Radiotherapy improves outcomes in patients with NSCLC receiving nivolumab.

Introduction
Despite significant advances in treatment in recent years, non–small cell lung cancer (NSCLC) remains a leading cause of cancer-related death, with most patients diagnosed at an advanced, incurable stage [1, 2].
In recent years, with the understanding of tumor immune evasion, multiple immune checkpoints for cancer immunotherapy treatment have been identified, including PD1/PD-L1, CTLA4, etc. [3]. Programmed cell death-1 (PD-1) inhibitors and programmed cell death ligand 1 (PD-L1) inhibitors, CTLA4 inhibitors are an effective treatment option in metastatic cancers and provide survival benefits [4, 5]. Nivolumab is a fully human IgG4 programmed death 1 (PD-1) immune checkpoint inhibitor (ICI) antibody indicated for the treatment of NSCLC [6, 7]. In two randomized, open-label phase III studies (CheckMate 017 and CheckMate 057), compared with docetaxel, nivolumab improved overall survival (OS) and safety in previously treated patients with advanced squamous and nonsquamous NSCLC [8, 9]. However, despite the widespread use of ICIs, only select patients benefit from ICIs, and multiple resistance mechanisms impair immune responses [10].
Given that immunotherapy is now a crucial component of the fight against cancer, additional efforts are needed to understand how it can best be combined with surgery, chemotherapy (CT), and RT [11]. RT, usually administered to alleviate local symptoms, can also induce tumour-specific immune effects both within and outside the irradiated field; this process is known as the abscopal effect [2]. RT can support tumour-specific immunity and synergize with ICIs to improve tumour control [2]. However, studies in the literature have shown that adding RT to ICI treatment has no survival benefit [12]. A review of the literature indicated that adding palliative RT to nivolumab is safe and feasible in some studies, but it has not been associated with increased efficacy [2]. Furthermore, studies have reported that the effects of RT within the tumour microenvironment (TME) may promote treatment resistance and regional or distant recurrence [13].
Our study aimed to evaluate the effect of prior RT on progression-free survival and overall survival in patients with advanced NSCLC treated with nivolumab.

Methods

Patients and characteristics
The study included patients aged 18 years and older with advanced clinical stage NSCLC (including squamous and non-squamous disease), ECOG PS ≤ 3, who progressed during or after first-line platinum-based CT, and who received nivolumab alone as second-line therapy. The patient selection process, inclusion and exclusion criteria, and formation of the final study cohort are summarized in Fig. 1. Patients with stage IV disease at diagnosis (AJCC 8th edition) who progressed after first-line platinum chemotherapy and were switched to nivolumab were included in the study. Patients with operable disease at diagnosis who underwent postoperative adjuvant chemotherapy or not, and who developed inoperable recurrence and metastases during follow-up, were also included in the study. All patients received platinum-based chemotherapy either at the time of metastatic diagnosis or during the course of advanced disease prior to nivolumab initiation. Patients with stage IV disease documented histologically or cytologically as stage III (AJCC 8th edition) who received concurrent CRT (≥ 2 cycles; total prescribed radiation dose typically 60 to 66 Gy in 30 to 33 fractions) and who did not receive maintenance ICI therapy and who progressed to nivolumab were also included in the study.

Oligoprogression was defined as the progression of limited (up to 2) metastatic sites during systemic therapy in the context of metastatic disease [14]. Data were recorded for patients who continued treatment with RT during oligoprogression. Additionally, patients who received RT for NSCLC at any time before the first course of nivolumab were coded as receiving RT along with those who received RT during the treatment period. Clinical protocols recorded that patients receiving chemoradiotherapy (CRT) received 60 Gy (Gy) RT in 30 fractions (fx). Patients with local lung progression received 39 Gy 13 fx, and patients with bone and brain metastases received 30 Gy 10 fx. Liver metastases received 10 Gy 4 fx stereotactic body radiotherapy (SBRT), and lymph node metastases received 5 Gy 5 fx.
Patients who received radiotherapy for NSCLC at any time before the initiation of nivolumab were classified as having a history of radiotherapy. In addition, patients who developed oligoprogression during nivolumab treatment and received radiotherapy while continuing nivolumab were also included in this group. Due to the retrospective design of the study and the heterogeneity of radiotherapy indications and timing, a predefined temporal cut-off analysis was not performed.
Data from patients followed at the oncology clinic of Balıkesir Atatürk City Hospital between January 1, 2015, and January 1, 2025, were retrospectively obtained. Data from 273 patients treated with nivolumab for lung cancer were analyzed. Molecular analyses were performed on formalin-fixed, paraffin-embedded (FFPE) tumor tissues obtained from patients via biopsy or surgical resection. Following DNA and RNA extraction, genetic analyses were performed using next-generation sequencing (NGS). The Illumina NextSeq 550 platform was used for analyses, and the QIAGEN Online kit was used as the targeted panel. PDL-1 scores were determined using the Dako kit and the PDL-1 IHK 28 − 8 pharmDx assay.
The study adhered to the tenets of the Declaration of Helsinki. The protocol was approved by the local institutional review boards. The authors used an opt-out approach approved by the local institutional review board as a substitute for written informed consent.

Statistical analysis
PFS was calculated as the time from the first nivolumab treatment date to the date of progression or the last follow-up date for patients without progression, and OS was calculated as the time from the first nivolumab treatment date to death or the last follow-up date for surviving patients.
Study analysis results are presented as numbers, percentages, and median (minimum–maximum range) values. Chi-square (Fisher’s Exact Test for inappropriate cases) was used to compare patient characteristics between groups that received and did not receive radiotherapy. Kaplan‒Meier survival analysis was used to analyse survival, along with estimated median survival times, and the log-rank method was used to compare variables. Univariate PFS and OS analyses were performed using the Cox regression model. Variables with a p value < 0.05 in univariate analysis were modelled using the forwards stepwise (likelihood ratio) method in multivariate Cox regression analysis. Findings with a p value < 0.05 were considered to be statistically significant at a Type 1 error level of less than 5%. All the statistical evaluations were performed with SPSS 24.0 (SPSS Inc., Chicago, IL, USA) and R software (R Core Team, 2024).

Ethical approval
This study was approved by the Balikesir Ataturk City Hospital’s Non-Invasive Clinical Research Ethics Committee (Approval No: E-30041352-514.19.99-272344044 2025/03/37 Date: 20.03.2025).

Patient consent
Since the study was a retrospective archive search, informed consent was not obtained from the patients.

Results

Patient characteristics
A total of 191 patients were included in the study. The median age of the patients was 64 years (min: 37; max: 83). A total of 47.6% (n = 91) of the patients had adenocarcinoma, 44.0% had squamous cell carcinoma (SCC) histology, and 8.4% (n = 16) had other histologies (adenosquamous, large cell carcinoma.). Adenosquamous carcinoma histology was detected in 6.2% (n = 12) of the patients and large cell carcinoma histology in 2% (n = 4) of the patients.
A total of 48.1 (n = 92) of patients had stage IV disease at diagnosis. A total of 51.8% (n = 99) of patients had stage I, II, or III disease at diagnosis. These patients subsequently developed progression and metastases. A total of 15.7% (n = 30) patients had undergone primary lung surgery (Table 1). According to TNM staging (AJCC 8th edition), 3.1% (n = 6) of patients were diagnosed with stage I disease, 12.5% ​​(n = 24) with stage II disease, and 36.1% (n = 69) with stage III disease.

Radiotherapy characteristics
A total of 41.8% (n = 80) had received radiation therapy to the tumour tissue at least once during any treatment period since diagnosis. A total of 34.5% (n = 66) had received CRT. Before nivolumab treatment in the metastatic setting, 9.4% (n = 18) of patients received RT to the bones, 10.9% (n = 21) to the brain, 4.1% (n = 8) to the lungs, 1% (n = 2) to the liver, and 1% (n = 2) to the lymph nodes. In 14 patients (7.3%) who developed oligoprogression while receiving nivolumab treatment, RT was administered to the oligoprogressive areas, and nivolumab treatment was continued in these patients. Among the 80 patients who received radiotherapy, 66 (82.5%) received radiotherapy before nivolumab initiation, while 14 (17.5%) received radiotherapy during nivolumab treatment due to oligoprogression.
Among these patients, RT was administered to bone metastases in 3 patients (1.5%), to the brain in 4 patients (2%), to the primary site in 6 patients (3.1%), and to the liver in 1 patient (0.5%). Patients who received RT during oligoprogression received conventional hypofractionated RT, and the goal of RT was symptom control. Radiotherapy dose and fractionation details are summarized as follows: patients who received definitive chemoradiotherapy were treated with total doses of 60–66 Gy in 30–33 fractions. Palliative radiotherapy consisted of 39 Gy in 13 fractions for lung lesions, 30 Gy in 10 fractions for bone and brain metastases, 10 Gy in 4 fractions using stereotactic body radiotherapy for liver metastases, and 5 Gy in 5 fractions for lymph node metastases.

Survival outcomes
The estimated PFS for all patients treated with second-line nivolumab was 6.9 months (95% CI: 5.2–8.6 months). The estimated OS after second-line therapy was 14.1 months (95% CI: 10.3–17.9 months). In patients with oligoprogression who received radiotherapy for progressive lesions (n = 14), the median was not reached; 6 patients progressed, and 8 patients did not progress at the 6-month follow-up. (Fig. 2 )

Patients with a history of palliative RT or definitive CRT before nivolumab treatment had a median PFS of 11.9 months (95% CI: 4.7–19.1 months) and a median OS of 31.8 months (95% CI: 11.9–49.4 months). Those who did not receive RT or CRT had a median PFS of 3.6 months (95% CI: 3.3–3.9 months) and a median OS of 8.3 months (95% CI: 4.7–14.2 months) (p < 0.001 for both) (Figs. 3 and 4).

Prognostic factors
Univariate Cox regression analysis revealed that age (HR 1.54, 95% CI 1.02–2.34, p = 0.042), ECOG performance status (HR 2.58, 95% CI 1.70–3.91, p < 0.001), BSA (HR 0.60, 95% CI 0.41–0.86, p = 0.006), stage at diagnosis (HR 1.98, 95% CI 1.38–2.83, p < 0.001), primary surgery (HR 0.56, 95% CI 0.33–0.96, p = 0.036), history of RT (HR 0.36, 95% CI 0.25–0.51, p < 0.001), presence of brain metastases (HR 1.77, 95% CI 1.07–2.92, p = 0.026), adrenal metastases (HR 3.53, 95% CI 2.12–5.87, p < 0.001), and liver metastases (HR 1.99, 95% CI 1.17–3.38, p = 0.011) were significant prognostic factors for PFS. ECOG performance status (HR: 2.75, 95% CI 1.76–4.31, p < 0.001), stage at diagnosis (HR: 2.06, 95% CI 1.39–3.06, p < 0.001), primary surgery (HR: 0.46, 95% CI 0.25–0.86, p = 0.015), history of RT (HR: 0.39, 95% CI 0.26–0.58, p < 0.001), brain metastasis (HR: 2.38, 95% CI 1.42–3.98, p = 0.001), adrenal metastasis (HR: 3.75, 95% CI 2.25–6.26, p < 0.001), liver metastasis (HR: 1.80, 95% CI 1.02–3.18, p = 0.041), and TP53 mutation (HR: 0.49, 95% CI 0.25–0.98, p = 0.043) were found to be significant prognostic factors for OS. No significant associations were found between other factors and PFS or OS (Table 2).

ECOG performance status, disease stage at diagnosis, primary surgery, history of RT, brain metastasis, adrenal metastasis, and liver metastasis were included in the multivariate analysis for both progression-free survival (PFS) and overall survival (OS). Additionally, age and BSA were included in the analysis of PFS, whereas TP53 mutation status was included in the OS analysis. ECOG (HR: 1.94, 95% CI: 1.26-3.00, p = 0.003), RT history (HR: 0.44, 95% CI 0.30–0.67, p < 0.001), and the presence of adrenal metastasis (HR: 2.05, 95% CI: 1.20–3.52, p = 0.009) constituted a model for PFS. ECOG performance status (HR: 1.85, 95% CI: 1.15–2.98, p = 0.011), history of primary surgery (HR: 0.49, 95% CI: 0.26–0.94, p = 0.032), history of RT (HR: 0.46, 95% CI: 0.30–0.70, p < 0.001), brain metastasis (HR: 2.50, 95% CI: 1.46–4.30, p = 0.001), and adrenal metastasis (HR: 2.52, 95% CI: 1.46–4.34, p = 0.001) constituted a prognostic model for OS (Table 3).

Site-specific radiotherapy analysis
Kaplan Meier analysis was used to compare patients who received brain radiotherapy with those who received other-site radiotherapy. The estimated PFS for patients who received brain radiotherapy was 3.3 months (95% CI: 1.98–4.62), and for patients who received other-site radiotherapy, the estimated PFS was 13.1 months (95% CI: 6.83–15.17) (p < 0.001). The estimated OS for those with brain metastases who received radiotherapy was 4.7 months (95% CI: 3.98–5.42), and the estimated OS for those who received RT for other-site metastases was 32.7 months (95% CI: 22.05–43.35) (p < 0.001).

Discussion
In our study, patients with a history of palliative or definitive CRT before or during nivolumab treatment had significantly longer median PFS (11.9 months) and median OS (31.8 months) than patients who did not receive RT (median PFS of 3.6 months and median OS of 8.3 months). In our study, PFS and OS were longer in patients who received RT compared with those who did not; however, this observation represents an association rather than evidence of a causal treatment effect. Consistent with these findings, multivariate analysis identified ECOG performance status, history of radiotherapy, and metastatic burden as key prognostic factors influencing survival outcomes. In the site-specific analysis, patients who received radiotherapy for brain metastases had significantly shorter PFS and OS compared with those who received radiotherapy for extracranial sites. This finding should not be interpreted as reduced intrinsic efficacy of immune checkpoint inhibitors in patients with brain metastases. Previous studies have shown that ICIs can provide clinical benefit in NSCLC patients with brain metastases, without clear evidence of inferior efficacy compared with patients without brain involvement [15]. Rather, brain metastases are generally considered a negative prognostic factor in NSCLC [16]. From a biological perspective, brain metastases exhibit a distinct immune microenvironment characterized by immune privilege and blood–brain barrier–related immune modulation, as well as differences in immune cell infiltration compared with extracranial metastatic sites, which may influence local treatment responses [17, 18]. Brain metastases have been shown to exhibit site-specific immune characteristics, including altered immune cell infiltration and local immune suppression compared with extracranial metastases [19].
In the phase 1 KEYNOTE-001 study, patients who had received RT for NSCLC before receiving pembrolizumab had significantly longer progression-free survival and overall survival than patients who had not received prior RT [15]. RT kills both cancerous and normal cells through irreparable DNA damage, leading to the release of tumour-associated antigens and damaging the tumour microenvironment [10]. Shaverdian et al. [15] attributed the longer PFS and OS in RT-treated patients to the release of antigenic peptides from tumours due to RT, increased antigen presentation due to the activation of dendritic cells, and increased antitumour T-cell activity. In summary, as noted in the study by Bozorgmehr et al. [2], RT promotes proinflammatory signalling. McLaughlin et al. [17] described this phenomenon as turning immunologically “cold” tumours into “hot” tumours. Formenti and Demaria [18] described it as transforming tissue into an in situ graft but reported that the process depends on many variables, including the host’s immune status, immunogenetic profile, tumour radiosensitivity and degree of genomic instability, and the type of cell death achieved. In our study, the prolonged PFS and OS results in patients with a history of palliative RT or definitive CRT before and during nivolumab treatment are consistent with the literature.
In our study, the estimated PFS with second-line nivolumab treatment was 6.9 months, and the estimated OS was 14.1 months, whereas the median was not reached in patients with oligoprogressive lesions who received RT. For limited metastatic disease, emerging data support local ablative therapies such as surgery or RT, but the optimal management of oligoprogression remains controversial because of the lack of prospective data [20]. However, the potential benefit of immunomodulatory radiotherapy appears to depend largely on patient- and tumor-specific factors rather than oligoprogression status alone. Despite the inherent limitations of its retrospective design, the present study provides clinically relevant insights. Our findings suggest that local ablative radiotherapy to oligo-progressive lesions during immune checkpoint inhibition may contribute to prolonged disease control by allowing continuation of effective systemic therapy. In this context, radiotherapy may serve as a strategy to overcome localized resistance while preserving the benefit of immunotherapy, which is consistent with previously reported evidence in the setting of oligoprogressive disease. In our study, the retrospective design and heterogeneity of disease burden precluded a dedicated analysis of oligoprogression, and therefore these findings should be interpreted with caution and considered hypothesis-generating.
Beyond disease burden, increasing evidence suggests that patient- and tumor-specific factors may influence the immunomodulatory effects of radiotherapy. Metabolic and nutritional status, tumor mutational burden, and immune-related biomarkers have been reported to affect the synergy between radiotherapy and immune checkpoint inhibitors. In this context, our univariate analysis demonstrated a significant association between body surface area (BSA) and progression-free survival, suggesting that host-related metabolic or nutritional factors may influence treatment outcomes in patients receiving radiotherapy and immune checkpoint inhibition. Although BSA is an indirect and imperfect surrogate of metabolic status, this finding should be interpreted as exploratory and hypothesis-generating, and is consistent with prior reports linking metabolic and nutritional factors with immunotherapy efficacy [21, 22]. The lack of consistent clinical efficacy of such approaches highlights the unmet need to identify the subset of patients most likely to benefit from the immunomodulatory effects of RT [21]. As noted by Chen et al. [22], even individual metabolic and nutritional status influences this synergy. Huang et al. [21] focused on cold tumours defined by low tumour mutational burden (TMB) (< 300 mutations per exome), lack of PDL1 expression, or the presence of mutations in the Wnt pathway. They demonstrated that radioimmunotherapy can overcome immunotherapy resistance in these tumours through its association with mutation-associated neoantigen (MANA)-reactive T-cell responses [21]. Preclinical studies have shown that PD-L1 expression increases in tumour cells after RT, leading to a synergistic antitumour effect of RT and PD-L1 blockade [23].
In their pooled analysis of the PEMBRO-RT and MDACC studies, Theelen et al. [24] reported that adding RT to pembrolizumab significantly improved responses and outcomes in patients with metastatic NSCLC. Similarly, a study by Bassanelli et al. [25] revealed that the combination of irradiation with nivolumab in the treatment of advanced NSCLC contributes to OS. In a secondary analysis of the KEYNOTE-001 trial by Shaverdian et al., the median interval between RT and pembrolizumab was 9.5 months (range, 1 to 106 months) [15]. This finding supports the idea that RT may cause persistent immunological changes that continue to influence treatment responses long after a course of radiation. Similarly, Jokimäki et al. reported that the timing or purpose of RT had no significant impact on time to progression with ICI therapy, suggesting that the immunomodulatory effects of RT may be long-lasting and biologically sustainable [26]. A study by Gagé et al. [27] also revealed that in cases of oligoprogression in patients with metastatic NSCLC, the combination of focal RT with continuation of nivolumab treatment leads to an increase in PFS. Our study results are consistent with the literature in this regard. However, as emphasized by Barker et al. [13] in their study, RT inflammation may facilitate tumour recurrence by inducing a wound healing response characterized by cancer-associated fibroblast (CAF) modulation and extracellular matrix (ECM) remodelling.
In the PEMBRO-RT study, the first dose of pembrolizumab was given sequentially less than 1 week after the last dose of RT (24 Gy in three fractions), whereas in the MDACC study, pembrolizumab was given concurrently with the first dose of RT (50 Gy in four fractions or 45 Gy in 15 fractions) [24].
There are several limitations in this study. In our study, patients who received RT before and after nivolumab were included. The heterogeneity of the study population is a limitation of the our study. Another limitation of our study is that it is retrospective and single-centered. Additionally, one of the limitations of our study is the variable time between RT and nivolumab administration in patients. And, it should not be overlooked that RT treatment may have been planned for patients with better ECOG PS. Despite these limitations, our findings are consistent with the existing literature. Due to the retrospective, non-randomized design, confounding by indication cannot be excluded. Patients receiving radiotherapy may represent a selected subgroup with more favorable baseline characteristics, and the observed associations should therefore be interpreted with caution.
Nivolumab is no longer a standard second-line treatment, as immunotherapy is now used in the first-line setting. Nevertheless, in Türkiye, ICIs were recently reimbursed by the social insurance institution for first-line treatment. Globally, not every patient has access to ICIs in first-line care. Therefore, while not standardized, we believe there are still many patients worldwide who access ICIs in second-line care. Our study results may be meaningful in guiding the treatment of these patient groups. In addition, the authors estimate that RT before and during treatment may improve outcomes in patients receiving ICI as first-line treatment, and they believe their study may provide insights into future studies on this subject.
Numerous studies are ongoing to better understand how to optimize combinations of RT and ICIs [28]. Careful patient selection is essential, and further studies are needed to better select those most likely to benefit.

Conclusion
In this retrospective single-center study, radiotherapy administered before or during nivolumab treatment was associated with improved progression-free and overall survival in patients with advanced non-small cell lung cancer. However, due to the non-randomized design and the potential for confounding by indication, causality cannot be inferred, and these findings should be interpreted with caution. Despite these limitations, our results are consistent with existing literature suggesting a potential interaction between radiotherapy and immune checkpoint inhibitors. Prospective, randomized studies are needed to clarify the true therapeutic contribution of radiotherapy, optimize treatment sequencing, and identify patients most likely to benefit from this combined approach.