ATTLAS, IMpower151 and ORIENT-31: Dusting off IMpower150 for Post-Osimertinib in EGFR-Mutated NSCLC?

Introduction

Immune checkpoint inhibitors (ICIs), including programmed death-ligand 1 (PD-L1), and programmed cell death (PD-1) inhibitors, as monotherapy or in combination with chemotherapy (Keynote-789, Checkmate-722, LIBRETTO-431 [control arm]) have limited efficacy in patients with oncogenic driver molecular alterations in advanced NSCLC.¹ Indeed, the use of immunotherapy as a monotherapy or in combination with chemotherapy for first-line treatment of advanced NSCLC excludes EGFR-mutant and ALK-rearranged NSCLC.² Until recently in the US, the treatment options for EGFR-mutant NSCLC post-EGFR-TKI were doublet platinum-based chemotherapies. In Europe in February 2019, the European Medical Agency (EMA) approved the IMpower150 quad regimen (carboplatin, paclitaxel, bevacizumab, atezolizumab) for treatment of EGFR-mutant NSCLC after progression with EGFR-TKI, based on a small subgroup retrospective analysis with both PFS and OS survival benefits^3,4 (Table 1). Hence, there is a need for consensus on effective post-osimertinib monotherapy.

Table 1 Comparison of IMpower150, ATTLAS and ORIENT-31 Trials

Mariposa-2

MARIPOSA-2 is the first randomized trial to investigate combination therapy after progression with osimertinib. Briefly, MARIPOSA-2 demonstrated that the quad regimen (platinum doublet chemotherapy + lazertinib [an alternate 3G EGFR-TKI] + amivantamab [an EGFR/MET bi-specific antibody]) (LACP) confers superior PFS over platinum doublet chemotherapy (CP) alone. The trial included a third investigation arm (chemotherapy + amivantamab, ACP), but was randomized in a 2:2:1, and was not the main primary endpoint for MARIPOSA-2. Essentially, LACP achieved superior PFS over CP (8.3 months vs 4.2 months; HR: 0.44; 95% CI: 0.35–0.56; P<0.001).⁵ However, LACP involves significantly higher adverse events requiring dose reduction and a higher incidence of venous thromboembolic disease. While the PFS of LACP was superior to that of CP, whether the regimen is easily administered to the vast majority of patients remains in question.⁶

IMpower151 and ATTLAS

It is within this context that we delve into the role of chemoimmunotherapy plus anti-angiogenic agents in the metastatic, EGFR-mutant NSCLC as an alternative to MARIPOSA-2. Previously, preclinical studies have shown that resistance to EGFR-TKIs may change the tumor microenvironment, resulting in favorable responses to immunotherapy.⁷ However, immunotherapy alone has no benefit in EGFR-mutant NSCLC,⁸ and the combination with chemotherapy failed to meet its primary endpoint in the KEYNOTE-789⁹ and CheckMate-72¹⁰ studies. The role of PD-1 or PD-L1 inhibitors is only supported with the addition of anti-angiogenic agents such as bevacizumab, a vascular endothelial growth factor A (VEGF-A) from the IMpower150⁴ and ORIENT-31 trials¹¹ (Table 1).

While ORIENT-31 was positive for the quad regimen over chemotherapy, it was conducted in a single country (China)¹¹ and the applicability of the results outside China remains to be determined.¹² Two similarly designed trials from China (IMpower151)¹³ and the Republic of Korea (ATTLAS)¹⁴ supported the role of the quad regimen. IMpower151 used a similar design to IMpower150, except that half of the patients had either an EGFR mutation or ALK rearrangement in IMpower151. ATTLAS had only two arms comparing platinum/paclitaxel + bevacizumab + atezolizumab (ABCP) to platinum/paclitaxel (CP) (Table 1).

The IMPOWER151 study was negative and did not meet its primary endpoint of investigator-assessed PFS in the intention-to-treat (ITT) population (9.5 vs 7.1 months, HR: 0.84). Of the 53% of patients enrolled with EGFR mutation and ALK rearrangement, there was no difference in PFS between the ABCP and BCP regimen (median PFS 8.5 vs 8.3 months, HR: 0.86). On the other hand, ATTLAS achieved its primary endpoint of improved PFS of ABPC over PC (median PFS, 8.48 months [95% CI: 8.18–10.28] versus 5.62 months [95% CI: 4.27–7.22; HR: 0.62, 95% CI: 0.45–0.86]). Importantly, the quad regimen in both trials achieved a median PFS of 8.5 months, which is within the range of LACP of MARIPOSA-2. The quad regimen theoretically achieves a PFS1 + PFS2 from osimertinib to IMpower151/ATTLAS of about 29 months, which is similar to the median PFS achieved in FLAURA-2 when osimertinib is combined with platinum-based chemotherapy.¹⁵

There was no subgroup analysis by PD-L1 expression for the EGFR-mutant NSCLC in IMpower151 from the WCLC 2023 presentation. The ABCP regimen in the ATTLAS trial showed PFS benefit as increased PD-L1 expression. Interestingly, exploratory analysis by PD-L1 (SP263) and distribution and density of tumor-infiltrating lymphocytes (TILs) in the tumor bed using artificial intelligence-powered Lunit SCOPE IO showed that an inflamed score of 20% or more demonstrated PFS benefit. However, this finding is hard to reproduce in clinical settings, and TILs remain largely exploratory. PD-L1 expression in EGFR-mutant NSCLC is highly heterogeneous, and is not a reliable predictive or prognostic biomarker for advanced non-squamous NSCLC without actionable genomic alteration.

FLAURA-2 and MARIPOSA

Recently, the improvement in PFS over osimertinib alone with the addition of chemotherapy (FLAUA-2)¹⁵ or with amivantamab (MARIPOSA)^16,17 has changed the treatment landscape for treatment-naïve patients. It is unknown whether chemotherapy or amivantamab were used in the front-line setting or if the second-line setting included chemotherapy and/or amivantamab rechallenge. In the future, specific second-line trials will have to be designed based on the initial treatment regimen and may complicate the sequential treatment landscape of advanced EGFR-mutant NSCLC.¹⁸

Proposed Second-Line Post-Osimertinib (3G) Clinical Trial Design

The IMpower151, ATTLAS and ORIEN-31 trials provided promising results suggesting that the quad regimen may be a effective for post-first-line EGFR-TKI. We thus propose a head-to-head comparison between the quad regimen of MARIPOSA-2 versus the quad regimen of IMpower151/Impower150/ATTLAS. We propose a randomized second-line immediate post-osimertinib/3G EGFR-TKI trial (Figure 1). Given the likely equivalency of both quad regimens, the primary endpoint could be non-inferiority design.

Figure 1 A hypothetical schema comparing the quad regimen of MARIPOSA-2 to the quad regimen of IMpower151/ATTLAS.

Summary

1. The treatment landscape for advanced EGFR-mutant NSCLC is rapidly evolving. The standard of care resulting from osimertinib as the first-line treatment is being challenged by the success of FLAURA-2 and MARIPOSA. These trials differ in that FLAURA-2 utilized chemotherapy upfront while MARIPOSA utilized “maximum” EGFR inhibition with both an EGFR-TKI and EGFR mAb.

2. All three trials (IMPower151, ATTLAS, and ORENT-031) were conducted in Asian countries. However, there is strength in numbers and all three trials point to a hypothesis-generating observation that chemotherapy in combination with anti-angiogenesis and ICI may be potentially an effective subsequent treatment, given the low frequency of EGFR-mutant NSCLC patients who had only exposure to 3G EGFR-TKI (N=33). Randomized trials comparing two quad regimens, MARIPOSA-2 versus IMpower151/ATTLAS post-osimertinib (or other 3G EGFR-TKI) progression, is urgently needed. We propose a simple head-to-head trial comparing the two quad regimens of MARIPOSA-2 and IMpower151/ATTLAS.