Benefit to Pembrolizumab Observed in Previously Treated Malignant Pleural Mesothelioma Irrespective of PD-L1 Status

Benefit to Pembrolizumab Observed in Previously Treated Malignant Pleural Mesothelioma Irrespective of PD-L1 Status

Systemic, Targeted, & Immune Therapies
Oct 29, 2021
Timothy A. Yap
Sanjay Popat
Dr. Timothy Yap and Dr. Sanjay Popat

Treatment options for patients with malignant pleural mesothelioma (MPM) are limited. While patients with advanced or inoperable MPM are traditionally treated with cisplatin and pemetrexed as a standard first-line treatment, there is still no established standard-of-care second-line therapy for these patients.1 Current second-line options after platinum-based chemotherapy have typically included either immune checkpoint inhibitors or chemotherapy options, such as vinorelbine or gemcitabine. On October 2, 2020, the U.S. Food and Drug Administration (FDA) approved the dual-immunotherapy combination of nivolumab plus ipilimumab as a first-line treatment for adult patients with unresectable MPM. This was followed by the approval of the same combination of immune checkpoint inhibitors by the European Medicines Agency’s Committee for Medicinal Products for Human Use on April 22, 2021. Given the latest drug approvals for MPM, the current implications for the rational sequencing of agents and optimal patient selection strategies remain to be defined.

Open questions include whether patients who do not have access to front-line nivolumab–ipilimumab should receive an immune checkpoint inhibitor upon relapse, and whether pembrolizumab would be an effective choice in that setting.

So, what do we know about the efficacy of the PD-1–inhibiting drug pembrolizumab for this population?

Expression of PD-L1 has been shown to range from 18% to 40% in patients with MPM, with relatively greater expression in non-epithelioid subtypes; however, the optimal PD-L1 assessment method (combined positive score or tumor propensity score) and cutoff remain uncertain. Two signal-seeking clinical trials have explored the antitumor efficacy and safety profile of pembrolizumab monotherapy in previously treated patients with MPM,2 ,3 and results from these studies have led to a small, randomized phase III trial of pembrolizumab monotherapy versus standard chemotherapy monotherapy.

The KEYNOTE-028 phase Ib basket trial investigated the antitumor response and safety profile of pembrolizumab monotherapy in previously treated MPM patients selected by PD-L1 status using a combination positive score.2 Results from a cohort of 25 patients whose MPM was PD-L1-positive showed an objective response rate (ORR) of 20%, median duration of 12 months (95% CI, 3.7–not reached), PFS of 5.5 months, and median OS of 18.7 months, and the suggestion of a tail for OS, with two patients completing 2 years of treatment. Although the results of KEYNOTE-028 demonstrated promising antitumor activity of pembrolizumab monotherapy in a biomarker-selected population, no conclusions regarding the predictive potential of PD-L1 expression were obtained because the eligibility criteria did not include PD-L1-negative patients. 

To assess whether patient selection on the basis of PD-L1 expression enriches for antitumor response, the phase II KEYNOTE-158 study enrolled 118 patients with MPM, irrespective of PD-L1 status, from 35 academic facilities and community-based institutions across 14 countries.3 The entire cohort had received at least one line of standard-of-care systemic therapy before entering the study, and 65% of the participants had PD-L1-positive tumors. The antitumor efficacy and safety outcomes of this cohort provided evidence that previously treated patients with advanced MPM could benefit from pembrolizumab with manageable toxicity and responses lasting more than 1 year, irrespective of PD-L1 status. At the time of data cutoff, 16% of patients (n = 17) had experienced a reduction of at least 30% in target lesions after receiving a median of 6 doses of pembrolizumab (range 1–35, interquartile range [IQR] 3–14), with a median treatment duration of 3.5 months (range 1–35, IQR 3-14). The ORR and median duration of the response were estimated to be 8% and 14.3 months, respectively. Importantly, the objective responses observed in this cohort occurred irrespective of PD-L1 status: an ORR of 8% (95% CI, 3–16) with a median response duration of 17.7 months was observed in patients with PD-L1-positive MPM, while an ORR of 13% (95% CI, 4–30) with a median response duration of 10.2 months was observed in patients who had PD-L1-negative MRM. The median PFS and median OS were 2.1 months (95% CI, 2.1–3.9) and 10 months (95% CI, 7.6–13.4), respectively. It was estimated that, 12 months after starting pembrolizumab treatment, at least 60% of the responses were ongoing and approximately 45% of the patients were alive.3

Due to differences in design, patient characteristics, and length of follow-up, it is difficult to define the role of PD-L1 status and compare the efficacy outcomes of trials that included patients with MPM who received pembrolizumab monotherapy after previous treatment. For instance, although antitumor activity was observed regardless of PD-L1 status in the KEYNOTE-158 study, the 95% CI between the subsets of patients with differing PD-L1 statuses overlaps; hence, no definite conclusion can be drawn from that study regarding the role of PD-L1 status. However, two studies investigating antitumor responses of single-agent nivolumab in patients with pretreated MPM reported objective response rates of 24% to 29%, regardless of PD-L1 status (p > 0.05). 

In parallel, the PROMISE-Meso academic, randomized, phase III trial compared pembrolizumab monotherapy with single-agent chemotherapy (vinorelbine or gemcitabine) in 144 previously treated patients with MPM. The trial was ambitiously designed to identify an improved PFS, with a cross-over at progression built in. This trial reported objective responses in 22% of the patients in the pembrolizumab group (16/73) and in 6% of the patients in the chemotherapy group (4/71). Yet, the primary outcome was not met because the median PFS rates for both groups were similar4 : 2.5 months in the pembrolizumab group versus 2.4 months in the chemotherapy group (HR 1.06; 95% CI, 0.73–1.53; p = 0.76). Multivariable analyses demonstrated no obvious difference in benefit between histologic subtypes and suggested poorer OS for non-epithelioid histology. Translational analyses of OS by PD-L1 status on surplus diagnostic biopsies demonstrated no predictive utility.

Recently, the CONFIRM phase III trial randomly assigned patients with MPM, most of whom had already received first- and second-line treatments, to either supportive care or supportive care with nivolumab, with results that reaffirmed the role of PD-1 inhibition.5 Here, a clear and important OS benefit for nivolumab over no active treatment was demonstrated (median 9.2 months vs. 6.6 months; HR 0.72; p = 0.02), underscoring the benefit of PD-1 inhibition in MPM. An OS benefit was restricted to the epithelioid subtype, with no benefit for non-epithelioid histology, which is in line with the results from PROMSE-Meso trial. Importantly, as observed in PROMISE-Meso, PD-L1 status from archived specimens did not predict survival. 

The role of CTLA-4 inhibition combined with PD-1 inhibition was explored in a non-comparative randomized phase II trial with parallel arms that was conducted in France.6 In this trial, 125 patients with relapsed MPM were randomly assigned to receive nivolumab or nivolumab–ipilimumab. Responses were observed in 19% (10 of 54) of patients who received nivolumab and 28% (15 of 54) of patients who received nivolumab–ipilimumab, with median PFS reported at 4 months and 5.6 months, respectively. Here, exploratory analyses of PD-L1 expression suggested a relationship between PD-L1 positivity (at the 1% cut point) and response. In fact, multivariable analyses suggested a predictive utility for PD-L1 for nivolumab, but not for nivolumab–ipilimumab. Similar to PROMISE-Meso and CONFIRM, this study suggested that OS was poorer for patients with non-epithelioid histology who received nivolumab but was in fact improved for patients in that population who received  nivolumab–ipilimumab. A separate single-arm phase II trial conducted in the Netherlands that evaluated nivolumab–ipilimumab in relapsed MPM demonstrated responses in 38% of participants (13 of 34). In this trial, pre-treatment biopsies were performed, and PD-L1 expression status was significantly associated with response at 12 weeks.7 Taken together, results from these trials suggest that the PD-1/PD-L1 pathway is a potentially effective therapeutic target for patients with previously treated MPM, although the studies offer conflicting evidence on the predictive utility of PD-L1 expression. 

In the first-line space, the registrational CHECKMATE-743 trial randomly assigned patients to a front-line regimen of either nivolumab–ipilimumab or platinum–pemetrexed chemotherapy. This trial demonstrated significant and consistent survival across both epithelioid and non-epithelioid subtypes with a median OS of around 18 months versus 14 months in the chemotherapy arm; it also demonstrated a markedly poorer chemotherapy efficacy in non-epithelioid subtypes.8 Analyses by PD-L1 subtype were exploratory and unpowered, and while only demonstrating an OS benefit in the PD-L1-positive population (1% tumor staining was the cut point), PD-L1 was not a stratification factor and outcomes were therefore likely confounded by hidden covariates. 

Based on what we know today, should all immunotherapy-eligible patients receive nivolumab–ipilimumab? 

The CHECKMATE-743 study was only powered to investigate the OS difference in the intent-to-treat  population between nivolumab-ipilimumab and chemotherapy, and not between histologic or other subgroups such as PD-L1 status. For non-epithelioid subtypes, the trial clearly demonstrated a large OS benefit for nivolumab–ipilimumab with an OS HR of 0.46 translating to a median survival of 8.8 months for patients who received chemotherapy versus 18.1 months for patients who received immunotherapy. For epithelioid subtypes, an OS benefit was still trending (HR 0.86; 95% CI, 0.69–1.08), at a median follow-up of 29.7 months. The 2-year landmark currently favors immunotherapy (42% vs. 33%). With greater follow-up, this benefit may become more apparent in both histologic subgroups. 

Taking all of the available data together, it is apparent that PD-1 inhibition is an important strategy in MPM, with the best outcomes observed in the front line with the combination of nivolumab–ipilimumab. As in small cell lung cancer, trial data suggest that up-front use of PD-1 inhibition is a superior strategy than its use as salvage therapy in patients with relapsed disease, but PD-L1 status remains uncertain as a method for patient selection. For non-epithelioid subtypes, CHECKMATE-743 has highlighted front-line chemotherapy’s limited activity and underscores that the greatest OS benefit is associated with front-line use of the nivolumab–ipilimumab combination. However, in the second-line setting and beyond there is a suggestion of relatively less efficacy for pembrolizumab or nivolumab monotherapy compared with chemotherapy or no chemotherapy from PROMISE-meso and CONFIRM; hence, the combination of nivolumab–ipilimumab is an alternative, particularly in patients with no prior exposure to checkpoint inhibitors. For epithelioid subtypes, the front-line combination of nivolumab–ipilimumab is often a preferred option where available, but when it is not, either single-agent pembrolizumab or nivolumab or the combination of nivolumab–ipilimumab are alternatives at relapse. 

Given the uncertain role of PD-L1 status in patient selection, predictive biomarkers of response that can effectively identify the optimal patient population that may benefit from single or dual immune checkpoint inhibitor therapy have not yet been established and are urgently required. Due to the early crossing of PFS curves observed in CHECKMATE-743, we await the results of three important, ongoing, prospective randomized front-line chemotherapy–immunotherapy combination trials (IND227, BEAT-MESO, and DREAM3R), which are evaluating the role of chemotherapy–pembrolizumab, chemotherapy–atezolizumab–bevacizumab, and chemotherapy–durvalumab combinations, respectively. These trials, along with their biomarker data, will undoubtedly add to our understanding of PD-1 inhibition in MPM. Now that it has dawned, this new era for mesothelioma drug treatment is expected to continue as immunotherapy retains a role as the backbone of future drug strategies.


  • 1. Yap TA, Aerts JG, Popat S, Fennell DA. Novel insights into mesothelioma biology and implications for therapy. Nat Rev Cancer. 2017;17(8):475-488.
  • 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.
  • 3. a. b. c. Yap TA, Nakagawa K, Fujimoto N, et al. Efficacy and safety of pembrolizumab in patients with advanced mesothelioma in the open-label, single-arm, phase 2 KEYNOTE-158 study. Lancet Respir Med. 2021;9(6):613-621.
  • 4. Popat S, Curioni-Fontecedro A, Dafni U, et al. A multicentre randomised phase III trial comparing pembrolizumab versus single-agent chemotherapy for advanced pre-treated malignant pleural mesothelioma: the European Thoracic Oncology Platform (ETOP 9-15) PROMISE-meso trial. Ann Oncol. 2020;31(12):1734-1745.
  • 5. Fennel D, Ottensmeier C, Califano R, et al. Nivolumab Versus Placebo in Relapsed Malignant Mesothelioma: The CONFIRM Phase 3 Trial. J Thorac Oncol. 2021;16(3):S62.
  • 6. Scherpereel A, Mazieres J, Greillier L, et al. Nivolumab or nivolumab plus ipilimumab in patients with relapsed malignant pleural mesothelioma (IFCT-1501 MAPS2): a multicentre, open-label, randomised, non-comparative, phase 2 trial. Lancet Oncol. 2019;20(2):238-253.
  • 7. Disselhorst MJ, Quispel-Janssen J, Lalezari F, et al. Ipilimumab and nivolumab in the treatment of recurrent malignant pleural mesothelioma (INITIATE): results of a prospective, single-arm, phase 2 trial. Lancet Respir Med. 2019;7(3):260-270.
  • 8. Baas P, Scherpereel A, Nowak AK, et al. First-line nivolumab plus ipilimumab in unresectable malignant pleural mesothelioma (CheckMate 743): a multicentre, randomised, open-label, phase 3 trial. Lancet. 2021;397(10272):375-386.


About the Authors

Timothy Yap

Timothy A. Yap

Associate Professor in the Department for Investigational Cancer Therapeutics (Phase I Program) and the Department of Thoracic/Head and Neck Medical Oncology, as well as the Medical Director of the Institute for Applied Cancer Science
Dr. Yap is an associate professor in the Department for Investigational Cancer Therapeutics (Phase I Program) and the Department of Thoracic/Head and Neck Medical Oncology, as well as the Medical Director of the Institute for Applied Cancer Science and the Associate Director of Translational Research in the Institute for Personalized Cancer Therapy at The University of Texas MD Anderson Cancer Center.
Dr. Sanjay Popat

Sanjay Popat

Consultant Thoracic Medical Oncologist; Professor of Thoracic Oncology
Dr. Popat is a consultant thoracic medical oncologist at The Royal Marsden, London, UK; and professor of Thoracic Oncology at the Institute of Cancer Research, London, UK.