Pembrolizumab Misses PFS Endpoint in Relapsed Malignant Pleural Mesothelioma

Pembrolizumab (Keytruda) did not show an improvement in progression-free survival compared with chemotherapy in patients with malignant pleural mesothelioma.

Sanjay Popat, BSc MBBS FRCP PhD

Pembrolizumab (Keytruda) did not show an improvement in progression-free survival (PFS) compared with chemotherapy in patients with malignant pleural mesothelioma (MPM), missing the primary endpoint of the European Thoracic Oncology Platform (ETOP 9-15) PROMISE-meso trial.1

The phase III findings, which were presented at the 2019 ESMO Congress, showed that the median PFS was 2.5 months for pembrolizumab compared with 3.4 months for chemotherapy (HR, 1.06; 95% CI, 0.73-1.53; P = .76). The 6-month PFS rates were 25.0% and 27.4% with pembrolizumab and chemotherapy, respectively.

“The trial did not meet the primary endpoint of improving PFS with pembrolizumab over single-agent chemotherapy in PD-L1 unselected patients,” said lead study author Sanjay Popat, BSc MBBS FRCP PhD, consultant medical oncologist at The Royal Marsden NHS Foundation Trust, in a presentation during the meeting. “Despite correcting for crossover, an overall survival benefit was not observed. Nevertheless, pembrolizumab was associated with a significantly improved objective response rate.”

MPM is an aggressive cancer with increasing mortality rates and a poor prognosis. Upon relapse on platinum-based chemotherapy, there are no treatment options that have demonstrated an improvement in survival. However, gemcitabine and vinorelbine are commonly used options.

Prior single-arm data with checkpoint inhibitors have demonstrated encouraging signals in patients with relapsed MPM. For example, an expansion cohort from the phase I KEYNOTE-028 trial showed a 20% objective response rate (ORR) and a median PFS of 5.4 months with pembrolizumab.2

In the phase III study, 144 patients with relapsed MPM were randomized 1:1 to receive 200 mg of fixed-dose pembrolizumab intravenously on day 1 of each 3-week cycle (n = 73) or physician’s choice chemotherapy of gemcitabine at 1000 mg/m2 on days 1 and 8 every 3 weeks, vinorelbine at 30 mg/m2 on days 1 and 8 every 3 weeks, or vinorelbine at 60 or 80 mg/m2 on days 1 and 8 every 3 weeks (n = 71).

Treatment was administered until disease progression by RECIST1.1 criteria up to 2 years, and patients were assessed every 9 weeks for 6 months, followed by every 12 weeks thereafter. Upon disease progression, those in the chemotherapy arm were permitted to cross over to receive the PD-1 inhibitor.

To be eligible for enrollment, patients with MPM had to have progressed following platinum-based chemotherapy; had to have an ECOG performance status of 0 or 1; measurable or evaluable disease according to RECIST 1.1 criteria; adequate hematologic, renal, and liver function; and the availability of tumor tissue for translational research.

Patients were stratified by epithelioid subtype or non-epithelioid subtype. The primary endpoint of the trial was PFS as assessed by blinded independent central review; secondary endpoints were ORR, time to treatment failure (TTF), overall survival (OS), investigator-assessed PFS, and adverse events (AEs). Outcomes by PD-L1 status served as a correlative endpoint.

Investigators had a target median PFS of 6 months for pembrolizumab for 80% power at a one-sided significance level of 2.5%; the targeted hazard ratio was 0.58. The data cutoff date was February 20, 2019.

Two patients on the pembrolizumab arm withdrew from treatment as did 1 patient on the chemotherapy arm. As of the data cutoff date, 8 patients remained on pembrolizumab and 5 were still receiving chemotherapy. The most common reasons for withdrawal on the pembrolizumab arm were disease progression (n = 56), death (n = 5), toxicity (n = 1), patient decision (n = 1), and other (n = 2). Sixty-three percent of patients crossed over to receive pembrolizumab upon disease progression (n = 45).

Baseline characteristics were similar between arms. The median age was 70 years, most patients had epithelioid histology (88.9%), and the majority had an ECOG performance status of 1 (75.1%). Prior therapies included carboplatin/pemetrexed, cisplatin/pemetrexed, platinum or without pemetrexed, and cisplatin/pemetrexed plus carboplatin/pemetrexed.

At a median follow-up of 11.8 months, the median OS was also not improved with pembrolizumab at 10.7 months versus 11.7 months with chemotherapy (HR, 1.04; 95% CI, 0.66-1.67; P = .85). Furthermore, 6-month OS rates were 72.9% with chemotherapy and 68.5% with pembrolizumab. Even when censoring for crossover, there was no OS benefit observed with pembrolizumab (HR, 1.44; 95% CI, 0.77-2.67; P = .25), nor for inverse probability weighting (HR, 1.07; 95% CI, 0.67-1.71; P = .79).

However, the ORR was 22% with pembrolizumab compared with 6% in those who received chemotherapy (P = .004). The 22% ORR consisted of 16 partial responses (PRs), 17 cases of stable disease (SD), 33 cases of progressive disease (SD), and 7 cases that were not evaluable (NE). In the chemotherapy arm, there were 4 cases of PRs, 23 SDs, 35 PD, and 9 that were NE.

The median DOR was 4.6 months and 11.2 months with pembrolizumab and chemotherapy, respectively (95% CI, 2.2-10.3; 95% CI, 6.2-15.3).

There did not appear to be a patient subgroup who experienced a PFS or OS benefit from pembrolizumab or chemotherapy, except for those with non-epithelioid histology who did worse on pembrolizumab (HR for PFS, 1.76; 95% CI, 0.58-5.33; HR for OS, 1.34; 95% CI, 0.38-4.74).

When stratifying by PD-L1 status, the median TTF was slightly improved with pembrolizumab at 2.8 months compared with 2.3 months for chemotherapy. However, PFS was similar in both PD-L1—positive and –negative subsets.

For those with a tumor proportion score (TPS) <1%, the median PFS was 4.2 months and 4.4 months with pembrolizumab and chemotherapy, respectively (HR, 1.26; 95% CI, 0.56-2.83; P = .57). The 6-month PFS rates were also higher with chemotherapy in this group at 46.3% versus 34.0% with pembrolizumab.

However, median OS was higher pembrolizumab in patients with PD-L1 TPS <1% at 11.7 months versus 9.9 months with chemotherapy (HR, 0.72; 95% CI, 0.26-2.00; P = .53). The 6-month PFS rates were 78.9% and 70.6% with pembrolizumab and chemotherapy, respectively.

In patients with TPS ≥1%, the median PFS was 3.2 months in both arms (HR, 1.06; 95% CI, 0.63-1.80; P = .82).

In the TPS ≥1% group, the median OS was 10.7 months with pembrolizumab and not reached with chemotherapy (HR, 1.47; 95% CI, 0.69-3.11; P = .32). The 6-month OS rates were 71.9% and 76.5% for pembrolizumab and chemotherapy, respectively.

Regarding safety, pembrolizumab’s profile was found to be similar to what has been previously reported. Treatment-related adverse events (TRAEs) occurred in 69.4% of pembrolizumab-treated patients and 72.9% of those on chemotherapy, and grade 3/5 AEs were 19.4% and 24.3%, respectively. TRAEs leading to treatment discontinuation were 8.3% for pembrolizumab and 7.1% for chemotherapy; those that led to death occurred in 1.4% of patients in each arm, with dyspnea with disease progression observed as the primary cause of death.

Exploratory translational research is ongoing to identify patient subgroups that could benefit from pembrolizumab, Popat concluded.


  1. Popat S, Cunoni-Fontecedro A, Polydoropoulou V, et al. A multicenter randomized phase III trial comparing pembrolizumab (P) vs single agent chemotherapy (CT) for advanced pre-treated malignant pleural mesothelioma (MPM)—results from the European Thoracic Oncology Platform (ETOP 9-15) PROMISE-meso trial. Presented at: 2019 ESMO Congress; September 27 to October 1, 2019. Abstract LBA91.
  2. Alley EW, Lopez J, Santoro A, et al. Clinical safety and activity of pembrolizumab in patients with malignant pleural mesothelioma (KEYNOTE-028): preliminary results from a non-randomised, open-label, phase 1b trial. Lancet Oncol. 2017;18(5):623-630. doi: 10.1016/S1470-2045(17)30169-9.

<<< View more from the 2019 ESMO Congress