Prognostic factors and quality of life in non-small cell lung cancer with leptomeningeal metastasis: a dichotomous pattern based on parenchymal involvement Authors: Yiyao Zhu, Zhiyuan Qiu, Lihua Zhu, Chao Lu, Tingjuan Zhang, Jun Chen, Yan Wang, Chaoyang Wu Published: 2025 ABSTRACT ObjectiveThis study retrospectively analyzed prognostic factors and quality of life in patients with Non-Small Cell Lung Cancer–Leptomeningeal Metastasis (NSCLC-LM), stratified by the presence or absence of concomitant Parenchymal Metastasis (PM), to inform clinical stratification and individualized therapy.MethodsWe retrospectively reviewed 43 NSCLC-LM patients diagnosed at the Affiliated People’s Hospital of Jiangsu University from September 2016 to April 2025, and classified them into LM-PM and LM-noPM groups according to brain parenchymal metastasis status. The primary endpoint was median overall survival (mOS). Prognostic factors were assessed, and quality of life was evaluated before and after treatment. Survival analysis was performed using the Kaplan-Meier method and log-rank test. Univariate and multivariate Cox regression analyses were conducted to identify independent prognostic factors. Quality of life scores were compared using paired sample t-tests. Scale reliability was assessed via Pearson correlation coefficients and Cronbach’s alpha. Statistical significance was defined as a two-sided P-value < 0.05.ResultsThe mOS was 5.0 months in both groups. Notably, the LM-noPM group demonstrated higher 1-year and 2-year survival rates, indicating a possible long-term survival advantage, although this trend did not reach statistical significance. Multivariate analysis identified non-receipt of anti-angiogenic therapy, adrenal metastasis, and elevated Platelet–Lymphocyte Ratio (PLR) as independent risk factors for poor prognosis. Among patients with good performance status, meningeal irritation signs also constituted an independent risk factor. Quality of life assessment revealed that comprehensive treatment significantly improved global health status and functional domains while reducing symptom burden.ConclusionAlthough NSCLC-LM patients without PM showed a non-significant trend toward improved long-term survival (reflected by numerically higher 1- and 2-year survival rates), the key independent risk factors associated with poor prognosis were the absence of anti-angiogenic therapy after LM diagnosis, the presence of adrenal metastasis, and an elevated PLR. These findings remained consistent among patients with good performance status, among whom meningeal irritation signs also independently predicted worse outcomes. Moreover, comprehensive treatment significantly alleviated symptom burden and enhanced quality of life across multiple domains, supporting longer survival with improved quality. Leptomeningeal metastasis (LM), a severe complication of metastatic cancer, occurs when tumor cells infiltrate the meningeal spaces. In advanced non-small cell lung cancer (NSCLC), LM occurs in approximately 3%-5% of patients. The development of LM markedly impairs patients’ quality of life and reduces overall survival (OS) [1]. The incidence is notably higher in patients with epidermal growth factor receptor (EGFR) mutations, reaching 9%-16% [2]. NSCLC-LM patients frequently experience rapid clinical deterioration and a poor prognosis, with a median OS typically between 1 and 3 months. As therapeutic options diversify, combined strategies-including systemic treatments like chemotherapy, immunotherapy, targeted therapy, and intrathecal injections alongside local interventions such as whole-brain radiotherapy (WBRT) and stereotactic radiosurgery (SRS), are now employed clinically. These approaches have extended patient survival from 1 to 3 months to 3–11 months [3–5], reflecting significant advances in the treatment of NSCLC. Together with refinements in imaging techniques, these improvements have increased the detection and diagnosis rates of LM. Previous research has primarily centered on patients with parenchymal metastasis (PM), often excluding those with LM. Consequently, dedicated research data on NSCLC-LM patients remain limited. This study retrospectively investigates prognostic factors and quality of life in NSCLC-LM patients to inform risk stratification and guide individualized treatment strategies for high-risk patients. Materials and methods Inclusion and exclusion criteria This study included patients diagnosed and treated for NSCLC-LM at the Affiliated People’s Hospital of Jiangsu University between September 2016 and April 2025. Patients were eligible if they had a pathologically confirmed NSCLC from either the primary tumor or a metastatic lesion and exhibited typical LM imaging findings on contrast-enhanced magnetic resonance imaging (MRI). Exclusion criteria included incomplete clinical or imaging data, inability to assess treatment efficacy or follow-up outcomes, and contraindications to MRI examination. Ultimately, a total of 43 patients from our institution were included in the study. The primary endpoint was OS, defined as the interval from the LM diagnosis date to death from any cause or the last follow-up, which was censored on September 30, 2025. We estimated the median survival time using the Kaplan-Meier method. The Medical Ethics Committee of the Affiliated People’s Hospital of Jiangsu University approved this study (Ethics Approval: SQK-2025127-W). Study design All patients underwent a comprehensive baseline assessment, which included a detailed medical history, thorough physical and neurological examinations, and imaging studies. We collected data on age, gender, performance status (PS) score, Tumor Node Metastasis (TNM) stage, comorbidities, primary tumor biological characteristics (including driver gene mutations and histological subtype), and treatments before and after LM diagnosis, such as surgery, chemotherapy, immune checkpoint blockers, targeted therapy, WBRT, and SRS. The quality of life of each patient was assessed using the Chinese version of the European Organization for Research and Treatment of Cancer Quality of Life Questionnaire (EORTC QLQ-C30) [6]. This study aimed to use these data to investigate prognostic factors in NSCLC-LM and to compare quality of life changes before and after treatment. The diagnosis was confirmed in all patients by contrast-enhanced MRI, and two or more experienced radiologists independently reviewed the images. The diagnostic criteria were based on established criteria for abnormal leptomeningeal enhancement [7]. Statistical analysis All statistical analyses were performed using SPSS software (version 27.0.1). Descriptive statistics were used to summarize baseline patient characteristics. Survival curves were estimated using the Kaplan-Meier method, with differences between groups compared via the log-rank test. Potential prognostic factors were evaluated using univariate and multivariate Cox proportional hazards regression analyses. Variables that demonstrated a potential association with OS in the univariate analysis (with a significance level of P < 0.05) were selected and included in the multivariate Cox proportional hazards model for further evaluation. The optimal cut-off values of the neutrophil-to-lymphocyte ratio (NLR) and platelet-to-lymphocyte ratio (PLR) for predicting OS were determined using receiver operating characteristic (ROC) curve analysis. For all subsequent survival analyses, patients were stratified into high- and low-value groups based on these cut-off points. Comparisons of quality of life scores before and after treatment were analyzed using paired-sample t-tests. Scale reliability was assessed with Pearson correlation coefficients and Cronbach’s alpha coefficient. All tests were two-sided, and a P-value < 0.05 was considered statistically significant. Results Baseline characteristics A total of 43 patients with NSCLC-LM were enrolled in this study, and their baseline characteristics are summarized in Table 1. The median age was 65 years, and adenocarcinoma was the most common histological type, representing 97.7% of cases. For the comparison of baseline characteristics presented in the main tables, EGFR status was categorized as ‘19 deletion’, ‘21 L858R’, or ‘Unknown’. The ‘Unknown’ category included patients with wild-type EGFR, those who were untested, and patients with an unspecified EGFR mutation. Among the 43 patients, 21 (48.8%) had a confirmed EGFR mutation, including 7 (16.3%) with an exon 19 deletion and 14 (32.6%) with the 21 L858R mutation. Of the remaining 22 patients, 8 (18.6% of the total cohort) had confirmed wild-type (EGFR-negative) tumors, 8 (18.6%) did not undergo genetic testing, and for 6 (14.0%), the testing was performed, but the specific mutation subtype could not be further specified beyond ‘EGFR-positive’. A detailed breakdown of the EGFR testing status is provided in Table 2. Subgroup analysis based on parenchymal involvement status showed that the LM-PM group comprised 32 patients (74.4%), with 28 deaths (87.5%). The LM-noPM group consisted of 11 patients (25.6%), with 9 deaths (81.8%). By the last follow-up, 37 patients (86.0%) had died. Table 1Baseline demographic and clinical characteristics [n (%)]CharacteristicLM-PM Group (n = 32)LM-noPM Group (n = 11)Total (n = 43)P-valueGender0.728 Male15(46.9%)4(36.4%)19(44.2%) Female17(53.1%)7(63.6%)24(55.8%)Age, years0.536 <6516(50.0%)5(45.5%)21(48.8%) ≥ 6516(50.0%)6(54.5%)22(51.2%)Smoking0.480 No18(56.2%)8(72.7%)26(60.5%) Yes14(43.8%)3(27.3%)17(39.5%)Hypertension0.719 No21(65.6%)6(54.5%)27(62.8%) Yes11(34.4%)5(45.5%)16(37.2%)PS Score0.637 ≤ 14(12.5%)2(18.2%)6(14.0%) ≥ 228(87.5%)9(81.8%)37(86.0%)EGFR Mutation0.805 19 del6(18.8%)1(9.1%)7(16.3%) 21L858R10(31.2%)4(36.4%)14(32.6%) Unknown16(50.0%)6(54.5%)22(51.1%) Pathology0.256Non-adenocarcinoma0(0.0%)1(9.1%)1(2.3%) Adenocarcinoma32(100.0%)10(90.9%)42(97.7%) Tumor Stage0.518 IA3(9.4%)1(9.1%)4(9.3%) IB-IIIB6(18.8%)3(27.3%)9(20.9%) IVA2(6.2%)2(18.2%)4(9.3%) IVB21(65.6%)5(45.4%)26(60.5%)Primary Tumor Surgery0.473 No22(68.8%)6(54.5%)28(65.1%) Yes10(31.2%)5(45.5%)15(34.9%)Brain Metastasis Surgery0.558 No29(90.6%)11(100.0%)40(93.0%) Yes3(9.4%)0(0.0%)3(7.0%)Hydrocephalus0.501 No16(50.0%)7(63.6%)23(53.5%) Yes16(50.0%)4(36.4%)20(46.5%)Meningeal Irritation Signs0.347 No4(12.5%)3(27.3%)7(16.3%) Yes28(87.5%)8(72.7%)36(83.7%)Cranial Nerve Involvement0.698 No25(78.1%)8(72.7%)33(76.7%) Yes7(21.9%)3(27.3%)10(23.3%)Signs of Increased Intracranial Pressure0.401 No5(15.6%)3(27.3%)8(18.6%) Yes27(84.4%)8(72.7%)35(81.4%)Spinal Cord and Nerve Root Irritation0.294 No16(50.0%)8(72.7%)24(55.8%) Yes16(50.0%)3(27.3%)19(44.2%)Number of Metastatic Sites0.515 <310(31.2%)4(36.4%)14(32.6%) ≥ 322(68.8%)7(63.6%)29(67.4%)Liver Metastasis0.744 No31(96.9%)11(100.0%)42(97.7%) Yes1(3.1%)0(0.0%)1(2.3%)Adrenal Metastasis0.558 No29(90.6%)11(100.0%)40(93.0%) Yes3(9.4%)0(0.0%)3(7.0%)Bone Metastasis0.080 No10(31.2%)7(63.6%)17(39.5%) Yes22(68.8%)4(36.4%)26(60.5%)Prior Systemic Chemotherapy (before LM)0.719 No12(37.5%)3(27.3%)15(34.9%) Yes20(62.5%)8(72.7%)28(65.1%)Prior Immunotherapy (before LM)0.583 No25(78.1%)9(81.8%)34(79.1%) Yes7(21.9%)2(18.2%)9(20.9%)Prior Targeted Therapy (before LM)0.457 No9(28.1%)5(45.5%)14(32.6%) Yes23(71.9%)6(54.5%)29(67.4%)Prior Anti-angiogenic Therapy (before LM)0.736 No14(43.8%)4(36.4%)18(41.9%) Yes18(56.2%)7(63.6%)25(58.1%)Prior WBRT (before LM)0.312 No26(81.2%)11(100.0%)37(86.0%) Yes6(18.8%)0(0.0%)6(14.0%)Note: PS: Performance Status; EGFR: Epidermal Growth Factor Receptor; LM: Leptomeningeal Metastasis; WBRT: Whole Brain Radiotherapy. Data are presented as n (%). Percentages are calculated based on the total number of patients in each column (LM-PM, n = 32; LM-noPM, n = 11; Total, n = 43) Table 1Baseline demographic and clinical characteristics [n (%)]CharacteristicLM-PM Group (n = 32)LM-noPM Group (n = 11)Total (n = 43)P-valueGender0.728 Male15(46.9%)4(36.4%)19(44.2%) Female17(53.1%)7(63.6%)24(55.8%)Age, years0.536 <6516(50.0%)5(45.5%)21(48.8%) ≥ 6516(50.0%)6(54.5%)22(51.2%)Smoking0.480 No18(56.2%)8(72.7%)26(60.5%) Yes14(43.8%)3(27.3%)17(39.5%)Hypertension0.719 No21(65.6%)6(54.5%)27(62.8%) Yes11(34.4%)5(45.5%)16(37.2%)PS Score0.637 ≤ 14(12.5%)2(18.2%)6(14.0%) ≥ 228(87.5%)9(81.8%)37(86.0%)EGFR Mutation0.805 19 del6(18.8%)1(9.1%)7(16.3%) 21L858R10(31.2%)4(36.4%)14(32.6%) Unknown16(50.0%)6(54.5%)22(51.1%) Pathology0.256Non-adenocarcinoma0(0.0%)1(9.1%)1(2.3%) Adenocarcinoma32(100.0%)10(90.9%)42(97.7%) Tumor Stage0.518 IA3(9.4%)1(9.1%)4(9.3%) IB-IIIB6(18.8%)3(27.3%)9(20.9%) IVA2(6.2%)2(18.2%)4(9.3%) IVB21(65.6%)5(45.4%)26(60.5%)Primary Tumor Surgery0.473 No22(68.8%)6(54.5%)28(65.1%) Yes10(31.2%)5(45.5%)15(34.9%)Brain Metastasis Surgery0.558 No29(90.6%)11(100.0%)40(93.0%) Yes3(9.4%)0(0.0%)3(7.0%)Hydrocephalus0.501 No16(50.0%)7(63.6%)23(53.5%) Yes16(50.0%)4(36.4%)20(46.5%)Meningeal Irritation Signs0.347 No4(12.5%)3(27.3%)7(16.3%) Yes28(87.5%)8(72.7%)36(83.7%)Cranial Nerve Involvement0.698 No25(78.1%)8(72.7%)33(76.7%) Yes7(21.9%)3(27.3%)10(23.3%)Signs of Increased Intracranial Pressure0.401 No5(15.6%)3(27.3%)8(18.6%) Yes27(84.4%)8(72.7%)35(81.4%)Spinal Cord and Nerve Root Irritation0.294 No16(50.0%)8(72.7%)24(55.8%) Yes16(50.0%)3(27.3%)19(44.2%)Number of Metastatic Sites0.515 <310(31.2%)4(36.4%)14(32.6%) ≥ 322(68.8%)7(63.6%)29(67.4%)Liver Metastasis0.744 No31(96.9%)11(100.0%)42(97.7%) Yes1(3.1%)0(0.0%)1(2.3%)Adrenal Metastasis0.558 No29(90.6%)11(100.0%)40(93.0%) Yes3(9.4%)0(0.0%)3(7.0%)Bone Metastasis0.080 No10(31.2%)7(63.6%)17(39.5%) Yes22(68.8%)4(36.4%)26(60.5%)Prior Systemic Chemotherapy (before LM)0.719 No12(37.5%)3(27.3%)15(34.9%) Yes20(62.5%)8(72.7%)28(65.1%)Prior Immunotherapy (before LM)0.583 No25(78.1%)9(81.8%)34(79.1%) Yes7(21.9%)2(18.2%)9(20.9%)Prior Targeted Therapy (before LM)0.457 No9(28.1%)5(45.5%)14(32.6%) Yes23(71.9%)6(54.5%)29(67.4%)Prior Anti-angiogenic Therapy (before LM)0.736 No14(43.8%)4(36.4%)18(41.9%) Yes18(56.2%)7(63.6%)25(58.1%)Prior WBRT (before LM)0.312 No26(81.2%)11(100.0%)37(86.0%) Yes6(18.8%)0(0.0%)6(14.0%)Note: PS: Performance Status; EGFR: Epidermal Growth Factor Receptor; LM: Leptomeningeal Metastasis; WBRT: Whole Brain Radiotherapy. Data are presented as n (%). Percentages are calculated based on the total number of patients in each column (LM-PM, n = 32; LM-noPM, n = 11; Total, n = 43) Baseline demographic and clinical characteristics [n (%)] Note: PS: Performance Status; EGFR: Epidermal Growth Factor Receptor; LM: Leptomeningeal Metastasis; WBRT: Whole Brain Radiotherapy. Data are presented as n (%). Percentages are calculated based on the total number of patients in each column (LM-PM, n = 32; LM-noPM, n = 11; Total, n = 43) Table 2Detailed breakdown of EGFR mutation testing statusCategoryNumberPercentage (%)Total enrolled patients43100Patients without completed genetic testing818.6(8/43)Patients with completed genetic testing3581.4(35/43) EGFR mutation negative822.9(8/35) EGFR mutation positive2777.1(27/35) Exon 19 deletion (19del)725.9(7/27) Exon 21 L858R mutation1451.9(14/27) Unknown mutation type622.2(6/27)Note: EGFR: Epidermal Growth Factor Receptor Table 2Detailed breakdown of EGFR mutation testing statusCategoryNumberPercentage (%)Total enrolled patients43100Patients without completed genetic testing818.6(8/43)Patients with completed genetic testing3581.4(35/43) EGFR mutation negative822.9(8/35) EGFR mutation positive2777.1(27/35) Exon 19 deletion (19del)725.9(7/27) Exon 21 L858R mutation1451.9(14/27) Unknown mutation type622.2(6/27)Note: EGFR: Epidermal Growth Factor Receptor Detailed breakdown of EGFR mutation testing status Note: EGFR: Epidermal Growth Factor Receptor Treatment profiles The median OS was 5.0 months in both the LM-PM and LM-noPM groups. Kaplan-Meier survival analysis showed a 1-year survival rate was 13.5% in the LM-PM group compared with 40.4% in the LM-noPM group. The 2-year survival rate was 4.3% in the LM-PM group and 17.1% in the LM-noPM group. Although the Log-rank test did not reach statistical significance (P = 0.239), this trend suggests a potential long-term survival difference (Fig. 1), indicating that a higher central nervous system metastatic burden may be associated with worse prognosis. Initial treatment before the diagnosis of LM was based on the tumor stage at initial presentation. Among the cohort, 4 patients (9.3%) with stage IA disease underwent radical lung cancer surgery alone. Of the 9 patients (20.9%) with stage IB-IIIB disease, 8 (18.6%) received radical surgery followed by postoperative adjuvant chemotherapy, while 1 (2.3%) received surgery with adjuvant targeted therapy. For the 4 patients with stage IVA (9.3%) and 26 patients with stage IVB (60.5%), systemic treatment was guided by molecular pathology: patients harboring sensitive driver gene mutations received targeted therapy, while those without detectable mutations were treated with immunotherapy, chemotherapy, or anti-angiogenic therapy. Parenchymal Brain Metastases (PM) and Local Therapy: Among the 32 patients in the LM-PM group, parenchymal brain metastases were diagnosed before LM in 28 cases (87.5%). This included 9 patients (28.1%) with brain metastases present at initial NSCLC diagnosis and 19 patients (59.4%) who developed brain metastases during subsequent treatment. The distribution across tumor stages was as follows: IA, 2 patients (6.2%); IB–IIIB, 6 patients (18.8%); IVA, 1 patient (3.1%); and IVB, 10 patients (31.3%). Local treatments for PM consisted of SRS in 11 patients (25.6%), WBRT alone in 14 patients (32.6%), and WBRT plus a local boost in 2 patients (4.7%). One patient in the latter group subsequently underwent palliative surgery for progressive intracranial metastases after radiotherapy. Additionally, one patient with stage IVB (2.3%) and initial oligometastatic intracranial involvement underwent surgery for both the primary lung lesion and intracranial metastasis, followed by WBRT plus a local boost when new brain metastases emerged. Following the diagnosis of LM, systemic treatments included chemotherapy in 25.6% of patients, targeted therapy in 55.8%, and anti-angiogenic therapy in 25.6%. Pemetrexed-based regimens were the predominant form of chemotherapy, third-generation EGFR-tyrosine kinase inhibitors were the main primary targeted therapy, and bevacizumab was the most commonly used anti-angiogenic drug. Regarding local treatment of LM, 21 patients (48.8%) received WBRT with or without a local boost. Two patients (4.6%) underwent Ommaya reservoir placement followed by intrathecal pemetrexed chemotherapy, while 4 patients (9.3%) received best supportive care only. Figure 2 illustrates the potential efficacy of comprehensive treatment based on neuroimaging findings from a representative case. The patient was diagnosed with lung adenocarcinoma with an EGFR exon 21 L858R mutation. Pre-treatment contrast-enhanced MRI (Figs. 2a-d) demonstrated diffuse, linear abnormal enhancement along the cerebral sulci, tentorium cerebelli, and lateral ventricular walls, consistent with the typical radiologic features of LM. After three months of combination therapy, follow-up contrast-enhanced MRI (Figs. 2e-h) revealed near-complete resolution of the previous abnormal leptomeningeal enhancement with normalized meningeal contours, indicating substantial radiological improvement. This imaging response corresponded with improvement of clinical symptoms and quality of life, indicating the importance of comprehensive treatment in NSCLC-LM. Fig. 1Kaplan-Meier curve for Overall Survival (OS). The median OS (mOS) was 5.0 months (95%CI༚3.495–6.505) in the LM-PM group and 5.0 months (95%CI: 3.989–8.011) in the LM-noPM group. The 1-year survival rates were 17.1% (95%CI: 0.073–0.404) and 40.4% (95%CI: 0.190–0.859), and the 2-year survival rates were 4.3% (95%CI: 0.006–0.287) and 13.5% (95%CI: 0.090–0809) for the LM-PM and LM-noPM groups, respectively. Comparison by log-rank test: P = 0.239. Hazard ratio (LM-noPM vs. LM-PM) = 0.635 (95%CI༚0.285–1.414). The table below the curve indicates the number of patients at risk at each time point Kaplan-Meier curve for Overall Survival (OS). The median OS (mOS) was 5.0 months (95%CI༚3.495–6.505) in the LM-PM group and 5.0 months (95%CI: 3.989–8.011) in the LM-noPM group. The 1-year survival rates were 17.1% (95%CI: 0.073–0.404) and 40.4% (95%CI: 0.190–0.859), and the 2-year survival rates were 4.3% (95%CI: 0.006–0.287) and 13.5% (95%CI: 0.090–0809) for the LM-PM and LM-noPM groups, respectively. Comparison by log-rank test: P = 0.239. Hazard ratio (LM-noPM vs. LM-PM) = 0.635 (95%CI༚0.285–1.414). The table below the curve indicates the number of patients at risk at each time point Fig. 2Contrast-enhanced brain MRI in a patient with NSCLC-LM before and after treatment. The sequence is Turbo Spin Echo. (a-d) Pre-treatment images reveal diffuse abnormal leptomeningeal enhancement (2024-08-23). (e-h) Post-treatment images, acquired 3 months after targeted therapy, demonstrate near-complete resolution of the abnormal enhancement (2024-11-28) Contrast-enhanced brain MRI in a patient with NSCLC-LM before and after treatment. The sequence is Turbo Spin Echo. (a-d) Pre-treatment images reveal diffuse abnormal leptomeningeal enhancement (2024-08-23). (e-h) Post-treatment images, acquired 3 months after targeted therapy, demonstrate near-complete resolution of the abnormal enhancement (2024-11-28) Analysis of prognostic factors Univariate and multivariate COX regression analyses (Table 3) identified three independent predictors of reduced OS: absence of anti-angiogenic therapy following LM diagnosis (P = 0.018), presence of adrenal metastasis (P = 0.043). Using an ROC-derived optimal cut-off value of 120.2 (AUC = 0.743; Fig. 3), patients with high PLR (≥ 120.2) showed significantly worse OS compared to those with low PLR (HR = 2.627,95%CI:1.166–5.922; P = 0.020). To control for the potential confounding impact of PS on treatment selection and prognosis, we conducted a subgroup analysis stratified by PS score (Table 4). This analysis showed that among patients with good PS (PS ≤ 1), the lack of anti-angiogenic therapy after LM diagnosis (P = 0.041), the presence of meningeal irritation signs (P = 0.045), and adrenal metastasis (P = 0.023) were independent adverse prognostic factors. Subsequent survival analysis indicated that neither a history of primary tumor resection nor brain metastasis surgery performed before LM diagnosis was significantly associated with patient prognosis. Similarly, the timing of radiotherapy relative to LM diagnosis and EGFR mutation status showed no significant impact on survival outcomes. Fig. 3Receiver operating characteristic (ROC) curve analysis for the platelet-to-lymphocyte ratio (PLR) in predicting overall survival Receiver operating characteristic (ROC) curve analysis for the platelet-to-lymphocyte ratio (PLR) in predicting overall survival Table 3Univariate and multivariate analysis of prognostic factorsFactorNo. of Patients (%)Univariate AnalysisMultivariate AnalysisP-valueP-valueHR95%CIGender0.697 Male19(44.2%) Female24(55.8%)Age, years0.225 <6521(48.8%) ≥ 6522(51.2%)Smoking0.617 No26(60.5%) Yes17(39.5%)Hypertension0.661 No27(62.8%) Yes16(37.2%)PS Score0.947 ≤ 16(14.0%) >137(86.0%)EGFR Mutation0.091 19 del7(16.3%) 21 L858R14(32.6%) Unknown22(51.1%) Pathology0.650Non-adenocarcinoma1(2.3%) Adenocarcinoma42(97.7%) Tumor Stage0.474 IA4(9.3%) IB-IIIB9(20.9%) IVA4(9.3%) IVB26(60.5%)Primary Tumor Surgery0.684 No28(65.1%) Yes15(34.9%)Brain Metastasis Surgery0.512 No40(93.0%) Yes3(7.0%)Hydrocephalus0.205 No23(53.5%) Yes20(46.5%)Meningeal Irritation Signs0.127 No7(16.3%) Yes36(83.7%)Cranial Nerve Involvement0.545 No33(76.7%) Yes10(23.3%)Signs of Increased Intracranial Pressure0.533 No8(18.6%) Yes35(81.4%)Spinal Cord and Nerve Root Irritation0.693 No24(55.8%) Yes19(44.2%)Number of Metastatic Sites0.135 <314(32.6%) ≥ 329(67.4%)Liver Metastasis0.803 No42(97.7%) Yes1(2.3%)Adrenal Metastasis0.0460.043 No40(93.0%)0.0270.076–0.960 Yes3(7.0%)1 (Ref)Bone Metastasis0.673 No17(39.5%) Yes26(60.5%)Prior Systemic Chemotherapy (before LM)0.200 No15(34.9%) Yes28(65.1%)Prior Immunotherapy (before LM)0.324 No34(79.1%) Yes9(20.9%)Prior Targeted Therapy (before LM)0.998 No14(32.6%) Yes29(67.4%)Prior Anti-angiogenic Therapy (before LM)0.538 No18(41.9%) Yes25(58.1%)Prior WBRT (before LM)0.888 No37(86.0%) Yes6(14.0%)Systemic Chemotherapy (after LM)0.127 No32(74.4%) Yes11(25.6%)Immunotherapy (after LM)0.185 No34(79.1%) Yes9(20.9%)Targeted Therapy (after LM)0.064 No19(44.2%) Yes24(55.8%)Anti-angiogenic Therapy (after LM)0.0330.018 No20(46.5%)2.4401.162–5.124 Yes23(53.5%)1 (Ref)WBRT (after LM)0.399 No22(51.2%) Yes21(48.8%) NLR0.133 high-NLR low-NLR PLR0.0110.020 high-PLR2.6271.116–5.922 low-PLR1 (Ref)Note: PS score: Performance Status; EGFR: Epidermal Growth Factor Receptor; LM: Leptomeningeal Metastasis; WBRT: Whole Brain Radiotherapy; NLR: Neutrophil-to-Lymphocyte Ratio; PLR: Platelet-to-Lymphocyte Ratio. Data in the ‘No. of Patients’ column are presented as n (%). Percentages are calculated based on the total cohort (N = 43) Table 3Univariate and multivariate analysis of prognostic factorsFactorNo. of Patients (%)Univariate AnalysisMultivariate AnalysisP-valueP-valueHR95%CIGender0.697 Male19(44.2%) Female24(55.8%)Age, years0.225 <6521(48.8%) ≥ 6522(51.2%)Smoking0.617 No26(60.5%) Yes17(39.5%)Hypertension0.661 No27(62.8%) Yes16(37.2%)PS Score0.947 ≤ 16(14.0%) >137(86.0%)EGFR Mutation0.091 19 del7(16.3%) 21 L858R14(32.6%) Unknown22(51.1%) Pathology0.650Non-adenocarcinoma1(2.3%) Adenocarcinoma42(97.7%) Tumor Stage0.474 IA4(9.3%) IB-IIIB9(20.9%) IVA4(9.3%) IVB26(60.5%)Primary Tumor Surgery0.684 No28(65.1%) Yes15(34.9%)Brain Metastasis Surgery0.512 No40(93.0%) Yes3(7.0%)Hydrocephalus0.205 No23(53.5%) Yes20(46.5%)Meningeal Irritation Signs0.127 No7(16.3%) Yes36(83.7%)Cranial Nerve Involvement0.545 No33(76.7%) Yes10(23.3%)Signs of Increased Intracranial Pressure0.533 No8(18.6%) Yes35(81.4%)Spinal Cord and Nerve Root Irritation0.693 No24(55.8%) Yes19(44.2%)Number of Metastatic Sites0.135 <314(32.6%) ≥ 329(67.4%)Liver Metastasis0.803 No42(97.7%) Yes1(2.3%)Adrenal Metastasis0.0460.043 No40(93.0%)0.0270.076–0.960 Yes3(7.0%)1 (Ref)Bone Metastasis0.673 No17(39.5%) Yes26(60.5%)Prior Systemic Chemotherapy (before LM)0.200 No15(34.9%) Yes28(65.1%)Prior Immunotherapy (before LM)0.324 No34(79.1%) Yes9(20.9%)Prior Targeted Therapy (before LM)0.998 No14(32.6%) Yes29(67.4%)Prior Anti-angiogenic Therapy (before LM)0.538 No18(41.9%) Yes25(58.1%)Prior WBRT (before LM)0.888 No37(86.0%) Yes6(14.0%)Systemic Chemotherapy (after LM)0.127 No32(74.4%) Yes11(25.6%)Immunotherapy (after LM)0.185 No34(79.1%) Yes9(20.9%)Targeted Therapy (after LM)0.064 No19(44.2%) Yes24(55.8%)Anti-angiogenic Therapy (after LM)0.0330.018 No20(46.5%)2.4401.162–5.124 Yes23(53.5%)1 (Ref)WBRT (after LM)0.399 No22(51.2%) Yes21(48.8%) NLR0.133 high-NLR low-NLR PLR0.0110.020 high-PLR2.6271.116–5.922 low-PLR1 (Ref)Note: PS score: Performance Status; EGFR: Epidermal Growth Factor Receptor; LM: Leptomeningeal Metastasis; WBRT: Whole Brain Radiotherapy; NLR: Neutrophil-to-Lymphocyte Ratio; PLR: Platelet-to-Lymphocyte Ratio. Data in the ‘No. of Patients’ column are presented as n (%). Percentages are calculated based on the total cohort (N = 43) Univariate and multivariate analysis of prognostic factors Note: PS score: Performance Status; EGFR: Epidermal Growth Factor Receptor; LM: Leptomeningeal Metastasis; WBRT: Whole Brain Radiotherapy; NLR: Neutrophil-to-Lymphocyte Ratio; PLR: Platelet-to-Lymphocyte Ratio. Data in the ‘No. of Patients’ column are presented as n (%). Percentages are calculated based on the total cohort (N = 43) Table 4Multivariate analysis of subgroups based on PS scoreFactorPS ≤ 1(n = 6)PS ≥ 2(n = 37)P-valueHR95%CIGender Male1(16.7%)18(48.6%) Female5(83.3%)19(51.4%)Age, years <651(16.7%)20(54.1%) ≥ 655(83.3%)17(45.9%)Smoking No6(100.0%)20(54.1%) Yes0(0.0%)17(45.9%)Hypertension No3(50.0%)24(64.9%) Yes3(50.0%)13(35.1%)EGFR Mutation 19 del1(16.7%)6(16.2%) 21 L858R2(33.3%)12(32.4%) Unknown3(50.0%)19(51.4%) PathologyNon-adenocarcinoma0(0.0%)1(2.7%) Adenocarcinoma6(100.0%)36(97.3%) Tumor Stage IA0(0.0%)4(10.8%) IB-IIIB1(16.7%)8(21.6%) IVA0(0.0%)4(10.8%) IVB5(83.3%)21(56.8%)Brain Parenchymal Metastasis No2(33.3%)9(24.3%) Yes4(66.7%)28(75.7%)Primary Tumor Surgery No5(83.3%)23(62.2%) Yes1(16.7%)14(37.8%)Brain Metastasis Surgery No5(83.3%)35(94.6%) Yes1(16.7%)2(5.4%)Hydrocephalus No4(66.7%)19(51.4%) Yes2(33.3%)18(48.6%)Meningeal Irritation Signs0.045 No1(16.7%)6(16.2%)2.7651.022–7.482 Yes5(83.3%)31(83.8%)1 (Ref)Cranial Nerve Involvement No5(83.3%)28(75.7%) Yes1(16.7%)9(24.3%)Signs of Increased Intracranial Pressure No1(16.7%)7(18.9%) Yes5(83.3%)30(81.1%)Spinal Cord and Nerve Root Irritation No5(83.3%)19(51.4%) Yes1(16.7%)18(48.6%)Number of Metastatic Sites <32(33.3%)12(32.4%) ≥ 34(66.7%)25(67.6%)Liver Metastasis No5(83.3%)37(100.0%) Yes1(16.7%)0(0.0%)Adrenal Metastasis0.023 No6(100.0%)34(91.9%)0.2200.060–0.809 Yes0(0.0%)3(8.1%)1 (Ref)Bone Metastasis No2(33.3%)15(40.5%) Yes4(66.7%)22(59.5%)Prior Systemic Chemotherapy (before LM) No2(33.3%)13(35.1%) Yes4(66.7%)24(64.9%)Prior Immunotherapy (before LM) No3(50.0%)31(83.8%) Yes3(50.0%)6(16.2%)Prior Targeted Therapy (before LM) No1(16.7%)31(83.8%) Yes5(83.3%)6(16.2%)Prior Anti-angiogenic Therapy (before LM) No2(33.3%)16(43.2%) Yes4(66.7%)21(56.8%)Prior WBRT (before LM) No5(83.3%)32(86.5%) Yes1(16.7%)5(13.5%)Systemic Chemotherapy (after LM) No5(83.3%)27(73.0%) Yes1(16.7%)10(27.0%)Immunotherapy (after LM) No5(83.3%)29(78.4%) Yes1(16.7%)8(21.6%)Targeted Therapy (after LM) No1(16.7%)18(48.6%) Yes5(83.3%)19(51.4%)Anti-angiogenic Therapy (after LM)0.041 No1(16.7%)19(51.4%)2.1311.033–4.394 Yes5(83.3%)18(48.6%)1 (Ref)WBRT (after LM) No3(50.0%)19(51.4%) Yes3(50.0%)18(48.6%)Note: EGFR: Epidermal Growth Factor Receptor; LM: Leptomeningeal Metastasis; WBRT: Whole Brain Radiotherapy. Data are presented as n (%). Percentages are calculated based on the total number of patients in each row subgroup (PS ≤ 1, n = 6; PS ≥ 2, n = 37) Table 4Multivariate analysis of subgroups based on PS scoreFactorPS ≤ 1(n = 6)PS ≥ 2(n = 37)P-valueHR95%CIGender Male1(16.7%)18(48.6%) Female5(83.3%)19(51.4%)Age, years <651(16.7%)20(54.1%) ≥ 655(83.3%)17(45.9%)Smoking No6(100.0%)20(54.1%) Yes0(0.0%)17(45.9%)Hypertension No3(50.0%)24(64.9%) Yes3(50.0%)13(35.1%)EGFR Mutation 19 del1(16.7%)6(16.2%) 21 L858R2(33.3%)12(32.4%) Unknown3(50.0%)19(51.4%) PathologyNon-adenocarcinoma0(0.0%)1(2.7%) Adenocarcinoma6(100.0%)36(97.3%) Tumor Stage IA0(0.0%)4(10.8%) IB-IIIB1(16.7%)8(21.6%) IVA0(0.0%)4(10.8%) IVB5(83.3%)21(56.8%)Brain Parenchymal Metastasis No2(33.3%)9(24.3%) Yes4(66.7%)28(75.7%)Primary Tumor Surgery No5(83.3%)23(62.2%) Yes1(16.7%)14(37.8%)Brain Metastasis Surgery No5(83.3%)35(94.6%) Yes1(16.7%)2(5.4%)Hydrocephalus No4(66.7%)19(51.4%) Yes2(33.3%)18(48.6%)Meningeal Irritation Signs0.045 No1(16.7%)6(16.2%)2.7651.022–7.482 Yes5(83.3%)31(83.8%)1 (Ref)Cranial Nerve Involvement No5(83.3%)28(75.7%) Yes1(16.7%)9(24.3%)Signs of Increased Intracranial Pressure No1(16.7%)7(18.9%) Yes5(83.3%)30(81.1%)Spinal Cord and Nerve Root Irritation No5(83.3%)19(51.4%) Yes1(16.7%)18(48.6%)Number of Metastatic Sites <32(33.3%)12(32.4%) ≥ 34(66.7%)25(67.6%)Liver Metastasis No5(83.3%)37(100.0%) Yes1(16.7%)0(0.0%)Adrenal Metastasis0.023 No6(100.0%)34(91.9%)0.2200.060–0.809 Yes0(0.0%)3(8.1%)1 (Ref)Bone Metastasis No2(33.3%)15(40.5%) Yes4(66.7%)22(59.5%)Prior Systemic Chemotherapy (before LM) No2(33.3%)13(35.1%) Yes4(66.7%)24(64.9%)Prior Immunotherapy (before LM) No3(50.0%)31(83.8%) Yes3(50.0%)6(16.2%)Prior Targeted Therapy (before LM) No1(16.7%)31(83.8%) Yes5(83.3%)6(16.2%)Prior Anti-angiogenic Therapy (before LM) No2(33.3%)16(43.2%) Yes4(66.7%)21(56.8%)Prior WBRT (before LM) No5(83.3%)32(86.5%) Yes1(16.7%)5(13.5%)Systemic Chemotherapy (after LM) No5(83.3%)27(73.0%) Yes1(16.7%)10(27.0%)Immunotherapy (after LM) No5(83.3%)29(78.4%) Yes1(16.7%)8(21.6%)Targeted Therapy (after LM) No1(16.7%)18(48.6%) Yes5(83.3%)19(51.4%)Anti-angiogenic Therapy (after LM)0.041 No1(16.7%)19(51.4%)2.1311.033–4.394 Yes5(83.3%)18(48.6%)1 (Ref)WBRT (after LM) No3(50.0%)19(51.4%) Yes3(50.0%)18(48.6%)Note: EGFR: Epidermal Growth Factor Receptor; LM: Leptomeningeal Metastasis; WBRT: Whole Brain Radiotherapy. Data are presented as n (%). Percentages are calculated based on the total number of patients in each row subgroup (PS ≤ 1, n = 6; PS ≥ 2, n = 37) Multivariate analysis of subgroups based on PS score Note: EGFR: Epidermal Growth Factor Receptor; LM: Leptomeningeal Metastasis; WBRT: Whole Brain Radiotherapy. Data are presented as n (%). Percentages are calculated based on the total number of patients in each row subgroup (PS ≤ 1, n = 6; PS ≥ 2, n = 37) Quality of life (QL) Responsiveness Quality of life was assessed by comparing pre- and post-treatment scores across various domains of the Chinese version of the EORTC QLQ-C30 in patients with meningeal metastasis. Paired-sample t-test results (Table 5) revealed a significant post-treatment improvement in quality of life across multiple domains. Among the functional scales, the QL score increased significantly compared with the pre-treatment level (P < 0.001). In contrast, scores for physical functioning (PF), role functioning (RF), emotional functioning (EF), cognitive functioning (CF), and social functioning (SF) all decreased significantly (P < 0.001), indicating meaningful improvement of functional limitations, with EF and SF showing the most pronounced improvements. On the symptom scales, significant improvements were observed in the core symptoms of fatigue (FA), nausea and vomiting (NV), and pain (PA) (P < 0.001). Other symptoms—including dyspnea, insomnia, appetite loss, constipation, diarrhea, and financial difficulties—also showed statistically significant alleviation (P < 0.05), indicating effective symptom control. Table 5Evaluation results of responsiveness for the Chinese version of the EORTC QLQ-C30Domain (Dimension)Pre-treatmentPost-treatmentt-testP-valueMeanSDMeanSDPhysical Functioning (PF)3.1530.7892.8510.8354.304<0.001Role Functioning (RF)3.2440.8122.7910.8544.832<0.001Emotional Functioning (EF)2.8550.6842.1570.7586.660<0.001Cognitive Functioning (CF)2.9300.8902.3260.8515.778<0.001Social Functioning (SF)3.1160.7222.4420.7266.766<0.001Quality of Life (QL)3.3300.9694.4001.094-6.950<0.001Fatigue (FA)3.2790.7662.4880.8059.201<0.001Nausea and Vomiting (NV)2.4650.7591.7090.6667.333<0.001Pain (PA)2.6740.8082.1280.7257.032<0.001Dyspnea (DY)2.0700.7371.7200.7343.3340.002Insomnia (SL)2.7400.9022.1900.8804.989<0.001Appetite Loss (AP)2.6300.7572.0500.8156.089<0.001Constipation (CO)1.4200.5871.1900.4502.8920.006Diarrhea (DI)1.3000.6381.1200.4982.7100.010Financial Difficulties (FI)2.8600.6752.4700.7673.930<0.001 Table 5Evaluation results of responsiveness for the Chinese version of the EORTC QLQ-C30Domain (Dimension)Pre-treatmentPost-treatmentt-testP-valueMeanSDMeanSDPhysical Functioning (PF)3.1530.7892.8510.8354.304<0.001Role Functioning (RF)3.2440.8122.7910.8544.832<0.001Emotional Functioning (EF)2.8550.6842.1570.7586.660<0.001Cognitive Functioning (CF)2.9300.8902.3260.8515.778<0.001Social Functioning (SF)3.1160.7222.4420.7266.766<0.001Quality of Life (QL)3.3300.9694.4001.094-6.950<0.001Fatigue (FA)3.2790.7662.4880.8059.201<0.001Nausea and Vomiting (NV)2.4650.7591.7090.6667.333<0.001Pain (PA)2.6740.8082.1280.7257.032<0.001Dyspnea (DY)2.0700.7371.7200.7343.3340.002Insomnia (SL)2.7400.9022.1900.8804.989<0.001Appetite Loss (AP)2.6300.7572.0500.8156.089<0.001Constipation (CO)1.4200.5871.1900.4502.8920.006Diarrhea (DI)1.3000.6381.1200.4982.7100.010Financial Difficulties (FI)2.8600.6752.4700.7673.930<0.001 Evaluation results of responsiveness for the Chinese version of the EORTC QLQ-C30 Reliability As shown in Table 6, reliability analysis of the scale indicated that the internal consistency, measured by Cronbach’s α, ranged from 0.687 to 0.913 across the main functional and symptom dimensions. The physical functioning (α = 0.913), role functioning (α = 0.842), and pain (α = 0.876) dimensions demonstrated particularly high reliability. Test-retest reliability analysis further showed that the intraclass correlation coefficients (γ) for all dimensions ranged from 0.509 to 0.841. The highest stability was observed in physical functioning (γ = 0.841), role functioning (γ = 0.728), and pain (γ = 0.784). These findings confirm that the QLQ-C30 scale used in this study demonstrates good reliability and stability in this patient population, indicating that the measurements are dependable. Table 6Reliability of the Chinese version of the EORTC QLQ-C30Domain (Dimension)Test-Retest Correlation (γ)Internal Consistency (α)Physical Functioning (PF)0.8410.913Role Functioning (RF)0.7280.842Emotional Functioning (EF)0.5500.707Cognitive Functioning (CF)0.6900.816Social Functioning (SF)0.5920.744Quality of Life (QL)0.5270.687Fatigue (FA)0.7440.852Nausea and Vomiting (NV)0.5570.711Pain (PA)0.7840.876Dyspnea (DY)0.565—Insomnia (SL)0.661—Appetite Loss (AP)0.685—Constipation (CO)0.509—Diarrhea (DI)0.711—Financial Difficulties (FI)0.588—Note: — Not applicable. Cronbach’s α was not calculated for single-item scales (DY, SL, AP, CO, DI, FI) as per the EORTC QLQ-C30 scoring manual Table 6Reliability of the Chinese version of the EORTC QLQ-C30Domain (Dimension)Test-Retest Correlation (γ)Internal Consistency (α)Physical Functioning (PF)0.8410.913Role Functioning (RF)0.7280.842Emotional Functioning (EF)0.5500.707Cognitive Functioning (CF)0.6900.816Social Functioning (SF)0.5920.744Quality of Life (QL)0.5270.687Fatigue (FA)0.7440.852Nausea and Vomiting (NV)0.5570.711Pain (PA)0.7840.876Dyspnea (DY)0.565—Insomnia (SL)0.661—Appetite Loss (AP)0.685—Constipation (CO)0.509—Diarrhea (DI)0.711—Financial Difficulties (FI)0.588—Note: — Not applicable. Cronbach’s α was not calculated for single-item scales (DY, SL, AP, CO, DI, FI) as per the EORTC QLQ-C30 scoring manual Reliability of the Chinese version of the EORTC QLQ-C30 Note: — Not applicable. Cronbach’s α was not calculated for single-item scales (DY, SL, AP, CO, DI, FI) as per the EORTC QLQ-C30 scoring manual Discussion LM is a severe complication of advanced malignant tumors, for which treatment remains primarily palliative. The primary goals of management are to alleviate neurological symptoms, improve quality of life, and prolong survival where possible [3]. This study retrospectively analyzed the efficacy and prognostic factors in patients with NSCLC-LM. The absence of anti-angiogenic therapy following LM diagnosis, the presence of adrenal metastasis, and elevated platelet-to-lymphocyte ratio (PLR) were identified as independent risk factors for poor prognosis. To further account for the potential confounding influence of performance status (PS) on treatment decisions and survival, we performed a subgroup analysis in patients with good performance status (PS ≤ 1). This analysis confirmed that the absence of anti-angiogenic therapy and the presence of adrenal metastasis remained robust and significant predictors of poor prognosis. Moreover, meningeal irritation signs emerged as an additional independent risk factor within this subgroup. Previous retrospective studies on prognostic factors in NSCLC-LM have suggested that elements such as tumor histology, LM subtype, and targeted therapy before LM diagnosis may be associated with prognosis [4– [8–15]; however, none of these factors were identified as independent predictors in our analysis. This study confirmed that the absence of anti-angiogenic therapy following a diagnosis of NSCLC-LM is an independent risk factor for poor prognosis, consistent with several clinical reports [16–21], and highlights the therapeutic significance of this treatment. Wu et al. [22] demonstrated that combining anti-angiogenic therapy with EGFR-TKI can provide survival benefits in certain NSCLC-LM patients, whereas other studies indicate that its combination with immune checkpoint inhibitors may also yield synergistic anti-tumor effects [23]. This may be attributed to the ability of anti-angiogenic agents to modulate the tumor microenvironment, thereby enhancing the efficacy of other treatments. Notably, the prognostic significance of this factor persisted in the subgroup with a PS score ≤ 1, suggesting that patients in good general condition without contraindications may derive substantial survival benefits from active anti-angiogenic therapy. Furthermore, an elevated PLR was identified as an independent predictor of poor prognosis in this study, in concordance with previous findings [24, 25]. As an integrated inflammatory index derived from platelet and lymphocyte counts, an increased PLR indicates not only a hypercoagulable and pro-angiogenic milieu but also a diminished lymphocyte-mediated anti-tumor immune response, collectively contributing to reduced treatment efficacy and poor survival. Our results indicate that baseline PLR remained significantly associated with survival outcomes even after adjusting for treatment strategies. Owing to its accessibility and low cost, PLR may serve as a practical tool for risk stratification in this population. As a key marker of systemic tumor spread, adrenal metastasis generally signifies advanced disease. Our findings establish it as an independent predictor of poor prognosis in NSCLC-LM, presumably due to a markedly elevated tumor burden. Furthermore, a meta-analysis revealed that advanced NSCLC patients with adrenal metastasis treated with PD-1 inhibitors tended to have inferior survival [26], highlighting the caution required when considering prognosis in this subgroup. The PS score is widely acknowledged as an important independent prognostic factor for LM [27, 28]. Patients with lower PS scores typically maintain superior overall health and physical capacity, which facilitates the administration of intensive therapies, including systemic treatment and radiotherapy. Enhanced treatment tolerance frequently results in improved outcomes from intensive, multimodal therapy, ultimately contributing to prolonged survival. To mitigate the potential confounding influence of physical status on treatment decisions and prognostic outcomes, we performed subgroup analyses stratified according to PS scores. Among patients with good PS (PS ≤ 1), the lack of post-LM anti-angiogenic therapy and the presence of adrenal metastasis persisted as significant and independent predictors of poor prognosis. Furthermore, meningeal irritation signs emerged as a new independent risk factor within this subgroup. This aligns with prior reports associating concurrent parenchymal brain metastases and leptomeningeal spread with worse prognosis [29–32]. The presence of meningeal irritation signs may indicate more aggressive tumor biology and a severely compromised blood-brain barrier, which can hinder therapeutic agents from reaching effective intracranial concentrations, ultimately reducing treatment efficacy and survival. These results suggest that, even among patients with good PS, those presenting with meningeal irritation signs represent a high-risk subgroup for whom more intensive and targeted therapeutic strategies should be considered. Using the rigorously validated Chinese version of the EORTC QLQ-C30 questionnaire, we assessed the short-term impact of comprehensive therapy on quality of life (QoL) in NSCLC-LM patients. According to the EORTC QLQ-C30 scoring manual, higher scores on the functional scales indicate better functioning, whereas higher scores on the symptom scales and the QL scale represent a greater severity of symptoms and a worse overall quality of life. Our results highlight the complex trade-offs associated with active treatment in this advanced patient population. We observed a significant improvement in patients’ perceived QL, along with a marked reduction in the burden of key symptoms, including fatigue, nausea/vomiting, and pain. These findings indicate that the comprehensive treatment regimen was highly effective in alleviating the most distressing physical complaints associated with LM [33–36]. However, this symptomatic relief was accompanied by a statistically significant decline in all functional domains, including physical, role, emotional, cognitive, and social functioning. Several factors may account for this apparent dichotomy. First, the toxicities associated with combined modality therapy–including chemotherapy, targeted therapy, and radiotherapy–can themselves impair a patient’s functional capacity, even while alleviating underlying disease symptoms. Second, patients with advanced cancer may recalibrate their internal standards and values in response to their illness and treatment. Thus, despite experiencing objective functional decline, patients may report an improved global QL, owing to substantial relief from debilitating symptoms such as pain and nausea, as well as the psychological benefit of actively engaging in treatment [37]. The marked alleviation of fatigue, one of the most pervasive and distressing symptoms in cancer, likely contributed substantially to the improvement in global health status, as it directly impacts a patient’s sense of well-being. These findings underscore the multifaceted nature of QL in terminal illness. They suggest that for NSCLC-LM patients, comprehensive treatment can successfully palliate specific symptoms and improve the overall sense of health, which patients value highly. However, clinicians must be aware of and proactively manage the potential functional trade-offs of such treatment, integrating robust supportive care measures to mitigate functional decline and help patients adapt to their changing capacities. This study is the first to systematically apply the “binary pattern” classification based on parenchymal involvement status in NSCLC-LM patients, assessing the clinical value of anti-angiogenic therapy, the prognostic significance of PLR, and the short-term impact of comprehensive treatment on QL. However, several limitations should be acknowledged. As a single-center retrospective study with a limited sample size, statistical power may be reduced, and the reliability of subgroup analyses may be constrained. Moreover, detailed data on drug dosage and treatment cycles were not comprehensively collected, which may have introduced residual confounding factors. Although radiological assessments were conducted by experienced radiologists, the absence of a blinded review may have introduced subjective bias. Future prospective, multi-center, large-sample studies are needed to validate these findings. Additional research should investigate the pharmacokinetic and pharmacodynamic properties of anti-angiogenic drugs in central nervous system metastases, as well as the interaction between systemic inflammatory markers and the tumor microenvironment, to further support precision treatment strategies for NSCLC-LM. In summary, this study found that while NSCLC-LM patients with and without PM had the same median OS, those without PM showed a non-significant trend toward better long-term survival, as reflected in higher 1- and 2-year survival rates. This observation requires validation in larger studies. Key independent risk factors for poor prognosis included the absence of anti-angiogenic therapy after LM diagnosis, adrenal metastasis, and a high PLR. These results were consistent in patients with good PS, for whom meningeal irritation signs also independently predicted worse outcomes. Additionally, comprehensive treatment not only alleviated symptom burden but also enhanced QL across multiple domains, contributing to longer high-quality survival.