In the setting of advanced non–small cell lung cancer, the choice of up-front immunotherapy or chemotherapy, or combinations of both, is aided by findings from several trials.
H. Jack West, MD
In the setting of advanced non—small cell lung cancer (NSCLC), the choice of up-front immunotherapy or chemotherapy, or combinations of both, is aided by findings from several trials discussed during a recent expert panel presentation on the OncLive® NewsNetwork.
Agents such as pembrolizumab (Keytruda), nivolumab (Opdivo), ipilimumab (Yervoy), and atezolizumab (Tecentriq), alone or in combination, offer the promise of improved response rates but need to be considered in the context of PD-L1 expression, the panel noted.
Some of the results from KEYNOTE-042, -189, and -407; CheckMate-227; and IMpower150 are practice-changing, according to the panel, moderated by H. Jack West, MD, medical director of the Thoracic Oncology Program at the Swedish Cancer Institute in Seattle, Washington. The panel discussed how these findings are likely to shape treatment and considered yet-unanswered questions.In patients with nondriver mutations, KEYNOTE-042 results showed the difference high PD-L1 expression can make in the efficacy of frontline immune monotherapy. “The dramatic benefit with pembrolizumab monotherapy was in the 50% [PD-L1] or higher subset, and it diluted progressively as you added patients with a 20% cutoff or a 1% cutoff. And when you look at just the group of patients with 1% to 49% PD-L1, there was really no difference,” West said.
In KEYNOTE-042, compared with chemotherapy, frontline pembrolizumab improved median overall survival (OS) in correlation with PD-L1 expression for patients with advanced non—small cell, squamous, or nonsquamous disease: tumor proportion score (TPS) ≥50% (20 vs 12.2 months; HR, 0.69; 95% CI, 0.56-0.85; P = .0003), TPS ≥20% (17.7 vs 13.0 months; HR, 0.77; 95% CI, 0.64-0.92; P = .002), and TPS ≥1% (16.7 vs 12.1 months; HR, 0.81; 95% CI, 0.71-0.93; P = .0018).1 However, across all patients with PD-L1 TPS of 1% to 49%, pembrolizumab improved median OS by just over 1 month (13.4 vs 12.1; HR, 0.92; 95% CI, 0.77-1.11).
In the phase III trial, patients with locally advanced or metastatic NSCLC were randomly assigned to pembrolizumab (n = 637) or chemotherapy (n = 637) with paclitaxel plus carboplatin or pemetrexed (Alimta) plus carboplatin. Patients with both squamous and nonsquamous cancers were included, but those who had cancers with genetic changes that could be treated with targeted therapies, such as EGFR and ALK inhibitors, were not eligible. West noted that there was no significant difference between the treatment groups for median OS in the TPS 1% to 49% population.
Crossover rates between arms were relatively low in KEYNOTE-042, less than would be expected in the United States, where access to advanced drugs is generally higher. West said that 19.8% of patients who were enrolled globally were allowed to cross over from the chemotherapy arm to the pembrolizumab arm.
Crossover rates would be higher in his own clinical practice, West said. “I think mostly in the United States, where we have much better access and an expectation that if you didn’t get it as first-line [treatment] and you were ever a candidate for immunotherapy, you’d get it second line,” he said.
Results of the phase III KEYNOTE-407 made a quiet splash at the 2018 American Society of Clinical Oncology Annual Meeting, owing to delayed publicity, but those data amounted to one of the most important presentations at the meeting, panelists said. Regardless of PD-L1 status, the combination of immunotherapy plus platinum-based chemotherapy improved OS and several other measures of response compared with chemotherapy alone. Also, adverse events were similar between the arms.
In adults with metastatic squamous NSCLC, pembrolizumab was given in combination with frontline carboplatin/paclitaxel or nab-paclitaxel (Abraxane) versus standard solvent-based paclitaxel with either placebo or pembrolizumab in patients with metastatic NSCLC.2 The median OS was 15.9 months (95% CI, 13.2-not evaluable) with pembrolizumab versus 11.3 months (95% CI, 9.5-14.8) with chemotherapy alone (HR, 0.64; 95% CI, 0.49-0.85; P = .0008). The OS benefit was observed regardless of PD-L1 expression level, choice of taxane, age, sex, and ECOG performance status.
Compared with chemotherapy alone, the addition of pembrolizumab resulted in a 1.6-month improvement in median PFS (6.4 vs 4.8 months; HR, 0.56; 95% CI, 0.45-0.70; P <.0001). Investigators observed a correlation between higher PD-L1 levels and greater PFS benefit, and the benefit was present across all PD-L1 expression levels.
The objective response rate (ORR) was 57.9% in the pembrolizumab arm with a complete response (CR) rate of 1.4% and a partial response (PR) rate of 56.5%. In the control arm, the ORR was 38.4%, with a CR rate of 2.1% and a PR rate of 36.3%. The duration of response was 7.7 months (range, 1.1+ to 14.7+) versus 4.8 months (1.3+ to 15.8+), favoring the pembrolizumab arm.
“Before late May, all we had heard was that there was a response rate difference which in itself didn’t impress me that much, except that in the abstract it was a 23% difference, which is enough to get noticed. But the trial also showed a major difference in progression-free survival and overall survival, both in the broad population and in all of the different subgroups across the PD-L1 spectrum,” West said.Panelists said they were impressed with KEYNOTE-189 because its results also demonstrated the strength of chemotherapy plus immunotherapy, regardless of PD-L1 status, over chemotherapy alone. Based on the strength of KEYNOTE-189, in August the FDA approved frontline pembrolizumab for use in combination with chemotherapy for metastatic nonsquamous NSCLC.
Presented at the American Association for Cancer Research Annual (AACR) Meeting 2018 and published in the New England Journal of Medicine, KEYNOTE-189 (N = 616) assessed frontline pembrolizumab or placebo in combination with pemetrexed and either cisplatin or carboplatin in patients with metastatic nonsquamous NSCLC. At a median follow-up of 10.5 months, the estimated 12-month OS rate was 69.2% (95% CI, 64.1%-73.8%) in the pembrolizumab arm compared with 49.4% (95% CI, 42.1%-56.2%) in the control group (HR, 0.49; 95% CI, 0.38-0.64; P <.001).3,4
The median OS was 11.3 months (95% CI, 8.7-15.1) in the chemotherapy-alone arm versus not reached (NR) in the pembrolizumab arm. PD-L1 status did not affect the OS benefit.
The study also met its co-primary endpoint for median PFS, which was 8.8 months (95% CI, 7.6-9.2) in the pembrolizumab group versus 4.9 months (95% CI, 4.7-5.5) in the control arm (HR, 0.52; 95% CI, 0.43 to 0.64; P <.001).
“When we look at the outcomes of KEYNOTE-189 regardless of PD-L1 status [and] regardless of performance status, patients did better if they got pembrolizumab plus chemotherapy rather than chemotherapy alone,” said Gregory J. Riely, MD, PhD, a medical oncologist and vice chair of clinical research in the Department of Medicine at Memorial Sloan Kettering Cancer Center in New York. “I think, regardless of how you approach initial therapy with regard to PD-L1 status, if you look at PD-L1 status or don’t, this applies to all from PD-L1 0%, 1%, or 50%. And so I think this was probably the most practice-changing trial we’ve seen in lung cancer over the last year or two.”
West pointed out that the triplet of pembrolizumab, pemetrexed, and carboplatin won FDA approval in May 2017 based on results from cohort G1 of the phase II KEYNOTE-21 trial. Riely replied that the approval was based on PFS results, and the KEYNOTE-189 findings provide stronger evidence supporting the triplet.
“The fact that KEYNOTE-189 showed a survival endpoint was really a big plus, and I think the other power in this study was that it was so much bigger. We’re able to look at PD-L1 status and see more clear signals,” Riely said. “In the original study, some of those PD-L1 numbers looked a little funny in that it seemed like some people with sort of midlevel PD-L1 didn’t do as well as people who had no PD-L1. It was all very confusing. But when we see the 189 data, we see very clear progression-free survival [and] overall survival signals that really are across PD-L1 status, and it’s very robust. I think it’s impressive stuff.”
Regarding added toxicity, the immuno-chemotherapy combination from KEYNOTE-189 is not a huge concern, said panelist Charu Aggarwal, MD, PhD, an assistant professor of medicine at the Hospital of the University of Pennsylvania, in Philadelphia. “Personally, in using the triplet, I find that it’s very well tolerated without adding much cumulative toxicity above and beyond what I would have obtained or had as a side effect with just chemotherapy alone, so I find that it’s really practice-changing and reinforces the data from the phase II trial.”
Jared Weiss, MD, section chief of Thoracic and Head and Neck Oncology and associate professor at the University of North Carolina School of Medicine, said the KEYNOTE-189 results leave him unsure of how to treat patients with PD-L1 expression ≤49%, but that he suspects that holding immunotherapy until the second line is a reasonable strategy for those patients.The phase III CheckMate-227 trial produced long-duration responses among small subsets of patients with the combination of nivolumab/ipilimumab versus chemotherapy, results that align with patient hopes to fall within the “tail end of the [mortality] curve,” Weiss said. “They’re not asking us for an average expectation of an extra month. They’ll take it, of course. But what they come in asking for is some chance at being alive 2 years down the road, 3 years down the road. So, from a human perspective, that’s extremely provocative.”
In part Ib of CheckMate-227 (N = 550), the combination of nivolumab and low-dose ipilimumab reduced the risk for progression or death by 52% compared with standard platinum doublet chemotherapy for patients with metastatic PD-L1—negative, tumor mutation burden (TMB)-high NSCLC.5
Irrespective of histology, the median PFS was 7.7 months among in the PD-L1—negative (<1% expression), TMB-high (≥10 mutations/megabase) subgroup assigned to nivolumab/ipilimumab compared with 6.2 months for nivolumab/chemotherapy and 5.3 months for chemotherapy alone. The 1-year PFS rate was 45% with nivolumab/ipilimumab compared with 27% for nivolumab/chemotherapy and 8% for chemotherapy.
In the TMB-high/PD-L1—negative group, 93% of patients continued to respond to the nivolumab/ipilimumab combination for ≥1 year compared with 33% of patients in the nivolumab/chemotherapy arm. The Kaplan-Meier curves for duration of response to chemotherapy did not extend past 12 months, and duration was not calculable in the chemotherapy-alone group.
The median PFS was 5.6 months in the nivolumab/chemotherapy arm across the full unselected population compared with 4.7 months for chemotherapy alone (HR, 0.74; 95% CI, 0.58-0.94). The median PFS with nivolumab/ipilimumab was 4.4 months (HR vs chemo, 0.79; 95% CI, 0.62-1.01). At 1-year, the PFS rate was 29% for nivolumab/ipilimumab, 26% for nivolumab/chemotherapy, and 14% for chemotherapy alone.
In the chemotherapy-free arm, nivolumab was given at 3 mg/kg every 2 weeks and ipilimumab was dosed at 1 mg/kg every 6 weeks. In the chemotherapy-immunotherapy group, nivolumab was administered at 360 mg every 3 weeks. Weiss said that TMB makes sense as a biomarker, noting the good reaction to the nivolumab/ipilimumab combination that TMB-high patients had.
“The more mutations you have, the more potential immunogenic neoantigens you’re going to have—potentially with a greater efficacy of these agents,” Weiss said. “In this study, for a [TMB] of at least 10, we saw the efficacy of [ipilimumab/nivolumab] compared with the other arms, both in PD-L1—low patients and in PD-L1–high patients and both in squamous and in nonsquamous histology.”
Unlike KEYNOTE-189, CheckMate-227 has not yet demonstrated an OS benefit, West said. “It’s encouraging that there’s clearly a pattern here [that] chemo-immunotherapy is superior,” he said. “But one of my thoughts in the wake of AACR, where you had the KEYNOTE-189 trial with a survival benefit and then -227 with the PFS but not OS, I thought, ‘You don’t bring a knife to a gunfight, and you don’t bring PFS to an OS fight.’ And I think if you have competing options and one delivers the big OS benefit, that’s going to get the nod in my mind.”
In May 2017, the FDA approved pembrolizumab in combination with pemetrexed and carboplatin for the treatment of patients with previously untreated metastatic nonsquamous NSCLC, based on findings from the open-label phase II KEYNOTE-021 cohort G study. The lack of long-term data at the time rendered this a controversial approval; however, in just a little over 1 year, multiple phase III trials have reported findings showing a significant advantage for immunotherapy plus chemotherapy over chemotherapy alone.
In updated long-term findings from the KEYNOTE-021 cohort G trial, median OS with the immunotherapy combination was not reached (95% CI, 24.5 months-NR) compared with 21.1 months (95% CI, 14.9-NR) for carboplatin and pemetrexed alone. The 2-year OS rate was 67% with pembrolizumab compared with 48% in the control arm, representing a 44% reduction in the risk of death with the immunotherapy combination (HR, 0.56; 95% CI, 0.32-0.95).IMPower150 (N = 1202), like KEYNOTE-189, delivered OS results, but Aggarwal was not convinced that the survival advantage conferred by the combination of atezolizumab, bevacizumab (Avastin), carboplatin, and paclitaxel (ABCP) compared with bevacizumab and chemotherapy (BCP) was robust enough to justify using the combination in patients with stage IV wild-type NSCLC with EGFR mutation or ALK rearrangement.6,7
“The survival numbers, or the advantage that we are seeing, do not really compare to what we can get with targeted therapy with much less adverse events and with much better ease of use for the patient and much better quality of life,” she said. “So, while these data are intriguing, I don’t think they are practice-changing for those 2 subsets of otherwise well and easy-to-treat lung cancer patients.”
The median OS for the quadruplet was 19.2 months (95% CI, 17.0-23.8) compared with 14.7 months (95% CI, 13.3-16.9) in the BCP arm (HR, 0.78; 95% CI, 0.64-0.96; P = .0164). The 24-month OS rate with atezolizumab was 43% compared with 34% for BCP. ABCP also improved median PFS by 1.5 months compared with BCP (8.3 vs 6.8 months; HR, 0.59; 95% CI, 0.50-0.70; P <.0001).
Data for patients with EGFR/ALK-mutated NSCLC were excluded from the coprimary endpoints of OS and PFS. Roughly 10% of patients were positive for EGFR mutation, and 2% to 5% had an ALK rearrangement. These patients had received at least 1 prior EGFR tyrosine kinase inhibitor prior to enrollment. When those patients were included in the intent-to-treat population, the median OS with ABCP increased to 19.8 months versus 14.9 for BCP (HR, 0.76; 95% Cl, 0.63-0.93).
Riely was glad that this patient population was included, but was unconvinced that the ABCP regimen “magically” improved outcomes for patients with EGFR/ALK-mutated NSCLC.
“It’s not clear that this quadruplet is somehow particularly effective for the EGFR-mutant patient or for an ALK-positive patient,” Riely said. “We haven’t seen those curves broken out specifically to look at just EGFR-mutant patients, and we haven’t seen even a table showing characteristics of these patients. Was there an imbalance in EGFR exon 19 deletions? We don’t know.”