Assessing the Impact of PD-L1 and BRAF Biomarkers on Long-Term Survivorship (LTS) Rates Among Treatment-Naive Advanced Melanoma Patients Receiving Immune Checkpoint Inhibitors (ICIs)
Author(s)
Murat Kurt, PhD, Paul Serafini, BA, Victoria Wan, MSc, Mir Sohail Fazeli, MD, PhD, Jean-Paul Collet, MD, PhD.
Evidinno Outcomes Research Inc, Vancouver, BC, Canada.
Evidinno Outcomes Research Inc, Vancouver, BC, Canada.
Presentation Documents
OBJECTIVES: BRAF and PD-L1 biomarkers have major prognostic role in identifying melanoma patients’ response to ICIs. This study investigated the impact of these biomarkers on LTS rates among treatment-naïve advanced melanoma patients receiving ipilimumab, nivolumab, and their combination.
METHODS: Publicly available Kaplan-Meier curves for melanoma-specific survival (MSS) from the Phase III CheckMate-067 study were digitized to reconstruct pseudo-patient-level data for PD-L1<5%, PD-L1≥5%, BRAF-Wild Type, and BRAF Mutant subgroups in each arm. Minimum follow-up was 10 years. Mixture cure models (MCMs) were applied to reconstructed MSS data for each subgroup in each arm. In the MCMs, patients were classified into two exclusive, latent subpopulations as cured (long-term survivors) and uncured, where cured patients were free from melanoma-related deaths. As definition of MSS censors non-melanoma-related deaths, MCMs did not require background mortality rates. MSS for the uncured was modeled via standard parametric distributions which were characterized simultaneously with LTS rates using maximum likelihood estimation. Statistical goodness-of-fit metrics, and visual inspection of candidate fits to reported Kaplan-Meier curves guided model selection.
RESULTS: Estimated LTS rates for (PD-L1<5%, PD-L1≥5%) subgroups from the best-fitting models were (16.5%, 34.1%), (43.9%, 55.4%) and (51.9%, 63.0%) in the ipilimumab, nivolumab and nivolumab+ipilimumab arms, respectively. Estimated LTS rates for (BRAF-Wild Type, BRAF-Mutant) subgroups from the best-fitting models were (15.7%, 20.3%), (46.0%, 41.0%) and (47.2%, 55.0%) in the ipilimumab, nivolumab and nivolumab+ipilimumab arms, respectively. In all arms, overlapping 95% CIs for LTS rates between the contrasting subgroups (PD-L1<5% vs. PD-L1≥5%; BRAF-Wild Type vs BRAF-Mutant) implied statistical insignificance for biomarkers’ impact on LTS rates.
CONCLUSIONS: PD-L1≥5% status had meaningful impact on LTS rates (≥11.1% across all arms) whereas BRAF-mutation status had relatively more modest impact on LTS rates (≥4.6% across ipilimumab containing arms). Results highlight clinical importance and predictive value of PD-L1 and BRAF biomarkers in selection of advanced melanoma patients for ICI treatment.
METHODS: Publicly available Kaplan-Meier curves for melanoma-specific survival (MSS) from the Phase III CheckMate-067 study were digitized to reconstruct pseudo-patient-level data for PD-L1<5%, PD-L1≥5%, BRAF-Wild Type, and BRAF Mutant subgroups in each arm. Minimum follow-up was 10 years. Mixture cure models (MCMs) were applied to reconstructed MSS data for each subgroup in each arm. In the MCMs, patients were classified into two exclusive, latent subpopulations as cured (long-term survivors) and uncured, where cured patients were free from melanoma-related deaths. As definition of MSS censors non-melanoma-related deaths, MCMs did not require background mortality rates. MSS for the uncured was modeled via standard parametric distributions which were characterized simultaneously with LTS rates using maximum likelihood estimation. Statistical goodness-of-fit metrics, and visual inspection of candidate fits to reported Kaplan-Meier curves guided model selection.
RESULTS: Estimated LTS rates for (PD-L1<5%, PD-L1≥5%) subgroups from the best-fitting models were (16.5%, 34.1%), (43.9%, 55.4%) and (51.9%, 63.0%) in the ipilimumab, nivolumab and nivolumab+ipilimumab arms, respectively. Estimated LTS rates for (BRAF-Wild Type, BRAF-Mutant) subgroups from the best-fitting models were (15.7%, 20.3%), (46.0%, 41.0%) and (47.2%, 55.0%) in the ipilimumab, nivolumab and nivolumab+ipilimumab arms, respectively. In all arms, overlapping 95% CIs for LTS rates between the contrasting subgroups (PD-L1<5% vs. PD-L1≥5%; BRAF-Wild Type vs BRAF-Mutant) implied statistical insignificance for biomarkers’ impact on LTS rates.
CONCLUSIONS: PD-L1≥5% status had meaningful impact on LTS rates (≥11.1% across all arms) whereas BRAF-mutation status had relatively more modest impact on LTS rates (≥4.6% across ipilimumab containing arms). Results highlight clinical importance and predictive value of PD-L1 and BRAF biomarkers in selection of advanced melanoma patients for ICI treatment.
Conference/Value in Health Info
2025-05, ISPOR 2025, Montréal, Quebec, CA
Value in Health, Volume 28, Issue S1
Code
MSR119
Topic
Methodological & Statistical Research
Disease
SDC: Oncology