Perioperative nivolumab or nivolumab plus ipilimumab in resectable diffuse pleural mesothelioma: a phase 2 trial and ctDNA analyses (Neoadjuvant Immune Checkpoint Blockade in Resectable Malignant Pleural Mesothelioma)
Phase 2N = 30NCT03918252
- Population
- 30 adults with previously untreated, resectable epithelioid or biphasic diffuse pleural mesothelioma and ECOG performance status 0 or 1
- Design
- Prospective, multicentre (three US centres), multiarm, non-comparative phase 2 trial (NCT03918252). Patients with previously untreated, resectable epithelioid or biphasic diffuse pleural mesothelioma were enrolled sequentially, not randomised, to neoadjuvant nivolumab 240 mg every 2 weeks for three cycles (Arm A, 16 patients) or nivolumab 3 mg/kg every 2 weeks for three cycles plus ipilimumab 1 mg/kg on cycle 1 (Arm B, 14 patients), followed by surgery, optional adjuvant platinum-pemetrexed chemotherapy and/or radiotherapy, and adjuvant nivolumab 480 mg every 4 weeks for up to one year. Enrolment 29 October 2019 to 22 January 2024. Co-primary endpoints feasibility (proceeding to surgery without a treatment-related delay of more than 24 days from the preplanned date; hypothesis 80% or more) and safety (dose-limiting toxicities); progression-free and overall survival and ctDNA analyses were exploratory. Complete macroscopic resection in 12 (Arm A) and 11 (Arm B) patients; among them extended pleurectomy decortication was performed in 10 of 12 (Arm A) and 9 of 11 (Arm B) and extrapleural pneumonectomy in 2 of 12 and 2 of 11. Median follow-up (intention-to-treat population) 43.2 months in Arm A and 24.1 months in Arm B. Not powered for efficacy.
- Primary endpoint
- Proceeded to surgery without treatment-related delay, Dose-limiting toxicity rate
- Arms
- Arm A: nivolumab (experimental, n = 16): Neoadjuvant nivolumab 240 mg every 2 weeks for three cycles, then surgery, optional adjuvant chemotherapy and/or radiotherapy, and adjuvant nivolumab
- Arm B: nivolumab plus ipilimumab (experimental, n = 14): Neoadjuvant nivolumab 3 mg/kg every 2 weeks for three cycles plus ipilimumab 1 mg/kg on cycle 1, then surgery, optional adjuvant chemotherapy and/or radiotherapy, and adjuvant nivolumab
- People
- Reuss JE, senior author Forde PM
Show the numbers
| Endpoint | Analysis | Arm A: nivolumab / Arm B: nivolumab plus ipilimumab | Control | Effect |
|---|---|---|---|---|
| Proceeded to surgery without treatment-related delayAll 30 enrolled patients (Arm A 16, Arm B 14) | primaryReuss 2025 | 81.3% | — | 13 of 16 and 12 of 14 patients. In the full text, feasibility means proceeding to surgery without a treatment-related delay of more than 24 days from the preplanned date, with the hypothesis that 80% or more would. Arms were enrolled one after the other and are not compared. Disease progression precluded surgery in three (Arm A) and two (Arm B) patients. |
| Dose-limiting toxicity rateAll 30 enrolled patients (Arm A 16, Arm B 14) | primaryReuss 2025 | 6.3% (0.2–30.2%) | — | One dose-limiting toxicity in each arm (pneumonitis in Arm A; raised alanine aminotransferase in Arm B), neither precluding resection. Grade 3 or higher immune-checkpoint-related adverse events occurred in 18.8% (Arm A) and 7.1% (Arm B) of patients. |
| PFSIntention-to-treat, each arm separately | exploratoryReuss 2025 | median 9.6 months (2.5–27.7) | — | Exploratory, in a trial not powered for efficacy. Arm B 95% CI is 7.1 months to not reached (upper limit not a number, so not entered). Arms were enrolled sequentially, differ in follow-up and baseline characteristics (for example 81% men in Arm A versus 57% in Arm B), and are not compared. Counted from the start of neoadjuvant therapy, not from randomisation (the arms were not randomised). |
| OSIntention-to-treat, each arm separately | exploratoryReuss 2025 | median 19.3 months (14.9–34.7) | — | Exploratory, in a trial not powered for efficacy. Arm B 95% CI is 20.4 months to not reached (upper limit not a number, so not entered). At data lock one patient was alive in Arm A and eight of 14 in Arm B. Arms were enrolled sequentially and are not compared. Counted from the start of neoadjuvant therapy, not from randomisation (the arms were not randomised). |
| Endpoint | Analysis | Arm A: nivolumab / Arm B: nivolumab plus ipilimumab | Control | Effect |
|---|---|---|---|---|
| Proceeded to surgery without treatment-related delayAll 30 enrolled patients (Arm A 16, Arm B 14) | primaryReuss 2025 | 81.3% | — | 13 of 16 and 12 of 14 patients. In the full text, feasibility means proceeding to surgery without a treatment-related delay of more than 24 days from the preplanned date, with the hypothesis that 80% or more would. Arms were enrolled one after the other and are not compared. Disease progression precluded surgery in three (Arm A) and two (Arm B) patients. |
| Dose-limiting toxicity rateAll 30 enrolled patients (Arm A 16, Arm B 14) | primaryReuss 2025 | 6.3% (0.2–30.2%) | — | One dose-limiting toxicity in each arm (pneumonitis in Arm A; raised alanine aminotransferase in Arm B), neither precluding resection. Grade 3 or higher immune-checkpoint-related adverse events occurred in 18.8% (Arm A) and 7.1% (Arm B) of patients. |
| PFSIntention-to-treat, each arm separately | exploratoryReuss 2025 | median 9.6 months (2.5–27.7) | — | Exploratory, in a trial not powered for efficacy. Arm B 95% CI is 7.1 months to not reached (upper limit not a number, so not entered). Arms were enrolled sequentially, differ in follow-up and baseline characteristics (for example 81% men in Arm A versus 57% in Arm B), and are not compared. Counted from the start of neoadjuvant therapy, not from randomisation (the arms were not randomised). |
| OSIntention-to-treat, each arm separately | exploratoryReuss 2025 | median 19.3 months (14.9–34.7) | — | Exploratory, in a trial not powered for efficacy. Arm B 95% CI is 20.4 months to not reached (upper limit not a number, so not entered). At data lock one patient was alive in Arm A and eight of 14 in Arm B. Arms were enrolled sequentially and are not compared. Counted from the start of neoadjuvant therapy, not from randomisation (the arms were not randomised). |
- Arm A: nivolumab (experimental)
- Arm B: nivolumab plus ipilimumab (experimental)
Median-only curves assume a constant hazard (exponential) and fit plateaus, as seen with some immunotherapies, poorly.
Data table (modelled values)
| Months | Arm A: nivolumab | Arm B: nivolumab plus ipilimumab |
|---|---|---|
| 0 | 100.0% | 100.0% |
| 12 | 42.0% | 65.7% |
| 25 | 16.4% | 41.7% |
| 37 | 6.9% | 27.4% |
| 50 | 2.7% | 17.4% |
- Arm A: nivolumab (experimental)
- Arm B: nivolumab plus ipilimumab (experimental)
Median-only curves assume a constant hazard (exponential) and fit plateaus, as seen with some immunotherapies, poorly.
Data table (modelled values)
| Months | Arm A: nivolumab | Arm B: nivolumab plus ipilimumab |
|---|---|---|
| 0 | 100.0% | 100.0% |
| 18 | 52.4% | 64.6% |
| 36 | 27.4% | 41.8% |
| 54 | 14.4% | 27.0% |
| 72 | 7.5% | 17.5% |