Postoperative course of cerebrospinal fluid diversion in the setting of leptomeningeal disease: a systematic review, meta-analysis, and meta-regression with an illustrative case

Background Management of hydrocephalus symptoms in the setting of leptomeningeal disease (LMD) includes cerebrospinal fluid (CSF) diversion, which can in the form of ventriculoperitoneal shunting (VPS) and lumboperitoneal shunting (LPS). However, the quantifiable postoperative course following this intervention is poorly defined. Correspondingly the aim of our study was to quantitatively define and analyze the pooled metadata regarding this topic. Methods Multiple electronic databases from inception to March 2023 were searched following PRISMA guidelines. Respective cohort-level outcomes were then abstracted and pooled by means of meta-analyses and analyzed by means meta-regression, both utilizing random-effects modeling. Post-hoc bias evaluation was then performed for all outcomes. Results A total of 12 studies were identified for inclusion, describing 503 LMD patients managed by CSF diversion – 442 (88%) by VPS and 61 (12%) by LPS. Median male percentage and age at diversion were 32% and 58 years respectively, with lung and breast cancer the most common primary diagnoses. Meta-analysis demonstrated pooled incidence of symptom resolution in 79% (95% CI 68–88%) of patients after index shunt surgery, and shunt revision required in 10% (95% CI 6–15%) of cases. Pooled overall survival from index shunt surgery was 3.8 mo (95% CI 2.9–4.6 mo) across all studies. Meta-regression demonstrated that studies published later trended towards significantly shorter overall survival from index shunt surgery (co-efficient=-0.38, P = 0.023), whereas the proportion of VPS to LPS in each study did not impact survival (P = 0.89). When accounting for these biases, overall survival from index shunt surgery was re-estimated to be shorter 3.1 mo (95% CI 1.7–4.4 mo). We present an illustrative case demonstrating the course of symptom improvement, shunt revision and an overall survival of 2 weeks from index CSF diversion. Conclusion Although CSF diversion in the setting of LMD can improve hydrocephalus symptoms in the majority of patients, there is a non-negligible proportion that will require shunt revision. Postoperatively, the prognosis of LMD remains poor irrespective of shunt type, and despite possible biases within the current literature, the expected median overall survival after index surgery is a matter of months. These findings support CSF diversion as an effective palliative procedure when considering symptoms and quality of life. Further research is required to understand how postoperative expectations can be managed to respect the best wishes of patients, their family, and the treating clinical team.