Paediatric Cancer Stage in Population-Based Cancer Registries: the Toronto Consensus Principles and Guidelines
Total Page:16
File Type:pdf, Size:1020Kb
Review Paediatric cancer stage in population-based cancer registries: the Toronto consensus principles and guidelines Sumit Gupta, Joanne F Aitken, Ute Bartels, James Brierley, Mae Dolendo, Paola Friedrich, Soad Fuentes-Alabi, Claudia P Garrido, Gemma Gatta, Mary Gospodarowicz, Thomas Gross, Scott C Howard, Elizabeth Molyneux, Florencia Moreno, Jason D Pole, Kathy Pritchard-Jones, Oscar Ramirez, Lynn A G Ries, Carlos Rodriguez-Galindo, Hee Young Shin, Eva Steliarova-Foucher, Lillian Sung, Eddy Supriyadi, Rajaraman Swaminathan, Julie Torode, Tushar Vora, Tezer Kutluk, A Lindsay Frazier Population-based cancer registries generate estimates of incidence and survival that are essential for cancer Lancet Oncol 2016; 17: e163–72 surveillance, research, and control strategies. Although data on cancer stage allow meaningful assessments of Division of Haematology/ changes in cancer incidence and outcomes, stage is not recorded by most population-based cancer registries. Oncology, Hospital for Sick The main method of staging adult cancers is the TNM classifi cation. The criteria for staging paediatric cancers, Children, Toronto, ON, Canada (S Gupta PhD, U Bartels MD, however, vary by diagnosis, have evolved over time, and sometimes vary by cooperative trial group. Consistency in the L Sung PhD); Department of collection of staging data has therefore been challenging for population-based cancer registries. We assembled key Paediatrics, Faculty of experts and stakeholders (oncologists, cancer registrars, epidemiologists) and used a modifi ed Delphi approach to Medicine, University of establish principles for paediatric cancer stage collection. In this Review, we make recommendations on which Toronto, Toronto, ON, Canada (S Gupta, U Bartels, L Sung); staging systems should be adopted by population-based cancer registries for the major childhood cancers, including Cancer Council Queensland, adaptations for low-income countries. Wide adoption of these guidelines in registries will ease international Fortitude Valley, Brisbane, comparative incidence and outcome studies. QLD, Australia (J F Aitken PhD); Department of Radiation Oncology, Princess Margaret Introduction cancer stage by population-based cancer registries, and Hospital, Toronto, ON, Canada Population-based cancer registries are a unique resource to recommend which staging system(s) should be used (J Brierley MBBS, for both cancer researchers and policy makers.1 Registry by cancer registries for 18 major childhood malignancies. M Gospodarowicz MD); Southern Philippines Medical data have been used for disease surveillance and to derive Staging systems should be able to be applied by registry Center, Bajada, Davao City, population-based estimates of incidence, prevalence, and staff using available records and should be suffi ciently Philippines (M Dolendo MD); outcome in both adults and children.2–5 The results of these detailed for analysis and interpretation of population Dana-Farber/Boston Children’s investigations have been used to plan and evaluate national cancer data, while respecting the diff erent capacities Cancer and Blood Disorders 1,3,6,7 Center, Boston, MA, USA cancer control strategies. Cancer stage is a core concept and resources of diff erent registries. The resultant (P Friedrich MD, A L Frazier MD, in oncology, providing a “common nomenclature on which Toronto Paediatric Cancer Stage guidelines have been C Rodriguez-Galindo MD); to base cancer management, research and information endorsed by the Union for International Cancer Control Hospital Nacional de Ninos exchange”.8,9 Accurate stage data are crucial when TNM Prognostic Factors Project. Our recommendations Benjamin Bloom, San Salvador, El Salvador comparing cancer outcomes between groups or over time. are not intended to replace staging systems in (S Fuentes-Alabi MD); Unidad Despite this, many population-based cancer registries clinical use. Nacional de Oncologia either do not record stage data at all or, in paediatric Pediatrica, Guatemala City, patients, record stage according to the adult TNM staging Methods Guatemala (C P Garrido MD); Evaluative Epidemiology Unit, classifi cation. The TNM classifi cation was developed by We assembled a panel of international experts and Fondazione IRCSS Instituto the Union for International Cancer Control, and is used to advocacy stakeholders, and undertook a modifi ed Delphi Nazionale dei Tumori, Milan, classify and code stage in many adult malignant diseases, approach to build consensus.12,13 Invited experts Italy (G Gatta MD); Center for but is not applicable to most paediatric cancers.8 represented diverse content skills (eg, clinicians and Global Health, National Cancer Institute, Bethesda, MD, USA Because of their heterogeneity and rarity, childhood epidemiologists), geography, and resource settings to (T Gross PhD); University of cancers already represent a particular data management ensure the most widely applicable and feasible Memphis, Memphis, TN, USA challenge for registries.10 Most childhood malignant recommendations.13 Representatives from selected (S C Howard MD); College of diseases are staged according to disease-specifi c staging cancer registries or registry associations were invited, Medicine, Blantyre, Malawi (E Molyneux PRCPCH); systems that often diff er between countries or clinical and invitees could nominate additional panellists. Argentinian Oncopediatric trial organisations.11 The usefulness of population-based Before starting the Delphi rounds, consensus Registry (ROHA), National registries to childhood cancer research and policy is workshop leaders (Sumit Gupta, A Lindsay Frazier) Cancer Institute, Buenos Aires, therefore limited by both the general paucity of cancer generated a list of candidate principles informed by the Argentina (F Moreno MD); Pediatric Oncology Group of stage data, the inadequacy of adult staging systems in principles that have been endorsed for the collection of Ontario, Toronto, ON, Canada showing the extent of disease in children, and the use cancer stage in adult malignancies by the Union for (J D Pole PhD); Institute of Child of many paediatric staging classifi cations for the International Cancer Control TNM Prognostic Factors Health, University College same malignant disease. Finally, cancer registries in Project, and through interviews with two experts in London, London, UK (K Pritchard-Jones PhD); low-income and middle-income countries face additional cancer registration (Oscar Ramirez, Lynn A G Ries), one Registro Poblacional de Cancer challenges in view of the unavailability or unaff ordability from a high-income country and one from a middle- de Cali, Universidad del Valle, of advanced imaging. income country.8 A prior search of Ovid MEDLINE Oficina, Cali, Valle, Colombia Our primary objectives were to identify the key revealed only studies pertaining to specifi c malignancy (O Ramirez MD); SEER Program, National Cancer Institute, principles that should guide the collection of childhood cohorts, and we identifi ed no articles that provided www.thelancet.com/oncology Vol 17 April 2016 e163 Review Bethesda, MD, USA principles for choosing one staging system over another. National Cancer Institute, or Surveillance, Epidemiology, (L A G Ries MS); Department of We then conducted two Delphi rounds by email to and End Results Program (appendix). 26 (93%) experts Pediatrics, Cancer Research gauge consensus on guiding principles relevant to the who accepted the invitation to participate represented Institute, Seoul National University College of Medicine, collection of paediatric cancer stage in population-based 17 countries across six continents. Ten (36%) panellists Seoul, South Korea registries. were from a low-income or middle-income country. (H Y Shin PhD); Section of In round one, we asked panellists to score each Of the 26 experts, 25 (96%) returned responses to the Cancer Surveillance, principle on a 5-point Likert scale.13 round one survey. All 25 of these respondents returned International Agency for We asked panellists Research on Cancer, to provide comments on each principle, and to suggest responses to the round two survey. All 26 participating Lyon, France additional principles for inclusion. In accordance with individuals, plus the two workshop leaders, attended the (E Steliarova-Foucher PhD); published guidelines, consensus was defi ned as 75% or meeting in Toronto, Canada. Pediatric Hematology Oncology Division, more of respondents either agreeing or strongly agreeing In round one, 18 principles were proposed; 13 achieved 13 Department of Pediatrics, with a principle (median score ≤2). Any principle with consensus and one was eliminated as consensus was not Dr Sardjito Hospital, Gadjah which 75% or more of respondents disagreed or strongly achieved (table 1). The four remaining original principles Mada University, Yogyakarta, disagreed with was eliminated. were modifi ed and one new principle was added Indonesia (E Supriyadi PhD); Epidemiology, Biostatistics Next, the consensus workshop leaders reviewed and according to round one comments, and included in the and Cancer Registry, Cancer revised any principles that had not achieved consensus, round two surveys. Of these fi ve second-round Institute, Chennai, India but had not been eliminated. Based on round one principles, three achieved consensus and two did not (R Swaminathan PhD); Union feedback, several entirely new principles were developed. (table 2). During the workshop, fi ve principles that had for International Cancer Control, Geneva, Switzerland In round