Prevalence and risk factors of nonyield brain biopsy: a 21-year experience with robot-assisted stereotactic biopsies

Magnetic resonance imaging-based, robot-assisted stereotactic brain biopsy is increasingly used worldwide. However, large series reporting nonyield biopsy rates of robot-assisted biopsies are lacking in the literature. The aim of this study was to report a 21-year-long experience on MRI-based, robot-assisted stereotactic biopsy for brain lesions.

The records from a single-center, retrospective, and consecutive collection of all adult patients undergoing MRI-based, robot-assisted stereotactic biopsy for a brain lesion in a tertiary neurosurgical center from December 2002 to January 2024 were reviewed.

A total of 911 patients (377 females, mean age at surgery 61.1 ± 16.7 years) were included. Of these patients, 15 (1.6%) had a nonyield biopsy. The nonyield biopsy rate remained stable over the 21-year-long study period (p = 0.224). The nonyield biopsy rate was significantly higher for neurological diseases (4/19, 21.5%) than infectious diseases (1/20, 5.0%) and tumors (10/872, 1.1%) (p < 0.001). There were significantly more nonyield biopsies for deep-seated lesions (8/261, 3.1%) than for superficial lesions (6/612, 1.0%) (p = 0.025) and non-contrast-enhanced lesions (6/99, 6.1%) than in contrast-enhanced lesions (9/797, 1.1%) (p < 0.001). Patients in the nonyield biopsy group had significantly smaller lesions on both contrast-enhanced 3D T1-weighted sequences (1.9 ± 2.2 vs 27.1 ± 29.5 cm3, p = 0.046) and FLAIR sequences (14.0 ± 17.0 vs 80.7 ± 73.3 cm3, p < 0.001). Preoperative corticosteroid administration (407/911, 44.7%; with the bias that corticosteroids were avoided in patients with suspected lymphoma), number of biopsy samples (mean 4.9 ± 2.4), neurosurgeon experience, and WHO classification versions were not associated with a higher risk of nonyield biopsy (p = 0.274, p = 0.053, p = 0.968, and p = 0.366, respectively).

MRI-based, robot-assisted stereotactic biopsy led to a low rate of nonyield biopsy. Neurological disease, non-contrast-enhanced lesions, and deep-seated lesions were more at risk of nonyield biopsy.