Multicenter Retrospective Study of the Impact of Localized Therapy on Prognosis in Patients With Oligometastatic Solid Tumors

ABSTRACT

Background

Local therapy has recently been applied to treat oligometastatic cancer across various malignant cancers. However, definitive evidence for local treatment regimens remains lacking. We aimed to evaluate the clinical outcomes of combining local therapy with systemic treatment for patients with oligometastatic solid tumors including breast cancer, liver cancer, and non‐small cell lung cancer (NSCLC) in a multicenter retrospective study.

Methods

From January 2013 to December 2023, 575 patients with fewer than or equal to five metastases of breast cancer, liver cancer, and NSCLC in three tertiary university hospitals were analyzed retrospectively. The primary endpoint was overall survival (OS). The secondary endpoints included progression‐free survival (PFS) and prognostic factors affecting the treatment outcomes.

Results

Before propensity score matching, significant differences were observed in both OS and PFS between the local therapy plus systemic therapy and systemic therapy alone groups (p < 0.001 and p < 0.001). After PSM, PFS remained significantly different (p = 0.017), but OS did not (p = 0.076). In liver cancer, before PSM analysis showed a significant PFS benefit (p = 0.011), which was absent after PSM for both OS and PFS (p = 0.568 and p = 0.363). For both breast cancer and NSCLC patients, no significant differences in OS or PFS were found after PSM.

Conclusions

Local treatment of solitary metastatic lesions showed that no statistically significant difference was observed in OS or PFS of patients with advanced oligometastatic breast, liver cancer, and NSCLC.

1. Introduction

Cancer metastasis was once regarded as a systemic, irreversible terminal event. In 1995, Hellman and Weichselbaum [1] introduced the notion of “oligometastasis,” which represents a novel and clinically relevant stage of cancer dissemination. This refers to a condition where the number and size of distant metastatic lesions are both limited, and the tumor metastasis is in a low‐burden state. Oligometastasis serves as a transitional phase, bridging localized disease and extensive systemic metastasis. In this state, lesions have the potential to attain complete remission via targeted local therapies. Local therapies like surgery or radiation may achieve complete remission. Currently, definitions of oligometastasis vary across cancer types. The 2022 ESTRO‐EORTC consensus defined oligometastases as ≤ 5 lesions in ≤ 3 organs, while the 2024 CSCO Lung Cancer Guidelines in China defined it as ≤ 3 lesions in ≤ 2 organs. The 2025 St. Gallen International Breast Cancer Conference (SGBCC) proposed the definition of oligometastatic breast cancer as an advanced subtype characterized by a limited number of metastatic lesions (typically ≤ 5) confined to specific anatomical sites (such as bone, liver, or lung).

Inconsistencies in the results of clinical trials have resulted in ambiguous guidelines for local treatment strategies across various types of cancer. For most cancers, radiotherapy remains the primary recommended approach, while other local therapies like surgery or trans‐arterial chemoembolization lack explicit endorsement. The 2024 NCCN guidelines establish the combination of stereotactic body radiotherapy (SBRT) and systemic therapy as a Class I evidence‐based component in managing oligometastatic lung cancer. The CURB study revealed that SBRT markedly improved the median PFS in NSCLC patients (10.0 vs. 2.2 months, HR = 0.41, p = 0.004), although no significant benefit was observed in the breast cancer group [2]. The SABR‐COMET study [3] and SARON study [4], studies further confirmed that SBRT improves overall survival (OS) in oligometastatic lung cancer patients. A Phase II clinical trial showed that SBRT increased median PFS from 3.5 to 9.7 months (p = 0.010) [5]. A review of 21 studies found that SBRT for up to five oligometastatic lesions resulted in a one‐year OS rate of 85.4% and a 1‐year PFS rate of 51.4% [6].

The efficacy of local consolidative therapy varies by the type of cancer. For oligometastatic hepatocellular carcinoma (HCC), the EASL 2023 guidelines, based on retrospective study data (such as a 2‐year overall survival rate of 80% after SABR [7]), suggest that SBRT may be considered. In breast cancer, despite ESMO 2023 issuing relevant recommendations, the key NRG‐BR002 study showed that local therapy did not significantly improve OS [8]. Nevertheless, several studies [9, 10, 11, 12, 13] still support the benefit of surgical intervention for patients with lung metastases. However, Niwinska et al. have found that significantly longer median survival in patients with isolated metastases undergoing local therapy combined with systemic therapy compared to those with solitary brain metastases but extracranial metastases (20 vs. 11 months, p = 0.010) [14].

Therefore, there is currently limited research in China comparing the survival outcomes between systemic therapy and local therapies combined with systemic therapy in solid tumors including breast cancer, liver cancer, and NSCLC presenting with oligometastatic disease. These three cancer types were selected for several reasons: their comparably radiation sensitivity ensures uniform radiobiological response; their similar oligometastatic disease patterns of predominant hematogenous leads to visceral organs metastasis such as the liver and lung, and bone. Additionally, these malignancies represent high‐incidence populations that facilitate efficient patient accrual and timely trial completion. Finally, each has well‐established systemic treatment guidelines, providing clear and standardized comparator arms for the study design. This study aims to evaluate whether adding local therapies with systemic treatment improves OS and PFS in patients with oligometastatic disease from breast cancer, liver cancer, or NSCLC. The results of our research might support a combined therapeutic strategy for advanced solid tumors with oligometastatic disease.

2. Material and Methods

2.1. Patients' Selection

2.1.1. Inclusion Criteria

Qualified participants were adults, aged 18 years or older, identified between January 2013 and December 2023.

Oligometastatic lesions were measured consistently. Oligometastatic disease was defined as the presence of five or fewer metastatic lesions, irrespective of the specific visceral organ(s) involved (liver, lung, or other), with allowance for concurrent bone and/or lymph node metastases.

Verifying through imaging or pathology in the lung (using CT, PET‐CT, bronchoscopy or biopsy), liver (via CT, MRI, PET‐CT), brain (through CT or MRI), or lymph nodes (using CT, PET‐CT or biopsy). These individuals achieved complete response, partial response, or stable disease for at least 3 months on systemic treatment and had sufficient organ function to endure this therapy.

The criteria for local therapy selection were consistent across all participating centers.

2.1.2. Exclusion Criteria

The exclusion criteria encompassed patients with a history of second primary malignancies, those lost to follow‐up, and individuals presenting with leptomeningeal or brainstem metastases.

2.2. Data Extraction

This retrospective study, conducted across multiple centers, examined records from Sun Yat‐sen Memorial Hospital, The First Affiliated Hospital of Soochow University, and West China Medical Center, covering the period from January 2013 to December 2023. Data on demographics, clinicopathology, and treatment were gathered from patient files, and survival information was kept current through clinical statistics and phone follow‐ups up until December 1,2024. OS was characterized as the period from the identification of oligometastasis until either the patient's death from any cause or the most recent follow‐up. PFS was described as the time from the diagnosis of oligometastasis until the first occurrence of disease progression or death from any cause. Disease progression encompasses either local recurrence or the spread of the disease to distant sites. The clinicopathological traits examined comprised age, ECOG performance score, histological classification, molecular category, PD‐L1 levels, patterns of metastasis, tumor load measures, approach to local treatment, and the interval of disease‐free survival.

2.3. Statistical Analysis

The attributes related to both clinical and pathological aspects were examined with Chi‐squared tests. For analyzing survival outcomes, Kaplan–Meier curves combined with log‐rank tests were employed, and multivariate Cox regression was used to determine independent prognostic indicators. Propensity score matching (PSM) (1:2 ratio, caliper = 0.2) balanced baseline characteristics including age, ECOG, pathological type, metastatic burden, treatment response, and DFI. Cancer‐specific analyses incorporated molecular subtypes for breast cancer and NSCLC, and Child‐Pugh scores for liver cancer. Every analysis was conducted utilizing IBM SPSS version 25.0 and STATA version 18.0, and statistical significance was determined at a p‐value of less than 0.05.

3. Results

In this research, 575 participants were included, with 252 undergoing both systemic and local treatments, while 323 were given only systemic therapy. The entire group was followed up for a median period of 26.0 months (range: 3.0–157.0), and the two categories had closely matched median ages (52 compared to 54). The combination therapy group predominantly consisted of NSCLC (48.2%, 122/252) and liver cancer (30.8%, 77/252), while the systemic therapy group was primarily breast cancer (76.2%, 246/323) with fewer NSCLC (17.3%, 56/323) and liver cancer (6.5%, 21/323) cases.

Prior to PSM, OS and PFS showed significant differences between the two treatment groups (Figure S1). Following PSM (248 patients in the combined therapy group vs. 142 in the systemic therapy group), the difference in OS was no longer statistically significant (median OS: 70.00 vs. 34.00 months; p = 0.076; HR 0.660 [0.488–0.893]; Figure 1a). However, a significant PFS benefit persisted for the combination therapy group (median PFS: 24.00 vs. 16.00 months; p = 0.017; HR 0.618 [0.492–0.777]; Figure 1b), indicating that local therapy combined with systemic therapy significantly prolonged PFS after matching. The love plot is shown in the supplement file (Figure S2).

FIGURE 1.

Survival curves for OS and PFS after PSM in solid tumors (a: OS; b: PFS).

Consistently significant univariate analyses before and after PSM revealed that metastatic lesion size, lesion number, and DFI were truly independent prognostic factors for OS, while primary tumor resection, lesion size, and DFI were truly independent prognostic factors for PFS (Table S1).

3.1. Liver Cancer

This group of liver cancer patients comprised 98 individuals in total, with 77 placed in the combination therapy category (systemic treatment alongside local therapy) and 21 allocated to the group receiving only systemic therapy. The median duration of follow‐up was 23.0 months, with a range from 3.0 to 135.0 months (Table S2).

Before PSM, PFS significantly favored the combined therapy group (median PFS: 18.00 vs. 8.00 months; p = 0.011), while OS showed no significant difference (median OS: 86.00 vs. 33.00 months; p = 0.061) (Figure S3). After PSM (22 patients in the combined therapy group vs. 16 in the systemic therapy group), no significant differences were observed in either PFS (p = 0.568; HR 0.743 [0.625–2.081]) or OS (p = 0.363; HR 0.739 [0.376–1.452]) between the two groups (Figure 2a,b). The median OS for the group receiving combined therapy was 30 months, compared to 24 months for those undergoing solely systemic therapy. Meanwhile, the median PFS was 13 months in the combined therapy group, whereas it was 8 months for the group receiving only systemic therapy. The love plot is shown in the supplement file (Figure S4).

FIGURE 2.

Survival curves for OS and PFS after PSM in liver cancer (a: OS; b: PFS).

Before PSM multivariate analysis identified a DFI < 24 months as the only independent risk factor for OS (HR 18.083, p < 0.001). Most importantly, after PSM, univariate analysis revealed that a maximum metastatic lesion diameter of < 5 cm was significantly associated with prolonged OS (HR 0.160, p = 0.002) (Table S3).

In the pre‐PSM assessment, DFI less than 24 months and ICC/mixed histology were significantly linked to inferior PFS in univariate analysis, while multivariable analysis confirmed DFI < 24 months (HR 8.955, p < 0.001) and age (HR 1.030, p = 0.006) as independent risk factors. Critically, after PSM, univariate analysis identified DFI 24 months (HR 8.605, p < 0.001), bone metastases (HR 2.381, p = 0.043), and SD response to systemic therapy (HR 1.856, p = 0.008) as significant risk factors for progression (Table S4).

Analyses of subgroups are shown in Figure S5.

3.2. Breast Cancer

Among this group of individuals diagnosed with breast cancer, there were 299 patients in total. Of these, 53 patients were part of the group receiving both systemic and local therapies, while 246 were treated with systemic therapy only. The median duration of follow‐up was 28.0 months, with a range from 3.0 to 157.0 months (Table S5).

Prior to PSM, Kaplan–Meier analyses indicated that there were no statistically significant variances in PFS (p = 0.130) or OS (p = 0.206) among the treatment cohorts, though the median OS was numerically higher (57.00 vs. 34.00 months) and PFS (17.00 vs. 15.00 months) were observed with combination therapy (Figure S6). After PSM (51 combination vs. 70 systemic therapy patients), no significant survival differences remained: mOS was 57.00 versus 34.00 months (p = 0.280; HR 0.757 [0.455–1.259]) and mPFS was 17.00 versus 18.00 months (p = 0.473; HR 0.863 [0.573–1.032]) (Figure 3a,b). The love plot is shown in the supplement file (Figure S7).

FIGURE 3.

Survival curves for OS and PFS after PSM in breast cancer (a: OS; b: PFS).

Before PSM, multivariable analysis identified that a DFI < 24 months (HR 3.291, p < 0.001), TNBC subtype (HR 2.017, p = 0.001), and distant lymph node metastasis (HR 1.574, p = 0.005) independently predicted a heightened risk. Conversely, the use of targeted therapy (HR 0.596, p = 0.003) and metastatic lesion size < 5 cm (HR 0.559, p = 0.012) were linked to a decreased risk. Post‐PSM analysis confirmed TNBC subtype (HR 2.209, p = 0.013), multi‐organ metastasis (HR 2.041, p = 0.023), 3–5 metastatic lesions (HR 2.020, p = 0.013), brain/lung/combined metastases (HR 24.167, p = 0.006), and DFI < 24 months (HR 3.851, p < 0.001) as significant risk factors, with metastatic lesion size < 5 cm (HR 0.387, p = 0.003) and targeted therapy (HR 0.430, p = 0.002) remaining protective. The consistency of DFI < 24 months, TNBC subtype, metastatic burden, and targeted therapy across analyses highlights their robust prognostic value (Table S6).

In the multivariate analysis performed before PSM, having a DFI < 24 months (HR 4.446, p < 0.001), maintaining stable disease status after systemic therapy (HR 1.490, p = 0.003), experiencing distant lymph node metastasis (HR 1.335, p = 0.026), and receiving second‐line or further treatment (HR 2.054, p < 0.001) were identified as independent indicators of reduced PFS. After PSM, univariate analysis verified DFI < 24 months (HR 6.119, p < 0.001), having 3 to 5 metastatic lesions (HR 1.811, p = 0.013), the presence of bone metastasis (HR 1.745, p = 0.012), and distant lymph node metastasis (HR 1.662, p = 0.023) were notable risk factors. The consistent identification of DFI < 24 months across both pre‐ and post‐PSM analyses underscores its robust prognostic value (Table S7).

Analyses of subgroups are shown in Figure S8.

3.3. NSCLC

This group of NSCLC patients comprised 178 participants in total, with 122 individuals placed in the cohort receiving both systemic and local treatments, while 56 patients were allocated to the systemic treatment‐only cohort. The median follow‐up period was 21.50 months (ranging from 3.0 to 125.0 months). Details of personal and clinical attributes can be found in Table S8 of the Supporting Information.

Before PSM, the combination therapy group demonstrated numerically longer median PFS (37.00 vs. 20.00 months) and OS (101.00 months vs. not reached) PFS showed a statistically significant difference (p = 0.008), whereas OS did not reach statistical significance (p = 0.166). (PFS p = 0.008, OS p = 0.166) (Figure S9). After PSM (89 patients receiving combination therapy versus 36 undergoing systemic therapy), there was no observed significant survival advantage with the combined method: the median OS was 104.00 months compared to not reached (p = 0.139; HR 0.597 [0.298–1.197]) and the median PFS was 52.00 months compared to not reached (p = 0.164; HR 0.690 [0.406–1.173]) (Figure 4a,b). The love plot is shown in the Supporting Information (Figure S10).

FIGURE 4.

Survival curves for OS and PFS after PSM in NSCLC (a: OS; b: PFS).

Pre‐PSM multivariable analysis identified DFI < 24 months (HR 5.862, p < 0.001) as an independent risk factor for OS. Post‐PSM univariate analysis confirmed DFI < 24 months (HR 4.044, p < 0.001) along with additional significant factors: poor ECOG status (HR 3.217, p < 0.001), non‐adenocarcinoma histology (HR 3.185, p < 0.001), > 2 metastatic lesions (HR 2.300, p = 0.006), and distant lymph node metastasis (HR 2.119, p = 0.045), while targeted therapy remained protective (HR 0.464, p < 0.05) (Table S9). Pre‐PSM multivariable analysis identified DFI < 24 months (HR 10.657, p < 0.001), second or further line treatment (HR 3.384, p < 0.001), SD response (HR 1.862, p = 0.016), and poor performance status (HR 2.498, p < 0.001) as independent risk factors, while PD‐L1 ≥ 50% (HR 0.075, p < 0.001), local surgery (HR 0.384, p = 0.004), radiotherapy/ablation (HR 0.499, p = 0.017), and metastatic lesion size < 5 cm (HR 0.277, p = 0.005) were protective. After PSM, high PD‐L1 expression (HR 0.104, p = 0.003), ≤ 2 metastatic lesions (HR 2.384, p = 0.001), no interruption of systemic therapy (HR 2.832, p = 0.046), and R0 primary resection (HR 0.582, p = 0.050) remained significant (Table S10).

Analysis of subgroups is shown in Figure S11.

4. Discussion

Advances in early cancer detection and systemic therapies have prolonged survival, highlighting the clinical challenge of oligometastatic disease. This multi‐center analysis evaluated the impact of local therapy combined with systemic treatment on survival outcomes in 575 patients with oligometastatic liver cancer, breast cancer, or NSCLC. The integration of local therapy significantly improved PFS but did not significantly enhance OS, partially aligning with the findings of the Phase II SABR‐COMET trial [3]. This study was explicitly designed to evaluate whether the addition of any local ablative therapy to systemic treatment improves outcomes in oligometastatic disease, rather than to compare the efficacy of specific local modalities. The choice of local therapy was at the treating physician's discretion based on lesion location, patient fitness, and institutional expertise, reflecting real‐world decision‐making. Consequently, the study is not designed to compare specific local treatment strategies.

Our research found that local therapy alongside systemic treatment improved PFS before PSM in liver cancer patients, aligning with the SABR‐COMET trial. However, several studies suggested local therapy might improve OS. Kim et al.'s meta‐analysis of 527 patients with oligometastatic HCC showed a superior 1‐year OS rate in the local therapy group (71.8% vs. 43.3%) [15]. Similarly, a Phase II clinical trial also confirmed that stereodirected ablation radiotherapy improved both OS and PFS (1‐year OS rate: 88.9% and 1‐year PFS rate: 21.2%) [16]. Other small sample sizes of studies have also reported that local treatments (surgical resection, microwave ablation, SBRT, etc.) for oligometastatic lesions may improve prognosis [17, 18]. The lack of significant PFS improvement after PSM in our study may be attributed to the insufficient statistical efficacy caused by the reduction in sample size after matching, and the lower acceptance rate of subsequent systemic treatment or treatment interruption of patients in the local treatment group—these are themselves adverse prognostic factors, which may mask the real impact of local treatment on oligometastasis. The current study included some intrahepatic cholangiocarcinoma cases, which is an aggressive cancer type with a 5‐year OS rate of only 9% [19]. Whose poor prognosis may have influenced the observed benefits in this study. Furthermore, local therapy is considered effective in delaying disease progression. Among the local treatment modalities, the majority of patients received TACE. We hypothesize that the tumor's aggressiveness, the intrinsic constraints of TACE in securing local control, the ongoing progression of underlying hepatic dysfunction, and the potential lack of efficacious systemic interventions may interact synergistically to affect overall survival. Future large‐scale trials are needed to validate the therapeutic advantage of integrating locoregional and systemic treatments in HCC. While our study observed an improvement in PFS, the clinical significance of this finding should be interpreted with caution for several reasons. First, this PFS benefit did not translate into a statistically significant improvement in OS. This discrepancy may be attributable to the influence of subsequent therapies or the relatively limited sample size, which could obscure a true OS effect. Second, the inherent nature of retrospective studies makes PFS assessment susceptible to bias, as the frequency of radiographic evaluation was not standardized and could vary among patients. Therefore, the observed PFS advantage should be considered hypothesis‐generating rather than definitive evidence of clinical benefit.

In individuals with oligometastatic breast cancer, our analysis revealed a numerical trend toward improved OS was observed in the combined local therapy group both before and after PSM, though not statistically significant. However, PFS showed no significant improvement. Regarding the choice of local treatment methods, there was a notable gap in surgical utilization before PSM. However, following PSM, no major distinctions were seen in the implementation of surgery, radiotherapy, or alternative local procedures, suggesting that under clinically balanced conditions, variations in local treatment under balanced conditions did not significantly improve survival outcomes. Our findings are consistent with the NRG‐BR002 trial, which also reported no PFS improvement from adding stereotactic body radiotherapy (SBRT) to standard systemic therapy in oligometastatic ER+/HER2− breast cancer [8]. Though, the NRG‐BR002 group restricted the applicability of its findings to other molecular subtypes. Although current evidence suggests that adding local treatment for oligometastatic foci may improve outcomes. In a cohort of 309 patients with oligometastatic breast cancer who received SRT to sites such as brain, lymph nodes, liver, or lung, OS was significantly influenced by certain intrinsic risk factors. Patients with 0–2 risk factors achieved a median OS of 23–40 months, whereas those with three or more risk factors had a poorer prognosis. The principal risk factors included non‐adenocarcinoma histology, intracranial metastases, synchronous metastases, and male sex, all of which significantly impacted OS [20]. When analyzed by specific metastatic sites, studies showed variation in the approach to local therapy. Evidence supports that SBRT can achieve clinical benefit for oligometastases in bone, adrenal glands, and liver [21, 22, 23, 24, 25, 26, 27], while surgical resection may be suitable for solitary pulmonary metastases [28, 29, 30]. For instances of brain oligometastases, some research [31, 32] indicates that selecting a local treatment should be based on the quantity of lesions and their anatomical location. Likewise, subgroup analysis demonstrated metastatic location had no significant prognostic influence on local therapy outcomes. This cohort consisted mainly of young Chinese women, differing from trials enrolling predominantly Caucasian populations, so the results may not be directly comparable. Therefore, future large‐scale studies focusing on younger Asian women and stratified analyses by metastatic site are needed to further investigate the prognostic implications of local treatment strategies.

Among the NSCLC group, survival outcomes showed no notable differences between different treatment, with numerical trends lacking statistical significance. However, the 2023 ASTRO/ESTRO guidelines [33] recommend that radical local therapy for all oligometastatic lesions (max 5 metastases) when feasible and safe, endorsing radiotherapy based on prior evidence. Pre PSM analysis linked surgery or radiotherapy to improved PFS, aligning with clinical trial like Gomez et al.'s [34] which demonstrated that PFS benefit from local consolidation therapy. The lack of OS benefit in our study potentially due to limited sample size or heterogeneity in subsequent systemic treatments. Surgery feasibility depended on R0 resection likelihood and disease confinement. Notably, most eligible NSCLC patients typically have ≤ 2 metastases, whereas half of our cohort had 2–5 lesions, possibly influencing outcomes. The role of biomarkers like EGFR mutation requires further study; emerging evidence supporting combined EGFR‐TKIs and radiotherapy, showing superior OS and PFS versus TKI monotherapy (e.g., SINDAS study [35] and Li et al. [36]). Other studies also affirm local therapy's role in EGFR‐mutant oligometastatic NSCLC [37, 38, 39]. Beyond genetically defined populations, comprehensive clinical evaluation remains crucial for identifying long‐term survivors. Post‐PSM analysis identified DFI < 24 months as a consistent independent risk factor for OS. This finding aligns with Vin et al.'s study in validating non‐adenocarcinoma histology and shorter DFI as OS predictors [20]. The association between high PD‐L1 expression and improved PFS suggests persistent immune regulation in oligometastatic setting. Our findings differ from CURB's [2], due to distinct disease states: CURB's oligoprogressive population derived clear 12‐month PFS benefit from salvage local therapy, while our new oligometastatic cohort received consolidation therapy with benefits possibly masked by effective systemic treatment. Additionally, CURB predated widespread immunotherapy, whereas our study spanned this era—improved systemic efficacy may have attenuated local therapy's relative contribution. Notably, improved OS with targeted therapies confirms precision systemic treatment as foundational for molecularly defined subtypes, while the identified clinical factors enable prognostic stratification beyond genetic markers. Integrating molecular classifications and immune characteristics will improve the comprehension of the significance and effectiveness of local therapeutic approaches in studies on oligometastatic solid tumors.

It is important to emphasize several key limitations of this study that may affect the interpretation of our results, such as inter‐group imbalance, treatment heterogeneity and residual confounding. This retrospective study, despite PSM to reduce baseline imbalances, residual confounding by tumor type and temporal treatment bias likely persists due to the inherent limitations of the retrospective design. Consequently, the observed associations may still be subject to inherent bias inherent to retrospective designs. Second, treatment heterogeneity represents a major limitation. In real‐world practice, local interventions were administered at various time points, ranging from upfront consolidative treatment after initial response to salvage therapy. This variability differs from the strict protocols of prospective trials and may introduce heterogeneity in the observed outcomes. Moreover, after PSM, the cohorts of the three cancer types became relatively small, which may limit the statistical power to make survival comparisons in the multivariate analyses; thus further studies with larger sample sizes may be warranted to confirm our findings. Immortal time bias is a critical methodological concern inherent to retrospective studies evaluating time‐dependent treatments. In our cohort, patients were required to survive and remain progression‐free until the administration of local treatment, which could potentially contribute to the observed PFS differences. Although we did not perform a landmark analysis to fully mitigate this bias due to sample size constraints and to avoid selection bias related to time‐point selection, we have explicitly acknowledged this limitation. The results regarding PFS should therefore be interpreted with this potential bias in mind. Several prognostic and treatment‐related factors including PD‐L1 expression, mutation status, and Child‐Pugh scores were not available for adjustment. Moreover, after PSM, the cohorts of the three cancer types became relatively small, thereby limiting statistical power to make survival comparisons in the multivariate analyses; thus further studies with larger sample sizes may be warranted to confirm our findings. Unlike SABR‐COMET, NRG‐BR002, or Gomez et al., our retrospective study lacked randomization, standardized protocols, and blinded outcome assessment. This limitation is inherent to the retrospective design and highlights the need for prospective studies with standardized data collection. Finally, while we performed stratified analyses by cancer type to enhance clinical interpretability, these exploratory findings should be interpreted with caution. They are not meant to imply definitive conclusions about differential treatment efficacy across tumor types. Since some analyses were exploratory, the findings should be interpreted with caution and validated in future prospective cohorts.

In summary, this multi‐center retrospective study has demonstrated that local therapy for oligometastatic lesions did not confer significant benefits in terms of PFS and OS for patients with advanced breast cancer, liver cancer, and NSCLC. However, a numerical trend toward improved OS in patients with advanced breast cancer was observed, although this did not reach statistical significance. Observed numerical trends of our study warrant attention and provide insights for future research. We eagerly anticipate the results of these studies to validate our conclusions.

Author Contributions

Xueman Deng: conceptualization, funding acquisition, supervision, writing – review and editing. Duoguang Wu: conceptualization, visualization, writing – original draft. Luting Ni: validation, writing – review and editing. Wei Li: validation, writing – review and editing. Zhijun Liu: validation, writing – review and editing. Jiewei Liu: conceptualization, supervision, writing – review and editing. Jieqiong Liu: conceptualization, funding acquisition, supervision, writing – review and editing.

Funding

This work was supported by grants from the National Major Science and Technology Projects of China [2025ZD0552502]; the Natural Science Foundation of China [82072906, 82573590]; CSCO Clinical Oncology Research Foundation [Y‐HR2020MS‐0200, Y‐Gilead2024‐PT‐0080]; Guangdong Provincial Clinical Research Center for Breast Diseases [2023B110005]; Sun Yat‐Sen Memorial Hospital Clinical Research 5010 Program [SYS‐5010‐202508]; Sun Yat‐Sen Memorial Hospital Medical Scientist Program [1320900074]; Guangdong Provincial Key Laboratory of Cancer Pathogenesis and Precision Diagnosis and Treatment [2024B1212030002]; and Beijing Zhongguancun Precision Medicine Foundation. The Sichuan Provincial Natural Science Fund (2024NSFSC0743).

Ethics Statement

This study was approved by the Sun Yat‐sen Memorial Hospital (SYSKY‐2023‐909‐01).

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Figure S1: Survival curves for OS and PFS before PSM in solid tumors (a: OS; b: PFS).

Figure S2: Love plot for solid tumors.

Figure S3: Survival curves for OS and PFS before PSM in liver cancer (a: OS; b: PFS).

Figure S4: Love plot for liver cancer.

Figure S5: Forest plot of OS subgroup analyses of liver cancer with hazard ratios. CI, confidence interval; HR, hazard ratio.

Figure S6: Survival curves for OS and PFS before PSM in breast cancer (a: OS; b: PFS).

Figure S7: Love plot for breast cancer.

Figure S8: Forest plot of OS subgroup analyses of breast cancer with hazard ratios. CI, confidence interval; HR, hazard ratio.

Figure S9: Survival curves for OS before PSM in NSCLC (a: OS; b: PFS).

Figure S10: Love plot for non‐small cell lung cancer.

Figure S11: Forest plot of OS subgroup analyses of NSCLC with hazard ratios. CI, confidence interval; HR, hazard ratio.

Table S1: Univariate analyses of OS and PFS for solid tumors before and after PSM.

Table S2: Patient clinicopathologic characteristics for liver cancer.

Table S3: Univariate and multivariate analyses of OS for liver cancer before PSM, and univariate analysis after PSM.

Table S4: Univariate and multivariate analyses of PFS for liver cancer before PSM, and univariate analysis after PSM.

Table S5: Patient clinicopathologic characteristics for breast cancer.

Table S6: Univariate and multivariate analyses of OS for breast cancer before PSM, and univariate analysis after PSM.

Table S7: Univariate and multivariate analyses of PFS for breast cancer before PSM, and univariate analysis after PSM.

Table S8: Patient clinicopathologic characteristics for NSCLC.

Table S9: Univariate and multivariate analyses of OS for NSCLC before PSM, and univariate analysis after PSM.

Table S10: Univariate and multivariate analyses of PFS for NSCLC before PSM, and univariate analysis after PSM.

Data Availability Statement

The data sets generated during and/or analyzed during the current study are not publicly available but are available from the corresponding author on reasonable request.