The Society for Immunotherapy of Cancer (SITC) is pleased to present highlights of the latest advances in cancer immunotherapy emerging from the 2025 ESMO Annual Meeting on Oct. 17 and 18, 2025.
2025 Scientific Highlights
Efficacy and safety of mRNA-4359, an mRNA cancer vaccine targeting PD-L1 and IDO1, in combination with pembrolizumab for immune checkpoint inhibitor resistant/refractory melanoma
1515MO. Clinical outcomes and PD-L1 expression analyses from a trial of mRNA-4359 plus pembrolizumab in checkpoint inhibitor-resistant/refractory (CPI-R/R) melanoma
David J. Pinato (Imperial College London Hammersmith Hospital, London, United Kingdom) presented results from an ongoing phase 1/2 trial of mRNA-4359 plus pembrolizumab for checkpoint inhibitor-resistant/refractory (CPI-R/R) melanoma. mRNA-4359 is an mRNA-based therapeutic vaccine that targets the immunomodulatory enzyme indoleamine 2,3-dioxygenase (IDO1) and PD-L1, drivers of anti-cancer immunity. mRNA-4359 is designed to elicit T cell responses against tumor and immunosuppressive cells, and previous studies indicate mRNA-4349 is well tolerated as monotherapy. 29 patients with CPI-R/R melanoma received pembrolizumab plus mRNA-4359 at 400 μg (n=14) or at 1000 μg (n=15) every 3 weeks for up to 9 doses, and the median follow-up was 20 weeks. mRNA-4359 plus pembrolizumab exhibited a manageable safety profile: most mRNA-4359-related adverse events (AEs) were grade 1 or 2, and 13.8% of patients experienced immune-related AEs. No dose-limiting toxicities or grade 4 or 5 AEs were observed. Of the 25 evaluable patients, the overall response rate (ORR) was 24% (n=6), the disease control rate (DCR) was 60%, and the median duration of response (DOR) was not reached. Among the 9 evaluable patients who were PD-L1 positive (TPS >= 1%), the ORR was 67%, and all 6 responders were included in this group. The DCR was 67% in PD-L1 positive patients, compared to 50% in patients who were PD-L1 negative (TPS <1%). Peripheral antigen-specific T cell responses and increased numbers of novel TCR clones were observed in patients after treatment with mRNA-4359. mRNA-4359 in combination with pembrolizumab is associated with a manageable safety profile, durable responses, enriched clinical efficacy in patients with PD-L1 positive disease, and peripheral T cell responses, supporting further clinical development and potentially meeting an unmet need in treating melanoma and other cancers that have acquired resistance to immune checkpoint inhibitors.
Personalized neoantigen cancer vaccine EVX-01 in combination with pembrolizumab for advanced melanoma
1516MO. EVX-01, a personalized cancer vaccine, induces potent T cell responses and durable disease control in advanced melanoma: 2-year follow-up
Muhammad A. Khattak (One Clinical Research, Hollywood Private Hospital & Edith Cowan University, Perth, Western Australia, Australia) presented the two-year follow-up of a phase 2 study of EVX-01, a cancer vaccine targeting mutation-derived neoantigens, in combination with pembrolizumab for advanced melanoma. 16 patients with stage III/IV unresectable melanoma received pembrolizumab every six weeks. After 12 weeks of induction therapy with pembrolizumab, patients received six priming doses of EVX-01 every two weeks, followed by four boosting doses every 12 weeks. No adverse events (AEs) of grade 3 or higher were associated with EVX-01, and the combination of EVX-01 with pembrolizumab was well-tolerated, with only one grade 3/4 AE, pancreatitis, occurring. The manufacturing success rate of EVX-01 was 100%. The best overall response rate of EVX-01 plus pembrolizumab was 75% (12/16). EVX-01 was associated with a 54% seroconversion rate, with 7 of 13 (54%) patients experiencing a deeper response upon initiation of EVX-01 therapy, and 92% of responders had continued their response at the 24-month follow-up. 81% of neoantigens in EVX-01 induced potent, specific T cell responses, and vaccine-specific T cell and T cell responses were induced and sustained in patients throughout the 24-month study. These results indicate that EVX-01 combined with pembrolizumab is well tolerated and induces potent and durable T cell and clinical responses, potentially providing a new treatment strategy for advanced melanoma that does not respond to immune checkpoint inhibitors. Clinical development of EVX-01 for high-risk melanoma, melanoma with an increased risk of metastasis, is ongoing.
Five-year follow-up and exploratory ctDNA analyses of CheckMate 274: Adjuvant nivolumab for high-risk muscle-invasive urothelial carcinoma
3068O. Adjuvant nivolumab vs placebo for high-risk muscle-invasive urothelial carcinoma: 5-year efficacy and ctDNA results from CheckMate 274
Matthew D. Galsky (Icahn School of Medicine at Mount Sinai, New York, New York, United States of America) discussed the 5-year follow-up results and exploratory circulating tumor DNA (ctDNA) data from CheckMate 274, a phase 3 clinical study of adjuvant nivolumab vs. placebo in patients with high-risk muscle-invasive urothelial carcinoma (MIUC) after radical surgery. Previous reports of CheckMate 274 indicated that adjuvant nivolumab was associated with significant improvement in disease free survival (DFS), making nivolumab the standard of care adjuvant treatment for patients with high-risk MIUC after radical surgery. 709 patients participated in the study, with 353 receiving adjuvant nivolumab (nivolumab arm) and 356 receiving placebo (placebo arm) after radical surgery. At five years of follow-up adjuvant nivolumab was associated with DFS benefits compared to placebo in all patients (median DFS 21.9 months and 11.0 months, respectively, HR 0.74), with a larger benefit in patients with PD-L1-positive tumors (TPS>=1%; median DFS 55.5 months and 8.4 months, respectively, HR 0.58). In the subgroup of patients with muscle-invasive urothelial cancer originating in the bladder (MIBC), median DFS in the nivolumab and placebo arms was 25.6 months and 9.4 months, respectively (HR 0.66), and in patients with MIBC and PD-L1-positive tumors, median DFS was 59.4 months with nivolumab, compared to 8.3 months with placebo (HR 0.50). Nivolumab-associated DFS benefits were observed in patients with MIBC regardless of whether they received neoadjuvant cisplatin, with hazard ratios of 0.63 among patients who had received neoadjuvant cisplatin and 0.70 in patients who had not received neoadjuvant cisplatin. Interim analyses indicate overall survival (OS) was also longer in patients who received nivolumab compared to placebo. In the total patient population, median OS was 75.0 months in the nivolumab arm, compared to 50.1 months in the placebo arm (HR 0.83). A similar pattern was observed in patients with PD-L1-positive tumors (HR 0.63), in patients with MIBC (HR 0.73), and in patients with MIBC and PD-L1 positive tumors (HR 0.51). Median disease-specific survival (DSS) was also longer with adjuvant nivolumab compared to pembrolizumab. In the total patient population, median DSS was not reached in either arm, and in patients with PD-L1 positive tumors, median DSS was not reached for the nivolumab arm and 92.1 months for the placebo arm. Exploratory ctDNA analysis was performed on cycle 1 day 1 samples from 133 patients. 40% of patients had detectable ctDNA, and median DFS was more than 10-fold higher in patients with undetectable ctDNA (52.1 months) compared to patients with detectable ctDNA (5.0 months). Median OS was also higher in patients with undetectable ctDNA (not reached) compared to those with detectable ctDNA (28.2 months). Among patients with detectable ctDNA (n=54), adjuvant nivolumab improved median DFS and median OS (7.4 months and 36.2 months, respectively), compared to placebo (median DFS 2.8 months and median OS 19.3 months). No new safety signals were observed, as patients had been off treatment for four or more years. In summary, at the five-year follow-up, adjuvant nivolumab was associated with extended DFS, OS, and DSS compared to placebo for every patient subgroup examined, even among those who were ctDNA positive, although conclusions are limited due to small sample size. These results provide further support for adjuvant nivolumab as the standard of care for high-risk MIUC after radical resection and has potential for curative outcomes. Recent approvals of subcutaneous nivolumab may provide a convenient dosing alternative for patients across tumor types, providing increased access to treatment.
First results of STARt-001: Invikafusp-alfa, a first-in-class bispecific dual T cell agonist, for antigen-rich solid tumors
LBA55. START-001: Initial phase II clinical activity of invikafusp alfa, a first-in-class T cell receptor (TCR) β-chain-targeted bispecific antibody as monotherapy in patients with antigen-rich solid tumors resistant to immune checkpoint blockade (ICB)
Antoine Italiano (Institut Bergonié, Bordeaux, France) reported the first results of phase 2 expansion of the STARt-001 study investigating invikafusp alfa, a bispecific T cell agonist targeting Vβ6/Vβ10 T cell receptors (TCRs) for antigen-rich solid tumors. Previous studies indicate that invikafusp alfa activates and expands Vβ6/Vβ10 T cell subsets, which represent the most prevalent T cell subsets in tumor-infiltrating lymphocytes. 63 patients with tumor mutation burden-high/microsatellite instability-high (TMB-H/MSI-H) tumor representing 21 different tumor types participated in the study. 67% of patients had received prior treatment with immune checkpoint inhibitors. Invikafusp alfa exhibited promising efficacy, with 52% of patients experiencing tumor regression and overall response rates of 20.5% in TMB-H tumors and 30% in MSI-H tumors. The disease control rate was 79%. Clinical benefits were observed in a broad range of tumor types, including those considered to be refractory, either primarily or secondarily, to checkpoint inhibitors, including microsatellite-stable colorectal cancer, head and neck cancers, breast cancers, and lung cancers. Invikafusp alfa was well-tolerated, and its safety profile reflected its mechanism of action, selective T cell activation and expansion. The most commonly observed treatment-related adverse event observed (86%) was grade 1-3 cytokine release syndrome, and no grade 4 or 5 toxicities were observed. Clinical activity was linked to expansion of Vβ6 T cells in peripheral blood and in the tumor. With its clinically meaningful activity and manageable safety profile, the bispecific dual T cell agonist invikafusp alfa represents a novel class of immunotherapy and a promising option for the treatment of TMB-H/MSI-H tumors that are resistant to immune checkpoint inhibitors. Ongoing clinical trials of invikafusp alfa include monotherapy for TMB-H/MSI-H/mismatch repair-defective tumors and invikafusp alfa in combination with antibody drug conjugates for breast cancer. The United States Food and Drug administration recently granted invikafusp alfa fast-track designation for TMB-H metastatic colorectal cancer.
Patient-level analysis of checkpoint-inhibitor pneumonitis
2799O. Patient-level analysis of pneumonitis from clinical trials of nivolumab-based regimens in solid tumours
Jarushka Naidoo (Beaumont RCSI Cancer Centre, Dublin, Ireland) presented pooled analysis of patients with reported checkpoint-inhibitor pneumonitis (CIP) from 24 phase 3 global clinical trials of nivolumab +/- ipilimumab +/- chemotherapy across nine solid tumors. CIP is an uncommon but potentially fatal immune-related adverse event (irAE). The natural history and clinical outcomes of CIP are poorly understood, and more guidance in diagnosis, management and monitoring CIP is needed. 716 patients in the study population of 11,777 (6.1%) developed CIP, representing 785 CIP events. Patient characteristics associated with increased risk of CIP include age of 65 years or more and a history of smoking. 40% of CIP cases had non-small cell lung cancer (NSCLC), and a higher incidence of CIP was associated with treatment with ipilimumab and nivolumab (8% of CIP cases) compared to nivolumab monotherapy (4.6%) or nivolumab + chemotherapy (6.4%). 68% of patients had grade 1 or 2 CIP, and the median time to onset was 17 weeks, with a range of 0.3 to 123.6 weeks. Median time to resolution was 17 weeks for all grades of CIP, with a range of 0.1 to 143 weeks. Immunotherapy was held or discontinued for the majority of cases of CIP, with only 34.8% of patients with grade 1 CIP, 4.5% of patients with grade 2 CIP, and 0.9% of patients with grade 3 CIP continuing treatment. 70% of CIP events were completely resolved, and 4% were fatal. 8.5% of patients with CIP experienced recurrent pneumonitis, defined by a subsequent pneumonitis event without immunotherapy rechallenge. Most patients experienced one recurrent event, and it was usually of low-grade severity. Recurrent pneumonitis was more common in patients who were of white race and in patients with renal cell carcinoma (representing 17.6% of cases) or with hepatocellular carcinoma (representing 13.3% of cases). The median time to recurrence was 6 weeks, and the median duration of a recurrence was 5.1 weeks. Recurrent pneumonitis was managed by discontinuing or holding immunotherapy in 67% of cases of grade 1 recurrent pneumonitis and in 100% of cases of grades 2 through grade 5. 75% of recurrent cases were completely resolved, and 4% were fatal. Chronic pneumonitis, pneumonitis lasting 12 or more weeks, was observed in 34.6% of patients who experienced CIP, and 94.5% of patients with chronic pneumonitis experienced a single chronic event. Regarding the management of CIP, 80.5% of cases of grade 1 pneumonitis (representing 33% of CIP cases) were not treated with corticosteroids (CS), and symptoms did not worsen in 74.3% of those cases, suggesting CS for grade 1 pneumonitis did not greatly impact outcomes. Conversely, treatment of grade 2 pneumonitis (43.3% of pneumonitis cases ) with CS was associated with beneficial outcomes. 86% of patients with grade 2 pneumonitis were treated with CS, and symptoms did not worsen in 90.4% of those cases. A median dose of 1.13 mg/kg of CS was used for managing any grade of pneumonitis, and a median dose of 1.39 mg/kg was used to manage cases of grade 3 or higher. The median duration of CS treatment was 7 weeks. Second-line immunosuppression was rarely used to manage pneumonitis, and outcomes were universally poor, with 10 of 21 events (47.6%) being fatal. In addition to identifying risk factors associated with CIP, this study of the largest reported patient dataset provides new insights surrounding CIP onset and duration, less common phenotypes of CIP, and CIP management, potentially impacting the monitoring, diagnosis, and management of CIP and other irAEs.
Safety and efficacy of tarlatamab, a bispecific T cell engager, in combination with chemoimmunotherapy, for first-line extensive stage lung cancer
2757O. Tarlatamab with first-line chemoimmunotherapy for extensive stage small cell lung cancer (ES-SCLC): DeLLphi-303 study
Martin Wermke (NCT/UCC Early Clinical Trial Unit, TU Dresden, Dresden, Germany) presented safety and efficacy results from the phase 1b DeLLphi-303 study (parts 2, 4, and 7) investigating tarlatamab, a DLL3 x CD3 bispecific T cell engager, in combination with first-line chemoimmunotherapy followed by tarlatamab and anti-PD-L1 maintenance therapy for extensive stage small cell lung cancer (ES-SCLC). Patients with ES-SCLC who had received one cycle of chemoimmunotherapy (platinum-etoposide and anti-PD-L1 therapy (atezolizumab or durvalumab) received induction therapy of tarlatamab combined with chemoimmunotherapy for one to three cycles, followed by maintenance therapy with tarlatamab and anti-PD-L1 therapy (atezolizumab or durvalumab) until progressive disease occurred. 96 patients participated in the study. Three dose-limiting toxicities occurred, i.e immune effector cell-associated neurotoxicity syndrome (ICANS) (n=1), platelet count decrease (n=1) and thrombocytopenia (n=1). 43% and 35% of patients experienced grade 3 and grade 4 adverse events (AEs), respectively, and most were cytopenias resulting from chemotherapy. One fatal treatment-related AE, sepsis, occurred, and it was caused by chemotherapy. Tarlatamab-related AEs led to discontinuation of tarlatamab in 6% of patients, and no tarlatamab-related deaths occurred. Immune-related adverse events other than cytokine release syndrome (CRS) and ICANS were rare, occurring in 2% of patients. Most toxicities occurred in the induction phase of therapy, when patients received tarlatamab with chemoimmunotherapy. Cases of CRS were limited to cycle 1 and cycle 2 of tarlatamab, and most cases were grade 1 or 2. 5 cases of ICANS occurred during cycle 1 of tarlatamab, and most cases were grade 2. At a median follow-up of 13.8 months, the overall response rate was 71%, with 5 patients experiencing a complete response. The median duration of response was 11.0 months, and responses were ongoing in 49% of patients at the time of data cutoff. The 12-month overall survival rate was 80.6%, and median OS was not reached. Median progression free survival was 10.3 months, and at 12 months, 43.1% of patients were alive and free of disease progression. Tarlatamab in combination with first-line chemoimmunotherapy and anti-PD-L1 maintenance therapy exhibited a manageable safety profile and encouraging clinical activity. These data support further investigation of this therapy regimen, and a phase 3 DeLLphi-312 study comparing this therapy regimen to the current standard of care chemoimmunotherapy regimen for ES-SCLC is ongoing.
Final results of COMPASSION-15: PD-1/CTLA-4 bispecific antibody cadonilimab plus chemotherapy for gastric/gastroesophageal junction adenocarcinoma
2098MO. Cadonilimab (Cado) plus chemotherapy (chemo) versus chemotherapy as first-line (1L) treatment for advanced gastric or gastroesophageal junction (G/GEJ) adenocarcinoma: Final results of the phase III COMPASSION-15 trial
Lin Shen (Peking University Cancer Hospital, Beijing, China) presented the final analysis of the phase 3 COMPASSION-15 trial, comparing the efficacy and safety of cadonilimab plus chemotherapy (XELOX) versus chemotherapy (XELOX) as first line treatment for advanced gastric or gastroesophageal junction (G/GEJ) adenocarcinoma. Cadonilimab is a humanized PD-1/CTLA-4 bispecific antibody, and interim analysis of COMPASSION-15 indicated that cadonilimab in combination with chemotherapy significantly improved overall survival (OS) and progression free survival (PFS) compared to chemotherapy alone in patients with G/GEJ adenocarcinoma. 610 patients with previously untreated, locally advanced unresectable or metastatic G/GEJ adenocarcinoma participated in the study. 305 patients in the cadonilimab + XELOX arm received up to six cycles of cadonilimab and XELOX, followed by cadonilimab maintenance therapy. 305 patients in the placebo + XELOX arm received up to six cycles of placebo + XELOX, followed by placebo. At a median follow-up of 33.9 months, cadonilimab improved overall survival (OS) regardless of PD-L1 expression. In the intent to treat (ITT) population, the median OS was 13.9 months in the cadonilimab arm, compared to 11.1 months in the placebo arm (HR 0.61, p<0.001). A similar pattern was observed in the PD-L1 CPS>=5 patient subgroup (HR 0.49, p<0.001), and a numeric increase in OS was observed in the PD-L1 CPS<5 patient subgroup (HR 0.76, p=0.019). OS benefits associated with cadonilimab were observed in all predetermined patient subgroups. Cadonilimab also improved duration of response (DOR) regardless of PD-L1 status. The median DOR in the ITT population was 8.9 months in the cadonilimab + XELOX arm and 4.5 months in the placebo + XELOX arm (HR 0.45, p<0.001). In patients with PD-L1 CPS>=5, median DOR was 8.5 months in the cadonilimab + XELOX arm and 4.9 months in the placebo + XELOX arm (HR 0.43, p<0.001), and in patients with PD-L1 CPS<5, median DOR was 9.4 months in the cadonilimab + XELOX arm and 4.3 months in the placebo + XELOX arm (HR 0.52, p<0.001). Cadonilimab also significantly improved progression free survival in patients regardless of PD-L1 status. In patients with PD-L1 CPS<5, median PFS was 6.9 months in the cadonilimab + XELOX arm, compared to 4.6 months in the placebo + XELOX arm (HR 0.61, p<0.001). No new safety signals were observed. With an additional 16 months of follow-up, these results indicate that cadonilimab plus chemotherapy continues to maintain clinically meaningful and significant survival benefits with a manageable safety profile compared to chemotherapy regardless of PD-L1 expression, further supporting cadonilimab plus chemotherapy as a standard of care for first line treatment of G/GEJ carcinoma.
Intratumor administration of the immune cytokine L19IL2/L19TNF (Daromun) for high-risk locally advanced basal cell carcinoma
LBA60. Favorable activity and safety of intralesional L19IL2/L19TNF in high-risk laBCC and lacSCC patients
Lukas Flatz (University of Tübingen, Tübingen, Germany) presented efficacy and safety results from a cohort of the phase II, single arm DUNCAN trial examining intratumoral administration of immunocytokines L19IL2/L19TNF (Daromun) in locally advanced basal cell carcinoma (LA-BCC). Daromun delivers cytokines IL2 and tumor necrosis factor (TNF) to the tumor, increasing and activating TILs and inducing tumor necrosis, respectively. The expanded cohort consisted of 71 patients with high-risk LA-BCC who refused or were not candidates for surgery. Patients received weekly injections of Daromun for four weeks, and post-treatment surgery was optional. Most patients had one target lesion, and 42.2% of patients were treatment-naïve. Among the 57 evaluable patients, the best overall response rate, i.e. complete response (CR) or partial response (PR), was 57.8%, with 43.8% of patients achieving a CR, and the disease control rate was 87.7%. 30 patients (52.6%) achieved a pathological complete response, which allowed them to avoid surgery. Daromun was well-tolerated; most adverse events were grade 1 or 2 and consistent with prior studies. No treatment-related deaths occurred. With a CR rate of 43.8%, the immunocytokine Daromun exhibits substantial anti-tumor activity against LA-BCC as well as a manageable safety profile. Additional studies of Daromun for melanoma and other skin cancers are ongoing.
First interim analysis of KEYNOTE-905/EV-303: Perioperative enfortumab vedotin and pembrolizumab for patients with muscle-invasive bladder cancer who are cisplatin-ineligible
LBA2. Perioperative (periop) enfortumab vedotin (EV) plus pembrolizumab (pembro) in participants (pts) with muscle-invasive bladder cancer (MIBC) who are cisplatin-ineligible: The phase III KEYNOTE-905 study
Christof Vulsteke (Integrated Cancer Center Ghent, AZ Maria Middelares, Gent, Belgium) presented the first interim analysis of KEYNOTE-905/EV-303, a phase 3 study of perioperative enfortumb vedotin (EV) in combination with pembrolizumab (pembro) and radical cystectomy and pelvic lymph node dissection (RC + PLND) versus RC + PLND in patients with muscle-invasive bladder cancer (MIBC) who are cisplatin-ineligible. Neoadjuvant cisplatin-based chemotherapy followed by RC + PLND is the standard of care for MIBC, but nearly 50% of patients with MIBC are ineligible for treatment with cisplatin. These patients have no options for neoadjuvant therapy, and RC + PLND alone is associated with poor outcomes. 344 patients participated in the study. 170 patients received EV and pembrolizumab before and after RC + PLND, (EV + pembro arm), and 174 patients received RC + PLND (control arm). 149 patients (87.6%) in the EV + pembro arm and 156 patients (89.7%) in the control arm underwent surgery, and 100 (57.4%) patients in the EV + pembro arm started the adjuvant phase of therapy. At a median follow-up time of 25.6 months, perioperative EV + pembro significantly improved event free survival (EFS). The median EFS was not reached for the EV + pembro arm and 15.7 months for the control arm (HR 0.40, p<0.0001). 24-month EFS rates were 74.7% and 39.4% for the EV + pembro arm and control arm, respectively. 16.7% of patients in the control arm (29/174) received adjuvant nivolumab as of the data cutoff per local guidelines. The EFS benefits with EV + pembro was consistent across patient subgroup, regardless of age, ECOG status, or PD-L1 status. EV + pembro also significantly improved overall survival (OS). Median OS was not reached in EV + pembro arm and 41.7 months in the control arm (HR 0.50, p=0.0002). 24-months OS rates for the EV + pembro arm and the control arm were 79.7% and 63.1%, respectively. Similar to EFS, OS benefits with EV + pembro were consistently observed across patient subgroups. The pathological complete response (pCR) rate was 57.1% for the EV + pembro arm and 8.6% for the control arm (p<0.000001). EV + pembro was associated with EFS benefits among patients who did achieve a pCR and among patients who did not achieve a pCR. Safety signals in the EV + pembro arm were similar to past studies of EV + pembro. Adverse events (AEs) leading to surgical delay were low in the EV + pembro arm (4.0%), and AEs leading to death were similar between both arms (7.8% for EV + pembro and 5.7% for control). Treatment-emergent adverse events (TEAEs) during the surgical phase were also similar between both arms, and most AEs related to EV were grade 1 or 2. KEYNOTE-905 is the first phase 3 study to show improved efficacy outcomes with perioperative therapy for patients with MIBC who are ineligible for cisplatin. With its significant survival benefits and manageable safety profile, perioperative EV + pembro combined with RC + PLND may represent new a new standard of care for a patient population with a high unmet clinical need.
First and second interim analyses of ENGOT-ov65/KEYNOTE-B96: Pembrolizumab plus weekly paclitaxel with or without bevacizumab for platinum-resistant recurrent ovarian cancer
LBA3. Pembrolizumab vs placebo plus weekly paclitaxel ± bevacizumab in platinum-resistant recurrent ovarian cancer: Results from the randomized double-blind phase III ENGOT-ov65/KEYNOTE-B96 study
Nicoletta Colombo (European Institute of Oncology, IRCCS, Milan, Italy) presented results from ENGOT-ov65/KEYNOTE-B96, a phase 3 clinical trial investigating the addition of pembrolizumab to weekly paclitaxel with or without bevacizumab for platinum-resistant recurrent ovarian cancer (PRROC). Prior studies have indicated that paclitaxel may boost anti-cancer immunity by inducing cell death, and it has been hypothesized that combining pembrolizumab with weekly chemotherapy may enhance the anti-tumor response for PRROC and improve outcomes. 643 patients with PRROC who had received one or two prior lines of therapy, with at least one platinum-based chemotherapy, were randomized; 322 patients in the pembrolizumab arm received 18 cycles of pembrolizumab every six weeks and weekly doses of paclitaxel with or without biweekly bevacizumab. 321 patients in the placebo arm received 18 cycles of placebo every six weeks and weekly doses of paclitaxel with or without biweekly bevacizumab. Approximately 72% of patients in each arm had PD-L1-positive tumors (CPS >=1). At the first interim analysis (median follow-up 15.6 months), pembrolizumab significantly improved progression free survival (PFS) in patients with PD-L1-positive tumors. Median PFS was 8.3 months in the pembrolizumab arm, compared to 7.2 months in the placebo arm (HR 0.72, p=0.0014). Similar results were observed in the intent-to-treat (ITT) population, with median PFS of 8.3 months in the pembrolizumab arm and 6.4 months in the placebo arm (HR 0.70, p<0.0001). PFS benefits associated with pembrolizumab were observed across prespecified patient subgroups, and benefits were observed regardless of bevacizumab use. Results from the second interim analysis (median follow-up 26.6 months) confirmed these findings. Pembrolizumab was associated with extended PFS in the PD-L1-positive population (HR 0.75) and in the ITT population (HR 0.73). Median overall survival (OS) was significantly higher in the pembrolizumab arm (18.2 months) compared to the placebo arm (14.0 months; HR 0.76, p=0.0053) in the PD-L1 positive patient population, and OS benefits were observed across patient subgroups, regardless of age, prior use of PARP inhibitors, and the length of patients’ platinum free interval. Pembrolizumab was also associated with a higher objective response rate (ORR) in the PD-L1-positive patient population, with ORR of 53.0% in the pembrolizumab arm and 46.6% in the placebo arm. Duration of response was also longer for PD-L1 positive patients in the pembrolizumab arm compared to the placebo arm, with 18-month response rates of 26.5% and 14.5%, respectively. Adverse events (AEs) were numerically higher in pembrolizumab arm, but patients in this arm had longer treatment exposure compared to the placebo arm. Immune mediated AEs occurred more frequently in pembrolizumab arm, but most were grade 1 or 2. Immune-mediated AEs caused two deaths, one from colitis and one from pneumonitis, in the pembroluzimab arm. Most treatment-related adverse events in the two arms were related to chemotherapy. Pembrolizumab combined with weekly paclitaxel demonstrated significant and clinically meaningful improvements in PFS regardless of PD-L1 status and in OS among participants with PD-L1 expressing tumors. ENGOT-ov65/EYNOTE-B96 is the first phase 3 study to report significant OS improvements with an immune checkpoint inhibitor-based treatment regimen for patients with ovarian cancer. These data support pembrolizumab plus weekly paclitaxel as a new standard of care for patients with PRROC, and the United States Food and Drug Administration recently accepted this treatment regimen for priority review.