Treatment-free survival with first-line nivolumab plus ipilimumab in metastatic non-small cell lung cancer: a critical analysis beyond PD-L1 expression.

Introduction
The recent CheckMate 227 Part 1 trial by Peters et al. [2025] showed that first-line nivolumab plus ipilimumab (NIVO + IPI) significantly improved treatment-free survival (TFS) compared with platinum doublet chemotherapy in patients with metastatic non-small cell lung cancer (NSCLC) across programmed death-ligand 1 (PD-L1) subgroups, with benefits sustained through 5 years. TFS, defined as the time between discontinuation of first-line therapy and the initiation of a new systemic therapy or death, is a clinically meaningful endpoint because it reflects not only the durability of immunotherapy benefit but also prolonged intervals off systemic therapy, translating into improved quality of life, fewer cumulative toxicities, and potential health-economic advantages. At 5 years, the Kaplan-Meier estimated overall survival (OS) rate was 20% with NIVO + IPI vs. 11% with chemotherapy. At the same landmark, 14% of patients in the NIVO + IPI arm remained alive and treatment free without grade ≥3 treatment-related adverse events (TRAEs), compared with 3% in the chemotherapy arm (1). The findings represent a meaningful shift towards patient-centered outcomes in lung cancer research, also highlighting the durable benefit of dual checkpoint blockade after discontinuing treatment. The questions now are whether TFS, specifically when adjusted for high-grade toxicity, should be broadly adopted as a complementary endpoint in evaluating long-term benefits, and, given that the United States Food and Drug Administration (U.S. FDA) has approved NIVO + IPI as a first-line treatment for patients with metastatic NSCLC whose tumors express PD-L1 ≥1% (2), whether NIVO + IPI alone should be considered a preferred first-line option in metastatic NSCLC regardless of PD-L1 expression status, noting that both National Comprehensive Cancer Network (NCCN) and European Society for Medical Oncology (ESMO) guidelines currently recommend chemo-immunotherapy rather than dual checkpoint inhibition for patients with PD-L1-negative disease (3,4).
In interpreting the role of NIVO + IPI, it is important to situate these findings within the existing first-line standards established by pembrolizumab-based regimens. KEYNOTE-024 and KEYNOTE-042 demonstrated that pembrolizumab monotherapy provides durable survival benefit for patients with PD-L1 ≥50% and ≥1%, respectively (5,6). Furthermore, chemo-immunotherapy combinations such as those evaluated in KEYNOTE-189 and KEYNOTE-407 remain foundational across PD-L1 subgroups, offering substantial improvements in response rates, progression-free survival (PFS), and OS compared with chemotherapy alone (7,8). Despite theoretical immunosuppressive effects of cytotoxic chemotherapy, clinical evidence does not indicate that adding chemotherapy diminishes the antitumor activity of immunotherapy; rather, biomarker analyses from KEYNOTE-189 show preserved or improved benefit across PD-L1 strata (9,10). The present commentary therefore focuses not on revisiting these well-established standards, but on examining the distinct contribution of dual checkpoint inhibition through the lens of TFS, which is an endpoint not captured in KEYNOTE trials yet increasingly relevant to long-term quality-of-life considerations.
The mechanical insights of both interventions are depicted in Figure 1, contrasting the immunosuppressive tumor microenvironment induced by platinum-based chemotherapy and the immune-activated phenotype promoted by dual immune checkpoint blockade. While cytotoxic chemotherapy primarily induces tumor cell death through DNA damage, it simultaneously fosters immune exhaustion, impaired antigen presentation, and expansion of regulatory and suppressive cell populations (e.g., regulatory T cells, tumor-associated macrophages, myeloid-derived suppressor cells) (11-14). In contrast, NIVO + IPI reprograms the tumor immune landscape by improving antigen presentation, reinvigorating cytotoxic T cells, and reducing immunosuppressive barriers (15,16), thereby providing a compelling mechanistic rationale for its evaluation as a first-line strategy to generate durable anti-tumor immunity in metastatic NSCLC.
It is important to note that while platinum-based chemotherapy may confer immunosuppressive effects at a mechanistic level, clinical studies do not show reduced immunotherapy responsiveness when chemotherapy is combined with programmed cell death protein 1 (PD-1) blockade (7,8); chemo-immunotherapy remains an effective and guideline-endorsed first-line approach.

Strengths and limitations of the study
The large, randomized, phase 3 CheckMate 227 Part 1 trial (NCT02477826) provides a long-term and comprehensive evaluation of TFS among 1,739 treatment-naive patients with stage IV or recurrent NSCLC and no EGFR or ALK alterations. TFS was defined as the interval between the end of first-line therapy and the initiation of subsequent systemic therapy or death. When adjusted for grade ≥3 TRAEs, this endpoint serves as a proxy for quality-adjusted survival. NIVO + IPI demonstrated superior TFS and OS compared to chemotherapy in both PD-L1 ≥1% and <1% subgroups. The 6-year OS rates were 29.3% vs. 23.7% in the PD-L1 ≥1% subgroup, and 27.8% vs. 18.5% in the PD-L1 <1% subgroup for NIVO + IPI vs. chemotherapy, respectively. Additionally, NIVO + IPI outperformed NIVO monotherapy in PD-L1 ≥1% patients (TFS: 12.6 vs. 9.6 months; difference: 3.0 months) and outperformed nivolumab plus chemotherapy in PD-L1 <1% patients (TFS: 11.3 vs. 7.2 months; difference: 4.1 months).
The durability of benefit observed in CheckMate 227 Part 1 aligns well with outcomes from other landmark trials in addition to providing novel insights into TFS and toxicity-adjusted quality of life. In metastatic NSCLC (stage IV or recurrent) without EGFR/ALK alterations and treated in the first-line setting, the combination of NIVO + IPI with two cycles of chemotherapy in CheckMate 9LA yielded a 5-year OS of 18% vs. 11% with chemotherapy alone [hazard ratio (HR): 0.73], with durable benefit across PD-L1 strata, histology types, and among patients with brain metastases (17). The IFCT-1701 DICIPLE trial offered insights into finite-duration immunotherapy: advanced NSCLC patients with disease control at 6 months who discontinued NIVO + IPI had a numerically longer PFS (35.2 vs. 20.5 months) and significantly fewer grade 3–5 TRAEs (2.9% vs. 28.6%) compared to those who continued therapy (18). However, given the observational design and relatively small sample size, these findings should be interpreted cautiously, and their generalizability may be limited. A pooled safety analysis from CheckMate 227, 817, and 568 further supported the tolerability of NIVO + IPI across >1,200 metastatic NSCLC patients, with 3-year OS reaching 50% among those who discontinued due to toxicity and nearly half of responders remaining in remission ≥2 years after cessation (19). Nevertheless, heterogeneity of the included populations and differences in trial design may confound pooled outcomes, warranting cautious interpretation. Real-world validation from CheckMate 817, which enrolled underrepresented populations such as patients with Eastern Cooperative Oncology Group performance status (ECOG PS) 2, brain metastases, or comorbidities, showed a 3-year OS of 20.5% in these high-risk groups, suggestive of the regimen’s broad applicability beyond ideal trial settings (20). Together, these data point towards durable benefit of first-line NIVO + IPI across PD-L1 subgroups and in the context of tumor mutational burden analyses from CheckMate 227.
The study has several important limitations. First, TFS measures time off systemic therapy only and does not typically account for local treatments like radiotherapy or palliative procedures unless specified, which may underrepresent the true treatment burden. Second, decisions around initiating subsequent therapy after discontinuation were not protocol-mandated and were left to clinical judgment, introducing variability across patients though this was balanced between arms. Third, while TFS offers meaningful insight into treatment-free durability, it remains an exploratory endpoint without regulatory validation. Finally, the 2-year cap on immunotherapy reflects trial protocol and may not fully capture real-world scenarios where extended treatment is considered in select patients.

Challenges in implementing first-line NIVO + IPI in metastatic NSCLC with PD-L1 <1%
Despite compelling long-term survival data from CheckMate 227 and 9LA, translating first-line NIVO + IPI into routine care for patients with PD-L1 <1% remains a challenge. This subgroup derives clear benefit as seen with a 5-year OS of 19% vs. 7% with chemotherapy in CheckMate 227 and 22% vs. 8% in 9LA but faces several unique implementation difficulties (17).
The absence of PD-L1 expression challenges traditional biomarker driven decision-making. Clinicians may hesitate to escalate to dual immunotherapy without molecular predictors, especially in patients with competing comorbidities or marginal performance status (21).
Immune-related adverse events occur in over one-third of patients, and up to 29% of older adults (≥75 years) discontinue therapy due to toxicity, as seen in pooled analyses from CheckMate 227/568/817 (19). In less specialized settings, the infrastructure to promptly identify and manage these events may be lacking.
Reimbursement and access remain barriers in many health systems, especially given the higher up-front cost of dual checkpoint blockade relative to chemo-immunotherapy (22). Additionally, real-world adherence to IPI dosing varies, and it is unclear whether reduced dosing compromises outcomes in this biomarker-low population (23,24).
The treatment duration question is unresolved. While protocol capped therapy at 2 years, data from trials like DICIPLE suggest early discontinuation may be feasible for some; but such strategies are not yet guideline supported.
In practice, treating PD-L1 <1% NSCLC with NIVO + IPI requires weighing durable benefit against uncertainty in biomarkers, toxicity, and resource burden; challenges that will need ongoing real-world validation to address.

Should TFS be broadly adopted and is NIVO + IPI alone ready for routine use in PD-L1 negative metastatic NSCLC?
TFS provides a meaningful view of durable benefit but remains an exploratory, non-validated endpoint. OS remains the gold standard for evaluating treatment efficacy, whereas TFS may be best positioned as a secondary or exploratory endpoint, particularly in quality-of-life-focused trials. While NIVO + IPI alone has shown long-term survival in PD-L1 negative NSCLC, its broader use is limited by toxicity, lack of biomarkers, and patient selection challenges. It may be a viable option in select cases, but routine use requires careful risk benefit assessment.

Conclusions
NIVO + IPI alone offers sustained survival and prolonged off-treatment intervals in metastatic NSCLC, including in patients with PD-L1 <1%. However, TFS is not a validated endpoint for regulatory approval and may not consistently reflect quality-adjusted life years. Moreover, NIVO + IPI without chemotherapy is not approved by the U.S. FDA for PD-L1-negative disease, and routine adoption in this setting would be off-label. Consistent with NCCN and ESMO guidelines, the current standard of care for PD-L1-negative NSCLC remains chemo-immunotherapy rather than dual checkpoint inhibition. Further prospective data are needed to support broader adoption.

Supplementary
The article’s supplementary files as