This retrospective study was conducted in accordance with the ethical principles of the Declaration of Helsinki. Approval was obtained from the local ethics committee of the Technical University of Munich (no. 5626:12) and the requirement for written informed consent was waived.

A retrospective review of institutional databases was performed and patients with KPS < 70 who underwent surgical resection for newly diagnosed BM between April 2007 and January 2020 were analyzed. Exclusion criteria were as follows: previous treatment for BM, biopsy, or missing data regarding postoperative treatment. All our patients had precocious and synchronous BM, so the surgery was prioritized before the beginning of adjuvant treatment in all patients independent of neurological symptoms. Metachronous BM were not included in further analysis. The KPS, on a scale of 0 to 100, with 100 representing the best status, was used to evaluate the quality of life and patients’ physical condition before and shortly after surgery during the same hospital stay. Patients were then grouped based on their KPS value. The groups were compared separably.

The following information was extracted from the electronic medical records: age at diagnosis, sex, tumor localization, preoperative tumor volume, number of BM, complete tumor resection, adjuvant therapy, systemic progression on therapy, and date of death or loss to follow-up. The decision again the adjuvant therapy was based on low functional status and family’s decision made toward palliative treatment. OS was also analyzed. Based on clinical characteristics, the pre- and postoperative RPA and Eastern Cooperative Oncology Group (ECOG) performance status scores were calculated.

Recommendations for surgical treatment were made by an interdisciplinary oncological board based on the presence of a new neurological deficit, tumor mass effect, intratumoral hemorrhage, large lesions in the posterior fossa, and an unknown histology. Epileptic seizures and motor deficits associated with brain edema, even in patients with extremely small lesions, were an indication for surgical resection. None of the patients received biopsy. Neurosurgical resection was performed using various intraoperative neuromonitoring techniques to achieve maximal tumor removal. Postoperative MRI was performed within 72 h of the surgical procedure. The T1-weighted images were analyzed using Origin^® Software (Brainlab, Ver 3.1, Brainlab AG, Munich, Germany), where any contrast-enhancing lesions were determined as tumor rests. The decision for subsequent systemic therapy and radiotherapy was made by the interdisciplinary oncological board based on the clinical status, histopathological results, extent of resection, and the patient’s choice. All patients included in our analysis were initially treatment naïve.

Statistical analyses were performed using SPSS (version 29.0.1.0; IBM, Chicago, IL, USA) and GraphPad Prism (version 8.3.1; La Jolla, CA, USA). In terms of prognosis, OS > 3 months after surgery was considered favorable, whereas OS ≤ 3 months was considered poor. According to the literature patients with extremely low clinical status show survival < 3 months. So, the cutoff 3 months was chosen to define individuals who lived longer than expected based on the current data (8, 15, 16). Recently, Park et el (26). suggested a model for survival of more and less than 3 months in order to identify the patients with low KPS who show better survival and who need further treatment. We found this grouping appropriate in this cohort of patients and tried to identify differences within these groups. The KPS were compared using the Wilcoxon test. Univariate analysis was performed to identify the risk factors that could influence survival after neurosurgical resection. Multivariate Cox regression analysis was used for previously reported significant variables. The Kaplan–Meier method was used to perform survival analysis. Statistical tests were two-sided, and p-values < 0.05 were considered significant.

A total of 140 patients with a KPS of < 70 who underwent surgical resection of BM between 2007 and 2020 met the inclusion criteria, death of 102 patients was documented. The demographic characteristics are summarized in Table 1 . The median age at the time of the first BM diagnosis was 66.1 years (range, 33–93 years). Sex distribution was equal, with 73 (52%) male and 67 (48%) female patients. Primary histology comprised non-small cell lung cancer (n = 66, 47.1%), breast cancer (n = 21, 15%), gastrointestinal tumors (n = 13, 9.3%), renal cell carcinoma (n = 9, 6.4%), prostatic adenocarcinoma (n = 7, 5%), cancer of unknown primary (n = 6, 4.4%), small cell lung cancer (n = 5, 3.6%), ovarian cancer (n = 4, 2.9%), thyroid cancer (n =3, 2.1%), endometrial cancer (n = 2, 1.4%), squamous cell carcinoma (n = 2, 1.4%), chondrosarcoma (n = 1, 0.7%), and sarcoma (n = 1, 0.7%). Regarding the number of ontracranial lesions, 73 (52.1%), 20 (14.3%), 30 (21.5%), and 17 (12.1%) patients had 1, 2, 3, and >3 intracranial lesions, respectively. A total of 111 (79.3%) lesions were supratentorial and 29 (20.7%) were infratentorial. Median preoperative tumor volume was 26.06 cm³ (range, 0.29–94.3). Complete resection was achieved in 83 (59.4%) patients; 77 (55%) underwent postoperative radiotherapy, 34 (24.3%) received adjuvant systemic chemotherapy or immunotherapy, and 32 (22.8%) showed systemic disease progression.

The median OS was 5.6 months (range, 0–58 months) with 52 (37.1%) patients having an OS > 3 months ( Figure 1 ). The median preoperative KPS was 50% (range, 10–60). The median postoperative KPS was 50% (range, 10–100). There was a significant improvement in the clinical status in the group of patients with preoperative KPS of 20% ( Figure 2 , preoperative, 20 vs. mean postoperative, 40 ± 20, p = 0.048). There was no significant difference in the mean ECOG performance status scores (3, range 2–4 vs. 3; range, 0–5) scores after surgical treatment. There was no difference in the preoperative KPS between the two groups, >3 and ≤ 3 months respectively (50; range, 20–60 vs. 50; range, 10–60, p = 0.077).

Several patients showed deterioration of KPS, so we further analyzed surgical complication in the whole cohort. Among 22 patients, who’s KPS reduced after surgery 11 (50%) developed intraoperative or postoperative complications (4 (18.2%) patients underwent a second surgery due to postoperative intracranial hemorrhage or experienced extreme intraoperative blood lost, 3 (13.6%) patients developed hydrocephalus, 2 (9.1%) meningitis, 2 (9.1%) uncontrolled intracranial pressure increase due to cerebral edema). 1 (4.5%) patient developed acute kidney injury shortly after surgery and 1 (4.5%) patient showed pulmonary embolism. Reduction in other 9 (41%) patients was associated with chronic heart or kidney disease or tumor systemic progression. Among 118 patients who showed stable or improved KPS postoperatively 16 (13.6%) experienced postoperative complications (3 (2.5%) meningitis, 2 (1.7%) CSF Leak, 2 (1.7%) hemorrhage, 2 (1.7%) stroke, 7 (6%) wound infection).

Postoperative radiotherapy (hazard ratio [HR], 0.255; confidence interval [CI], 0.164–0.396, p < 0.001) and systemic therapy (HR 0.404; CI 0.235–0.695, p = 0.001) were the prognostic factors associated with favorable survival. In the better survival and poorer prognosis groups, 44 (84.6%) and 32 (36.4%) patients received adjuvant radiation, respectively. Systemic disease progression was another prognostic factor for poor survival (HR 3.638; CI 1.979–6.687, p < 0.001) ( Table 2 ). Seventy-one (80.7%) patients in the group with an OS ≤ 3 months showed disease progression, whereas only 36 (50%) patients in the group with an OS > 3 months presented with systemic progression. Preoperative tumor volume, anatomical localization, number of lesions, and the extent of resection did not affect survival.

In the multivariate Cox hazard regression test, adjuvant radiotherapy (HR 0.0363; CI 0.197–0.670, p = 0.001), adjuvant chemotherapy and immunotherapy (HR 0.474; CI 0.263–0.854, p = 0.013), and absence of systemic disease progression (HR 5.975; CI 2.610–13.677, p < 0.001) were independent factors associated with better survival ( Table 3 ).

Kaplan–Meier curves were used to present the distribution of OS for adjuvant radiotherapy ( Figure 3A ), adjuvant chemotherapy and immunotherapy ( Figure 3B ), and systemic disease progression ( Figure 3C ).

The limitations of our study include its retrospective design, the moderate size of the cohort, and the various histopathological profiles of the tumors treated at our center. The stratification of patients into different pathologies and treatment regimens for each one is limited due to the small number of patients with brain metastases and low KPS. It is difficult to make a single statement regarding the treatment regimens for all the spectrum of pathologies that comprise the oncologic diseases with brain metastases. Nowadays, driver mutations and immune checkpoint expression are extremely relevant for interpretation of treatment-related outcomes. Molecular analysis was not part of routine diagnostics in the early years. We, unfortunately, cannot provide the mutation and expression status for all our patients. Not only the heterogeneity in tumor histology and treatment modalities, but also anatomical localization could further complicate data interpretation. Our study focused exclusively on surgically treated patients with initially rapid KPS deterioration preoperatively associated with BM. Comparison with relapsed patients would strengthen the study.

However, the lack of information about the role of surgery in patients with BM and low KPS makes the data remarkably relevant for decision-making in patients with advanced stages of oncological disease.

Surgical resection of singular or multiple BM can be considered as an efficient treatment modality in patients with low KPS. Surgery may improve functional status and facilitate further tumor-specific treatment even despite the possible surgical complications. Combined treatment with adjuvant radiotherapy and systemic therapy was associated with improved survival in this cohort of patients. Systemic tumor progression has been identified as an independent factor of poor prognosis.

The current study suggests that the benefit of initial surgical resection on the clinical outcome and OS in patients with BM is significant. Our findings challenge the current paradigm of BM management, raising the opportunity to perform further randomized clinical trials to investigate the role of initial systematic surgical resection of BM in patients with low KPS.