Association of salvage radiotherapy with survival outcomes in extracranial oligoprogression after first-line therapy for extensive-stage small-cell lung cancer: a multicenter retrospective cohort study

Abstract

Background:

While first-line systemic therapy is the standard of care for extensive-stage small-cell lung cancer (ES-SCLC), therapeutic failure remains nearly universal, with survival heavily dictated by progression patterns. Extracranial oligoprogression represents a distinct treatment failure phenotype lacking standardized management.

Objectives:

To evaluate the efficacy and survival impact of salvage radiotherapy (RT) in ES-SCLC patients experiencing extracranial oligoprogression after first-line therapy.

Design:

A multicenter retrospective cohort study.

Methods:

We retrospectively enrolled patients with ES-SCLC developing extracranial oligoprogression after first-line therapy from three cancer centers in China between January 2020 and March 2025. Patients were stratified into RT and non-RT groups based on the administration of radiotherapy post-progression. Propensity score matching (PSM) was employed to balance baseline characteristics. Primary endpoints included second-line overall survival (OS2) and progression-free survival (PFS2), estimated via Kaplan–Meier analysis and compared using log-rank tests.

Results:

A total of 232 Chinese patients experienced extracranial oligoprogression, with a median time to progression of 6.1 months (range: 1.0–37.8 months). In the matched cohort, RT significantly prolonged median PFS2 (6.1 vs 3.3 months; p = 0.007) and OS2 (13.4 vs 8.4 months; p = 0.029). Subgroup analyses revealed more pronounced survival benefits in patients characterized by prior immunotherapy, progression of existing lesions, intrathoracic progression, and delayed progression (PFS1 >6 months). Furthermore, among patients who received RT, continuation of first-line systemic therapy was associated with superior PFS2 (9.6 vs 5.4 months; p = 0.017) and OS2 (14.4 vs 13.4 months; p = 0.030) compared with switching to an alternative systemic therapy.

Conclusion:

This multicenter study is one of the first to indicate that radiotherapy is associated with a survival benefit in patients with ES-SCLC and extracranial oligoprogression. Notably, combining radiotherapy with continuation of first-line systemic therapy offered greater clinical utility than switching to an alternative systemic therapy.

Plain language summary

Salvage radiotherapy for patients with advanced small cell lung cancer who have limited new tumor growth after initial treatment

Advanced small cell lung cancer is a fast-growing type of cancer. Many patients respond well to initial chemotherapy or immunotherapy, but some later develop a small number of new tumors outside the brain. There is no clear best treatment for this stage. We studied 232 such patients from three major cancer centers in China. We compared those who received radiotherapy (targeted radiation) with those who did not, using statistics to balance patient differences like age and health. We found that patients who had radiotherapy lived longer without their cancer worsening and had better overall survival. The benefit was greatest for those who had immunotherapy first, whose cancer grew in existing tumors, in the chest, or who stayed healthy for over 6 months after initial treatment. Continuing the first treatment alongside radiotherapy also worked better than switching to a new one. These results show radiotherapy is a valuable option for these patients, helping doctors choose the best care to improve outcomes.

Introduction

Extensive-stage small-cell lung cancer (ES-SCLC) is a highly aggressive neuroendocrine malignancy characterized by rapid proliferation, early systemic dissemination, and a poor long-term prognosis. Historically, the standard of care was limited to platinum-etoposide chemotherapy.^1–3 In 2019, IMpower133 ⁴ and CASPIAN ⁵ trials established the use of programmed death-ligand 1 (PD-L1) inhibitors in combination with platinum-etoposide chemotherapy as the new first-line standard of care. While this chemoimmunotherapy (Chemo-IO) regimen has significantly improved survival outcomes, disease progression remains nearly universal, and optimal management following first-line failure remains a critical clinical challenge.

In recent years, the therapeutic landscape of ES-SCLC has evolved rapidly in both maintenance and salvage settings. The phase III IMforte trial showed that lurbinectedin plus atezolizumab significantly improved progression-free and overall survival (OS) as first-line maintenance compared with atezolizumab alone. ⁶ Notably, the delta-like ligand 3-targeted bispecific T-cell engager tarlatamab has recently established a new standard of care for relapsed SCLC based on the landmark DeLLphi-301 trial. ⁷ In addition, novel antibody–drug conjugates (ADCs) such as ifinatamab deruxtecan (I-DXd) have shown significant promise in clinical trials.^8–10 Nevertheless, most of these novel agents are restricted to later lines and remain unavailable in routine clinical practice in many regions, including China.

With the evolution of modern first-line therapeutic strategies, the prognosis for patients with ES-SCLC has improved significantly. Although systemic progression remains the predominant failure pattern after first-line treatment, an increasing proportion of patients present with oligometastatic recurrence or oligoprogression in clinical practice. Notably, management strategies for oligoprogression in SCLC often diverge based on the anatomical site of failure. For intracranial progression, stereotactic radiosurgery (SRS) has emerged as a promising approach, with retrospective data demonstrating an encouraging median OS of 11 months for patients with a single brain metastasis (BM).^11,12 In contrast, the clinical role of local therapy specifically for extracranial oligoprogression—a distinct clinical entity with potentially different tumor biology—remains poorly defined. In non-small cell lung cancer, the landmark phase II CURB trial demonstrated that oligoprogression can be effectively controlled by adding stereotactic body radiotherapy (SBRT) to standard-of-care systemic therapy, resulting in a more than fourfold prolongation of progression-free survival (PFS) compared with systemic therapy alone. ¹³ However, evidence specifically addressing oligoprogression in SCLC remains limited. A retrospective study of 20 SCLC patients who received SBRT for extracranial oligoprogression or oligorecurrence demonstrated an impressive 1-year local control rate of 85%; however, distant recurrence remained the predominant failure pattern. Notably, the survival benefit was significantly more pronounced in patients initially diagnosed with limited-stage SCLC (LS-SCLC) compared to those with ES-SCLC (median OS: 23.2 vs 8.6 months). ¹⁴ These findings underscore the potential of local ablative therapy in oligoprogressive SCLC but also suggest its value as a strategic “bridge” to stabilize disease and extend the window of opportunity for next-generation systemic treatments.

Accordingly, this multicenter retrospective cohort study evaluated the efficacy of salvage radiotherapy (RT) in ES-SCLC patients presenting with extracranial oligoprogression after first-line therapy. By analyzing the co-primary endpoints of second-line progression-free survival (PFS2) and overall survival (OS2), we aimed to clarify the role of salvage RT in ES-SCLC management in the Chemo-IO era and provide an evidence-based rationale for integrating local and systemic therapies in this distinct clinical setting.

Materials and methods

Study design and patients

This retrospective, multicenter cohort study was conducted across three independent cancer centers in China: Shandong Cancer Hospital (SDCH); Shanghai Chest Hospital (SHCH); and Tianjin Cancer Hospital (TJCH). Patients with ES-SCLC presenting without baseline BMs who subsequently developed extracranial oligoprogression following first-line therapy were enrolled between January 2020 and March 2025. Oligoprogression was defined as limited disease progression (1–3 progressive metastatic lesions confined to a single organ) occurring after achieving disease control with first-line systemic therapy, while all other non-progressive lesions remained stable or demonstrated response.

Inclusion criteria were as follows: (1) pathological and clinical diagnosis of ES-SCLC according to the 8th tumor-node-metastasis staging system published by the AJCC; (2) age ⩾18 years; (3) Eastern Cooperative Oncology Group (ECOG) performance status (PS) score ⩽2; (4) absence of BMs at initial diagnosis, confirmed by contrast-enhanced cranial MRI or CT scans; (5) receipt of first-line standard therapy, development of extracranial oligoprogression, and subsequent second-line treatment. Exclusion criteria were as follows: (1) ECOG score >2; (2) initial diagnosis of baseline BMs; (3) patients who received a first-line combined radiotherapy strategy; (4) unclear site of progression; (5) development of intracranial progression or extracranial multiprogression after first-line therapy; (6) incomplete follow-up data. Notably, patients with baseline BMs or intracranial progression following first-line therapy were excluded to maintain the homogeneity of the study population and to minimize the confounding survival impact of central nervous system (CNS)-directed local therapies (e.g., whole-brain radiotherapy or SRS).

A total of 1112 ES-SCLC patients were initially enrolled from the three centers (SDCH: 598; SHCH: 269; TJCH: 245). Of these, 232 patients developed extracranial oligoprogression after first-line therapy. Figure 1 presents the patient inclusion process and the exclusion criteria.

Figure 1.

The patient inclusion process and the exclusion criteria.

Treatment protocol

During first-line therapy, patients with ES-SCLC received at least one cycle of EP chemotherapy (etoposide + cisplatin/carboplatin); a subset of patients concurrently received PD-L1/PD-1 inhibitors (including atezolizumab, durvalumab, or sintilimab) as first-line therapy, with maintenance therapy continued in responders or those with stable disease (SD). After developing oligoprogression following first-line treatment, the subsequent systemic therapy administered in conjunction with salvage RT (either concurrently or sequentially) was categorized into two distinct strategies: Continuous Therapy (CTh) and Alternative Therapy (ATh). CTh was defined as the continuation of the initial treatment paradigm, which encompassed either maintaining first-line immunotherapy or re-challenging with the original chemotherapy doublet. Conversely, ATh referred to switching to a different systemic agent (e.g., second-line chemotherapy such as topotecan or taxanes) upon the initiation of salvage RT. This classification enabled comparison between retaining prior systemic treatments and adopting new therapeutic regimens during local consolidative therapy.

Patients were then divided into RT and non-RT groups based on whether they received radiotherapy after first-line progression. Radiotherapy strategies were formulated by a multidisciplinary tumor board based on the NCCN and CSCO guidelines, considering tumor location, size, and proximity to organs at risk. All patients underwent CT simulation positioning with a slice thickness of 3–5 mm. For thoracic or abdominal lesions affected by respiratory motion, four-dimensional CT simulation or respiratory gating techniques were employed. The Gross Tumor Volume was delineated based on planning CT and fused with diagnostic PET/CT or MRI images where available. The Clinical Target Volume and Planning Target Volume were expanded according to institutional protocols. Intensity-modulated radiotherapy or volumetric modulated arc therapy was administered using 6-MV linear accelerators. SBRT was prioritized for small, well-circumscribed oligoprogressive lesions to achieve ablative control.

The prescribed dose and fractionation regimens were individualized based on the site of progression: (1) Intrathoracic progression (Lung and Regional Lymph Nodes): Patients received thoracic radiotherapy with doses ranging from 30 Gy in 10 fractions (3 Gy/f) to 60 Gy in 30 fractions (2 Gy/f). For small peripheral lung nodules suitable for SBRT, a dose of 48–60 Gy in 4–8 fractions was prescribed. (2) Bone metastases: The standard regimen was 30 Gy in 10 fractions. For spinal instability or oligometastatic bone lesions requiring higher biological effective doses, SBRT (e.g., 24–30 Gy in 3–5 fractions) or hypofractionated regimens (e.g., 40 Gy in 20 fractions) were utilized. (3) Abdominal/Visceral progression (Liver, Adrenal gland, Spleen): SBRT was the preferred modality for liver and adrenal oligoprogression, typically delivering 40–50 Gy in 5–10 fractions. For larger lesions or those close to the bowel, a hypofractionated regimen of 45–50 Gy in 15–25 fractions was applied. (4) Soft tissue/Pleura/Skin: These lesions were treated with conventional fractionation (e.g., 30–40 Gy in 10–20 fractions) or electron beam therapy for superficial skin metastases, aiming for local control while minimizing toxicity. Normal tissue constraints strictly adhered to the quantitative analyses of normal tissue effects in the clinic (QUANTEC) recommendations.

Clinical response assessment

Patients were followed until death or the final follow-up deadline. Systemic responses were assessed using contrast-enhanced CT and brain MRI. Tumor status was evaluated every two cycles for 4–6 cycles post-treatment; thereafter, assessments occurred every 6 weeks for the first year and every 2–3 months until progression. Responses were assessed using the Response Evaluation Criteria in Solid Tumors, version 1.1 (RECIST) v1.1.

End points

In this study, primary endpoints included second-line overall survival (OS2, from the initiation of the second-line treatment strategy to death from any cause) and second-line progression-free survival (PFS2, from initiation of second-line treatment to subsequent progression, death, or last follow-up). OS from initial diagnosis was analyzed as a secondary outcome.

Statistical analysis

Baseline characteristics between the RT and non-RT groups were compared using the Chi-square test or Fisher’s exact test, as appropriate. To minimize selection bias and ensure comparability between groups, propensity score matching (PSM) was performed using the nearest-neighbor algorithm with a caliper of 0.02 and a 1:1 matching ratio. Covariates included in the matching model were selected based on clinical relevance and potential prognostic value: gender, age, smoking status, ECOG PS, metastatic sites (lung, liver, kidney/adrenal, and bone), number of first-line EP cycles, first-line immunotherapy status, and best initial response. Covariate balance was assessed after matching.

Survival endpoints (PFS, OS, PFS2, OS2) and the cumulative incidence of extracranial oligoprogressive disease were estimated using the Kaplan–Meier method. Given the adequate balance achieved in baseline characteristics after matching (all p-values >0.05), the primary efficacy endpoints (PFS2 and OS2) were compared directly between the matched groups using the log-rank test. To address potential immortal time bias, a landmark sensitivity analysis was performed. ¹⁵ The landmark was set at 2 months post-progression, and only patients surviving without progression until the landmark were included in this specific analysis. Hazard ratios (HRs) and 95% confidence intervals (CIs) were estimated using the Cox proportional hazards model. All statistical tests were two-sided, and a p-value <0.05 was considered statistically significant. Analyses were conducted using SPSS v26.0 (IBM Corporation, Armonk, NY, USA) and GraphPad Prism v8.0 (GraphPad Software, San Diego, CA, USA).

Reporting guidelines

This retrospective cohort study is reported in accordance with the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement for cohort studies. ¹⁶ The completed STROBE checklist is provided as Supplemental File.

Results

Patient characteristics

Between January 2020 and March 2025, 1112 Chinese patients with ES-SCLC and no baseline BMs were screened across 3 participating centers. Ultimately, 232 patients presenting with extracranial oligoprogression following first-line therapy were enrolled. The median follow-up duration was 16.2 months (range: 3.0–50.9 months). The cohort was stratified based on the receipt of post-progression radiotherapy: 60 patients (25.9%) in the RT group and 172 (74.1%) in the non-RT group. Baseline characteristics are summarized in Table 1.

Table 1.

Baseline characteristics of patients before and after PSM.

Characteristic	Before matching			After matching
	RT (n = 60)	Non-RT (n = 172)	p-Value	RT (n = 53)	Non-RT (n = 53)	p-Value
Gender, n(%)			0.292			1.000
Female	13 (21.67)	27 (15.70)		10 (18.87)	10 (18.87)
Male	47 (78.33)	145 (84.30)		43 (81.13)	43 (81.13)
Age, years, n(%)			0.967			0.845
<65	34 (56.67)	98 (56.98)		30 (56.60)	29 (54.72)
⩾65	26 (43.33)	74 (43.02)		23 (43.40)	24 (45.28)
Smoking status, n(%)			0.647			0.685
No	20 (33.33)	63 (36.63)		18 (33.96)	20 (37.74)
Yes	40 (66.67)	109 (63.37)		35 (66.04)	33 (62.26)
ECOG, n(%)			0.531			1.000
0–1	56 (93.33)	156 (90.70)		49 (92.45)	49 (92.45)
2	4 (6.67)	16 (9.30)		4 (7.55)	4 (7.55)
Baseline lung metastases, n(%)			0.465			0.558
No	58 (96.67)	169 (98.26)		51 (96.23)	52 (98.11)
Yes	2 (3.33)	3 (1.74)		2 (3.77)	1 (1.89)
Baseline liver metastases, n(%)			0.010			0.800
No	50 (83.33)	113 (65.70)		43 (81.13)	44 (83.02)
Yes	10 (16.67)	59 (34.30)		10 (18.87)	9 (16.98)
Baseline bone metastases, n(%)			0.025			0.405
No	45 (75.00)	101 (58.72)		38 (71.70)	34 (64.15)
Yes	15 (25.00)	71 (41.28)		15 (28.30)	19 (35.85)
Baseline kidney/adrenal gland metastases, n(%)			0.008			0.647
No	53 (88.33)	167 (97.09)		51 (96.23)	50 (94.34)
Yes	7 (11.67)	5 (2.91)		2 (3.77)	3 (5.66)
First-line EP cycles, n(%)			0.029			0.566
<4/>6	12 (20.00)	16 (9.30)		6 (11.32)	8 (15.09)
4–6	48 (80.00)	156 (90.70)		47 (88.68)	45 (84.91)
First-line immunotherapy, n(%)			0.958			0.323
No	25 (41.67)	71 (41.28)		24 (45.28)	19 (35.85)
Yes	35 (58.33)	101 (58.72)		29 (54.72)	34 (64.15)
Best efficacy evaluation, n(%)			0.018			0.814
CR	4 (6.67)	1 (0.58)		0 (0.00)	0 (0.00)
PR	44 (73.33)	139 (80.81)		42 (79.25)	41 (77.36)
SD	12 (20.00)	32 (18.61)		11 (20.75)	12 (22.64)

Bold values indicate statistically significant intergroup differences.

ECOG, Eastern Cooperative Oncology Group; EP, etoposide plus platinum; PR, partial response; PSM, propensity score matching; RT, radiotherapy; SD, stable disease.

Before PSM, gender, age, smoking status, ECOG score, first-line immunotherapy, and lung metastasis were balanced between the two groups. The median age was 66 years (range: 39–79). Most patients were male (192/232, 82.8%), and 149 (64.2%) had a smoking history. The majority (204/232, 87.9%) received 4–6 cycles of etoposide plus platinum (EP) chemotherapy; only 12.1% received <4 or >6 cycles. Compared with the RT group, more patients in the non-RT group received 4–6 cycles (90.7% vs 80.0%, p = 0.029). Liver (34.3% vs 16.7%, p = 0.010) and bone metastases (41.3% vs 25.0%, p = 0.025) were unbalanced, with higher proportions in the non-RT group. In addition, the proportion of kidney/adrenal gland metastases was significantly higher in the RT group compared to the non-RT group (11.67% vs 2.91%, p = 0.008). According to RECIST v1.1, 78.9% achieved partial response (PR) and 19.0% SD after first-line therapy, with a significant difference between groups (p = 0.018). After PSM, 53 patients from each group were included in the final analysis, and all baseline characteristics were well balanced (Table 1).

The patterns of post-first-line extracranial oligoprogression

At the data cutoff, 232 patients exhibited extracranial oligoprogression, with a median time to progression of 6.1 months (range: 1.0–37.8 months). The most frequent sites of progression were the lung (109 patients, 47.0%), regional lymph nodes (37 patients, 15.9%), liver (33 patients, 14.2%), bone (33 patients, 14.2%), pleura (10 patients, 4.3%), kidney/adrenal gland (5 patients, 2.2%), skin (3 patients, 1.3%), and spleen (2 patients, 0.9%).

Based on concordance between the first-line oligoprogressive site and baseline metastatic sites, progressive patterns were classified as progression of existing lesions (179/232, 77.2%) or development of new lesions (53/232, 22.8%). Among patients with progression of existing lesions, primary lung progression was the most common (109 patients, 60.9%); among those with development of new lesions, bone progression predominated (17 patients, 32.1%) prior to matching. Figure 2 illustrates the distribution of baseline extracranial metastatic sites and first-line extracranial oligoprogressive lesions among the 232 enrolled patients. During first-line treatment, 136 patients received immunotherapy plus chemotherapy (IO + ChT), and 96 received chemotherapy alone (ChT group). First-line immunotherapy was significantly associated with a reduced incidence of extracranial oligoprogression (HR, 0.76; 95% CI, 0.58–0.99; p = 0.037; Figure S1).

Figure 2.

Schematic illustration of baseline extracranial metastatic sites and first-line extracranial oligoprogressive lesions among the 232 enrolled patients.

Survival outcomes

OS2 and PFS2 in the overall cohort

Among the 232 patients with ES-SCLC who experienced extracranial oligoprogression after receiving first-line treatment, 60 patients were in the RT group, and 172 patients were in the non-RT group. Survival analysis results showed that RT markedly extended PFS2 and translated into longer survival time. Before PSM, the median PFS2 was significantly longer in the RT group (7.3 months; 95% CI, 5.1–9.4) than in the non-RT group (3.6 months; 95% CI, 2.8–4.4; p = 0.015, Figure 3(a)). The median OS2 was 13.4 months (95% CI, 11.4–15.3) and 8.9 months (95% CI, 7.0–10.8) in the RT group and the non-RT group, respectively (p = 0.026, Figure 3(b)). After PSM, among 106 patients with extracranial oligoprogression (53 in the RT group; 53 in the non-RT group), radiotherapy significantly prolonged median PFS2 (6.1 vs 3.3 months; HR, 0.58; 95% CI, 0.38–0.87; p = 0.007, Figure 3(c)) and improved survival (mOS2: 13.4 vs 8.4 months; HR, 0.62; 95% CI, 0.40–0.96; p = 0.029, Figure 3(d)). Cumulative OS from diagnosis showed a consistent trend, with the RT group having significantly longer median OS (20.3 vs 14.6 months; p = 0.028, Figure 4).

Figure 3.

PFS2 (a) and OS2 (b) in patients between the RT group and the non-RT group before PSM; PFS2 (c) and OS2 (d) in patients between the RT group and the non-RT group after PSM.

OS, overall survival; PFS, progression-free survival; PSM, propensity score matching; RT, radiotherapy.

Figure 4.

OS (a) in patients between the RT group and the non-RT group before PSM; OS (b) in patients between the RT group and the non-RT group after PSM.

OS, overall survival; PSM, propensity score matching; RT, radiotherapy.

Sensitivity analysis: 2-Month landmark analysis

To evaluate the impact of potential immortal time bias, a landmark sensitivity analysis was performed on the PSM cohort. The landmark was set at 2 months after the initial first-line progression to account for the time required for radiotherapy planning and initiation. A total of 20 patients who experienced events (3 deaths and 17 disease progressions) within the 2-month landmark period were excluded. In the remaining landmark cohort (n = 86; RT group: 46; non-RT group: 40), the survival benefit of salvage radiotherapy remained statistically significant, with time-zero adjusted to the landmark date. The RT group demonstrated a significantly longer OS2 (median: 12.4 vs 6.8 months; HR, 0.59; 95% CI, 0.36–0.96; p = 0.023) and PFS2 (median: 5.7 vs 2.5 months; HR, 0.60; 95% CI, 0.38–0.96; p = 0.023) compared to the non-RT group (Figure S2). These results confirm that the observed therapeutic benefit of radiotherapy was not primarily driven by immortal time bias.

Subgroups analysis of survival in the matched cohort

To further identify the potential population benefiting from RT after first-line extracranial oligoprogression, we conducted a subgroup analysis of survival, which revealed that the therapeutic benefits in the RT group were highly consistent across the majority of clinical characteristic subgroups in matched patients (Figure 5).

Figure 5.

Subgroup analysis of OS2 (a) and PFS2 (b) after PSM.

OS, overall survival; PFS, progression-free survival; PSM, propensity score matching.

Outcomes in IO + ChT versus ChT alone subgroups

In the subgroup of 63 patients who received first-line immunotherapy, the survival benefits were also significant in the RT group compared to the non-RT group (mPFS2: 7.7 vs 3.1 months, HR, 0.58; p = 0.004, Figure S3(a); mOS2: 14.4 vs 7.4 months, HR, 0.62; p = 0.001, Figure S3(b)). In addition, for 43 patients who received first-line chemotherapy alone, although the median PFS2 in the RT group was numerically superior to that in the non-RT group (6.0 vs 4.5 months, p = 0.423, Figure S3(c)), the difference was not statistically significant. The results of median OS2 showed no significant trend or statistical difference between the two groups (11.1 vs 12.9 months, p = 0.586, Figure S3(d)).

Outcomes in existing lesions versus new lesions progression subgroups

The 106 patients were classified by whether extracranial oligoprogressive sites overlapped with baseline metastatic sites: progression of existing lesions (81, 76.4%) or development of new lesions (25, 23.6%). In the existing-lesion group, progression most commonly involved the primary lung or regional lymph nodes (64/81, 79.0%), followed by liver (12/81, 14.8%) and bone (5/81, 6.2%). Survival analysis showed that in this subgroup, the RT group had significantly longer PFS2 (mPFS2: 7.7 vs 3.2 months, HR, 0.55; 95% CI, 0.36–0.83; p = 0.001, Figure S4(a)) and OS2 (mOS2: 14.4 vs 8.9 months, HR, 0.59; 95% CI, 0.38–0.91; p = 0.009, Figure S4(b)). In contrast, no significant differences in PFS2 or OS2 were observed between groups in the new-lesion subgroup (Figure S4(c) and (d)). Bone was the most frequent site of new progression, occurring in 11 of 25 patients (44.0%).

Outcomes in intrathoracic versus extrathoracic progression subgroups

Among 106 patients with extracranial oligoprogression after first-line therapy, 67 (63.2%) had intrathoracic progression and 39 (36.8%) had extrathoracic progression. In the intrathoracic progression subgroup, the RT group showed greater PFS2 and OS2 benefits than the non-RT group (median PFS2: 7.7 vs 3.1 months; HR, 0.54; 95% CI, 0.35–0.82; p = 0.015; median OS2: 21.6 vs 7.9 months; HR, 0.63; 95% CI, 0.41–0.98; p = 0.019, Figure S5(a) and (b)). In contrast, no significant differences in PFS2 or OS2 were observed in the extrathoracic progression subgroup (median PFS2: 5.2 vs 3.5 months; p = 0.097; median OS2: 10.6 vs 8.4 months; p = 0.654, Figure S5(c) and (d)).

Outcomes in delayed versus early progression subgroups

The 106 matched patients were divided into two subgroups based on whether first-line extracranial oligoprogression occurred after 6 months: delayed progression (PFS1 >6 months, n = 51) or early progression (PFS1 ⩽6 months, n = 55) subgroup. RT was associated with significant survival benefits in the delayed-progression subgroup (median PFS2: 9.6 vs 4.4 months; HR, 0.63; 95% CI, 0.40–0.97; p = 0.009; median OS2: 14.7 vs 7.1 months; HR, 0.57; 95% CI, 0.36–0.92; p = 0.001, Figure S6(a) and (b)), but not in the early-progression subgroup, where no significant differences were observed (median PFS2: 5.4 vs 3.0 months; p = 0.511; median OS2: 12.5 vs 9.5 months; p = 0.643, Figure S6(c) and (d)).

Second-line combined systemic treatment strategy

To identify the most effective systemic strategy when combined with radiotherapy, we analyzed the RT group (n = 60 before PSM), which consisted of a continuous systemic therapy subgroup (CTh + RT, n = 32) and an ATh subgroup (ATh + RT, n = 28). Following PSM, the matched RT cohort (n = 53) included 26 patients in the CTh + RT group and 27 patients in the ATh + RT group (Figure 1).

In the matched cohort analysis, the CTh + RT strategy demonstrated superior outcomes. Regarding the choice of systemic partner for salvage RT, the detailed regimens for the CTh (n = 26) and ATh (n = 27) groups are summarized in Table 2. In the CTh group, 53.85% (14/26) of patients received immunotherapy maintenance. In the ATh group, the most common regimens were topotecan (44.44%) and Albumin-bound paclitaxel (37.04%). Survival analysis of the matched cohort showed that CTh + RT was associated with significantly longer PFS2 and OS2 compared to the ATh alone group (p < 0.001 for both, Figure 6). Furthermore, when compared directly to switching systemic therapy combined with RT (ATh + RT), the CTh + RT group continued to show superior median PFS2 (9.6 vs 5.4 months, HR, 0.48; 95% CI, 0.26–0.89; p = 0.017, Figure 6(c)) and median OS2 (14.4 vs 13.4 months, HR, 0.47; 95% CI, 0.24–0.94; p = 0.030, Figure 6(d)). These results suggest that maintaining the original systemic regimen concurrently with radiotherapy may be the optimal strategy for oligoprogression.

Table 2.

Details of systemic therapy regimens combined with salvage radiotherapy in the PSM cohort.

Regimen categories	Specific regimens	Patients, n (%)
CTh (n = 26)	Immunotherapy maintenance	14 (53.85)
	Re-challenge with first-line chemotherapy	8 (30.77)
	Re-challenge with first-line chemoimmunotherapy	4 (15.38)
ATh (n = 27)	Topotecan	12 (44.44)
	Albumin-bound paclitaxel monotherapy	10 (37.04)
	Irinotecan	3 (11.11)
	Anlotinib	2 (7.41)

ATh, alternative therapy; ChT, continuous therapy; PSM, propensity score matching.

Figure 6.

PFS2 (a) and OS2 (b) in patients among CTh + RT, ATh + RT, and ATh alone group before PSM. PFS2 (c) and OS2 (d) in patients among CTh + RT, ATh + RT, and ATh alone group after PSM.

ATh, alternative therapy; ChT, continuous therapy; OS, overall survival; PFS, progression-free survival; PSM, propensity score matching; RT, radiotherapy.

Multivariate analysis of factors associated with survival in the matched cohort

To identify independent prognostic factors for OS2 and PFS2 within the propensity score-matched cohort, univariate and multivariate Cox proportional hazards analyses were conducted, as detailed in Tables 3 and 4. Given the limited sample size in the matched cohort, variables for the final multivariate models were carefully selected based on their significance in the univariate analysis and their clinical relevance, aiming to prevent model overfitting.

Table 3.

Univariate and multivariate analyses of factors influencing second-line overall survival in patients with extracranial oligoprogression after first-line treatment after PSM.

Variables	p	Univariate analysis	p	Multivariate analysis
		HR (95% CI)		HR (95% CI)
Gender
Female		1.00 (Reference)
Male	0.759	1.11 (0.58–2.09)
Age, years
<65		1.00 (Reference)
⩾65	0.228	1.31 (0.85–2.01)
ECOG
2		1.00 (Reference)		1.00 (Reference)
0–1	0.006	0.33 (0.15–0.73)	0.001	0.24 (0.11–0.54)
Baseline lung metastases
No		1.00 (Reference)
Yes	0.301	0.35 (0.05–2.54)
Baseline liver metastases
No		1.00 (Reference)
Yes	0.009	1.85 (1.17–2.94)
Baseline bone metastases
No		1.00 (Reference)
Yes	0.012	1.75 (1.13–2.71)
Baseline kidney/adrenal gland metastases
No		1.00 (Reference)
Yes	0.951	0.98 (0.45–2.12)
Radiotherapy after progression
No		1.00 (Reference)		1.00 (Reference)
Yes	0.031	0.62 (0.40–0.96)	0.053	0.64 (0.41–1.01)
Intrathoracic progression
No		1.00 (Reference)
Yes	0.033	0.62 (0.40–0.96)
Progression of existing lesions
No		1.00 (Reference)
Yes	0.170	0.71 (0.44–1.16)
First-line progression-free survival, months
⩽6		1.00 (Reference)
>6	0.351	0.81 (0.53–1.26)

Bold values indicate statistically significant intergroup differences.

CI, confidence interval; ECOG, Eastern Cooperative Oncology Group; HR, hazard ratio; PSM, propensity score matching.

Table 4.

Univariate and multivariate analyses of factors influencing second-line PFS in patients with extracranial oligoprogression after first-line treatment after PSM.

Variables	p	Univariate analysis	p	Multivariate analysis
		HR (95% CI)		HR (95% CI)
Gender
Female		1.00 (Reference)
Male	0.115	1.66 (0.88–3.12)
Age, years
<65		1.00 (Reference)
⩾65	0.723	0.93 (0.61–1.41)
ECOG
2		1.00 (Reference)
0–1	0.719	0.87 (0.39–1.90)
Baseline lung metastases
No		1.00 (Reference)
Yes	0.385	0.42 (0.06–3.00)
Baseline liver metastases
No		1.00 (Reference)		1.00 (Reference)
Yes	0.003	1.99 (1.27–3.12)	0.010	1.82 (1.15–2.87)
Baseline bone metastases
No		1.00 (Reference)
Yes	0.069	1.47 (0.97–2.21)
Baseline kidney/adrenal gland metastases
No		1.00 (Reference)
Yes	0.799	0.91 (0.42–1.96)
Radiotherapy after progression
No		1.00 (Reference)		1.00 (Reference)
Yes	0.008	0.58 (0.38–0.87)	0.027	0.62 (0.41–0.95)
Intrathoracic progression
No		1.00 (Reference)
Yes	0.089	0.70 (0.46–1.06)
Progression of existing lesions
No		1.00 (Reference)
Yes	0.716	0.92 (0.57–1.47)
First-line PFS, months
⩽6		1.00 (Reference)
>6	0.006	0.56 (0.37–0.85)

CI, confidence interval; ECOG, Eastern Cooperative Oncology Group; HR, hazard ratio; PFS, progression-free survival; PSM, propensity score matching.

Bold values indicate statistically significant intergroup differences.

Regarding OS2 (Table 3), univariate analysis indicated that ECOG PS, baseline liver metastases, baseline bone metastases, intrathoracic progression, and radiotherapy after progression were significantly associated with survival outcomes (all p < 0.05). In the multivariate model, ECOG PS 0–1 was confirmed as a highly significant independent predictor of prolonged OS2 (HR, 0.24; 95% CI, 0.11–0.54; p = 0.001). Notably, radiotherapy after progression demonstrated a strong and clinically meaningful trend toward improved OS2 (HR, 0.64; 95% CI, 0.41–1.01; p = 0.053), corresponding to a 36% reduction in the risk of mortality.

In terms of PFS2 (Table 4), univariate analysis revealed that baseline liver metastases, first-line progression-free survival (PFS1), and radiotherapy after progression were significantly correlated with disease progression (all p < 0.05). After adjusting for key clinical covariates, the multivariate analysis confirmed radiotherapy after progression as an independent prognostic factor for prolonged PFS2 (HR, 0.62; 95% CI, 0.41–0.95; p = 0.027). Conversely, baseline liver metastasis was identified as an independent predictor of inferior PFS2 (HR, 1.82; 95% CI, 1.15–2.87; p = 0.010).

Safety and toxicity

All adverse events (AEs) were assessed per CTCAE v5.0 and summarized in Table 5. Safety outcomes were stratified into Chemo-IO and chemotherapy-alone subgroups to evaluate the impact of immune checkpoint inhibitors on radiotherapy-related toxicity.

Table 5.

Treatment-related AEs in the propensity score-matched cohort (CTCAE v5.0).

AEs, n(%)	RT group (n = 53)			Non-RT group (n = 53)			p-Value (RT vs non-RT)
	Total (n = 53)	Chemo alone (n = 24)	Chemo-IO (n = 29)	Total (n = 53)	Chemo alone (n = 19)	Chemo-IO (n = 34)
Radiotherapy-related AEs				–	–		–
Radiation pneumonitis				–	–		–
Grade 1–2	5 (9.43)	2 (8.33)	3 (10.34)	–	–		–
Grade ⩾3	2 (3.77)	1 (4.17)	1 (3.45)	–	–		–
Radiation esophagitis				–	–		–
Grade 1–2	3 (5.66)	2 (8.33)	1 (3.45)	–	–		–
Grade ⩾3	1 (1.90)	0 (0.00)	1 (3.45)	–	–		–
Radiation dermatitis (thoracic)	4 (7.55)	2 (8.33)	2 (6.90)	–	–		–
Systemic therapy-related AEs
Fatigue	21 (39.62)	8 (33.33)	13 (44.83)	19 (35.85)	8 (42.11)	11 (32.35)	0.652
Leukopenia	15 (28.30)	7 (29.17)	8 (27.59)	13 (24.53)	4 (21.05)	9 (26.47)	0.621
Thrombocytopenia	8 (15.09)	3 (12.50)	5 (17.24)	7 (13.21)	2 (10.53)	5 (14.71)	0.785
Nausea/vomiting	10 (18.87)	6 (25.00)	4 (13.79)	9 (16.98)	4 (21.05)	5 (14.71)	0.800
Anorexia	7 (13.21)	5 (20.83)	2 (6.90)	6 (11.32)	3 (15.79)	3 (8.82)	0.759
Total AEs
Any grade	36 (67.92)	16 (66.67)	20 (68.97)	30 (56.60)	11 (57.89)	19 (55.88)	0.213
Grade ⩾3	9 (16.98)	3 (12.50)	6 (20.69)	7 (13.21)	3 (15.79)	4 (11.76)	0.578
Grade 5 AEs (treatment-related death)	0 (0.00)	0 (0.00)	0 (0.00)	0 (0.00)	0 (0.00)	0 (0.00)	–

AEs, adverse events; IO, immunotherapy; RT, radiotherapy.

In the PSM cohort, salvage radiotherapy was well tolerated. No significant differences in all-grade AEs (p = 0.213) or Grade ⩾3 AEs (p = 0.578) were found between RT and non-RT groups. No Grade 5 treatment-related deaths occurred, supporting the safety of salvage radiotherapy in this cohort.

In the RT subgroup, concurrent immunotherapy did not increase severe radiation-related toxicity. Grade 3 radiation pneumonitis rates were similar in Chemo-IO (3.45%, 1/29) and Chemo-alone subgroups (4.17%, 1/24), with low and balanced Grade 1–2 pneumonitis rates. All radiation-induced esophagitis and dermatitis were manageable with symptomatic and corticosteroid treatment, with no permanent organ injury.

Systemic toxicity profiles were comparable across subgroups. Fatigue was the most frequent AE, with a slightly higher rate in the RT + Chemo-IO group. Grade ⩾3 AEs in this subgroup were limited to reversible hematological events responsive to supportive care. No severe immune-radiation overlapping toxicities or permanent systemic therapy discontinuation cases were observed.

Discussion

To our knowledge, this study is one of the first multicenter cohort studies to demonstrate a survival benefit of salvage radiotherapy in patients with ES-SCLC presenting with extracranial oligoprogression following first-line therapy. Our findings indicate that radiotherapy significantly prolonged both OS2 and PFS2 in this population. These results address a critical evidence gap regarding second-line management strategies for ES-SCLC and underscore the therapeutic potential of radiotherapy in the setting of oligoprogression.

In this study, we strictly included patients with extracranial oligoprogression while excluding those with baseline BMs or intracranial progression following first-line therapy. From a clinical perspective, intracranial metastases are typically managed with distinct modalities such as whole-brain radiotherapy or SRS. These CNS-directed therapies carry distinct prognostic implications,^17,18 and have been shown to provide substantial survival benefits for patients with brain involvement. ¹⁹ By excluding these individuals, we effectively avoided the confounding survival impact of intracranial interventions, allowing for a more homogeneous cohort and a more precise evaluation of the target population that specifically benefits from extracranial radiotherapy. However, this exclusion introduces a notable selection bias, which warrants careful interpretation of our findings. Biologically, it is widely hypothesized that patients without early CNS involvement presumably represent a distinct subgroup, exhibiting a less aggressive tumor phenotype or a distinct genomic landscape that lacks the molecular drivers predisposing to early hematogenous spread to the brain. ²⁰ By excluding patients with baseline BMs, our cohort is inherently restricted to a “favorable-risk” subgroup with a potentially more indolent natural disease course. While this design enables a cleaner evaluation of extracranial treatment synergy without being confounded by the high mortality associated with CNS progression, the survival benefits observed in our study could partly reflect the inherently less aggressive biological spectrum of this selected ES-SCLC subset. Accordingly, caution is recommended when extrapolating these results to the broader, more heterogeneous ES-SCLC population with baseline BMs. Within this carefully selected extracranial cohort, the liver and bone were the most frequently observed sites of disease progression (33 patients each), excluding the primary tumor and regional lymph nodes. This pattern is consistent with previously reported findings in ES-SCLC, in which patients receiving second-line or subsequent immunotherapy and experiencing multiple disease progressions demonstrate a predominant pattern of liver involvement. ²¹

Currently, evidence evaluating the efficacy of radiotherapy specifically in the second-line setting for ES-SCLC remains extremely scarce. A small retrospective study of 20 SCLC patients with oligoprogression or oligorecurrence reported that while SBRT achieved an impressive local control rate, the survival benefit was primarily limited to patients with an initial diagnosis of limited-stage disease, and distant failure remained common. ¹⁴ In contrast, our multicenter study, with a significantly larger and more homogeneous ES-SCLC cohort, demonstrates that salvage RT can indeed translate into meaningful PFS2 and OS2 improvements for patients with extracranial oligoprogression after first-line therapy. Biologically, this benefit aligns with the “oligometastatic/oligoprogressive” hypothesis, positing that an intermediate state of tumor spread exists where local control can alter the disease trajectory. ²² Our findings are consistent with and further extend the principles established in prior studies on first-line cTRT. The phase II RTOG-0937 trial, which included 146 patients with oligometastatic disease, found no significant difference in 1-year OS (p = 0.21), but showed a longer time to progression (p = 0.01) and improved local control in the cTRT group (62.5% vs 25.8%). ²³ A secondary analysis of the CREST trial also reported superior OS in patients with ⩽2 metastases and significantly improved PFS (p = 0.003). ²⁴ Our study extends these survival benefits to treatment strategies following first-line progression. Notably, the second-line benefit was substantial: median PFS2 nearly doubled, and OS2 was significantly prolonged.

Our multivariate analysis in the PSM cohort provided further insights into the survival benefit of salvage radiotherapy. Consistent with previous studies, ECOG PS and the absence of liver metastases (a key visceral metastatic site) were identified as robust prognostic factors. Most importantly, radiotherapy emerged as an independent predictor of improved PFS2 (p = 0.027). Regarding OS, although the p-value for radiotherapy reached marginal significance (p = 0.053), the HR (0.64) was remarkably consistent with that of PFS2 (HR, 0.62). This suggests that the statistical insignificance for OS2 may be attributable to the relatively limited sample size of the matched cohort (N = 106) rather than a lack of therapeutic efficacy. Furthermore, OS in the second-line setting is often confounded by the efficacy of subsequent third-line therapies, which might dilute the statistical signal of local treatment. Nevertheless, the 36%–38% reduction in the risk of progression or death observed in our matched cohort strongly supports the incorporation of radiotherapy into the treatment paradigm for extracranial oligoprogressive ES-SCLC.

Historically, second-line treatment for ES-SCLC has relied on switching systemic agents, such as lurbinectedin or topotecan. ²⁵ However, the therapeutic landscape has evolved with the emergence of novel agents such as tarlatamab, which demonstrated an ORR of 40% and a median OS of 14.3 months in the DeLLphi-301 trial, ⁷ and promising ADCs such as I-DXd. In this shifting landscape, our findings suggest that local ablative RT for oligoprogression may serve as a strategic “bridge” to these next-generation therapies. Specifically, by nearly doubling the median PFS2 (6.1 vs 3.3 months, HR, 0.58, p = 0.007) and achieving an mOS2 of 13.4 months in our RT cohort, radiotherapy can effectively stabilize resistant clones and extend the treatment window. This gain in progression-free time is clinically decisive, as it preserves a patient’s PS and ensures they remain eligible to benefit from these highly effective, emerging systemic options. Furthermore, our results indicate that RT may allow for the continuation of first-line immunotherapy, thereby delaying the need for a systemic switch and preserving these potent novel agents for subsequent use.

Biologically, this survival benefit potentially aligns with the growing understanding of the synergistic mechanisms between radiation and immune checkpoint inhibitors. Radiotherapy is proposed to induce immunogenic cell death, leading to the release of tumor-associated antigens and priming of cytotoxic T lymphocytes, thereby functioning as an in situ vaccine. ²⁶ In oligoprogression, acquired resistance is often driven by clonal evolution in specific sites while systemic dominance remains controlled. Local ablative radiotherapy not only eradicates these resistant clones but may also reinvigorate the exhausted immune microenvironment, potentially restoring sensitivity to PD-L1/PD-1 blockade. ²⁷ Consequently, continuing the established immune checkpoint inhibitor maintains the systemic “brake release” on T-cells, amplifying the abscopal-like effects triggered by local radiation. Future prospective studies incorporating translational biomarkers, such as dynamic circulating tumor DNA (ctDNA) monitoring and tumor mutation burden analysis, are warranted to further validate these biological hypotheses. However, we acknowledge that without translational data—such as dynamic ctDNA kinetics or detailed immune profiling—in the present cohort, these mechanistic insights remain speculative. Therefore, our observations regarding the biological synergy between radiotherapy and systemic therapy should be considered strictly hypothesis-generating, warranting further validation in prospective trials with integrated biomarker analyses.

Furthermore, we identified distinct patterns of progression that serve as predictors for radiotherapy efficacy. In the subgroup with existing-lesion progression, especially in the intrathoracic progression subgroup, the OS2 benefit was the most significant. However, in the subgroup with new-lesion progression, RT did not bring about significant survival benefits. This differential therapeutic effect underscores the pivotal role of tumor biological heterogeneity in treatment decision-making: progression within pre-existing lesions likely represents a localized resistance mechanism that remains amenable to focal intervention, whereas the emergence of new metastatic sites reflects a systemic, aggressive disease phenotype associated with widespread therapeutic failure—rendering local modalities such as radiotherapy unlikely to confer meaningful clinical benefit. In addition, compared to the PFS1 ⩽6 month subgroup, the greatest improvement in PFS2 was observed in the PFS1 >6 months subgroup. This result indicates that a longer PFS1 period is more likely to be associated with more indolent biological characteristics of the tumor, resulting in potentially higher sensitivity to radiotherapy.

Several limitations should be acknowledged. First, the retrospective design introduces potential immortal time bias, as patients in the RT group had to survive long enough to receive treatment. While we proactively addressed this by performing a 2-month landmark sensitivity analysis, which confirmed the robustness of our results, the inherent nature of retrospective data still warrants cautious interpretation. Second, the exclusion of patients with baseline BMs might have introduced a selection bias favoring a population with a more indolent biological phenotype and better prognosis, potentially limiting the generalizability of our findings. Third, although PSM was applied, residual selection bias may persist due to unmeasured confounders such as baseline lactate dehydrogenase levels, comorbidities, and total tumor burden, which were not fully captured. Fourth, there was heterogeneity in RT regimens (dose and fractionation) across centers, and the lack of centralized imaging or pathology review might introduce inter-institutional variability in defining oligoprogression. Fifth, routine follow-up and disease reassessment were not performed with PET/CT in this real-world Chinese cohort, largely due to high out-of-pocket costs and limited medical insurance reimbursement for surveillance imaging. Consequently, occult distant metastases might have been underestimated, although standardized contrast-enhanced CT was uniformly adopted for progression evaluation and radiotherapy planning across all participating centers. Finally, the study was conducted in a Chinese cohort, where treatment patterns (e.g., specific PD-1/PD-L1 inhibitors) and drug accessibility (e.g., lurbinectedin) may differ from Western populations, potentially limiting the generalizability of our results. Prospective global trials are needed to validate these preliminary findings.

Conclusion

In conclusion, this multicenter cohort study is one of the first to demonstrate that salvage radiotherapy is associated with significantly improved survival in patients with ES-SCLC experiencing extracranial oligoprogression after first-line therapy. Among those receiving radiotherapy, continued systemic therapy conferred greater benefit—particularly following first-line immunotherapy. While our findings provide a valuable therapeutic strategy for managing this specific pattern of progression, prospective trials are warranted to validate these results in a broader patient population.