ESMO 2026: The Key Pembrolizumab Trials to Know

ESMO 2026 presented new results and long-term updates on pembrolizumab-based treatment across lung, urothelial, breast, renal, melanoma, and esophageal cancers.

This overview highlights efficacy, safety, and biomarker findings, including positive studies, negative trials, and exploratory analyses that may inform future treatment strategies.

KEYNOTE-975: No EFS Benefit With Pembrolizumab Added to Definitive Chemoradiotherapy

The randomized, double-blind phase III KEYNOTE-975 trial enrolled 703 patients with previously untreated, unresectable esophageal or gastroesophageal junction cancer to pembrolizumab plus definitive chemoradiotherapy (dCRT) or placebo plus dCRT. Manish Shah presented the second interim analysis at ESMO 2026.

At a median follow-up of 47.1 months, pembrolizumab did not significantly improve EFS in patients with PD-L1 CPS ≥10:

• Median EFS: 15.2 versus 14.7 months (HR 0.97; 95% CI 0.77–1.23; P = 0.3994).
• Median OS: 32.2 versus 35.0 months (HR 1.00; 95% CI 0.77–1.30). OS was not formally tested because the preceding EFS endpoint did not meet the prespecified significance threshold.
• Adverse-event-related discontinuation: 34% versus 27%.

The study was stopped. These findings do not support adding pembrolizumab to dCRT using the KEYNOTE-975 regimen.

KEYNOTE-866: Perioperative Pembrolizumab Does Not Significantly Improve EFS in Cisplatin-Eligible MIBC

The randomized, double-blind phase III KEYNOTE-866 trial enrolled 907 patients with cisplatin-eligible muscle-invasive bladder cancer. Patients received neoadjuvant pembrolizumab or placebo plus gemcitabine–cisplatin, followed by radical cystectomy with pelvic lymph node dissection and adjuvant pembrolizumab or placebo. Jens Bedke presented the second interim analysis at ESMO 2026.

At a median follow-up of 60.9 months, pembrolizumab did not significantly improve the primary EFS endpoint:

• EFS: Median 74.1 months versus not reached (HR 0.88; 95% CI 0.71–1.10; P = 0.1366).
• Pathological complete response: 44.4% versus 35.8%, with an estimated difference of 8.9 percentage points (95% CI 2.5–15.1).
• Overall survival: Median not reached in either arm (HR 1.00; 95% CI 0.79–1.27).
• Safety: Grade 3–5 treatment-emergent adverse events occurred in 80.2% versus 75.1%.

Despite a higher pCR rate, this analysis did not establish an EFS or OS benefit from adding perioperative pembrolizumab to neoadjuvant gemcitabine–cisplatin and surgery.

KEYNOTE-B15/EV-304: Higher ctDNA Clearance With Perioperative Enfortumab Vedotin Plus Pembrolizumab

An exploratory analysis of the phase III KEYNOTE-B15/EV-304 trial evaluated circulating tumor DNA (ctDNA) in cisplatin-eligible muscle-invasive bladder cancer. The trial randomized 808 patients to perioperative enfortumab vedotin plus pembrolizumab or neoadjuvant gemcitabine–cisplatin, with radical cystectomy in both groups. Jonathan A. Chatzkel presented the analysis at ESMO 2026.

ctDNA was assessed using the Signatera assay before and after neoadjuvant treatment and following surgery.

• Neoadjuvant ctDNA clearance: 82.1% with enfortumab vedotin plus pembrolizumab versus 66.7% with chemotherapy.
• Postoperative ctDNA positivity: 4.8% versus 13.5%, respectively.
• Clinical associations: ctDNA clearance was associated with more favorable EFS and pCR outcomes in both arms.
• EFS by ctDNA status: Hazard ratios favored enfortumab vedotin plus pembrolizumab across baseline and postoperative ctDNA subgroups, although some confidence intervals crossed 1.

These exploratory findings support the prognostic relevance of ctDNA status and clearance. They do not establish ctDNA as a predictive biomarker for selecting treatment or justify treatment de-escalation based on ctDNA negativity.

neoHIP: Combined Biomarkers May Identify Greater Pembrolizumab Benefit in HER2-Positive Early Breast Cancer

An exploratory biomarker analysis from the phase II neoHIP trial evaluated whether hormone receptor (HR) status, PD-L1 expression, and stromal tumor-infiltrating lymphocytes (sTILs) could identify greater benefit from adding pembrolizumab to taxane, trastuzumab, and pertuzumab (THP). David Page presented the findings at ESMO 2026.

Tissue biomarkers were evaluable in 82 of 116 patients with stage II/III HER2-positive breast cancer: 45 receiving THP and 37 receiving THP plus pembrolizumab.

• Overall trial result: Adding pembrolizumab increased pCR from 48% to 67%, an absolute improvement of 19 percentage points (P = 0.03).
• Individual biomarkers: HR status, PD-L1, and sTILs were associated with pCR but did not individually clearly enrich for incremental pembrolizumab benefit.
• Combined assessment: A rule incorporating HR negativity, PD-L1 CPS ≥10, or sTILs ≥10% identified 67% of the biomarker-evaluable population and showed enrichment for pembrolizumab benefit.

These findings suggest that combined biomarker assessment may help refine patient selection, but prospective phase III validation is needed before it can guide treatment or support omission of additional cytotoxic therapy.

neoHIP Trial: Biomarker-Guided Immunotherapy in HER2-Positive Breast Cancer

EV-103/KEYNOTE-869: Long-Term Responses and Survival With Enfortumab Vedotin Plus Pembrolizumab

Long-term follow-up from the phase Ib/II EV-103/KEYNOTE-869 study showed sustained responses and survival with first-line enfortumab vedotin plus pembrolizumab in cisplatin-ineligible locally advanced or metastatic urothelial carcinoma. Jose Pablo Maroto Rey presented the findings at ESMO 2026.

The analysis pooled 121 combination-treated patients from cohorts DE/A and K. Cohort K also included 73 patients receiving enfortumab vedotin alone; no formal between-arm statistical comparisons were planned.

• Response durability: Median duration of response was 29.6 months; estimated response-duration rates were 37.0% at five years and 30.8% at seven years.
• Progression-free survival: Median PFS was 10.4 months, with estimated five- and seven-year rates of 24.3% and 20.2%.
• Overall survival: Median OS was 27.7 months. Estimated five- and seven-year OS rates were 33.0% and 28.2%.
• Safety: Grade ≥3 treatment-related adverse events occurred in 64.5% of combination-treated patients. Peripheral neuropathy fully resolved in 21.8% of affected patients.

These findings demonstrate prolonged benefit in a subset of patients. The pooled analysis provides long-term context for the regimen, but does not establish comparative efficacy versus enfortumab vedotin alone.

LITESPARK-012: Adding Belzutifan or CTLA-4 Blockade Does Not Improve Outcomes Over Pembrolizumab–Lenvatinib

The randomized, open-label phase III LITESPARK-012 trial enrolled 1,688 patients with untreated advanced clear cell renal cell carcinoma. It compared pembrolizumab–lenvatinib plus belzutifan, or coformulated quavonlimab–pembrolizumab plus lenvatinib, with pembrolizumab–lenvatinib alone. Toni K. Choueiri presented the findings at ESMO 2026.

At a median follow-up of 43.6 months, neither experimental regimen improved efficacy:

• Belzutifan-containing triplet: Median PFS was 20.8 versus 23.7 months with control (HR 1.08; 95% CI 0.93–1.26). The OS HR was 1.04 (95% CI 0.87–1.25).
• Quavonlimab-containing regimen: Median PFS was 20.7 months (HR versus control 1.07; 95% CI 0.92–1.25). The OS HR was 1.12 (95% CI 0.94–1.35). No benefit was established in the IMDC intermediate/poor-risk population.
• Objective response: ORR was 69.1%, 69.0%, and 76.3% in the belzutifan, quavonlimab, and control arms, respectively.
• Treatment discontinuation: All-cause adverse events led to discontinuation of at least one drug in 51.6%, 51.4%, and 36.2%, respectively.

The findings do not support adding belzutifan or quavonlimab to first-line pembrolizumab–lenvatinib. Higher discontinuation rates and shorter treatment exposure in the experimental arms may have influenced outcomes, but do not establish the reason for the lack of benefit.

PERSEE: Adding Chemotherapy to Pembrolizumab Improves PFS in PD-L1–High Advanced NSCLC

The randomized, open-label phase III PERSEE trial compared pembrolizumab plus platinum-based chemotherapy with pembrolizumab alone in 321 patients with untreated advanced NSCLC, PD-L1 expression ≥50%, ECOG performance status 0–1, and no EGFR/ALK alterations. Chantal Decroisette presented the findings at ESMO 2026.

At a median follow-up of 38.7 months, the combination significantly improved the primary endpoint of progression-free survival assessed by blinded central review:

• Median PFS: 16.1 versus 6.9 months (HR 0.56; 95% CI 0.43–0.74; P <0.001).
• Overall survival: Median OS was 50.2 versus 29.9 months, but a statistically significant benefit was not established (HR 0.77; 95% CI 0.57–1.05).
• Objective response: ORR was 77.5% versus 50.3%.
• Response duration: Median duration of response was 30.0 versus 15.6 months.
• Safety: Grade ≥3 treatment-related adverse events occurred in 53.8% versus 19.3%.

PERSEE provides direct randomized evidence that adding chemotherapy improves PFS and response rates even in PD-L1–high disease. The increased toxicity and absence of an established OS benefit remain important considerations when selecting between combination treatment and pembrolizumab monotherapy.

DeLLphi-305: Tarlatamab Plus Durvalumab Improves Survival as Maintenance Therapy in ES-SCLC

The randomized phase III DeLLphi-305 trial enrolled 563 patients with extensive-stage small cell lung cancer without progression after first-line durvalumab plus platinum–etoposide. Patients received maintenance tarlatamab plus durvalumab or durvalumab alone. Jacob Sands presented the primary analysis at ESMO 2026.

At approximately 16 months of median follow-up:

• Overall survival: Median OS was 21.1 versus 12.7 months (HR 0.42; 95% CI 0.32–0.54; P <0.001).
• Progression-free survival: Median PFS was 7.4 versus 1.9 months (HR 0.34; 95% CI 0.28–0.42; P <0.001).
• Objective response: ORR was 26.4% versus 4.1%.
• Safety: Grade ≥3 treatment-emergent adverse events occurred in 46.8% versus 40.6%.
• CRS and neurotoxicity: Cytokine release syndrome occurred in 49.6% of combination-treated patients, including grade ≥3 events in 0.4%. ICANS occurred in 1.1%, with no grade ≥3 events. No fatal CRS or ICANS events were reported.

Adding tarlatamab to maintenance durvalumab significantly improved survival and disease control. OS and PFS were measured from maintenance randomization, rather than the start of induction treatment.

DeLLphi-305: Tarlatamab–Durvalumab Redefines First-Line Maintenance in Extensive-Stage SCLC

PLUG-IN GEM 2301: Preliminary Activity With Enfortumab Vedotin Plus Pembrolizumab in Refractory Melanoma

The multicenter, single-arm phase II PLUG-IN GEM 2301 trial evaluated enfortumab vedotin plus pembrolizumab in 60 patients with advanced melanoma progressing after immunotherapy and, for BRAF-mutated disease, targeted therapy. Enrollment did not require a specific NECTIN-4 expression level. Guillermo de Velasco presented the findings at ESMO 2026.

At the April 2026 abstract cutoff, 53 patients were efficacy-evaluable:

• Confirmed ORR: 17%, with response confirmation ongoing.
• Best observed response: Approximately 28%, including two complete and 13 partial responses. This measure should be distinguished from confirmed ORR.
• Treatment status: Treatment remained ongoing in 53.3% of enrolled patients.
• Safety: Grade 3–4 toxicities occurred in 23.3%. Enfortumab vedotin dose reductions and interruptions occurred in 25% and 50%, respectively.

The abstract showed preliminary antitumor activity, but did not report whether the prespecified primary-endpoint criterion was met. The single-arm design and ongoing response confirmation limit conclusions about efficacy and durability.

KEYNOTE-564: Adjuvant Pembrolizumab Maintains DFS and OS Benefits at Seven Years

The final protocol-prespecified analysis of the phase III KEYNOTE-564 trial showed sustained disease-free survival (DFS) and overall survival (OS) benefits with adjuvant pembrolizumab in clear cell renal cell carcinoma at increased risk of recurrence after surgery. Toni K. Choueiri presented the findings at ESMO 2026.

The trial randomized 994 patients following nephrectomy, with or without metastasectomy, to pembrolizumab or placebo every three weeks for up to 17 cycles.

At a median follow-up of 85.9 months:

• Disease-free survival: Median DFS was not reached versus 68.3 months (HR 0.68; 95% CI 0.56–0.82). Seven-year DFS was 57.2% versus 44.6%.
• Overall survival: Median OS was not reached in either arm (HR 0.70; 95% CI 0.53–0.92). Seven-year OS was 82.3% versus 76.6%.
• Safety: No new serious treatment-related adverse events were reported.

These long-term findings reinforce the previously established DFS and OS benefits and support adjuvant pembrolizumab for eligible patients at increased risk of recurrence.