Should chemotherapy plus immune checkpoint inhibition be the standard front-line therapy for metastatic NSCLC?

Front-line therapy for NSCLC: the shift from chemotherapy to immunotherapy

Immunotherapy has revolutionized the way we approach patients with metastatic NSCLC.¹ Platinum-based chemotherapy has been the standard first-line treatment for advanced NSCLC for the past several decades. Refinements to chemotherapy have been made over the subsequent years, including improvements to the chemotherapy doublet partner,^2–3 the addition of the VEGF inhibitor bevacizumab,⁴ and continuation of chemotherapy beyond induction therapy into maintenance therapy.⁵ These changes have increased survival incrementally for certain subsets of patients with NSCLC. However, a major paradigm shift occurred in 2016 when the PD-L1 inhibitor pembrolizumab was demonstrated to have a superior survival compared to platinum-based chemotherapy when used as first-line therapy for patients with NSCLC with PD-L1 expression >50%.⁶ Since that time, pembrolizumab has become the standard treatment option for patients with both squamous and non-squamous NSCLC with high PD-L1 expression, while those with low PD-L1 expression continued to receive chemotherapy.

Rationale for combining chemotherapy with immune checkpoint inhibitors

Ideally, one would have preferred to combine PD-1 and PD-L1 inhibitors with more specific immunotherapeutic agents targeting known mechanisms of checkpoint inhibitor sensitivity and resistance,⁷ however at present none of these combinations have been elevated to standard of care. Instead, the combination of chemotherapy with checkpoint inhibitors has successfully advanced through clinical trial development.

There are several possible reasons to consider combining chemotherapy with a checkpoint inhibitor as first-line treatment for patients with advanced NSCLC. The most straight-forward is that many patients with lung cancer will not have the opportunity to receive a second-line therapy, typically due to increasing symptom burden and declining performance status. For example, in the trial comparing pembrolizumab with chemotherapy, only 43.7% of patients who received chemotherapy as first-line therapy crossed over to receive pembrolizumab.⁶ Therefore giving multiple agents as initial therapy may increase the chance of the patient responding to one of the agents administered.⁸ Clearly, we want as many patients as possible to have the opportunity to benefit from immunotherapy.

The more elegant reason to combine chemotherapy with immunotherapy is the possible synergy between the two treatments. Although the mechanism is not entirely clear, chemotherapy has been shown to modulate the immune system by depleting myeloid-derived suppressor cells⁹ and increasing the ratio between effector T cells and regulatory T cells.¹⁰ Interestingly, studies have demonstrated that different chemotherapy agents can impact the immune response to varying degrees,^9–11 which may explain certain outcomes seen on trials with various chemotherapy regimens.

Clinical data on chemotherapy plus immunotherapy for non-squamous NSCLC

Different combinations of chemotherapy and immune checkpoint inhibitors have been tested in phase I trials, and the first combination to emerge from a phase 2 trial was carboplatin/pemetrexed/pembrolizumab. The KEYNOTE-021 trial was a multi-arm study that included a cohort of 123 patients with advanced, chemotherapy-naïve non-squamous NSCLC without an EGFR mutation or ALK rearrangement. Patients were randomized to carboplatin/pemetrexed plus pembrolizumab for 4 cycles followed by maintenance therapy with pemetrexed/pembrolizumab, or carboplatin/pemetrexed for 4 cycles followed by maintenance pemetrexed. Initial results demonstrated an improvement in objective response rate (ORR) and progression-free survival (PFS) for the chemo plus pembro arm (ORR 55% vs 29%, p=0.0016; median PFS 13 vs 8.9 months (p=0.010).¹² These results led to the US FDA approval of the combination of carboplatin/pemetrexed/pembrolizumab for patients with previously untreated metastatic non-squamous NSCLC. The follow-up phase 3 trial, KEYNOTE-189, that randomized a similar population of patients to the same two arms was performed to confirm these findings. The primary endpoints were overall survival (OS) and PFS, and importantly, cross-over to pembrolizumab in the chemotherapy arm was permitted. This trial also demonstrated striking results: OS at 12 months was 69.2% in the chemo plus pembrolizumab group versus 49.4% in the placebo-chemotherapy group (HR 0.49, 95% CI, 0.38 to 0.64, P<0.001).¹³ PFS was also improved in the immunotherapy-containing arm. High-grade toxicity was similar between the two groups. Interestingly, OS was improved in all sub-groups of PD-L1 expression including <1%, 1–49%, and >50%, however PFS was improved in only the 1–49% and >50% groups.

Additional trials have combined chemotherapy with immunotherapy. The CheckMate-227 trial was a phase 3 trial that enrolled patients with advanced, chemotherapy-naive NSCLC regardless of PD-L1 expression. Those with PD-L1 expression <1% were enrolled to nivolumab plus chemotherapy, nivolumab plus ipilimumab, or chemotherapy alone. In this subset of patients with PD-L1 < 1%, those who received nivolumab plus chemotherapy had an improvement in PFS compared to chemotherapy alone (median PFS 5.6 vs 4.7 months, HR 0.74), a finding that was most pronounced in the non-squamous population (HR 0.68) as well as those with high tumor mutation burden (HR 0.56).¹⁴ The IMpower150 trial combined atezolizumab with chemotherapy with or without bevacizumab and was unique in two ways: it was the first trial of chemotherapy plus immunotherapy to also add a VEGF inhibitor, and also allowed the inclusion of EGFR-mutated and ALK-rearranged patients after progression on targeted therapy.¹⁵ 1202 patients with non-squamous NSCLC were randomized to atezolizumab plus carboplatin/paclitaxel with or without bevacizumab or carboplatin/paclitaxel/bevacizumab. Among the wild-type patients, PFS was improved with the addition of the atezolizumab to carboplatin/paclitaxel/bevacizumab, with a median of 8.3 vs 6.8 months (HR 0.62, P<0.0001). OS was also improved with a median OS 19.2 vs 14.7 months (HR 0.78, P=0.02), a trend that was maintained regardless of PD-L1 status, although stronger in the PD-L1 high group. Importantly, the subgroup of patients with EGFR mutations or ALK rearrangements also appeared to benefit from the addition of atezolizumab (median 9.7 vs 6.1 months, HR 0.59, 95% CI 0.37–0.94), although the total number of patients was small and confirmatory studies are likely needed. During the interim analysis of the atezolizumab plus chemotherapy versus bevacizumab plus chemotherapy arms, there was no significant benefit in terms of overall survival between the two groups.

Certainly there is no obvious biological reason why these effects should be histology-specific. In fact, two recent phase III trials were completed specifically examining chemo-immunotherapy combinations as first-line treatment for patients with metastatic squamous cell carcinoma of the lung. The KEYNOTE-407 randomized patients with any PD-L1 status to carboplatin plus either paclitaxel or nab-paclitaxel with or without pembrolizumab. This trial demonstrated a benefit in terms of OS, PFS, and ORR with the addition of pembrolizumab to chemotherapy. Although the magnitude of benefit appears less than in the similar trial with non-squamous patients, this still marks the first time that the addition of a checkpoint inhibitor to chemotherapy has demonstrated benefit for patients specifically with squamous NSCLC, regardless of PD-L1 status.¹⁶ A similarly designed trial, IMpower131, randomized patients with squamous cell lung cancer to carboplatin/paclitaxel/atezolizumab, carboplatin/nab-paclitaxel/atezolizumab, or carboplatin/nab-paclitaxel. The results of the two arms containing nab-paclitaxel were presented with an interim OS analysis demonstrating similar survival between the two groups but an improvement in PFS and response rate with the addition of atezolizumab.¹⁷ Table 1 summarizes the key phase 3 trials of chemotherapy plus immunotherapy in first-line NSCLC.

Table 1.

Phase 3 trials of chemotherapy plus immunotherapy in first-line NSCLC.

	Treatment	Patient population	PD-L1 Status	EGFR/ALK	Cross-over allowed	OS (median)	PFS (median)	ORR (%)
KEYNOTE-18913	Carboplatin/ pemetrexed/ pembrolizumab vs Carboplatin/ pemetrexed	Treatment-naïve, advanced non-squamous NSCLC	Any	Not eligible	Yes	NR vs 11.3 (HR, 0.49, 95% CI 0.38–0.64, p<0.001)	8.8 vs 4.9 (HR 0.52, 95% CI 0.43–0.64, p<0.001)	47.6 vs 18.9 (p<0.001)
IMpower15015	Carboplatin/ paclitaxel/ bevacizumab/ atezolizumab vs Carboplatin/ paclitaxel/ bevacizumab	Treatment-naïve, advanced non-squamous NSCLC	Any	Eligible after 1 prior TKI	No	19.2 vs 14.7 (HR 0.78, 95% CI 0.64–0.96), p=0.02	8.3 vs 6.8 (HR 0.62, 95% CI 0.52–0.74, p<0.0001)	63.5 vs 48
CheckMate-22714	Chemotherapy/ nivolumab vs chemotherapy	Treatment-naïve, advanced NSCLC (~75% non-squamous)	Any; PD-L1 <1% reported	Not eligible	N/A	N/A	5.6 vs 4.7 (HR 0.74, 95% CI 0.58–0.94)	36.7 vs 23.1
KEYNOTE-40716	Carboplatin/ paclitaxel or nab-paclitaxel/ pembrolizumab vs Carboplatin/ paclitaxel or nab-paclitaxel	Treatment-naïve, advanced squamous NSCLC	Any	N/A	Yes	15.9 vs 11.3 (HR 0.64, 95% CI 0.49–0.85, P=0.0008)	6.4 vs 4.8 (HR 0.56, 95% CI 0.45–0.70, P<0.0001)	57.9 vs 38.4
IMpower 13117	Carboplatin/ nab-paclitaxel/ atezolizumab vs Carboplatin/ nab-paclitaxel	Treatment-naïve, advanced squamous NSCLC	Any	N/A	No	14 vs 13.9 (HR 0.96, 95% CI 0.78–1.18, P=0.6931) *First interim OS analysis	6.3 vs 5.6 (HR 0.71, 95% CI 0.60–0.85, P=0.0001)	49 vs 41

Should chemotherapy plus immune checkpoint inhibition be the standard front-line therapy for metastatic NSCLC?

To answer the question posed in the title: yes, at least for most patients. Chemotherapy plus an immune checkpoint inhibitor should be considered a standard first-line treatment for patients with advanced NSCLC without an EGFR or ALK alteration. Compared to chemotherapy alone, there is a survival advantage in patients with both non-squamous and squamous NSCLC, regardless of PD-L1 expression. High-grade toxicity is not significantly increased with the addition of a checkpoint inhibitor, although care must be taken to monitor patients closely as side effects from both drug classes can occur.

However, it appears PDL-1 status remains important, and there are, of course, still several outstanding questions regarding the exact patients who are most likely to benefit and the particular regimen to use. First, until chemotherapy plus immunotherapy is compared to immunotherapy alone, an unanswered question is which regimen is superior for patients with high PD-L1 expression (>50%). At this time, patients with PD-L1 expression > 50% can also be given pembrolizumab alone based on the results of the KEYNOTE-024 trial which demonstrated a survival advantage compared to chemotherapy. This will likely involve discussions and shared decision-making between patients and physicians and depend on such variables as performance status and co-morbidities. Second, some patients are unlikely to benefit from combination therapy, yet identifying these patients remains a challenge. For patients without PD-L1 expression in their tumor, there did appear to be an improvement in survival with the addition of immunotherapy to chemotherapy in several trials, yet the benefit is less pronounced than in the PD-L1 expressers. Possibly those not deriving benefit are the patients without PD-L1 expression and low tumor mutation burden; additional studies are needed in this regard but this population may be able to receive chemotherapy alone. Finally, the exact regimen to use is still an area of investigation as several trials are ongoing.

Currently the only FDA-approved regimen is carboplatin/pemetrexed/pembrolizumab for non-squamous NSCLC, but several other combination regimens have demonstrated promise. Whether the addition of bevacizumab will provide incremental benefit is still an unanswered question, however it is intriguing that the combination of carboplatin/pemetrexed/bevacizumab/atezolizumab appeared to improve survival even in patients with EGFR mutations and ALK rearrangements,¹⁵ a population that typically does not benefit from immunotherapy and has been excluded from all other combination trials. Bevacizumab may be specifically important in these patients with oncogene addicted tumors¹⁸ and should be studied in additional trials. We have now seen benefit from carboplatin/paclitaxel or nab-paclitaxel/pembrolizumab in patients with squamous NSCLC,¹⁶ providing these patients with a much-needed additional treatment option beyond chemotherapy alone. However, it is intriguing that the magnitude of benefit does not appear as great compared to the non-squamous trial with carboplatin/pemetrexed/pemetrexed; whether this is due to tumor histology, different chemotherapy regimen, or the continuation of maintenance chemotherapy along with maintenance immunotherapy is unknown at this time.

In summary, recent data has clearly demonstrated the superior outcomes for the combination of chemotherapy with immune checkpoint inhibitors in patients with both non-squamous and squamous NSCLC. Additional follow-up will be necessary to confirm the long-term benefit for combination therapy, as well as to define which patients are most likely to benefit. In either case, the future for immune checkpoint inhibitors in NSCLC continues to look bright.