Safety and diagnostic yield of robotic-assisted stereotactic biopsy for pediatric brainstem lesions: a systematic review and single-arm meta-analysis.

Abstract

Pediatric brainstem lesions pose significant diagnostic challenges due to their deep-seated location and critical neuroanatomical surroundings. Robotic-assisted stereotactic biopsy has emerged as a promising minimally invasive strategy, yet its safety and efficacy in this vulnerable population remain incompletely defined. We conducted a systematic review and single-arm meta-analysis in accordance with PRISMA guidelines (PROSPERO registration: CRD420251135743). PubMed, Embase, Web of Science, and Cochrane Library were searched from inception to May 2025. Eligible studies included pediatric patients (≤ 18 years) with radiologically confirmed brainstem lesions who underwent robot-assisted stereotactic biopsy with reported histopathological and postoperative outcomes. Data were pooled using either a fixed-effects or random-effects model (DerSimonian-Laird) with Freeman-Tukey double arcsine transformation, depending on the degree of heterogeneity. From 72 records, 15 studies comprising 215 pediatric patients were included. Robotic systems evaluated included Neuromate, ROSA, Sino-Precision, and Autoguide, using both frameless and frame-based approaches. The pooled rate of technically efficacious biopsy was 99% (95% CI 98-100%), and the conclusive diagnostic yield was 98% (95% CI 93-98%). Molecular profiling was feasible, with an H3 K27-altered positivity rate of 77% (95% CI 54-94%). The pooled rate of temporary complications was 6% (95% CI 2-11%), mainly transient cranial nerve palsies or mild weakness, while permanent morbidity occurred in only 2 patients (1 hemiparesis and 1 cranial nerve XI/XII palsy), both following frameless procedures. No procedure-related mortality was reported. Subgroup analyses confirmed comparable diagnostic yield between frameless and frame-based systems, though permanent morbidity was observed only in the frameless cohort. The mean hospital stay was 2.08 days (95% CI 1.91-2.25 days), reflecting rapid postoperative recovery. Robotic-assisted stereotactic biopsy of pediatric brainstem lesions provides near-universal diagnostic yield with a favorable safety profile, supporting its role as a reliable minimally invasive approach in this high-risk setting. Both frameless and frame-based systems are effective, although rare permanent deficits have been reported with frameless techniques. Future multicenter randomized trials are warranted to directly compare robotic and conventional stereotactic methods, standardize perioperative outcome reporting, and further evaluate the contribution of robotic biopsy to molecular diagnostics and precision medicine in pediatric neuro-oncology.