Susan K. Parsons1,2,3,4, Michael J. Kelly1,5, Joshua T. Cohen2,3,6, Sharon M. Castellino7, Tara O. Henderson8, Kara M. Kelly9, Frank G. Keller7, Tobi J. Henzer3, Anita J. Kumar2,3,4, Peter Johnson10, Ralph M. Meyer11, John Radford12, John Raemaekers13, David C. Hodgson14, Andrew M. Evens15
1 Department of Pediatrics, Tufts University School of Medicine, Boston, MA, USA, 2 Department of Medicine, Tufts University School of Medicine, Boston, MA, USA, 3 Institute for Clinical Research and Health Policy Studies, Tufts MC, Boston, MA, 4 Division of Hematology/Oncology, Tufts MC, Boston, MA, USA, 5 Division of Pediatric Hematology/Oncology, The Floating Hospital for Children at Tufts Medical Center (MC), Boston, MA, USA, 6 Center for the Evaluation of Value and Risk in Health, Tufts MC, Boston, MA, USA, 7 Aflac Cancer and Blood Disorders Center, Children's Healthcare of Atlanta, Emory University School of Medicine, Atlanta, GA, USA, 8 Department of Pediatrics, Section of Hematology, Oncology and Stem Cell Transplantation, University of Chicago, Chicago, IL, 9 Department of Pediatrics, Roswell Park Cancer Institute, University of Buffalo School of Medicine and Biomedical Sciences, Buffalo, NY, 10 Cancer Research UK Centre, Southampton, United Kingdom, 11 Department of Oncology, Juravinski Hospital and Cancer Centre and McMaster University, Hamilton Ontario, Canada, 12 University of Manchester and the Christie NHS Foundation Trust, Manchester Academic Health Science Centre, Manchester, UK, 13 Department of Hematology, Radboud University Medical Center, Nijmegen, the Netherlands, 14 Radiation Medicine Programme, University of Toronto, Princess Margaret Cancer Centre, Toronto, ON, Canada, 15 Division of Blood Disorders, Rutgers Cancer Institute of New Jersey, New Brunswick, New Jersey, USA
Introduction: Helping clinicians and patients (pts) assess alternative HL treatment options is challenging, especially with consideration of the incidence and impact of late effects (LEs) on outcomes. We constructed a novel simulation model analyzing and integrating multiple data sets to project long-term outcomes with contemporary early-stage HL (ESHL) therapy, namely combined modality therapy (CMT) vs. chemotherapy alone (CA) via PET response-adaption.
Methods: The model consists of a series of health states: 1) at risk for relapse; 2) relapse; 3) cured without relapse; 4) cured with relapse; 5) cured with late effects; and 6) dead. During each model cycle (a period of 1 year in the model), simulated subjects can transition from their current health state to other health states (Figure). Whether a subject transitions to another health state depends on the transition pathway probability connecting the current and destination states. The 6 health states in the model have utility weights ranging from zero (dead) to 0.80 (cured without relapse). The model incorporated 3-year progression-free survival (PFS) estimates (Radford et al NEJM 2015); probability of cure with/without relapse; 35-year probability of LEs; and frequency of severe LEs. We generated estimates for quality-adjusted life years (QALYs) and unadjusted survival (life years = LY) and used model projections to compare outcomes for CMT vs. CA for two index pts. Pt #1: a 25-year-old male with favorable ESHL (stage IA); pt #2: a 25-year-old female with unfavorable ESHL (stage IIB). Multiple sensitivity analyses assessed the impact of alternative assumptions for LE probabilities.
Results: For pt #1, CMT was superior to CA (CMT incremental gain = 0.11 QALYs, 0.21 LYs). For pt #2, CA was superior to CMT (CA incremental gain = 0.37 QALYs, 0.92 LYs). As the proportion of pts with LEs with severe outcomes was reduced from its base case value of 20% to 5% in sensitivity analysis, the relative advantage of CMT for pt #1 increased to 0.15 QALYS and 0.43 unadjusted, undiscounted LYs. Increasing the severity proportion for pt #2's LEs from 20% to 80% showed that these alternative assumptions increased the CA advantage vs.CMT to as much as 1.1 QALYs (13 months in perfect health) and 6.5 unadjusted, undiscounted LYs.
Conclusions: Collectively, this detailed and dynamic simulation model quantified the impact that alternative treatment options have on long-term survival for individual, varying pts with ESHL.