- and 5-mm advancement groups, no intraoperative complications developed. No infections, hema- Received September 1, 2016, and accepted for publication, after revision March 30, 2017. tomas, or other complications were noted postoperatively in either group. From the *University of Alabama at Birmingham School of Medicine; †Division of Plastic Surgery, Department of Surgery, University of Alabama at Birmingham, Statistical analyses comparing preoperative and postoperative Birmingham, AL; ‡Department of Pediatrics, University of California, San Diego, MRD1 measurements using paired t tests were performed. The average La Jolla, CA; and §Morsani School of Medicine, University of South Florida, time at follow-up was 5.3 months, with a range of 1 to 26 months. Tampa, FL. Conflicts of interest and sources of funding: none declared. Reprints: Sherry S. Collawn, MD, PhD, Division of Plastic Surgery, Department RESULTS of Surgery, FOT 1102, 1960 6th Ave S, Birmingham, AL 35233. E-mail: For 4-mm advancement (n = 14), the average preoperative [email protected]. Copyright © 2017 Wolters Kluwer Health, Inc. All rights reserved. MRD1 was 1.2 mm, with a range of 0 to 3.5 mm. The average postop- ISSN: 0148-7043/17/7806–S279 erative MRD1 was 3.57 mm, with a range of 3 to 5 mm. It is important DOI: 10.1097/SAP.0000000000001143 to note that postoperative MRD1 of less than 2 mm is classified as mild

Annals of Plastic Surgery • Volume 78, Supplement 5, June 2017 www.annalsplasticsurgery.com S279 Makeeva et al Annals of Plastic Surgery • Volume 78, Supplement 5, June 2017

advancements, the percent effectiveness was 27%. In conclusion, the percent effectiveness of the procedure was lower in the 5-mm group, as compared with the 4-mm population, and even lower as advancement distance increased, although there are not enough data points in the 6- to 8-mm population to recognize a trend. In summary, our results indicate that there is a statistically significant difference between preoperative and postoperative MRD1 measurements for both the 4-mm and the 5-mm advancements. The average preoperative MRD1 was 1.2 mm for the 4-mm advancement group and was 0.857 mm in the 5-mm advancement group, indicating that, on average, patients with lower preoperative MRD1 values need to undergo the 5-mm over the 4-mm advancement. However, the wide range of preoperative and postoperative MRD1 and low numbers of FIGURE 1. MRD1 illustration. MRD1 is light reflex to lid margin. participating patients make conclusions of postoperative MRD1 improvement from advancement distance more difficult. That withstanding, upper lid ptosis. All patients resulted in postoperative MRD1s greater our numbers still indicate that an advancement of 4 mm resulted in an than 2 mm, indicating that the surgical procedure was successful. Of average eyelid elevation of 2.26 mm and an advancement of 5 mm re- the 4-mm advancement, 9 postoperative MRD1 improvements were be- sulted in eyelid margin elevation of 2.74 mm above the mid pupil. low the 3.57-mm average, ranging from 3 to 3.5 mm, and 6 postoperative MRD1 improvements were above the 3.57 mm average with a range of 4 to 5 mm. There was a statistically significant difference be- DISCUSSION tween preoperative and postoperative MRD1 measurements (P < 0.05), Levator advancement is a commonly performed surgical recon- with results considered statistically significant when P is less than 0.05. structive procedure in the aging plastic surgery patient population. It Preoperative and postoperative measurements on patients who is therefore critical that the surgeon knows how to predict postoperative underwent 4-mm levator advancements are shown in Figure 2. Overall, outcome to avoid overcorrection or undercorrection. As judgment of based on our numbers, an advancement of 4 mm resulted in an average the surgical endpoint becomes extremely difficult when levator function eyelid elevation of 2.26 mm. is compromised intraoperatively, methods to determine the surgical For the 5-mm advancement (n = 15), the average preoperative endpoint using perioperative measurements are necessary.21 The results MRD1 was 0.857 mm, with a range of 0 to 2 mm. The average postop- here suggest that perioperative measurements including MRD1 as well erative MRD1 was 3.6 mm, with a range of 3 to 4.5 mm. All patients as levator advancement distance may be able to aid clinicians in resulted in postoperative MRD1s greater than 2 mm, indicating that predicting postoperative eyelid height. Perioperative MRD1 has been the surgical procedure was successful. Of the 5-mm advancements, 8 shown to have only minimal, clinically acceptable interobserver and postoperative MRD1 improvements were below the 3.6 mm average intraobserver variability among experienced clinicians, and can there- with a range of 3 to 3.5 mm and 6 were above with a range of 4 to fore be accurately used for perioperative planning.22 In conclusion, 4.5 mm. The average postoperative improvement in MRD1 in all pa- based on our numbers from this study, an advancement of 4 mm re- tients was 3.569 mm, with a range of 3 to 5 mm. There was a statisti- sulted in an average eyelid elevation of 2.26 mm and an advancement cally significant difference between preoperative and postoperative of 5 mm resulted in an elevation of the eyelid margin 2.74 mm above MRD1 measurements (P < 0.05), with results considered statistically the mid pupil. These results can be used to plan the amount of advance- significant when P is less than 0.05. Preoperative and postoperative ment required to elevate the eyelid a desired distance above the pupil. measurements on patients who underwent 5-mm levator advancements Future directions for study include further correlating preopera- are shown in Figure 3. Overall, based on our numbers, an advancement tive and postoperative MRD1 with higher powered investigations in- of 5 mm resulted in an average eyelid elevation of 2.74 mm. cluding a greater number of patients as one limitation of this study Perioperative photographs of 2 patients who underwent levator was a small number of patients, making definitive conclusions between advancement are shown in Figure 4. measurements and outcomes difficult. In addition, we calculated the percent effectiveness of the proce- The timing of follow-up can also be a factor in determination of dure, which we defined as follows: the percent ratio of MRD1 improve- postoperative MRD1, because there is generally a lower amount of eye- ment divided by the advancement distance. In the 4-mm patient lid elevation less than 1 month postsurgery as compared with 3 months population the procedure was 67% effective, whereas in the 5-mm pop- or longer. Longer follow-up times can lead to more accurate measure- ulation, it was only 48% effective. In patients who had 6- to 8-mm ments in future studies.23

FIGURE 2. Number of procedures done for 4-mm advancement.

FIGURE 3. Number of procedures done for 5-mm advancement.

FIGURE 4. Perioperative photographs of blepharoplasty procedure done at our institution. The 2 patients shown underwent bilateral upper eyelid levator advancements. The upper patient had 4-mm advancements bilaterally. The lower patient had an advancement of 4 mm on the left and 5 mm on the right.

More studies are necessary to determine the relationship be- 8. Liu D. Ptosis repair by single suture aponeurotic tuck. Surgical technique and tween the levator advancement distance and improvement in MRD1 long-term results. Ophthalmology. 1993;100:251–259. measurement. Although this study may suggest the potential decrease 9. Anderson RL, Dixon RS. Aponeurotic ptosis surgery. Arch Ophthalmol.1979;97: – in effectiveness of the procedure as the advancement distance increases, 1123 1128. 10. McCulley TJ, Kersten RC, Kulwin DR, et al. Outcome and influencing factors of more patient data would be necessary to ascertain this inverse relation- external levator palpebrae superioris aponeurosis advancement for blepharoptosis. ship. In addition, correlation of etiology of blepharoptosis with advance- Ophthal Plast Reconstr Surg. 2003;19:388–393. ment distance and perioperative MRD1 measurements will be necessary. 11. Tucker SM, Verhulst SJ. Stabilization of eyelid height after aponeurotic ptosis Ptosis has a diversity of etiologies including acquired and congenital repair. Ophthalmology. 1999;106:517–522. causes. Other future areas for study will be examining levator advance- 12. Bartley GB, Lowry JC, Hodge DO. Results of levator-advancement blepharoptosis ment and perioperative MRD1 measurements in those patients in whom repair using a standard protocol: effect of epinephrine-induced eyelid position – such measurements may present challenges, such as those with deep change. Trans Am Ophthalmol Soc. 1996;94:165 173. 13. Berlin AJ, Vestal KP. Levator aponeurosis surgery. A retrospective review. eyes, brow ptosis, facial paralysis, or asymmetrical advancements. Ophthalmology. 1989;96:1033–1036. 14. Linberg JV, Vasquez RJ, Chao GM. Aponeurotic ptosis repair under local anesthe- ACKNOWLEDGMENT sia. Prediction of results from operative lid height. Ophthalmology. 1988;95: 1046–1052. The artist’s drawing in Figure 1 was by Matthew K. Collawn. 15. Codner MA. Discussion: a systematic review of comparison of upper eyelid invo- Permission was obtained to use his artwork in this article. lutional ptosis repair techniques: efficacy and complication rates. Plast Reconstr Surg. 2012;129:158–159. 16. McCord CD, Seify H, Codner MA. Transblepharoplasty ptosis repair: three-step REFERENCES technique. PlastReconstrSurg. 2007;120:1037–1044. 1. Nemet AY. Accuracy of marginal reflex distance measurements in eyelid surgery. 17. Jones LT, Quickert MH, Wobig JL. The cure of ptosis by aponeurotic repair. Arch J Craniofac Surg. 2015;26:e569–e571. Ophthalmol. 1975;93:629–634. 2. Putterman AM, Urist MJ. Müller muscle-conjunctiva resection. Technique for 18. Older JJ. Levator aponeurosis surgery for the correction of acquired ptosis: anal- treatment of blepharoptosis. Arch Ophthalmol. 1975;93:619–623. ysis of 113 procedures. Ophthalmology. 1983;90:1056–1059. 3. Chang S, Lehrman C, Itani K, et al. A systematic review of comparison of upper 19. Berris CE. Adjustable sutures for the correction of adult-acquired ptosis. Ophthal eyelid involutional ptosis repair techniques: efficacy and complication rates. Plast Plast Reconstr Surg. 1988;4:171–173. Reconstr Surg. 2012;129:149–157. 20. Carraway JH, Vincent MP.Levator advancement technique for eyelid ptosis. Plast 4. Baroody M, Holds JB, Sakamoto DK, et al. Small incision transcutaneous levator Reconstr Surg. 1986;77:394–403. aponeurotic repair for blepharoptosis. Ann Plast Surg. 2004;52:558–561. 21. Lucarelli MJ, Lemke BN. Small incision external levator repair: technique and 5. Litwin AS, Patel B, McNab AA, et al. Blepharoptosis surgery in patients with my- early results. Am J Ophthalmol. 1999;127:637–644. asthenia gravis. Br J Ophthalmol. 2015;99:899–902. 22. Ahmad SM, Della Rocca RC. Blepharoptosis: evaluation, techniques, and compli- 6. Hwang K, Ko YS. Improvement of the marginal reflex distance-1 in blepharoptosis cations. Facial Plast Surg. 2007;23:203–215. Surgeries. J Craniofac Surg. 2016;27:455–460. 23. Thomas GN, Chan J, Sundar G, et al. Outcomes of levator advancement and 7. Wada Y, Hashimoto T, Kakizaki H, et al. What Is the best way to handle the invo- Müller muscle-conjunctiva resection for the repair of upper eyelid ptosis. Orbit. lutional blepharoptosis repair? J Craniofac Surg.2015;26:e377–e380. 2017;36:39–42.

Dermal Autograft Using Donor Breast as Alternative to Acellular Dermal Matrices in Tissue Expander Breast Reconstruction A Comparative Review William Darden North, MD,* Christopher S. Kubajak, BS,† Brad St. Martin, MPH,† and Brian Rinker, MD*

n 2016, an estimated 246,660 cases of invasive breast cancer will be Background: Shifting preference for implant-based breast reconstruction has re- I diagnosed in the United States, many of which will require therapeu- sulted in an increased use of acellular dermal matrix (ADM) in tissue-expander tic mastectomy.1 In addition, there has been a trend in recent years for breast reconstruction. The benefits afforded by ADM must be weighed against – – women to choose to undergo prophylactic mastectomy of the non a potential increased risk for postoperative complications. Dermal autograft cancer-bearing contralateral breast.2 Breast reconstruction becomes assisted breast reconstruction using autograft harvest from the lower abdomen a vital step in the recovery process for these women, with 41% of has been shown to result in equivalent aesthetics and patient satisfaction com- women electing to have some form of reconstruction. A paradigm pared with ADM at a lower cost, with fewer complications. The purpose of this shift has occurred in the last decade with a larger proportion of study was to review a series of patients who underwent bilateral mastectomy – women choosing immediate implant-based reconstruction over and immediate dermal autograft assisted tissue expander (TE) breast reconstruc- autologous reconstruction.2,3 tion using the non-cancerous breast as a donor site, comparing the outcomes with With the shift toward implant-based reconstruction, the use of a concurrent cohort of patients undergoing ADM-assisted reconstruction to deter- acellular dermal matrix (ADM) in breast reconstruction has grown in mine the relative safety, cost, and effectiveness of the 2 procedures. popularity. In 2015, over half of women receiving implant-based breast Methods: The study population included all patients who underwent dermal 4 – reconstruction received ADM. The ability to create an inferior sling autograft assisted TE breast reconstruction, using the contralateral cancer-free with ADM provides several distinct advantages over prior techniques, breast as the source of dermal autograft, between 2010 and 2015. The ADM co- including improved lower pole coverage, enlargement of the subpectoral hort consisted of patients who underwent bilateral mastectomy and immediate space, better control of the inframammary fold (IMF), and maintenance ADM-assisted TE breast reconstruction during the same period. Univariate anal- of medial implant position.5–12 However, ADM usage has been associ- ysis was performed for demographic data, complications, operative cost, and ated in multiple studies with increased rates of postoperative complica- operative time. Data were compared using the Wilcoxon rank sum test for non- χ2 tions, including seroma, hematoma, infection, skin flap necrosis, and parametric data and analyses for continuous and categorical variables. Signif- implant loss.13–18 icance was defined as P value less than 0.05. The advantages of a lower pole sling are undeniable. Dermal au- Results: Seventeen patients received dermal autograft using the non-cancerous tograft is a readily available, biocompatible, and inexpensive alternative breast donor site. Twenty-seven patients who underwent ADM-assisted recon- to ADM. Dermal autograft–assisted tissue expander (TE) breast reconstruction during the same period were identified. Significantly higher cost was struction using dermal autograft harvested from the lower abdomen has demonstrated between groups (ADM, US $9999.87; autograft, US $3924.19; been previously described. The procedure has been shown to result in P < 0.0001). No significance difference existed operative time (autograft, 97 equivalent aesthetics and patient satisfaction rates as ADM-assisted re- min; ADM, 120 min). No difference was found in wound healing complications construction, at a lower cost and with fewer complications.19–21 (ADM, 14.8%; autograft, 23.53%; P = 0.47). No significant difference was found A larger number of women are opting for bilateral prophylactic in major complications (ADM, 26%; autograft, 17.65%; P=0.52) or infectious mastectomies or prophylactic mastectomy of their contralateral non– complications (ADM, 26%; autograft, 17.65%; P = 0.52). – cancer-bearing breast. Many of these patients suffer from ptosis or Conclusions: Dermal autograft assisted breast reconstruction using the contralat- macromastia, where a skin-reducing mastectomy is incorporated into eral non-cancerous breast as the source of dermal autograft harvest represents a lower the ablative and reconstructive plan. In these cases, the lower pole of cost alternative to ADM without increased risk of postoperative complications. the breast can be a potential donor site for dermal autograft harvest, Key Words: dermal autograft, acellular dermal matrices, AlloDerm, thus, obviating the need for a separate abdominal donor site. tissue expander, breast reconstruction The purpose of this study was to review a series of patients who underwent bilateral mastectomy and immediate dermal autograft– (Ann Plast Surg 2017;78: S282–S285) assisted TE breast reconstruction using the non-cancerous breast as a donor site and to compare the outcomes with a concurrent cohort of patients undergoing ADM-assisted reconstruction, to determine the relative safety, cost, and effectiveness of the 2 procedures. Received October 11, 2016, and accepted for publication, after revision December 24, 2016. From the *University of Kentucky Department of Surgery, Division of Plastic and METHODS Reconstructive Surgery; and †University of Kentucky College of Medicine, Lexington, KY. Presented at the Annual Meeting of the Southeastern Society of Plastic Reconstructive Chart Review Surgeons, June 6, 2016, Lake Buena Vista, FL. Conflicts of interest and sources of funding: No external sources of funding were used The study was reviewed by and met the standards of the spon- to assist in the preparation of this article. soring institutional review board for studies involving human subjects. Reprints: William D. North, MD, Division of Plastic Surgery, University of Kentucky The study population included all patients who underwent dermal College of Medicine, Kentucky Clinic K454, 740 S Limestone, Lexington, KY autograft–assisted TE breast reconstruction, using the contralateral, 40536-0284. E-mail: [email protected]. Copyright © 2017 Wolters Kluwer Health, Inc. All rights reserved. cancer-free breast as the source of dermal autograft, between 2010 ISSN: 0148-7043/17/7806–S282 and 2015, inclusive. All operations were performed at one university DOI: 10.1097/SAP.0000000000001041 medical center by one senior surgeon (B.D.R.). The ADM cohort

S282 www.annalsplasticsurgery.com Annals of Plastic Surgery • Volume 78, Supplement 5, June 2017 Annals of Plastic Surgery • Volume 78, Supplement 5, June 2017 Dermal Autograft Breast Reconstruction

FIGURE 1. Preoperative image demonstrating grade 3 ptosis (left). Markings for harvest of dermal autograft. Preoperative marking of a 23 Â 10-cm horizontally oriented ellipse of skin on the lower pole of the contralateral non-cancerous breast on the skin between the nipple-areolar complex and IMF (middle). The ellipse is de-epithelialized with Metzenbaum scissors in situ (right). consisted of patients who underwent bilateral mastectomy and immedi- previously described techniques.20 For the autograft group, all patients ate ADM-assisted TE breast reconstruction, by the same surgeon dur- had significant ptosis, with the entire nipple-areolar complex positioned ing the same time period. Demographic data, body mass index, inferior to the IMF when standing, and were candidates for a Passot- chemotherapy history, smoking status, and other comorbid factors were pattern horizontal skin-reducing mastectomy.22 In a standing position, recorded. Wound healing complications were recorded for patients the donor site was marked out as a horizontally oriented ellipse of skin experiencing dehiscence, skin flap necrosis, and delayed wound on the lower pole of the contralateral non-cancerous breast, on the skin healing. Infectious complications were recorded for patients experienc- between the nipple-areolar complex and IMF. The horizontal and verti- ing cellulitis, breast abscess, and TE infection. Major complications cal dimensions of the graft were determined by the dimensions of the were defined as any unplanned readmission or return to the operating breast, but the width of the grafts measured between 22 cm and room. Surgical logs were accessed for each patient to determine the to- 30 cm and the height measured between 6 cm and 12 cm in the study tal operative time from incision to closing. Operative cost was calcu- period. The graft harvest began by scoring the perimeter of the graft lated using the product of operative time and the medical center's with a scalpel (Fig 1). The graft was then de-epithelialized with long- average cost per hour of operating time. In patients undergoing recon- handled Metzenbaum scissors and harvested in the subcutaneous plane struction using ADM, the cost per sheet of ADM was added to the op- with a scalpel. The graft was defatted with scissors on a back table and erative cost calculation. split into two halves. The autograft was then sutured internally, superficial side up, to the IMF,lateral mammary fold, and inferior border of the pectoralis major muscle, covering the inferior pole of the TE (Fig 2). Operative Technique Patients were classified into 2 groups: those receiving ADM for lower pole expander coverage (ADM group) and those receiving der- Data Analysis mal autograft from the contralateral non-cancerous breast (autograft SAS Enterprise Guide version 5.1 was used to perform all anal- group). All patients in the ADM group received AlloDerm (LifeCell yses. Univariate analysis was performed for all demographic data, com- Corporation, Branchburg, NJ), with placement in each case using plications, operative cost, and operative time. Data collected were

FIGURE 2. Dermal autograft harvested in the subcutaneous plane with a scalpel (left). The graft is then defatted with scissors. Graft sutured internally, superficial side up to the IMF, lateral mammary fold, and inferior border of the pectoralis major muscle (right).

autograft group. However, this difference was not statistically significant. TABLE 1. Treatment Group Comparative Data No statistically significant difference was observed in wound healing complications between the ADM group (14.8%) and autograft (23.53%, P Autograft Group (%) ADM Group (%) P = 0.47). Neither major complications (26% for ADM vs 17.65% No. patients 17 27 for autograft) nor infectious complications (26% for ADM vs 18% Age, y 53 (46–59) 52 (41–61) 0.84 for autograft, P = 0.52) demonstrated significant difference (Table 2). BMI 32 (28.5–36.6) 25 (22.3–29) 0.01 Obese 10 (58) 6 (22) 0.14 DISCUSSION Smokers 6 (35) 7 (25) 0.51 A paradigm shift has occurred in breast reconstruction in the Received chemotherapy 9 (53) 15 (56) 0.87 United States, with implant-based breast reconstruction being favored over autologous reconstruction. With this shift has occurred a comple- Medical comorbidities 2–4,23 Diabetes 3 (17.6) 3 (11.1) 0.54 mentary increase in ADM-assisted TE breast reconstruction. The benefits of ADM must be weighed against a potential increased risk Hypertension 6 (35.3) 10 (37) 0.91 – of infection, seroma, hematoma, and skin flap necrosis.13 18 As the COPD 3 (17.6) 3 (11.1) 0.54 number of women diagnosed with breast cancer continues to rise, so COPD, chronic obstructive pulmonary disease. have the rates of mastectomy and breast reconstruction. In a climate Continuous variable median values reported with interquartile range. of ever-increasing health-care costs, safe and cost-effective methods Categorical variables reported as n (%). for breast reconstruction should be sought. Dermal autograft harvest from the abdomen in tissue-expander reconstruction has been described previously, demonstrating an im- 19–21 compared using the Wilcoxon rank sum test for nonparametric data and proved complication profile than the use of ADM. The aim of χ2 analyses for continuous and bivariate variables. Significance was de- the present study was to describe a technique using the contralateral fined as P value less than 0.05. non-cancerous breast removed during prophylactic mastectomy as the source for dermal autograft to be used as an alternative to ADM in immediate TE reconstruction. This donor site is available only in RESULTS women undergoing a bilateral procedure and with sufficient ptosis to require a sizeable skin-reducing mastectomy. We avoided using dermis from a known cancer-bearing breast, but using skin from such a Comparison of Treatment Groups breast is probably oncologically sound, if taken from a site distant to Seventeen patients (34 breasts) were identified who underwent the tumor (Fig 3). – dermal autograft assisted TE breast reconstruction using the non- Our results demonstrate that for women with large ptotic breasts cancerous breast donor site during the study period. For historical providing adequate tissue for dermal autograft harvest, operative cost control, 27 patients were identified who underwent ADM-assisted re- can be significantly decreased over the use of ADM, without added op- construction during the same period. In the autograft population ages erative time. This corroborates the findings of a recent published review ranged from 37 to 68 years with a median age of 53 years. In the of the cost-effectiveness of autologous dermal flaps compared with ADM population ages ranged from 26 to 71 years with a median age ADM.24 In addition, rates of wound healing and infectious complica- of 52 years. No statistical significance existed in respect to age between tions showed no difference between the 2 groups, and no difference the 2 populations. No statistical significance was found between the 2 was found in the rate of major complications requiring unplanned re- populations with respect to smoking status, diabetes, hypertension, or admission or return to the operating room. There was no significant pre-existing chemotherapy treatment. Body mass index (BMI) was sig- difference in complications between the 2 groups, despite the fact that nificantly higher in the autograft population (32 ± 9.6) compared with there was a higher average BMI in the autograft group. In previous the allograft population (25.7 ± 5.7, P = 0.01) (Table 1). studies, BMI has been a strong predictor for postoperative complications.17,25,26 A previous study using dermal autograft harvested from Outcomes the abdomen demonstrated improvements in infectious and wound The average cost per patient was significantly higher in the healing complications as well as major complication rates for patients ADM group than in the autograft group (US $9999.87 ± 2587.13 vs when compared with the use of ADM. Although our study was not able US $3924.19 ± 1618.87, P < 0.0001). The average operative time in to demonstrate statistical significance, similar trends were seen for in- the ADM group was 120 minutes, compared with 97 minutes in the fectious and major complications in the dermal autograft group.

TABLE 2. Recorded Outcomes Assessment

Autograft Group (%) ADM Group (%) P No. patients 17 27 Operative cost, US $ 3924.19 (3573–4144) 9999.87 (5686–10681) <0.05 Operative time, min 97 (1.4–1.77) 120 (1.5–2.5) 0.11 Complications Wound healing 4 (23.5) 4 (14.8) 0.47 Infection 3 (17.6) 7 (26) 0.52 Major 3 (17.6) 7 (26) 0.52 Continuous variable median values reported with interquartile range. Categorical variables reported as n (%).

FIGURE 3. Postoperative photographs after completion of expansion and exchange for permanent implant with subsequent nipple-areolar complex reconstruction.

The study is a retrospective cohort review, with the inherent lim- 11. Vardanian AJ, Clayton JL, Roostaeian J, et al. Comparison of implant-based im- itations. Due to the relatively narrow indications for the procedure, the mediate breast reconstruction with and without acellular dermal matrix. Plast – sample size was small. The treatment groups were not randomly Reconstr Surg. 2011;128:403e 410e. 12. Basu CB, Jeffers L. The role of acellular dermal matrices in capsular contracture: a assigned, but rather assigned based on availability of donor tissue and review of the evidence. Plast Reconstr Surg. 2012;130(suppl 2):118S–124S. patient preference. In respect to operative time, investigators were lim- 13. Ho G, Nguyen TJ, Shahabi A, et al. A systematic review and meta-analysis of ited to reviewing data from surgical logs, which used “incision” and complications associated with acellular dermal matrix-assisted breast reconstruc- “closing” for start and end time. This makes timing of various aspects tion. Ann Plast Surg. 2012;68:346–356. of the procedure, such as autograft harvest, difficult to assess accurately, 14. Salzberg CA, Dunavant C, Nocera N. Immediate breast reconstruction using por- as the plastic surgery team and surgical oncology team work in concert cine acellular dermal matrix (Strattice™): long-term outcomes and complications. – at our institution. This potentially accounts for the trend toward de- J Plast Reconstr Aesthet Surg. 2013;66:323 328. 15. Chun YS, Verma K, Rosen H, et al. Implant-based breast reconstruction using creased operative time in the autograft group compared with the acellular dermal matrix and the risk of postoperative complications. Plast ADM group, although no significant difference was found. Reconstr Surg. 2010;125:429–436. 16. Sbitany H, Serletti JM. Acellular dermis-assisted prosthetic breast reconstruction: a systematic and critical review of efficacy and associated morbidity. Plast CONCLUSIONS Reconstr Surg. 2011;128:1162–1169. In this present study, we demonstrated that dermal autograft 17. Lanier ST, Wang ED, Chen JJ, et al. The effect of acellular dermal matrix use on breast reconstruction using the contralateral non-cancerous breast as complication rates in tissue expander/implant breast reconstruction. Ann Plast the dermal autograft donor site represents a lower cost alternative to Surg. 2010;64:674–678. ADM without increased risk of postoperative complications. 18. Kim JY, Davila AA, Persing S, et al. A meta-analysis of human acellular dermis and submuscular tissue expander breast reconstruction. Plast Reconstr Surg. 2012;129:28–41. REFERENCES 19. Rinker B. The use of dermal autograft as an adjunct to breast reconstruction with tissue expanders. Plast Reconstr Surg. 2012;130:1179–1185. 1. American Cancer Society. Cancer Facts & Figures 2016. Available at: http://www. cancer.org/acs/groups/content/@research/documents/document/acspc-047079. 20. Lynch MP, Chung MT, Rinker BD. Dermal autografts as a substitute for acellular pdf. Accessed September 24, 2016. dermal matrices (ADM) in tissue expander breast reconstruction: a prospective comparative study. J Plast Reconstr Aesthet Surg. 2013;66:1534–1542. 2. Albornoz CR, Bach PB, Mehrara BJ, et al. A paradigm shift in U.S. breast reconstruction: increasing implant rates. Plast Reconstr Surg. 2013;131:15–23. 21. Lynch MP, Chung MT, Rinker BD. A comparison of dermal autograft and acellular dermal matrix in tissue expander breast reconstruction: long-term aesthetic out- 3. Cagli B, Segreto F,Santoro S, et al. A paradigm shift in U.S. breast reconstruction: comes and capsular contracture. Ann Plast Surg. 2015;74(suppl 4):S214–S217. part 2. The influence of changing mastectomy patterns on reconstructive rate and method. Plast Reconstr Surg. 2013;132:320–326. 22. Rinker B, Thornton BP.Skin-sparing mastectomy and immediate tissue expander breast reconstruction in patients with macromastia using the Passot breast reduc- 4. American Society of Plastic Surgeons. Plastic Surgery Statistics Report 2015. tion pattern. Ann Plast Surg. 2014;72:S158–S164. Available at: https://www.plasticsurgery.org/news/plastic-surgery-statistics#section- title. Accessed September 27, 2016. 23. American Society of Plastic Surgeons. Plastic Surgery Statistics Report 2014. Available at: https://www.plasticsurgery.org/news/plastic-surgery-statistics#section- 5. Spear SL, Parikh PM, Reisin E, et al. Acellular dermis-assisted breast reconstruc- title. Accessed September 27, 2016. tion. Aesthetic Plast Surg. 2008;32:418–425. 24. Krishnan NM, Chatterjee A, Van Vliet MM, et al. A comparison of acellular der- 6. Breuing KH, Warren SM. Immediate bilateral breast reconstruction with implants mal matrix to autologous dermal flaps in single-stage, implant-based immediate and inferolateral AlloDerm slings. Ann Plast Surg. 2005;55:232–239. breast reconstruction: a cost-effectiveness analysis. J Plast Reconstr Aesthet Surg. 7. Salzberg CA. Nonexpansive immediate breast reconstruction using human acellu- 2014;67:468–476. lar tissue matrix graft (AlloDerm). Ann Plast Surg.2006;57:1–5. 25. Antony AK, McCarthy CM, Cordeiro PG, et al. Acellular human dermis implan- 8. Zienowicz RJ, Karacaoglu E. Implant-based breast reconstruction with allograft. tation in 153 immediate two-stage tissue expander breast reconstructions: deter- Plast Reconstr Surg. 2007;120:373–381. mining the incidence and significant predictors of complications. Plast Reconstr 9. Gamboa-Bobadilla GM. Implant breast reconstruction using acellular dermal ma- Surg. 2010;125:1606–1614. trix. Ann Plast Surg.2006;56:22–25. 26. Liu AS, Kao HK, Reish RG, et al. Postoperative complications in prosthesis-based 10. Kim JY, Connor CM. Focus on technique: two-stage implant-based breast recon- breast reconstruction using acellular dermal matrix. Plast Reconstr Surg.2011; struction. Plast Reconstr Surg. 2012;130(suppl 2):104S–115S. 127:1755–1762.

The Horseshoe Labiaplasty Problems and Pearls Adam J. Oppenheimer, MD, FACS

epinephrine. Once this small infiltration has been achieved, a penetrat- Abstract: Aesthetic vulvovaginal surgery has become the final frontier of cos- ing towel clip is used to grasp the tissues, and a pinch is performed to metic surgery. Accordingly, there has been an increased interest in plastic surgical mark the amount of wedge resection. Markings are joined on the inner procedures for the female genitalia. Labiaplasty is the most common among such labial surface and extended cephalad in the sulcus between majora and procedures, and the number of practitioners and techniques has varied widely in minora, towards the clitoral hood. Here, a band of hood tissue can be the medical literature. This article will focus on the problems and pearls encoun- included in the markings as needed based on the clitoral hood mor- tered by the author in treating enlargement of the labia minora and clitoral hood. phology, thus, completing the horseshoe. These areas are infiltrated, The content will focus on the technical aspects of the labiaplasty surgery and and after allowing for epinephrine effect, they are demucosalized only, ways to avoid common complications encountered during this procedure. maintaining the submucosa layer in its entirety as much as possible. Key Words: labiaplasty, clitoral hood reduction, labia minora, clitoris, Once hemostasis is achieved, closure is performed with a key “U” clitoral hood, aesthetic vaginal surgery, vaginal rejuvenation, female genitalia, stitch at the labial margin using a 3-0 vicryl suture (Ethicon/Johnson aesthetic vulvovaginal surgery & Johnson, Somerville, NJ), followed by interrupted, buried, submu- cosal 4-0 vicryl suture in 2 layers, on the medial and lateral aspects of (Ann Plast Surg 2017;78: S286–S288) the wedge resection. An additional 5-0 monofilament polyglycolic acid esthetic vulvovaginal surgery represents the “final frontier” of aes- suture (Monocryl, Ethicon/Johnson & Johnson, Somerville, NJ) is run A thetic surgery.1 There has been a growing interest in cosmetic pro- in subcuticular fashion on each side, beginning at the clitoral hood and cedures on the female genitalia, with 44% and 16% increases year over ending in a through-and-through horizontal mattress suture at the labia year in the number of labiaplasties performed since 2013.2 This interest minora edge. Antibiotic ointment and cold pack are then applied with is not limited to patients: a broad range of subspecialists have been an absorbable pad and mesh undergarment. drawn to the labiaplasty trend.3 Accordingly and with the adoption of any new procedure, there have been varied results following the use RESULTS of 2 divergent labiaplasty techniques. This review aims to address the There are several problems that may be encountered in the problems and pearls associated with labiaplasty surgery using modifi- labiaplasty procedure. One of the technical variety is in aligning the cations of the extended wedge technique, presenting the author's pre- wedge after tissue resection. The ideal wedge location starts superiorly ferred approach for managing enlarged labia minora and prominence at the junction of the clitoral frenulum with the labia minora. This key of the clitoral hood. point must be aligned with the lower wedge margin, which is more easily identified. This process is akin to aligning the white roll in cleft lip MATERIALS AND METHODS surgery, and ink markings or tattoos may be implemented to ensure Over a 3-year period, the author's technique evolved, combining alignment. The size of excision is similarly a challenge and is best mea- the extended wedge approach defined by Dr Alter4,5 and the manage- sured with a pinch maneuver: when tension on the posterior fourchette ment of the clitoral hood described by Dr Hamori.6 A unified technique for a total vulvoplasty was defined based on contiguous circumferential recontouring of the labia minora and clitoral hood.7 This approach, termed the horseshoe labiaplasty because of the pattern of tissue resection (Video 1, http://links.lww.com/SAP/A225), allows for the management of both commonly associated anatomical findings of the hypertrophic female genitalia. The labiaplasty procedure can be performed under a wide range of anesthetic options, from local to general anesthesia. Most cases are performed under intravenous sedation by an anesthesiologist. The first step is anesthetizing the most distal tip of the labia minora with a local anesthetic. The author's preference is to use .25% bupivacaine with

Received October 30, 2016, and accepted for publication, after revision December 26, 2016. From the Oppenheimer Plastic Surgery, Orlando, FL. Conflicts of interest and sources of funding: No financial support was received in the production of this article. Reprints: Adam J. Oppenheimer, MD, FACS, Oppenheimer Plastic Surgery, 3872 Oakwater Circle, Orlando, FL 32806. E-mail: [email protected]. Supplemental digital content is available for this article. Direct URL citations appear in the printed text and are provided in the HTML and PDF versions of this article on the journal's Web site (www.annalsplasticsurgery.com). Copyright © 2017 Wolters Kluwer Health, Inc. All rights reserved. FIGURE 1. Preoperative appearance in lithotomy position of ISSN: 0148-7043/17/7806–S286 enlarged labia minora with labia reduplication (folds) and DOI: 10.1097/SAP.0000000000001028 minimal clitoral hood excess.

S286 www.annalsplasticsurgery.com Annals of Plastic Surgery • Volume 78, Supplement 5, June 2017 Annals of Plastic Surgery • Volume 78, Supplement 5, June 2017 The Horseshoe Labiaplasty

FIGURE 2. Standing view of same patient. FIGURE 4. Postoperative appearance standing, demonstrating “tucked in” look of the labia minora inside of the majora as desired by most patients. is visualized, the surgeon can be made aware that the planned resection may be too great. The management of the clitoral hood is an additional problem, striking, intimidating to the patient, and quite severe. Bed rest is recom- mostly diagnostic and not technical in nature. When clitoral hood en- mended for a period of 2 to 3 days postoperatively. largement is identified, a simple extension of each wedge across the Admittedly, the horseshoe extension across the hood inevitably hood can be performed to connect the 2 resection specimens. Because interrupts some lymphatics, resulting in increased swelling and, there- the procedure is a demucosalization/deepithelialization technique, the fore, tension on the wedge closure. One option in cases with severe risk of nerve injury is as much as could be encountered in the harvest hood redundancy is the “Edge-Wedge” approach as described by Dr of a skin graft. Lara Devgan, MD, FACS7 (written communication, December The most commonly encountered postoperative complications 2016). This additional modification allows for a wedge or ellipse (or involving the extended wedge and horseshoe labiaplasty techniques even connected “horseshoe”) to be taken from the hood and extending are dehiscence at the labial margin. A variety of factors contribute to to the labia minora margin, while then performing an edge resection this, namely, friability of the delicate labial edge, moisture and friction, from the labia minora proper, extending from clitoral frenulum to distal blood supply created by a wedge approach, yeast infections (no the posterior fourchette. This resection pattern comes to a 3-point antibiotics are given pre- or perioperatively), repetitive trauma, over- “T” on each side, much like the intersecting ellipses of a Wise pattern aggressive wedge excision, and postoperative swelling. The use of the breast reduction. The risk of dehiscence is all but eliminated in this ap- bidirectional barbed suture has drastically reduced the frequency of proach, although the scar burden is greater, extending along the entire edge separation when compared to interrupted suture techniques used labial edge. previously. The author has also implemented a vigorous regimen of ice and elevation postoperatively to manage edema, which can be CONCLUSIONS The anatomical variation in the female genitalia is extraordinary, and therefore, the ease of a simple surgical approach is misleading. Much like the seemingly straightforward procedure of implant-based breast augmentation, individual anatomic differences—both between and within patients—largely dictate the outcome of a labiaplasty. It is paramount that the correct diagnosis is first made, namely, addressing the 2 related but distinct anatomical findings: labia minora hypertrophy and clitoral hood excess. Although the edge labiaplasty is more straightforward, the most common pitfall is leaving an untreated clitoral hood, when present.6 It is the author's opinion that, although a edge labiaplasty may be suitable for some patients, an extended wedge labiaplasty with or without a “horseshoe” modification is suitable for all patients.7 Using the approach outlined above and the pearls described, many of the problems encountered during a labiaplasty procedure can be avoided, yielding high patient satisfaction and low complication rates, ultimately achieving the goal of the labiaplasty procedure: no labia minora or clitoral hood “show” on frontal view of the female genitalia (Figs. 1–4).

REFERENCES FIGURE 3. Postoperative appearance in lithotomy. 1. Mann D. Cosmetic surgery's final frontier. Plast Surg Pract. 2015.

2. Cosmetic Surgery National Data Bank Statistics. The American Society for Aes- 5. Alter GJ. Aesthetic labia minora and clitoral hood reduction using extended central thetic Plastic Surgery. 2015. wedge resection. Plast Reconstr Surg. 2008;122:1780–1789. 3. Hamori CA. Teen labiaplasty: a response to the May 2016 American College of 6. Hamori CA. Postoperative clitoral hood deformity after labiaplasty. Aesthet Surg J. Obstetricians and Gynecologists (ACOG) recommendations on labiaplasty in ado- 2013;33:1030–1036. lescents. Aesthet Surg J. 2016;36:807–809. 7. Oppenheimer A. Contiguous, circumferential recontouring of the labia minora 4. Alter GJ. A new technique for aesthetic labia minora reduction. Ann Plast Surg. and clitoral hood. Paper presented at: 2nd Annual ASPS Aesthetica Meeting; 1998;40:287–290. June 2016.

Single Institution Review of Patients With Prior Breast Augmentation Undergoing Breast Conservation Therapy for Breast Cancer Sangeetha Prabhakaran, MD,* Joshua B. Elston, MD,† Amina Lleshi, MPH,* Ambuj Kumar, MD,† Weihong Sun, MD,* Nazanin Khakpour, MD,* and Deniz Dayicioglu, MD*†

radiation therapy. The consideration of lumpectomy in women with Background: Increasing number of patients with preexisting breast implants de- prior breast augmentation is an increasingly popular option; however, sire breast conservation therapy for breast cancer. There is paucity of comparative this carries risk of radiation-induced contractures requiring reoperation, data on tumor margins and re-excisions in these patients. High re-excision rates reported in some series as high as 65%.2 up to 25% have been reported in breast conservation therapy patients; efforts to Women who have undergone implant placement, frequently also obtain cosmesis and avoid implant rupture might increase this further. We ana- have small breast volumes, with resultant concerns about the ability to lyzed tumor margins, re-excision rates, and recurrence in previously augmented achieve negative margins after BCT. However, retrospective reviews versus non-augmented patients undergoing lumpectomy for breast cancer. We have shown that BCT after implant breast augmentation achieved ac- preserved preexisting implants if feasible with oncologic clearance and cosmesis. – ceptable negative pathologic margins and, inclusive of radiotherapy, Methods: Institutional review board approved retrospective analysis was per- produced acceptable cosmetic results in about two thirds of the pa- formed on patients undergoing lumpectomy with history of prior breast augmen- tients.3 In a series of 12 patients with prior breast augmentation who de- tation (N = 52) and consecutively selected non-augmented patients (N = 51). veloped breast cancer, 6 patients underwent BCT; negative pathologic Based on tumor distance to inked margin, we grouped margins as negative ≥ margins were obtained in all patients, 1 patient had local recurrence ( 2 mm), close (<2 mm), and positive. Patients were followed up clinically and and 1 patient developed distant recurrence after initial surgery.4 with imaging in the outpatient clinic, and recurrences were documented. In another series, among 20 women with prior augmentation Results: Patients in the non-augmented group were significantly more likely to who were treated with BCT, 3 had microscopically positive margins. have larger tumors (T2 and above; P = 0.05) compared with the augmented No local recurrences were noted after a median follow-up of 3.9 years.5 group. Although more patients in the augmented group had positive margins, this High re-excision rates up to 25% have been reported in non- was not statistically significant (6 vs 3, P = 0.86). No difference was noted be- augmented patients undergoing BCT for breast cancer, and efforts to tween re-excision rates among the augmented versus non-augmented groups obtain cosmesis and avoid implant rupture in the presence of preexisting (21.1% vs 19.6%, respectively; odds ratio, 0.91; 95% confidence interval, 6,7 – implants can potentially increase this number. The current evidence 0.35 2.37; P = 0.85); these remained unchanged even when adjusting for tumor regarding tumor margins and re-excision rates in patients with prior stage (P =.75)andmargins(P = 0.73). Although more patients in the augmented breast augmentation is scant, mostly consisting of retrospective reviews group recurred (4 vs 0), this was not statistically significant (P = 0.1). and case series. Conclusions: Our results indicate that, from the oncological standpoint, patients We wanted to review our single institutional experience of BCT with prior augmentation can undergo lumpectomy with equivalent tumor margins in patients with prior breast augmentation and compare them with non- and re-excision rates. To the best of our knowledge, this is the first reported com- augmented patients with specific focus on tumor margins, re-excision parative study between these 2 groups. rates, and recurrence. Key Words: prior breast augmentation, breast conservation therapy, lumpectomy, breast cancer, augmentation mammoplasty, implants METHODS (Ann Plast Surg 2017;78: S289–S291) Institutional review board approval was obtained. We performed a retrospective analysis on 52 patients with prior breast augmentation reast augmentation is the most prevalent surgical cosmetic proce- and 51 consecutive non-augmented patients who underwent BCT for B dure performed in the United States with close to 280,000 cases breast cancer at Moffitt Cancer Center. We analyzed tumor margins performed in the year 2015.1 With the increasing use of this procedure, and re-excision rates in these patients. Based on tumor distance to inked it could be postulated that the number of women with breast cancer with margin, we grouped margins as negative when margins were greater history of prior augmentation is also likely to increase. These patients than 2 mm, close (margins <2 mm), and positive (if tumor was present form a unique subset with several options available for surgical treat- at the inked margin). Patients were followed up clinically and with im- ment of their breast cancer. Breast conservation therapy (BCT) includes aging in the outpatient clinic, and incidence of tumor recurrence was noted. breast-conserving surgery or lumpectomy followed by moderate-dose Among patients with prior augmentation, the majority [33 (63.5%)] had submuscular implants and 18 (34.6%) had subglandular implants. Received October 14, 2016, and accepted for publication, after revision December 24, We preferred to preserve preexisting implants if feasible with oncologic 2016. clearance and cosmesis. The association between the dependent and in- From the *H. Lee Moffitt Cancer Center and Research Institute; and †Division of dependent variables was assessed using the Fisher exact test. For all as- Plastic Surgery, Department of Surgery, Morsani College of Medicine, University of South Florida, Tampa, FL sociations, the results were summarized as odds ratio (OR) and 95% Conflicts of interest and sources of funding: none declared. confidence intervals (CIs). Statistical significance was set at 5% for all Reprints: Deniz Dayicioglu, MD, Division of Plastic Surgery, Department of Surgery, comparisons. To adjust for multiple comparisons, Bonferroni correction Morsani College of Medicine, University of South Florida, 2 Tampa General was applied. Statistical analyses were performed using SPSS v22 software. Circle, 7th Floor, Tampa, FL 33606. E-mail: [email protected]. Presented at the 59th Annual Scientific Meeting of the Southeastern Society of Plastic and Reconstructive Surgeons, June 11–15, 2016 in Lake Buena Vista, Florida. RESULTS Copyright © 2017 Wolters Kluwer Health, Inc. All rights reserved. ISSN: 0148-7043/17/7806–S289 Fifty-two patients with prior breast augmentation who underwent DOI: 10.1097/SAP.0000000000001040 lumpectomy for breast cancer between December 2002 and November

Annals of Plastic Surgery • Volume 78, Supplement 5, June 2017 www.annalsplasticsurgery.com S289 Prabhakaran et al Annals of Plastic Surgery • Volume 78, Supplement 5, June 2017

TABLE 1. Comparison of Tumor Stage, Margins, and Re-excision Rates After Breast Conservative Surgery Between Patients With Prior Breast Augmentation and Non-augmented Patients

Variable Non-augmented group (N = 51), N (%) Augmented group (N = 52), N (%) P Tumor stage Tis (in-situ) 5 (9.8) 5 (9.6) 0.05 T1 28 (54.9) 36 (69.2) T2 and higher 18 (35.3) 6 (11.5) Margins Negative 41 (80.4) 34 (65.4) 0.86 Close 7 (13.7) 9 (17.3) Positive 3 (5.9) 6 (11.5) Re-excision Yes 10 (19.6) 11 (21.2) 0.85 No 41 (80.4) 41 (78.8)

2014 were compared with 51 consecutive non-augmented patients who There were 4 recurrences (7.7%) in the prior breast augmentation underwent lumpectomy between June 2010 and March 2011. group. Three patients had locoregional recurrences, and 1 patient had Median age of patients in the prior breast augmentation group at both local and systemic sites of recurrence. Two patients were treated time of lumpectomy was 56.9 years (range, 34.7–72.7 years) and, in the with mastectomy, 1 with axillary dissection and 1 with systemic ther- non-augmented group, was 62.1 years (26.7–84.8 years). In the prior apy. None of these recurrences were noted in patients who had their breast augmentation group, 36 (69.2%) had T1 stage versus 28 implants removed (Table 2). Because the rate of recurrence in the (54.9%) in the non-augmented group. More patients in the non- non-augmented group was zero, we could not calculate a summary OR augmented group had T2 stage or higher compared with the prior breast and 95% CI. However, the difference was statistically nonsignificant augmentation group [18 (35.3%) vs 6 (11.5%), P = 0.05; Table 1]. The (P = 0.1; Table 3). predominant tumor histology in the prior breast augmentation group was invasive ductal carcinoma [44 (84.6%)]. DISCUSSION Although more prior breast augmentation patients had positive margins, 6 (11.5%) compared with 3 (5.9%) of the non-augmented Breast conservation surgery is certainly feasible in patients with group, this was not statistically significant (P = 0.86). There was thus prior breast augmentation with preservation of implant. Figure 1 shows no statistically significant overall difference in distribution of margins a patient who had prior breast augmentation and had 2 areas success- between these 2 groups (Table 1). fully excised using wire localization. One of these target areas was pos- We compared re-excision rates between these groups. Eleven terior to the implant and could not be visualized but was still successfully (21.2%) of patients in the prior breast augmentation group and 10 excised after wire localization. Figure 2 shows a patient with prior aug- (19.6%) in the non-augmented group underwent re-excision of margins, mentation who developed a local recurrence after BCT and was subse- but this was not statistically significant (P = 0.85; Table 1). The re- quently treated with mastectomy. Our complication rate after BCT in excision rate of 21.2%, in our prior breast augmentation group, was the patients with prior breast augmentation was 3.8% (2/52), which re- comparable to the currently reported re-excision rates in patients under- flects the safety of this procedure. going lumpectomy for breast cancer. The OR of re-excision was 0.91 There is paucity of data in the literature on tumor margins, re- (95% CI, 0.35–2.37; P = 0.85). These results remained unchanged excisions, and recurrence rates in patients with preexisting implants when adjusted for tumor stage (OR, 0.85; 95% CI, 0.31–2.33; who undergo BCT. We sought to study these variables in our patient P = 0.75) and margins (OR, 0.84; 95% CI, 0.32–2.21; P = 0.73). population and compare them to non-augmented patients. Our results Two patients in the prior breast augmentation group developed do not show any statistically significant differences in tumor margins, complications. One had postoperative wound infection, and the other re-excision rates or recurrence between the 2 groups. Non-augmented developed a systemic allergic reaction requiring readmission. patients have statistically significant larger tumors compared with Six patients (11.5%) had implants removed in the lumpectomy the augmented group. The reason for this might be that patients with group [2 prior to lumpectomy, 3 during lumpectomy (including 1 patient larger tumors might be preferably treated with mastectomy for requested), and 1 prior to radiation]. Four of these implant removals were oncologic clearance. performed in patients with preexisting subglandular implants and 2 in There were higher recurrences noted in the augmented group patients with submuscular implants. compared with the non-augmented group (4 vs 0), although this was Median period of follow-up in the prior breast augmentation not statistically significant. This might be owing to insufficient power group was 100.3 months (26.1–418.5 mo) and, in the non-augmented due to smaller sample size, and we plan to continue this study in larger group, was 141 months (0.3–170.4 mo). groups of patients. It has also been noted that patients seeking cosmetic breast augmentation may have greater aesthetic expectations.8 No definitive

TABLE 2. Type of Recurrence Noted in Patients in the Prior Breast TABLE 3. Comparison of Tumor Recurrence Rates After Augmentation Group who Recurred and Treatment Performed Breast Conservative Surgery Between Patients With Prior Breast Augmentation and Non-augmented Patients Patient No. Type of Recurrence Treatment of Recurrence 1 Local recurrence Mastectomy Non-augmented group Augmented group Variable (N = 51), N (%) (N = 52), N (%) P 2 Local + distant recurrence Systemic therapy 3 Ipsilateral axilla Axillary dissection Recurrence No 51 (100) 47 (90.4) 0.1 4 Local recurrence Mastectomy Yes 0 (0) 4 (7.7)

S290 www.annalsplasticsurgery.com © 2017 Wolters Kluwer Health, Inc. All rights reserved. Annals of Plastic Surgery • Volume 78, Supplement 5, June 2017 Lumpectomy in Patients With Prior Breast Augmentation

FIGURE 1. Postprocedure mammogram of a patient with prior breast augmentation. A, Postprocedure mammogram (right mediolateral oblique view). B, Post procedure mammogram (right craniocaudal view). These images represent the feasibility of performing lumpectomy in patients with prior breast augmentation. This patient had prior breast augmentation and had 2 areas successfully excised with wire localization. One of these target areas is posterior to the implant but was still successfully excised after wire localization.

FIGURE 2. Photographs of a patient with prior breast augmentation treated with breast conservative surgery subsequently treated with mastectomy for local recurrence. predictive factors for unfavorable cosmetic outcome have been identi- REFERENCES fied in the literature because of the small numbers of patients analyzed 9 1. American Society of Plastic Surgeons. National Clearinghouse of Plastic Surgery in individual studies. Procedural Statistics 2015. Available at: https://www.plasticsurgery.org/news/ Our study has some inherent disadvantages. It is a retrospective plastic-surgery-statistics. Accessed October 10, 2016. study with smaller number of patients and a relatively shorter period of 2. Handel N, Lewinsky B, Jensen JA, et al. Breast conservation therapy after augmen- follow-up. The impact of radiation causing problems such as capsular tation mammaplasty: is it appropriate? PlastReconstrSurg. 1996;98:1216–1224. contracture has not been studied at this time, and we will monitor for 3. Gray RJ, Forstner-Barthell AW, Pockaj BA, et al. Breast-conserving therapy and sentinel lymph node biopsy are feasible in cancer patients with previous implant this during long-term follow-up of these patients. breast augmentation. Am J Surg. 2004;188:122–125. 4. Tuli R, Flynn RA, Brill KL, et al. Diagnosis, treatment, and management of breast cancer in previously augmented women. Breast J. 2006;12:343–348. 5. Guenther MJ, Tokita KM, Guiliano AE. Breast conserving surgery and radiation CONCLUSIONS after augmentation mammoplasty. Cancer. 1994;73:2613–2618. Use of BCT for management of breast cancer in patients with 6. McCahill LE, Privette A, James T, et al. Quality measures for breast cancer surgery: prior breast augmentation is safe and feasible. From the oncological initial validation of feasibility and assessment of variation among surgeons. Arch – standpoint, equivalent tumor margins and re-excision rates can be achieved Surg. 2009;144:455 462. 7. McCahill LE, Single RM, Aiello Bowles EJ, et al. Variability in reexcision follow- in patients with preexisting implants undergoing BCT for breast cancer ing breast conservation surgery. JAMA. 2012;307:467–475. with implant preservation. Larger studies need to be performed to fur- 8. Birtchnell S, Whitfield P, Lacey JH. Motivational factors in women requesting ther confirm these results. Future efforts should also focus on under- augmentation and reduction mammaplasty. J Psychosom Res. 1990;34:509–514. standing the long-term effects of radiation therapy in patients with 9. Karanas YL, Leong DSM, Da Lio A, et al. Surgical treatment of breast cancer in prior augmentation undergoing BCT. previously augmented patients. Plast Reconstr Surg. 2003;111:1078–1083.

Use of Processed Nerve Allografts to Repair Nerve Injuries Greater Than 25 mm in the Hand Brian Rinker, MD, FACS,* Jozef Zoldos, MD,† Renata V.Weber, MD,‡ Jason Ko, MD,§ Wesley Thayer, MD, PhD,|| Jeffrey Greenberg, MD,¶ Fraser J. Leversedge, MD,# Bauback Safa, MD,** and Gregory Buncke, MD**

igital nerve injuries are among the most common conditions Abstract: Processed nerve allografts (PNAs) have been demonstrated to have D treated by hand surgeons. The conditions for a successful nerve improved clinical results compared with hollow conduits for reconstruction of repair are well known and consist of generous trimming of the nerve un- digital nerve gaps less than 25 mm; however, the use of PNAs for longer gaps til healthy substance is observed, minimal handling, a tension-free co- warrants further clinical investigation. Long nerve gaps have been traditionally aptation, and maintaining a well-vascularized wound bed.1 In many hard to study because of low incidence. The advent of the RANGER registry, a – cases, however, the condition of a tension-free repair cannot be met. large, institutional review board approved, active database for PNA This occurs frequently in cases with a crushing or avulsing mechanism, (Avance Nerve Graft; AxoGen, Inc, Alachua, FL) has allowed evaluation of with a wide zone of nerve injury. In these cases, the options include lower incidence subsets. The RANGER database was queried for digital nerve re- repairing under tension; positioning in flexion; or bridging the gap with pairs of 25 mm or greater. Demographics, injury, treatment, and functional out- a nerve autograft, processed nerve allograft, or conduit. Tension has comes were recorded on standardized forms. Patients younger than 18 and been shown to inhibit axonal outgrowth and Schwann cell activity,2 those lacking quantitative follow-up data were excluded. Recovery was graded likely through its effect on nerve microcirculation.3,4 Positioning of according to the Medical Research Council Classification for sensory function, the digits in flexion can lead to flexion contractures or can interfere with with meaningful recovery defined as S3 or greater level. Fifty digital nerve inju- postoperative rehabilitation protocols. ries in 28 subjects were included. There were 22 male and 6 female subjects, and Beginning in the 1940s, nerve autografts have been the mainstay the mean age was 45. Three patients gave a previous history of diabetes, and there of nerve gap repair. Autografts are readily available, inexpensive, and were 6 active smokers. The most commonly reported mechanisms of injury were biocompatible.5 However, they add time and operative complexity to saw injuries (n = 13), crushing injuries (n = 9), resection of neuroma (n = 9), the reconstruction and are associated with donor site sensory loss, amputation/avulsions (n = 8), sharp lacerations (n = 7), and blast/gunshots which in some cases can be longstanding and debilitating.6,7 To avoid (n = 4). The average gap length was 35 ± 8 mm (range, 25-50 mm). Recovery these complications, surgeons have sought an alternative to autografting. to the S3 or greater level was reported in 86% of repairs. Static 2-point discrim- Nerve conduits of various composition, including polyglycolic acid ination (s2PD) and Semmes-Weinstein monofilament (SWF) were the most com- (PGA), collagen, and autologous vein, have been used in the clinical set- mon completed assessments. Mean s2PD in 24 repairs reporting 2PD data was ting, with variable outcomes.8–15 Processed nerve allografts are an alter- 9 ± 4 mm. For the 38 repairs with SWF data, protective sensation was reported native to nerve autografts and conduits, without donor site morbidity. in 33 repairs, deep pressure in 2, and no recovery in 3. These data compared They consist of human nerves which are prepared with enzymatic favorably with historical data for nerve autograft repairs, with reported levels of decellularization but with preservation of the epineurium and internal fas- meaningful recovery of 60% to 88%. There were no reported adverse effects. cicular architecture. In experimental studies, they have been shown to Processed nerve allograft can be used to reconstruct long gap nerve defects in revascularize rapidly, repopulate with host cells, and provide a supportive the hand with consistently high rates of meaningful recovery. Results for milieu for nerve regeneration.16–18 Several large retrospective and PNA repairs of digital nerve injuries with gaps longer than 25 mm compare prospective clinical studies have demonstrated levels of recovery after favorably with historical reports for nerve autograft repair but without donor processed nerve allograft reconstruction of nerve gaps equaling or site morbidity. exceeding that reported for nerve autograft or conduit reconstruc- Key Words: digital nerves, nerve repair, nerve gap, processed nerve allograft tions.19–22 However, most critical nerve gaps fall in the shorter end of the spectrum, between 5 and 15 mm, and most of the large retrospec- (Ann Plast Surg 2017;78: S292–S295) tive series to date have contained a proportionately larger number of these shorter gaps.8,9 Because of the relative scarcity of longer nerve gaps, it is harder to draw conclusions from the existing literature regarding the relative effectiveness of the available treatment modalities. Received October 26, 2016, and accepted for publication, after revision The RANGER Study registry, opened in 2007, is an institutional December 24, 2016. – From the *Division of Plastic Surgery, University of Kentucky, Lexington, KY; †Arizona review board approved, active database that collects data on the use and Center for Hand Surgery, Phoenix, AZ; ‡Institute for Nerve, Hand, and Reconstructive outcomes of processed nerve allografts using standardized collection Surgery, Rutherford, NJ; §Division of Plastic and Reconstructive Surgery, University methods. The registry pools data from a wide cross-section of patient of Washington, Harborview Medical Center, Seattle, WA; ||Department of Plastic demographics, nerve injury types, mechanisms, and practice settings. Surgery, Vanderbilt University, Nashville, TN; ¶Indiana Hand to Shoulder Center, Indianapolis, IN; #Duke University Department of Orthopedic Surgery, Durham, The advent of the RANGER processed nerve allograft registry allows NC; and **The Buncke Clinic, San Francisco, CA. investigators to specifically query outcomes in lower incidence sub- Presented as a Podium Presentation in the “Hot Topics” Session at the Annual Meeting groups. The purpose of this study was to review a large database of pa- of the Southeastern Society of Plastic and Reconstructive Surgeons, Lake Buena tients who underwent nerve reconstruction with processed nerve Vista, FL, June, 2016. Conflicts of interest and sources of funding: Funding for this research was provided by allograft to determine the degree of sensory recovery after repair of dig- AxoGen, Inc. The authors receive research support from Axogen, Inc., and are ital nerve injuries with critical gaps greater than 25 mm and to compare members of the RANGER study group. these results with historical controls for nerve autograft repairs. Reprints: Brian Rinker, MD, FACS, Division of Plastic Surgery, University of Kentucky, Kentucky Clinic, K454 Lexington, KY. E-mail: [email protected]. METHODS Copyright © 2017 Wolters Kluwer Health, Inc. All rights reserved. ISSN: 0148-7043/17/7806–S292 This investigation was performed with approval from our institu- DOI: 10.1097/SAP.0000000000001037 tional review boards and in accordance with Good Clinical Practices.

S292 www.annalsplasticsurgery.com Annals of Plastic Surgery • Volume 78, Supplement 5, June 2017 Annals of Plastic Surgery • Volume 78, Supplement 5, June 2017 Nerve Allografts for Long Gaps

All consenting subjects at least 18 years old and who had been implanted with processed allografts at participating sites were eligible for the TABLE 2. Summary of Subjects With Concomitant Injuries study. Standardized case report forms were used for each subject to nor- malize chart information, and data were collected to the extent available Concomitant Injuries Subjects in the medical records. Chart reviews were completed in an observa- Vascular 2 tional fashion to collect subject demographics, details of the nerve in- Tendon 3 jury and repair(s), and concomitant injuries and treatments. Each Fracture 4 center followed its own standard practices with regard to postoperative Tendon and vascular 7 care, and functional outcomes data were collected in both a retrospective and prospective manner from all available follow-up evaluations. Fracture and vascular 4 Additionally, information was collected on adverse experiences or com- Tendon and fracture 3 plications related to the nerve graft occurring intraoperatively Vascular, tendon, and fracture 12 or postoperatively. All collected data were entered into the centralized RANGER Semmes-Weinstein monofilament (SWF) were the most common Study registry database. This database was queried for subjects present- completed quantitative assessments with a mean follow-up time of ing with digital nerve repairs of greater than 25 mm with quantitative 11 months. Of subjects reporting 2PD, 92% reported return of s2PD follow-up data. Medical Research Council Classification (MRCC) for with a mean of 9 ± 4 (4-15) mm. Of the subjects reporting SWF data sensory function with meaningful recovery was defined as S3 or greater 23 (n = 38), protective sensation or greater was reported in 33 repairs, deep level. Outcomes were compared with historical data for nerve autograft. pressure in 2, and no recovery in 3. These data compared favorably with historical data for nerve autograft repairs, with reported levels of meaningful recovery of 60% to 88%. There were no reported adverse RESULTS effects (Fig. 1). Subject Demographic, Injury, and Operative Data DISCUSSION Twenty-two male and 6 female subjects were identified in the Digital nerve injuries are among the most common condition database presenting with 50 digital nerve repairs with gaps greater than treated by hand surgeons, but most digital nerve injuries are either ame- 25 mm. The mean age was 45 ± 24 (22-78) years. Forty-two subjects nable to direct suture or present with gaps shorter than 15 mm.24,25 Be- reported no pertinent medical history that may affect the outcome of cause of the relative scarcity of digital nerve injuries with gaps greater the repair. In the remaining 8 subjects, 5 reported hypertension, 1 re- than 25 mm, there is a paucity of data on the relative effectiveness of ported diabetes, and 2 reported both hypertension and diabetes. Six the different modalities for reconstruction of these injuries, and it subjects were active smokers. Mechanisms of injury included saw has been correspondingly difficult to make informed decisions (n = 13), crushing injuries (n = 9), resection of neuroma (n = 9), about treatment. amputation/avulsions (n = 8), sharp lacerations (n = 7), and blast/ In 1991, Frykman and Gramyk reported recovery to at least the gunshots (n = 4). Injuries included all common and proper digital S3+ level on the MRCC scale in 88% of 141 patients with digital nerve nerves in the hand with a majority also reporting concomitant bony, ten- gaps treated with autografting. Patients with gap lengths up to 5 cm don, and/or vascular injuries. Table 1 shows the distribution of nerves were included.26 In 1993, Kallio presented the results of a series of repaired, and Table 2 summarizes the number of repairs with concom- 254 digital nerve repairs in 95 patients, performed over a 16-year itant injuries. Most subjects were repaired acutely with the median time period. One hundred three nerve gaps in 37 patients were treated with to repair at 6 (0-2514) days. The nerve gap injuries were reconstructed nerve autografts, with gap lengths up to 5 cm. The source of autograft with an appropriately sized nerve allograft in a tensionless manner ac- was either the antebrachial cutaneous or sural nerve. The size of the cording to instructions for use. The average gap length was 35 ± 8 study group allowed subset analysis by gap length. Meaningful recov- (27-50) mm. ery, defined as recovery to the S3 or greater level, was observed in 100% of patients with gap lengths less than 21 mm, but this recovery Outcomes Data rate fell to 67% for gap lengths between 21 and 49 mm and to 9% for 25 Recovery to S3 or greater level was reported in 86% of repairs gap lengths greater than 49 mm. A more recent study comparing with 32 of these repairs reaching S3+ or S4 levels of recovery. These results from autografting of digital nerve defects using either the poste- outcomes were found to be consistent across the entire gap range of rior interosseous or medial antebrachial cutaneous nerve reported re- up to 50 mm (Table 3). Static 2-point discrimination (2PD) and covery of S3 or greater in 86% of 28 repairs, with a mean gap length of 22 mm.27

TABLE 1. Distribution of Repaired Nerves in the Hand TABLE 3. Distribution of MRCC Scores for Return of Location No. Repairs Sensory Function

Thumb 4 UDN/4 RDN Sensory Gaps Gaps Cumulative Index finger 5 UDN/6 RDN MRCC Score 26–39 mm 40–50 mm Group Long finger 4 UDN/1 RDN S0 2 2 4 Ring finger 2 UDN/5 RDN S1 1 0 1 Small finger 6 UDN/4 RDN S2 1 0 1 Second Web Space 2 S3 11 1 12 Third Web Space 5 S3+ 18 10 28 Fourth Web Space 1 S4 3 1 4 UND indicates ulnar digital nerve; RDN, radial digital nerve. MRCC ≥ S3 89% 86% 86%

FIGURE 1. A, A 39-year-old man sustained a saw injury to the nondominant left thumb with laceration of the thenar muscles and complete severance of both digital nerves with a wide zone of injury. Appearance after preparation of the nerves by trimming under magnification back to healthy nerve substance. Arrows show nerve stumps. Gap lengths were 25 and 23 mm for the ulnar digital nerve and radial digital nerve, respectively. B, After reconstruction of the nerve gaps with processed nerve allografts. A 1 to 2 mm diameter nerve was used for the radial digital nerve, and a 2 to 3 mm diameter nerve was used for the ulnar digital nerve, with good size match. At 5 months postsurgery, the patient regained sensory recovery to the S3 level in both nerve distributions.

Studies reporting results of long gap digital neve repairs using multiple centers, across a wide cross section of geographic areas and synthetic conduits are rarer still. In 1990, Mackinnon and Dellon pre- practice settings. Standardized entry forms are used, but each participat- sented a series of 15 patients who underwent secondary reconstruction ing site follows its own standard of care for intraoperative and postop- of digital nerve gaps using woven PGA conduits. Gaps ranged from 5 to erative management. The large number of patients included in the 30 mm, with a mean of 17 mm. Recovery to the S3 or greater level was database allows for subset analysis with numbers in excess of that observed in 86% of the repairs.28 In 2005, Battiston et al reported a se- which has previously been possible in peripheral nerve research. For ries of 19 primary and secondary digital nerve reconstructions using the present study, 50 nerve repairs were identified using processed PGA conduits. The gap lengths ranged from 10 to 40 mm, with a mean nerve allografts for digital nerve gaps between 25 and 50 mm length. of 20 mm.29 Recovery to the S3 level or greater was observed in 75% of Of these, 36 (86%) resulted in measured recovery at the S3 or greater the 12 repairs with a gap length ≥ 20 mm, but this level of recovery was level. These results compare favorably with historical data for autografts seen in only 60% of the 5 repairs with a gap ≥ 25 mm. More recently, and conduits.25–30 Unlike previous studies using autografts and syn- Lohmeyer et al reported recovery to the S3 or greater level in 9 (75%) thetic conduits, which showed a degradation in recovery at longer gap of 12 patients who underwent repair of digital nerve gaps using a hollow lengths,25,29 the high level of recovery in this study was preserved, even collagen conduit. The average gap length in this series was 12.2 mm, in the subset of patients with gap lengths between 40 and 50 mm (86% and no gap in the series was longer than 18 mm.30 recovery). The observed advantage of allografts over autografts may re- Nerve allografts have a history which exceeds that of conduits late to the processing, which removes the cellular material, which is still and autografts, dating to 1885 when Eduard Albert of Vienna used a present in autografts, but must be removed through phagocytosis. The fresh allograft from an amputated limb to bridge a 3-cm median nerve observed advantage to conduits likely results from the presence of gap resulting from resection of a sarcoma.31 Nerve allografts did not intact nerve basal lamina or possibly growth factors, which gain broad acceptance for more than a century, however, because of promote regeneration. the need for immunomodulatory therapy and the risk of disease trans- This study is not without limitations. It is retrospective in nature. mission. Techniques have been developed, such as irradiation, enzy- The patients were not randomized to treatment groups, but rather the matic treatment, cryopreservation, and detergent-processing, to render type of treatment was left to the surgeon’s discretion. Surgeons were allografts acellular, thus eliminating the need for immunosuppres- not held to standardized treatment protocols but rather followed the sion.19 Processed nerve allografts have many advantages in the clin- standard of care for their treatment site; therefore, suture materials, su- ical setting. They are biocompatible, easy to use, and avoid a donor turing techniques, use of microscopy, and postoperative assessments site deficit. First made available in 2008, the Avance Nerve Graft is were not standardized. Surgeons and those performing the assessments the only processed nerve allograft approved by the US Food and Drug were not blinded. However, the registry remains ongoing and continues Administration for clinical use in the United States. It undergoes a com- to collect outcomes data. This should allow for an even more granular bination of detergent and enzymatic processing, which is balanced to subset analysis in the future. In addition, a multicenter, prospective ran- mitigate cellular burden while maintaining the basal laminar proteins domized study is underway to compare processed allograft repairs with and microarchitecture of the nerve, both of which are believed to be conduit repair, which should provide even more clarity to one of the un- critical in supporting nerve regeneration.19 This provides theoretical answered questions in hand surgery: the optimal treatment for a critical support for the superiority of processed allograft over other means of re- nerve gap. In the meantime, surgeons should feel confident using proc- construction of a long nerve gap. Additional support for this notion has essed nerve allografts in injuries of the digital nerves, even for long been provided by animal studies. In a rat sciatic nerve model, processed nerve gaps. nerve allograft was found to be superior to nerve isografts and collagen conduits in the reconstruction of a 10-mm gap.18 Processed nerve allograft repairs resulted in improved functional recovery and higher REFERENCES axon counts.18 1. Artico M, Cervoni L, Nucci F, et al. Birthday of peripheral nervous system The RANGER registry was initiated in 2008 to capture data on surgery: the contribution of Gabriele Ferrara (1543-1627). Neurosurgery. the use of the Avance Nerve Graft. It is an active database used by 1996;39:380–382.

2. Yi C, Dahlin LB. Impaired nerve regeneration and Schwann cell activation after 18. Whitlock EL, Tuffaha SH, Luciano JP,et al. Processed allografts and type I colla- repair with tension. Neuroreport. 2010;21:958–962. gen conduits for repair of peripheral nerve gaps. Muscle Nerve. 2009;39:787–799. 3. Driscoll PJ, Glasby MA, Lawson GM. An in vivo study of peripheral nerves in 19. Karabekmez FE, Duymaz A, Moran SL. Early clinical outcomes with the use of continuity: biomechanical and physiological responses to elongation. JOrthop decellularized nerve allograft for repair of sensory defects within the hand. Hand Res. 2002;20:370–375. (N Y).2009;4:245–249. 4. Boyd KU, Nimigan AS, Mackinnon SE. Nerve reconstruction in the hand and 20. Brooks DN, Weber RV, Chao JD, et al. Processed nerve allografts for peripheral upper extremity. Clin Plast Surg. 2011;38:643–660. nerve reconstruction: a multicenter study of utilization and outcomes in sensory, 5. Konofaos P, Ver Halen JP. Nerve repair by means of tubulization: past, pres- mixed, and motor nerve reconstructions. Microsurgery. 2012;32:1–14. – ent, future. J Reconstr Microsurg. 2013;29:149 164. 21. Cho MS, Rinker BD, Weber RV,et al. Functional outcome following nerve repair 6. Ehretsman RL, Novak CB, Mackinnon SE. Subjective recovery of nerve graft in the upper extremity using processed nerve allograft. JHandSurgAm.2012;37: donor site. Ann Plast Surg. 1999;43:606–612. 2340–2349. 7. IJpma FF, Nicolai JP, Meek MF. Sural nerve donor-site morbidity: thirty-four 22. Means KR Jr, Rinker BD, Higgins JP, et al. A multicenter, prospective, random- years of follow-up. Ann Plast Surg. 2006;57:391–395. ized, pilot study of outcomes for digital nerve repair in the hand using hollow con- 8. Weber RA, Breidenbach WC, Brown RE, et al. A randomized prospective study duit compared with processed allograft nerve. Hand (N Y). 2016;11:144–151. of polyglycolic acid conduits for digital nerve reconstruction in humans. Plast 23. Mackinnon SE, Dellon AL. Results of nerve repair and grafting. In: Surgery of the – Reconstr Surg. 2000;106:1036 1045. Peripheral Nerve.1sted.NewYork:ThiemeMedical;1988. 9. Rinker B, Liau JY. A prospective randomized study comparing woven 24. de Medinaceli L, Prayon M, Merle M. Percentage of nerve injuries in which pri- polyglycolic acid and autogenous vein conduits for reconstruction of digital nerve – mary repair can be achieved by end-to-end approximation: review of 2,181 nerve gaps. JHandSurgAm. 2011;36:775 781. lesions. Microsurgery. 1993;14:244–246. 10. Chiu DT, Janecka I, Krizek TJ, et al. Autogenous vein graft as a conduit for nerve regeneration. Surgery. 1982;91:226–233. 25. Kallio PK. The results of secondary repair of 254 digital nerves. JHandSurgBr. 1993;18:327–330. 11. Walton RL, Brown RE, Matory WE Jr, et al. Autogenous vein graft repair of digital nerve defects in the finger: a retrospective clinical study. Plast Reconstr Surg. 26. Frykman G, Gramyk K. Results of nerve grafting. In: Gelberman R, ed. Operative – 1989;84:944–949. Nerve Repair and Reconstruction. Philadelphia: JB Lippincott; 1991:553 567. 12. Chiu DT, Strauch B. A prospective clinical evaluation of autogenous vein grafts 27. Stang F, Stollwerck P, Prommersberger KJ, et al. Posterior interosseous nerve vs. used as a nerve conduit for distal sensory nerve defects of 3 cm or less. Plast medial cutaneous nerve of the forearm: differences in digital nerve reconstruction. Reconstr Surg. 1990;86:928–934. Arch Orthop Trauma Surg. 2013;133:875–880. 13. Schlosshauer B, Dreesmann L, Schaller HE, et al. Synthetic nerve guide implants 28. Mackinnon SE, Dellon AL. Clinical nerve reconstruction with a bioabsorbable in humans: a comprehensive survey. Neurosurgery. 2006;59:740–747. polyglycolic acid tube. Plast Reconstr Surg. 1990;85:419–424. 14. Dienstknecht T, Klein S, Vykoukal J, et al. Type I collagen nerve conduits for me- 29. Battiston B, Geuna S, Ferrero M, et al. Nerve repair by means of tubulization: lit- dian nerve repairs in the forearm. JHandSurgAm. 2013;38:1119–1124. erature review and personal clinical experience comparing biological and syn- 15. Jiang X, Lim SH, Mao HQ, et al. Current applications and future perspectives of thetic conduits for sensory nerve repair. Microsurgery. 2005;25:258–267. artificial nerve conduits. Exp Neurol. 2010;223:86–101. 30. Lohmeyer JA, Siemers F, Machens HG, et al. The clinical use of artificial nerve 16. Hudson TW, Zawko S, Deister C, et al. Optimized acellular nerve graft is immu- conduits for digital nerve repair: a prospective cohort study and literature review. nologically tolerated and supports regeneration. Tissue Eng. 2004;10:1641–1651. J Reconstr Microsurg. 2009;25:55–61. 17. Sondell M, Lundborg G, Kanje M. Regeneration of the rat sciatic nerve into allo- 31. Schmidt G. Eduard Albert and the beginning of human nerve grafting. Acta grafts made acellular through chemical extraction. Brain Res. 1998;795:44–54. Chirurgica Austriaca. 1993;25:287–288.

Fluorescein Isothiocyanate A Novel Application for Lymphatic Surgery Lisa Spiguel, MD,* Christiana Shaw, MD,* Adam Katz, MD,† Lifei Guo, MD, PhD,‡ Hung-Chi Chen, MD,§ Bernard T. Lee, MD,|| and Dhruv Singhal, MD†||

after ALND are highly variable however usually are quoted between Abstract: The Lymphatic Microsurgical Preventing Healing Approach (LYMPHA) 20% and 40%3–6 and have been reported as high as 77%.7 The chal- procedure entails performing a lymphovenous bypass (LVB) at the time of axil- lenge of the LYMPHAprocedure is visualizing healthy cut lymphatics lary lymph node dissection to reduce lymphedema risk. The two most common lateral to the level 1 lymph nodes after an ALND. Dr. Campisi’s team fluorophores utilized in LVB are blue dye and indocyanine green. We developed was able to identify these lymphatics by injecting blue dye into the ip- a novel application of fluorescein isothiocyanate for intraoperative lymphatic silateral proximal upper arm. However, breast surgeons often prefer to mapping. Our goal is to demonstrate the safety and efficacy of fluorescein iso- use a dual tracer method including both blue dye and technetium sul- thiocyanate for this application. We reviewed a prospectively collected database fur colloid for sentinel lymph node (SLN) identification. This is espe- on breast cancer patients who underwent LYMPHA from March to September cially important in cases where neoadjuvant chemotherapy has 2015. Fluorescein isothiocyanate was used to identify arm lymphatic channels af- previously been administered.8,9 Therefore, a different dye was sought ter axillary lymph node dissection to perform an LVB between disrupted lym- for arm lymphatic mapping to differentiate staining from arm versus phatics and axillary vein tributaries. Data on preoperative and intraoperative breast lymphatics. variables were analyzed. Thirteen patients underwent LYMPHA with intraopera- The most common method of lymphatic vessel mapping cur- tive fluorescein isothiocyanate lymphatic mapping from March to September rently in use is indocyanine green (ICG). However, the challenge 2015. Average patient age was 50 years with a mean body mass index of 28. with ICG is that the dye is near-infrared and therefore excited in On average, 3.4 lacerated lymphatic channels were identified at an average dis- – the nonvisible spectrum. This limits the usefulness of ICG for visu- tance of 2.72 cm (range, 0.25 5 cm) caudal to the axillary vein. On average, alization and simultaneous dissection because the dye is displayed as 1.7 channels were bypassed per patient. Eleven anastomoses were performed to a white signal on a black background and cannot be concurrently visu- the accessory branch of the axillary vein and 1 to a lateral branch. In 1 patient, alized through the binoculars of a microscope (Fig. 1A). Fluorescein a bypass was not performed due to poor lymphatic caliber and inadequate length isothiocyanate (FITC), on the other hand, is excited in the visible of the harvested vein tributary. No intraoperative adverse events were noted. spectrum and routinely used in the operating room. Neurosurgeons Fluorescein isothiocyanate is a safe and effective method for intra-operative inject this dye intravenously and use microscopes equipped with lymphatic mapping. Fluorescein isothiocyanate imaging allows for simulta- filter technology to visualize tumors while maintaining life-like color neous dissection and lymphatic visualization, making it an ideal agent for of the surrounding tissues allowing for simultaneous magnification lymphatic mapping and dissection in open surgical fields, such as in the and tissue dissection (Fig. 1B). This is a powerful property for the LYMPHA procedure. lymphatic surgeon. Although FITC has been used outside the operating Key Words: FITC, LYMPHA, lymphatic surgery room for lymphatic mapping in the skin,11–13 to our knowledge, it had never been used in the operating room for lymphatic mapping. More (Ann Plast Surg 2017;78: S296–S298) recently, a single case report was published from France where FITC was used to perform a lymphovenous bypass (LVB) in the superficial n 2009, Dr. Campisi introduced a technique to prevent lymphedema 14 I termed the Lymphatic Microsurgical Preventing Healing Approach tissues of the arm in a patient with chronic lymphedema. (LYMPHA).1 In this technique, lymphatics draining the arm are identi- We hypothesized that FITC would be a safe and highly effective fied and bypassed into an axillary vein tributary at the time of an axil- dye for lymphatic mapping and dissection in open surgical fields, such lary dissection. Dr. Campisi’s team has reported an unprecedented 5% as in the LYMPHA procedure. lymphedema rate after axillary lymph node dissection (ALND) and LYMPHA over a four year follow-up.2 Historical rates of lymphedema METHODS We reviewed our prospectively collected Lymphedema Reposi- tory data on all breast cancer patients who underwent the LYMPHA Received December 13, 2016, and accepted for publication, after revision December procedure from March to September 2015. All patients that underwent 24, 2016. the LYMPHA procedure with FITC were included in the review, and the From the *Department of Surgery; †Division of Plastic and Reconstructive Surgery, Department of Surgery, University of Florida School of Medicine, Gainesville, charts were reviewed for demographic information (age, body mass in- FL; ‡Department of Plastic and Reconstructive Surgery, Lahey Clinic, dex) and perioperative data (number of lymphatic channels visualized Burlington, MA; §Department of Plastic and Reconstructive Surgery, China and bypassed, distance of channels from axillary vein, name of targeted Medical University Hospital, China Medical University, Taichung, Taiwan; and vein, and adverse events). ||Division of Plastic and Reconstructive Surgery, Department of Surgery, Beth Israel Deaconess Medical Center, Harvard Medical School, Boston, MA. Conflicts of interest and sources of funding: none declared. Surgical Technique Presented at the Plastic Surgery Research Council (New York, NY), May, 2016. Presented at the Southeastern Society for Plastic and Reconstructive Surgeons, Immediately before the ALND, 2 mL of a modified 2% fluo- June, 2016. rescein solution are injected intradermally and along the muscle fascia Reprints: Dhruv Singhal, MD, Division of Plastic and Reconstructive Surgery, Beth of the ipsilateral upper arm (Fig. 2A). Our solution is modified from Israel Deaconess Medical Center, 110, Francis Street, Suite 5A, Boston, MA the stock AK-FLUOR 10% (Akorn Inc, Lake Forest, Ill) solution by 02215. E-mail: [email protected]. Copyright © 2017 Wolters Kluwer Health, Inc. All rights reserved. diluting 2 cc with 7.5 cc of normal saline and 0.5 cc of AlbuRx5 ISSN: 0148-7043/17/7806–S296 (CSL Behring Inc, King of Prussia, Pa). The ALND is performed with DOI: 10.1097/SAP.0000000000001034 careful attention to preserve a superficial accessory vein tributary which

S296 www.annalsplasticsurgery.com Annals of Plastic Surgery • Volume 78, Supplement 5, June 2017 Annals of Plastic Surgery • Volume 78, Supplement 5, June 2017 FITC and Lymphatic Surgery

FIGURE 1. Comparison of ICG and FITC Imaging. A, Image of an superficial circumflex iliac artery donor flap harvest site with reverse lymphatic mapping10 of the abdomen with ICG. Residual lymphatics draining the abdomen are identified as a white signal on a black background. Of note, despite the microscope being equipped with ICG capabilities, this image cannot be visualized in real-time through the binoculars and instead is displayed on a secondary screen. B, FITC imaging demonstrating a spinal meningioma. Note the life like color of the surrounding tissues and ability to visualize the target through the binoculars of the microscope allowing for simultaneous magnification and dissection of the meningioma.(Figure 1B courtesy of Dr. Maryam Rahman, Neurosurgery, UF Health). longitudinally traverses the level 1 lymph nodes. The superior dissection After completion of the axillary dissection, a Pentero 900D of the level 1 axillary contents along the axillary vein is performed with Microscope (Carl Zeiss Inc, Germany) equipped with the YELLOW identification of the accessory vein tributary which is typically found 560 package is used to identify and map the cut lymphatic channels anterior to the thoracodorsal neurovascular bundle. The vein is then draining the arm. The harvested vein is prepared per standard microsur- dissected free from the level 1 axillary contents and clipped distally to gical technique. Using 9–0 nylon suture, a “U” stitch is placed to capture provide maximal length. Routine completion of the levels 1 and 2 the anterior wall of the vein and parachute in the lymphatic channels ALND is then performed. chosen for bypass.1 (Fig. 2B) 10-0 nylon is then used to suture the

FIGURE 2. Schematic of the modified LYMPHA technique using FITC. A, Both blue and nuclear dyes are reserved for breast sentinel lymph node identification. FITC is injected in the proximal upper inner arm. B, After completion of the axillary dissection and removal of levels 1 and 2 lymph nodes, arm lymphatic channels, now “glowing” from the FITC injection, are identified and re-routed into an axillary vein tributary.

used commonly by our neurosurgery and ophthalmology colleagues TABLE 1. Advantages and Disadvantages of the 2 Most with a high safety profile.15–17 No serious adverse reaction has ever Commonly Used Fluorophores in Lymphatic Surgery been documented, to our knowledge, when injected intradermally. (Blue Dye and ICG) in Comparison to FITC Moreover, FITC does not permanently stain surrounding tissues, as opposed to ICG and blue dyes, which facilitates dissection of the Dye Advantages Disadvantages lymphatic channels. The primary advantage of FITC over ICG in Blue dye Technical Technical lymphatic surgery is the ability to allow for simultaneous visualization ✓Visualized through ✗No depth of penetration and dissection of lymphatic channels because FITC is excited in the binoculars (live surgery) ✗Permanent training visible spectrum, making it an ideal dye to be used in open surgical ✓No specialized equipment Safety fields. The limitations of FITC are that its depth of penetration is a necessary ✗Adverse reactions quarter that of ICG (therefore, ineffective for transdermal visualization), and it does require specialized equipment for visualization. However, as Skin necrosis (methylene blue) we noted at our institution, this specialized equipment if often already Anaphylaxis (isosulfan blue) readily available. ✗Cross-reactivity Sulfa drugs (isosulfan blue) SSRI (methylene blue) CONCLUSIONS ICG Technical Technical Fluorescein isothiocyanate is a safe and effective technique for ✓Depth of penetration ✗Unable to visualize through lymphatic mapping in the LYMPHA technique. We encourage our col- =20mm binoculars (no live surgery) leagues to explore the power of FITC in open surgical fields to simulta- Safety ✗Permanent staining neously visualize, magnify, and dissect lymphatic channels. ✓No adverse reactions ✗Requires specialized (dermal) equipment REFERENCES ✓No cross-reactivities 1. Boccardo F, Casabona F, De Cian F, et al. Lymphedema Microsurgical Preventive FITC Technical Technical Healing Approach: a new technique for primary prevention of arm lymphedema ✓Visualized through ✗Requires specialized after mastectomy. Ann Surg Oncol. 2009;16:703–708. binoculars (live surgery) equipment 2. Boccardo F, Casabona F, De Cian F, et al. Lymphatic Microsurgical Preventing ✓Depth of penetration Healing Approach (LYMPHA) for primary surgical prevention of breast cancer- =5mm related lymphedema: Over 4 years follow-up. Microsurgery. 2014. ✓ 3. Husted Madsen A, Haugaard K, Soerensen J, et al. Arm morbidity following No permanent staining sentinel lymph node biopsy or axillary lymph node dissection: a study from Safety the Danish Breast Cancer Cooperative Group. Breast. 2008;17:138–147. ✓No adverse 4. Leidenius M, Leivonen M, Vironen J, et al. The consequences of long-time arm reactions (dermal) morbidity in node-negative breast cancer patients with sentinel node biopsy or axillary clearance. JSurgOncol. 2005;92:23–31. ✓No cross-reactivities 5. Blanchard DK, Donohue JH, Reynolds C, et al. Relapse and morbidity in patients SSRI, selective serotonin reuptake inhibitor. undergoing sentinel lymph node biopsy alone or with axillary dissection for breast cancer. Arch Surg. 2003;138:482–488. 6. Haid A, Köberle-Wührer R, Knauer M, et al. Morbidity of breast cancer patients wall of the vein to the perilymphatic tissue. Channels not bypassed following complete axillary dissection or sentinel node biopsy only: a comparative – are clipped. Lymphatic flow filling the vein can be visualized with evaluation. Breast Cancer Res Treat.2002;73:31 36. 7. Rönkä R, von Smitten K, Tasmuth T, et al. One-year morbidity after sentinel node the filter activated 1 hour after anastomosis. (Supplemental Digital biopsy and breast surgery. Breast. 2005;14:28–36. Content 1, http://links.lww.com/SAP/A232. Video demonstrating 8. Boughey JC, Suman VJ, Mittendorf EA, et al. Factors affecting sentinel lymph the modified LYMPHA procedure with FITC for arm lymphatic node identification rate after neoadjuvant chemotherapy for breast cancer patients visualization and confirmation of flow into the venous system). enrolled in ACOSOG Z1071 (Alliance). Ann Surg. 2015;261:547–552. 9. Boughey JC, Suman VJ, Mittendorf EA, et al. Sentinel lymph node surgery after RESULTS neoadjuvant chemotherapy in patients with node-positive breast cancer: the ACOSOG Z1071 (Alliance) clinical trial. JAMA. 2013;310:1455–1461. Thirteen patients underwent LYMPHA with intraoperative FITC 10. Dayan JH, Dayan E, Smith ML. Reverse lymphatic mapping. Plast Reconstr Surg. lymphatic imaging from March to September 2015. Average patient age 2015;135:277–285. was 50 years with a mean body mass index of 28. On average, 3.4 lac- 11. Bollinger A, Jäger K, Sgier F, et al. Fluorescence microlymphography. Circulation. erated lymphatic channels (range, 1–8) were identified at an average 1981;64:1195–1200. distance of 2.72 cm (range, 0.25–5 cm) caudal to the axillary vein. 12. Leu AJ, Husmann MJ, Held T, et al. Measurement of the lymphatic clearance of 1.7 channels were bypassed per patient (0–4). Eleven anastomoses were the human skin using a fluorescent tracer. JVascRes. 2001;38:423–431. performed to the accessory branch of the axillary vein and 1 to a lateral 13. Keo HH, Husmann M, Groechenig E, et al. Diagnostic accuracy of fluorescence microlymphography for detecting limb lymphedema. Eur J Vasc Endovasc Surg. branch. In 1 patient, a bypass was not performed due to poor lymphatic 2015;49:474–479. caliber and inadequate length of the vein tributary. No intraoperative or 14. Ayestaray B, Bekara F. Fluorescein sodium fluorescence microscope-integrated 30-day postoperative adverse events were noted. LYMPHA added an lymphangiography for lymphatic supermicrosurgery. Microsurgery.2015;35: average of 67 minutes (45–120 minutes) to the oncologic procedure. 407–410. 15. Lipson BK, Yannuzzi LA. Complications of intravenous fluorescein injections. – DISCUSSION Int Ophthalmol Clin. 1989;29:200 205. 16. Yannuzzi LA, Rohrer KT, Tindel LJ, et al. Fluorescein angiography complication This study demonstrates that FITC is a safe and effective dye for survey. Ophthalmology. 1986;93:611–617. the LYMPHA technique. In comparison to ICG and blue dye, FITC has 17. Karhunen U, Raitta C, Kala R. Adverse reactions to fluorescein angiography. many advantages.(Table 1) Fluorescein isothiocyanate is a safe dye Acta Ophthalmol (Copenh). 1986;64:282–286.

A Comparison of 4 Analgesic Regimens for Acute Postoperative Pain Control in Breast Augmentation Patients Pamela Tan, MD,* Morgan Sparks Martin, MD,* Nina Shank, BA,* Leann Myers, PhD,† Emily Wolfe, MD,* John Lindsey, MD,* and Stephen Metzinger, MD*

from anesthesia. Factors that will delay discharge include nausea, Purpose: Patients undergoing breast augmentation are treated with multiple vomiting, and unrelieved pain. There are significant consequences of combinations of medications for pain control including ketorolac, liposomal inadequate perioperative pain that include a delay in recovery, a delayed bupivacaine, bupivacaine, and intravenous and oral narcotics. There is no current return to normal daily living, and reduced patient satisfaction.2 Tradi- consensus on the optimal combination; therefore, all are used at the discretion of tional systemic methods of pain control include opioids; however, the the surgeon. adverse outcomes associated with opioids continue to be a major Methods: This was a single-center, retrospective study. The total number of pa- problem in outpatient anesthesia. Potential side effects include altered tients included was 132. Comparisons were made between 4 groups: bupivacaine mental status, narcotic dependence, nausea, emesis, impairment of bowel only (B); bupivacaine and liposomal bupivacaine (BL); bupivacaine and liposo- function, constipation, and respiratory depression.3 Multimodal analge- mal bupivacaine plus intraoperative ketorolac (BLKi); and bupivacaine and lipo- sics, where local and intravenous analgesics are combined, is a pre- somal bupivacaine plus postoperative ketorolac (BLKp). Average pain scores ferred over monotherapies because of lower pain scores, reduced opioid immediately postoperative and before discharge were recorded and correlated usage, and opioid-related adverse effects. to percentage of patients who received narcotic in the post-anesthesia care unit Breast augmentation was the most common cosmetic surgical (PACU). Additional end points noted were side effects including nausea and time procedure (286,000) as per American Society of Plastic Surgeons 2014 spent in PACU postoperatively. Plastic Surgery Statistics Report. Our goal is to provide adequate post- Results: Those receiving intraoperative ketorolac had the lowest pain on dis- operative analgesia after outpatient breast augmentation while meeting charge (P < 0.0001) and the lowest percentage of patients receiving narcotics the criteria for discharge. We will evaluate the efficacy of multimodal (P = 0.009) out of all 4 groups. There was no significant difference between therapy comparing 4 different techniques: bupivacaine only (B); the 4 groups in terms of time spent in PACU, pain immediately after the proce- bupivacaine and liposomal bupivacaine (BL); bupivacaine and liposo- dure, or amount of antiemetic given. No bleeding complications were noted for mal bupivacaine plus intraoperative ketorolac (BLKi); and bupivacaine those who did or did not receive ketorolac. and liposomal bupivacaine plus postoperative ketorolac (BLKp). We Conclusions: When given options for pain control in breast augmentation, intra- hypothesize that intraoperative use of ketorolac will produce prolonged operative ketorolac should be considered, because its inclusion was significant in analgesia, decrease postoperative supplemental narcotic use, and facil- decreasing use of narcotics and pain upon discharge. Addition of other costly itate discharge from post-anesthesia care unit (PACU). We also hypoth- drugs such as liposomal bupivacaine may not provide additional benefit in the esize that intraoperative use of liposomal bupivacaine will produce immediate postoperative setting for procedures with a short recovery period such prolonged analgesia in PACU. as breast augmentation. Key Words: pain management, breast augmentation, ketorolac, bupivacaine, MATERIALS AND METHODS liposomal bupivacaine (Ann Plast Surg 2017;78: S299–S304) Study Population This was a single-center, retrospective study that analyzed ver the past 4 decades, there have been an increasing number of patients who underwent elective outpatient surgery for breast aug- O outpatient procedures because of advances in anesthetic and sur- mentation. All patients were female and 18 years or older. Patients gical techniques and escalating healthcare costs. Improvement in anes- that underwent surgery for this procedure over a 7-year period were thesia enabled patients to regain consciousness more quickly with fewer – 1 included (2009 2015). Those who did not receive a combination of after effects and better analgesics for relief of pain. Discharge of bupivacaine and liposomal bupivacaine with or without ketorolac patients undergoing outpatient procedures often depends on recovery were excluded. Two surgeons performed all operations with similar techniques. Data were collected via a database located at the outpatient surgery center where the procedures were performed. This Received August 22, 2016, and accepted for publication, after revision March 23, database held all scanned medical records pertaining to the periop- 2017. From the *Division of Plastic and Reconstructive Surgery, and †Department of erative care of the patient. The study was approved by the Biomed- Biostatistics and Bioinformatics, Tulane University, New Orleans, LA. ical Institutional Review Board of Tulane University New Orleans Conflicts of interest and sources of funding: The authors have no financial or commercial with waiver of consent. associations that might create a conflict of interest with the information presented in this article. Tulane University Department of Global Biostatistics and Data Sciences. Dr. Myers Data Collection was supported in part by 1 U54 GM104940 from the National Institute of For the purpose of this study, information was collected re- General Medical Sciences of the National Institutes of Health, which funds the Louisiana Clinical and Translational Science Center. garding type of analgesia used intravenously or intramuscularly by Presented at the Southern Society of Plastic and Reconstructive Surgery Annual anesthesia and subcutaneously by the surgeon intraoperatively. meeting 2016. These data were combined with the medications given intravenously Reprints: Pamela Tan, MD, 1430 Tulane Ave, SL22, Room 8207, New Orleans, LA and by mouth in the PACU. Timing of the narcotics given postoper- 70112. E-mail: [email protected]. Copyright © 2017 Wolters Kluwer Health, Inc. All rights reserved. atively was also analyzed. ISSN: 0148-7043/17/7806–S299 Comparisons were made between 4 groups: bupivacaine only (B); DOI: 10.1097/SAP.0000000000001132 bupivacaine and liposomal bupivacaine (BL); bupivacaine and liposomal

Annals of Plastic Surgery • Volume 78, Supplement 5, June 2017 www.annalsplasticsurgery.com S299 Tan et al Annals of Plastic Surgery • Volume 78, Supplement 5, June 2017

in 100 (75.8%) out of 132 patients, and periareolar incisions were placed in 32 (24.2%) out of 132 patients. The injection of local anesthetic for each group was as follows: Patients in the B group (Fig. 1) were given a mean of 40 mL of bupivacaine with 0.25% epinephrine per patient, and this was injected into 3 areas: (1) into the marked incision before incision, (2) intercostal nerve block before incision, and (3) into the breast pocket after dissection. Patients in the BL group (Fig. 2) were given a mixture of 20 mL of bupivacaine, 20 mL of liposomal bupivacaine with 0.25% epinephrine, and 20 mL of 0.9% normal saline, and this was injected into 3 areas: (1) into the marked incision before incision, (2) intercostal nerve block before incision, and (3) into the breast pocket after dissection. Patients in the BLKi group (Fig. 3) were given a mixture of 20 mL of bupivacaine, 20 mL of liposomal bupivacaine with 0.25% epinephrine, and 20 mL of 0.9% normal saline, and this was injected into 3 areas: (1) into the marked incision before incision, (2) intercostal nerve block before incision, and (3) into the breast pocket after dissec- FIGURE 1. Patients in the B group were given a mean of 40 mL of tion. In addition, ketorolac 30 mg intravenous (IV) and 30 mg intramus- bupivacaine with 0.25% epinephrine per patient, and this was cular (IM) were given postincisionally but before the end of procedure. injected into 3 areas: (1) into the marked incision before incision, Patients in the BLKp group (Fig. 4) were given a mixture of (2) intercostal nerve block before incision, and (3) into the 20 mL of bupivacaine, 20 mL of liposomal bupivacaine with 0.25% breast pocket after dissection. epinephrine, and 20 mL of 0.9% normal saline, and this was injected into 3 areas: (1) into the marked incision before incision, (2) intercostal bupivacaine plus intraoperative ketorolac (BLKi); and bupivacaine and nerve block before incision, and (3) into the breast pocket after dissec- liposomal bupivacaine plus postoperative ketorolac (BLKp). Outcomes tion. In addition, ketorolac 30 mg IV and 30 mg IM were given at the of interest were number of patients that required narcotics in the PACU, end of procedure. amount of narcotic received postoperatively, amount of antiemetic re- It should be noted that the use of liposomal bupivacaine as a required postoperatively, pain scores upon arrival to PACU, pain scores gional nerve block is off-label and has not been approved by the Food upon discharge, time in PACU, and complications. The generic unidi- and Drug Administration. mensional numeric rating scale was used with “no pain” represented by a score of 0 and “worst imaginable pain” representedbyascoreof 10. Patient age, body mass index (BMI), and weight were also assessed. Statistical Analysis All complications after surgery were recorded to determine the rate of hematoma formation within each set of patients. For the continuous measures, data were analyzed using analy- Initial query of the database of all patients undergoing breast sis of variance methods. When assumptions of normality were not augmentation during appropriate time period identified 132 patients. met, results were verified by analysis with the Kruskal-Wallis test. Division based on combination of analgesic resulted in the following: For categorical measures, differences in frequencies were assessed bupivacaine only (N = 59); bupivacaine and liposomal bupivacaine using Pearson χ2 test. Subhypotheses were assessed using Bonferroni- (N = 26); bupivacaine and liposomal bupivacaine plus intraoperative adjusted P values. ketorolac (N = 34); and liposomal bupivacaine and bupivacaine plus postoperative ketorolac (N = 13). Patients were excluded if they had a subglandular augmentation. All remaining patients included in the study received a submuscular augmentation.

Technique Primary breast augmentation was performed in all patients in the subpectoral plane. Patients were first seen and marked preoperatively. After appropriate induction of anesthesia, skin prepping, and sterile draping, the operation was initiated with injection of local anesthetic into the incision and as an intercostal nerve block, specific to each group as detailed below. An incision was then made at the inframammary fold or periareolar. Dissection was performed such that a subpectoral plane was created using a dual plane technique with complete division of pectoralis origins along the inferior border. Once adequate hemostasis was achieved, injection of local anesthetic into the breast pocket was performed. After implant selection, pocket was irrigated, and permanent saline or silicone implants were placed. No drains were used. FIGURE 2. Patients in the BL group were given a mixture of The mean implant volume for the study population was 363 cc 20 mL of bupivacaine, 20 mL of liposomal bupivacaine with with a range of 175 to 550 cc. There was no difference in the mean vol- 0.25% epinephrine, and 20 mL of 0.9% normal saline, and this ume of implant placed between the groups. All implants were round was injected into 3 areas: (1) into the marked incision before and smooth. Out of 132 patients, 68 (51.5%) had saline implants and incision, (2) intercostal nerve block before incision, and (3) into 64 (48.5%) had silicone implants. Inframammary incisions were used the breast pocket after dissection.

S300 www.annalsplasticsurgery.com © 2017 Wolters Kluwer Health, Inc. All rights reserved. Annals of Plastic Surgery • Volume 78, Supplement 5, June 2017 Postoperative Pain Control in Breast Augmentation

urinary retention, ileus,13–15 headache, constipation, sleep disturbance, and altered mental status, as well as tolerance and abuse that often occur. To reduce use, other modes of analgesia such as nonsteroidal anti- inflammatory drugs (NSAIDs) and local anesthetics are advised. Local anesthetics have the ability to limit central sensitization by limiting impulses to afferent fibers and diminishing spinal cord stimulation and pain processing.8,16 This can contribute to a significant pain- free period postoperatively. Bupivacaine has a half-life of 2.7 hours, and this duration is prolonged when combined with epinephrine.12 It is also comparatively inexpensive and can be used to obtain this pain- free period.10,17 An alternative that is even longer acting is liposomal bupivacaine. This liposomal formulation is administered intraoperatively and releases bupivacaine for up to 96 hours. This has been demonstrated in other studies as an advantageous addition to multimodal analgesia.18–22 Nonsteroidal anti-inflammatory drugs have the ability to work FIGURE 3. Patients in the BLKi group were given a mixture of both centrally and peripherally. The local inflammatory response from injury is reduced, attenuating peripheral sensitization, which also 20 mL of bupivacaine, 20 mL of liposomal bupivacaine with 10 0.25% epinephrine, and 20 mL of 0.9% normal saline, and this decreases spinal nociceptive processing. was injected into 3 areas: (1) into the marked incision before Centrally, they prevent spinal prostanoid synthesis. By reducing incision, (2) intercostal nerve block before incision, and (3) into both peripheral and central sensitization, pain is reduced in the postoper- the breast pocket after dissection. In addition, ketorolac 30 mg ative period, requiring fewer additional analgesics including decreased total dose of opioids even after major surgery.8,16,23 Nonsteroidal anti- IV and 30 mg IM were given postincisionally but before the 23–25 end of procedure. inflammatory drugs are likely to be inadequate when used alone. Other studies have demonstrated that a combination of NSAIDs such as ketorolac with other modalities including local anesthetics is safe RESULTS – and superior than the use of each product alone.8,23 25 A total of 132 patients were analyzed in this study and fell into 1 One study comparing intravenous ketorolac versus placebo for of 4 groups: group B (n = 59), group BL (n = 26), group BLKi (n = 34), pain control after surgery found that NSAIDs combined with opioids and group BLKp (n = 13). There were no significant differences in age resulted in not only less opioid consumption and better analgesia but or BMI across the groups, and all breast augmentations took place in the also decreased postoperative nausea, vomiting, and sedation.26–28 This same surgical center. Procedures were all performed by 1 of 2 surgeons. gives NSAIDs an advantage over opioids in the perioperative period.10 Patients in group BLKi had lower pain scores upon discharge at 1.15 Ketorolac has even been shown to provide similar analgesic effect to (P < 0.0001). Additionally, the smallest number of patients requiring morphine.29,30 For mild to moderate postoperative pain, NSAIDS should narcotics in PACU was also found in the BLKi group at 29.4% (10 out be the drug of choice.3,13 of 34 patients; P = 0.0094). There was no significant difference be- Multimodal analgesia has become the standard in clinical prac- tween the groups when comparing time in PACU, pain immediately tice in recent decades in attempts to prevent peripheral and central postoperative, number of patients who received antiemetics postop- stimulation and control postoperative pain.26,31–33 This approach eratively and time in the PACU for patients before they needed pain utilizes combinations of analgesic drugs to optimize pain relief. Drugs medications (Table 1). with different mechanisms of action can be used to target distinct DISCUSSION Breast augmentation has steadily increased in popularity, and parallel to it, the need to optimize postoperative pain control, mitigate cost, and decrease adverse side effects.4 Placement of the implant into the retromuscular pocket has also increased pain associated with the procedure because of the partial detachment of the sternal and costal origins of the pectoralis major muscle. In our study, we have aimed to demonstrate the best combination of drugs for optimal pain control in this patient population. Studies show that pain is triggered by sensitization of neurons in the pain pathway. This sensitization can be either central or peripheral.5 During tissue injury such as surgery, cytokine and prostaglandins are produced. This increases inflammation and, therefore, increases excit- ability of central neurons in the spinal cord, giving the sensation of pain.6 During surgery, massive stimulation occurs of peripheral afferent fibers.7,8 Other studies also show that C-fibers are activated by brief noxious stimulation, and this may alter central neural function by pro- longing their alteration.9,10 General anesthesia attenuates this stimula- FIGURE 4. Patients in the BLKp group were given a mixture of tion, but does not prevent it.8 Traditional analgesic regimens have 20 mL of bupivacaine, 20 mL of liposomal bupivacaine with involved narcotics both intravenously or orally; however, systemic opi- 0.25% epinephrine, and 20 mL of 0.9% normal saline, and this oid use alone is insufficient to block nociceptive neurons and prevent was injected into 3 areas: (1) into the marked incision before central sensitization.10,11 In addition, there are side effects such as incision, (2) intercostal nerve block before incision, and (3) into postoperative nausea and vomiting. The spare use of opioids can decrease the breast pocket after dissection. In addition, ketorolac 30 mg this effect as well as sedation, disorientation,10,12 respiratory depression, IV and 30 mg IM were given at the end of procedure.

TABLE 1. Summary of Outcomes

B (n = 59) BL (N = 26) BLKi (N = 34) BLKp (N = 13) Variable Mean SD Mean SD Mean SD Mean SD P Age 30.8 9.9 31.9 9.3 32.1 8.4 28.1 6.9 0.5309 BMI 21.7 2.7 22.5 4.4 22.3 3.0 20.8 2.3 0.4604 Time PACU 90.9 18.7 93.8 23.7 86.4 16.9 96.9 12.9 0.5644 Pain po (0–10) 0.56 1.74 0.23 1.18 0.53 1.94 0.00 0.00 0.6832 Pain dc (0–10) 2.22 1.91 2.81 1.30 1.15 1.78 3.08 1.32 <0.0001 Time meds 33.9 19.5 38.6 37.2 24.6 12.1 41.4 23.2 0.4085 Amount of meperidine given (mg) in PACU 37.1 18.9 34.2 18.6 34.0 32.0 23.2 8.6 N=32 N=19 N=10 N=7 0.5521 N%N%N% N % P Received meperidine 33 55.9 19 73.1 9 26.5 7 53.8 0.0044 Received antiemetics 30 50.8 15 44.1 9 26.5 6 46.2 0.0706 B group, bupivacaine only; BL group, bupivacaine and liposomal bupivacaine; BLKi group, ketorolac given intraoperatively plus bupivacaine and liposomal bupivacaine; BLKp group, ketorolac given postoperatively plus bupivacaine and liposomal bupivacaine. Variables listed above include age; BMI; time PACU, time spent in the PACU; pain po, pain score immediately upon PACU arrival; pain dc, pain score upon discharge; time meds, timing of narcotic administration; amount of meperidine given (mg); patients receiving meperidine; and patients receiving antiemetics. pathways. This method of postsurgical pain control is preferred to before “primary” hemostasis was achieved.45,46 There is currently no singular methods.18,34–36 evidence to suggest any bleeding complications from use of ketorolac Numerous studies have shown that a multimodal regimen incor- after hemostasis was achieved.47 It is commonly given near the end of porating several different mechanisms of action has clear advantages surgery.13 In our study, ketorolac was used postincisional in breast over one that employs a single medication.37 In our study, we looked at augmentation with no bleeding complications. In both clinical and 4 different pain regimens: group B, group BL, group BLKi, and group laboratory studies, bleeding has been demonstrated because of a re- BLKp. Our study shows that intraoperative ketorolac with liposomal versible inhibition of platelet aggregation causing prolonged bleeding bupivacaine and bupivacaine (BLKi) performed the best in terms of times.30,48 This, however, does not conclude that ketorolac is unsafe lowering narcotic usage and pain scores upon discharge (Fig. 5). for use. In patients who are not otherwise coagulopathic, it has even Additionally, bupivacaine is much more cost-effective, be- been demonstrated as safe with respect to bleeding in surgery patients cause liposomal bupivacaine costs one hundred times more than including those undergoing spinal procedure.13,49–52 standard bupivacaine.38 There is also a concern among the surgical community that As discussed above, avoiding peripheral and central sensitization ketorolac can cause acute kidney injury postoperatively. This is a is an optimal method to decrease post-injury pain. Preventive analgesia rare event and occurs in approximately 1 in 1000 to 100,000.53 Its use utilizes this theory by using analgesia before the painful stimulus to does not significantly elevate serum creatinine levels and can be used decrease the intensity of the pain.39 This has been demonstrated in safely.23,54 Some studies have shown that the transient fall in postoper- animal studies with substantial results.5 Because of NSAIDs demon- ative glomerular filtration rate and, thus, renal function is insignificant strating antinociceptive effects both on peripheral and central neurons, enough to warrant its use preoperatively.55 ketorolac is an optimal drug for use in preventive analgesia. Local anesthetic can be combined for additional local, peripheral nociceptor blockade to attenuate sensitization before the surgical stimulus.6,40 Based on the underlying principal of preemptive analgesia, ketorolac as well as local anesthetic should be given before incision for optimal pain prevention. Ketorolac should be given as preventive analgesia after primary hemostasis has been achieved. Our study supports that the timing of ketorolac administration greatly influences outcomes. With regards to discharge pain scores and postoperative narcotic usage, our results suggest that intravenous ketorolac given systemically after surgery is less effective than ketorolac given intraoperatively, near the end of surgery. This is supported by a study in hand surgery. Here, preincisional use of both local anesthetics and NSAIDs as opposed to postincisional or postsurgical was superior for adequate pain control.41,42 Nonsteroidal anti-inflammatory drugs are still thought to be relatively contraindicated when used preincisionally because of the effect on platelet function, which is crucial for surgical hemostasis.13 Overall, it is accepted for use preincisionally in small procedures but is still unclear in larger more invasive procedures because of lack of sufficient evidence in its safety.16,43 In the neurosurgical population, it is has been associated with higher risk for postoperative hematoma.44 Others have FIGURE 5. Pain score at time of discharge. In pairwise reported significant blood loss intraoperatively if ketorolac was given comparisons, BLKi < BL, and BLKi < BLKp (all P < 0.05).

S302 www.annalsplasticsurgery.com © 2017 Wolters Kluwer Health, Inc. All rights reserved. Annals of Plastic Surgery • Volume 78, Supplement 5, June 2017 Postoperative Pain Control in Breast Augmentation

Studies have found that parenteral ketorolac is associated with 18. Baxter R, Bramlett K, Onel E, et al. Impact of local administration of liposome upper gastrointestinal adverse events including gastritis, duodenitis, bupivacaine for postsurgical analgesia on wound healing: a review of data from – peptic ulcers, and hemorrhage compared with nonuse.56 Ketorolac should ten prospective, controlled clinical studies. Clin Ther. 2013;35:312 320.e5. be avoided in patients with malignancy and coagulopathies, as well as 19. Golf M, Daniels SE, Onel E. A phase 3, randomized, placebo-controlled trial of DepoFoam® bupivacaine (extended-release bupivacaine local analgesic) in chronic kidney disease. bunionectomy. Adv Ther. 2011;28:776–788. This study found no significant difference in postoperative nau- 20. Gorfine SR, Onel E, Patou G, et al. Bupivacaine extended-release liposome injection sea, vomiting, renal injury, or hemorrhage between these 4 groups. for prolonged postsurgical analgesia in patients undergoing hemorrhoidectomy: a multicenter, randomized, double-blind, placebo-controlled trial. Dis Colon Rectum. 2011;54:1552–1559. CONCLUSIONS 21. Smoot JD, Bergese SD, Onel E, et al. The efficacy and safety of DepoFoam Intraoperative ketorolac should be considered as part of the bupivacaine in patients undergoing bilateral, cosmetic, submuscular augmentation analgesic regimen for acute postoperative pain control in breast aug- mammaplasty: a randomized, double-blind, active-control study. Aesthet Surg J. – mentation. Its inclusion was significant in decreasing patient use of 2012;32:69 76. narcotics and pain upon discharge. Other points to consider include 22. Haas E, Onel E, Miller H, et al. A double-blind, randomized, active-controlled study for post-hemorrhoidectomy pain management with liposome bupivacaine, issues of cost, overall patient satisfaction, and safety as compared to a novel local analgesic formulation. Am Surg. 2012;78:574–581. narcotics. Addition of other costly drugs such as liposomal bupivacaine 23. Pavy TJ, Paech MJ, Evans SF. The effect of intravenous ketorolac on opioid re- may not provide further benefit in the immediate postoperative setting quirement and pain after cesarean delivery. Anesth Analg. 2001;92:1010–1014. for procedures with a short recovery period such as breast augmenta- 24. Gin T, Kan AF,Lam KK, et al. Analgesia after caesarean section with intramuscu- tion. Studies have found that compared with narcotics, intravenous lar ketorolac or pethidine. Anaesth Intensive Care. 1993;21:420–423. ketorolac produced a lower overall mean cost per patient, as well as a 25. Tzeng JI, Mok MS. Combination of intramuscular ketorolac and low dose epidu- higher average patient satisfaction score.57 ral morphine for the relief of post-caesarean pain. Ann Acad Med Singapore. – Limitations of this study include that these data have been ret- 1994;23:10 13. rospectively analyzed, which may have more sources of bias than if 26. Gan TJ, Daniels SE, Singla N, et al. A novel injectable formulation of diclofenac compared with intravenous ketorolac or placebo for acute moderate-to-severe pain it was studied prospectively. Additionally, 2 surgeons performed all after abdominal or pelvic surgery: a multicenter, double-blind, randomized, multiple- operations. This study does not account for differences in skill, tech- dose study. Anesth Analg. 2012;115:1212–1220. nique, and experience between the 2. 27. Elia N, Lysakowski C, Tramèr MR. Does multimodal analgesia with acetamino- phen, nonsteroidal antiinflammatory drugs, or selective cyclooxygenase-2 inhibi- tors and patient-controlled analgesia morphine offer advantages over morphine REFERENCES alone? Meta-analyses of randomized trials. Anesthesiology. 2005;103:1296–1304. 1. National Health Statistics Reports Ambulatory Surgery in the United States, 2006. 28. Marret E, Kurdi O, Zufferey P, et al. Effects of nonsteroidal antiinflammatory Available at: http://www.cdc.gov/nchs/data/nhsr/nhsr011.pdf. drugs on patient-controlled analgesia morphine side effects: meta-analysis of ran- 2. Joshi GP, Ogunnaike BO. Consequences of inadequate postoperative pain relief domized controlled trials. Anesthesiology. 2005;102:1249–1260. and chronic persistent postoperative pain. Anesthesiol Clin North America. 2005; 29. Le Roux PD, Samudrala S. Postoperative pain after lumbar disc surgery: a com- 23:21–26. parison between parenteral ketorolac and narcotics. Acta Neurochir (Wien). 1999; 3. American Society of Anesthesiologists Task Force on Acute Pain Management. 141:261–267. Practice guidelines for acute pain management in the perioperative setting: 30. Resman-Targoff BH. Ketorolac: a parenteral nonsteroidal antiinflammatory drug. an updated report by the American Society of Anesthesiologists Task Force DICP. 1990;24:1098–1104. – on Acute Pain Management. Anesthesiology. 2012;116:248 273. 31. Acute Pain Management: Scientific Evidence. 3rd ed. Melbourne, Australia: 4. Stanley SS, Hoppe IC, Ciminello FS. Pain control following breast augmentation: Australian and New Zealand College of Anaesthetists and Faculty of Pain a qualitative systematic review. Aesthet Surg J. 2012;32:964–972. Medicine; 2010. 5. Kelly DJ, Ahmad M, Brull SJ. Preemptive analgesia I: physiological pathways and 32. Ashburn MA, Caplan RA, Carr DB, et al. Practice guidelines for acute pain man- pharmacological modalities. Can J Anaesth. 2001;48:1000–1010. agement in the perioperative setting: an updated report by the American Society of 6. Gutta R, Koehn CR, James LE. Does ketorolac have a preemptive analgesic Anesthesiologists Task Force on Acute Pain Management. Anesthesiology.2004; effect? A randomized, double-blind, control study. J Oral Maxillofac Surg. 100:1573–1581. – 2013;71:2029 2034. 33. Carr DB, Jacox AK, Chapman CR, et al Clinical Practice Guideline Number 1. 7. Ferreira SH, Nakamura M, de Abreu Castro MS. The hyperalgesic effects of AHCPR Publication No. 92 0032. In: Acute Pain Management: Operative or prostacyclin and prostaglandin E2. Prostaglandins. 1978;16:31–37. Medical Procedures and Trauma. Rockville, MD: Agency for Healthcare Re- 8. Rundshagen I, Kochs E, Schulte am Esch J. Surgical stimulation increases search and Quality, Public Health Service, US Department of Health and Human median nerve somatosensory evoked responses during isoflurane-nitrous oxide Services; 1992. anaesthesia. Br J Anaesth. 1995;75:598–602. 34. Apfelbaum JL, Ashburn MA, Connis RT, et al. Practice guidelines for acute pain 9. Melzack R,WallPD.The Challenge of Pain.2nded.NewYork:Basic management in the perioperative setting: an updated report by the American Soci- Books; 1988. ety of Anesthesiologists Task Force on Acute Pain Management. Anesthesiology. – 10. Mahabir RC, Peterson BD, Williamson JS, et al. Locally administered ketorolac 2012;116:248 273. and bupivacaine for control of postoperative pain in breast augmentation patients. 35. McGuire L, Heffner K, Glaser R, et al. Pain and wound healing in surgical Plast Reconstr Surg. 2004;114:1910–1916. patients. Ann Behav Med. 2006;31:165–172. 11. Katz J, Clairoux M, Redahan C, et al. High dose alfentanil pre-empts pain after 36. Jin F, Chung F.Multimodal analgesia for postoperative pain control. J Clin Anesth. abdominal hysterectomy. Pain. 1996;68:109–118. 2001;13:524–539. 12. Swerdlow M, Jones R. The duration of action of bupivacaine, prilocaine and 37. Ziemann-Gimmel P, Hensel P, Koppman J, et al. Multimodal analgesia re- lignocaine. Br J Anaesth. 1970;42:335–339. duces narcotic requirements and antiemetic rescue medication in laparoscopic 13. Magni G, La Rosa I, Melillo G, et al. Intracranial hemorrhage requiring surgery Roux-en-Y gastric bypass surgery. Surg Obes Relat Dis. 2013;9:975–980. in neurosurgical patients given ketorolac: a case-control study within a cohort 38. Noviasky J, Pierce DP, Whalen K, et al. Bupivacaine liposomal versus bupivacaine: (2001–2010). Anesth Analg. 2013;116:443–447. comparative review. Hosp Pharm. 2014;49:539–543. 14. Litvak KM, McEvoy GK. Ketorolac, an injectable nonnarcotic analgesic. Clin 39. Woolf CJ, Chong MS. Preemptive analgesia—treating postoperative pain by Pharm. 1990;9:921–935. preventing the establishment of central sensitization. Anesth Analg. 1993;77: 15. DeAndrade JR, Maslanka M, Reines HD, et al. Ketorolac versus meperidine for 362–379. pain relief after orthopaedic surgery. Clin Orthop Relat Res. 1996;325:301–312. 40. Savage MG, Henry MA. Preoperative nonsteroidal anti-inflammatory agents: re- 16. O'Hanlon DM, Thambipillai T, Colbert ST, et al. Timing of pre-emptive view of the literature. Oral Surg Oral Med Oral Pathol Oral Radiol Endod.2004; tenoxicam is important for postoperative analgesia. Can J Anaesth. 2001; 98:146–152. 48:162–166. 41. Katz J, Kavanagh BP, Sandler AN, et al. Preemptive analgesia. Clinical evidence 17. Thorne AC. Local anesthetics. In: Aston SJ, Beasley RW,Thorne CH, eds. Grabb of neuroplasticity contributing to postoperative pain. Anesthesiology. 1992;77: and Smith's Plastic Surgery. 5th ed. Philadelphia: Lippincott-Raven; 1997. 439–446.

42. Ejlersen E, Andersen HB, Eliasen K, et al. A comparison between preincisional 50. Lieh-Lai MW, Kauffman RE, Uy HG, et al. A randomized comparison of and postincisional lidocaine infiltration and postoperative pain. Anesth Analg. ketorolac tromethamine and morphine for postoperative analgesia in critically 1992;74:495–498. ill children. Crit Care Med. 1999;27:2786–2791. 43. Mahabir RC, Peterson BD, Williamson JS, et al. Locally administered ketorolac 51. Forrest JB, Camu F, Greer IA, et al. Ketorolac, diclofenac, and ketoprofen are and bupivacaine for control of postoperative pain in breast augmentation patients: equally safe for pain relief after major surgery. Br J Anaesth. 2002;88:227–233. part II. 10-day follow-up. Plast Reconstr Surg. 2008;121:638–643. 52. Gillis JC, Brogden RN. Ketorolac. A reappraisal of its pharmacodynamic and 44. Palmer JD, Sparrow OC, Iannotti F. Postoperative hematoma: a 5-year survey and pharmacokinetic properties and therapeutic use in pain management. Drugs. identification of avoidable risk factors. Neurosurgery. 1994;35:1061–1064. 1997;53:139–188. 45. De Oliveira GS Jr, Agarwal D, Benzon HT. Perioperative single dose ketorolac to 53. Parker RK, Holtmann B, Smith I, et al. Use of ketorolac after lower abdominal prevent postoperative pain: a meta-analysis of randomized trials. Anesth Analg. surgery. Effect on analgesic requirement and surgical outcome. Anesthesiology. 2012;114:424–433. 1994;80:6–12. 46. White PF, Raeder J, Kehlet H. Ketorolac: its role as part of a multimodal analgesic 54. Myles PS, Power I. Does ketorolac cause postoperative renal failure: how do we regimen. Anesth Analg. 2012;114:250–254. assess the evidence? Br J Anaesth. 1998;80:420–421. 47. Magni G, La Rosa I, Gimignani S, et al. Early postoperative complications after 55. Lee A, Cooper MC, Craig JC, et al. Effects of nonsteroidal anti-inflammatory intracranial surgery: comparison between total intravenous and balanced anesthe- drugs on post-operative renal function in normal adults. Cochrane Database Syst sia. J Neurosurg Anesthesiol. 2007;19:229–234. Rev. 2001;2:CD002765. 48. Singer AJ, Mynster CJ, McMahon BJ. The effect of IM ketorolac tromethamine 56. Chang CH, Chen HC, Lin JW,et al. Risk of hospitalization for upper gastrointes- on bleeding time: a prospective, interventional, controlled study. Am J Emerg tinal adverse events associated with nonsteroidal anti-inflammatory drugs: a Med. 2003;21:441–443. nationwide case-crossover study in Taiwan. Pharmacoepidemiol Drug Saf. 2011; 49. Cassinelli EH, Dean CL, Garcia RM, et al. Ketorolac use for postoperative pain 20:763–771. management following lumbar decompression surgery: a prospective, random- 57. Rainer TH, Jacobs P, Ng YC, et al. Cost effectiveness analysis of intravenous ized, double-blinded, placebo-controlled trial. Spine (Phila Pa 1976).2008;33: ketorolac and morphine for treating pain after limb injury: double blind randomised 1313–1317. controlled trial. BMJ. 2000;321:1247–1251.

Clinical, Biomechanical, and Anatomic Investigation of Colles Fascia and Pubic Ramus Periosteum for Use During Medial Thighplasty Martin J. Carney, BS,* Tim Matatov, MD,† Matthew Freeman, MD,‡ John Miller, BS,§ Rahul Vemula, MD,† Jason Schuster, MSBME,|| Michael Dancisak, PhD,|| John Lindsey, MD,† and Guenevere Rae, PhD¶

Key Words: medial thighplasty, body contouring, cosmetic, Introduction: The medial thighplasty is a procedure where patients may attain superficial fascial system, plastic surgery superior mobility, hygiene, and cosmesis. Most surgeons use attachment of the superficial fascial system (SFS) of the thigh flap to the Colles fascia, whereas (Ann Plast Surg 2017;78: S305–S310) others attach the SFS to the pubic ramus periosteum. Because of a high complication profile, we aim to elucidate the clinical, biomechanical, and anatomic qual- ithin the auspice of an obesity epidemic nationally, there has ities of the Colles fascia versus the pubic ramus periosteum. W been increased popularization of bariatric surgery with a resul- Materials and Methods: We performed a 17-year retrospective review tant need for postoperative body contouring.1–3 The massive weight documenting clinical complications, a biomechanical analysis of sutures placed loss achieved after gastric restrictive procedures produces severe skin in different tissue layers of the thigh, and a histologic analysis surrounding the laxity, lipodystrophy, and deformities best addressed through skin exci- ischiopubic ramus. Separate suture pull-out strength testing was conducted on ca- sion and reapproximation.4,5 On a psychosocial level, massive weight daveric tissue using an Admet MTEST Quattro with no. 1 Vicryl suture and tissue loss can negatively affect quality of life, self-esteem, and body image, grips at a displacement rate of 2.12 mm/s. Simultaneous displacement and force while also exacerbating physical discomfort.6 First described by were acquired at 100 Hz and with measurements obtained at regular intervals be- Lewis,7,8 medial thighplasty helps to improve mobility, hygiene, and tween the pubic symphysis and the ischial tuberosity in both the Colles fascia and aesthetics within the indicated patient population. This traditional tech- the deeper periosteal layers of the thigh. A histologic analysis was performed at 3 nique, however, remained unpopular within the standard cosmetic plas- points along the ischiopubic ramus using paraffin-embedded large mount tissue tic surgery clinic for many years because of scar migration, wound sections stained with hematoxylin, eosin, and Gomori trichrome. dehiscence, vulvar distortion, lymphatic disruption, and early recur- Results: Thirty-nine patients underwent medial thighplasty with a 46.16% com- rence of thigh ptosis.9 plication rate. Suture pull-out force of the suspected superficial Colles fascia sites It was not until the notoriety of Lockwood10 and his descrip- was, on average, 72.8% less than values from the deeper periosteum tissue. An- tion of the superficial fascial system (SFS), and its interconnected- chor points in the Colles fascia elongated 17.4% further before failure than those ness throughout the body, that the medial thigh lift was seen as a in the periosteum. There was noticeable variability between anchor points and manageable and beneficial procedure. The SFS is thought to be across samples. The histologic sections suggest that the Colles fascia from the dif- widely distributed throughout the body because it is largely respon- ferent regions of the ischiopubic ramus varies considerably in both continuity and sible for tethering the skin and fat to underlying tissues, creating collagen fiber content with no discernible pattern. The periosteal and muscular many of the topographic landmarks in body-surface anatomy.11 fascial layers were more continuous histologically with direct attachments into Lockwood10–13 described a dense area of the Colles fascia at the the pubis and ischium. junction of the perineum and medial thigh as the anatomic shelf— Conclusions: Anchoring of the SFS to the periosteum did not improve our com- an idea that remained a standard throughout all future publications plication profile when compared with the literature. Both the biomechanical and and medial thighplasty technique descriptions. Even today, the histologic analyses demonstrate that the Colles fascia is highly variable in organi- deep-tissue anchoring maneuver into Colles fascia is technically im- zations with coincident variability in tissue strength. Our results require further portant in both horizontal and vertical thighplasty. Current practice study to identify the optimal surgical technique for medial thighplasty. still reflects this advancement but does not distinguish a difference between anchoring of the SFS to the superficial perineal fascia or the ischiopubic ramus to reinforce thighplasty and prevent scar migration.1 Figure 1 depicts an illustration of the Lockwood anatomic Received September 10, 2016, and accepted for publication, after revision December description of the medial thigh. 6, 2016. From the *Tulane University School of Medicine, †Division of Plastic Surgery, Tulane These anatomic and surgical advancements of the thigh lift as University School of Medicine, New Orleans, LA; ‡Division of Plastic Surgery, well as the addition of intraoperative liposuction have yet to appreciably Mount Sinai Hospital, New York, NY; §Louisiana State University Health decrease postoperative morbidity.14 Depending on study criteria, com- Sciences Center; ||Center for Anatomical and Movement Sciences, Tulane plication rates range from an average of 42.72% to as high as 74%, University; and ¶Department of Cell Biology and Anatomy, Louisiana State and most commonly include dehiscence, seroma formation, wound in- University Health Sciences Center, New Orleans, LA. 6,15 Conflicts of interest and sources of funding: none declared. fection, or hematoma. Massive weight loss patients are predisposed This work was accepted for presentation at the Southeastern Society of Plastic and to impaired wound healing and often have underlying comorbidities Reconstructive Surgeons—59th Annual Scientific Meeting, Orlando, Fla, that may further complicate postoperative outcomes.16 June 2016; the American Society of Plastic Surgeons—Plastic Surgery The Meeting, Los Angeles, Calif, September 2016; and abstract publication in Medial thighplasty remains a distinct challenge for even the most PRS Global Open. experienced plastic surgeon and holds distinct room for improvement. Reprints: John Lindsey, MD, Division of Plastic Surgery, Tulane University School There is a paucity of literature surrounding the anatomic and biome- of Medicine, 4288 Houma Blvd, no. 500, Metairie, LA 70006. E-mail: chanical properties of Colles fascia compared with the periosteum of [email protected]. Copyright © 2017 Wolters Kluwer Health, Inc. All rights reserved. the inferior pubic ramus, even as key components to a successful repair. ISSN: 0148-7043/17/7806–S305 Noting a central lack of understanding and extreme complication pro- DOI: 10.1097/SAP.0000000000001033 file, we attempt to distinguish a clinically relevant picture of the inner

Annals of Plastic Surgery • Volume 78, Supplement 5, June 2017 www.annalsplasticsurgery.com S305 Carney et al Annals of Plastic Surgery • Volume 78, Supplement 5, June 2017

FIGURE 1. An illustration of the Lockwood anatomic description of the medial thigh. Oblique view of the pelvis including the muscles of the medial compartment of the thigh (red). The SFS (yellow) as described by Lockwood has been cut and reflected to display the clinically relevant anatomy of the medial thigh lift. The SFS of the medial thigh is continuous with the perineal fascia. The thick condensation or band of the Colles fascia is shown at the junction of the medial thigh and the perineum, overlying the inferior pubic ramus. This thick band continues cephalad toward the pubic symphysis. Not pictured: anteriorly, the SFS continues over the pubic symphysis as the Scarpa fascia of the abdomen. Posteriorly, the SFS is continuous with the urogenital diaphragm. thigh SFS, Colles fascia, and inferior pubic ramus periosteum using su- tissue layers along the inferior ramus of the pubis (IRP). Case 1 used the ture pull-out strengths and histologic mapping. left side of the pelvis to anchor 7 sutures into periosteal tissue along the IRP,and the right side of the pelvis from the same specimen was used to METHODS anchor 7 sutures into the suspected Colles fascia. All anchor sites were placed at 1-cm intervals. Case 2 used 3 sites at 3-cm intervals with dis- crete anchor sites in the Colles fascia, the adductor fascia, and the deep Clinical Analysis periosteal layer along the IRP. The 3-site interval was used to reduce A 17-year retrospective chart review was conducted for all pa- proximity effects when pulling adjacent sutures. In both cases, a single tients from our chief surgeon's (J.L.) private practice who underwent surgeon with more than 20 years of clinical experience placed all su- medial thighplasty. Inclusion criteria included all patients older than tures to reduce variability due to clinician preference. An Admet 18 years who received medial thighplasty with anchoring to the inferior MTEST Quattro mechanical test stand with specialized suture and tis- pubic ramus periosteum. Exclusion criteria were patients younger than sue grips was used to hold and obtain data. Sutures were preloaded to 18 years. Demographic data such as age, sex, comorbid disease, and less than 0.1 kg to remove slack in the tissue, and the system was then body mass index were tabulated. Complications were subdivided into zeroed. All specimens were pulled at a displacement rate of 2.12 mm/s, major and minor criteria. Major complications were categorized as vul- and simultaneous displacement and force data were acquired at 100 Hz. var distortion, scar migration, hematomas, seromas requiring drainage, All specimens were allowed to reach room temperature (68–70°F) be- sepsis, death, and 3 or more minor complications. Minor complications fore testing. Figure 2 depicts the Admet MTEST Quattro equipment us- included cellulitis, wound dehiscence, and seroma not requiring drainage during a suture-pull trial. age. Institutional board review was acquired (protocol no. 9256).

Biomechanical Analysis Histologic Analysis Uniaxial biomechanical parameters were tested on 2 un- A fresh female cadaver (age, 79 years) was acquired for the his- embalmed cadavers on separate occasions and included suture pull- tologic evaluation of the SFS overlying the superior and inferior out strength and displacement of tissue at load failure. Tissue pull-out ischiopubic rami and the ischial tuberosity. Both the right and left is- strength and elongation were acquired using no. 1 Vicryl suture an- chial and pubis portions of the os coxae bones were removed with the chored at regular intervals between the pubic symphysis and the ischial soft tissues intact. The anterior margin of the removed specimen was tuberosity in both the superficial Colles fascia and the deeper periosteal the epidermis, and the posterior margin was the periosteum and soft

FIGURE 2. The tissue grip of Admet MTEST Quattro during trial of the specimen. tissue structures attached to the bone. Care was given to restrict separa- 62.38 mm. Maximal load for all periosteal sites ranged from a high of tion of the soft tissues from each other during the extraction process 20.56 kg to a low of 2.37 kg, and the displacement ranged from a high to ensure that the specimens remained as true to anatomic of 62.38 mm to a low of 13.28 mm (Table 1). The difference in the max- position as possible. imal load for suture pull-out between the Colles site and the periosteal The specimens were decalcified over a 5-day period using Calex site was 73.5%, and at the minimal loads, the difference was 50.2%. and then sliced using a cell path surgical knife at 1.5-cm intervals per- The 2 highest loads were recorded at the periosteal site approximately pendicular to the long axis of the bone. Sections from 3 regions were chosen for anatomic investigation (Fig. 3). The tissue segments were paraffin embedded using a 3-day embedding process and sectioned at 6 μm using a Leica RM2125 microtome. The sections were then mountedona3Â 2–in glass microscope slide and stained with hematoxylin and eosin and Gomori trichrome. Photomicrographs were taken using a Nikon Eclipse Ni of 3 tissue segments labeled A to C (Figs. 4–6). These photomicrographs and microscope slides were visually analyzed for collagen continuity and organization by an expert histologist.

RESULTS Thirty-nine patients underwent medial thighplasty with attachment of SFS to the periosteum and met the inclusion criteria for this study. Demographics included a mean age of 42 years and an average body mass index of 28.21, and the study group included 36 women and 3 men. Overall, there was a 46.16% complication rate, with 5 patients (12.6%) involving major complications and 13 patients (33.3%) involving minor complications. Primarily, patients experienced wound dehiscence (25.64%), cellulitis or erythema (20.5%), and seroma (2.56%). Notably, this population did not experience any scar migration, labial distortion, or severe complications such as sepsis or death. Biomechanical analysis is reported on a case-specific basis.

Case 1 The Colles fascia showed a maximal load of 5.45 kg at a failure displacement of 52.33 mm at the site 1 cm from the ischial tuberosity. This was the highest load for any Colles site tested in either case spec- FIGURE 3. The representative regions of the os coxae that were imen. Maximal load for all Colles sites ranged from a high of 5.45 kg to compared. The areas indicated in red depict the representative a low of 1.18 kg, and the displacement ranged from a high of 62.23 mm regions in which microanatomy was compared. The posterior to a low of 22.82 mm (Table 1). In contrast, maximal load for inferior pubic ramus (A), the midinferior pubic ramus (B), and the periosteal sites reached 20.56 kg at a failure displacement of the anterior inferior pubic ramus (C).

midline of the pubic symphysis. Maximal load for all Colles sites ranged from a high of 2.64 kg to a low of 0.91 kg, and the displacement ranged from a high of 26.94 mm to a low of 4.92 mm (Table 1). In contrast, maximal load for the periosteal sites reached 12.24 kg at a failure displacement of 55.83 mm. Maximal load for all periosteal sites ranged from a high of 12.24 kg to a low of 3.49 kg, and the displacement ranged from a high of 55.83 mm to a low of 9.01 mm (Table 1). The difference in the maximal load for suture pull-out between the Colles site and the periosteal site was 78.4%, and for the minimal loads, the difference was 73.9%. The 2 highest loads were recorded at the periosteal site immediately adjacent to the pubic symphysis and approximately 1 cm from the symphyseal border. In contrast, the lowest load was recorded at the periosteal site located 1.5 cm lateral and posterior to the high- load anchor site.

Left Side The Colles fascia had a maximal load of 2.04 kg at a failure displacement of 34.99 mm at the site approximately 3.0 cm from the midline of the pubic symphysis. Maximal load for all Colles sites ranged from a high of 2.04 kg to a low of 1.56 kg, and the displacement ranged from a high of 34.99 mm to a low of 18.64 mm (Table 1). The maximal load for the periosteal sites reached 20.71 kg at a failure displacement of 42.81 mm. Maximal load for all periosteal sites ranged from a high of FIGURE 4. Microanatomy of region A. This low-power photomicrograph is a transverse section through the superior pubic ramus. *Epidermis and dermal layers. AD Mag indicates the adductor magnus muscle; hypo, the hypodermis/layer of subcutaneous tissue; IT, the ischial tuberosity.

1 and 2 cm from the symphyseal border. In contrast, the 2 lowest loads were recorded at the periosteal sites located 1 and 2 cm lateral and posterior to the high-load anchor sites (approximately 1 and 2 cm medial and anterior to the ischial tuberosity).

Case 2 Right Side The Colles fascia showed a maximal load of 2.64 kg at a failure displacement of 26.94 mm at the site approximately 1.5 cm from the

FIGURE 5. Microanatomy of region B. This low-power photomicrograph is a transverse section through the inferior pubic ramus. Dermis indicates the reticular layer of the dermis. FIGURE 6. Microanatomy of region C. This low-power *Membranous layer of subcutaneous tissue (the Colles fascia). photomicrograph is a transverse section through the superior AB indicates the abductor brevis muscle; bone, the inferior pubic pubic ramus. Ad L indicates the adductor longus muscle; Cr, the ramus; C, the crus of the clitoris; G, the gracilis muscle; OE, the crus of the clitoris; hypo, the hypodermis/subcutaneous tissue; obturator externus muscle. pubis, the bone; S, the epidermis and dermal layers.

TABLE 1. Biomechanical Experimental Data

Periosteum Colles Fascia Suture Location Maximum Load, kg Extension at Max Load, mm Maximum Load, kg Extension at Max Load, mm Case 1 Anterior at the symphyseal surface 10.98 23.24 3.33 52.35 12.56 40.01 —— 20.56 62.38 1.18 22.82 2.75 13.28 4.80 53.37 2.37 25.34 3.60 26.34 4.88 36.60 5.45 52.33 Posterior at the ischial tuberosity 3.85 13.45 3.99 62.23 Case 2—right pelvis Anterior at the symphyseal surface 11.55 52.76 1.43 26.32 Anterior inferior ramus 12.24 55.83 2.64 26.94 Posterior inferior ramus 0.91 4.92 Posterior at the ischial tuberosity 3.49 9.01 2.34 12.94 Case 2—left pelvis Anterior at the symphyseal surface 6.92 34.33 1.56 18.64 Medial at the inferior ramus 20.71 42.81 2.04 34.99 Posterior at the ischial tuberosity 6.98 44.68 1.81 32.00

20.71 kg to a low of 6.92 kg, and the displacement ranged from a high factor of hypodermal thickness, the fibroelastic bands were at least of 44.68 mm to a low of 34.33 mm (Table 1). The difference in the max- double in thickness to region B and more organized. The composi- imal load for suture pull-out between the Colles site and the periosteal tion of the fibroelastic bands seems to have a composition of in- site was 90.1%, and at the minimal loads, the difference was 77.4%. creased elastic fibers when compared with region A, but less than The highest load was recorded at the periosteal site midway between region B. the pubic symphysis and the ischial tuberosity. In contrast, the lowest load was recorded at the periosteal site located adjacent to the pubic symphysis. DISCUSSION In addition to the trials to determine suture pull-out loads in tis- The SFS is anatomically variable in the area of the medial thigh sues, 2 unused sutures were tested using specifically designed yarn in both biomechanical analysis and histologic observation. We also did holders for the Admet MTEST Quattro to test for failure strength and not note a difference in complication rates within this series as com- were found consistent with manufacturer data. Based on manufacturer pared with the literature. Because medial thighplasty continues to prog- data, the maximal load was estimated at 12.1 kg at failure, which was ress technically, a foundational understanding of the SFS must be a top similar to our trials that found loads of 12.71 kg, and 12.69 kg at failure. priority. The ideal patient with localized skin excess and minimal adipos- The histologic microanatomy of the SFS in regions A to C was ity of the thighs is often not the day-to-day reality of body contouring.17 consistent on both the left and right sides of the cadaver. Even with the addition of liposuction-assisted lifts improving postopera- A: The subcutaneous tissue had thick congruent bands of fibroelastic tive thigh contour in both the bariatric and nonbariatric patients, compli- tissue spanning from the most superficial region of the hypodermis to cation rates remain unreasonably high.6,9,14,15,18 the tendons inserting into the periosteum of the ischial tuberosity Massive weight loss from bariatric surgery creates an aesthetic (Fig. 4). Most of the fibroelastic bands were composed of collagen sequelae of excess skin best corrected through body-contouring proce- and oriented in a direction that was parallel to the surface of the body. dures such as medial thighplasty.19 Skin redundancies in the groin will The ratio of collagen throughout the section was markedly increased often cause dermatologic disease, which diminishes personal hygiene compared with regions B and C. and the ability to exercise—severely hindering activities of daily liv- B: The subcutaneous layer of these sections had thin interconnecting ing.2 Although medial thighplasty consistently demonstrates minor bands of fibroelastic tissue spanning from the most superficial region complications, patients receive a significant improvement in quality of the hypodermis to the superficial fascia of the underlying muscles of life.19,20 Even outside of our country's obesity epidemic, the western- and the crus of the clitoris (Fig. 5). There was a larger, more orga- ization of diets in certain Asian countries has made body contouring nized band of fibroelastic tissue superficial to the fascia of the gracilis more common in recent years.21 These questions and complications muscle; however, this band was discontinuous throughout the entire have global implications. Yet, the morbidity burden after medial specimen. Furthermore, the fibroelastic bands within the hypodermis thighplasty suppresses enthusiasm on the topic and often removes the were organized both parallel and perpendicular to the body's surface. procedure from a plastic surgeon's armamentarium. The content of collagen versus elastic fibers within the hypodermis Although the importance of the SFS has been well documented seems to have more elastic fibers when compared with both regions in the plastic surgery literature, there is also evidence to suggest that the A and B, but less collagen than region A. composition of the subcutaneous layer (SFS) may be anatomically var- C: The subcutaneous layer of the sections taken from this region was iable on an individual basis and within certain body regions.22,23 There approximately half the thickness of sections taken from regions A is also evidence to support the claim that the membranous layer of the and B (Fig. 6). The hypodermis had fibroelastic bands of tissue that SFS is not a continuous layer.24,25 The comparison of the microanatomy were organized parallel to the body's surface. Even removing the of regions A to C revealed that the fiber composition of these regions

varies according to location along the ischiopubic ramus. This observa- 2. Moreno CH, Neto HJ, Junior AH, et al. Thighplasty after bariatric surgery: eval- tion has been seen in other regions of the body such as the abdomen and uation of lymphatic drainage in lower extremities. Obes Surg.2008;18: – extremities.22,25 Although the distance between these regions is less 1160 1164. than 2.5 cm, there are anatomic properties that likely affect the orienta- 3. Richter DF, Stoff A. The Scarpa lift—a novel technique for minimal invasive medial thigh lifts. Obes Surg. 2011;21:1975–1980. tion of the fibroelastic bands that develop within the hypodermis. The 4. Simone P, Carusi C, Del Buono R, et al. Medial thigh lift in post-bariatric patients: muscle attachments change from all 3 of these regions. Section A is our encouraging experience. J Plast Surg Hand Surg. 2016;50:359–366. the attachment site for mainly the adductor magnus and the obturator 5. Schoeller T, Meirer R, Otto-Schoeller A, et al. Medial thigh lift free flap for autol- externus on the lateral aspect of the bone; section B is the attachment ogous breast augmentation after bariatric surgery. Obes Surg. 2002;12:831–834. site for the gracilis, adductor brevis, and obturator externus laterally 6. Sisti A, Cuomo R, Zerini I, et al. Complications associated with medial thigh lift: a and the obturator internus on the internal aspect; section C is the attach- comprehensive literature review. J Cutan Aesthet Surg. 2015;8:191–197. ment site for the adductor longus on the lateral aspect and the obturator 7. Lewis JR Jr. The thigh lift. J Int Coll Surg. 1957;27:330–334. 26 internus on the internal aspect of the bone. These differences may al- 8. Lewis JR Jr. Correction of ptosis of the thighs: the thigh lift. Plast Reconstr Surg. ter the stresses placed on the soft tissue structures surrounding these 1966;37:494–498. muscles and the composition of the hypodermis. In addition, because 9. Armijo BS, Campbell CF, Rohrich RJ. Four-step medial thighplasty: refined and of the complex functionality of the perineal region, gravity itself places reproducible. Plast Reconstr Surg. 2014;134:717e–725e. stress on the cutaneous and subcutaneous tissue layers because the re- 10. Lockwood TE. Fascial anchoring technique in medial thigh lifts. Plast Reconstr mainder of the lower extremity draws the skin downward. Furthermore, Surg. 1988;82:299–304. the sitting position makes this region somewhat weight bearing. These 11. Lockwood TE. Superficial fascial system (SFS) of the trunk and extremities: a factors most likely vary by individual and result in a subcutaneous anat- new concept. Plast Reconstr Surg. 1991;87:1009–1018. omy that is also variable. 12. Lockwood T. Lower body lift with superficial fascial system suspension. Plast – This study is not without limitations. Because of this being a Reconstr Surg. 1993;92:1112 1122. – single-surgeon, single-practice site literature review, we had limited 13. Lockwood T. Medial thigh surgery. Aesthet Surg J. 1996;16:94 96. 14. Le Louarn C, Pascal JF. The concentric medial thigh lift. Aesthetic Plast Surg. number of patients for our clinical analysis. Retrospective reviews are – inherently limited to relying on charting amenable to human error. 2004;28:20 23. Some surgeons also contend increased pain with periosteum attachment, 15. Schmidt M, Pollhammer MS, Januszyk M, et al. Concomitant liposuction reduces complications of vertical medial thigh lift in massive weight loss patients. Plast although our clinical series could not prove this. As well, there are differ- Reconstr Surg. 2016;137:1748–1757. ences between fresh frozen cadaver specimen and animated tissue, 16. Gusenoff JA, Coon D, Nayar H, et al. Medial thigh lift in the massive weight loss which may affect both biomechanical and histologic interpretations. population: outcomes and complications. Plast Reconstr Surg. 2015;135:98–106. 17. Aboueldahab AK. Liposuction-assisted medial thigh lift in obese and non obese CONCLUSIONS patients. J Cutan Aesthet Surg. 2013;6:217–225. 18. Labardi L, Gentile P, Gigliotti S, et al. Medial thighplasty: horizontal and vertical Medial thighplasty remains a distinct challenge for plastic sur- – geons because of the high morbidity rate postoperatively and insuffi- procedures after massive weight loss. J Cutan Aesthet Surg. 2012;5:20 25. cient data driven agreement on technique. Our results preliminarily 19. Bertheuil N, Thienot S, Chaput B, et al. Quality-of-life assessment after medial thighplasty in patients following massive weight loss. Plast Reconstr Surg. support the conclusion that there is anatomic variability of the SFS 2015;135:67e–73e. — within the medial thigh region histologically and mechanically 20. Bertheuil N, Carloni R, De Runz A, et al. Medial thighplasty: current concepts and although anchoring of the SFS to the inferior pubic ramus periosteum practices. Ann Chir Plast Esthet. 2016;61:E1–E7. did not improve our complication profile in comparison with previous 21. Kim SW, Han HH, Seo JW,et al. Two cases of lower body contouring with a spiral studies. A more in-depth evaluation of the anatomic and biomechanical and vertical medial thigh lift. Arch Plast Surg. 2012;39:67–70. properties within the thigh region may facilitate the design of optimal 22. Abu-Hijleh MF, Roshier AL, Al-Shboul Q, et al. The membranous layer of super- suturing techniques for individualized patient surgical plans. ficial fascia: evidence for its widespread distribution in the body. Surg Radiol Anat. 2006;28:606–619. 23. Markman B, Barton FE Jr. Anatomy of the subcutaneous tissue of the trunk and ACKNOWLEDGMENTS – The authors thank Stephanie Preston for her contribution and cre- lower extremity. Plast Reconstr Surg. 1987;80:248 254. ation of the artwork represented in Figure 1. 24. Tobin CE, Benjamin JA. Anatomic and clinical re-evaluation of Camper's, Scarpa's, and Colles' fasciae. Surg Gynecol Obstet. 1949;88:545–559. 25. Lancerotto L, Stecco C, Macchi V,et al. Layers of the abdominal wall: anatomical REFERENCES investigation of subcutaneous tissue and superficial fascia. Surg Radiol Anat. – 1. Kolker AR, Xipoleas GD. The circumferential thigh lift and vertical extension cir- 2011;33:835 842. cumferential thigh lift: maximizing aesthetics and safety in lower extremity 26. Gray H, Clemente C. Anatomy of the Human Body. 30th ed. Philadelphia, PA: contouring. Ann Plast Surg. 2011;66:452–456. Lea & Febiger; 1985.

Plasticity of Auricular Cartilage in Response to Hormone Therapy Patrick S. Cottler, PhD, Matthew D. McLeod, MD, MPH, Jesse I. Payton, BS, Angela Pineros-Fernandez, MD, and Jonathan S. Black, MD

Cartilage elasticity has been shown to be dependent on the con- Introduction: Correction of auricular deformities can be accomplished through centration of the proteoglycan aggregate, and hyaluronic acid, a con- splinting within the first few weeks of life. This is hypothesized to be due to stituent of proteoglycan aggregate, is increased by estrogens.6 The retained circulating maternal estrogens decreasing the structural density of colla- literature hypothesizes that remaining maternal estrogen within the gen; however, this has not been fully tested. Cartilage elasticity is dependent on first weeks of life increases hyaluronic acid and other extracellular the concentration of the proteoglycan aggregate, and hyaluronic acid, a constitu- components leading to a decrease in elasticity and an increase in car- ent of proteoglycan aggregate, is increased by estrogens. Nonsurgical correction tilage malleability.7–9 This transient malleability of the auricular carti- of these deformities in more developed patients has the potential to change clin- lage is the mechanism behind the early success of external molding. ical practice and eliminate surgical risks. Previous studies have demonstrated pre- Though widely accepted, this hypothesis has not been fully tested, liminary promise with the use of injectable estrogen to treat auricular deformities. and there is limited literature on the topic. Previous studies have For this study, we have validated an animal model and demonstrated the feasibil- shown chemical component changes in the auricular cartilage matrix ity of a more therapeutically appropriate topical estrogen treatment in restoring and ear appearance in response to the injections various steroid hor- neonatal plasticity of auricular cartilage. mones including estrogen7,9 as well as increased malleability of ear Methods: Ears of 12 New Zealand rabbits were folded and splinted, and assigned cartilage in response to enzymatic treatment targeting hyaluronic acid an experimental group (estrogen, placebo, and untreated control) (n = 8 ears). and elastin.10 Though these studies provide the foundation for the hor- Treatment ears received topical estrogen or placebo cream daily for 4 weeks, monal treatment of auricular deformities, injection-based delivery whereas controls received no treatment. The splints were removed following 2 ad- limits the clinical relevance of this approach. ditional weeks, and photographs were taken to calculate the retained fold angle. Nonsurgical correction of auricular deformities beyond the neo- Biopsies were also taken for histologic analysis. natal period has the potential to change clinical practice and eliminate Results: The 8 control ears showed a statistically increased angle from a folded surgical risks. The purpose of this study is to further validate an animal orientation of 46.6 degrees to return of ear position to a normal upright position model to evaluate the efficacy and mechanisms of exogenous hormonal of 151.2 degrees by the fourth day after splint removal. Both the estrogen-treated therapy to the external ear and to achieve plastic deformation of the and placebo-treated ears responded to splinting with maintained folding (36.6 degrees auricular cartilage for effective ear recontouring. Topical estrogen and 32.5 degrees, respectively). Auricular cartilage thickness trended toward was then investigated in this model as a potential agent to promote car- thicker in ears treated with estrogen, consistent with increased matrix components. tilage plasticity to test of the hypothesis that effective nonsurgical Conclusions: Estrogen and placebo treatment with splinting of ears lead to a sig- recontouring of auricular cartilage is due to circulating residual mater- nificant change to the cartilage configuration, validating the model. The results of nal estrogens. The results can be compared with previous reports of this study are very encouraging and provide the foundation for a noninvasive ther- successfully altering auricular cartilage by increasing the hexosamine apeutic approach for correcting auricular deformities. Future work will include a content in the chondromucoprotein compared with controls with intra- more detailed mechanistic study evaluating the dosing of estrogen and the effi- muscular estrogen injection.7 ciency of dermal penetration as well as evaluating the long-term outcomes and molecular mechanism-associated cartilaginous responses to estrogen. Key Words: auricular deformities, nonsurgical correction, estrogen METHODS Animal experiments were performed under a protocol approved (Ann Plast Surg 2017;78: S311–S314) by the University of Virginia Institutional Animal Care (Animal Welfare Assurance A3245-01) and Use Committee in accordance with the very year approximately 1 of 3800 infants are born with an ear mal- 1 National Institutes of Health's Guide for the Care and Use of Laboratory E formation and 1 of 6000 has a defect of the outer ear. Correction of Animals. Rabbits were housed in an Association for Assessment and external auricular deformities is possible through nonsurgical molding 2–4 Accreditation of Laboratory Animal Care-accredited facility. All proce- within the first 3 weeks of life. Molding with taping or commercially dures were performed under anesthesia, and all efforts were made to available devices has not been successful beyond this early neonatal minimize pain or suffering. period.4 Beyond the first weeks of life, surgical correction is the only reliable solution; however, it is not performed until the age of 5 Splint Placement to7years.5 Twelve male New Zealand White Rabbits (2–2.5 kg) were used for this study due to their easily manipulated ears, the similarity of car- 10 Received October 7, 2016, and accepted for publication, after revision December 6, 2016. tilage to human ears, and a similar previously described model. On From the Department of Plastic Surgery, University of Virginia Medical Center, the day of the procedure, animals were anesthetized to a surgical plane Charlottesville, VA. with inhaled isoflurane (2–2.5%), and each ear was shaved, depilated, Conflicts of interest and sources of funding: None declared. and prepped for aseptic surgery with iodophor solution and 70% alco- This article was presented in the resident section of the 59th Annual Meeting of the Southeastern Society of Plastic and Reconstructive Surgeons on June 12, 2016 hol, and draped for surgery. in Lake Buena Vista, FL. Each ear was folded 5 cm from its tip. A 2-0-Prolene suture Reprints: Patrick S. Cottler, PhD, Department of Plastic Surgery University of (Ethicon, Somerville, NJ) was then passed through a cotton dental roll Virginia Health System P.O. Box 800376 Charlottesville, VA 22908. E-mail: on the medial dorsal side of the ear, through the now adjacent proximal [email protected]. Copyright © 2017 Wolters Kluwer Health, Inc. All rights reserved. and distal ear, and then through another dental roll on the ventral side of ISSN: 0148-7043/17/7806–S311 the ear. The suture followed the reverse path on the lateral side of the ear DOI: 10.1097/SAP.0000000000001032 and was tied so that the ear would retain a near linear conformation.

Annals of Plastic Surgery • Volume 78, Supplement 5, June 2017 www.annalsplasticsurgery.com S311 Cottler et al Annals of Plastic Surgery • Volume 78, Supplement 5, June 2017

Topical Treatments

The 12 animals were divided into 2 groups. The initial 4 animals (n = 8 ears) were used as an untreated control group. Splints were maintained for 6 weeks, and the ears did not receive any treatment. At the completion of the 6 weeks, animals were reanesthetized and the splints were removed and digital photographs were taken. Control ears were also photographed daily for 6 days after the removal of the splints. After the final photograph, animals were euthanized through an anesthetic overdose and 3-mm biopsies were then taken through the entire ear thickness at the medial and lateral aspect of the auricular cartilage of the fold of each ear and fixed in 10% formalin for histological processing. The remaining 8 animals were prepped and received bilateral splints in the same fashion as described, and used to compare topical estrogen treatment to a placebo. 0.5 g of topical estrogen (Premarin [0.625 mg/g]; Pfizer NY, NY) was applied to the external surface of the left ear of each rabbit along the created fold just superior to its splint and gently massaged into the skin (n = 8 ears). Additionally, 0.5 g of a placebo cream with similar inert ingredients to Premarin (CVS Eczema Cream, CVS, Woonsocket, RI) was applied the created fold of the right in the same fashion (n = 8 ears). Both ears underwent daily application for 4 weeks. After 4 weeks of treatment, the splints were maintained for an additional 2 weeks without topical therapy. The 6 weeks of splinting is consistent with previously described suc- FIGURE 1. Ear with the cotton dental roll and monofilament 4,8 splint to create the auricular fold. cessful splinting techniques in neonates. At the conclusion of 6 weeks, animals were reanesthetized, the Splints were placed bilaterally on both ears on each animal (Fig. 1). An- splints were removed, digital photographs were taken, and animals were imals were then placed in soft Elizabethan collars for 2 to 3 days euthanized through an anesthetic overdose. Three-millimeter biopsies postprocedure to ensure the splints maintained position. Animals were then taken through the entire ear thickness at the medial and lateral were given postoperative analgesics and allowed to fully recover aspect of the auricular cartilage of the fold of each ear and fixed in 10% from anesthesia. formalin for histological processing.

FIGURE 2. Demonstration of how the postsplinting angles were defined on the ears.

Treatment Analysis To determine the effect of the splinting and treatment, the angle of each ear was calculated by analyzing the digital images using ImageJ photo analysis software (Fig. 2).11 The fixed tissue biopsies were sectioned into 5 mm slices and stained with hematoxylin-eosin, and 20Â images were obtained. The thickness of the cartilage was also calculated using ImageJ software.

RESULTS Splints were successfully placed on all the ears of the 12 animals and were able to be maintained for 6 weeks (Fig. 1). All ears maintained an acute angle after splint removal without immediate rebound. At the time of splint removal, there was no statistical difference in the treatment arm between the untreated control ears (46.6 ± 10.8 degrees), the placebo-treated ears (32.5 ± 20.8 degrees), and the estrogen-treated ears (36.6 ± 11.4 degrees) (Fig. 3). The angle of the untreated control ears increased daily and returned to a presplinted orientation (151.2 ± 9.3 degrees) after 4 days (Fig. 4) with no additional change FIGURE 4. Demonstration of the maintained elasticity of the seen out to 7 days (data not shown). nontreated control ears. Four days after the splints were The auricular cartilage in the rabbit ear thins as it transitions from removed, the ear angle was able returned to a more normal the medial to the lateral aspect. To determine if the cartilage thickness was configuration. *Statistically significant increase in ear angle altered due to increased extracellular matrix deposition, measurements (P <0.05). were taken at both the lateral and medial aspects of each ear. Though 15,16 not statistically significant, the cartilage trended toward being thicker in many parents to pursue correction. Earlier treatment in with non- the medial cartilage samples treated with estrogen (400.9 ± 147.4 μm), invasive methods would address many of the surgical disadvantages. compared with the placebo-treated samples (292.0 ± 94.0 μm) and un- Nonsurgical molding of auricular deformities has reported suc- 2–5,7,8,17–19 treated control samples (261.2 ± 62.8 μm), as well as the lateral cartilage cess in the neonatal period. Remaining maternal estrogen is samples treated with estrogen (274.8 ± 99.7 μm), compared with the thought to be responsible for this plasticity but scientific data are 8,9 placebo-treated samples (213.3 ± 67.2 μm) and untreated control samples scarce. Previous models exist documenting change in ear shape using 7 6,9 (204.8 ± 57.7 μm) (Fig. 5). intramuscular and direct estrogen injections. This demonstrates promise in altering the malleability of ear cartilage, but specifics regarding the animal models used are unclear. Additionally, these reports DISCUSSION using an invasive injection of the treatment modality provide limited Surgical molding of auricular deformities involves invasive su- data as to the long-term efficacy. Though Oh et al6 used an ointment de- turing methods combined with cartilage excision and/or scoring. These livery of estrogen to ear deformities, it is unknown whether topical techniques provide reliable correction, but not until the ear is near adult methods of delivery for these treatments would be efficacious and pro- size (age, 5–7years).12 Surgery has numerous disadvantages beyond vide a clinically applicable long-term noninvasive approach. the delay of repair including infection, hematoma, and most commonly The control arm of our study further clarified the animal model recurrence of the deformity.13 Cost is a significant concern as well be- needed to study ear plasticity. At the end of 6 weeks of splinting, untreated cause most insurance carriers classify repair as “cosmetic.”14 The psy- control ears exhibited an immediate preservation of the acute splinting chosocial consequences of severe deformities can be profound leading angle. However, there was not any long-term plastic recontouring,

FIGURE 3. The average angle of the ears for the control, placebo-treated, and estrogen-treated ears immediately after FIGURE 5. The average thickness of the auricular cartilage the splints were removed. There was not any statistical difference trended toward being thicker in the cartilage samples treated between the groups (P <0.05). with estrogen. No significant differences were seen (P <0.05).

because this effect was temporary in the untreated control ears as they cartilage plasticity would provide exciting improvements to our knowl- returned to a normal configuration by the fourth day after the splints edge of cartilage mechanics and better treatment options for children. were removed (Fig. 4). These data provide a clear foundation for the animal model for future study looking at lasting change to ear shape. ACKNOWLEDGMENT Comparisons to treatments will provide evidence of their efficacy. The authors would like to thank Lisa Salopek, LVT, RLATG for Both the estrogen-treated and placebo-treated ears demonstrated the excellent standard of care given to all the animals used during the acute effect from splinting. The ear angle from each treatment group this study. was not statistically different from the control group immediately after splint removal (Fig. 3); however, it is unclear whether that effect is permanent. It is also unclear whether the placebo ears would demonstrate REFERENCES same shape as the treatment ears over time. It is possible that the estro- 1. Bartel-Friedrich S, Wulke C. Classification and diagnosis of ear malformations. gen application on one ear led to a systemic treatment effect on the GMS Curr Top Otorhinolaryngol Head Neck Surg. 2007;6:Doc05. placebo-treated ear. This potential systemic effect of estrogen needs to 2. Yotsuyanagi T, Yokoi K, Sawada Y. Nonsurgical treatment of various auricular deformities. Clin Plast Surg. 2002;29:327–332. be further evaluated. Additionally, the dosing of the estrogen will need 3. Yotsuyanagi T, Yokoi K, Urushidate S, et al. Nonsurgical correction of congenital further investigation. To be a clinically relevant therapeutic approach, auricular deformities in children older than early neonates. Plast Reconstr Surg. the application of estrogen would need to be topically applied; there- 1998;101:907–914. fore, these future dosing studies will include more detailed evaluation 4. Byrd HS, Langevin CJ, Ghidoni LA. Ear molding in newborn infants with of the dermal penetration. auricular deformities. Plast Reconstr Surg. 2010;126:1191–1200. Though the thickness of the cartilage was not statistically dif- 5. Janis JE, Rohrich RJ, Gutowski KA. Otoplasty. Plast Reconstr Surg. 2005;115: ferent between ears, the trend toward increased thickness in the 60e–72e. estrogen-treated ears could indicate it is consistent with increased levels 6. Oh K, Kim JH, Kim JT, et al. The effect of estrogen on transformation of rabbit ear. J Korean Soc Plast Reconstr Surg. 1999;26:99–104. of extracellular matrix components such as hyaluronic acid as has been 6,8,9 7. Kyriazis AP, Tsaltas TT. Changes in the chemical composition of ear cartilage previously suggested. matrix after administration of various steroid hormones. Studies in New There were several limitations to this initial study evaluating a Zealand albino rabbits. Am J Pathol. 1971;63:149–160. model to test treatments for postneonatal nonsurgical ear contouring. 8. Matsuo K, Hayashi R, Kiyono M, et al. Nonsurgical correction of congenital In efforts to determine the microscopic structure of the cartilage due auricular deformities. Clin Plast Surg. 1990;17:383–395. to treatment, biopsies were taken immediately after the splints were re- 9. Matsuo K, Hirose T, Tomono T, et al. Nonsurgical correction of congenital auric- moved. Therefore, the long-term effects of the treatments are still un- ular deformities in the early neonate: a preliminary report. Plast Reconstr Surg. 1984;73:38–51. clear. However, this study provides an excellent foundation for future 10. Massengill PL, Goco PE, Norlund LL, et al. Enzymatic recontouring of auricular mechanistic studies on auricular cartilage plasticity. cartilage in a rabbit model. Arch Facial Plast Surg. 2005;7:104–110. There is much promise in the use of estrogen to allow for in- 11. Schneider CA, Rasband WS, Eliceiri KW. NIH Image to ImageJ: 25 years of creased ear malleability beyond the first weeks of life. We plan future image analysis. Nat Methods. 2012;9:671–675. study to evaluate the many possibilities we have seen in this pilot 12. Pawar SS, Koch CA, Murakami C. Treatment of prominent ears and otoplasty: a study. Using our validated animal model, we plan to evaluate the last- contemporary review. JAMA Facial Plast Surg. 2015;17:449–454. ing effect of estrogen treatment on ear shape by following the animals 13. Staindl O, Siedek V. Complications of auricular correction. GMS Curr Top over time without disruption of their ear architecture through tissue Otorhinolaryngol Head Neck Surg. 2007;6:Doc03. harvest. We plan to evaluate the ear cartilage and systemic blood for 14. Volk A, Agullo FJ, Palladino H. Early detection of auricular deformities in newborns may allow non-surgical correction.pdf. El Paso Physician. 2014;37:8–10. estrogen levels with a comparison to injectable estrogen. This would 15. Brent B. Ear deformity can sap child's self-esteem: auricular construction restores aid in determining systemic effect of estrogen and efficacy of topical psychological well-being. AAP News.1998;14:14–15. application in dermal penetration. 16. Petersson RS, Recker CA, Martin JRK, et al. Identification of congenital auricular deformities during newborn hearing screening allows for non-surgical correction: CONCLUSIONS a Mayo Clinic pilot study. Int J Pediatr Otorhinolaryngol. 2012;76:1406–1412. 17. Razook SJ. Nonsurgical treatment as an adjunct to surgery. Oral Maxillofac Surg We have presented an improved, validated animal model for the Clin North Am. 2006;18:305–310. study of auricular molding through external approaches. Although the 18. Thorne CH, Brecht LE, Bradley JP,et al. Auricular reconstruction: indications for results are promising, further study is needed to determine the efficacy autogenous and prosthetic techniques. Plast Reconstr Surg. 2001;107:1241–1252. of dermal penetration of estrogen to the perichondrium and if the effect 19. Ullmann Y, Blazer S, Ramon Y, et al. Early nonsurgical correction of congenital of estrogen is long lasting. The ability to extend the period of neonatal auricular deformities. Plast Reconstr Surg. 2002;109:905–907.

Abdominal Wall Allograft Preclinical Biomechanical Investigation of a Novel Reconstructive Adjunct Brent R. DeGeorge, Jr, MD, PhD, Jonathan Bryce Olenczak, MD, Angela Pineros-Fernandez, MD, Raymond F.Morgan, MD, Patrick S. Cottler, PhD, and David B. Drake, MD

abdominal wall tissues or bacterial burden make successful repair Introduction: Acellular dermal matrices have revolutionized abdominal wall re- increasingly difficult.6,7 construction; however, device failure and hernia recurrence remain significant In accordance with concepts of the reconstructive ladder, when problems. Fascia grafts are a reconstructive adjunct with increased tensile strength primary closure of the abdominal wall cannot be achieved, the technique compared with acellular dermal matrices; however, clinical use is limited by in- of local tissue rearrangement with component separation or its deriva- sufficient donor material and donor site morbidity. To this end, we investigate tives provides a means for abdominal wall reconstruction.8–10 However, the biomechanical properties of human abdominal wall allografts (AWAs) when these techniques have been exhausted, the reconstructive surgeon consisting of the anterior rectus sheath from xiphoid to pubis. is faced with a substantial dilemma in restoring the continuity of the ab- Methods: After cadaveric procurement of 6 human AWAs, the tissue was divided dominal wall with tissue of adequate tensile strength and revasculariza- horizontally and a matched-sample study was performed with specimens ran- tion potential.11,12 The emergence of acellular dermal matrices (ADMs) domized to 2 groups: fresh, unprocessed versus processed with gamma irradia- as a key component of the surgical armamentarium of the reconstruction and decellularization. Specimens were evaluated for physical properties, tive surgeon has revolutionized the management of complex abdominal DNA content, tensile strength, and electron microscopy. wall defects. Acellular dermal matrices provide an ample tissue source Results: All AWA donors were male, with a mean age of 55.2 years (range, when autologous tissues are unavailable and offer a construct resistant 35–74 years). Procured AWAs had a mean length of 21.70 ± 1.8 cm, width of 2 to infection when reconstruction is performed in a contaminated envi- 14.30 ± 1.32 cm, and area of 318.50 cm , and processing resulted in a 98.3% ronment.13,14 Manufacturers have utilized different proprietary tech- reduction in DNA content. Ultimate tensile strength was significantly in- niques for tissue processing, decellularization, and sterilization, to creased after tissue processing, and after subcutaneous implantation, proc- optimize the ADM properties, as well as host interaction with respect essed AWA demonstrated 4-fold increased tensile strength compared with to biocompatibility and foreign body response, to facilitate maximum unprocessed AWAs. ADM cellular incorporation and neovascularization. These biomaterial Conclusions: Acellular AWAs represent a novel reconstructive adjunct for ab- “ ” scaffolds serve as a regenerative framework to support host cell integra- dominal wall reconstruction with the potential of replacing like with like with- tion and new collagen deposition.15–17 Furthermore, as neovasculariza- out additional donor site morbidity or antigenicity. tion occurs and these materials become incorporated by the host, they Key Words: acellular, abdominal wall allograft, abdominal wall reconstruction, show an increase in tensile strength over time. Whereas the application biomechanical properties of ADMs has been a major advancement in the field of abdominal wall reconstruction, the incidence of complications including hernia recur- (Ann Plast Surg 2017;78: S315–S321) rence remains clinically significant.18–22 ndications for abdominal wall reconstruction include trauma, necro- Because they are derived from the dermis, ADMs have a struc- I tizing soft tissue infection, tumor resection, intra-abdominal sepsis, ture and biomechanical profile different from that of abdominal wall and incisional hernia repair.1 With over 300,000 ventral hernia repairs fascia. Although ADMs have well-established benefits in resistance to performed each year in the United States, and with a hernia recurrence infection and biocompatibility, the innate microstructure of dermis is rate of 5% to 40%, there is substantial opportunity to reduce morbidity not ideally suited to the functional stresses exerted on the abdominal through advances in hernia repair and abdominal wall reconstruction.2–4 wall, which include constant loading from intra-abdominal pressure When reconstructing the abdominal wall, the general and plastic sur- as well as cyclical loading with respiration and during strenuous phys- geons often face a hostile reconstructive environment with a paucity ical activity. The impact of these forces on the tensile durability of of good quality autologous tissue or contamination of the surgical field. ADMs has been clinically borne out in multiple studies that document a substantial rate of hernia recurrence and abdominal wall laxity with The technical challenge of abdominal wall reconstruction is further 23–25 compounded by comorbidities in this patient population that contribute use of ADMs in abdominal wall reconstruction. Before the introduction of ADMs, there was a long history of ab- to a high rate of postoperative complications including bulging, hernia 26–28 recurrence, and wound infection—each of which exceeds 20%.5 These dominal wall reconstruction with fascia flaps. The indications for fascia flaps are similar to those for ADMs, including deficient autolo- complications often necessitate reoperation in an environment more 29–31 challenging than that experienced at the index surgery, where inadequate gous tissue and a contaminated surgical field. Animal models of abdominal wall reconstruction with fascia flaps have shown no decrease in tensile strength at 1 year, and human abdominal wall recon- Received October 19, 2016, and accepted for publication, after revision December 6, 2016. struction with fascia lata allografts has been performed with no major From the Department of Plastic Surgery, University of Virginia Medical Center, signs of recurrence, laxity, or infection at a mean follow-up of approx- Charlottesville, VA. imately 2 years.32,33 With the goal of improving options and outcomes Conflicts of interest and sources of funding: LifeNet performed allograft processing at no cost to the authors. No external funding was necessary for the execution of in abdominal wall reconstruction, we performed a preclinical biome- this project. chanical investigation of an acellular abdominal wall allograft (AWA). Reprints: Patrick S. Cottler, PhD, Department of Plastic Surgery, University of This novel reconstructive adjunct has the potential to not only reconsti- Virginia Health System, PO Box 800376, Charlottesville, VA 22908. E-mail: tute deficient tissue on a gross anatomic scale but also restore the struc- [email protected]. Copyright © 2017 Wolters Kluwer Health, Inc. All rights reserved. tural properties that are essential to long-term durable function of the ISSN: 0148-7043/17/7806–S315 abdominal wall while avoiding the problems of antigenicity and donor DOI: 10.1097/SAP.0000000000001029 site morbidity.

Annals of Plastic Surgery • Volume 78, Supplement 5, June 2017 www.annalsplasticsurgery.com S315 DeGeorge et al Annals of Plastic Surgery • Volume 78, Supplement 5, June 2017

MATERIALS AND METHODS TABLE 1. AWA Physical Properties

AWA Procurement Demographics of Donor Tissue Six AWAs were procured from organ donors according to guide- n8 lines for human cellular and tissue-based products. Briefly, the anterior Age, y 55.2 ± 5.6 rectus sheath fascia was isolated from the xyphoid to the pubis and laterally from semilunar line to semilunar line. The linea alba was pre- Sex (% male) 100.0 served during isolation of the construct (Fig. 1). The constructs were then characterized for maximum length, maximum width, surface Before tissue processing Mean SEM area, and weight in grams (Table 1). The constructs were then divided Length, cm 21.7 1.80 transversely perpendicular to the linea alba, and transferred to LifeNet Width, cm 14.3 1.32 Health on ice for proprietary tissue processing according to approved Surface area, cm2 318.5 47.34 protocols (Fig. 1). Mass, g 41.2 5.30 DNA, ng DNA/mg 81.2 12.10 After tissue processing Experimental Groups and Tissue Processing % DNA reduction 98.3 0.23 Two experimental groups were investigated—control AWAs (Fig. 1; right panel, top) and Matracell-processed acellular AWAs SEM, standard error of the mean. (Fig. 1; right panel, bottom). Both groups are handled in a similar fashion including freezing and gamma irradiation with the exception of Matracell tissue processing to directly compare the effects of tissue Tissue Response to Implantation of AWA acellularization on the physical and mechanical properties of the AWA constructs. After cadaveric procurement of human AWAs, the tis- Animal experiments were performed under a protocol approved sue was sent on ice to LifeNet Health for tissue processing using by the University of Virginia Institutional Animal Care (Animal Wel- gamma irradiation for sterilization as well as Matracell tissue process- fare Assurance No. A3245-01) and Use Committee in accordance with ing, which is an efficient 2-day decellularization protocol, which dem- the National Institutes of Health's Guide for the Care and Use of Labo- onstrates a more than 98% reduction in donor DNA with multiple ratory Animals. Rats were housed in a facility accredited by Association disinfecting agents targeting both cellular and genetic components to for Assessment and Accreditation of Laboratory Animal Care. All pro- produce a safe bio-implant. The protocol includes validated United cedures were performed under anesthesia, and all efforts were made to States Pharmacopeia chapter 71 endpoint microbial testing and does minimize pain or suffering. Abdominal wall allograft specimens, proc- not utilize any animal-derived reagents. The tissue processing protocol essed and unprocessed, were thawed in room temperature phosphate has been used previously for commercially available products for ortho- buffered saline (1Â), and three 1.0 Â 5.5 cm samples of each specimen pedic, cardiac, and vascular procedures. were sharply cut perpendicular to the linea alba of the donor and placed into an iodophor solution. Three female Sprague-Dawley rats (250–300 g) were utilized for this portion of the study. Animals were DNA Quantification of AWAs anesthetized and maintained under inhaled isoflurane (2%–2.5%). Once The DNA reduction after tissue processing and acellularization anesthetized, the animals were positioned supine, shaved, depilated, were assessed on 1 Â 1 cm sections of AWA constructs for DNA using and prepared for aseptic surgery with iodophor scrub followed by the DNeasy kit (QIAgen). The DNA concentration was determined 70% alcohol and iodophor solution, and then draped for surgery. A with a spectrophotometer and was normalized to AWA dry weight, as 4 cm midline skin incision was created in the mid abdomen and the dis- well as compared with the unprocessed and acellular AWA groups. section proceeded through the panniculus carnosus muscle to expose

FIGURE 1. Abdominal wall allografts were procured from 8 human donors. AWA before tissue processing (left). AWA is sectioned transversely perpendicular to the linea alba and randomized to 1 of 2 groups (right): (1) unprocessed, frozen; and (2) processed, − Allowash XG (LifeNet Health) low-temperature gamma irradiation, 25 kGy, with a sterilization assurance level of 10 6.

FIGURE 2. Processed and unprocessed AWAs were subjected to mechanical strength testing. Processed AWAs demonstrated a significant increase in ultimate tensile strength (P < 0.05, n = 6 per group). the abdominal wall fascia and linea alba. Samples of human AWA (1.0 Â 5.5 cm2) that had been soaking in iodophor solution for 10 minutes and rinsed with sterile water were placed longitudinally on the anterior surface of the abdominal wall and secured proximally and distally 4-0 Vicryl in interrupted fashion at the cranial and caudal ends. Unprocessed AWA samples were placed into the animals' right side, and processed AWA samples were placed into the in the animals' left side. After implantation, the skin flaps were reapproximated in a tension free manner overlying the implanted AWAs and secured with 4–0 Prolene sutures and the animal was circumferentially bandaged. Animals were also given a subcutaneous injection of buprenorphine (0.5 mg/kg) for pain, recovered and returned to their cages. Female animals were utilized to allow for ease of urination with the bandaging. One week after implantation, the bandages and sutures were removed. Three weeks after implantation, the animals were reanesthetized with inhaled isoflurane (2%–2.5%) and the midline was reopened and the skin was carefully dissected and retracted to expose the AWA implants. The implants were sharply excised. A 2- to 3-mm strip was sharply dissected from the short axis and placed into 10% formalin for histology and scanning electron microscopy. The remainder of the specimen was utilized for material strength testing. Animals were then euthanized with an anesthetic overdose of Euthasol.

Characterization of the Effect of Mode of Tissue Processing on AWA Tensile Strength FIGURE 3. Processed and unprocessed AWAs were subjected to The AWA constructs were subjected to repeat baseline character- scanning electron microscopy to determine the effect of tissue ization after tissue processing, and then, they were dissected into 1 cm processing on ultrastructural architecture and collagen by 5 cm strips perpendicular to the linea alba but not including the linea cross-linking. Processed AWAs demonstrated a notable increase alba in the specimen. Mechanical tensile strength testing was performed in collagen cross-linking with a significant increase in the number on an Instron mechanical tester (Model No. 5943) equipped with a of collagen branch points as well as an increase in the density of 100-N load cell (Model No. 2530-427) and 1 kg screw action clamps collagen fibers per high-power field. Representative images depict (Norwood, Mass). The AWA strips were loaded onto a 4 Â 2cm2 sheet the collagen ultrastructure at 500Â and 5000Â, respectively.

of 100-grit sandpaper pretreated with cyanoacrylate. The AWA strips were imaged on a Zeiss Sigma HD scanning electron microscope at within the clamp segments were covered with cyanoacrylate followed a voltage of 3.0 kV and a working distance of 13.4 mm. by sandpaper folded over to ensure solid clamp contact. The AWAstrips were loaded into mechanical clamps with a gauge length of 20 mm and Histologic Characterization of AWA clamp length of 15 mm for the studies before tissue implantation. For AWA constructs (1 Â 1 cm sections) were formalin fixed and studies performed after subcutaneous implantation in a rat model, be- embedded in paraffin. Sections with a concentration of 5 μM were cause of tissue availability, samples were loaded with a 1-cm clamp subjected to staining with hematoxylin and eosin to cellular infiltration length and 20 mm gauge length. Specimens were prestretched with after implantation and compared with the unprocessed and acellular 10 cycles of 5 mm strain at 10 mm/s and then pulled to failure at AWA groups. Cell counting and quantification were performed using 100 mm/min to ensure a midsubstance tear. End of test was determined ImageJ software. by 80% decrease from peak load or maximum 10 cm of extension. Pri- mary outcomes include maximum load (N), tensile stress at maximum Statistical Analysis load (MPa), and elastic modulus. All values recorded are presented as mean ± standard error of the mean from independent experiments from given n sizes. Statistical significance of multiple treatments was determined by analysis of variance Scanning Electron Microscopy for followed by the Bonferroni post hoc test when appropriate. Statistical Ultrastructural Characterization significance between 2 groups was determined by using the 2-tailed Samples of unprocessed and processed AWAs before and after Student t test. P values of less than 0.05 are considered significant. implantation were evaluated through scanning electron microscopy to evaluate the microscopic structure. After the AWA samples were fixed in 10% formalin for 5 days, they were washed (3 Â 10 min) with RESULTS 0.1 M cacodylate buffer and then treated with 2% osmium tetroxide for All AWA donors were male, with a mean age of 55.2 years 60 min. After 2 Â 10 min washes with 0.1 M cacodylate buffer and (range, 35–74 years). Physical data for Matracell-processed acellular distilled H2O, the samples were dehydrated in a series of 10-minute AWAs are presented in Table 1. Procured AWAs had a mean length of ethanol treatments (30%, 50%, 70%, 95%, and 100%). Samples were 21.70 ± 1.8 cm, width of 14.30 ± 1.32 cm, and area of 318.50 cm2. then placed into a critical point dryer with a 15-minute purge time and The average weight before and after decellularization was 41.26 g and then mounted on microscope stubs with carbon stickers and sputter 22.42 g, respectively. There was a 98.3% reduction in DNA after the coated with gold, 200 seconds at 60 mA. Once coated, the samples Matracell processing of the AWAs.

FIGURE 4. Processed and unprocessed AWAs were implanted in a suprafascial position on rat abdominal walls and maintained for 4 weeks to assess the effect of tissue processing on mechanical strength, collagen ultrastructure, and recellularization. Abdominal wall allografts were secured in suprafascial fashion to the abdominal wall. After abdominal wall implantation, processed and unprocessed AWAs strength testing was performed. Processed AWAs demonstrated a significant increase in ultimate tensile strength (P <0.05, n = 3 per group).

a significant increase in ultimate tensile strength after tissue processing (P < 0.05, n = 6 per group) (Fig. 2A). Young modulus was 31.39 ± 17.82 mPA for the unprocessed AWA group and 38.25 ± 25.76 mPA for the processed AWA group (Fig. 2B). Maximum extension at failure was 5.91 ± 1.84 mPA for the unprocessed AWA group and 11.46 ± 3.91 mPA for the processed AWA group, which also demonstrated a significant increase after tissue processing (P < 0.05, n = 6 per group) (Fig. 2C). Collagen ultrastructure was assessed at baseline to assess the effect of tissue processing on the composition of the AWAwith scanning electron microscopy (Fig. 3). Processed AWAs demonstrated a notable increase in collagen cross-linking with an increase in the number of collagen branch points as well as an increase in the density of collagen fibers per high-power field (Fig. 3). Mechanical strength testing was then performed after subcutaneous implantation of the AWAs in a rat model as described. After tissue implantation, the maximum load at failure was 0.66 ± 0.15 kgf in the unprocessed AWA group and 2.64 ± 0.54 kgf in the processed AWA group, which demonstrates a significant increase in ultimate tensile strength in the processed group (P < 0.05, n = 6 per group) (Fig. 4). Both groups experienced a decrease in maximal tensile strength after 3 weeks of subcutaneous implantation; however, the effect was significantly decreased in the processed AWA group. After abdominal wall implantation, the processed and unprocessed AWAs were procured and were subjected to scanning electron FIGURE 5. After abdominal wall implantation, the processed and microscopy to determine the effect of tissue processing on ultrastruc- unprocessed AWAs were procured and were subjected to tural architecture. Processed AWAs demonstrated a notable increase scanning electron microscopy to determine the effect of tissue in surface cellularity (Fig. 5). Collagen architecture is maintained in processing on ultrastructural architecture. Processed AWAs both groups after 3 weeks of implantation. To further confirm the find- demonstrated a notable increase in surface cellularity. Collagen ings of increased surface cellularity, histologic analysis was performed architecture is maintained in both groups after 4 weeks of after 3 weeks of subcutaneous implantation. After abdominal wall implantation. Representative images depict the collagen implantation, processed AWAs demonstrate a significant increase ultrastructure at 500Â and 5000Â, respectively (n = 3). in cellularity in both the center of the construct (0.59 ± 0.28% and 2.34 ± 0.77% for unprocessed AWA and processed AWA, respec- Mechanical testing was performed on matched unprocessed tively) and the AWA-host tissue border zone (3.53 ± 1.61% and and processed AWAs before soft tissue implantation (Fig. 2). The 11.61 ± 4.31% for unprocessed AWA and processed AWA, respec- maximum load was 4.72 ± 1.80 kgf in the unprocessed AWA group tively) (P < 0.05, n = 3); however, both groups demonstrated the and 6.8 ± 0.74 kgf in the processed AWA group, which demonstrates presence of basophilic cells throughout the construct (Fig. 6).

FIGURE 6. After abdominal wall implantation, processed AWAs demonstrate a significant increase in cellularity in both the center of the construct and the AWA-host tissue border zone (P < 0.05, n = 3).

DISCUSSION thickness of the abdominal wall, in contrast to ADM, which can be procured in thicker or thinner fashion to meet specific reconstructive needs. Given the substantial problem posed by hernia recurrence, with Furthermore, the procurement of AWA fascia is not practical in donors rates approaching 40% in the literature, advances in the field of ab- intended for solid intra-abdominal organ donation, because procure- dominal wall reconstruction have the potential to impact the care of ment of the fascia would delay procurement of liver, kidney, or pancreas thousands of patients annually.2–4 The availability of suitable autolo- tissues by the organ procurement team. gous tissues with the tensile and cellular properties necessary to facilitate healing of the abdominal wall repair and withstand the forces CONCLUSIONS applied to the abdominal wall during the recovery period is generally Acellular AWAs have potential to serve as a novel reconstruc- deficient. This fact is often compounded by a patient population with tive adjunct for abdominal wall reconstruction with the potential of re- multiple comorbidities and previous operative procedures, which fur- placing “like with like” without additional donor site morbidity or ther compromise available tissues, leading to a postoperative major 5 antigenicity. There was a decrease in maximum load at failure seen af- complication rate approaching 20%. Acellular dermal matrices serve ter implantation. Therefore, further testing is required in a clinically as a regenerative biomaterial scaffold to support host cell integration relevant model of abdominal wall repair to assess its biomechanical and collagen remodeling and are used in a variety of reconstructive properties after implantation in comparison with existing modalities surgical applications. Acellular dermal matrices have been processed for abdominal wall reconstruction. to render the tissues sterile and acellular to facilitate biocompatibility and revascularization. The ADMs for abdominal wall reconstruction has revolutionized the management of complex abdominal wall de- ACKNOWLEDGMENTS The authors wish to thank LifeNet Health for harvesting and pro- fects with the provision of an ample tissue source when autologous tis- cessing the abdominal wall fascia samples used during this study. We sues are insufficient. However, ADMs are derived from dermis and, would also like to thank Lisa Salopek, LVT, RLATG, for the excellent therefore, have mechanical properties disparate from that of native ab- standard of care given to all the animals used during this study. dominal wall fascia. In search of a way to replace abdominal wall fascial defects in “like with like” fashion, a variety of autologous approaches using expendable donor fascia through separation of com- REFERENCES ponents for local abdominal wall fascia8,34 or regional flaps using 1. Janis JE, O'Neill AC, Ahmad J, et al. Acellular dermal matrices in abdominal wall thigh-based fascial flaps28 have been proposed. However, these tech- reconstruction: a systematic review of the current evidence. Plast Reconstr Surg. – niques are technically demanding, are associated with increased donor 2012;130(suppl 2):183S 193S. 2. Poulose BK, Shelton J, Phillips S, et al. Epidemiology and cost of ventral hernia site morbidity, and are not always sufficient to restore fascial continu- – 35,36 repair: making the case for hernia research. Hernia. 2012;16:179 183. ity to the abdominal wall. 3. Luijendijk RW, Hop WC, van den Tol MP, et al. A comparison of suture repair To this end, we describe and provisionally characterize the utility with mesh repair for incisional hernia. N Engl J Med. 2000;343:392–398. of human AWA, which consists of the anterior rectus sheath from semi- 4. Millbourn D, Cengiz Y,Israelsson LA. Effect of stitch length on wound complica- lunar line to semilunar line and from xiphoid to pubis, and subjected to tions after closure of midline incisions: a randomized controlled trial. Arch Surg. tissue processing to render it sterile and acellular. We have demonstrated 2009;144:1056–1059. that our tissue processing technique results in increased tensile strength 5. Albino FP, Patel KM, Nahabedian MY,et al. Does mesh location matter in abdominal wall reconstruction? A systematic review of the literature and a summary of compared with unprocessed fascia, which is substantiated ultrastructur- recommendations. Plast Reconstr Surg. 2013;132:1295–1304. ally with the presence of increased collagen cross-linking. We have per- 6. Adetayo OA, Salcedo SE, Bahjri K, et al. A meta-analysis of outcomes using acel- formed preclinical characterization in a rat subcutaneous implantation lular dermal matrix in breast and abdominal wall reconstructions: event rates and model and have demonstrated that processed acellular AWAsretain ten- risk factors predictive of complications. Ann Plast Surg.2011;77:e31–e38. sile strength to a greater degree than unprocessed AWAs and, despite 7. Ghazi B, Deigni O, Yezhelyev M, et al. Current options in the management of the increased collagen cross-linking and preserved tensile strength, complex abdominal wall defects. Ann Plast Surg. 2011;66:488–492. these constructs allow cellular penetration and bio-incorporation. 8. Ramirez OM,RuasE,DellonAL.“Components separation” method for closure of abdominal-wall defects: an anatomic and clinical study. Plast Reconstr Surg. This study is intended to serve as an initial preclinical evaluation 1990;86:519–526. of the biophysical characteristics of AWAs as a potential alternative to 9. Koltz PF, Frey JD, Bell DE, et al. Evolution of abdominal wall reconstruction: de- other adjuncts available for abdominal wall reconstruction. The tissue velopment of a unified algorithm with improved outcomes. Ann Plast Surg.2013; procurement and processing described herein result in the procurement 71:554–560. of abdominal wall fascial grafts of significant size, 318 cm2 in surface 10. Mericli AF, Bell D, DeGeorge BR Jr, et al. The single fascial incision modification area, and relatively uniform thickness. The tissue processing described of the “open-book” component separation repair: a 15-year experience. Ann Plast – herein, is analogous to tissue processing used for commercially avail- Surg. 2013;71:203 208. 11. Rodriguez ED, Bluebond-Langner R, Silverman RP,et al. Abdominal wall recon- able ADMs (DermACELL, LifeNet Health, Inc, Virginia Beach, Va) struction following severe loss of domain: the R Adams Cowley Shock Trauma with similar reduction in total DNA content and alterations in collagen Center algorithm. Plast Reconstr Surg. 2007;120:669–680. ultrastructure. Future studies will elucidate the potential of this con- 12. Bluebond-Langner R, Keifa ES, Mithani S, et al. Recurrent abdominal laxity fol- struct for abdominal wall reconstruction in animal and human models lowing interpositional human acellular dermal matrix. Ann Plast Surg.2008;60: of abdominal wall hernia to determine its bio-incorporation proper- 76–80. ties, suture pullout strength, and mechanical failure strength before 13. Albino FP, Patel KM, Nahabedian MY, et al. Immediate, multistaged approach to infected synthetic mesh: outcomes after abdominal wall reconstruction with por- clinical applications. cine acellular dermal matrix. Ann Plast Surg. 2015;75:629–633. Limitations of the AWA approach include difficulty associated 14. Garvey PB, Martinez RA, Baumann DP,et al. Outcomes of abdominal wall recon- with procuring the abdominal wall without the creation of fascial de- struction with acellular dermal matrix are not affected by wound contamination. fects, in contrast to ADM, which can be procured with the use of a me- JAmCollSurg. 2014;219:853–864. chanical dermatome. Furthermore, donors must be appropriate as all 15. Burns NK, Jaffari MV,Rios CN, et al. Non-cross-linked porcine acellular dermal abdominal walls are not created equal. History of previous abdominal matrices for abdominal wall reconstruction. PlastReconstrSurg. 2010;125: – hernia or bulge, history of intra-abdominal surgery, or obesity should 167 176. 16. Campbell KT, Burns NK, Rios CN, et al. Human versus non-cross-linked porcine be considered contraindications to abdominal wall procurement. The acellular dermal matrix used for ventral hernia repair: comparison of in vivo fibro- size of the abdominal wall is also limited by the size of the donor, and vascular remodeling and mechanical repair strength. Plast Reconstr Surg.2011; more importantly, the thickness of the construct is limited by the 127:2321–2332.

17. Deeken CR, Melman L, Jenkins ED, et al. Histologic and biomechanical evalua- 27. Caffee HH. Reconstruction of the abdominal wall by variations of the tensor fas- tion of crosslinked and non-crosslinked biologic meshes in a porcine model of ciae latae flap. Plast Reconstr Surg. 1983/03/01. 1983;71:348–353. – ventralincisionalherniarepair.JAmCollSurg. 2011;212:880 888. 28. Williams JK, Carlson GW, deChalain T, et al. Role of tensor fasciae latae in 18. Espinosa-de-los-Monteros A, de la Torre JI, Marrero I, et al. Utilization of human abdominal wall reconstruction. Plast Reconstr Surg. 1998/03/21. 1998;101: cadaveric acellular dermis for abdominal hernia reconstruction. Ann Plast Surg. 713–718. 2007;58:264–267. 29. Disa JJ, Klein MH, Goldberg NH. Advantages of autologous fascia versus syn- 19. Patel KM, Nahabedian MY, Gatti M, et al. Indications and outcomes following thetic patch abdominal reconstruction in experimental animal defects. Plast complex abdominal reconstruction with component separation combined with Reconstr Surg. 1996/04/01. 2001;7:2086–2087. porcine acellular dermal matrix reinforcement. Ann Plast Surg. 2012;69:394–398. 30. Disa JJ, Goldberg NH, Carlton JM, et al. Restoring abdominal wall integrity in 20. Sbitany H, Kwon E, Chern H, et al. Outcomes analysis of biologic mesh use for contaminated tissue-deficient wounds using autologous fascia grafts. Plast abdominal wall reconstruction in clean-contaminated and contaminated ventral Reconstr Surg. 1998/03/26. 1998;101:979–986. hernia repair. Ann Plast Surg. 2015;75:201–204. 31. Silverman RP, Singh NK, Li EN, et al. Restoring abdominal wall integrity in con- 21. Rosen MJ, Krpata DM, Ermlich B, et al. A 5-year clinical experience with single- taminated tissue-deficient wounds using autologous fascia grafts. Plast Reconstr staged repairs of infected and contaminated abdominal wall defects utilizing bio- Surg. 2004;113:673–675. logic mesh. Ann Surg. 2013;257:991–996. 22. Patel KM, Bhanot P.Complications of acellular dermal matrices in abdominal wall 32. Matloub HS, Jensen P, Grunert BK, et al. Characteristics of prosthetic mesh and reconstruction. Plast Reconstr Surg. 2012;130(suppl 2):216S–224S. autogenous fascia in abdominal wall reconstruction after prolonged implantation. Ann Plast Surg. 1992/12/01. 1992;29:508–511. 23. Diaz JJ Jr, Conquest AM, Ferzoco SJ, et al. Multi-institutional experience using human acellular dermal matrix for ventral hernia repair in a compromised surgical 33. Tiengo C, Giatsidis G, Azzena B. Fascia lata allografts as biological mesh in ab- field. Arch Surg. 2009;144:209–215. dominal wall repair: preliminary outcomes from a retrospective case series. Plast Reconstr Surg. 2013;132:631e–639e. 24. Kissane NA, Itani KM. A decade of ventral incisional hernia repairs with biologic acellular dermal matrix: what have we learned? Plast Reconstr Surg.2012;130 34. Borud LJ, Grunwaldt L, Janz B, et al. Components separation combined with ab- (suppl 2):194S–202S. dominal wall plication for repair of large abdominal wall hernias following bariatric surgery. Plast Reconstr Surg. 2007;119:1792–1798. 25. Bochicchio GV, De Castro GP, Bochicchio KM, et al. Comparison study of acellular dermal matrices in complicated hernia surgery. JAmCollSurg. 35. Heller L, Chike-Obi C, Xue AS. Abdominal wall reconstruction with mesh and 2013;217:606–613. components separation. Semin Plast Surg.2012;26:29–35. 26. Wagensteen OH. Repair of recurrent and difficult hernias and other large defects 36. De Vries Reilingh TS, Van Goor H, Charbon JA, et al. Repair of giant midline ab- of the abdominal wall employing the iliotibial tract of fascia lata as a pedicled flap. dominal wall hernias: “components separation technique” versus prosthetic repair: Surg Gynecol Obs. 1934;59:766. interim analysis of a randomized controlled trial. World J Surg. 2007;31:756–763.

Identifying Risk Factors Leading to Unanticipated Postoperative Readmission Peter A. Felice, MD,* David T. Kerekes, BS,† and Bruce A. Mast, MD*

program, the Centers for Medicare and Medicaid Services (CMS) have Introduction: Unanticipated postoperative readmissions are a grading metric authority to reduce reimbursement to any CMS-funded institution or in- directly linked to both the quality of patient care and physician reimbursement. dividual found to have an excessive 30-day postoperative readmission However, little data exist to define factors responsible for these readmissions in ratio when compared with the national average of comparable providers the plastic surgery patient population. This study aims to identify patient risk and patients.5–7 factors contributing to unanticipated postoperative readmissions to optimize This study represents an investigation into our own postoperative perioperative patient care and mitigate negative financial impact upon providers. readmission rates to determine whether specific patient risk factors Methods: We present an institutional review board-approved study retrospective contribute to unanticipated postoperative readmissions. Identifying review of 819 plastic surgery patients undergoing operative procedures per- these risk factors is a crucial step toward providing appropriate resources formed at our institution between January 1, 2013, and December 31, 2014. and management for the plastic surgery patient, preventing costly post- All unanticipated readmissions within 30 days of an operation were identified operative readmissions and operative intervention, and satisfying quality and subjected to statistical analysis in an effort to determine whether these read- metrics linked to provider compensation. missions were associated with identifiable patient risk factors. Results: One hundred forty-nine (18.1%) of the 819 investigated patients underwent readmission, reoperation, or both within 30 postoperative days. Seventy-four METHODS (9%) patients required hospital readmission, alone; 55 (6.7%) underwent read- We performed an IRB-approved retrospective review evaluating mission with operative intervention; and 20 (2.4%) required outpatient operative all plastic surgery patients undergoing operative procedures at our insti- intervention without readmission. Readmitted patients were significantly more tution during a 2-year period between January 1, 2013, and December likely to have a positive smoking history (P = 0.009), hypertension (P = 0.0008), 31, 2014. Our patient population represented a broad range of general congestive heart failure (P = 0.0015), chronic obstructive pulmonary disease plastic and reconstructive surgical interventions and included hand, (P = 0.023), a higher mean age (P = 0.0001), and a higher Charlson Comorbidity reconstructive breast, head and neck, microsurgical with free flap, Score (P = 0.0001). and aesthetic/cosmetic procedures performed in both inpatient and Conclusions: These results identify risk factors associated with unanticipated outpatient surgical settings. From a pool of 2,000 patients, 819 were postoperative readmissions specific to a plastic surgery patient population. randomly selected to undergo investigation of their 30-day postoperative With this information, practitioners can allocate appropriate perioperative re- course. Any patient readmitted within 30 days of an operation underwent sources and planning for patients at increased risk for readmission, thereby analysis to determine whether any particular risk factors or demographics improving delivery of patient care and satisfying quality metrics linked to were associated with the readmission. Statistical analysis was performed practitioner reimbursement. using SPSS software (IBM, Armonk, NY). We used the Wilcoxon text Key Words: postoperative readmission, unanticipated readmission, for comparing patient comorbidity scores and the χ2 test for categorical hospital readmission reduction program, quality metric, variables comparing readmitted to non-readmitted patients. Statistical physician reimbursement, plastic surgery patient significance was defined as a P value less than 0.05. (Ann Plast Surg 2017;78: S322–S324) RESULTS nanticipated postoperative readmissions impose a tremendous eco- U nomic, physical, and emotional burden on both patients and the Of the 819 patients investigated in this study, 149 (18.1%) practitioners responsible for providing perioperative evaluation and patients experienced unanticipated readmission, reoperation, or both care.1,2 Hospital readmissions and operating room take-backs are costly within the 30-day postoperative period. 74 (9.0%) patients were and detrimental consequences of surgical procedures performed in both readmitted for nonsurgical issues requiring nonoperative management, elective and critically-ill patient populations.3 Additionally, readmission alone, while 55 (6.7%) were readmitted and underwent operative inter- rates are a prominent quality marker measuring the adequacy of the pa- vention during that same readmission. Lastly, 20 (2.4%) patients tient's postoperative care, as well as a metric directly associated with underwent outpatient operative procedures but did not require a hospital practitioner reimbursement.2,4 readmission. This breakdown of our study population is summarized A link between postoperative readmission rates and provider in Figure 1. compensation was defined in 2012 when the Affordable Care Act es- Several risk factor variables pertaining to both patient demo- tablished the Hospital Readmissions Reduction Program.4,5 Under this graphic and health history information were chosen for statistical analysis (Fig. 2). When comparing these variables for readmitted and non-readmitted patients, we discovered that a positive smoking his- Received December 29, 2016, and accepted for publication, after revision January 7, 2017. tory (P = 0.009), hypertension (P = 0.0008), congestive heart failure From the *Division of Plastic and Reconstructive Surgery, Department of Surgery, and †University of Florida College of Medicine, Gainesville, FL. (P = 0.0015), and chronic obstructive pulmonary disease (P = 0.023) Conflicts of interest and sources of funding: none declared. were significantly elevated in our population of readmitted plastic Reprints: Peter A. Felice, MD, Division of Plastic & Reconstructive Surgery, surgery patients. Additionally, the mean age for readmitted patients University of Florida, PO Box 100138, 1600 Archer Road, Gainesville, FL was 52.4 years, a value significantly higher than the mean age of non- 32610-3003. E-mail: [email protected]. Copyright © 2017 Wolters Kluwer Health, Inc. All rights reserved. readmitted patients at 38.3 years (P = 0.0001). Readmitted patients also ISSN: 0148-7043/17/7806–S322 had a significantly higher mean Charlson Comorbidities Index score at DOI: 10.1097/SAP.0000000000001061 0.94 than did non-readmitted patients at 0.39 (P = 0.0001).8 No

S322 www.annalsplasticsurgery.com Annals of Plastic Surgery • Volume 78, Supplement 5, June 2017 Annals of Plastic Surgery • Volume 78, Supplement 5, June 2017 Unanticipated Postoperative Readmission

practitioner reviews are readily available through performance-based websites, social media postings, and patient-sponsored internet surveys, which are all increasingly used by patients when it comes to choosing a surgical provider.9 This attention to outcome-based and quality-based measures and the direct link both have to physician reimbursement places a considerable onus on the practitioner to provide optimal patient care. Many studies investigate patient risk factors associated with unanticipated readmissions for various general surgery patient populations.10–14 Although their findings may translate to various other medical and surgical subspecialties, we set out to determine exactly which risk factors were responsible for compromising the postoperative course of the plastic surgery patient population. We evaluated various patient demographic and health history data points based on previously published data as well as our authors' own prediction that significant medical comorbidities would contribute to a complicated postoperative course.10–13 When comparing these variables in FIGURE 1. Our study population was derived from a pool of our statistical analysis of readmitted and non-readmitted patients, we 2000 patients undergoing plastic and reconstructive observed that readmitted patients were significantly more likely to procedures at our institution over a 2-year period. From this be tobacco users, hypertensive, and have a previous diagnoses of original population, 819 patients were selected to undergo congestive heart failure or chronic obstructive pulmonary disease. analysis of their 30-day postoperative course. A total of 149 Additionally, advanced age and a higher health severity score calculated unanticipated postoperative readmissions were encountered, by the Charlson Comorbidities Index (CCI) put patients at a signifi- of which 74 patients required readmission without reoperation, cantly greater risk for 30-day readmission. The CCI is calculated using 55 required readmission with reoperation, and 20 required age, albumin level, history of myocardial infarction, cerebrovascular outpatient reoperation without a hospital readmission. disease, as well as other systemic disease processes, and was a relevant variable to include because it reinforced many of the other individual risk factor variables evaluated in our study population.8 In addition to statistically significant differences were noted between readmitted and defining the risk factors associated with readmission, just as enlighten- non-readmitted patients for any of the other variables investigated. ing was the discovery of which risk factors were not. For instance, patients with a history of radiation administration, cancer history, DISCUSSION immunosuppressed conditions, and greater travel distances to access Quality metrics are a prominent component in how physicians care centers, factors we assumed would lend to unanticipated readmis- are evaluated and gauge overall practitioner performance and reputa- sions, were not significantly different between the readmitted and non- tion. Furthermore, readmission rates have become a quality indicator readmitted patient populations. used to determine practitioner reimbursement from the Centers for Our study has some limitations we are currently working to im- Medicare and Medicaid Services.2,5–7 These metric assessments and prove. We are expanding the number of subjects and incorporating

FIGURE 2. Patients with unanticipated postoperative readmissions within 30-days of an initial procedure were compared to non-readmitted patients with respect to the above demographic and health history risk factors. Bold and asterisked risk factors were found to be significantly different between the 2 patient groups (P = 0.05 or less).

additional categorical variables to bolster study power and statistical REFERENCES significance. We also plan on exploring whether distinctions exist be- 1. Jencks SF, Williams MV,Coleman EA. Rehospitalizations among patients in the tween readmitted and non-readmitted patients based on the type and Medicare fee-for-service program. N Engl J Med. 2009;360:1418–1428. complexity of plastic surgery procedures performed. It may also be use- 2. Damle RN, Cherng NB, Flahive JM, et al. Clinical and financial impact of hospi- ful to parse out whether significant readmission differences are seen be- tal readmissions after colorectal resection: predictors, outcomes, and costs. tween ambulatory and inpatient operative settings, as well as when Dis Colon Rectum. 2014;57:1421–1429. 3. Merkow RP, Ju MH, Chung JW,et al. Underlying reasons associated with hospital comparing procedures performed by junior or senior attending physi- – cians. Future project aims include initiating a prospective study to deter- readmission following surgery in the United States.". JAMA. 2015;313:483 495. 4. Glass G, Lisk C, Stensland J. Refining the Hospital Readmissions Reduction mine whether the risk factors defined by this investigation are in fact Program. Medicare Payment Advisory Commission September 7, 2012. Avail- associated with future readmissions. Additionally, we are in the process able at http://www.medpac.gov/docs/default-source/reports/jun13_ch04.pdf. of constructing a Postoperative Readmission Risk Calculator to quanti- Accessed September 1, 2015. tatively predict the percentage likelihood of 30-day postoperative read- 5. James J. Health Policy Brief: Medicare Hospital Readmissions Reduction missions based on known patient demographics and risk factors. Program. Health Affairs November. 2013;12. 6. Berenson RA, Paulus RA, Kalman NS. Medicare's readmissions-reduction Ultimately, our vision is that this information be used at the time — – of the initial consultation between physicians and patients and be reiter- program a positive alternative. N Engl J Med. 2012;366:1364 1366. 7. Readmissions Reduction Program (HRRP) Available at https://www.cms.gov/ ated frequently throughout the perioperative course. Practitioners and medicare/medicare-fee-for-service-payment/acuteinpatientpps/readmissions-re- patients can work together to ameliorate modifiable patient risk factors duction-program.html. Accessed November 1, 2016. and have a free and open dialogue where anticipated risks, benefits, and 8. Roffman CE, Buchanan J, Allison GT. Charlson comorbidities index. J Physiother. expectations of an operative procedure are fully disclosed. With this ex- 2016;62:171. change and understanding, the patient and physician both take owner- 9. 24 outstanding statistics & figures on how social media has impacted the health ship of, and responsibility for, the challenges and goals of the care industryHonigman B. Available at https://getreferralmd.com/2013/09/health- postoperative course to arrive at an optimal result. care-social-media-statistics Accessed December 15, 2016. 10. Spolverato G, Ejaz A, Kim Y, et al. Readmission incidence and associated factors after a hepatic resection at a major hepato-pancreatico-biliary academic centre. HPB (Oxford). 2014;16:972–978. CONCLUSIONS 11. Kassin MT, Owen RM, Perez SD, et al. Risk factors for 30-day hospital readmis- This study identifies specific risk factors responsible for unantic- sion among general surgery patients. JAmCollSurg. 2012;215:322–330. ipated postoperative readmissions in a plastic surgery patient popula- 12. Damle M, Rachelle N, Karim Alavi. Risk factors for 30-d readmission after colo- tion. With this information, our goal is to further optimize patient care rectal surgery: a systematic review. JSurgRes. 2016;200:200–207. Web. 15 Dec. by providing appropriate perioperative resources and planning for pa- 2016. 13. Fischer JP, Wes AM, Nelson JA, et al. Factors associated with readmission follow- tients with characteristics that place them at an increased risk for post- ing plastic surgery. Plast Reconstr Surg. 2013;132:666–674. Web. 15 Dec. 2016. operative complication and readmission. In addition, application of 14. Tahiri Y, Fischer JP,Wink JD, et al. Analysis of risk factors associated with 30-day this information can help satisfy quality-directed metrics linked to reim- readmissions following pediatric plastic surgery: a review of 5376 procedures. bursement for institutions and individuals responsible for patients' care. Plast Reconstr Surg. 2015;135:521–529.

Preserving the Legitimacy of Board Certification Michael S. Hanemann, Jr, MD, FACS,*†‡ Holly C. Wall, MD, FACS,* and John A. Dean, MD, FACS*†‡

and residency programs is necessary to ensure that high standards for Objectives: The aims of this discussion were to inform the medical community quality education and training are upheld. about the American Board of Cosmetic Surgery's ongoing attempts in Louisiana The national standards for independent oversight of medical spe- to achieve equivalency to American Board of Medical Specialties (ABMS) member cialty boards are the American Board of Medical Specialties (ABMS) boards so that its diplomates may use the term “board certified” in advertising and and the American Osteopathic Association (AOA). All 24 ABMS to ensure public safety by upholding the standards for medical board certification. member boards require their diplomates to successfully complete resi- Background: In 2011, Louisiana passed a truth in medical advertising law, dency training programs supervised by the Accreditation Council for which was intended to protect the public by prohibiting the use of the term “board Graduate Medical Education (ACGME) in the specialty that the board certified” by improperly credentialed physicians. An American Board of is certifying. Graduates are then required to pass rigorous examinations Cosmetic Surgery diplomate petitioned the Louisiana State Board of Medical to obtain initial board certification and to participate in accredited Examiners to approve a rule that would establish a pathway to equivalency for Maintenance of Certification programs for the duration of their medical non-ABMS member boards, whose diplomates have not completed training ap- careers. This exacting system was developed to ensure that board- proved by the Accreditation Council for Graduate Medical Education (ACGME) certified physicians have undergone the appropriate scope, content, in the specialty they are certifying. Physicians and physician organizations representing multiple specialties (facial plastic and reconstructive surgery, otolar- and duration of training and that they have developed the validated yngology [head and neck surgery], orthopedic spine surgery, pediatric neurosur- skills necessary to practice safely and competently. gery, dermatology, and plastic surgery) urged the Louisiana State Board of By June 30, 2020, the ACGME will be the single graduate Medical Examiners to clarify its advertising policy, limiting the use of the term medical education accreditation system in the United States. The “board certified” to physicians who have completed ACGME-approved training single accreditation system allows graduates of allopathic and oste- in the specialty or subspecialty named in the certificate. opathic medical schools to complete their residency and/or fellow- Discussion: The public equates the term “board certified” with the highest level ship education in ACGME-accredited programs and demonstrate of expertise in a medical specialty. When a certifying board does not require com- achievement of common milestones and competencies. Accredita- pletion of ACGME or American Osteopathic Association (AOA)–accredited tion Council for Graduate Medical Education accreditation is over- training in the specialty it certifies, the result is an unacceptable degree of seen by a residency review committee made up of volunteer specialty variability in the education and training standards applied to its diplomates. experts from the field that set standards and provide peer evaluation of sponsoring institutions and specialty and subspecialty residency Independent, third-party oversight of certifying boards and training programs 2 is necessary to ensure quality standards are upheld. Any system that assesses and fellowship programs. a non-ABMS member or non–AOA-certified board for equivalency approval Patients have a right to know about their physician's qualifications, education, and training. When physicians advertise themselves must ensure that the training and qualifications required by the non-ABMS or “ ” AOA board are equivalent in scope, content, and duration to those required by to the public as board certified, the public cannot be expected to dif- the ABMS and AOA. This issue must not be misconstrued as a “turf battle” ferentiate between a legitimate medical board and a substandard one. between physicians of 2 competing specialties. Preserving the legitimacy of Louisiana Act 337, signed into law in 2011, was designed to protect the public by prohibiting physicians from using the term board certification is incumbent upon all medical specialties and subspe- “ ” cialties. This argument is a truthful, principled defense of the legitimacy of board certified in advertising unless the board meets any of the board certification. following criteria: 1. The board is an ABMS member board or an AOA certifying board. Key Words: board certification, plastic surgery, ABMS, ACGME, ASPS, ABPS, 2. The board has been approved by the Louisiana State Board of truth in advertising, facial plastic surgery, state medical board, Louisiana, Medical Examiners (LSBME). cosmetic surgery 3. The board requires an ACGME- or AOA-approved postgraduate (Ann Plast Surg 2017;78: S325–S327) training program that provides complete training in that specialty or subspecialty.3 mong the general public, the term “board certified” denotes the A highest level of achievement in medical training and expertise. In cases where the LSBME considers a non-ABMS member When individuals seek medical care, the knowledge that their doctor board for equivalency approval, the state medical board must ensure is board certified assures them that they are engaging a physician with that the training and qualifications required by the non-ABMS or significant direct training and demonstrated skill in the specialty certifi- AOA board are equivalent in scope, content, and duration to those re- cation advertised.1 Third-party regulatory oversight of medical boards quired by the ABMS and AOA. The ACGME- and AOA-approved residency programs require specific standards in regard to instructor qualifications, a minimum number and variety of types of procedures Received October 31, 2016, and accepted for publication, after revision performed, and progressive patient care responsibility. When a certify- December 24, 2016. ing body does not require completion of ACGME- or AOA-accredited From the *Louisiana Society of Plastic Surgeons, Shreveport; †Department of Plastic training in the specialty it certifies, the result is an unacceptable de- ‡ Surgery, Louisiana State University School of Medicine; and Department of gree of variability in the education and training standards applied to Plastic Surgery, Tulane University School of Medicine, New Orleans, LA. 1 Conflicts of interest and sources of funding: none declared. its diplomates. Reprints: Michael S. Hanemann Jr, MD, FACS, Plastic Surgery Consultants, 5233 The American Board of Cosmetic Surgery (ABCS) is not an Dijon Dr, Baton Rouge, LA 70808. E-mail: [email protected]; ABMS member board and does not have an ACGME training pro- [email protected]. Copyright © 2017 Wolters Kluwer Health, Inc. All rights reserved. gram. As such, the only criteria of Louisiana's truth in advertis- ISSN: 0148-7043/17/7806–S325 ing law the ABCS can potentially meet is number 2—approval by DOI: 10.1097/SAP.0000000000001035 the LSBME.

Annals of Plastic Surgery • Volume 78, Supplement 5, June 2017 www.annalsplasticsurgery.com S325 Hanemann et al Annals of Plastic Surgery • Volume 78, Supplement 5, June 2017

In April 2015, an ABCS diplomate, an otolaryngologist who has Equivalency is measured by scope, content, and duration of expanded his scope of practice to include cosmetic surgery of the training of a related ABMS specialty. To become a diplomate of the breast, trunk, and extremities, petitioned the LSBME to enact a rule that ABFPRS, one must first successfully complete an ACGME-approved would establish a pathway for non-ABMS member boards to be recog- otolaryngology residency, which provides complete training in facial nized as equivalent to ABMS- and AOA-approved boards. plastic surgery. This request was modeled after a rule approved by the Texas Board-certified otolaryngologists who have completed an addi- Board of Medical Examiners on April 12, 2013, through a non- tional 1-year fellowship in facial plastic surgery are eligible to take transparent process, which allowed for the issue to be voted on as one the certifying examination for the ABFPRS. A facial plastic surgery of numerous “consent agenda” items. No public notice of a meeting fellowship is a focused continuation of subspecialty training that or hearing of the Ad Hoc Committee on Advertising can be found in was received in ACGME- or AOA-approved otolaryngology resi- the public record. The rule was passed without a public hearing or dis- dency programs. Diplomates of the ABFPRS have already received cussion of the public safety, training, and expertise issues related to complete training in facial plastic surgery during their ACGME- board certification.4 Section A2e of the rule reads as the following: supervised otolaryngology residencies. Requires all physicians seeking certification to have satisfacto- Comparably, the American Board of Spine Surgery certifies or- rily completed postgraduate training that is accredited by the thopedic surgeons and neurosurgeons who have received complete ACGME or the AOA and that provides substantial and identifi- training in spine surgery while in their ACGME-approved orthopedic able supervised training of comprehensive scope in the specialty surgery or neurosurgery residencies. The additional 1-year fellowship 5 or subspecialty certified. these surgeons undertake is a focused continuation of the spine surgery training they receive in their ACGME-approved residencies. A reasonable person would interpret this as requiring all physi- Two of the three common ACGME residency pathways for cer- cians seeking certification to complete ACGME postgraduate training tification by the ABCS—general surgery and OB/GYN—do not contain in the specialty being certified. In contrast, the cosmetic surgeons con- case requirements for the breadth of cosmetic surgical procedures. The tend that this requirement consists of 2 independent clauses, the first of third, otolaryngology, provides ACGME approved training in cos- which is met by the completion of any previous ACGME-approved metic surgery within the head and neck but not cosmetic surgery of training, such as general surgery or obstetrics and gynecology (OB/ the breast, trunk, and extremities. Otolaryngologists who complete GYN). The second stipulation of the clause, they assert, is met by – a 1-year general cosmetic surgery fellowship have not undergone non ACGME-accredited training from a cosmetic surgery fellowship ACGME-approved training in cosmetic surgery of the breast, trunk, or apprenticeship. or extremities. It is irrelevant whether a cosmetic surgeon has undergone pre- Unlike a facial plastic surgery, spine surgery, or pediatric neuro- vious ACGME residency training in general surgery or OB/GYN surgery fellowship, a cosmetic surgery fellowship is not a focused con- because neither of these residencies provides complete training in cos- tinuation of ACGME-approved residency training in the same specialty. metic surgery. A cosmetic surgery fellowship charts an entirely new course. There is The leadership of the Texas Society of Plastic Surgeons wrote a letter to the LSBME, urging them to “(1) maintain the highest level of no foundation of ACGME training in cosmetic surgery on which to fo- standards and qualifications when awarding the use of ‘board certified’ cus or upon which to build. to physicians, (2) that any and all discussions and/or votes regarding the Increasing government regulations, decreasing insurance reim- useoftheterm‘board certified’ be done in the ‘light of day’ with proper bursement, and various economic and lifestyle pressures have caused some physicians to change or expand their area of practice, a phenom- and timely notice to the public and all interested physicians and parties, “ ” and (3) that the LSBME employ a stakeholder process that includes all in- enon known as practice drift. The Federation of State Medical Boards' terested parties when considering the use of the term ‘board certified.’”4 policy states that patients should seek information about prospective Independent, third-party regulatory oversight of a medical physicians, including the level of expertise they have with particular training program is necessary to ensure high-quality education procedures, their education (especially graduate medical education), and training. To secure this necessary oversight, physicians from and any board certifications held. The Federation of State Medical multiple specialties (plastic surgery, facial plastic surgery, orthopedic Boards also recommends that physicians should be prepared to provide spine surgery, pediatric neurosurgery, and dermatology), representing information about their qualifications and any additional training under- multiple state and national organizations, have urged in spoken and taken that has prepared them to provide treatment that falls outside their original area of practice and should provide this information to patients written testimony that the LSBME strike the previously mentioned 6 language in section A2e of the proposed rule and replace it with as part of the informed consent process. the following: Louisiana is currently experiencing the scenario that occurred in Requires successful completion of a postgraduate training pro- California in 2008, which was decided at the appellate court level. “ gram approved by the ACGME or the AOA that provides com- This case involves an 11 year effort by the ABCS, to gain spe- plete training in the specific specialty or subspecialty named in cialty board approval by the Medical Board of California, so that physi- the certificate. cians certified by the ABCS may advertise themselves as board certified in cosmetic surgery.”7 Under this language, non-ABMS or AOA boards can still be The state medical board enlisted an independent expert, deemed equivalent. For example, diplomates of the American Board of Dr Ronald Tompkins, to review ABCS fellowship programs. He Facial Plastic and Reconstructive Surgery (ABFPRS), the American concluded, “because of the…wide variations in prerequisites for, and Board of Pain Medicine, the American Board of Pediatric Neurosurgery, scope and content of, the training programs of the American Board of and the American Board of Spine Surgery all certify in an ABMS- Cosmetic Surgery, it is not possible for this reviewer to assign equiva- recognized specialty that provides complete ACGME-approved training lency to an ABMS board or to a non-ABMS board approved in the subspecialty named in the subspecialty certificate. in California.” A central argument of the ABCS's case for board equivalency in The California medical board denied the petitioner's applica- Louisiana is that the ABCS is analogous to other certifying bodies that tion after determining that the ABCS's requirements were not equiv- have rightly been deemed equivalent to ABMS or AOA boards, specif- alent to those of an ABMS member board. Dissatisfied, the ABCS ically, the ABFPRS. filed a discrimination lawsuit against the California Medical Board.

S326 www.annalsplasticsurgery.com © 2017 Wolters Kluwer Health, Inc. All rights reserved. Annals of Plastic Surgery • Volume 78, Supplement 5, June 2017 Preserving the Legitimacy of Board Certification

The appellate court's 2008 decision from The American Board of ACGME methodologies or approaches. Moreover, it also felt it is Cosmetic Surgery versus the Medical Board of California states, “Since highly unlikely that a state agency, such as the LSBME, would have the record reflects the reasons for the Board's decision, the reasons are the ability, expertise, or resources to duplicate or reproduce the pro- rationally related to the regulatory requirements and are supported cesses utilized by the ABMS, AOA, or the ACGME. by ample evidence, we conclude that the Medical Board did not Therefore, in summation, the committee makes the abuse its discretion by denying the American Board of Cosmetic following recommendations: Surgery's application.”7 In order for a physician (Doctor of Medicine of Doctor of Oste- On April 26, 2016, Georgia enacted a truth in advertising law opathy) to indicate a standing of ‘board certified,’ at a minimum,the similar to Louisiana's, but it closed the pathway of allowing the state Board Certification should include: medical board to recognize a non-ABMS member board whose diplo- 1. Successful completion of an ACGME-accredited training program in mates do not undergo ACGME training in the specialty named in the that specialty and/or subspecialty; and 8 certificate as equivalent. 2. Successful completion and passage of a psychometrically In a press release response to Georgia's truth in advertising law, valid examination.” the American Board of Cosmetic Surgeons attempts to undermine the importance of the ABMS and the AOA because they are private organi- “ The Louisiana Society of Plastic Surgeons trusts that the zations that are not federally mandated to uphold safety and ethics in LSBME will affirm the committee's findings and codify the necessity healthcare.” They label the law as special interest legislation that at- “ of completing ACGME-accredited training in the specialty or subspe- tempts to limit the ability of qualified physicians to advertise their cialty named in a certifying board's certificate. services and credentials—under the pretense of truthful advertising ”9 Lessons we have learned or reaffirmed in Louisiana from this and patient safety. process are the following: The only “special interest” of this legislation is to protect the public by ensuring that the quality standards for medical board certifi- 1. Public safety is the core issue and of paramount importance. cation are upheld. 2. The legitimacy of board certification must be preserved and protected. There are 24 ABMS member boards, 38 ABMS general spe- 3. Independent, third-party oversight of certifying boards and training cialty certificates, and 88 subspecialty certificates available. Nowhere programs is necessary and must be mandatory. among this huge number of potential certification options is there one 4. Board certification in a specialty or subspecialty must require com- for cosmetic surgery. This is because no ACGME-approved cosmetic pletion of an ACGME-accredited training program in the specialty surgery fellowship programs exist. Cosmetic surgery is an integral or subspecialty named in the certificate, followed by successful completion of a certifying examination in that specialty. and inextricable part of plastic surgery and otolaryngology training – (head and neck only). Residents must perform a minimum number of 5. Any system that assesses a non-ABMS member or non AOA- cosmetic surgeries in various anatomic categories within ACGME- certified board for equivalency approval must ensure that the training approved residency programs. and qualifications required by the non-ABMS or AOA board are Consider the vastly disparate credentials of the 6 Louisiana dip- equivalent in scope, content, and duration to those required by the ABMS and AOA. lomates of the ABCS: a colorectal surgeon, a dermatologist, 2 otolaryn- “ ” gologists, an oral surgeon, and a general surgeon. It does not make 6. This issue must not be misconstrued as a turf battle between phy- sense that the training of physicians from such markedly different fields sicians of 2 competing specialties. Preserving the legitimacy of can be certified as equivalent under a single board without them having board certification is incumbent upon all medical specialties completed any additional recognized training in that specialty. and subspecialties. This argument is a truthful, principled defense Only one of the six current ABCS diplomates in Louisiana has of the legitimacy of board certification by physician leaders from completed the very cosmetic fellowship training they contend is an es- multiple specialties, representing multiple state and national sential requirement for certification. This “grandfathering” of certifica- medical societies. tion through an experience track has never been an option for certification by an ABMS member board. Establishing and maintaining high standards for board certifica- REFERENCES tion in the practice of medicine protect the public. Lowering standards would be detrimental to public safety. This argument is not about “re- 1. Finley E, Williams R, Madara J. Letter From Leaders of Multiple Medical Orga- straint of trade” or “limiting patient choice.” Patients can seek care from nizations to LSBME. 2015. 2. Single GME. Accreditation System. Available at: www.acgme.org Accessed any physician they choose, but patients have a right to know the truth October 30, 2016. about their doctor's qualifications, education, and training. 3. Michot M. R.S. 37:1285 (A) (32)—Louisiana Act 337—Truth in Medical Advertising. Granting equivalency to a substandard board will incentivize 4. Pownell P, Barrera A, Capelo J. Letter From Texas Society of Plastic Surgeons more physicians to opt for the “shortcut” route to board certification, to LSBME. eschewing the ABMS, AOA, and ACGME for alternate training of in- 5. Mouton C. Louisiana State Board of Medical Examiners. May 2015 Regulatory adequate scope, content, and duration. In addition to confusing the pub- Report—Draft of Title 46 Professional and Occupational Standards. Part XLV. lic and endangering their safety, lowering standards denigrates the Medical Professions. Subpart 3. Practice. Chapter 76. Definition of Enforcement meaning of board certification. and Other Terms. Subchapter C. Board Certification. 6. The Federation of State Medical Boards. Public Policy Compendium 2016. Avail- The LSBME assembled an ad hoc committee of physicians to able at: www.fsmb.org. Accessed October 30, 2016. determine whether there are methods to assess the adequacy of 7. American Board of Cosmetic Surgery v. Md. Bd., 75 Cal. Rptr. 3d 573, 162 Call. training/demonstration of competency outside traditional ACGME- App. 4th 534 (Cal App., 2008). accredited residency/fellowship programs. 8. Kelly and Cooper. Georgia HB 1043. Accessed April 26, 2016. The investigators' report states, “The committee determined that 9. American Board of Cosmetic Surgery opposes legislation that restricts patient no other vehicle or mechanism is comparable to the ABMS, AOA, or choice. PRNewswire.April29,2016.

Bridging the Gap Engineered Porcine-derived Urinary Bladder Matrix Conduits as a Novel Scaffold for Peripheral Nerve Regeneration Lyly Nguyen, MD,*§ Ashkan Afshari, MD,*|| Nathaniel D. Kelm, BS,†‡ Alonda C. Pollins, MS,* R. Bruce Shack, MD,* Mark D. Does, PhD,†‡ and Wesley P.Thayer, MD, PhD*

morbidity and the availability of expendable donor nerves. For this rea- Purpose: This study aims to compare engineered nerve conduits constructed son, various conduits have been studied as alternatives to autografts.3–5 from porcine-derived urinary bladder matrix (UBM) with the criterion-standard Current Food and Drug Administration–approved biodegradable con- nerve autografts, for segmental loss peripheral nerve repairs. duits in the market have had variable results in comparison with auto- Methods: Forty-eight Sprague-Dawley rats were divided into 2 groups. All grafts both clinically and experimentally. These variable results have underwent a 10-mm sciatic nerve gap injury. This was repaired using either (1) re- precluded conduit-based repair full acceptance in bridging segmental verse autograft—the 10-mm cut segment was oriented 180 degrees and used to 2,3,6,7 — peripheral nerve loss, leading bioengineers continue to search for coapt the proximal and distal stumps or (2) UBM conduit the 10-mm nerve conduit material that will consistently outperform autografts. gap was bridged with UBM conduit. Behavior assessments such as sciatic func- Extracellular matrix (ECM) has been studied as a tissue scaffold tion index and foot fault asymmetry scores were performed weekly. At 3- or and has demonstrated favorable outcomes in many reconstructive 6-week time endpoints, the repaired nerves and bilateral gastrocnemius/soleus fields.8–12 It has a complex array of functional proteins, such as colla- muscles were harvested from each animal. Nerves were evaluated using immuno- gen and laminin, that provides structural support and serves as an or- histochemistry for motor and sensory axon staining and with diffusion tensor chestrator to signaling molecules for cell growth and differentiation, imaging. The net wet muscle weights were calculated to assess the degree of angiogenesis, and modulates immune cells for wound healing.8,13 Be- muscle atrophy. cause of its versatility and benefits seen in wound healing, we study Results: The UBM group demonstrated significantly improved foot fault asym- conduits constructed of ECM derived from porcine urinary bladder metry scores at 2 and 4 weeks, whereas there was no difference in sciatic function for the repair of peripheral nerves with gaps. index. The net muscle weights were similar between both groups. Motor axon counts proximal/inside/distal to the conduit/graft were similar between UBM conduits and reverse autografts, whereas sensory axon counts within and distal METHODS to the conduit were significantly higher than those of the autograft at 6 weeks. Sensory axonal regeneration seemed to be adherent to the inner surface of the Experimental Design/Animal Model UBM conduit, whereas it had a scattered appearance in autografts. Diffusion ten- Adult female Sprague-Dawley rats (n = 48; 200–250 g; Charles sor imaging parameters between groups were similar. River Lab) were used in the experiment. The objective of this study is to Conclusions: Urinary bladder matrix conduits prove to be at least similar to evaluate behavior, histology, and diffusion tensor imaging (DTI) param- nerve autografts for the repair of peripheral nerve injuries with a short gap. The eters between 2 groups: (1) autologous grafts (n = 24) and (2) conduits matrix perhaps serves as a scaffold to augment sensory nerve growth. constructed of porcine-derived urinary bladder matrix (UBM) (n = 24) Clinical Relevance: In a clinical setting, UBM may eliminate the donor site mor- for the repair of peripheral nerve gaps in a rat sciatic nerve model. All bidity and increased operative time associated with nerve autografting. rats underwent postoperative behavioral studies. After reaching 3- or Key Words: conduit, diffusion tensor imaging, extracellular matrix, 6-week time endpoints, the left sciatic nerve was harvested and evalu- peripheral nerve, urinary bladder matrix ated with (1) immunohistochemistry (IHC) or (2) DTI. Additionally, bilateral gastrocnemius/soleus muscles were harvested to obtain net wet (Ann Plast Surg 2017;78: S328–S334) weights (Fig. 1). Rats were housed in a central animal care facility and were pro- ost peripheral nerve injuries are traumatic in nature, and debride- vided food and water ad libitum. All experimental procedures were M ment of damaged nerves outside the zone of injury is often nec- approved and performed in accordance with the standards of our insti- 1 essary for improved regeneration. The potential nerve gap created tutional animal care and use committee. results in difficulties to achieve tensionless primary neurorrhaphy.2 Au- tografts are the criterion standard to treat segmental peripheral nerve Sciatic Nerve Injury gaps; however, autologous nerve grafting has been limited by donor site Each rat was anesthetized via inhaled isoflurane (2%). The left hindquarter prepped aseptically. A 3-cm incision was made at the left Received August 29, 2016, and accepted for publication, after revision sciatic notch, parallel and posterior to the left femur. The biceps femoris December 24, 2016. was then reflected to expose the sciatic nerve. A 10-mm nerve gap was From the Departments of *Plastic Surgery, and †Biomedical Engineering, and created in the midportion of the sciatic nerve with microsurgical ‡Institute of Imaging Science, Vanderbilt University Medical Center, Nashville, scissors. The gap was repaired with either of the following: (1) The TN; §Department of Surgery, Morristown Medical Center, Morristown, NJ; ||Department of Surgery, University of South Carolina, Columbia, SC. transected portion of the nerve was reversed 180 degrees and used as Conflicts of interest and sources of funding: ACELL provided funding for materials in an autograft to coapt the proximal and distal stumps using 9–0nylonsu- performance of this experiment. ture (Ethicon, Somerville, NJ) in an end-to-end fashion using microsur- Reprints: Lyly Nguyen, MD, Medical Center North, 1161 21st Ave S, S2221 gical technique. (2) A 10 mm (L) Â 1.5 mm (D) UBM conduit was used Nashville, TN 37232. E-mail: [email protected]. – Copyright © 2017 Wolters Kluwer Health, Inc. All rights reserved. to bridge the sciatic nerve gap and secured using 9 0 nylon sutures. The ISSN: 0148-7043/17/7806–S328 muscle and skin were closed to cover the nerve using 5–0 monocryl DOI: 10.1097/SAP.0000000000001042 suture (Ethicon).

S328 www.annalsplasticsurgery.com Annals of Plastic Surgery • Volume 78, Supplement 5, June 2017 Annals of Plastic Surgery • Volume 78, Supplement 5, June 2017 Porcine-derived Urinary Bladder Matrix Conduits

FIGURE 1. Experimental design.

Behavioral Testing equally to undergo either DTI processing or histology via IHC staining. Postoperatively, foot fault (FF) asymmetry and sciatic functional The gastrocnemius and soleus muscles were then dissected from each index (SFI) behavioral assessments were performed at weekly intervals. hind limb and weighed (in grams) to obtain net wet weights for To prevent possible bias, behavior testers were blinded to the surgical each animal. protocols of each animal.

Foot Fault Histology Animals were allowed to roam freely on a wire mesh grid mea- Immunohistochemical staining for motor and sensory axons was Â Â suring 45 30 cm, with square openings measuring 2.5 2.5 cm. performed as previously described using commercial antibodies specif- The grid was elevated 2 cm above a solid base. Trials of 50 total steps ically directed against carbonic anhydrase II (Abcam, Cambridge, per hind limb were recorded for each animal. A full fault was recorded Mass) and choline acetyltransferase (Millipore, Temecula, Calif).17 if the animal's hind limb fell through the opening in the grid and For carbonic anhydrase II, primary antibody was used at 1:200 over- touched the floor, whereas a partial fault was recorded if the foot did night. For choline acetyltransferase staining, primary antibody was used not touch the floor. An FF asymmetry score was calculated using the at 1:200 for 1 hour. Digital images of stained sections were acquired Â following equation: composite FF score = (n partial faults 1) + (n full using Image-Pro Plus 7.0 software (Media Cybernetics, Bethesda, Â Â faults 2), %FF = (composite FF score/total number of steps) 100%, Md) with an Olympus C-35AD-4 microscope at low magnification − FF asymmetry score = %FF (normal hind limb) %FF (surgical (10Â). Digital images were loaded on ImageJ (Wayne Rasband, Na- hind limb). tional Institute of Mental Health, Bethesda, Md); however, axon counts were obtained by manual count of the entire nerve area. Sciatic Functional Index Before surgery, habituation trials were performed for animals to navigate an inclined beam into their home cage without hesitation.14 Each hind limb was inked, and animals walked the beam until 6 consec- Diffusion Tensor Imaging utive (3 from each limb) foot prints were recorded. Print measurements Tissue Sample Preparation include the following: normal print length (NPL), normal toe spread After nerve excision, reverse autografts were incubated in 4% (NTS), normal intermediary toe spread (NIT), experimental print length glutaraldehyde/0.5% paraformaldehyde in phosphate-buffered saline (EPL), experimental toe spread (ETS), and experimental intermediary (PBS) at 4°C and nerves repaired with a UBM conduit were incubated toe spread (EIT). Sciatic functional index scores were the calculated in 4% glutaraldehyde/2% paraformaldehyde in PBS at 4°C for better − − using the following formula: SFI = 38.3([EPL NPL]/NPL) + 109.5 penetration of the conduit material. After a minimum of 24 hours of − − − 15,16 ([ETS NTS]/NTS) + 13.3([EIT NIT]/NIT) 8.8. postfixation, excised nerves were placed in PBS + 2 mM gadolinium diethylene triamine pentaacetic acid (Magnevist, Bayer HealthCare, Wayne, NJ) at 4°C for at least 24 hours before imaging. For imaging, Sciatic Nerve and Gastrocnemius/Soleus excised nerves were placed in glass capillary tubes (3-mm outer diam- Muscle Harvest eter) filled with a perfluropolyether liquid (Fomblin, Solvay Solexis, When the animals reached their final time endpoint, they were Thorofare, NJ) for susceptibility matching, preventing tissue dehydra- euthanized with Euthasol (Virbac AH, Fort Worth, Tex) intracardiac in- tion, and a signal-free background. For higher throughput, 6 nerves in jections. The full length of the nerve was harvested and distributed a hexagonal arrangement were imaged simultaneously.

Diffusion Tensor Magnetic Resonance Imaging Magnetic resonance imaging was performed on a 4.7-T Agilent Direct Drive scanner (Agilent Technologies, Santa Clara, Calif) for reverse autografts and a 9.4-T Agilent Direct Drive scanner for conduit- repaired nerves. Diffusion tensor imaging data were acquired using a 3-dimensional diffusion-weighted spin-echo sequence with repetition time/echo time of 170/23.0 ms, 12 signal averages, and field of view of 9.6 Â 9.6 Â 14.4 mm3 for reverse autografts and repetition time/ echo time of 160/23.0 ms, 10 signal averages, and field of view of 9.6 Â 9.6 Â 18.0 mm3 for conduit-repaired nerves. The nominal resolution for both groups was 100 Â 100 Â 450 μm3 (450 μm along the nerve). Diffusion weighting was achieved with δ/Δ =4/12ms,apre- scribed b value of 2000 s/mm2, and 6 directions. One b = 0 image was acquired for a total of 7 images in a scan time of approximately 12 hours.

DTI Post-processing and Analysis FIGURE 3. Sciatic function index scoring. A more negative score Image data reconstruction was performed using in-house written indicates worse impairment. code in MATLAB (MathWorks, Natick, Mass), and DTI analysis was 18 implemented using ExploreDTI. Three-dimensional image volumes Nerve Histology With IHC were zero-padded 2Â in each direction during reconstruction from k-space data. Diffusion tensors were estimated voxel-wise using a linear When harvesting the injured nerves, the UBM conduit segment least-squares approach. From the diffusion tensor, fractional anisotropy in the 6-week endpoint group seemed more similar in size to the prox- (FA), axial diffusivity (AD), and radial diffusivity (RD) were computed imal and distal ends of the sciatic nerve than the 3-week conduits. The on a voxel-wise basis and DTI tractography was performed. 3-week conduits seemed bulkier. This was accounted for by partial bio- degradation of the conduit in 6-week group compared with the 3-week Statistical Analysis animals. No samples demonstrated evidence of neuroma formation. Statistical significance was determined using nonparametric After staining with choline acetyltransferase for motor axons and Mann-Whitney U test and designated a P value of less than 0.05. carbonic anhydrase II for sensory axons, cross sections proximal to, within, and distal to the conduit/graft were evaluated. There was no significant difference in motor axon counts proximal to, within, and distal to the conduit/graft in UBM conduits and autografts in the 6-week time RESULTS point (Fig. 5a). However, the UBM conduit had greater sensory axons within the conduit (455 ± 31 vs 140 ± 34, P < 0.01) and distal to the Behavioral Evaluations of Sciatic Nerve Function conduit (253 ± 27 vs 77 ± 14, P < 0.01) in comparison with the autograft group at 6 weeks (Fig. 5b). Animals repaired with a UBM conduit significantly performed Within the 3-week time point, UBM conduits had similar axon better in FF than the reverse autografts in the 2-week and 4-week time counts to the reverse autografts in both motor and sensory axons prox- point (Fig. 2). When evaluating SFI scoring, the UBM conduit had imal and within the conduit/graft. However, it is important to note that 2 behavior results comparable with the reverse autograft at all time samples in the 3-week UBM conduit group had uncharacteristically points (Fig. 3). lower sensory counts compared with the other 3-week UBM samples (but similar to autografts) because of collapse, which resulted in a loss Gastrocnemius/Soleus Net Wet Weights of statistical significance. Because of uncleared cellular debris from There was no statistically significant difference in net muscle Wallerian degeneration in the distal end at 3 weeks, axon counts at those weight between the UBM conduit and autografts at 3 weeks or segments for both study groups were deemed unreliable and, thus, were 6 weeks (Fig. 4). not included in this study.

FIGURE 2. Foot fault scoring. *P < 0.05, statistically significant. A more negative score indicates worse impairment. FIGURE 4. Gastrocnemius/soleus muscle net wet weight.

basement membrane complex also helps regulate cell growth, differentiation, and migration during tissue development and reconstruction.7,8 The beneficial regenerative aspects of UBM scaffolds were demonstrated in this study because we found improved sensory axon numbers in comparison with the autograft. Uniquely, sensory axons also seemed adherent to the inner surface lining of the UBM conduit. Al- though the mechanism in which sensory axons favor growth along the conduit is unknown, structural proteins (ie, laminin, fibronectin, collagen) known to be present in ECM may be responsible. In vitro as- says on human fetal sensory neurons have shown that laminin enhances Schwann cell response and antibodies to laminin suppress sensory neurite growth.20,21 Similarly, fibronectin has been shown to act as a chemoattractant for migrating Schwann cells and provide a direct sub- strate for adhesion and outgrowth of regenerating sensory axons.20–23 The UBM conduit basement membrane may possess more of these ECM structural proteins that support Schwann cell and sensory axon infiltration over autografts; however, further studies would need to be performed to identify the causative factor. Additionally, motor axon growth was not significantly different between groups at both 3- and 6-week time points. In relation to motor axon regeneration, UBM conduits pre- sumably have a similar structural milieu to autografts to augment axonal outgrowth. Although histomorphological parameters give information regarding nerve regeneration on a microscopic level, it is important to translate this information to functional recovery as histological evaluation after repair is not possible in humans. The FF scoring system has historically been used to determine the sensorimotor function in animals.24 The improved FF score at 2 and 4 weeks within the UBM conduit could be explained by higher sensory axon counts, in which animals may have more sensory feedback for the next step and grip the rung of the grid with fewer faults. However, despite a more robust sensory axon count, there is question of the accuracy of the FF at 2 weeks postoperatively, as axons are unlikely to have made complete FIGURE 5. A, Motor axon count at 6 weeks. B, Sensory axon functional connections.25 De Medinaceli et al15 had described difficulty count at 6 weeks. *P < 0.05, statistically significant. in evaluating sensory function because of overlapping innervation and indirect measurements. Therefore, SFI scoring is regarded as a more re- Motor axons in both groups seemed to be uniformly distributed liable and reproducible quantitative method for the assessment of func- in all segments (Fig. 6a). Interestingly, the sensory axons within the tional outcome in the recovery of the sciatic nerve in rats.14,15,26–28 In UBM conduit seemed to grow adherent to the inner lining of the conduit this study, we found that UBM conduits performed no worse than that and then demonstrated a scattered appearance once they reached the of autografts in SFI at any time point. distal segment. Sensory axons in the reverse autografts seemed to be The gastrocnemius net weight was similar between both groups, scattered in both the graft and distal segments (Fig. 6b). demonstrating likely similar degrees of muscle reinnervation. As expected, there was greater muscle atrophy in the injured limb in compar- Diffusion Tensor Imaging ison with its control limb. Martins et al29 described that one must Diffusion tensor imaging of excised sciatic nerves was evaluated consider that not all regenerated fibers are viable and misdirected axon in 3 regions of interest: 1.5 mm proximal to the proximal graft site, the growth can lead to aberrant muscle reinnervation. This consideration, in middle of the conduit/autograft, and 1.5 mm distal to the distal graft site. addition to the time it takes regenerating axons to reach the motor end Fractional anisotropy, AD, and RD were measured for each region of plate, explains the difference in the degree of atrophy between the ex- interest. Within the 3- or 6-week UBM group, there was no difference perimental and the control limb. in FA, AD, and RD, in all segments compared with the autografts This study is the first to monitor the efficacy of a nerve conduit (Table 1). Some UBM conduit samples had evidence of partial or near in nerve regeneration using DTI technology. DTI is an advanced neuro- complete collapse without visualization of distal tracks. Figure 7 shows imaging tool with the ability to evaluate peripheral nerves noninva- tractography demonstrating proximo-distal axonal growth at 6 weeks sively, objectively, and quantitatively. Parameters yielded by DTI, for noncollapsed UBM conduits and reverse autografts. such as FA, AD, and RD, reflect the anisotropic diffusion of water molecules along and perpendicular to the axon bundles, which provide valuable information on axonal regeneration and degeneration.30 The DISCUSSION UBM conduits performed equally in all parameters of DTI in compari- With advances in understanding nerve regeneration pathophysi- son with autografts demonstrating similar axonal integrity in all time ology, tissue engineering research has focused on developing aniso- periods. Diffusion tensor imaging can further be expanded to create tropic ECM scaffolds that augment axonal regeneration and Schwann tractography images, which allow visualization of the longitudinal cell migration.19 In this study, we evaluate ECM derived from porcine extent of nerve regeneration. This study demonstrated track forma- bladder matrix to serve as a novel scaffold. In comparison with other tion proximally to distally in the UBM conduit similar to reverse au- xenogeneic ECM, porcine UBM has demonstrated fewer invasions of fi- tografts. However, because of partial collapse found in some UBM broblasts because of its intact basement membrane complex on its surface, samples, further experiments will need to be performed with less thereby decreasing scar tissue formation.8 The advantage of an intact deformable models.

FIGURE 6. A, Motor axon histomorphology at 6 weeks. Motor axons seem to be uniformly distributed in all segments in both groups. B, Sensory axon histomorphology at 6 weeks. Sensory axons within the conduit seem to grow adherent to the inner lining of the UBM conduit.

The timing of outcome assessment in animal studies is critical to been described to have rapid neuroregenerative capacity compared with recognizing the true differences between study groups. Rodents have humans and higher mammals. Studying outcomes at earlier time points

TABLE 1. Diffusion Tensor Imaging Parameters in Each Segment Between UBM Conduit and Reverse Autograft at 6 Weeks

Proximal SEM P Within SEM P Distal SEM P FA, μm2/ms 0.57 1.00 0.08 UBM conduit 0.578 0.019 0.593 0.022 0.481 0.010 Reverse autograft 0.524 0.016 0.562 0.014 0.545 0.018 AD, μm2/ms 0.57 0.08 0.57 UBM conduit 0.910 0.037 0.098 0.054 0.890 0.079 Reverse autograft 0.082 0.026 0.800 0.038 0.791 0.020 RD, μm2/ms 1.00 0.08 0.57 UBM conduit 0.360 0.029 0.488 0.053 0.440 0.037 Reverse autograft 0.387 0.021 0.357 0.028 0.363 0.015 P < 0.05, statistical significance.

In this study, UBM conduits demonstrated to be at least similar to nerve autografts for the repair of peripheral nerve injuries. The matrix perhaps serves as a scaffold to augment sensory nerve growth. In a clinical setting, these promising results may eliminate the donor site morbidity and increased operative time associated with nerve autografting.

ACKNOWLEDGMENTS We would like to thank Nancy Cardwell, Justine S. Kim, and Ravinder Bamba for their help with this study. We would also like to ac- knowledge the Animal Care staff and Amy Nunnally for their assistance in animal care.

REFERENCES 1. Scholz T, Krichevsky A, Sumarto A, et al. Peripheral nerve injuries: an international survey of current treatments and future perspectives. J Reconstr Microsurg. 2009;25:339–344. 2. Isaacs J, Browne T. Overcoming short gaps in peripheral nerve repair: conduits and human acellular nerve allograft. Hand.2014;9:131–137. 3. Moore AM, Kasukurthi R, Magill CK, et al. Limitations of conduits in peripheral nerve repairs. Hand (N Y). 2009;4:180–186. 4. Griffin JW, Hogan MV,Chhabra AB, et al. Peripheral nerve repair and reconstruction. J Bone Joint Surg Am. 2013;95:2144–2151. 5. Deal DN, Griffin JW, Hogan MV.Nerve conduits for nerve repair or reconstruction. JAmAcadOrthopSurg. 2012;20:63–68. 6. Pabari A, Lloyd-Hughes H, Seifalian AM, et al. Nerve conduits for peripheral nerve surgery. Plast Reconstr Surg. 2014;133:1420–1430. FIGURE 7. Diffusion tensor imaging tractography at 6 weeks 7. Pfister BJ, Gordon T, Loverde JR, et al. Biomedical engineering strategies for peripheral nerve repair: surgical applications, state of the art, and future challenges. demonstrating nerve regeneration in (A) UBM conduits that Crit Rev Biomed Eng. 2011;39:81–124. did not collapse and (B) reverse autografts. 8. Brown B, Lindberg K, Reing J, et al. The basement membrane component of biologic scaffolds derived from extracellular matrix. Tissue Eng. 2006;12:519–526. provide better clinical correlation as the experimental and control 9. Badylak SF, Vorp DA, Spievack AR, et al. Esophageal reconstruction with ECM – groups are still significantly different. Later time point (ie, 15 weeks) and muscle tissue in a dog model. JSurgRes. 2005;128:87 97. 10. Wainwright DJ. Use of an acellular allograft dermal matrix (Alloderm) in the man- observations may show experimental and control groups to be equiva- agement of full-thickness burns. Burns. 1995;21:243–248. lent because of the enhanced regenerative capacity of the rodent ner- 11. Clark JM, Saffold SH, Israel JM. Decellularized dermal grafting in cleft palate re- vous system, and this could mask any differences seen between pair. Arch Facial Plast Surg. 2003;5:40–44. 3,6,31,32 groups. At 6 weeks, Wallerian degeneration is complete, yet ax- 12. Valerio IL, Campbell P, Sabino J, et al. The use of urinary bladder matrix in the onal regeneration is still ongoing, providing the ability to more accu- treatment of trauma and combat casualty wound care. Regen Med. 2015;10: rately compare recovery differences between both groups. Therefore, 611–622. this time point strengthens the validity of improved sensory axons at 13. Badylak SF. The extracellular matrix as a scaffold for tissue reconstruction. Semin – least early in the nerve recovery process when utilizing the UBM con- Cell Dev Bio. 2002;13:377 383. 14. Sarikcioglu L, Demirel BM, Utuk A. Walking track analysis: an assessment duit. Further studies with later time points would need to be performed method for functional recovery after sciatic nerve injury in the rat. Folia Morphol to ascertain if this still remains true when axonal regeneration is (Warsz). 2009;68:1–7. considered complete. 15. de Medinaceli L, Freed WJ, Wyatt RJ. An index of the functional condition of rat In comparing gross appearance of 3- and 6-week conduits sciatic nerve based on measurements made from walking tracks. Exp Neuro.1982; groups, the latter conduits revealed that the conduit had partially de- 77:634–643. graded over time. A conduit resorption rate that mirrors the speed of 16. Bain JR, Mackinnon SE, Hunter DA. Functional evaluation of complete sciatic, peroneal, and posterior tibial nerve lesions in the rat. Plast Reconstr Surg.1989; nerve regeneration would be ideal as this would provide stability of 83:129–136. the conduit walls for nerve regeneration, while being degradable to pre- 7,33 17. Riley DC, Bittner GD, Mikesh MA, et al. Polyethylene glycol-fused allografts pro- vent scar tissue accumulation and nerve compression. These duce rapid behavioral recovery after ablation of sciatic nerve segments. JNeurosci features were demonstrated in the UBM conduit as both axonal regen- Res. 2015;93:572–583. eration and conduit degradation occurred over time. 18. Leemans A, Jeurissen B, Sijbers J, et al. ExploreDTI: a graphical toolbox for pro- The need to develop an ideal conduit that can perform superiorly cessing, analyzing, and visualizing diffusion MR data. Proc Intl Soc Mag Reson to autografts is important, particularly for the deficit seen in conduits for Med. 2009;17:3537. 2,6,7 19. Houschyar KS, Momeni A, Pyles MN, et al. The role of current techniques and gaps over 30 mm. Furthermore, the diameter of the injured nerve concepts in peripheral nerve repair. Plast Surg Int. 2016;2016:4175293. and conduit employed in a repair significantly affects nerve regenera- 20. Evercooren A, Kleinman HK, Ohno S, et al. Nerve growth factor, laminin, and fi- tion because larger conduits may cause a dilution effect of neurotropic bronectin promote neurite growth in human fetal sensory ganglia cultures. factors and are subject to collapse.2,3,6,25 Few studies have been con- JNeurosciRes. 1982;8:179–193. ducted for large diameter nerve repairs, and many have reported the in- 21. de Luca AC, Lacour SP,Raffoul W, et al. Extracellular matrix components in pe- feriority of the use of any conduit in comparison with autografts.3,6,7,34,35 ripheral nerve repair: how to affect neural cellular response and nerve regenera- – In this study, we were able to demonstrate at least comparable results tion? Neural Regen Res. 2014;9:1943 1948. 22. Armstrong SJ, Wiberg M, Terenghi G, et al. Laminin activates NF-kB in Schwann for segmental gaps of 10 mm with a conduit diameter of 1.5 mm. The cells to enhance neurite outgrowth. Neurosci Lett. 2008;439:42–46. need to study this model in larger gaps, with larger diameter nerves, is 23. Gardiner NJ. Integrins and the extracellular matrix: key mediators of development necessary to evaluate the UBM conduit's potential against autografts and regeneration of the sensory nervous system. Dev Neurobiol. 2011;71: in larger animals. 1054–1072.

24. Schaar KL, Brenneman MM, Savitz SI. Functional assessments in the rodent 30. Lehmann HC, Zhang J, Mori S, et al. Diffusion tensor imaging to assess axonal stroke model. Exp Transl Stroke Med. 2010;2:11. regeneration in peripheral nerves. Exp Neurol. 2010;223:238–244. 25. Chamberlain LJ, Yannas IV, Hsu HP, et al. Near-terminus axonal structure and 31. Brenner MJ, Moradzadeh A, Myckatyn TM, et al. Role of timing in assessment of function following rat sciatic nerve regeneration through a collagen-GAG matrix nerve regeneration. Microsurgery. 2008;28:265–272. in a ten-millimeter gap. JNeurosciRes. 2000;60:666–677. 32. Waitayawinyu T, Parisi DM, Miller B, et al. A comparison of polyglycolic acid 26. Monte-Raso VV, Barbieri CH, Mazzer N, et al. Is the sciatic functional index al- versus type 1 collagen bioabsorbable nerve conduits in a rat model: an alternative ways reliable and reproducible? J Neurosci Methods. 2008;170:255–261. to autografting. JHandSurgAm. 2007;32:1521–1529. 27. Brown CJ, Evans PJ, Mackinnon SE, et al. Inter- and intraobserver reliability of 33. Ichihara S, Inada Y, Nakamura T. Artificial nerve tubes and their application for walking-track analysis used to assess sciatic nerve function in rats. Microsurgery. repair of peripheral nerve injury: an update of current concepts. Injury. 2008;39 1991;12:76–79. (suppl 4):29–39. 28. Brown CJ, Mackinnon SE, Evans PJ, et al. Self-evaluation of walking track mea- 34. Shin RH, Friedrich PF,Crum BA, et al. Motor outcomes of segmental nerve defect surement using sciatic function index. Microsurgery. 1989;10:226–235. in the rat using bioabsorbable synthetic nerve conduits: a comparison of commer- 29. Martins RS, Siqueira MG, da Silva CF, et al. Correlation between parameters of cially available conduits. J Bone Joint Surg Am. 2009;91:2194–2204. electrophysiological, histomorphometric and sciatic functional index evaluations 35. Isaacs J, Mallu S, Yan W, et al. Consequences of oversizing: nerve-to-nerve tube after rat sciatic nerve repair. Arq Neuropsiquiatr. 2006;64:750–756. diameter mismatch. J Bone Joint Surg Am. 2014;96:1461–1467.

Effects of Collagenase Digestion and Stromal Vascular Fraction Supplementation on Volume Retention of Fat Grafts Jonathan B. Olenczak, MD,* Scott A. Seaman, PhD,† Kant Y. Lin, MD,* Angela Pineros-Fernandez, MD,* Catherine E. Davis,‡ Lisa S. Salopek, LVT, RLATG,* Shayn M. Peirce, PhD,*† and Patrick S. Cottler, PhD*

members found that of those performing the procedure (69% of respon- Objective: The use of autologous fat as a soft tissue filler has increased over the dents), 7.9% believe fat graft volume retention is 75% to 100%, past decade in both reconstructive and aesthetic surgeries. Enhancement of autol- whereas 56% believe fat graft volume retention is 50% to 75%.10 To ogous fat grafts with the addition of the stromal vascular fraction (SVF) has been combat this resorption, overcorrection, or injected more fat than is reported to improve long-term volume retention. Stromal vascular fraction is needed, is performed by 78% of plastic surgeons.10 Furthermore, the most commonly isolated using enzymatic digestion, but it is unknown what effect ensuing volume loss leads 46% of plastic surgeons to perform repeat the digestion process has on the adipocytes and SVF cells that comprise the graft. fat grafting procedures 3 to 6 months after the initial procedure. The Some clinicians have reported use of enzymatically digested fat grafts to alter the practices of fat graft harvesting, processing, and injection are being physical properties of the tissue in specialized applications. We have previously studied to determine methods for maximizing fat graft retention at each reported that increasing collagenase digestion duration adversely affects the via- step of the procedure.11–17 However, the American Society of Plastic bility of adipocytes and SVF cells. Here, we aimed to determine if collagenase Surgeons Fat Graft Task Force reports that a lack of data precludes spe- digestion of adipocytes before grafting is detrimental to long-term graft retention cific recommendations for the clinical use of fat grafts with regards to and if SVF supplementation can abrogate these potential deleterious effects. harvesting, preparation or injection.1 Although the use of autologous Methods and Results: We used a published xenograft model in which human fat in the clinical realm is rising, challenges in optimizing long-term lipoaspirate was implanted into the scalp of immunocompromised mice to study fat graft volume retention continue to drive research and innovation. the effects of collagenase digestion on in vivo graft survival after 12 weeks. Although the mechanism(s) responsible for fat graft survival re- We used 4 experimental groups: grafts composed of collagenase-digested and main unclear, proper vascularization and subsequent sufficient oxygen/ nondigested adipocytes (50-minute digestion) and grafts with and without SVF nutrient delivery appear to be critical to fat graft survival.18–21 Clini- supplementation. We used microcomputed tomography to serially and noninva- cians and scientists have aimed to improve the long-term retention rate sively quantify graft volume, in conjunction with hematoxylin-eosin staining of of fat grafts, and these approaches have been the subject of many pub- histological cross-sections of implanted and excised grafts to assess overall tissue lished studies. Central to many recent studies is the “enhancement” of viability. We found that adipocytes that were collagenase-digested before implan- fat grafts with the stromal vascular fraction (SVF), or cultured and seri- tation had significantly lower retention rates at 12 weeks and poorer tissue health, ally passaged SVF, which is described in the field as adipose-derived which was assessed by quantifying the number of intact adipocytes, the number stem cells (ASCs).22–24 The SVF, composed of adipose-resident stro- of cystic formations, and by scoring the degree of inflammation and fibrosis. Fur- mal cells that are easily isolated from lipoaspirate, is thought to include ther, we found that SVF supplementation of the digested grafts improved graft preadipocytes, ASCs, endothelial progenitor cells, macrophages, and survival, but not to the level observed in undigested grafts. vascular cells.25,26 It is believed that the progenitor and stem cells within Conclusions: We conclude that collagenase digestion adversely affects the long- the SVF are responsible for the enhanced fat graft retention after im- term volume retention of fat grafts, but that graft retention is improved by SVF plantation by providing a source for new adipocytes, vascular cells, supplementation. These experimental results can serve as an initial framework and support cells in the grafted tissue.27,28 The SVF is most commonly to further elucidate the reported efficacy and safety of using collagenase- isolated from the adipose tissue using an enzyme such as collagenase, digested fat grafts and SVF in the clinical setting. or an enzymatic “cocktail” which contains various collagenases (ie, Key Words: fat grafting, collagenase, stromal vascular fraction, Liberase Blendzyme).29–31 Other techniques used to isolate SVF in- adipose-derived stem cells clude centrifugation, mechanical isolation via vigorous shaking, and decanting after density gradient separation.32–34 Herein, we focus on (Ann Plast Surg 2017;78: S335–S342) fat grafting techniques that use enzymatic digestion, specifically colla- at grafting is a technique used to achieve soft tissue volume aug- genase, to harvest SVF and enhance fat grafts before grafting. F mentation and contour improvement. There are nearly 50 000 fat Previous studies have reported the use of enzymatic digestion of grafting procedures performed per year in the United States, with an- lipoaspirate material as a means to separate the SVF from adipocytes other 400 000 performed globally.1 Autologous fat has the advantages before using both the SVF and adipocytes as part of the autologous of being plentiful and innately biocompatible without the risk of foreign graft. Typically, the enzymatically isolated SVF is added to adipo- body reaction or allergic reaction.2–4 Despite its promise, autologous cytes that have not been enzymatically digested (reserved portion of lipoaspirate), but reports have described the use of collagenase- fat grafting is hindered by highly variable long-term fat graft volume 22,35–37 retention.5–9 A 2013 survey of American Society of Plastic Surgeons digested adipocytes as part of the graft. In their practice using collagenase-digested adipocytes to surgically correct infraorbital dark circles, Younet al37 cite that collagenase digestion of the adipose tissue Received September 14, 2016, and accepted for publication, after revision January 7, 2017. increased the contouring ability of the fat graft by making it a ”more From the Departments of *Plastic Surgery, †Biomedical Engineering, and ‡Biology, ” University of Virginia, Charlottesville, VA. liquid filler, and they reported safe and efficacious results. Al- J.B.O. and S.A.S. contributed equally to this work. though the use of SVF/adipocytes harvested from collagenase-digested Conflicts of interest and sources of funding: none declared. lipoaspirate is not currently approved for use by the United Stated Food Reprints: Patrick S. Cottler, PhD, Department of Plastic Surgery, P.O. Box 800376, and Drug Administration (FDA), collagenase has gained acceptance in Charlottesville, VA 22908-0376. E-mail: [email protected]. 38,39 Copyright © 2017 Wolters Kluwer Health, Inc. All rights reserved. the treatment of Dupuytren contracture and Peyronie disease. Fur- ISSN: 0148-7043/17/7806–S335 thermore, with the increasing use of enhanced fat grafting, specifically DOI: 10.1097/SAP.0000000000001063 in large volume reconstructions and in cases with patients who have

Annals of Plastic Surgery • Volume 78, Supplement 5, June 2017 www.annalsplasticsurgery.com S335 Olenczak et al Annals of Plastic Surgery • Volume 78, Supplement 5, June 2017

minimal adiposity, it will be important to study the effect of collagenase Isolation of Adipocytes From Collagenase-Digested digestion on the adipocytes. In such circumstances, it may be necessary Human Adipose Tissue to use all of the fat harvested from the patient, including that used to ob- The human adipose tissue was digested in a collagenase contain- tain SVF, to have adequate volume for grafting. Future studies con- ing buffer as described in our previous studies.40,46 Briefly, human tissue ducted on collagenase-digested adipocytes and SVF may provide was digested at a concentration 1 g tissue/mL of collagenase-containing grounds for FDA approval of collagenase-isolated SVF and adipocytes digestion buffer consisting of 0.1% (weight/volume) collagenase Type I in fat grafts. (Worthington Biochemical Corporation, Lakewood, NJ), 2.5% (weight/ In our previously published research we reported the in vitro ef- volume) bovine serum albumin (Jackson ImmunoResearch, West fect of collagenase digestion on adipocytes and SVF for both human Grove, Pa), 20 mM HEPES (Life Technologies, Grand Island, NY), and murine adipose tissue.40 We found that increasing the duration of 200 nM adenosine (Sigma, St. Louis, Mo), 1.2 mM KH2PO4 (Sigma), collagenase digestion is detrimental to both the SVF and the adipocytes, 1.2 mM MgSO ·7H O (Sigma), 120 mM NaCl (Sigma), 4.7 mM KCl causing decreases in the viability of both interstitial cells (which com- 4 2 (Sigma), 1.3 mM CaCl2·2H2O (Fisher Scientific, Pittsburgh, Pa). The prise the SVF compartment) and adipocytes. Collagenase digestion human adipose tissue was incubated in a 37°C water bath under con- affected both adipocyte and interstitial cell viability in a dose- and stant agitation for 50 minutes. After digestion was complete, the sam- time-dependent manner. For human tissue, the viability of adipocytes ples were washed with PBS 3 times for 5 minutes each wash to and SVF was significantly decreased when collagenase digestion times remove residual collagenase. After PBS washes, samples were centri- exceeded 50 minutes. Although our in vitro studies were the first to indi- fuged at 1100 RPM for 5 minutes and the adipocytes were isolated cate the potentially detrimental effects of collagenase digestion on SVF using a micropipette. These adipocytes were termed collagenase- and adipocytes, it is imperative to consider the effect of collagenase diges- digested adipocytes and were stored in sterile microcentrifuge tubes at tion on in vivo graft survival. To our knowledge, no studies to date have room temperature until use. All procedures were performed in a examined the effect of collagenase digestion on in vivo graft survival. biosafety cabinet. In this study, we adopted a previously published xenograft model in which human fat grafts were implanted in the scalps of immunocompromised mice to study the effects of collagenase digestion on in vivo Isolation of Adipocytes From Untreated Human graft survival.41 We used microcomputed tomography (micro-CT) to Adipose Tissue noninvasively monitor the volume of implanted fat grafts over the Adipocytes were also derived from undigested adipose tissue in course of 12 weeks. The micro-CT data was coupled with hematoxylin- the same way as above, but the samples were incubated in collagenase- eosin (H&E) staining to provide information about the overall tissue free buffer for 50 minutes. After incubation in the collagenase-free health and viability. buffer, samples were washed 3 times with PBS as stated above and were A better understanding of the impact of collagenase digestion on centrifuged at 1100 RPM for 5 minutes. The untreated adipocytes were in vivo graft survival is of importance to clinicians and will help clarify isolated using a micropipette and were stored in sterile microcentrifuge the efficacy of using digested adipocytes/SVF during autologous fat tubes at room temperature until use. All procedures were performed in a grafting. This study builds on our in vitro findings and aims to deter- biosafety cabinet. mine the potential detrimental effects of collagenase digestion on in vivo graft retention. SVF Isolation After the collagenase-digested adipocytes were removed via mi- METHODS cropipette as stated above, the remaining portion of the sample was used to isolate the SVF for subsequent studies. After centrifugation, the sam- — Human Adipose Tissue Harvest ple stratifies into 3 layers based on densities the top layer consisted of adipocytes (used as stated above), the middle layer is the tumescent All procedures were performed in accordance with the Univer- fluid and washing fluid, and the bottom layer contains the SVF and sity of Virginia Institutional Review Board. At the time of elective lipo- blood cells. The adipocytes (top layer) were removed for use and the suction, human adipose tissue was obtained from the abdomen, flank – middle layer was aspirated leaving the bottom pellet containing the and thigh using the Coleman technique.42 45 Briefly, through a small in- SVF. One milliliter of DMEM/F12 media (ThermoFisher Scientific, cision a mixed tumescent solution (0.5% lidocaine with 1:200,000 of Waltham, Mass) was added to the pellet and gently mixed by pipetting. epinephrine in Lactated Ringer solution) was infiltrated into the donor The solution was centrifuged at 1100 RPM for 5 minutes and 500 μL site using a blunt Lamis infiltrator (Byron Medical, Inc., Tucson, Ariz.). was removed from the sample. 1 mL of erythrocyte lysis buffer The solution was infiltrated at a ratio of 1 mL of solution per cubic cen- (eBioscience, San Diego, Calif) was added and the solution was gently timeter of fat graft to be harvested. The fat grafts were harvested vortexed to resuspend the pellet. The sample was incubated for through the same incisions made previously. The harvesting cannula 5 minutes at room temperature, which was followed by the addition (3 mm diameter) was connected to a 10-mL Luer-Lok syringe. Gently of 1.5 mL of DMEM/F12. This solution was sterile filtered through a pulling back on the plunger of a 10-mL syringe, with suction generated 40-μm centrifuge tube mesh and centrifuged at 1100 RPM for by no more than 1 to 2 mL of pull on the plunger, provided gentle neg- 5 minutes. After centrifugation, the supernatant was removed, and the ative pressure while the cannula was advanced and retracted through the pellet was resuspended in 1 mL of DMEM/F12. The cells were counted harvest site. After filling the syringe with harvested tissue, the cannula using a hemocytometer and were resuspended in sterile PBS into a was disconnected from the syringe. A Luer-Lok plug was twisted onto known volume for the concentration desired for the subsequent studies. the syringe to seal the Luer-Lok aperture, the plunger was removed from the barrel of the syringe and the body of the filled syringe was placed into a centrifuge (Byron Medical) and spun at 3000 rpm for Preparation of Adipose Grafts 3 minutes. After centrifugation, the oil layer (upper level) was decanted Four separate groups were used for these studies: (1) undigested and the aqueous layer (lower level) was also drained out of the syringe. adipocytes without SVF supplementation, (2) undigested adipocytes The middle layer, composed predominantly of adipose tissue was used with SVF supplementation, (3) digested adipocytes without SVF sup- for this research. The fat graft was stored in a sterile container with plementation, and (4) digested adipocytes with SVF supplementation phosphate-buffered saline (PBS) and transported to the University of (Fig. 1A). All implanted adipose tissue grafts were 250 μL, and those Virginia Plastic Surgery Laboratory. grafts with SVF supplementation contained 12,500 SVF cells/250 μL

S336 www.annalsplasticsurgery.com © 2017 Wolters Kluwer Health, Inc. All rights reserved. Annals of Plastic Surgery • Volume 78, Supplement 5, June 2017 Fat Grafting with Collagenase and SVF Supplement

postimplantation, 4 weeks postimplantation, 8 weeks postimplantation, and 12 weeks postimplantation (Fig. 1B).

Harvest of Adipose Tissue Grafts At the terminal time point of 12 weeks postimplantation, mice were euthanized via CO2 asphyxiation. The skin above the scalp was carefully excised, and the adipose graft was exposed. An image of the adipose graft was taken to provide a macroscopic view of the graft. After imaging, the adipose graft was carefully excised from the skin, and the mass of the graft was recorded. The graft was then fixed in 4% paraformaldehyde at 4°C for 7 days. In some recipient mice (group 3), the adipose graft was undetectable and was not harvested.

Volumetric Analysis of Micro-CT Scans Fiji, an image processing package based on ImageJ (National Institutes of Health, Bethesda, Md), was used to process the DICOM files generated from the micro-CT scans. “Segmentation Editor,” a plugin within Fiji, was used to analyze the DICOM files and to quantify the graft volume. The DICOM files were imported into Fiji as an “ ” FIGURE 1. Experimental groups and experimental outline of Image Sequence and converted to 8-bit grayscale (Fig. 2A). The “ ” xenograft model. A, Experimental groups 1–4 consisting of imported stack of images was loaded into Segmentation Editor and collagenase-digested and undigested adipocytes with or without the freehand tracing tool was used to outline the adipose graft on SVF supplementation. B, Experimental timeline depicts serial every fifth slice until the adipose graft was no longer visualized micro-CT scanning time points and the terminal time point (Fig. 2B). The adipose graft was distinct from the surrounding skin at which harvested tissue is massed and processed for based on the color and contrast of the image. After tracing every fifth H&E staining. slice, interpolation was performed to create selections/traces for slices between the previously traced slices. A 3D projection of these slices was constructed, and the volume of the adipose tissue graft was of fat.24 Grafts were prepared based on the groups above and were loaded calculated after setting the voxel size for scaling purposes (Figs. 2C, D). into 1-mL syringes coupled to a sterile 16-gauge catheter tubing for injection. H&E Staining and Scoring After fixation, adipose grafts were processed for paraffin em- Implantation of Adipose Grafts bedding, and 5 μm samples were cut and stained with H&E. Indi- vidual 100X images were acquired using an EVOS XL microscope Mouse experiments were performed under a protocol approved (ThermoFisher Scientific) and montages were made of the entirety of by the University of Virginia Institutional Animal Care and Use Com- the adipose graft for analysis. An established and widely applied scor- mittee (Animal Welfare Assurance A3245-01) in accordance with the ing system was used to assess histologic parameters as follows: pres- National Institutes of Health's Guide for the Care and Use of Laboratory ence of intact and nucleated fat cells; presence of cysts and vacuoles; Animals. Mice were housed in an AAALAC-accredited facility. All inflammation, as evidenced by infiltration of lymphocytes and macro- procedures were performed under anesthesia, and all efforts were made phages; and the presence of fibrosis and other components of connec- to minimize pain or suffering. – tive tissue.47 52 Each of these parameters were graded on a scale of 0 A xenograft adipose tissue graft model was modified in which to 5 by evaluation of the presence of the histologic parameters as fol- adipose grafts were implanted in the scalp of recipient mice.41 NOD- lows: 0 = absence, 1 = minimal presence, 2 = minimal to moderate pres- SCID mice (Charles River, Washington, Mass) were anesthetized with ence, 3 = moderate presence, 4 = moderate to extensive presence, 2% isoflurane and the hair on the top of head and around the ears was 5 = extensive presence. Montaged images were scored by 3 blinded, shaved and depilated. After depilation, the scalp was sterilized with trained observers. Representative examples of fibrosis, cysts/vacuoles, 3 alternating ethanol and povidone-iodine wipes. A small access inci- and inflammation are depicted in Figure 3A-C as well as average scores sion as made at the base of the skull and a subcutaneous tunnel was cre- in Figure 3D. ated for implantation of the fat graft. 250 μL of isolated fat was injected with a 1-mL syringe through a 16-gauge Teflon cannula in a linear ret- rograde fashion. The incision was closed with 5-0 nylon sutures and the RESULTS animal was recovered, monitored postoperatively, and returned to the animal vivarium. Collagenase Digestion of Adipocytes Before Implantation Decreases Volume Retention, But SVF Supplementation Microcomputed Tomographic Imaging Aids in Volume Retention of Collagenase-Digested Grafts A Scanco vivaCT 40 micro-CT instrument (Scanco Medical, To determine the effect of collagenase digestion on the volume Brüttisellen, Switzerland) was used to scan recipient mice at various retention of the fat grafts, we used micro-CT to calculate graft volumes time points after implantation of human adipose tissue grafts as postimplantation. Calculation of adipose graft volumes from 3D projec- depicted in Figure 1B. Micro-CT scanning allowed for noninvasive tions revealed that collagenase digestion of adipocytes before grafting quantification of the graft volume and provided topographical informa- significantly reduced volume retention post-implantation. The graft tion of the adipose graft. Mice were anesthetized with 2% isoflurane be- volume of the collagenase treated group with SVF supplementation fore and during the entirety of the scanning procedure. Settings used on (group 4, 83.90 mm3) was roughly one third of the initial grafted vol- the micro-CT instrument were as follows: tube voltage, 55 kV; tube cur- ume (250 mm3) 1 week after implantation (Fig. 2D). The group that rent, 145 pA; power, 8 W; integration 1 week postimplantation, 2 weeks was collagenase digested without SVF supplementation (group 3) was

FIGURE 2. micro-CT scanning of fat grafts. Micro-CT allows for noninvasive tracking of graft volume up to 12 weeks. A, Example image of micro-CT scan performed on recipient mice. Scale bar = 1 cm. B, Red area depicts outlined adipose graft using segmentation editor. Graft is distinguished from skin by contrast. Scale bar = 1 cm. C, Three-dimensional volume render constructed from outlines of adipose graft. Calculation of graft volume is performed in Segmentation Editor and is reported in Panel D. Scale bar = 0.4 cm. D, Average graft volumes for the experimental groups at time points post-implantation. Two-way ANOVA, *P value < 0.001. All data are mean + standard deviation. n = 6 animals for each group. ANOVA, analysis of variance. nearly undetectable via micro-CT scanning after 1 week, with an average SVF supplemented group being slightly lower, but not statistically volume of only 0.87 mm3. The groups containing adipocytes that lower, than the unsupplemented groups at all time points (Fig. 2D). were not collagenase digested before implantation (group. 1 and 2) had much higher volumes immediately postimplantation (Fig. 2D). The group with undigested adipocytes without SVF supplementation Terminal Adipose Graft Masses Are Consistent With (group 1) had an average volume of 192.06 mm3, whereas the group Graft Volumes with undigested adipocytes with SVF supplementation (group 2) had After 12 weeks, the adipose grafts were harvested and weighed an average volume of 163.21 mm3 (Fig. 2D). to calculate the terminal graft volume. The masses of the undigested This decrease in graft volume for digested adipocytes was ob- groups without SVFand with SVF were 0.116 and 0.104 g, respectively served for all time points postimplantation up to 12 weeks. At the termi- (Fig. 3E). The initial graft was assumed to be 0.225 g (0.9 g/mL density nal time point of 12 weeks, the collagenase-digested group without of adipose tissue34), this equated to roughly 40% volume retention rate SVF (group 3) was undetectable and the collagenase-digested group after 1 weeks, which is consistent with the literature.35 These masses with SVF supplementation (group 4) had an average volume of were also consistent with the trends of the graft volumes at 12 weeks 4.98 mm3. The undigested group without SVF had an average volume and the 40% to 50% volume retention rate (above). The unsupplemented of 109.69 mm3, whereas the undigested group with SVF had an average grafts that were collagenase digested before implantation (group 3) volume of 86.01 mm3 (Fig. 2D). Based on these data, it appears that were undetectable at 12 weeks and not harvested (Fig. 3E). Consistent collagenase digestion before grafting is detrimental to the retention of with the graft volumes determined using micro-CT scanning, the the graft (Fig. 2D). These data also suggest that SVF supplementation collagenase digested grafts with SVF supplementation (group 4) had of the collagenase-digested adipose grafts increased the volume an average mass of 0.017 g (Fig. 3E). Taken together, these data retention and mitigated the resorption of the graft post-implantation. corroborate that this level of collagenase digestion is detrimental to Interestingly, SVF supplementation of undigested grafts (group. 1 and graft survival and that SVF supplementation of digested grafts aids in 2) did not increase the volume retention, with average volumes of the volume retention.

S338 www.annalsplasticsurgery.com © 2017 Wolters Kluwer Health, Inc. All rights reserved. Annals of Plastic Surgery • Volume 78, Supplement 5, June 2017 Fat Grafting with Collagenase and SVF Supplement

FIGURE 3. Histological scoring of fat grafts. Collagenase-digested grafts were qualitatively less healthy as compared to undigested grafts. A, Example image of H&E stain depicting the presence of cysts and vacuoles. *Cyst or vacuole. B, Example image of H&E stain depicting the presence of fibrosis. +Presence of fibrosis. C, Example image of H&E stain depicting the presence of inflammation. # Presence of inflammation. D, Average histological scores for each of the 4 metrics quantified. One-way ANOVA between each metric, *P <0.05,**P < 0.001. E, Average mass of grafts at 12 weeks compared to assumed mass of grafts at initial time point. Grafts were not detectable for group 3. One-way ANOVA, **P < 0.001. All scale bars = 100 μm. All data are mean + standard deviation. n = 6 animals for each group.

Collagenase Digestion Decreases Overall Tissue Health as on the overall tissue health. These are in agreement with volume Evidenced by H&E Scoring Metrics retention outcomes and suggest that collagenase digestion is detrimental After harvest, we scored the overall health of the fat grafts by to graft retention and overall tissue health. using H&E staining and previously published protocols.47–52 The collagenase-digested grafts (group. 3 and 4) had, on average, higher scores for inflammation, cysts/vacuole presence, and fibrosis when DISCUSSION compared with grafts that had not been collagenase digested before The use of autologous fat grafting to address soft tissue defects implantation (group. 1 and 2) (Fig. 3D). Because there were no grafts has gained widespread acceptance in plastic surgery over the past de- remaining at week 12 for group 3 (collagenase digested, no SVF supple- cade.4,53 Autologous fat is biocompatible, easy to harvest, and allows mentation), H&E scoring was not performed on this group. There was a clinicians to avoid the use of synthetic or foreign implant materials. significant increase in the amount of inflammation present, the number Long-term resorption rates of autologous fat grafts range from 25% to of cysts/vacuoles observed, and the amount of fibrosis in collagenase- 80% and are unpredictable, which has prompted clinicians and scien- digested when compared to undigested grafts (Fig. 3D). There was tists to develop strategies to combat this adverse resorption process. Cli- also a striking difference in the membrane integrity of the groups. The nicians have started to “enhance” fat grafts by supplementing with the undigested grafts (group. 1 and 2) had much more uniform, intact SVF that is most commonly isolated from lipoaspirate using enzymatic adipocytes than the collagenase-digested group (Fig. 3D). There were digestion (ie, collagenase digestion). Although other techniques can be no statistically significant differences between group. 1 and 2 for any used to isolate the SVF (mechanical, centrifugation, and so on), our cur- metric, suggesting that SVF supplementation did not have an impact rent studies focused on fat grafting techniques that utilize enzymatic

digestion, specifically collagenase, before fat grafting. Many fat observed an improvement in volume retention for SVF-treated grafts grafting studies have reported the use of enzymatic digestion to isolate that were collagenase-digested before implantation. Indeed, supple- the SVF and subsequent supplementation of the adipocyte portion with mentation of collagenase-digested grafts with SVF before implanta- enzymatically isolated SVF.22,35,37 Additionally, other groups have used tion delayed the resorption and increased the volume of the graft at the collagenase-digested adipocyte portion as part of the graft without each time point. Our findings are consistent with many other groups having a thorough understanding of the effect of digestion on the viabil- that have reported increased volume retention and delayed resorption ity of adipocytes.37 In our previous study, we found that increasing the after the supplementation of grafts with SVF.24,33 It is also possible duration of collagenase digestion decreased the viability of adipocytes that the collagenase digestion “stripped” the adipose tissue graft of and the interstitial cells within the adipose tissue.40 In the present study, the interstitial cells, and supplementing the adipose graft with SVF cells we examined the effect of collagenase digestion on adipocytes and the “repopulated” the interstitial cell compartment, which appears to be impact of digestion on in vivo graft survival. We modified a published necessary for graft survival. We have previously reported that increas- xenograft model and implanted human lipoaspirate as a fat graft and ing collagenase digestion decreases the number of interstitial cells then serially evaluated volume retention and tissue health over a within adipose tissue.40 This “stripping” of the interstitial cell compart- 12-week time course.41 We found that collagenase digestion before ment may also provide further explanation as to why collagenase diges- grafting significantly decreased the long-term volume retention of the tion negatively impacts graft retention. implanted adipose tissue, as evidenced by noninvasive micro-CT, and Interestingly, SVF supplementation of grafts that were not colla- decreased the overall tissue health, as determined by H&E staining. Fur- genase digested had no effect on the overall volume retention for the ther, we found that SVF supplementation of collagenase-digested grafts experimental time points. This finding differs from what has been re- slightly improved the volume retention, but not to the level observed ported in the literature, which is that SVF supplementation improves using undigested grafts. volume retention.59 The cell dosage that we used (12,500 SVF cells Collagenase digestion of the adipocytes before implantation not for each 250 μL fat graft) was based on a similar study conducted by only adversely affected the long-term volume retention and overall tis- Paik et al. where the concentration of SVF cells within adipose tissue sue health, but micro-CT scans and volume calculations revealed rapid was titrated to achieve an optimal effect.24 They reported that concen- resorption of collagenase-digested adipose at the earliest time point trations of SVF cells above a certain threshold (10 million cells/mL) (1 week). The collagenase-digested grafts were nearly undetectable, decreased tissue health and caused greater lipodegeneration in a similar and the collagenase digested grafts with SVF supplementation retained xenograft model. Paik et al. further reported that 50,000 SVF cells/mL only 40% of their initial volume. The H&E evaluation revealed an over- adipose tissue was an ideal concentration, and hence this was the SVF all decrease in tissue health for collagenase-digested grafts, with the ad- concentration used in our study. However, this concentration is much ipocytes having lower membrane integrity scores (more fragmented lower than concentrations demonstrated in other studies to signifi- adipocytes), higher levels of inflammation/fibrosis, and a more significantly increase volume retention. For example, Gentile et al. supple- cant presence of cysts and vacuoles when compared to adipocytes that mented clinical fat grafts with 250,000 SVF cells/mL and reported were not collagenase digested. The mass of the collagenase-digested less volume loss over 18 months.59 Studies which supplement grafts grafts was also significantly lower at the terminal time point. Taken to- with ex vivo expanded ASCs have also used much higher concentra- gether, these findings suggest that this level of collagenase-digestion tions of cells per volume of adipose tissue, with concentrations as high before grafting is detrimental to the long-term volume retention of as 20 Â 106 ASCs/mL fat.2 Because we did not conduct a dose- fat grafts. response study, we cannot draw conclusions about whether higher (or Currently, it is unclear why the collagenase digestion of adipo- lower) doses of SVF might yield an improvement in fat graft volume re- cytes before implantation/grafting decreases the volume retention, tention in this murine model. but previous studies may provide some insight. In our previous in There are several other caveats to our study. First, we examined vitro studies, for example, we observed that increasing durations of the effect of collagenase digestion on in vivo graft retention by using exposure to collagenase had an increasingly adverse effect on human a single digestion duration (50 minutes). This digestion duration was adipocytes and interstitial cell viability.40 Our previous in vitro exper- based on our previous study in which we found that 50 minutes was iments suggest that in the present study, the collagenase-digested the threshold beyond which adipocytes and interstitial cells demon- grafts had reduced viability at the onset of grafting. In another study strated decreased viability.40 In future work, it will be important to eval- conducted by Ruan et al,54 isolation of primary adipocytes from mu- uate both shorter and longer collagenase digestion times to more fully rine epididymal fat pads via collagenase digestion increased several elucidate the effects of collagenase digestion on in vivo graft survival. genes that encoded several proinflammatory mediators. Specifically, Indeed, it is possible that shorter collagenase digestion times (eg, less Ruan et al reported a significant increase in TNF-α expression by than 50 minutes) may be beneficial to graft survival without jeopardiz- primary adipocytes that were isolated using collagenase digestion. ing the viability of the adipocytes within the graft. Short digestion du- Based on these findings, we speculate that collagenase digestion of rations may, in fact, benefit graft survival by increasing the surface the adipocytes before grafting induces an inflammatory environment area-to-volume ratio (eg, smaller parcels of fat), which could promote within the graft via TNF-α secretion that leads to increased resorp- more rapid revascularization. Another important consideration in our tion of the graft. In support of this theory, Thompson et al55 showed work is the fact that all studies were conducted in immunocompro- that IL-6 levels are also altered in adipocytes during collagenase di- mised mice (NOD-SCID mice), which lack a complete immune sys- gestion and that varying collagenase concentrations/rotational tem. Although the xenograft model allows for the grafting of human speeds during digestion modulates the IL-6 expression. IL-6 and adipose tissue into a mouse without rejection, the lack of a complete TNF-α have been linked to obesity-related inflammation and are immune system omits a significant factor in the grafting and inflam- generally considered pro-inflammatory cytokines.56,57 Therefore, fu- matory process. Specifically, these mice have defective T and B cell ture studies investigating the effects of collagenase on adipose tissue development and have deficient natural killer cell function, and it is retention should examine levels of these two pro-inflammatory cyto- possible that these cell types play a significant role in the resorption kines as a function of digestion time as well as in the tissue after process.60,61 Despite these caveats, this study constitutes an important graft implantation.58 follow-on to our previous in vitro experiments, and builds on those re- Both preclinical and clinical studies have reported significant sults by highlighting what could be a clinically important drawback of improvements in fat graft volume retention by “enhancing” or collagenase digestion that is only partially combatted by supplementa- supplementing the fat grafts with SVF before implantation. We tion with SVF.

S340 www.annalsplasticsurgery.com © 2017 Wolters Kluwer Health, Inc. All rights reserved. Annals of Plastic Surgery • Volume 78, Supplement 5, June 2017 Fat Grafting with Collagenase and SVF Supplement

CONCLUSIONS 22. Yoshimura K, Sato K, Aoi N, et al. Cell-assisted lipotransfer for cosmetic breast augmentation: supportive use of adipose-derived stem/stromal cells. Aesthetic This study extends our previous in vitro finding, that collagenase Plast Surg. 2008;32:48–55. discussion 56–47. digestion is detrimental to adipocyte and interstitial cell viability, to a 23. Garza RM, Rennert RC, Paik KJ, et al. Studies in fat grafting: part IV. clinically relevant in vivo model. The current study suggests that colla- Adipose-derived stromal cell gene expression in cell-assisted lipotransfer. Plast genase digestion of adipose grafts before implantation adversely affects Reconstr Surg. 2015;135:1045–1055. the long-term in vivo retention over the course of 12 weeks. Further, we 24. Kevin J, Paik AB, Elizabeth R, et al. Studies in fat grafting: Part V.Cell-assisted show that supplementing collagenase-digested fat grafts with SVF en- lipotransfer to enhance fat graft retention is dose dependent. Plast Reconstr hances volume retention and delays the resorption process. Our initial Surg. 2015. preclinical findings begin to help to clarify the safety and efficacy of 25. Bourin P, Bunnell BA, Casteilla L, et al. Stromal cells from the adipose tissue- derived stromal vascular fraction and culture expanded adipose tissue-derived using enzymatically digested adipocytes and SVF during autologous stromal/stem cells: a joint statement of the International Federation for Adipose fat grafting. Further, this study helps to define the risks and benefits that Therapeutics and Science (IFATS) and the International Society for Cellular Ther- different adipose treatments pose and may guide protocols for process- apy (ISCT). Cytotherapy. 2013;15:641–648. ing and preparing fat grafts in the clinical setting. 26. Gimble JM, Bunnell BA, Chiu ES, et al. Concise review: adipose-derived stromal vascular fraction cells and stem cells: let's not get lost in translation. Stem Cells. 2011;29:749–754. 27. Fu S, Luan J, Xin M, et al. Fate of adipose-derived stromal vascular fraction cells REFERENCES after co-implantation with fat grafts: evidence of cell survival and differentiation in – 1. Surgeons, A. S. o. P.Fat Transfer/Fat Graft and Fat Injection ASPS Guiding Prin- ischemic adipose tissue. Plast Reconstr Surg. 2013;132:363 373. ciples. http://www.plasticsurgery.org, 2009. 28. Gimble JM, Katz AJ, Bunnell BA. Adipose-derived stem cells for regenerative – 2. Kølle SF, Fischer-Nielsen A, Mathiasen AB, et al. Enrichment of autologous fat medicine. Circ Res. 2007;100:1249 1260. grafts with ex-vivo expanded adipose tissue-derived stem cells for graft survival: 29. Agrawal H, Shang H, Sattah AP,et al. Human adipose-derived stromal/stem cells a randomised placebo-controlled trial. Lancet. 2013;382:1113–1120. demonstrate short-lived persistence after implantation in both an immunocompe- 3. Trojahn Kølle SF, Oliveri RS, Glovinski PV, et al. Importance of mesenchymal tent and an immunocompromised murine model. Stem Cell Res Ther. 2014;5:142. stem cells in autologous fat grafting: a systematic review of existing studies. 30. Gimble J, Guilak F. Adipose-derived adult stem cells: isolation, characterization, J Plast Surg Hand Surg. 2012;46:59–68. and differentiation potential. Cytotherapy. 2003;5:362–369. 4. Coleman SR. Structural fat grafting: more than a permanent filler. Plast Reconstr 31. Aronowitz JA, Lockhart RA, Hakakian CS. Mechanical versus enzymatic iso- Surg. 2006;118:108S–120S. lation of stromal vascular fraction cells from adipose tissue. Springerplus. 5. Delay E, Garson S, Tousson G, et al. Fat injection to the breast: technique, results, 2015;4:713. and indications based on 880 procedures over 10 years. Aesthet Surg J.2009;29: 32. Dong Z, Luo L, Liao Y,et al. In vivo injectable human adipose tissue regeneration 360–376. by adipose-derived stem cells isolated from the fluid portion of liposuction aspi- 6. Niechajev I,Sevćuk O. Long-term results of fat transplantation: clinical and histo- rates. Tissue Cell. 2014;46:178–184. logic studies. Plast Reconstr Surg. 1994;94:496–506. 33. Raposio E, Caruana G, Bonomini S, et al. A novel and effective strategy for the 7. Rubin A, Hoefflin SM. Fat purification: survival of the fittest. Plast Reconstr isolation of adipose-derived stem cells: minimally manipulated adipose-derived Surg. 2002;109:1463–1464. stem cells for more rapid and safe stem cell therapy. Plast Reconstr Surg.2014; – 8. Wolf GA, Gallego S, Patrón AS, et al. Magnetic resonance imaging assessment of 133:1406 1409. gluteal fat grafts. Aesthetic Plast Surg. 2006;30:460–468. 34. Shah FS, Wu X, Dietrich M, et al. A non-enzymatic method for isolating human – 9. Zocchi ML, Zuliani F. Bicompartmental breast lipostructuring. Aesthetic Plast adipose tissue-derived stromal stem cells. Cytotherapy. 2013;15:979 985. Surg. 2008;32:313–328. 35. Gentile P, De Angelis B, Pasin M, et al. Adipose-derived stromal vascular fraction 10. Cheng A, Lakhiani C, Rojas K, et al. Current fat grafting practices for cosmetic cells and platelet-rich plasma: basic and clinical evaluation for cell-based therapies and reconstructive breast surgery—a national survey of ASPS members.pdf. Plast in patients with scars on the face. J Craniofac Surg. 2014;25:267–272. Reconstr Surg. 2013;132:134. 36. Li J, Gao J, Cha P, et al. Supplementing fat grafts with adipose stromal cells for 11. Gir P, Brown SA, Oni G, et al. Fat grafting: evidence-based review on autologous cosmetic facial contouring. Dermatol Surg. 2013;39:449–456. fat harvesting, processing, reinjection, and storage. Plast Reconstr Surg.2012; 37. Youn S, Shin JI, Kim JD, et al. Correction of infraorbital dark circles using 130:249–258. collagenase-digested fat cell grafts. Dermatol Surg. 2013;39:766–772. 12. Erdim M, Tezel E, Numanoglu A, et al. The effects of the size of liposuction 38. Gilpin D, Coleman S, Hall S, et al. Injectable collagenase Clostridium cannula on adipocyte survival and the optimum temperature for fat graft histolyticum: a new nonsurgical treatment for Dupuytren's disease. JHandSurg storage: an experimental study. J Plast Reconstr Aesthet Surg. 2009;62: [Am]. 2010;35:2027.e1–2038.e1. 1210–1214. 39. Honig SC. Intralesional collagenase in the treatment of Peyronie's disease. Ther 13. Fisher C, Grahovac TL, Schafer ME, et al. Comparison of harvest and processing Adv Urol.2014;6:47–53. techniques for fat grafting and adipose stem cell isolation. Plast Reconstr Surg. 2013;132:351–361. 40. Seaman SA, Tannan SC, Cao Y, et al. Differential effects of processing time and duration of collagenase digestion on human and murine fat grafts. Plast Reconstr 14. Smith P, Adams WP Jr, Lipschitz AH, et al. Autologous human fat grafting: effect Surg. 2015;136:189e–199e. of harvesting and preparation techniques on adipocyte graft survival. Plast Reconstr Surg. 2006;117:1836–1844. 41. Atashroo DA, Paik KJ, Chung MT, et al. Assessment of viability of human fat injection into nude mice with micro-computed tomography. J Vis Exp. 15. Cheriyan T, Kao HK, Qiao X, et al. Low harvest pressure enhances autologous fat 2015:e52217. graft viability. Plast Reconstr Surg. 2014;133:1365–1368. 42. Coleman SR. Facial recontouring with lipostructure. Clin Plast Surg. 1997;24: 16. Condé-Green A, de Amorim NF, Pitanguy I. Influence of decantation, washing 347–367. and centrifugation on adipocyte and mesenchymal stem cell content of aspirated adipose tissue: a comparative study. J Plast Reconstr Aesthet Surg.2010;63: 43. Coleman SR. Hand rejuvenation with structural fat grafting. Plast Reconstr Surg. – – 1375–1381. 2002;110:1731 1744. discussion 1745 1737. 17. Xie Y, Zheng D, Li Q, et al. The effect of centrifugation on viability of fat grafts: an 44. Coleman SR. Structural fat grafts: the ideal filler? Clin Plast Surg.2001;28: evaluation with the glucose transport test. J Plast Reconstr Aesthet Surg.2010;63: 111–119. 482–487. 45. Pu LL, Coleman SR, Cui X, et al. Autologous fat grafts harvested and refined by 18. Sommer B, Sattler G. Current concepts of fat graft survival: histology of aspirated the Coleman technique: a comparative study. Plast Reconstr Surg. 2008;122: adipose tissue and review of the literature. Dermatol Surg. 2000;26:1159–1166. 932–937. 19. Topcu A, Aydin OE, Ünlü M, et al. Increasing the viability of fat grafts by vascular 46. Mendel TA, Clabough EB, Kao DS, et al. Pericytes derived from adipose-derived endothelial growth factor. Arch Facial Plast Surg. 2012;14:270–276. stem cells protect against retinal vasculopathy. PLoS One. 2013;8:e65691. 20. Kerrigan CL. Skin flap failure: pathophysiology. Plast Reconstr Surg.1983;72: 47. Shoshani O, Livne E, Armoni M, et al. The effect of interleukin-8 on the viability 766–777. of injected adipose tissue in nude mice. Plast Reconstr Surg. 2005;115:853–859. 21. Agha RA, Goodacre T, Orgill DP.Use of autologous fat grafting for reconstruction 48. Garza RM, Paik KJ, Chung MT, et al. Studies in fat grafting: Part III. Fat grafting postmastectomy and breast conserving surgery: a systematic review protocol. irradiated tissue—improved skin quality and decreased fat graft retention. Plast BMJ Open. 2013;3:e003709. Reconstr Surg. 2014;134:249–257.

49. Atik B, Oztürk G, Erdoğan E, et al. Comparison of techniques for long-term 56. Makki K, Froguel P,Wolowczuk I. Adipose tissue in obesity-related inflammation storage of fat grafts: an experimental study. Plast Reconstr Surg. 2006;118: and insulin resistance: cells, cytokines, and chemokines. ISRN Inflamm.2013; 1533–1537. 2013:139239. 50. Lee JH, Kirkham JC, McCormack MC, et al. The effect of pressure and shear on 57. Shi C, Zhu L, Chen X, et al. IL-6 and TNF-α induced obesity-related inflamma- autologous fat grafting. Plast Reconstr Surg. 2013;131:1125–1136. tory response through transcriptional regulation of miR-146b. JInterferonCyto- 51. Ramon Y, Shoshani O, Peled IJ, et al. Enhancing the take of injected adipose tissue by kine Res. 2014;34:342–348. a simple method for concentrating fat cells. Plast Reconstr Surg. 2005;115:197–201. 58. Coppack SW. Pro-inflammatory cytokines and adipose tissue. Proc Nutr Soc. 52. Yi C, Pan Y, Zhen Y, et al. Enhancement of viability of fat grafts in nude mice by 2001;60:349–356. endothelial progenitor cells. Dermatol Surg. 2006;32:1437–1443. 59. Gentile P, Orlandi A, Scioli MG, et al. A comparative translational study: the com- 53. Gir P, Oni G, Brown SA, et al. Human adipose stem cells: current clinical applica- bined use of enhanced stromal vascular fraction and platelet-rich plasma improves tions. Plast Reconstr Surg. 2012;129:1277–1290. fat grafting maintenance in breast reconstruction. Stem Cells Transl Med.2012;1: – 54. Ruan H, Zarnowski MJ, Cushman SW,et al. Standard isolation of primary adipose 341 351. cells from mouse epididymal fat pads induces inflammatory mediators and down- 60. Christianson SW, Greiner DL, Hesselton RA, et al. Enhanced human CD4+ T cell regulates adipocyte genes. JBiolChem. 2003;278:47585–47593. engraftment in beta2-microglobulin-deficient NOD-scid mice. JImmunol.1997; – 55. Thompson AC, Nuñez M, Davidson R, et al. Mitigation of isolation-associated 158:3578 3586. adipocyte interleukin-6 secretion following rapid dissociation of adipose tissue. 61. Shultz LD, Ishikawa F, Greiner DL. Humanized mice in translational biomedical J Lipid Res. 2012;53:2797–2805. research. Nat Rev Immunol. 2007;7:118–130.

A Formula for Planning and Predicting Postoperative Mammoplasty Results Mary G. Smithson, BS,* Sherry S. Collawn, MD, PhD,† Mina S. Mousa, MD,‡ and Carly M. Bramel, PA-C§

reverting skeletal disturbances.2,3 Mammoplasty techniques were first Purpose: For women with macromastia, reduction mammoplasty is a safe and ef- described by Lötsch and Dartigues in 1923 and 1924, respectively, fective solution to increasing quality of life through alleviating pain and improv- and were later reintroduced in 1970 by Lassus, using a vertical ap- ing aesthetics. This study developed a way to combine a surgeon’s view of breast 4,5 ’ proach. The vertical approach led to a cascade of new techniques in- measurement (volume) with a patient s view of breast measurement (distance be- cluding Marchac and de Olarte’s inverted T-incision in 1982, Lejour’s tween nipple and notch, inframammary fold, or midline) to provide patients with introduction of liposuction in the 1990s, Hammond’sSPAIR*tech- a better understanding of expected surgical outcomes after breast reduction with a nique in 1998, the Hall-Findlay technique in 1999, and the SPEAR** medial superior pedicle. technique, which combines Hammond’s skin incision with Hall- Patient and Methods: An institutional review board approved retrospective chart Findlay’s utilization of the superomedial dermoglandular pedicle, in review was performed on all medial superior pedicle reduction mammoplasties 2003.5 Whereas the inverted T-incision or Wise pattern closure with performed by a single surgeon at a university medical center from 2008 to the inferior pedicle has been preferred for large breast reductions,6 the 2016, and a total of 133 patients were identified. Measurements of interest for this SPEAR technique has commonly been used for mild to moderate re- study were nipple to sternal notch (N-S), nipple to inframammary fold (N-I), nip- ductions,5 but the superomedial pedicle vertical scar breast reduction ple to midline (N-M), and breast diameter (BD). The average bilateral change per can also be used for larger reductions.7 After our review of the literature, measurement was calculated for each patient in centimeters. Change was aver- quantitative measurements of change per volume have yet to be identi- aged for left and right breasts for N-S, N-I, N-M, and BD per patient. Grams re- fied when using a technique similar to the Spear technique, and we will moved for left and right breasts were also averaged. Each measurement of average now present our measurements. change was divided by the gram average and multiplied by 100 to obtain centime- Quantitative measurements are important not only for the sur- ter change per 100 grams. Individual patient measurements per type of measure- geon but also for the patient. Volume removed is the norm for quantify- ment were averaged to achieve a final improvement reported in centimeters per ing and discussing reductions among surgeons and is how expectations 100 g tissue removed per breast. are communicated to patients.8–11 However, follow-up and outcomes Results: The average change in the N-S distance was calculated to be a decrease are typically measured using patient satisfaction surveys.12 Although of 1.5 ± 0.8 cm/100 g of breast tissue removed. The average change in N-I was surgeons feel comfortable with this, many patients cannot envision re- calculated to be an overall decrease of 0.7 ± 0.5 cm/100 g. The average change sults before surgery based on volume alone and are forced to rely on in N-M was calculated to be a decrease of 0.1 ± 0.3 cm/100 g. Finally, the average past patient subjective satisfaction surveys. This is important because, change in BD was calculated to be 0.0 ± 0.4 cm/100 g. Conclusions: Asurgeon’s expression of breast measurements in terms of volume can be difficult for a patient to understand and visualize. This study determined the impact volume has on length of typical breast measurements to increase patients’ understanding of expected outcomes. In summary, patients can be told to expect to see a nipple elevation of 1.5 cm per 100 grams of breast tissue removed using this medial superior pedicle technique. Key Words: macromastia, medial superior pedicle reduction, surgical outcomes, volume and breast measurements, nipple to sternal notch (N-S), nipple to inframammary fold (NI), nipple to midline (N-M), and breast diameter (BD), nipple sensation, nipple elevation (Ann Plast Surg 2017;78: S343–S346)

or women with macromastia, reduction mammoplasty is a safe and F effective solution to increasing quality of life through alleviating pain and improving aesthetics.1 Breast reductions have been shown to improve cervical and lumbar lordosis, thoracic kyphosis, and lumbosa- cral angle as well as decreasing patient neck, back, and lumbar pain and

Received October 14, 2016, and accepted for publication, after revision December 24, 2016. From the *UAB School of Medicine, Plastic Surgery, †UAB Division of Plastic Surgery/Department of Surgery, Birmingham, AL; ‡University of South Florida Morsani College of Medicine, Tampa, FL; and §University of Alabama School of Health Professions, Birmingham, AL. Conflicts of interest and sources of funding: none declared. Reprints: Sherry S. Collawn, MD, PhD, University of Alabama, FOT 1102 1720, 2nd Ave, South Birmingham, AL 35294. E-mail: [email protected]. Copyright © 2017 Wolters Kluwer Health, Inc. All rights reserved. ISSN: 0148-7043/17/7806–S343 DOI: 10.1097/SAP.0000000000001038 FIGURE 1. Medial superior pedicle suspension to pectoralis fascia.

Annals of Plastic Surgery • Volume 78, Supplement 5, June 2017 www.annalsplasticsurgery.com S343 Smithson et al Annals of Plastic Surgery • Volume 78, Supplement 5, June 2017

Measurements of interest for this study were nipple to sternal notch (N-S), nipple to inframammary fold (N-I), nipple to midline (N-M), and breast diameter (BD) (Fig. 2). Only 31 of the 133 patient charts included both preoperative and postoperative measurements for at least one of the measurements of interest. All 31 patients had preoperative and 4-week postoperative measurements of N-S, 25 of 31 had preoperative and postoperative measurements of N-I, 16 of 31 had preoperative and postoperative measurements of N-M, and 14 of 31 had preoperative and postoperative measurement of BD. Patients were also asked about bilateral nipple sensation pre-op and post-op with 27 of the 133 reporting either an increase, decrease, or the same nipple sensation as before surgery. Figure 3 demonstrates a patient undergoing breast reduction with this technique. The average bilateral change per measurement was calculated for each patient in centimeters. Change was averaged for left and right breasts for N-S, N-I, N-M, and BD per patient. Figure 4 shows the FIGURE 2. Measurements of interest in this study. changes in N-S measurements. Grams removed for left and right breast were also averaged. Each measurement of average change was divided by the gram average and multiplied by 100 to obtain centimeter change although many patients are happy with their results, there is a subset that per 100 g. Individual patient measurements per type of measurement would like either smaller or larger breasts.13 One way to alleviate this were averaged to achieve a final improvement reported in centimeters unhappiness with the results experienced by some patients is to provide per 100 g of tissue removed per breast. a better way for patients to understand and envision their outcomes. This study developed a way to combine a surgeon’sviewof breast measurement (volume) with a patient’s view of breast measure- RESULTS ment (distance between nipple and notch, fold, or midline) to provide For the 31 patients with preoperative and postoperative measure- patients with a better understanding of expected outcomes. ments of N-S, the average change in N-S per patient ranged from an increase of 0.3 cm/100 g to a decrease of 4.1 cm/100 g. Individual patient averages were combined and averaged to obtain an overall decrease of PATIENTS AND METHODS 1.5 ± 0.8 cm/100 g. Twenty-five patients had preoperative and postop- An institutional review board–approved retrospective chart re- erative measurements of N-I. The average change in N-I per patient view was performed on all mammoplasties performed by a single sur- ranged from no change (0.0 cm/100 g) to a decrease of 1.5 cm/100 g. geon at a university medical center from 2008 to 2016. A total of Patient averages were averaged to obtain an overall decrease of 133 patients undergoing reduction mammoplasty using a medial supe- 0.7 ± 0.5 cm/100 g. Sixteen patients had preoperative and postoperative rior pedicle technique with suture suspension of the medial superior measurements of N-M. The average change in N-M per patient ranged pedicle to the pectoralis fascia (Fig. 1) were identified. The median from an increase of 0.4 cm/100 g to a decrease of 0.6 cm/100 g. Patient weight of tissue excised per breast was between 400 and 500 g, with averages were averaged to obtain an overall decrease of 0.1 ± 0.3 cm/ a range of 200 to 1000 g per breast. Skin closure was done using vertical 100 g. Fourteen patients had preoperative and postoperative measure- approach in 40% of patients, inverted-T in 40%, and short-L in 20%. ments of BD. The average change in N-M per patient ranged from an

FIGURE 3. Patient undergoing breast reduction using medial superior pedicle suspension to pectoralis fascia.

FIGURE 4. Change in bilateral N-S measurement in centimeters.

increase of 0.6 cm/100 g to a decrease of 0.5 cm/100 g. Patient averages this limitation would be to take serial measurements multiple times were averaged to obtain an overall decrease of 0.0 ± 0.4 cm/100 g. Of throughout the healing process to establish a more finalized number the 54 nipples (2 per patient) reported, over 94% had the same or many months after surgery. This may take into account any drop in nip- increased sensation. ple location further along in the healing process as well.

Future Studies DISCUSSION One idea for future studies would be attempting to determine if actually knowing the measurements beforehand has a positive effect Measurements on patient satisfaction and if predication is correct. Patients undergoing this technique could be split into 2 groups: one given the average mea- In this study, quantitative measurements for all 4 measures of in- sures beforehand and the other just receiving traditional volume lingo. terest were established. N-S and N-I had measureable changes, whereas Patients would take a survey asking about overall satisfaction and if N-M and BD had very little average change. N-S was the greatest measurements were accurate. Another idea would be to compare other change, which seems reasonable when considering the volume re- reduction mammoplasty techniques to the SPEAR technique to see if moved. An average measure that can be communicated to patients other techniques have different average changes or any difference in along with standard information about volume reduction was estab- N-M and BD. lished. During a preoperative consultation, a physician could still men- tion volume but also discuss what has been found for N-S change on average. The surgeon can then demonstrate maximums and minimums CONCLUSIONS on new nipple placement. This is beneficial for patients in establishing a A surgeon’s expression of breast measurements in terms of vol- better understanding of outcomes and expectations and in decreasing ume can be difficult for a patient to understand and visualize. In this the likelihood of unhappy results. N-I also has a measureable change study, we established the impact volume has on length of typical breast that can be demonstrated to patients before operation to better prepare measurements to increase patients’ understanding of expected outcomes. patients for the surgery. N-M had no average change, which is also worth mentioning to patients. Although nipples in large breasts are generally lower than desired, the nipple is still located in the middle of the ACKNOWLEDGMENTS breast. This can reassure patients that nipple placement will remain in The artist’s drawing in Figure 1 was by Matthew K. Collawn. Per- approximately the same place. The BD also had no average change, mission was obtained to use his artwork in this paper. which is an interesting find. Initially, this study hypothesized that BD would decrease, but that was not the case. A possible explanation is that the pedicle flap and the desire to retain breasts proportional to body type REFERENCES resulted in the small changes in BD. This can easily be explained to 1. Kalliainen LK; ASPS Health Policy Committee. ASPS clinical practice guideline patients, especially those that fear being too small or out of proportion summary on reduction mammaplasty. Plast Reconstr Surg. 2012;130:785–789. to their body type. These measurements can help surgeons communi- 2. Berberoğlu Ö, Temel M, Türkmen A. Effects of reduction mammaplasty operations on the spinal column: clinical and radiological response. Aesthetic Plast cate better with patients and also allow patients to have a better idea of Surg. 2015;39:514–522. surgical outcomes. 3. Karabekmez FE, Gokkaya A, Isik C, et al. Does reduction mammaplasty revert skeletal disturbances in the vertebral column of patients with macromastia? A pre- – Limitations liminary study. Aesthetic Plast Surg. 2014;38:104 112. 4. Atiyeh BS, Rubeiz MT, Hayek SN. Refinements of vertical scar mammaplasty: There were a few limitations to this study. Although 133 patients circumvertical skin excision design with limited inferior pole subdermal underwent this surgery using the SPEAR technique, a much smaller undermining and liposculpture of the inframammary crease. Aesthetic Plast Surg. number actually had preoperative and postoperative measurements. In 2005;29:519–531. the future, an increased sample size may help provide more accurate 5. Spear SL, Davison SP, Ducic I. Superomedial pedicle reduction with short scar. – measurements. Another limitation is the fact that only 4-week measure- Semin Plast Surg. 2004;18:203 210. – ments were recorded. This does not account for individual responses to 6. Hall-Findlay EJ. Vertical breast reduction. Semin Plast Surg. 2004;18:211 224. 7. Antony AK, Yegiyants SS, Danielson KK, et al. A matched cohort study of surgery and length of postoperative edema in an individual. Some pa- superomedial pedicle vertical scar breast reduction (100 breasts) and traditional in- tients may have had falsely low numbers because of residual swelling, ferior pedicle Wise-pattern reduction (100 breasts): an outcomes study over which may have already resolved in other patients. A way to mitigate 3 years. Plast Reconstr Surg. 2013;132:1068–1076.

8. Akyurek M. Contouring the inferior pole of the breast in vertical mammaplasty: 11. Reekie TA, Wharton E, Sugden P.Intra-operative breast volume estimation during suction-assisted lipectomy versus direct defatting. Plast Reconstr Surg.2011; breast reduction: validation of a commonly used technique. J Plast Reconstr 127:1314–1322. Aesthet Surg. 2013;66:1008–1009. 9. Chae MP, Hunter-Smith DJ, Spychal RT, et al. 3D volumetric analysis for plan- 12. Xue AS, Wolfswinkel EM, Weathers WM, et al. Breast reduction in adolescents: ning breast reconstructive surgery. Breast Cancer Res Treat. 2014;146:457–460. indication, timing, and a review of the literature. J Pediatr Adolesc Gynecol.2013; 10. Quan M, Fadl A, Small K, et al. Defining pseudoptosis (bottoming out) 3 years 26:228–233. after short-scar medial pedicle breast reduction. Aesthetic Plast Surg.2011;35: 13. Grewal NS, Fisher J. Why do patients seek revisionary breast surgery? Aesthet 357–364. Surg J. 2013;33:237–244.

The Sternum-Nipple Distance is Double the Nipple-Inframammary Fold Distance in Macromastia Thomas N. Steele, BS,* Julian J. Pribaz, MD,† and Frank H. Lau, MD*

eduction mammoplasty is one of the most commonly performed Background: Reduction mammaplasty is one of the most commonly performed R plastic surgery procedures, with 114,470 reductions performed in plastic surgery operations. For a majority of techniques, the most common long- 2015.1 The benefits of reduction mammaplasty have been extensively term complication is pseudoptosis. It has previously been proposed that upper documented and include pain relief, improved posture, improved qual- breast suspensory ligaments (SL) are weaker than lower breast SL. We tested this ity of life, and an increase in quality-adjusted life years.2–6 hypothesis through anthropometry of the proxies for upper and lower SL strength: Technical refinements have improved the safety of breast reduc- the sternal notch-nipple (SN-N) distance and the nipple-inframammary fold tions, and innovation now focuses on long-term aesthetic outcomes. (N-IMF) distance, respectively. – The most common long-term complication is pseudoptosis, where the Methods: An institutional review board approved retrospective review of pa- breast parenchyma migrates to the lower pole independent of nipple po- tients undergoing reduction mammoplasty in an academic faculty practice be- sition (Fig. 1). This contrasts with ptosis, which involves rotation of the tween 2008 and 2015 was conducted. Patient demographics included age, race, breast and nipple as a unit towards the inframammary fold. and body mass index (BMI); patient comorbidities included smoking status, dia- While the pathogenesis of pseudoptosis remains controversial, betes, and hypertension. Breast anthropometric measurements included SN-N we hypothesize that disruption of the lower suspensory ligaments and N-IMF. Sternal notch-nipple was used as the primary metric of the upper (SL) of the breast is a critical factor. Reduction techniques that dis- SL strength, whereas N-IMF was used as the primary metric of the lower SL rupt the lower SL have higher reported pseudoptosis rates. For exam- strength. Intraoperative details included reduction technique and resection mass. ple, the rate in Wise pattern with an inferior pedicle is 50%.7,8 In Postoperative complications were recorded, including nipple areola complex ne- techniques that preserve the inferior SL, such as the Boston modifica- crosis and hematoma. Linear regression analysis was performed with the primary tion of the Robertson technique (BMRT), pseudoptosis rates are re- endpoint of the relationship between SN-N and N-IMF distance in macromastia. duced (0%–13%).9,10 Results: Data from 208 patients, totaling 400 individual breast measurements, Together, these data suggest that the upper SL are weaker than were collected. The mean SN-N length was 35.1 cm, mean N-IMF length was the lower SL. This differential would explain the rotation of the breast 16.0 cm, and mean resection weight was 1094 g. Linear regression found that in natural or “true” ptosis and the differing rates of pseudoptosis follow- N-IMF distance could be predicted as 45% of the SN-N distance (N-IMF = 0.454 * ing reduction mammaplasty. We tested this hypothesis through anthro- SN-N). This was a strong relationship, demonstrated by univariate analysis of pometry of the proxies for upper and lower SL strength: the sternal SN-N and N-IMF (R, 0.624; P < 0.001). A Wise pattern was used in 89.9% of notch-nipple (SN-N) distance and the nipple-inframammary fold (N- cases; an inferior pedicle was used in 83.7% of cases. Nipple areola complex ne- IMF) distance, respectively. We thus performed a retrospective cohort crosis occurred in 15 breasts (3.75%). Sternal notch-nipple (R, 0.127; P =0.011) review of all breast reductions performed in an academic faculty prac- and N-IMF (R, 0.119; P = 0.017) were both predictive of nipple areola complex tice over a 7-year period. necrosis (Table 4). Conclusions: In our series, the N-IMF distance increased 0.45 cm for every 1 cm increase in the SN-N distance. This relationship strengthens our primary hypoth- PATIENTS AND METHODS esis that the lower pole ligaments stretch at a significantly slower rate than the up- An institutional review board–approved retrospective review of per pole ligaments. Taking this into consideration, we suggest that surgeons consecutive reduction mammaplasties performed in an academic fac- seeking to minimize pseudoptosis rates should favor techniques that minimally ulty practice between May 2008 and August 2015 was conducted. A disrupt the lower SL. search of all procedures that were coded as reduction mammaplasties Key Words: macromastia, mammaplasty, reduction mammaplasty, (Current Procedural Terminology code 19318) was performed, and data suspensory ligaments, Cooper ligaments, breast ligaments, breast anthropometry, were collected from electronic medical records. sternal notch-nipple, SN-N, nipple-inframammary fold, N-IMF, breast ptosis, Patient demographics included age, race, and body mass index pseudoptosis, bottoming out (BMI). Patient comorbidities included smoking status, diabetes, and hypertension. Breast anthropometric measurements included SN-N and (Ann Plast Surg 2017;78: S347–S350) N-IMF. Intraoperative details included reduction technique and resection mass. Postoperative complications of nipple areola complex (NAC) necrosis and hematoma were recorded. The data were manually reviewed to remove duplicated patients Received October 14, 2016, and accepted for publication, after revision December and incomplete or incorrect entries. IBM SPSS Statistics for Mac 23.0 6, 2016. was used for statistical analysis. Linear regression was performed to ex- From the *Division of Plastic and Reconstructive Surgery, School of Medicine, amine the relationship between SN-N and N-IMF. Bivariate correlation Louisiana State University Health Sciences Center, New Orleans, LA; and analysis was performed to analyze any potential correlation of vari- †Division of Plastic Surgery, Brigham and Women's Hospital, Harvard Medical School, Boston, MA. ables with anthropometric measurements including age, BMI, type Conflicts of interest and source of funding: none declared. of resection, resection mass, NAC necrosis, and medical comorbidities Reprints: Frank H. Lau, MD, Division of Plastic and Reconstructive Surgery, Department (smoking status, hypertension, and diabetes). Univariate generalized of Surgery, Louisiana State University Health Sciences Center, 1542 Tulane Ave, linear model analysis was conducted to evaluate for correlations with Room 734, New Orleans, LA 70112. E-mail: [email protected]. Copyright © 2017 Wolters Kluwer Health, Inc. All rights reserved. increasing N-IMF and to analyze potential correlations between NAC ISSN: 0148-7043/17/7806–S347 necrosis and selected variables. A P value of less than 0.05 was consid- DOI: 10.1097/SAP.0000000000001031 ered statistically significant. Breasts were analyzed independently to

Annals of Plastic Surgery • Volume 78, Supplement 5, June 2017 www.annalsplasticsurgery.com S347 Steele et al Annals of Plastic Surgery • Volume 78, Supplement 5, June 2017

FIGURE 1. Visualization of true ptosis (first, second, and third degree) compared to pseudoptosis.

FIGURE 2. Patient exclusion criteria. account for asymmetry and because outcomes in bilateral breasts were mellitus, 48 patients (23.1%) had a diagnosis of hypertension, and not necessarily the same. 27 patients (13.0%) identified as everyday smokers at the time of preoperative assessment. Mean SN-N length was 35.1 cm, mean N-IMF length was 16.0 RESULTS cm, and mean resection mass was 1094 g (Table 1A). Sternal notch- During the study period, a total of 422 patients undergoing re- nipple distances and N-IMF distances were then linearly regressed, duction mammaplasty were identified. The procedures were performed and a predictive formula was derived: the N-IMF distance increases at 5 institutions by 7 academic plastic surgeons. After manually filtering 0.45 cm for every 1 cm increase in SN-N distance (N-IMF = 0.454 * the data to exclude 111 duplicate procedures and 103 patients with in- SN-N, Fig. 3). complete, incorrect, or absent measurements, a total of 208 patients In all patients, only one reduction technique was performed per (400 individual breasts) who received reduction mammaplasty ulti- patient. One hundred seventy-four reductions used an inferior pedicle, mately met the selection criteria (Fig. 2). and 34 reductions used a superomedial pedicle. Of the 174 inferior ped- The mean age at time of resection was 38 years; mean BMI was icle reductions, 161 (92.5%) of these implemented a Wise pattern inci- 34.1 (Table 1A). A total of 163 patients (78.4%) self-identified as sion and 13 (7.5%) used a Boston modification of Robertson technique African-American, 41 patients (19.7%) as Caucasian, and 2 patients pattern. In an independent t test, the mean SN-N length was 35.7 cm in each (0.96%) as Hispanic and Middle Eastern (Table 1B). In regards breasts undergoing inferior pedicle reductions and 32.7 cm in superomedial to medical comorbidities, 20 patients (9.6%) had a diagnosis of diabetes pedicle reductions (P < 0.001, Table 2). Of the 33 superomedial pedicle

TABLE 1A. Descriptive Statistics (N = 400 individual breasts) TABLE 1B. Self-identified Race of Patients Entered Into Study (N = 208 patients) Variable Mean Range SD Race N % SN-N distance, cm 35.14 21–56 5.80 N-IMF distance, cm 16.00 7–35 3.76 African-American 163 78.37 Mass of breast resection, g 1094.39 119–5078 707.90 Caucasian 41 19.31 BMI 34.10 19.62–67.49 7.24 Hispanic 2 0.96 Age, y 38.2 15–75 13.64 Middle-Eastern 2 0.96

TABLE 3. Univariate Analysis to Assess Correlation Between N-IMF Distance and Selected Variables

Variable F P Mass of resection 132.976 <0.001 SN-N 14.412 <0.001 BMI NS NS Age NS NS Diabetes NS NS Smoker NS NS FIGURE 3. Linear regression of SN-N distance and N-IMF NS, not significant. distance shows a positive correlation. quantitative relationship is very similar to that reported by Jackson reductions, 25 (75.8%) used a Wise pattern incision and 8 (24.2%) used et al20 in a smaller cohort of patients. We hypothesize that this relation- a vertical pattern incision. ship is owing to weaker SL in the upper breast versus the lower breast. In a univariate analysis, the N-IMF distance demonstrated no Chalekson et al9 argued that the decreased rate of pseudoptosis in the significant correlation with age, BMI, diabetes mellitus, hypertension, modified Robertson technique is owing to preservation of N-IMF pa- or smoking status. However, as one might assume, there is a positive renchyma and the lower SL. Furthermore, an anatomic basis for this correlation between N-IMF and SN-N distance and also between differential may exist: Würinger21 described a horizontal ligamentous N-IMF distance and resection mass (Table 3). septum of dense connective tissue that originates at the level of the Nipple areola complex necrosis occurred in 15 individual breasts fifth rib and divides the breast at the level of the nipple into cranial (3.75%). Two patients experienced bilateral NAC necrosis. Postoper- and caudal segments. This division may signify a transition point in ative hematoma occurred in 5 total breasts. In a univariate analysis SL strength. (Table 4), a significant association was observed between NAC necro- On this basis, we propose that techniques such as the BMRT sis and resection mass (P < 0.001), N-IMF distance (P = 0.017), and should be preferred by surgeons seeking to avoid pseudoptosis.10 The SN-N distance (P = 0.011). There was no significant difference in BMRT combines a broad-based inferior pedicle encompassing the en- N-IMF distance, resection mass, and rates of NAC necrosis between tire inframammary fold with a superior apron flap to conceal the trans- the 2 pedicles. However, with only 33 reductions based on a super- verse scar along the inferior portion of the breast. Resection occurs in a omedial pedicle, this study is not sufficiently powered to make a bell-shaped pattern superior, medial, and lateral to the NAC. One crite- definitive conclusion. rion for selection of this technique is NAC transposition of higher than 5 cm, indicating that this technique is appropriate for moderate-large reductions. Movassaghi et al10 reports that the BMRT demonstrates de- DISCUSSION creased postoperative hematoma formation when compared to the A primary objective of reduction mammaplasty is a long-lasting, Wise pattern reduction (0% vs 10%, P = 0.016) and also subjectively aesthetically pleasing breast shape. Unfortunately, breast shape in the reports minimal to no postreduction pseudoptosis, although further immediate postoperative period still varies unpredictably from the final scrutiny is necessary. Both Movassaghi et al10 and Chalekson9 hypoth- – breast shape.11 13 Pseudoptosis, or “bottoming out,” remains a too- esize that the undisturbed dermis and breast tissue in the N-IMF region common long-term complication of all reduction mammoplasty tech- are the primary factor for prevention of pseudoptosis. Other advantages niques.12,14 In the most commonly performed reduction technique, of the BMRT include decreased operative time and elimination of the the Wise pattern with an inferior pedicle, pseudoptosis occurs in 50% “triple point,” the most common site of postoperative dehiscence.10 of patients. In the technique with the lowest rates, the vertical pattern with Our study has several limitations. First, as an anthropometric pi- a superomedial pedicle, pseudoptosis still occurs in 20% of patients.7 lot study, our data did not include pseudoptosis and other complica- A critical barrier to further reducing the pseudoptosis rate is the tions. The rates of these complications should be studied as a absence of a known pathogenic mechanism. This is reflected in the di- multicentered, randomized, controlled prospective trial. Second, our versity of methods that have been proposed to combat pseudoptosis, in- study was unable to biomechanically and histologically examine the cluding placing the NAC below the traditional location of Pitanguy point,15 the use of acellular dermal matrix as an internal brassiere,16 muscle flaps,17 and fascial suspension.18,19 None have proven to be ef- TABLE 4. Univariate Analysis to Examine Variables Demonstrating ficacious in preventing pseudoptosis. Positive Correlation With NAC Necrosis In this study, we show a differential rate of stretch in macromastia: for every 1 cm increase in upper SL length (SN-N distance), Variable F P the lower SL length (N-IMF distance) increases by 0.45 cm. This SN-N distance, cm 5.592 0.019 N-IMF distance, cm 4.438 0.036 Mass of breast resection, g 12.547 <0.001 TABLE 2. Independent t Test Comparing Pedicle Selection. Inferior Pedicle Demonstrated a Significantly Larger Mean SN-N Distance BMI 3.909 0.049 Versus Superomedial Pedicle (P <0.001) Age NS NS Diabetes NS NS Technique N Mean SN-N, cm SD Smoker NS NS Hypertension NS NS Inferior pedicle 333 35.65 5.804 Superomedial pedicle 67 32.65 5.138 NS, not significant.

upper and lower SL of the breast. This should be performed as either a 9. Chalekson CP, Neumeister MW, Zook EG, et al. Outcome analysis of reduction cadaver study or a study of discarded specimens from prophylactic mammaplasty using the modified Robertson technique. Plast Reconstr Surg. – – mastectomies. Lastly, our patient population differed demographi- 2002;110:71 79. discussion 80 1. 10. Movassaghi K, Ting V, Matros E, et al. Eliminating the vertical scar in breast cally from the average United States population, thus, limiting the reduction-Boston modification of the Robertson technique. Aesthet Surg J. generalizability of our results. Our patients were more obese (mean 2006;26:687–696. BMI, 34.1), predominantly African American, and had more severe 11. Choi M, Small K, Tepper O, et al. Defining the kinetics of breast pseudoptosis af- macromastia (mean resection mass, 1094 g per breast) than most stud- ter reduction mammaplasty. Ann Plast Surg. 2009;62:518–522. – ies of macromastia.22 26 12. Quan M, Fadl A, Small K, et al. Defining pseudoptosis (bottoming out) 3 years after short-scar medial pedicle breast reduction. Aesthetic Plast Surg.2011;35: 357–364. CONCLUSIONS 13. Small KH, Tepper OM, Unger JG, et al. Re-defining pseudoptosis from a 3D perspective after short scar-medial pedicle reduction mammaplasty. J Plast Reconstr In conclusion, this study demonstrates that N-IMF distance in- – creases 0.45 cm for every 1 cm increase in SN-N distance. This rela- Aesthet Surg. 2010;63:346 353. 14. Nahai FR, Nahai F. MOC-PSSM CME article: breast reduction. Plast Reconstr tionship strengthens our primary hypothesis that the lower pole SL Surg. 2008;121:1–13. stretch at a significantly slower rate than the upper pole SL. Taking 15. Meshulam-Derazon S, Barnea Y,Zaretski A, et al. Large-volume breast reduction: this into consideration, we suggest that surgeons seeking to minimize long-term results. Scand J Plast Reconstr Surg Hand Surg. 2009;43:65–70. pseudoptosis rates should favor techniques that minimally disrupt the 16. Brown RH, Izaddoost S, Bullocks JM. Preventing the ‘bottoming out’ and ‘star- lower SL. gazing’ phenomena in inferior pedicle breast reduction with an acellular dermal matrix internal brassiere. Aesthetic Plast Surg. 2010;34:760–767. 17. Caldeira AM, Lucas A. Pectoralis major muscle flap: a new support approach to REFERENCES mammaplasty, personal technique. Aesthetic Plast Surg. 2000;24:58–70. 1. American Society of Plastic Surgeons. 2015 Plastic Surgery Statistics Report. 18. Lockwood T. Reduction mammaplasty and mastopexy with superficial fascial American Society for Aesthetic Plastic Surgery. 2015. system suspension. Plast Reconstr Surg. 1999;103:1411–1420. 2. Blomqvist L, Eriksson A, Brandberg Y. Reduction mammaplasty provides long- 19. Widgerow AD. Breast reduction with inferior pedicle fascial suspension. Aesthetic term improvement in health status and quality of life. Plast Reconstr Surg. Plast Surg. 2005;29:532–537. 2000;106:991–997. 20. Jackson IT, Bayramicli M, Gupta M, et al. Importance of the pedicle length mea- 3. Chadbourne EB, Zhang S, Gordon MJ, et al. Clinical outcomes in reduction surement in reduction mammaplasty. Plast Reconstr Surg. 1999;104:398–400. mammaplasty: a systematic review and meta-analysis of published studies. Mayo 21. Würinger E. Refinement of the central pedicle breast reduction by application of Clin Proc. 2001;76:503–510. the ligamentous suspension. Plast Reconstr Surg. 1999;103:1400–1410. 4. Chao JD, Memmel HC, Redding JF,et al. Reduction mammaplasty is a functional 22. Antony AK, Yegiyants SS, Danielson KK, et al. A matched cohort study of operation, improving quality of life in symptomatic women: a prospective, superomedial pedicle vertical scar breast reduction (100 breasts) and traditional in- single-center breast reduction outcome study. Plast Reconstr Surg. 2002;110: ferior pedicle wise-pattern reduction (100 breasts): an outcomes study over three 1644–1652. discussion 1653–4. years. PlastReconstrSurg. 2013;132:1068–1076. 5. Spector JA, Karp NS. Reduction mammaplasty: a significant improvement at any 23. Appel JZ 3rd, Wendel JJ, Zellner EG, et al. Association between preoperative mea- size. Plast Reconstr Surg. 2007;120:845–850. surements and resection weight in patients undergoing reduction mammaplasty. 6. Thoma A, Sprague S, Veltri K, et al. A prospective study of patients undergoing Ann Plast Surg. 2010;64:512–515. breast reduction surgery: health-related quality of life and clinical outcomes. Plast 24. Cunningham BL, Gear AJ, Kerrigan CL, et al. Analysis of breast reduction com- Reconstr Surg. 2007;120:13–26. plications derived from the BRAVO study. Plast Reconstr Surg. 2005;115: 7. Bouwer LR, van der Biezen JJ, Spronk CA, et al. Vertical scar versus the 1597–1604. inverted-T scar reduction mammaplasty: a 10-year follow-up. J Plast Reconstr 25. Manahan MA, Buretta KJ, Chang D, et al. An outcomes analysis of 2142 breast Aesthet Surg. 2012;65:1298–1304. reduction procedures. Ann Plast Surg. 2015;74:289–292. 8. Rohrich RJ, Gosman AA, Brown SA, et al. Current preferences for breast reduc- 26. O'Grady KF, Thoma A, Dal Cin A. A comparison of complication rates in large tion techniques: a survey of board-certified plastic surgeons 2002. Plast Reconstr and small inferior pedicle reduction mammaplasty. Plast Reconstr Surg.2005; Surg. 2004;114:1724–1733. 115:736–742.