Melanoma Trial Design Patterns
Definition
Melanoma Phase 3 trials are stratified by: disease stage (unresectable/metastatic Stage IV vs. resected Stage III/IV adjuvant), BRAF mutation status (BRAF V600E/K mutant ~50% vs. wild-type), PD-L1 expression, LDH level, and tumor burden. The treatment landscape is divided into two distinct strategies: immune checkpoint inhibitor (ICI) monotherapy or combination (anti-PD-1 ± anti-CTLA-4), and BRAF/MEK inhibitor combinations for BRAF V600-mutant tumors.
Trial patterns: OS primary predominates in metastatic melanoma (IO and targeted therapy). RFS/DFS primary in adjuvant settings.
Regulatory Position
Metastatic melanoma: OS is the gold standard primary endpoint. Key IO approvals based on OS: ipilimumab (MDX010-20, OS HR 0.66), nivolumab+ipilimumab (CHECKMATE-067, OS HR 0.55). BRAF+MEK inhibitor combinations: OS primary in BRIM-3, coBRIM, COMBI-d/v.
Adjuvant melanoma: RFS (relapse-free survival) or DFS is the standard primary endpoint for adjuvant Phase 3. OS is key secondary; follow-up of 10+ years required for OS maturity given long post-recurrence survival.
Accelerated approval: ORR used for accelerated approval in specific BRAF V600-mutant settings and tumor-agnostic approvals (pembrolizumab MSI-H, larotrectinib NTRK+).
Status: FDA Cancer Endpoints 2018 = Final
When to Use Each Endpoint
| Setting | Primary Endpoint | Key Secondary | Notes |
|---|---|---|---|
| 1L metastatic — IO (PD-1 mono) | OS | PFS, ORR | KEYNOTE-006 (pembro vs ipi, OS HR 0.68) |
| 1L metastatic — IO combo (nivo+ipi) | OS | PFS, ORR | CHECKMATE-067 (nivo+ipi vs nivo vs ipi; OS HR 0.52 combo vs ipi) |
| 1L metastatic — BRAF V600+ (BRAF+MEK) | OS | PFS, ORR | COMBI-d (dab+tram, OS HR 0.71); COMBI-v; coBRIM |
| 2L metastatic — IO post-BRAF/MEK | OS or PFS | ORR | |
| Adjuvant Stage III (high-risk) | RFS | OS | KEYNOTE-054 (pembro, RFS HR 0.57); CHECKMATE-238 (nivo, RFS HR 0.66) |
| Adjuvant BRAF V600+ Stage III | RFS | OS | COMBI-AD (dab+tram adjuvant, RFS HR 0.47) |
| Neoadjuvant (resectable Stage III) | pCR/EFS | RFS, OS | SWOG S1801, NeoTRIPaPDL1 (emerging) |
Design Considerations
Non-Proportional Hazards (IO Trials)
IO monotherapy and combination trials in melanoma demonstrate characteristic non-proportional hazards:
- Early crossing or overlap of survival curves (patients who progress early may do so faster on IO due to hyperprogression)
- Late plateau (long-term survivors, immune memory — 5-year OS ~40–50% with nivo+ipi)
Statistical methods required:
- RMST as pre-specified sensitivity for OS
- Landmark OS at 1, 2, 3, 5 years as key secondary
- Cox HR reported with acknowledgment of PH violation if Schoenfeld residuals p < 0.05
BRAF Testing and Stratification
- Central BRAF testing required for BRAF-targeted therapy trials
- Pre-stratification by BRAF status in IO combination trials (some patients receive BRAF/MEK if V600+)
- BRAF V600E vs. V600K distinction: V600K (~10–15% of BRAF+) may have different response to BRAF inhibitors — should be pre-specified for subgroup analysis
Adjuvant RFS Assessment
Adjuvant melanoma RFS assessment:
- Physical examination every 3 months for 2 years, then every 6 months for years 3–5
- Imaging (CT or PET-CT): Every 6 months for 5 years (though practice varies); frequency is less standardized than solid tumor trials with RECIST endpoints
- RFS vs. DFS: In melanoma, RFS typically excludes deaths unrelated to melanoma (all-cause deaths may or may not be included — must be pre-specified). FDA 2018 guidance recommends all-cause death; some melanoma-specific analyses censor non-melanoma deaths.
Intercurrent Events
Crossover from adjuvant ipilimumab control to PD-1 therapy: Post-trial crossover when PD-1 therapies became available affected OS in some adjuvant trials. Treatment policy for OS (count all deaths); RPSFT as sensitivity if formal crossover occurred.
Subsequent BRAF/MEK therapy after IO progression: In BRAF V600+ patients progressing on IO, subsequent BRAF/MEK inhibitors are standard. Treatment policy for OS; dilutes OS signal.
Immunotherapy-related discontinuation (irAE): Patients stopping IO due to irAE may have durable responses (immune memory). Treatment policy strategy essential — events counted regardless of irAE-related discontinuation.
Regulatory Precedent
| Trial | Drug | Setting | Primary Endpoint | HR / Result | Approval |
|---|---|---|---|---|---|
| MDX010-20 | Ipilimumab vs. gp100 | 2L metastatic | OS | HR 0.66 (3-year OS 22% vs 12%) | Regular (2011) |
| CHECKMATE-067 | Nivo+ipi vs. nivo vs. ipi | 1L metastatic | OS | HR 0.52 (nivo+ipi vs ipi; 5-year OS 52%) | Regular (2015) |
| KEYNOTE-006 | Pembrolizumab vs. ipilimumab | 1L metastatic | OS + PFS | OS HR 0.68; PFS HR 0.58 | Regular (2015) |
| COMBI-d | Dabrafenib + trametinib | 1L BRAF V600E+ | PFS → OS | PFS HR 0.75; OS HR 0.71 | Regular (2014) |
| COMBI-AD | Dabrafenib + trametinib adjuvant | Adjuvant BRAF V600+ | RFS | HR 0.47 (3-year RFS 58% vs 39%) | Regular (2018) |
| KEYNOTE-054 | Pembrolizumab adjuvant | Adjuvant Stage III | RFS | HR 0.57 (1-year RFS 75% vs 61%) | Regular (2019) |
| CHECKMATE-238 | Nivolumab vs. ipilimumab adjuvant | Adjuvant Stage III/IV | RFS | HR 0.66 | Regular (2017) |
Limitations and Pitfalls
Long post-recurrence survival in adjuvant setting: Active IO treatment available at recurrence means OS follow-up of 10+ years required. RFS primary analyses at 2–3 years with immature OS are the norm.
BRAF testing heterogeneity: BRAF V600E and V600K have different inhibitor sensitivity. V600E-specific CDx required; off-label use in V600K requires protocol-specified eligibility criteria.
IO combination toxicity: Nivo+ipi has ~55% Grade 3/4 AE rate. Any OS benefit claim must be weighed against toxicity in benefit-risk analysis. FDA has accepted OS benefit despite high toxicity given the magnitude (HR 0.52 at 5 years).
Backlinks
Source: FDA Cancer Endpoints 2018 (Final) Compiled from FDA guidance + known Phase 3 trial results