Analysis of Long-Term Survivorship (LTS) Rates for Stage 2B/2C Melanoma Using Published Real-World Data (RWD) and Data From Randomized Controlled Trials (RCTS)
Author(s)
Alagoz O1, Winge-Main A2, Srinivasan S3, Dyer M4, Wolf R5, Thybo SH5, May JR4, Moshyk A3, Kurt M3
1University of Wisconsin-Madison, Madison, WI, USA, 2Oslo University Hospital, Oslo, Not applicable, Norway, 3Bristol Myers Squibb, Lawrenceville, NJ, USA, 4Bristol Myers Squibb, Uxbridge, UK, 5Bristol Myers Squibb, Sorgenfri, Capital Region, Denmark
Presentation Documents
OBJECTIVES: Despite the curability of Stage IIB/IIC melanoma with resection, recurrence rates of the disease remain high but heterogeneous due to plateaus in the recurrence free survival (RFS) examined by several RCTs and RWD sources. This study explored the LTS rates among Stage IIB/IIC melanoma patients via mixture cure models (MCMs) using published RCTs and RWD sources.
METHODS: MCMs were employed separately to digitized RFS data from 3 real-world cohort studies (2 from the US [n1 = 90, n2 = 567], 1 from Norway [n=2317]), data pooled across 10 RCTs (n=3927) identified by a targeted literature review, and data pooled across the surgery alone arms of a restricted subset of these RCTs published after year 2000 (n = 1635). In the MCMs, patients were classified in two latent subgroups as cured (long-term survivors) and uncured. Cured subgroup was subject to only non-disease-related mortality whereas the uncured subgroup was subject to both recurrence and all-cause mortality. For each cohort, survival of the cured was estimated using corresponding baseline demographic information, and age- and sex-adjusted background mortality rates reported by WHO. Time-to-event outcomes for the uncured were modeled by standard parametric distributions. Model selection was guided by statistical and visual fit to the reported data.
RESULTS: Lognormal MCMs provided the best fit for all data sets with relatively more conservative LTS-rates than other candidate models. The estimated LTS rates (95% CI) were 0.34 (0.14-0.61) and 0.31 (0.21-0.44) for the US RWD sources; 0.56 (0.52-0.60) for the Norwegian RWD; and 0.46 (0.43-0.50) and 0.52 (0.48-0.56) for the pooled RCT data and its restricted subset, respectively. Estimated LTS rates varied <3% between the top-2 fitting MCMs across all data sets.
CONCLUSIONS: The variability in estimated LTS-rates across different cohorts and treatment settings highlights a high unmet need for novel systemic therapies in postoperative treatment of Stage IIB/IIC melanoma.
Conference/Value in Health Info
Value in Health, Volume 26, Issue 11, S2 (December 2023)
Code
MSR12
Topic
Clinical Outcomes, Methodological & Statistical Research, Study Approaches
Topic Subcategory
Decision Modeling & Simulation, Relating Intermediate to Long-term Outcomes
Disease
Oncology