Dear Editor, We read with great interest the article by Yang et al, “Comparative effectiveness for Early-Stage NSCLC Without Lymph Node Involvement Based on Prospective Studies”1. The authors conducted a network meta-analysis of prospective trials comparing different treatment modalities for stage IA–IIB N0 NSCLC. They found that immunotherapy plus SBRT (I-SBRT) improved outcomes in inoperable patients, comparable to adjuvant chemotherapy, and that EGFR-mutant patients benefited most from targeted therapy. This study provides valuable evidence for individualized treatment in early-stage NSCLC, though certain limitations warrant further investigation. The paper follows the TITAN 2025 Guidelines on Artificial Intelligence declaration and usage2. First, one important aspect not addressed in this study is the lack of stratification by PD-L1 expression. Immunotherapy efficacy is closely linked to PD-L1 status, and pooling patients with different levels may obscure subgroup differences, producing only an “average” estimate of I-SBRT effectiveness. Moreover, the latest National Comprehensive Cancer Network (NCCN) guidelines recommend adjuvant immunotherapy only for patients with PD-L1 ≥1%3. Accordingly, incorporating PD-L1 stratification would therefore enhance the generalizability and clinical relevance of the conclusions. Second, while this study included neoadjuvant immunotherapy, it did not address adjuvant immunotherapy. The IMpower010 trial showed that adjuvant atezolizumab after chemotherapy improved DFS in resected stage II–IIIA NSCLC with PD-L1 ≥1% and has been adopted in NCCN guidelines3,4. Therefore, consideration of such evidence would better align the analysis with current practice and strengthen guidance for early-stage and biomarker-defined subgroups. In addition, the analysis of clinical staging could be further refined. Although patients with stage IA–IIB N0 disease were included, prognostic differences across these stages are considerable; for example, outcomes for stage IA versus IIB, or T3N0 with larger tumors or chest wall invasion, vary substantially5,6. The study did not stratify efficacy by these subgroups, leaving uncertainty about the applicability of I-SBRT across all early-stage N0 patients. More detailed stage-specific analyses in future studies would help define optimal strategies for different risk groups. On the other hand, further evidence is required to clarify the effects on distant relapse and long-term survival. As the analysis relied mainly on DFS, the short-term benefit of I-SBRT in distant metastasis control cannot confirm durable systemic efficacy, and the absence of overall survival data limits validation. Extended follow-up focusing on recurrence dynamics would allow a more complete evaluation of long-term outcomes across treatment strategies. Lastly, the molecular subgroup analysis was limited to EGFR mutations. Yet KRAS, ALK, ROS1 and other alterations together comprise a substantial share of NSCLC cases and already have targeted therapies in advanced disease7. Including these driver subtypes in early-stage analyses would improve the generalizability of the findings and support precision strategies across diverse molecular profiles. In summary, we highly appreciate the authors’ systematic comparison of treatment strategies for early-stage NSCLC without lymph node involvement. The study underscores the promise of I-SBRT and confirms the benefit of adjuvant TKI therapy in EGFR-mutant patients, while refinement is still needed in areas such as PD-L1 stratification, adjuvant immunotherapy, stage-specific analyses, long-term outcomes, and other driver alterations. Future large-scale prospective trials integrating molecular and clinical factors will be crucial to advance individualized treatment.
Jiang et al. (Tue,) studied this question.