Atezolizumab/Bevacizumab Plus Chemotherapy Sustains Survival Benefit After 20 Months in EGFR+ NSCLC

Martin Reck, MD, PhD

Atezolizumab (Tecentriq) in combination with bevacizumab (Avastin), carboplatin, and paclitaxel (ABCP) demonstrated continued progression-free survival (PFS) and overall survival (OS) benefit at 20 months of follow-up compared with bevacizumab, carboplatin, and paclitaxel (BCP) in patients with chemotherapy-naïve metastatic nonsquamous non–small cell lung cancer (NSCLC) who harbor an EGFR mutation, according to updated efficacy results from a sub-analysis the phase 3 IMpower150 clinical trial (NCT02366143).

IMpower150 was launched to address failure to tyrosine kinase inhibitors (TKIs), which are the recommended treatment option in the frontline setting. After failure on TKIs, the PFS achieved with these agents in the second- and later-line settings are significantly lower. Prior results from IMpower150 did demonstrate a survival benefit with an 8.3-month median PFS in the atezolizumab arm compared with 6.8 months in the control arm (95% CI, 0.52-0.74; P < .001). The initial results from the study led to the FDA approval of the regimen in patients with metastatic nonsquamous NSCLC with an EGFR mutation.

In the randomized, open-label, international study, 1202 patients were randomized 1:1:1 to receive either ABCP, atezolizumab plus carboplatin and paclitaxel (ACP), or BCP every 3 weeks. Atezolizumab was administered at 1200 mg intravenously (IV), carboplatin AUC 6 was also administered IV, and was bevacizumab at 15 mg/kg. The dose level for paclitaxel was 200 mg/m2.

In the intention-to-treatment population in the subgroup analysis, 87% of patients had wild-type disease, 13% had an EGFR or ALK mutation. Specifically, 10% in the mutated group were EGFR-positive and 3% were ALK-positive.

Across all subgroups explored in the study, patients with EGFR mutations who were treated with ABCP had a median PFS of 10.2 months compared with 7.1 months in the BCP arm (HR, 0.56; 95% CI, 0.34-0.91). Among those with sensitizing EGFR mutation, the median PFS was 10.3 months with ABCP compared with 6.1 months in the BCP arm (HR, 0.38; 95% CI, 0.21-0.68). Additionally, in the subset of 50 patients who received prior treatment with a TKI, the median PFS was 9.7 months in the ABCP group versus 6.1 months in the BCP group.

Comparing treatment with ACP versus BCP, the median PFS in patients with an EGFR mutation was 6.9 months compared with 7.1 months, respectively (HR, 0.93; 95% CI, 0.60-1.44). In the group of 65 patients with an EGFR sensitizing mutation, the median PFS with ACP was 6.5 months versus 6.1 months with BCP (HR, 0.88; 95% CI, 0.53-1.47). The 56 patients who received prior TKI therapy had a median PFS of 5.7 months in the ACP arm versus 6.1 months in the BCP arm (HR, 1.24; 95% CI, 0.72-2.15).

OS in the intention-to-treat population favored ABCP compared with BCP in patients with an EGFR mutation. The median OS was 20.3 months (95% CI, 13.4-33.6) in the ABCP arm versus 26.1 months (95% CI, 17.0 41.4) in the BCP arm (HR, 0.91; 95% CI, 053-1.59). The ACP arm was slightly inferior to the BCP arm in terms of OS though. The median OS was 20.3 months (95% CI, 13-4-33.6) in the ACP arm compared with 21.4 months (95% CI, 14.0-31.4) in the BCP arm (HR, 1.16; 95% CI, 0.71-1.89).

The BCP group demonstrated better OS than both ABCP and ACP among the subgroup of patients who had an EGFR sensitizing mutation. The median OS was 18.1 months (95% CI, 11.7-27.8) in the ABCP arm compared with 29.4 months (95% CI, 24.9-not evaluable). In the BCP arm (HR, 0.60l 95% CI, 0.31-1.14). The median OS observed with ACP was 18.1 months (95% CI, 11.7-27.8) versus 19.0 months (95% CI, 13.5-28.5) with BCP (HR, 1.00; 95% CI, 0.57-1.74).

Finally, BCP had much longer OS compared with both ABCP and ACP among patients with sensitizing EGFR-positive disease who received prior treatment with a TKI. The median OS observed was 18.1 months (95% CI, 12.3-27.8) in the ABCP arm versus 27.8 months (95% CI, 18.6-41.4) in the BCP arm (HR, 0.74; 95% CI, 0.38-1.46). Treatment with ACP showed a median OS of 14.9 months (95% CI, 13.1-22.3) compared with 18.1 months (95% CI, 12.3-27.8) in the BCP arm (HR, 1.22; 955 CI, 0.68-2.22).

Objective responses were higher and the duration of response (DOR) longer in patients who were treated with ABCP compared with those who received BCP. The objective response rate (ORR) was 73.5% in the ABCP arm with a median DOR of 11.1 months (95% CI, 6.8-16.1) versus 40. (% with a median DOR of 4.7 (95% CI. 4.2-7.1) in the BCP arm, and 37.8% with a median DOR of 6.9 months (95% CI, 4.2-14.5) in the ACP arm. The difference between ABCP and BCP was demonstrated with a hazard ratio (HR) of 029 (95% CI, 0.14-0.58), and the difference between ACP and BCP was demonstrated with an HR of 0.47 (95% CI, 0.22-1.01). Notably, there was no change in the ORR following treatment with either ACP versus BCP, but there was a small increase in the DOR.

Overall, 44 patients from the ACP arm, 33 patients from the ABCP arm, and 43 patients from the BCP arm were evaluated for safety. Considering the longer follow-up in the BCP arm, these patients received more subsequent therapies that were unrelated to the study protocol. Specifically, 79.1% of the BCP received non-protocol treatment compared with 45.5% in the ABCP arm and 70.5% in the ACP arm. In addition, 27.3% of the ABCP arm and 43.2% of the ACP received targeted therapy, and 21.2%, and 18.2%, respectively were given post-progression osimertinib (Tagrisso).

The safety profile overall was consistent with each drug individually. Treatment was well-tolerated in the EGFR-positive population, despite a longer treatment duration. In all study arms, grade 3 o 4 treatment-related adverse events (AEs) were reported, including in 57% of the ACP arm, 67% of the ABCP arm, and 56% of the BCP arm. Only 1 patient in the BCP arm had a grade 5 AE. Serious AEs were reported in 36.4% of the ACO arm, 42.4% of the ABCP arm, and 20.9% of the BCP arm. Some patients withdrew from the study due to AEs, including 13.6 from the ACP arm, 42.4% from the ABCP arm, and 16.3% from the BCP arm.

Immune-related AEs were also reported in more than 5 patients. The most common in the ACP, ABCP, and BCP arms, respectively, were rash (36.4%, 33.3%, and 11.6%) hepatitis (6.8%, 15.2%, and 7.0%), and hypothyroidism (4.5%, 21.2%, and 2.3%).

The investigators of IMpower150 lead by Marin Reck, MD of Lugen Clinic, conclude from this sub-analysis that ABCP may be a novel treatment option for patients with EGFR-mutant NSCLC who failed treatment with a TKI.

_References:

_{Reck M, Mok T, Socinski M, et al. IMpower150: Updated efficacy analysis in patients with EGFR mutations. Presented at: 2020 European Society of Medical Oncology Virtual Congress. September 19-21, 2020; Virtual. Abstract 1293P.}

_{Socinski MA, Jotte RM, Cappuzzo F, et al. Atezolizumab for first-line treatment of metastatic nonsquamous NSCLC. N Engl J Med. 2018; 378(24):2288-230. doi: 10.1056/NEJMoa171694}