Why This Study Matters
BRAF and MEK inhibitors improve progression-free survival and overall survival relative to other therapies and remain approved options for patients with BRAF V600-mutant advanced melanoma. Immune checkpoint inhibitors are a separate standard-of-care option, regardless of a tumor's driver-mutation status. Because targeted therapy tends to produce rapid but often short-lived responses, while checkpoint immunotherapy can produce slower but more durable responses, researchers have long hypothesized that combining the two approaches could capture the best of both. The COMBI-I trial tested exactly that hypothesis by adding the anti-PD-1 antibody spartalizumab to dabrafenib and trametinib.
This report is not a new trial — it is the final, long-term overall survival analysis of a trial whose primary endpoint, progression-free survival, was already reported as not statistically significant. That distinction matters for how the findings below should be read: because the primary endpoint was missed, overall survival in this analysis was not formally tested for statistical significance, and the reported P value is exploratory and descriptive rather than confirmatory.
Study Design
COMBI-I is a phase 3, randomized, double-blind, placebo-controlled trial. Adult patients with histologically confirmed unresectable or metastatic BRAF V600-mutant cutaneous melanoma were randomized 1:1 to spartalizumab 400 mg intravenously every 4 weeks plus dabrafenib 150 mg orally twice daily plus trametinib 2 mg orally once daily, or to placebo plus the same doses of dabrafenib and trametinib. Patients were enrolled between September 2017 and July 2018. In the original analysis, the primary endpoint of progression-free survival did not reach statistical significance: median progression-free survival was 16.2 months with the triplet versus 12.0 months with dabrafenib-trametinib alone (hazard ratio 0.82, 95% CI 0.66–1.03; P=.042, one-sided, nonsignificant). Because that primary endpoint was not met, overall survival in this final analysis was not formally tested for statistical significance — the P value reported for overall survival is exploratory and descriptive, from a one-sided stratified log-rank test.
Patient Population
Baseline characteristics were balanced between the two arms. Median age was 56 years in the spartalizumab arm and 55 years in the placebo arm; roughly three-quarters of patients in each arm had an ECOG performance status of 0. Most patients in both arms had stage IV M1c disease (62% and 65%, respectively), and BRAF V600E was the most common mutation subtype (88% and 89%). PD-L1 positivity (≥1%) was present in 52% of the spartalizumab arm and 48% of the placebo arm.
Final Overall Survival Analysis
Because COMBI-I's primary progression-free survival endpoint was not met, this overall survival result is a descriptive, exploratory finding rather than a confirmed, statistically validated benefit. The authors report that Kaplan-Meier curve separation emerged around 6 months and was sustained through follow-up.
Subgroup Analyses
The published report provides only a qualitative subgroup statement rather than individual subgroup statistics: in most subgroups of patient and tumor characteristics, the treatment effect favored the spartalizumab triplet (hazard ratio below 1). Specific subgroup-level hazard ratios, confidence intervals, or P values — for example by BRAF mutation subtype, PD-L1 status, or LDH level — were not reported in the article text reviewed for this piece; they may appear only in an online-only appendix.
Safety Profile
Adverse events of any grade occurred in 99.3% of patients on the spartalizumab triplet and 97.3% of patients on dabrafenib-trametinib alone (safety analysis set: n=267 vs n=264). Treatment-related adverse events occurred in 98.5% versus 88.6% of patients (all grades) and 57.3% versus 36.7% (grade ≥3). The authors state that safety findings in this final analysis were consistent with the primary analysis and that no new safety signals were observed.
| Adverse Event | Triplet, All Grades (n=267) | Triplet, Grade ≥3 | Dab-Tram Alone, All Grades (n=264) | Dab-Tram Alone, Grade ≥3 |
|---|---|---|---|---|
| Any adverse event | 265 (99.3%) | 200 (74.9%) | 257 (97.3%) | 172 (65.2%) |
| Pyrexia | 193 (72.3%) | 15 (5.6%) | 147 (55.7%) | 10 (3.8%) |
| Diarrhea | 97 (36.3%) | 3 (1.1%) | 71 (26.9%) | 9 (3.4%) |
| Nausea | 91 (34.1%) | 2 (0.7%) | 78 (29.5%) | 1 (0.4%) |
| Arthralgia | 88 (33.0%) | 5 (1.9%) | 80 (30.3%) | 5 (1.9%) |
| Blood CPK increased | 76 (28.5%) | 22 (8.2%) | 72 (27.3%) | 19 (7.2%) |
| Lipase increased | 66 (24.7%) | 35 (13.1%) | 38 (14.4%) | 17 (6.4%) |
| Neutropenia | 44 (16.5%) | 21 (7.9%) | 37 (14.0%) | 13 (4.9%) |
| Hypertension | 21 (7.9%) | 9 (3.4%) | 34 (12.9%) | 17 (6.4%) |
Serious adverse events occurred in 56.2% of patients on the spartalizumab triplet versus 46.6% on dabrafenib-trametinib alone, with pyrexia again the most common (17.2% versus 6.1%). The most common adverse event of special interest attributable to spartalizumab was skin reactions, seen in 54.0% of patients overall (61.0% versus 47.0% by arm), with grade ≥3 skin reactions in 4% of patients overall (6.7% versus 1.1% by arm).
The safety results in the final analysis were consistent with those reported in the primary analysis. No new safety signals were observed.
— Discussion, COMBI-I final overall survival analysis, Journal of Clinical Oncology
Interpretation and Broader Context
The authors frame this as the longest-follow-up dataset among trials that combined a checkpoint inhibitor with BRAF/MEK-targeted therapy in BRAF-mutant melanoma. The roughly 20-month numerical difference in median overall survival, and a hazard ratio of 0.760 favoring the triplet, are notable — but because COMBI-I's primary progression-free survival endpoint was not met, the authors explicitly describe this overall survival finding as a trend rather than a confirmed benefit. Other trials, including the phase 2 SECOMBIT trial and the DREAMseq trial, have supported checkpoint immunotherapy as the preferred first-line approach for most patients with BRAF-mutant advanced melanoma, underscoring that this triplet strategy remains investigational rather than a new standard of care.
Limitations
The authors identify several limitations specific to this final analysis. A large proportion of patients were censored from the overall survival analysis after study termination interrupted survival data collection — 140 patients (52.4%) in the spartalizumab arm and 114 patients (43.0%) in the placebo arm. Use of immune checkpoint inhibitors after progression was imbalanced between arms, with only about two-thirds of placebo-arm patients who progressed going on to receive subsequent immunotherapy. Treatment discontinuation due to grade ≥3 adverse events and nonadherence to protocol-specified dosing schedules also occurred, which the authors note may have been influenced by the onset of pyrexia. Because the trial's primary progression-free survival endpoint was not met, the overall survival result reported here was not formally tested for statistical significance and should be read as an exploratory trend, not a confirmed treatment benefit.
COMBI-I, a phase 3 trial in BRAF V600-mutant advanced melanoma, missed its primary endpoint of progression-free survival. This final analysis, at a median follow-up of 76.9 months, reports a numerically longer overall survival with spartalizumab added to dabrafenib and trametinib (61.5 months) versus dabrafenib and trametinib alone (41.6 months; hazard ratio 0.760), but because the primary endpoint was not met, this survival difference was not formally tested and is described by the authors as a trend rather than a confirmed benefit. Toxicity, including immune-related skin reactions, was meaningfully higher with the triplet, and more than half of the patients on the triplet were censored from the survival analysis after the trial was terminated early.