ONGOING TRIALS OF PEMBROLIZUMAB IN COMBINATION
Pembrolizumab was also evaluated in combination with chemotherapy as first-line NSCLC treatment, regardless of PDL1 expression, in a Phase I/II trial (KEYNOTE-021) comprised of eight cohorts:12,34cohort A, pembrolizumab 2/10 mg/kg + carboplatin AUC6 + paclitaxel 200 mg/m2; cohort B, pembrolizumab 2/10 mg/kg + carboplatin AUC6 + paclitaxel 200 mg/m2 + bevacizumab 15 mg/kg; cohort C, pembrolizumab 2/10 mg/kg + carboplatin AUC5 + pemetrexed 500 mg/m2; cohort D, pembrolizumab 2 mg/kg + ipilimumab 1 mg/kg; cohort E, pembrolizumab 2 mg/kg + erlotinib 150 mg; cohort F, pembrolizumab 2 mg/kg + gefitinib 250 mg; cohort G, carboplatin AUC5 + pemetrexed 500 mg/m2 with/without pembrolizumab 200 mg; and cohort H, extension of cohort D.
KEYNOTE-021 results are available only for cohorts A, B, D, G, and H, but not for cohorts C, D, or E. KEYNOTE-021 cohort G had the strongest clinical impact: the combination of pembrolizumab (200 mg IV) with pemetrexed (500 mg/m2 IV)–carboplatin (PPC; AUC5) Q3W achieved a better ORR than chemotherapy alone.12 Patients with unmutated EGFR or unrearranged ALK NSCLC were treatment-naïve and PS 0–1, without brain metastases or interstitial lung disease, and only systemic corticosteroid doses <10 mg/day were allowed. The primary end point was ORR, and secondary end points were PFS, OS, and pembrolizumab safety and efficacy according to PDL1 status. The 123 enrolled patients were randomized to the PPC arm (n=60) or to receive pemetrexed–carboplatin (PC) chemotherapy alone (n=63). A 26% ORR benefit was observed with pembrolizumab (ORR 55% with PPC vs 29% with PC, P=0.0016), with respective median PFS of 13.0 vs 8.9 months (HR 0.53, 95% CI 0.31–0.91; P=0.0102). OS did not differ between the two arms (HR 0.90, 95% CI 0.42–1.91), with 1-year OS at 75% and 72% for the PPC and PC arms, respectively. No effect of PDL1 status at the <1% threshold was observed. With the 50% threshold, ORR differed: 26% for patients with 1%–49% PDL1 (n=19) and 80% for the subgroup with ≥50% PDL1 expression (n=20).
All-grade AEs occurred in 93% of the PPC recipients and 90% of those given PC, among which 39% and 26%, respectively, were grade >2.12 The main PPC-associated AEs reported were fatigue (64%), nausea (58%), anemia (32%), cutaneous eruptions (27%), and vomiting (27%). IRAEs in the pembrolizumab-treated group were hypothyroidism (15%), hyperthyroidism (8%), and interstitial pneumopathy (5%). Those results, submitted to the FDA in January 2017, led to the approval of pembrolizumab in combination with (pemetrexed–carboplatin) chemotherapy for first-line treatment of advanced, nonsquamous-cell NSCLC without EGFR mutations or ALK rearrangements, regardless of PDL1 status.
Updated data were presented at the 2018 American Society of Clinical Oncology (ASCO) Annual Meeting, with median follow-up at 24 months (range 0.8–29.0 months).35 ORR was 57% with PPC vs 30% for PC (P=0.0016). PFS was significantly improved with PPC (24.0 months) vs PC (9.3 months; HR 0.53, 95% CI 0.33–0.86; P=0.0049). Median OS was not reached for PPC or PC, respectively: (24.5–not recorded [NR]) and 21.1 (14.9–NR) months (HR 0.56, 95% CI 0.32–0.95; P=0.0151). Twenty-four-month OS rates were 67% for the pembrolizumab arm and 48% for the placebo arm.35These updated data confirmed the first published results that had led to FDA approval.
The Phase III KEYNOTE-189 trial included 616 patients (randomized 2:1) to confirm the results of the KEYNOTE-021 trial on the efficacy of the pembrolizumab–pemetrexed–platinum combination for chemotherapy-naïve NSCLC without EGFR mutation or ALK rearrangement (Table 1).36 The main end points were PFS and OS. The recently published results supported the use of that combination. PFS was prolonged by 3.9 months for the pembrolizumab arm (median PFS 8.8 months) compared to the placebo arm (median PFS 4.9 months) with HR 0.52 (95% CI 0.43–0.64, P<0.001), with respective 1-year OS rates of 69.5% and 49.4%. According to subgroup analyses, no PFS benefit was obtained for PDL1 <1% (HR 0.75, 95% CI 0.53–1.05). Those findings stand out, regardless of tumor PDL1 status. After centralized review, ORRs were 46.7% for pembrolizumab recipients vs 18.9% for those given placebo (P<0.001).36
During the trial, 41% of placebo arm patients crossed over to receive pembrolizumab monotherapy or other immunotherapies after progression. Grade ≥3 toxicity rates for the pembrolizumab and placebo arms, respectively, were 67.3% and 65.2%, with respective 20.2% and 10.7% AE rates leading to discontinuation, and 8.9% and 4.5% grade ≥3 IRAE rates. The most frequent secondary grade ≥3 IRAEs were interstitial pneumopathy (2.7%, with three deaths), cutaneous reactions (2%), and nephritides (1.5%).36
For KEYNOTE-021 cohort A (pembrolizumab 2/10 mg/kg + carboplatin AUC6 + paclitaxel 200 mg/m2) Q3W, four cycles followed by pembrolizumab Q3W up to 2 years or progression, 13 of 25 (52%) patients achieved partial responses. For cohort B (pembrolizumab 2/10 mg/kg + carboplatin AUC6 + paclitaxel 200 mg/m2 + bevacizumab 15 mg/kg Q3W), four cycles also obtained partial responses. During the follow-up of one arm by pembrolizumab–bevacizumab maintenance, 12 of 25 (48%) patients had partial responses.
The pembrolizumab–ipilimumab combination given to KEYNOTE-021 cohorts D (pembrolizumab 2 mg/kg, ipilimumab 1 mg/kg) and H (cohort extension) led to 25% ORRs for 45 of the assessable patients given the pembrolizumab (2 mg/kg)–ipilimumab (1 mg/kg) combination, a rate similar to that obtained with pembrolizumab alone, and a 64% disease-control rate.37 The response was not associated with PDL1 status. Compared to immunotherapy alone, the combination’s safety was not as good, with 49% of patients experiencing grade 3/4 AEs, necessitating treatment discontinuation for 9%. All-grade IRAEs occurred in 40% of the patients. Principal ongoing trials are summarized in Table 2.
(To view a larger version of Table 2, click here.)
The randomized, open, Phase III KEYNOTE-042 trial was planned to include 1,240 patients in comparing first-line pembrolizumab vs chemotherapy for patients with advanced NSCLC with >1% PDL1 expression. Patients were randomized at a 1:1 ratio between pembrolizumab 200 mg Q3W for 2 years or until disease progression, with chemotherapy chosen by the physician. The results of this Phase III trial were presented in a presidential session at ASCO 2018.38 The primary objective of the study was OS of patients with PDL1 ≥50%, ≥20%, and ≥1%, with a statistically hierarchical analysis starting with patients with PDL1 ≥50%. A total of 1,274 patients were included (637 in each arm), including 599 PDL1 patients ≥50% (47%) and 818 PDL1 patients ≥20%. The median follow-up at the time of analysis was 12.8 months. OS was significantly increased in the pembrolizumab arm for patients with PDL1 >1% (16.7 vs 12.1 months, HR 0.81, 95% CI 0.71–0.93; P=0.0018), PDL1 ≥20% (17.7 vs 13.0 months, HR 0.77, 95% CI 0.64–0.92; P=0.0020), and PDL1 ≥50% (20.0 vs 12.2 months, HR 0.69, 95% CI 0.56–0.84; P=0.0003) compared to chemotherapy. OS gain was 4.5 months, 4.7 months, and 7.8 months for patients with PDL1 ≥1%, PDL1 ≥20%, and PDL1 ≥50% respectively. OS benefit was essentially driven by PDL1 >50% patients. For patients with PDL1 of 1%–49%, an exploratory analysis found no difference in OS between pembrolizumab and chemotherapy (HR 0.92, 95% CI 0.77–1.11). It should be noted that only 19.3% of patients on the chemotherapy arm benefited from a crossover to second-line immunotherapy. The toxicity profile was acceptable, with less grade ≥3 toxicity in the pembrolizumab arm (17.8% vs 41%).38
For patients with advanced squamous-cell NSCLC, the randomized, double-blind, Phase III KEYNOTE-407 trial was designed to compare the efficacy and safety of a first-line pembrolizumab–platinum chemotherapy combination, regardless of PDL1 expression. The primary end point was PFS, with OS and PFS secondary-outcome criteria for the population with PDL1 ≥1% expression, ORR, DOR, and safety. The study recruited 560 patients, randomized 2:1 between pembrolizumab 200 mg IV QW3 carboplatin–nab-paclitaxel and chemotherapy alone. Results of the second interim analysis were presented at ASCO 2018 by Paz-Ares et al,39 and are summarized in Table 1. OS increased significantly in the pembrolizumab + chemotherapy arm (median 15.9 months vs 11.3 months, HR 0.64, 95% CI 0.49–0.85; P=0.0008). Similarly, median PFS increased from 4.8 months to 6.4 months (HR 0.56, 95% CI 0.45–0.70; P<0.001). ORR increased with pembrolizumab (57.9% vs 38.4%). These positive results were observed regardless of the level of PDL1 expression. Tolerance was identical in both arms, with the same rate of grade 3–5 side effects (one treatment-related death in each arm) and of course more autoimmune events in the pembrolizumab arm (28.8% vs 8.6%).39
For patients with advanced all-histology PDL1 ≥50% NSCLC, a randomized Phase-III trial with 542 planned inclusions compared fixed-dose 200 mg pembrolizumab and a combination pembrolizumab fixed-dose 200 mg Q3W–ipilimumab 1 mg/kg Q6W. The auxiliary PEARLS (KEYNOTE-091) trial, enrolling 1,380 patients, compared in a Phase III setting pembrolizumab (200 mg IV Q3W for 1 year) vs placebo in stage IB/IIIA NSCLC patients who had undergone surgical resections (lobectomy or pneumonectomy), with adjuvant chemotherapy or without. The principal outcome measure was disease-free survival. In the neoadjuvant setting, with the objective of improving disease-free survival and OS, the Phase-III KEYNOTE-671 trial, with 789 planned inclusions, evaluated the safety of pembrolizumab in combination with cisplatin–pemetrexed or cisplatin–gemcitabine chemotherapy depending on tumor histology before surgery, followed postoperatively by pembrolizumab alone in subjects with resectable stage IIB/IIIA NSCLC.