Pembrolizumab Vibostolima

MK-7684A is Merck's co-formulation of pembrolizumab (Keytruda, anti-PD-1) and vibostolimab (anti-TIGIT) in a single syringe. The thesis was that TIGIT would be the next immune checkpoint pillar after PD-1, with dual blockade freeing more T cells to attack tumors than either alone. Merck ran the largest TIGIT program in industry, with at least eight KEYVIBE Phase 2/3 trials across lung, head and neck, and hematologic malignancies [1][2]. The program is now effectively dead. Merck discontinued KEYVIBE-008 in August 2024 for futility on overall survival plus excess immune-related adverse events [6], then discontinued KEYVIBE-003 and KEYVIBE-007 on December 16, 2024 after both also met pre-specified futility criteria for OS, and at the same time announced the discontinuation of KEYVIBE-006 and the remaining vibostolimab studies [9]. Roche's tiragolumab had already failed SKYSCRAPER-02 in ES-SCLC (March 2022) [5] and SKYSCRAPER-01 in PD-L1-high NSCLC (final OS analysis November 26, 2024) [4]. Gilead/Arcus's domvanalimab program has produced mixed Phase 2 readouts. The commercial logic for follow-on products to Keytruda - $29.5B in 2024 [3], loss of exclusivity in 2028 - remains intact, but TIGIT will not be the answer. CITYSCAPE first lit the field up in 2020 [8]; in 2026 the dual-brake-release thesis is effectively falsified across three major sponsors.

MK-7684A is a fixed-dose co-formulation of two products: pembrolizumab is approved as Keytruda across 30+ indications, but vibostolimab is investigational and has never been approved anywhere. The combination is regulatorily novel. Merck never disclosed breakthrough therapy or accelerated approval designations for the co-formulation. The Phase 3 program has been wound down. KEYVIBE-008 in first-line extensive-stage SCLC (NCT05224141) was discontinued August 8, 2024: the independent Data Monitoring Committee concluded the regimen met pre-specified futility criteria on overall survival and showed a worse safety profile than atezolizumab + chemotherapy [6]. KEYVIBE-003 in PD-L1-positive metastatic NSCLC (NCT04738487) and KEYVIBE-007 in perioperative NSCLC were discontinued December 16, 2024; both also met pre-specified futility criteria for OS at pre-planned analyses, and Merck simultaneously announced it would discontinue the Phase 3 KEYVIBE-006 trial and the remaining vibostolimab studies given the totality of data [9]. KEYVIBE-010 in resected high-risk melanoma had been discontinued earlier. There is no remaining active Phase 3 program for MK-7684A. There will be no FDA filing. Merck has also discontinued the parallel KEYFORM program for favezelimab (anti-LAG-3) [9]. The relevant question is no longer 'when does this read out' - it is 'what does Merck publish about why it failed, and which trials produce useful biomarker data for the field.'

PD-1 (programmed cell death protein 1) sits on the surface of T cells like a brake pedal. When tumors express PD-L1, the ligand engages PD-1 and tells the T cell to stand down, letting the tumor escape immune attack. Pembrolizumab blocks PD-1 from binding PD-L1, releasing the brake. This works spectacularly in some patients and not at all in others, and the unmet need is everyone who doesn't respond. TIGIT (T-cell immunoreceptor with Ig and ITIM domains) is another brake: a separate inhibitory receptor on T cells and NK cells that's upregulated in exhausted tumor-infiltrating lymphocytes. The biological bet was that many PD-1 non-responders have T cells exhausted by multiple brakes at once. Block PD-1 alone and TIGIT still suppresses them. Block both and you might recover responses. Animal models showed clean synergy, and early Phase 2 data with tiragolumab plus atezolizumab in PD-L1-high NSCLC (CITYSCAPE) showed an ORR jump that lit the field up [8]. Then the Phase 3 readouts came in and the synergy disappeared. Leading hypotheses for why dual blockade failed in the clinic: (1) clinical TIGIT activity is Fc-gamma receptor dependent - efficacy requires antibody-dependent cellular cytotoxicity (ADCC) via NK cells, and isotype differences across tiragolumab (Fc-active IgG1), vibostolimab (IgG1), and domvanalimab (Fc-silent IgG1) may explain inconsistent signals; (2) terminal T-cell exhaustion in lung cancer is too deeply established for incremental checkpoint release to rescue cytotoxic function - exhausted TILs co-express many inhibitory receptors and blocking one more produces diminishing returns; (3) the CITYSCAPE PD-L1-high enrichment strategy selected a population where PD-1 monotherapy already captures most responders, leaving little residual TIGIT-responsive benefit to detect. Bottom line: PD-1 remains the most validated target in oncology, supporting $29.5B in Keytruda revenue in 2024 [3]. TIGIT, after three years of Phase 3 failures across Roche, Merck, and likely Gilead/Arcus, is not. The biology is real but the magnitude of clinical benefit looks small to none.

KEYVIBE-008 (NCT05224141) was the Phase 3 trial in first-line extensive-stage small cell lung cancer: randomized, double-blind, MK-7684A plus etoposide/platinum chemotherapy versus atezolizumab (Tecentriq) plus etoposide/platinum, with co-primary endpoints of overall survival and progression-free survival and target enrollment around 460 patients [1]. The comparator was well-chosen because atezolizumab + chemo is the ES-SCLC standard from IMpower133 [7]. The design closely paralleled Roche's SKYSCRAPER-02 (also ES-SCLC, also vs atezo+chemo, primary OS failure announced March 29, 2022 [5]). Merck announced discontinuation on August 8, 2024 - the IDMC concluded MK-7684A met pre-specified futility criteria on OS and added immune-related adverse events without improving survival [6]. KEYVIBE-003 (NCT04738487) compared MK-7684A vs pembrolizumab monotherapy in PD-L1-positive metastatic NSCLC, primary endpoint OS in PD-L1 TPS ≥50% [2]; on December 16, 2024 Merck disclosed the trial met futility criteria for OS at a pre-planned analysis and was being discontinued [9]. KEYVIBE-007 in perioperative NSCLC was also discontinued December 16, 2024 on the same futility grounds [9]. Merck simultaneously terminated KEYVIBE-006 and the other remaining vibostolimab studies given the totality of the data. Trial design was not the problem; the comparators, endpoints, and enrollment targets were appropriate for the questions being asked. The drug did not beat control. The class-wide pattern across SKYSCRAPER, KEYVIBE, and ARC trials argues for a biological failure of the thesis, not an executional failure of any single trial.

Our model puts this drug's approval odds at 25%. That starts from a baseline of about 48%-the historical approval rate for Phase 3 drugs in this area-then adjusts based on ten facts about the trial and sponsor. The number is helped by more secondary endpoints than typical and a sponsor with a strong approval track record, but is pulled down by weak earlier-phase results and a randomized trial design. The remaining factors fall close to average, so they leave the estimate near where the baseline put it.

Efficacy risk has fully resolved to the downside. The TIGIT class has now failed in PD-L1-high NSCLC (SKYSCRAPER-01 final OS Nov 26, 2024 [4]; KEYVIBE-003 futility Dec 16, 2024 [9]), in ES-SCLC (SKYSCRAPER-02 [5]; KEYVIBE-008 [6]), and in perioperative NSCLC (KEYVIBE-007 [9]). The most likely explanation is that the dual-checkpoint thesis was overweight on translational rationale and underweight on the real biology: exhausted T cells in solid tumors carry many inhibitory receptors simultaneously, and blocking one more checkpoint does not unlock meaningfully more cytotoxic activity beyond what PD-1 blockade already achieves. The biomarker question remains formally open - no validated patient selection strategy has emerged across multiple sponsors with deep biomarker programs - but the practical answer for MK-7684A is moot since the program is wound down. Safety risk is documented: KEYVIBE-008 was halted partly for excess immune-mediated adverse events, the on-target toxicity expected from doubling up checkpoint blockade [6]. Execution risk is no longer prospective - Merck has completed the unenviable task of shutting the program down. Commercial risk is now structural rather than program-specific: with TIGIT eliminated as a Keytruda life-extension lever, Merck's options for post-2028 LOE protection narrow to subcutaneous Keytruda (Keytruda Qlex, FDA approved September 19, 2025 [10]), the Daiichi Sankyo ADC partnership (datopotamab deruxtecan, ifinatamab deruxtecan, raludotatug deruxtecan), the Moderna mRNA-4157 cancer vaccine collaboration, and smaller-molecule plays. The biosimilar threat is real: at least two pembrolizumab biosimilar programs are in late-stage development targeting the 2028 LOE window.

Reclassify from 'watching' to 'closed.' MK-7684A is not a future readout to monitor - it is a closed chapter in immuno-oncology, useful primarily as a negative control for the dual-checkpoint thesis and as a class-level data point on how many shots on goal a target deserves before the field walks away. The investor-relevant question is no longer about vibostolimab. It is about what fills the post-Keytruda gap, with the TIGIT lever permanently off the table. The pieces Merck has left: (1) subcutaneous Keytruda (Keytruda Qlex, pembrolizumab + berahyaluronidase alfa-pmph, FDA approved September 19, 2025 [10]) extends the franchise through formulation-based exclusivity and convenience but does not break new ground biologically; (2) the Daiichi Sankyo ADC partnership covers three assets across lung, breast, and other solid tumors with multi-billion-dollar peak-sales potential per asset under analyst consensus, and is now the primary checkpoint-adjacent growth story; (3) earlier-stage bets including mRNA vaccines and bispecifics. Biosimilar entry is the timeline risk: at least two late-stage pembrolizumab biosimilar programs are positioned for the 2028 LOE window, which will compress Keytruda IV revenue regardless of how well Keytruda Qlex defends share. For investors: stop modeling any vibostolimab contribution. Reweight Merck's post-2028 story toward ADCs and subQ formulation defense. The next near-term catalysts to watch are ADC Phase 3 readouts and any Merck pipeline-day disclosures at ASCO 2026 (June) and JPM 2027 (January). No additional vibostolimab readouts are expected; if Merck publishes the failed-trial biomarker analyses, those will be scientifically interesting but not commercially relevant to MK-7684A specifically.