Immune checkpoint inhibitors (ICIs) have revolutionized the standard treatment strategies for patients with advanced and metastatic non-small cell lung cancer (NSCLC). ICIs, such as anti-programmed cell death protein 1 (PD-1) and programmed cell death ligand 1 (PD-L1), have been demonstrated to prolong overall survival (OS) in patients with metastatic NSCLC and have been approved as first-line systemic therapy for advanced NSCLC (1–3). However, the efficacy of ICIs is limited by both primary and acquired resistance (4). Preclinical studies have demonstrated that combining radiotherapy (RT) with ICIs offers synergistic benefits, including RT-induced tumor debulking, triggering immunogenic cell death, releasing tumor-associated antigens, activating tumor-associated dendritic cells, remodeling immunosuppressive tumor microenvironments, and modulating immune checkpoint molecule upregulation (5–8). Recently, an increasing number of clinical trials have explored the efficacy of combining ICIs with RT for NSCLC. Notably, the phase III PACIFIC (9)and GEMSTOM 301 (10) trials established the combination of RT and ICIs as a treatment modality for unresectable, stage III NSCLC. A phase I/II trial demonstrated that pembrolizumab, with or without concurrent RT, did not significantly affect the objective response rates (ORR) or progression-free survival (PFS) in patients with metastatic NSCLC (11). Numerous studies have suggested that combining RT and ICIs may benefit NSCLC treatment, though some have reported conflicting results (12, 13). These discrepancies are likely attributable to patient heterogeneity and variations in the modalities of combination therapy. Therefore, we aimed to evaluate the clinical outcomes of combining thoracic radiotherapy (TRT) and ICIs in patients with advanced NSCLC in real-world clinical practice and identify subgroups that may benefit most from this approach.

In our study, the combination of TRT and ICIs significantly prolonged OS and enhanced the 24-month and 36-month OS rates in patients with advanced NSCLC compared with ICIs alone. This highlights the efficacy of this combination strategy for treating advanced NSCLC. A secondary analysis of the KEYNOTE-001 trial showed that patients who had previously received radiotherapy exhibited longer PFS (4.4 months vs. 2.1 months, P = 0.019) and OS (10.7 months vs. 5.3 months, P = 0.026) than did those who did not, with an acceptable safety profile (17). Theelen et al. performed a secondary analysis of the MD Anderson Cancer Center (phase 1/2) and Pembrolizumab and Radiotherapy (phase 2) trials, where the thorax was the most common site targeted by radiotherapy (18). The combination of pembrolizumab and radiotherapy demonstrated a higher OS rate than that of pembrolizumab alone (19.2 months vs. 8.7 months, P = 0.0004), and PFS was significantly improved in the combination group (9.0 months vs. 4.4 months, P = 0.045). Additional radiotherapy significantly enhanced the out-of-field response rates (abscopal response rate: 41.7% vs. 19.7%, P = 0.0039; abscopal disease control rate: 65.3% vs. 43.4%, P = 0.0071). Moreover, neither of the two aforementioned trials achieved the predefined primary endpoints (11, 19). The phase III PACIFIC trial (9) demonstrated that the consolidation therapy with durvalumab following platinum-based concurrent chemoradiotherapy significantly improved PFS (P < 0.0001) and OS (P = 0.00251) in patients with unresectable stage III NSCLC. In conclusion, our findings are consistent with the current research, suggesting that the combination of RT with ICIs has considerable potential for the treatment of advanced NSCLC.

Although the clinical characteristics were not perfectly balanced between the two groups, a greater proportion of patients in the TRT+ICIs group received ICIs as second-line or later-line therapy, were treated with ICIs monotherapy, and showed a higher NLR at the start of treatment. According to the KEYNOTE-001 trial, enrolled patients with confirmed locally advanced or metastatic NSCLC received pembrolizumab (20). Treatment-naive patients demonstrated superior outcomes, with a median OS of 22.3 months (95% CI 17.1-32.3) compared to 10.5 months (95% CI 8.6-13.2) in previously treated patients. The 5-year OS rates were 23.2% and 15.5% for treatment-naive and previously treated cohorts, respectively. A retrospective cohort study by Heyward et al. analyzing SEER-Medicare data (2013–2019) from 17,681 individuals found that ICI plus chemotherapy showed significantly reduced mortality risk compared to ICI monotherapy in the first-line treatment setting (21). A Meta-analysis included data from 17 Phase III clinical trials involving 10,283 patients and demonstrated that immunotherapy significantly improves survival in lung cancer patients regardless of smoking status (never-smokers, former, or current smokers), with no significant interaction effects observed between treatment outcomes and smoking history (22). Ksienski et al. found that higher NLR significantly correlated with shorter OS in stage III NSCLC patients treated with durvalumab (23). Similarly, Lin et al. observed that increased NLR predicted worse OS in NSCLC patients receiving neoadjuvant therapy followed by surgical resection (24). All the aforementioned studies consistently demonstrated that the TRT+ICI group had poorer predicted survival outcomes. However, contrary to these expectations, the OS in the TRT+ICIs group was significantly longer than that in the ICIs-only group, providing further evidence for the effectiveness of this combined treatment strategy.

Preclinical evidence indicates that the combination of RT and ICIs produces synergistic effects through multiple mechanisms involving radiation-induced tumor cell killing that generates an in situ vaccine effect, enhances antigen presentation and dendritic cell activation, reprograms the immunosuppressive tumor microenvironment, and modulates immune checkpoint molecule expression (5, 6, 8, 25, 26). This study further confirms the therapeutic efficacy of the combined treatment.

The subgroup analysis of OS between the TRT+ICIs and ICIs-only groups identified factors associated with improved survival, including male sex (27), former smoking, ECOG performance status 0–1 (23, 28), stage IIIb-c (9, 10), high albumin level (29), and low NLR level (23, 28, 30). These findings almost align with those of previous studies, supporting the validity of our results. Only former smoking demonstrated a survival benefit inconsistent with the aforementioned meta-analysis results (22). This discrepancy may relate to the limited sample size in this study. In addition, the subgroup findings should be interpreted cautiously rather than definitive conclusions, particularly given non-significance after multiplicity correction. These results require prospective validation in adequately powered studies.

In the univariate survival analysis, TRT, PD-L1 expression, albumin level, and NLR were statistically significant. However, in the multivariate analysis, only TRT, PD-L1 expression, and NLR were independent predictors of prognosis. A significant association with OS was observed only in patients with PD-L1 < 1% and ≥ 50%. At the baseline level, the absolute proportion of PD-L1 ≥ 50% in the TRT + ICI group was indeed higher than that in the ICIs-only group (28.7% vs. 17.4%). However, there was no statistically significant difference between the two groups (P = 0.059). Moreover, in the multivariate analysis, the regression model was constructed to adjust for other confounding factors. This adjustment strategy ensures the reported protective effect of PD-L1 ≥50% is independent of baseline group differences. Patients in the PD-L1 ≥ 50% group were more likely than those in the PD-L1 < 1% group to benefit from treatment with ICIs, which is consistent with previous findings (3). No significant association was observed for the PD-L1 1-49% group, potentially due to the limited sample size. Ksienski et al. reported that elevated NLR was significantly associated with shorter OS in patients with stage III NSCLC treated with durvalumab (23).

Few studies have evaluated the optimal parameters for RT. In our study, we categorized RT techniques primarily into CFRT and SBRT. Our findings suggest that both CFRT and SBRT may enhance immunological efficacy. In the PACIFIC trial (9), patients received consolidation therapy with durvalumab following platinum-based concurrent chemoradiotherapy, with a typical CFRT dose of 60–66 Gy delivered in 30–33 fractions. A previous retrospective study concluded that SBRT (P = 0.013) and concurrent RT combined with ICIs (P = 0.002) were significantly associated with improved outcomes (31). Li et al. demonstrated that both CFRT (n = 75) and SBRT (n = 42), as well as intrathoracic RT (n = 76) and extrathoracic RT (n = 43), when combined with immunotherapy, could enhance survival (27). However, their study was limited by its relatively small sample size. Verma et al. reported that TRT in combination with ICIs is safe for patients, regardless of the different RT techniques and fractionation schedules (32).

In our study, no statistically significant difference was observed in the overall rate of treatment-related AEs between the TRT+ICIs and ICIs-only groups. Similarly, the incidences of grade 3 or higher AEs were comparable between the two groups. Although the overall incidence of pneumonitis was significantly higher in the TRT+ICIs group (12.2% vs. 2.9%, P = 0.003) than in the ICIs-only group, the majority of cases were mild to moderate (grade 1 or 2). Importantly, the incidence of clinically significant grade 3 or higher pneumonitis did not differ substantially between the two groups (2.4% vs. 1.4%; P = 0.841). Collectively, these findings suggest that the safety profile of thoracic radiotherapy combined with immunotherapy is well-tolerated and acceptable.

However, this study has some limitations. First, it was a retrospective, single-center study, which may have introduced some selection bias. Second, the PD-L1 expression status was not available for some patients. Third, in clinical practice, differentiating between radiation pneumonitis and immune-related pneumonitis remains a significant diagnostic challenge due to overlapping radiographic features, frequent concurrent administration of corticosteroids and limitations in retrospective study. Consequently, this study was virtually impossible to accurately distinguish between these two types of pneumonitis. Further well-designed randomized controlled trials are necessary to investigate the synergistic effects of RT and ICIs.

Our study supports the efficacy and safety of combining thoracic radiotherapy with immunotherapy in patients with advanced NSCLC. Additionally, this study delineates specific patient characteristics that may predict a favorable response to the proposed combination therapy, thereby providing a foundation for future research.