Background
Phase III, open-label RCT (IMpower150). N=1,202 with stage IV non-squamous NSCLC, no prior chemotherapy for metastatic disease, ECOG PS 0–1. Three-arm design. EGFR/ALK-altered patients eligible (after targeted therapy failure); PD-L1-low/negative included. Tested whether adding atezolizumab to bevacizumab + carboplatin + paclitaxel (BCP) yields additive benefit across PD-L1 and genomic subgroups.
Interventions and follow up
Arm A (ACP): Atezolizumab + carboplatin + paclitaxel × 4–6 cycles, then atezolizumab maintenance
Arm B (ABCP): Atezolizumab + bevacizumab + carboplatin + paclitaxel × 4–6 cycles, then atezolizumab + bevacizumab maintenance
Arm C (BCP, control): Bevacizumab + carboplatin + paclitaxel × 4–6 cycles, then bevacizumab maintenance
Primary endpoint: PFS and OS in WT (EGFR/ALK wild-type) and Teff-high WT subgroups
mFollow up: 15.2 months (PFS analysis)
Arm B (ABCP): Atezolizumab + bevacizumab + carboplatin + paclitaxel × 4–6 cycles, then atezolizumab + bevacizumab maintenance
Arm C (BCP, control): Bevacizumab + carboplatin + paclitaxel × 4–6 cycles, then bevacizumab maintenance
Primary endpoint: PFS and OS in WT (EGFR/ALK wild-type) and Teff-high WT subgroups
mFollow up: 15.2 months (PFS analysis)
Results
PFS (ABCP vs BCP, WT): 8.3 vs 6.8 mo, HR 0.62 (95% CI 0.52–0.74), P<.001
OS (ABCP vs BCP, WT): 19.2 vs 14.7 mo, HR 0.78 (95% CI 0.64–0.96), P=.02
PFS (ABCP vs BCP, EGFR/ALK-altered): HR 0.59 — benefit maintained regardless of EGFR/ALK status
PFS (ABCP vs BCP, PD-L1 negative): HR 0.77
ORR (ABCP vs BCP, WT): 63.5% vs 48.0%
OS (ABCP vs BCP, WT): 19.2 vs 14.7 mo, HR 0.78 (95% CI 0.64–0.96), P=.02
PFS (ABCP vs BCP, EGFR/ALK-altered): HR 0.59 — benefit maintained regardless of EGFR/ALK status
PFS (ABCP vs BCP, PD-L1 negative): HR 0.77
ORR (ABCP vs BCP, WT): 63.5% vs 48.0%
Adverse events
Overall (grade ≥3): 58.5% (ABCP) vs 50.0% (BCP)
Immune-mediated (any grade): 39.9% vs 24.7%
Cardiovascular: grade ≥3 hypertension 8.3% vs 6.5%
Dermatologic: rash 30.0% vs 22.2%
Hepatic: hepatotoxicity grade ≥3 2.2% vs 0.4%
Pulmonary: pneumonitis grade ≥3 1.2% vs 0%
GI: perforation 0.6% vs 0.4%
Discontinuation (AE): 16.7% vs 12.0%
Immune-mediated (any grade): 39.9% vs 24.7%
Cardiovascular: grade ≥3 hypertension 8.3% vs 6.5%
Dermatologic: rash 30.0% vs 22.2%
Hepatic: hepatotoxicity grade ≥3 2.2% vs 0.4%
Pulmonary: pneumonitis grade ≥3 1.2% vs 0%
GI: perforation 0.6% vs 0.4%
Discontinuation (AE): 16.7% vs 12.0%
Conclusions
Atezolizumab + bevacizumab + carboplatin + paclitaxel (ABCP) significantly improved PFS and OS versus BCP in 1L metastatic non-squamous NSCLC, with benefit regardless of PD-L1 expression and — notably — in patients with EGFR/ALK alterations after targeted therapy failure.
Key Limitations
Open-label design. Complex four-drug regimen with higher toxicity and infusion burden vs pembrolizumab+chemo doublets. The EGFR/ALK and PD-L1-negative benefits derive from exploratory/subgroup analyses, not powered comparisons. ACP (no bevacizumab) arm showed weaker benefit, leaving the relative contribution of bevacizumab uncertain. Paclitaxel-based backbone less commonly used than pemetrexed in non-squamous practice.
Clinical Context
ABCP received FDA approval (2018) for 1L metastatic non-squamous NSCLC; EMA approval followed. ESMO-MCBS 3. Historically the principal IO-containing regimen with evidence in EGFR/ALK-altered patients post-targeted therapy, though pemetrexed-based doublets (KEYNOTE-189) are generally preferred for IO-eligible non-squamous patients given simpler toxicity.
References