For patients with advanced non-small cell lung cancer (NSCLC) and high programmed cell death protein 1 ligand 1 (PD-L1) expression (tumor proportion score [TPS] ≥50%), the optimal first-line treatment strategy remains elusive. While PD-(L)1 inhibitor monotherapy is commonly prescribed and offers the appeal of a chemotherapy-free approach with a less severe toxicity profile, chemoimmunotherapy approaches that combine the cytotoxic activity of platinum-based chemotherapy with immune checkpoint blockade may overcome resistance to anti–PD-(L)1 monotherapy. However, it has been unclear whether the efficacy benefit of adding chemotherapy in this biomarker-selected population justifies the added toxicity. Di Federico et al (JAMA Oncol. 2026) addressed this evidence gap with a systematic review and reconstructed an individual patient data (IPD) meta-analysis of 24 phase III randomized controlled trials (RCTs) encompassing 5,546 patients with PD-L1–high advanced NSCLC, concluding that chemoimmunotherapy is profoundly beneficial relative to anti–PD-(L)1 monotherapy in this setting.
The authors systematically searched PubMed, Embase, and major oncology meeting abstracts (ESMO, ASCO, WCLC, ELCC) for phase III RCTs comparing PD-(L)1 inhibitor monotherapy or chemoimmunotherapy against platinum-based chemotherapy alone in previously untreated advanced NSCLC. High PD-L1 expression was defined as TPS ≥50% across all trials. The primary endpoint was overall survival (OS) and the secondary endpoint was progression-free survival (PFS). Safety, including treatment-related adverse events (TRAEs) and discontinuation rates, was an exploratory endpoint. The analysis employed multiple complementary methods—conventional meta-analysis, meta-regression, network meta-analysis, and reconstructed IPD analysis—to strengthen the robustness of the findings.
In the reconstructed IPD analysis, chemoimmunotherapy (n = 704) was associated with significantly longer median OS compared with PD-(L)1 inhibitor monotherapy (n = 1,706): 29.2 months (95% CI, 25.2–35.4) vs 19.8 months (95% CI, 18.3–21.7) (hazard ratio [HR] 0.74; 95% CI, 0.66–0.82; P <.001). Median PFS was likewise significantly longer with chemoimmunotherapy: 11.3 months (95% CI, 10.3–13.5) vs 6.8 months for immune monotherapy (95% CI, 6.2–7.1) (HR 0.67; 95% CI, 0.60–0.75; P <.001). Both strategies improved survival over chemotherapy alone, but tests for subgroup differences and meta-regression analyses consistently favored chemoimmunotherapy over immune monotherapy for both OS and PFS. On the safety side, PD-(L)1 inhibitor monotherapy was associated with a significantly lower risk of TRAEs compared with chemotherapy alone (risk ratio [RR] 0.76; 95% CI, 0.72–0.80; P <.001), while the risk with chemoimmunotherapy was similar to chemotherapy alone (RR 1.01; 95% CI, 1.00–1.02; P = .08). Grade ≥3 TRAEs and treatment discontinuation rates due to TRAEs were also significantly higher with chemoimmunotherapy than with immune monotherapy.
The authors acknowledge several important limitations, most notably the inherent lack of a direct comparison between the 2 strategies of interest. Rather, all comparisons rely on reconstructed IPD from Kaplan-Meier curves rather than source-level data, a method that introduces small but unavoidable imprecisions. To this end, the meta-analysis was conducted in accordance with Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines and registered with PROSPERO to ensure adherence to statistical methodology standards. Nonetheless, the granularity of published subgroup data was insufficient to determine whether patients with very high PD-L1 expression (≥90%), particular histologies, smoking histories, metastatic involvement patterns, or key mutations might have different benefit/toxicity patterns. The investigators stress that a definitive prospective head-to-head trial remains necessary.
These results provide the most comprehensive evidence to date favoring chemoimmunotherapy over immune monotherapy in PD-L1–high NSCLC, with distinct implications for oncologists across practice settings.
High Level
This meta-analysis represents the highest level of evidence currently available comparing chemoimmunotherapy with PD-(L)1 inhibitor monotherapy in PD-L1–high NSCLC, and the consistency of results across 4 analytic approaches lends support to the conclusion that combination therapy has notably superior efficacy. The nearly 10-month median OS difference (29.2 vs 19.8 months) is clinically significant and challenges the assumption that high PD-L1 expression alone is sufficient to justify a chemotherapy-free approach. However, the inability to perform meaningful subgroup analyses is a critical gap: Clinicians still lack the evidence to determine whether patients with TPS ≥90%, specific co-mutations such as STK11 or KEAP1, or particular metastatic patterns (eg, brain, liver) derive differential benefit from one strategy over the other. The authors’ call for a prospective RCT with prespecified subgroup analyses by PD-L1 expression level, histology, metastatic sites, and molecular variants defines the clear next step. Until that study is completed, clinicians must weigh the survival advantage of adding chemotherapy against its toxicity on a patient-by-patient basis, using judgment rather than level 1 evidence to guide individual decisions.
Ground Level
For community oncologists, this meta-analysis offers insights into the treatment of patients with advanced NSCLC and PD-L1 TPS ≥50%, indicating that chemoimmunotherapy was associated with significantly longer survival than PD-(L)1 inhibitor monotherapy in the first-line setting, with a median OS difference of nearly 10 months. While both approaches improve outcomes over chemotherapy alone, the consistency of the benefit across multiple analytic methods may warrant consideration of combination therapy for some patients who can tolerate platinum-based chemotherapy and who have the necessary support. At the same time, the safety data are an important counterbalance, as immune monotherapy was associated with significantly fewer TRAEs and lower discontinuation rates, reinforcing its role for patients with borderline performance status, significant comorbidities, or a preference for a less toxic approach. The practical takeaway is that individualized treatment selection remains critical, and while monotherapy remains a valid and evidence-based option, chemoimmunotherapy may benefit patients with high PD-L1 expression who are fit and motivated to maximize survival. Therefore, there remains no “universal” best approach to treatment of patients with advanced NSCLC and PD-L1 TPS ≥50%.