Predictive factors for clinical outcome of palliative ventriculoperitoneal shunt in the treatment of leptomeningeal metastasis.

Background: Leptomeningeal metastasis (LM) is associated with poor prognosis and affects the quality of life (QOL) of end-stage cancer patients, particularly when accompanied by hydrocephalus. Ventriculoperitoneal shunt (VPS) is recognized as an effective therapy for patients with this condition, but previous studies have focused on clinical effects. This study aimed to explore the predictors of clinical outcome in patients with LM who underwent VPS. Methods: Thirty-seven patients who underwent VPS after being diagnosed with LM at Huashan Hospital, Fudan University, from June 2018 to June 2024 were enrolled. Clinical outcomes were evaluated using the Karnofsky performance status (KPS) at baseline and one week postoperatively. A good outcome was defined as a ≥30% improvement in KPS, while a poor outcome was <30%. Patients were accordingly grouped into good and poor outcome categories. Potential predictive factors, including the type of primary malignancy, presence of hydrocephalus, location and extent of LM, preoperative lumbar puncture pressure (measured in mmH2O), and shunt equipment manufacturer, were initially analyzed using univariate methods. Factors that reached statistical significance were further examined with multivariate analyses to identify predictors of clinical outcomes. Results: Two factors showed significant differences between the good- and poor-outcome groups: the hydrocephalus (P=0.008) and lumbar puncture pressure (P=0.003). Multivariate analyses revealed that lumbar puncture pressure (P=0.006) was the only significant predictor of clinical outcome. Conclusions: The lumbar puncture pressure was the only independent predictive factor of the clinical outcomes of VPS for the treatment of patients with LM.