In this article, Priyanka Bhateja, MD, and Neelesh Sharma, MD, PhD review the completed trials and ongoing clinical development of pembrolizumab in NSCLC.
Recent approval of immune checkpoint inhibitors for the treatment of melanoma and lung cancer has generated a new excitement in the field of cancer therapeutics. PD-1/ PD-L1 pathway is an important regulator of immune tolerance in the tumor microenvironment. Pembrolizumab is highly selective, humanized monoclonal IgG4-kappa antibody against PD-1 receptor that promotes anti-tumor immune response by preventing interaction of PD-1 with its ligands PD-L1 and PD-L2. The FDA granted accelerated approval for Keytruda (pembrolizumab) in October 2015 to treat patients with advanced nonsmall-cell lung cancer (NSCLC) whose disease has progressed after other treatments and with tumors that express PD-L1 as determined by a companion diagnostic, the PD-L1 IHC 22C3 pharmDx test. This approval was based on the demonstration of durable objective response rate and acceptable safety in the NSCLC subgroup of a large phase I clinical trial (Keynote-001). Subsequently, the Keynote-010 study also demonstrated overall survival benefit in a randomized study comparing pembrolizumab to docetaxel in PD-L1 positive previously treated advanced NSCLC. In this article, we review the completed trials and ongoing clinical development of pembrolizumab in NSCLC.
Lung cancer is the leading cause of cancer death world-wide.1-3An estimated 221,200 new cases of lung cancer were expected to be diagnosed in the United States in 2015 with 158,040 deaths estimated.2There have been advances in treatment of the subset of patients with driver mutations such as epidermal growth factor receptor (EGFR) and anaplastic lymphoma kinase (ALK) rearrangement. Despite these advances, the 5-year survival for patients with stage IV (metastatic) disease remains dismal. Platinum doublet therapy continues to be the standard first-line treatment for advanced nonsmall-cell lung cancer (NSCLC) patients without driver mutations.4In 2015, immune checkpoint inhibitors were added to the armamentarium for treatment of NSCLC.
Cancer cells multiply by evading immune response. One such mechanism of interest is upregulation of pro- grammed cell death protein ligands (PDL-1 and PDL-2) on tumor cells and tumor-invading inflammatory cells that bind to PD-1 receptors on cytotoxic T lymphocytes causing down regulation of the immune response.5-11Pembrolizumab (Keytruda) is a highly selective, humanized monoclonal IgG4-kappa isotype antibody against PD-1. It was developed by grafting a high affinity mouse antihuman PD-1 antibody into a human IgG4 with a stabilizing S228 P Fc alteration. It disrupts the binding of PD-1 with its ligands (PDL-1 and PDL-2) causing recognition of tumor cells by cytotoxic T cells.
Keynote-001: Phase I Trial in Advanced Treatment Refractory NSCLC
Keynote-001 trial is a landmark phase I trial that was granted breakthrough therapy designation in October 2014 and accelerated FDA approval in October 2015 for pembrolizumab as a second line treatment for patients with NSCLC.12It is approved after progression of disease on or after platinum-based chemotherapy in NSCLC (EGFR receptor mutation negative and no ALK rearrangement) whose tumors exhibit high-level PD-L1 expression (greater than 50% based on 22C3 IHC assay). The dose approved by FDA is 2mg/kg every 3 weeks until disease progression or unacceptable toxicity.13In this Phase I trial, 495 patients were assigned to receive pembrolizumab at a dose of 2mg or 10mg every 3 weeks or 10mg every 2 weeks. The adverse effects didn’t vary according to dose or schedule and included fatigue, decreased appetite, and pruritis. The overall incidence of pneumonitis was less than 4%. The overall response rate (ORR) evaluated by ECIST criteria was 19.4% (95% confidence interval [CI], 16, and 23.2) and median duration of response was 12.5 months with a median OS of 12 months. The ORR in previously treated patients was 18% (95% CI 14.4 to 22.2) and 24.8% (95% CI 16.7 to 34.3) in previously untreated patients. The response rate was similar across different doses, schedules, or histologic subtypes. Current or former smokers had ORR of 22.5% and never smokers had an ORR of 10.3%.
Keynote 010: Randomized Phase II/III Trial in Previously Treated NSCLC
Keynote 010 was a recent randomized, multicentre, open label phase II/III trial.14In this trial, previously treated NSCLC patients with PDL-1 expression on at least 1% of tumor cells were randomized to receive pembrolizumab 2mg/kg, pembrolizumab 10mg/kg, or docetaxel 75mg/m2 every 3 weeks. The primary end- points were overall survival (OS) and progression-free survival (PFS) for total population and in patients with PD-L1 expression on at least 50% of tumor cells. The median OS was 10.4 and 12.7 months for pembrolizumab 2mg/kg and 10mg/kg respectively and 8.5 months for the docetaxel arm. OS was significantly longer for pembrolizumab 2mg/kg versus docetaxel [Hazard ratio (HR) 0.71, 95% CI 0.58-0.88;P= 0.0008] and pembrolizumab 10mg/kg versus docetaxel (HR 0·61, 95% CI 0·490·75;P<0·0001). Median PFS for unselected patients was not significantly different in the three arms. Patients with at least 50% of tumor cells showing expression of PDL-1 had significantly higher OS with pembrolizumab 2 mg/kg than with docetaxel (median 14·9 months vs 8·2 months; HR 0·54, 95% CI 0·380·77;P= 0·0002) and with pembrolizumab 10 mg/kg than with docetaxel (17·3 months vs 8·2 months; 0·50, 0·36 0·70;P<0·0001). In this subgroup of patients with high PDL-1 expression, PFS was also significantly longer with pembrolizumab 2 mg/kg than with docetaxel (median 5·0 months vs 4·1 months; HR 0·59, 95% CI 0·44 0·78;P= 0·0001) and with pembrolizumab 10 mg/kg than with docetaxel (5·2 months vs 4·1 months; 0·59, 0·450·78;P<0·0001). Pembrolizumab had lower incidence of grade 35 adverse effects (AEs) compared to docetaxel (13% and 16% respectively in pembrolizumab 2 mg/kg arm and 10 mg/kg arm, and 35% in docetaxel arm). Both doses of pembrolizumab showed better OS with a more favorable adverse effect profile compared to docetaxel. The benefit seemed to be magnified for patients with more than 50% of tumor cells showing PDL-1 expression and translated into a significant difference in PFS in this subset.
First-Line Single Agent Trials
Two large international phase III trials, Keynote-02415 and Keynote-042,16are comparing pembrolizumab to standard of care platinum-based chemotherapy in the first-line setting. Both these trials are randomizing patients with previously untreated NSCLC without EGFR mutations or ALK translocation with tumors showing PDL-1 expression (≥1%) to receive pembrolizumab (200mg fixed dose every 3 weeks) or platinum-based doublet therapy (carboplatin/cisplatin plus paclitaxel/gemcitabine or carboplatin plus premetrexed for non-squamous histology). The Keynote-024 trial is accruing patients with high level PDL-1 expression (≥50%) only, whereas the Keynote-042 will enroll patients with ≥1% PD-L1 expression and stratify them based on weak (1- 49%) or strong (≥50%) PDL-1 expression. Chemotherapy will be given for maximum of 6 cycles and may be followed by optional pemetrexed maintenance in patients with non-squamous histology. The patients randomized to Pembrolizumab will be treated for up to 35 cycles or disease progression. Crossover from standard of care platinum doublet to pembrolizumab is allowed for disease progression.
The Keynote-024 trial will continue until approximately 300 patients with metastatic NSCLC are randomized. The primary endpoint is PFS; secondary endpoints are ORR, OS, and safety. The primary endpoint for Keynote-042 trial is OS in the PD-L expression ≥50% stratum and secondary endpoints include PFS for patients with PD-L expression ≥50% and PFS and OS for all the patients. The trial will continue until the target of 1240 patients is reached. The primary endpoint is OS, and the secondary endpoint is PFS.
Combination Trials with Chemo, Immuno, or Targeted Therapy
Keynote 021 is multicohort phase I/II study of pembrolizumab in combination with chemotherapy or immunotherapy in patients with stage IIIB or metastatic NSCLC.17-18Cohorts A and C are evaluations of pembrolizumab plus platinum-based doublet chemotherapy, cohort E and F are evaluations of pembrolizumab and erlotinib or geftinib, and cohort D is evaluation of pembrolizumab and ipilimumab. The results from cohort A and C and those of Cohort D were reported in two separate abstracts at ASCO 2015. Treatment-naive NSCLC patients were randomized 1:1 to receive pembrolizumab (2mg/kg or 10mg/kg every 3 weeks) plus carboplatin (AUC of 6) and paclitaxel (200mg/m2) (cohort A = 20 patients) or carboplatin (AUC of 5) and premetrexed (500mg/m2) (Cohort C = 24 patients with nonsquamous histology). Four cycles of combination chemo/immunotherapy was followed by pembrolizumab monotherapy or maintenance therapy with pemetrexed and pembrolizumab. The ORR was 28% in cohort A and 58% in cohort C. The disease control rate with pembrolizumab of 10mg/kg was 80% in cohort A and 100% in cohort C, and with Pembrolizumab 2mg/ kg was 60% in cohort A and 92% in cohort C. One dose-limiting toxicity: Grade 3 rash requiring hospitalization and drug discontinuation was reported in cohort C with pembrolizumab dose of 10mg/kg. Grade 3-4 AEs were reported in 32% patients in cohort A and 38% patients in cohort C. Immune-related AEs were reported in 2 (8%) patients in cohort A (one case each of grade 3 rash and grade 2 colitis) and 4 (17%) in cohort C (one case each of grade 3 colitis, grade 2 hypothyroidism, and grade 1 colitis and hyperthyroidism). Based on its acceptable safety profile and encouraging clinical activity, the combination of pembrolizumab, carboplatin, and pemetrexed is being evaluated in larger cohort. Efficacy was not significantly different across different pembrolizumab dose levels, suggesting a flat dose/exposure relationship. Phase II cohorts will examine a flat dose of pembrolizumab 200mg in combination with chemotherapy.
Cohort D was a phase I study that examined the combination of pembrolizumab (10mg/kg or 2mg/kg) with anti-CTLA-4 antibody ipilimumab (0.3, 1 or 3mg/kg) every 3 weeks as second-line treatment in advanced NSCLC patients who have progressed on 2 or fewer lines of treatment. Combination immunotherapy for 4 cycles was followed by pembrolizumab monotherapy until progression or up to 2 years. Based on initial safety data from other immunotherapy combination trials, the protocol was amended to enroll patients on a pembrolizumab 2mg/kg and ipilimumab 1mg/kg every 3 weeks regimen. Three grade 3 events (17%) were reported: adrenal insufficiency, drug eruption, and maculopapular rash. Overall incidence of drug-related AEs was 33% and two treatment related-discontinuations were reported. There was no dose-limiting toxicity at the pembrolizumab 2mg/kg and ipilimumab 1mg/kg dose. In 18 response-evaluable patients, ORR was 39% and included 1 patient with complete response. Based on these results, the combination is being evaluated in an expansion cohort of 32 patients.
The combination of pembrolizumab (2mg/kg) with erlotinib 150mg daily or gefitinib 250mg daily is being evaluated in NSCLC patients with EGFR mutations. Efficacy results are not currently available for these cohorts.
Various other combinations of pembrolizumab with standard platinum chemotherapy, radiation therapy, post-surgery in adjuvant setting, with EGFR and ALK inhibitor and other targeted drugs, are being studied in several ongoing or soon-to-start phase I/II/III studies. Some of these trials are studying the role of pembrolizumab outside of the advanced/ metastatic NSCLC setting.
Pembrolizumab Activity in Subgroups of NSCLC
Hellmann et al reported activity of pembrolizumab in key subgroups of NSCLC at the 16th World Conference of Lung Cancer (WCLC) in Denver, CO.19Advanced NSCLC patients (n = 550) treated with different doses of pembrolizumab 2 mg/kg every 3 weeks, 10 mg/kg every 3 weeks, or 10 mg/kg every 2 weeks in the Key- note-001 trial were included. Patients with squamous histology (n=95) had ORR of 26.3% (50% for PD-L1 ex- pression tumor proportion score [TPS] ≥50%, 17.2% for TPS 1-49%, and 0% for TPS <1%) compared to non-squamous histology (n = 446) with ORR of 19.1% (35.9% for TPS ≥50%, 11.1% for TPS 1-49%, and 12.3% for TPS <1%). Median PFS for patients with squamous histology was 6.1 months (10.3 months for TPS ≥50% and 6 months for TPS 1-49%) compared to 3 months (4.3 months for TPS ≥50% and 2.3 months for TPS 1-49%) for patients with non-squamous histology. Patients with prior smoking history had OS of 14.3 months (26.8 months for TPS ≥50% and 10.6 months for TPS 1-49%) compared to 8.8 months (8.3 months for TPS ≥50% and 6.9 months for TPS 1-49%) for never smokers. Patients with EGFR mutation (n = 77) had ORR of 7.8% (20% for TPS ≥50%, 8.7% for TPS 1-49%) compared to ORR of 21.6% (39.8% for TPS ≥50%, 12.2% for TPS 1-49%) for EGFR wild type (n = 450). Similar results have been reported in other trials suggesting better activity for smokers and squamous histology compared to never smokers and patients with positive EGFR mutation.
Brain metastases are common among patients with advanced NSCLC and pose a treatment challenge due to the concern for the ability of drugs to penetrate the bloodbrain barrier. Patients with untreated brain metastases are generally excluded from most trials. Goldberg et al reported preliminary data at WCLC 2015 from an ongoing phase II trial evaluating response to pembrolizumab in 18 patients with PDL-1 expression and small, asymptomatic, brain metastases that are previously untreated or progressing after local treatment.20Brain metastases response rate and systemic response rate were 33%. All patients with systemic response also had response in brain mets except for 1 patient who progressed in brain after a transient response and required radiation therapy. There were no neurological aderse events greater than grade 1.
PD-L1 as Biomarker
In an unselected population, approximately 1 in 5 patients show durable response to checkpoint inhibitors.12Developing reliable, reproducible biomarkers to identify this subset of patients who are likely to show a response remains a lucrative target. In the Keynote-001 trial, there was an observed relationship between high level PD-L1 and response to pembrolizumab, so the initial protocol was amended to include a coprimary endpoint to assess the efficacy of pembrolizumab in patients with a high PDL-1 expression. The ORR was 45.2% (95% CI, 33.5-57.3) in patients with a proportional score of at least 50%, compared to 16.5% (95% CI, 9.9-25.1) in patients with performance status (PS) of 1-49% and 10.7% (95% CI, 2.3-28.2) for those with a PS score of less than 1%. The higher response rate for patients with a PS score of greater than 50 was sustained when divided by previously treated patients (P<0.001) and untreated patients (P= 0.01). Of all the patients who had samples that could be evaluated, 23.2 % of patients had a proportion score of at least 50%, 37.6% had a score of 1-49%, and 39.2% had a score of less than 1%. This finding of higher PD-L1 expression and response to pembrolizumab was also reported by Herbst et al as mentioned above. In a meta-analysis, Wang et al reported that high PD-L1 expression was significantly associated with poor differentiation of tumor and worse OS.21
There are several limitations in assessing the role of PD-L1 expression as a biomarker. Many different assays and definitions of positivity have been used to express PD-L1 positivity for different checkpoint inhibitors, which makes it difficult to compare results across trials. Moreover, PDL-1 expression can be variable within a tumor or in an individual at a different site of metastatic disease. As the tumor cells develop resistance and multiply post-treatment, it is likely that the expression of PD-L1 will change over time. In addition to the tumor cells, the tumor-invading dendritic and macrophage cells express PD-L1 and do not show uniform tumor invasion. Consequently, depending on the biopsy, the specimen might not be a true representative of actual PDL-1 expression.
Occasionally, interpretation of clinical benefit could be more complex with pembrolizumab or other checkpoint inhibitors as pseudoprogression can result in an increase in size of the target lesion before actual response evaluation criteria in solid tumors (RECIST) response occurs. Systematic criteria, defined as immune-related response criteria (irRC) have been created to capture these additional response patterns seen with immune therapy beyond those described by RECIST criteria. irRC takes into account total tumor burden and allows treatment even in the presence of new lesion and initial increase in the target lesion. Pseudoprogression has been encountered less frequently with NSCLC compared to melanoma and early rescanning in 4 to 6 weeks helps to differentiate pseudoprogression from true progression.
The treatment of NSCLC is changing rapidly and immune check point inhibitors such as pembrolizumab show promising results in the second-line setting with tolerable toxicities. Several ongoing phase II/III trials will help identify the role of pembrolizumab in the first-line setting alone or in combination with chemotherapy or CTLA-4 inhibitor. In multiple trials with pembrolizumab or other checkpoint inhibitors, there seems to be greater effect on OS compared to PFS, indicating continued benefit after treatment discontinuation with subsequent therapy. It is possible that PFS may not be an appropriate endpoint for clinical trials involving checkpoint inhibitors. As more data from ongoing clinical trials become available, it is likely that subsets within the NSCLC will be identified with more or less likely to benefit with pembrolizumab. For example, the subset analysis of Keynote 001 trial revealed that patients with sensitizing EGFR mutation-positive NSCLC are not the most likely group of patients to benefit from pembrolizumab. The toxicity profile of pembrolizumab is quite favorable compared to traditional chemotherapy and similar to other checkpoint inhibitors. Immune related adverse effects can occur any time during the treatment and have also been reported after treatment discontinuation. Most serious immune reactions can be adequately managed with discontinuation of treatment and prompt administration of corticosteroids. Future trials will likely explore other biomarkers in associaion with PD-L1 to enrich the patient population likely to benefit from this treatment.