The most challenging issue in cancer patients is the occurrence of brain metastases(1); in these patients, the type of treatment, survival rate, disease control, and the incidence of complications are of great importance(2). Surgical removal of the lesion and focused stereotactic radiosurgery (SRS) are used for treatment(3). Studies have shown that local lesion control is possible after SRS without a significant increase in neurological complications(4, 5). However, there is evidence of local recurrence and a risk of leptomeningeal disease (LMD)(6); statistics have shown that 12% of patients develop tumor spread to the cerebrospinal fluid after treatment(3, 7). In recent years, the introduction of preoperative or neoadjuvant SRS has provided more precise targeting of the healthy tumor(8). Preliminary studies suggest that this approach results in a reduced incidence of radiation necrosis and LMD(9). a study showed that pre-SRS compared with post-SRS had lower rates of two-year LMD (3.2% vs. 16.6%) and symptomatic radioactive necrosis (4.9% vs. 16.4%), while having no effect on overall survival(10). It was also reported in a study with a two-year follow-up period that pre‑SRS was associated with only a 4.8% rate of symptomatic necrosis(11).

A meta-analysis of six pre-SRS and 33 post-SRS studies reported that the local recurrence rate in pre-SRS was 11% (95% CI:4.9-13.7) and in post-SRS was 17.5% (95% CI:15.1-19.9); the risk of leptomeningeal disease in pre-SRS was 4.4% (95% CI:2.6-6.2) compared with 12.3% (95% CI:8.9-15.7) in post-SRS; however, overall survival was not different(12). Another meta-analysis has shown that local control rates of over 89% and low rates of radiation complications are observed at one- and two-year follow-up periods. Maroufi et al., 2025  reported that the he preoperative group had lower risks of radiation necrosis and leptomeningeal disease in one-year follow-up(13). 

Previous meta-analyses have examined one or two indicators, such as local control or overall survival. Whereas the current study considered overall survival, local control, local recurrence, leptospirosis, and radiation necrosis. Many previous meta-analyses have examined data from articles up to 2020, which is the most recent data in this study. Previous studies have mostly examined SRS without distinguishing preoperative vs postoperative. The innovation of the present study is the direct comparison of the effect of SRS timing on several important clinical indicators. By collecting comprehensive data and direct comparison, the present study can provide evidence-based guidance for choosing the best SRS timing, which has been less addressed in previous meta-analyses. Given the gap in evidence, particularly the lack of pooled data from different studies with adequate sample sizes and robust methodology, a systematic review and meta-analysis focused on a rigorous comparison of clinical outcomes between pre- and post-SRS is warranted. Present study can better determine the advantages and limitations of each approach and provide evidence-based guidance for clinical decision-making and the design of future studies, especially on the scale of randomized trials that are currently being planned. Therefore, the aim of the present study is Comparative Outcomes of Preoperative and Postoperative Stereotactic Radiosurgery in Patients with Brain Metastases.

 Method

Information sources and Search strategy

The present study was conducted according to the PRISMA 2020 guidelines. A search was conducted in the PubMed, Embase, Web of Science, Scopus, and Cochrane Library databases over the last ten years (January 2015 to November 2025) to find emerging evidence.

Mesh keywords:

("Brain Neoplasms"[Mesh]) OR ("Brain Neoplasms/complications"[Mesh] OR  "Brain Neoplasms/diagnosis"[Mesh] OR  "Brain Neoplasms/mortality"[Mesh] OR  "Brain Neoplasms/prevention and control"[Mesh] OR  "Brain Neoplasms/surgery"[Mesh] OR  "Brain Neoplasms/therapy"[Mesh]) AND "Radiosurgery"[Mesh]) OR ( "Radiosurgery/adverse effects"[Mesh] OR  "Radiosurgery/methods"[Mesh] OR  "Radiosurgery/mortality"[Mesh] OR  "Radiosurgery/statistics and numerical data"[Mesh]) AND "Preoperative Period"[Mesh]) AND "Neoadjuvant Therapy"[Mesh]) AND "Postoperative Period"[Mesh]) AND ( "Chemotherapy, Adjuvant"[Mesh] OR  "Radiotherapy, Adjuvant"[Mesh] )) AND ( "Survival"[Mesh] OR "Mortality"[Mesh] OR  "Survival Rate"[Mesh] )) AND "Diffuse Noxious Inhibitory Control"[Mesh]) AND ( "Recurrence"[Mesh] OR  "Neoplasm Recurrence, Local"[Mesh]) AND "Meningeal Neoplasms"[Mesh]) AND "Necrosis"[Mesh]) AND "Progression-Free Survival"[Mesh]) AND "Disease-Free Survival"[Mesh]) AND ( "complications" [Subheading] OR  "Postoperative Complications"[Mesh]).

Eligibility Criteria

The study inclusion strategy was determined based on the PICOS framework, as follows:

ü  Population (P): Adult patients (≥18 years) with brain metastases.

ü  Intervention (I): Preoperative SRS.

ü  Comparison (S): Postoperative SRS.

ü  Outcomes (O): Overall survival (OS), local lesion control, local recurrence, LMD, and radiation necrosis.

ü  Study type (S): Observational studies, randomized clinical trials (RCTs), cohort studies.

Exclusion criteria: Studies without quantitative data, single case reports, all types of review studies, animal studies, letters to the editor, conference and symposium abstracts, non-English language articles, articles without full text, sparse data, articles without a specified control group.

Study Selection and Data Extraction

All titles and abstracts were reviewed by two independent investigators. Studies that met the inclusion criteria were selected for full-text review. Disagreements were resolved by a third investigator. Two independent investigators extracted data using a pre-prepared form. The extracted data included study characteristics (year and study design), patient characteristics (number, age, gender, primary tumor type), intervention details, and clinical outcomes (overall survival, local control, local recurrence, LMD, radiation necrosis).

 Risk of Bias Assessment

To ensure the validity and quality of the data, randomized controlled trials (RCTs) were assessed with the Cochrane Risk of Bias 2.0 (RoB 2.0) tool(14) and non-randomized interventional studies with the Risk Of Bias In Non-randomized Studies of Interventions (ROBINS-I) tool(15). RoB 2.0 examines five key areas of bias including randomization process, deviation from planned intervention, incomplete outcome data, outcome measurement, and selective reporting of outcomes, and each area is categorized with three levels of risk: Low risk, Some concerns, and High risk.

 Statistical Analysis

All analyses were performed with Stata/MP. v17 software, and the significance level was considered to be P<0.05. Hazard Ratio (HR) was used for binary data and risk Ratio (RR) was used for proportional data. The fixed effects model was used using the Mantel-Haenszel method and inverse variance. Study weighting was performed based on sampling variance and, in the random model, between-study variance, and heterogeneity was measured using I² and Cochran's Q. Sensitivity and publication bias were examined with Funnel plots and Egger's test.

 Result

Literature Search, and Characteristics of included studies:

The search strategy was carried out according to the selected keywords and based on PRISMA 2020 protocol. Initially, 284 articles were found; After removing duplicates and screening titles, the abstracts of 853 articles were reviewed, and 714 articles unrelated to the inclusion criteria and in accordance with the exclusion criteria were eliminated at this stage. The full text of 139 articles was reviewed. Articles with scattered data, low quality in methodology, unclear sample size, lack of grouping of studies, lack of presentation of accurate results, lack of access to full text, contradictory data, and biased data were excluded from the research. Finally, ten articles that met the inclusion criteria for the present study were selected (Figure 1).

The total number of included patients in preset studies was 908, with approximately 356 in the Pre-SRS group and 552 in the Post-SRS group.  Patients were between 56 and 64.5 years old. The male-female gender ratio was approximately balanced (49–55% male and 45–51% female). The primary tumor type was similar in both groups, and the most common included lung cancer, breast cancer, melanoma, gastrointestinal cancers, urogenital cancers, kidney cancer, gynecological cancers, head and neck tumors, sarcoma, and other rare tumors. The characteristics of the brain lesions were also similar between the two groups. Lesions were mainly located in the supraantrenal regions including the frontal, parietal, temporal, occipital, and cerebellar lobes. The number and location of lesions were similar between the Pre-SRS and Post-SRS groups (Table 1).

 Table 1. Characteristics of the included RCT and cohort studies.

ü  RCs: Retrospective Cohort.

ü  BC: Breast cancer.

ü  SRS: Radiosurgery.

ü  GKRS: Gamma Knife Radiosurgery.

ü  LINAC: Linear Accelerator.

ü  GU: Genitourinary.

ü  GI: Gastrointestinal.

ü  HNT: Head, Neck, Throat.

ü  GYN: Gynecologic.

ü  CRC: Colorectal.

ü  NR: Not Reported.

ü  Me: Melanoma.

Table 2. continued…

Overall survival

The overall survival between the Pre-SRS and Post-SRS groups was 0.891 with a 95% confidence interval of 0.554 to 1.228. This finding indicates that there was no statistically significant difference in overall survival between the two groups (Figure 2).

Figure 2. forest plot showed overall survival between the Pre-SRS and Post-SRS

local failure 

The hazard ratio for local failure was 1.40 with a 95% confidence interval of 1.063 to 1.738. This finding indicates no significantly increased risk of local failure in the Post-SRS group compared to the c group (p>0.05) (Figure 3).

According to subgroup meta-analysis, At 6 months, the risk ratio for local failure was -0.83 with a 95% confidence interval of -1.51 to -0.15. This finding suggests that patients in the Post-SRS group have not a significantly higher risk of local failure than those in the Pre-SRS group (p>0.05). The risk ratio for the one-year period was -0.10 with a 95% confidence interval of -0.47 to 0.26, indicating that there is no statistically significant difference in local failure control between Pre-SRS and Post-SRS in the long term. The overall risk ratio of -0.30 and the 95% confidence interval of -0.62 to 0.02 confirmed that there was no significant difference between subgroups (p=0.06) (Figure 4).

Figure 3. forest plot showed local failure between the Pre-SRS and Post-SRS

Figure 4. Forest plot showed subgroup meta-analysis of 6-month and one-years follow-up of local failure between the Pre-SRS and Post-SRS

Distant failure

According to subgroup meta-analysis, at six months, the risk ratio for distant failure was -0.155 with a 95% confidence interval of -0.416 to 0.106. This indicates that there was no statistically significant difference between Pre-SRS and Post-SRS in the incidence of distant failure of the lesion during this period. After one year, the risk ratio was -0.129 with a 95% confidence interval between -0.358 and 0.099, which, similar to the first interval, did not show a statistically significant difference between the groups (Figure 5).

The overall risk ratio was -0.140 and the 95% confidence interval was -0.312 to 0.032, indicating that SRS timing had no significant effect on the incidence of distant failure over the entire treatment period. The test for difference between groups (p = 0.89) confirmed this result, indicating that stratification of studies based on follow-up period had no significant effect (Figure 5).

Figure 5. Forest plot showed subgroup meta-analysis of 6-month and one-years follow-up of distant failure between the Pre-SRS and Post-SRS

Radiation necrosis

The hazard ratio for the occurrence of radiation necrosis was 0.689 with a 95% confidence interval of 0.352 to 1.026. This finding suggests that patients in the Pre-SRS group are less likely to develop radiation necrosis compared with the Post-SRS group, and that preoperative SRS timing may reduce the risk of this important complication (p<0.05) (Figure 6).

Figure 6. Forest plot showed radiation necrosis between the Pre-SRS and Post-SRS

leptomeningeal disease (LMD)

The hazard ratio for LMD was 0.540 with a 95% confidence interval of 0.202 to 0.877. This finding suggests that preoperative SRS is associated with a significant reduction in the risk of LMD compared with postoperative SRS (Figure 6).

Figure 7. Forest plot showed LMD between the Pre-SRS and Post-SRS