Longitudinal trends in Gamma Knife radiosurgery utilization across multiple technological generations: A Single-Institution Observational Study

Abstract

Background

Gamma Knife radiosurgery (GKRS) has been a foundational modality for the treatment of intracranial diseases for several decades, with successive platform generations introducing substantial technological advancements. Improvements in source configuration, patient positioning systems, imaging guidance, and immobilization techniques have expanded clinical capabilities. Despite these ongoing innovations, few long-term institutional analyses have evaluated whether GKRS utilization has remained stable, increased, or shifted with changes in disease patterns and retreatment needs over time.

Purpose

Gamma Knife radiosurgery (GKRS) has undergone multiple technological evolutions over the past two decades. This study provides a retrospective analysis of longitudinal utilization at a single institution, examining trends in case volume, disease indications, retreatment frequency, and platform era.

Methods

A retrospective analysis of Gamma Knife treatments delivered at one institution between 1998 and 2025 was performed. Annual counts of first, repeat, and total GKRS treatments were analyzed using negative binomial regression with a log link and Newey–West standard errors; platform-adjusted models accounted for the sequential relationship between calendar time and platform era. Segmented negative binomial regression of monthly counts evaluated slope changes at platform transitions.

Results

Across the study period, 6,201 unique patients received 7,267 delivered treatment courses. Median annual treatment volume was 276 procedures (range, 71–362); median age at first treatment was 60 years (IQR, 50–69), and brain metastasis was the primary indication in 2,991 patients (48.2%). Total volume increased in unadjusted modeling (count ratio [CR], 1.03; p < 0.001), and repeat treatments showed the strongest platform-adjusted growth (CR 1.14; p < 0.001). Metastatic disease was the dominant adjusted disease-specific trend (CR 1.08; p = 0.01). Repeat metastatic treatments increased more than overall repeat treatments (CR 1.15; p < 0.001). Segmented regression showed significant growth during the baseline Model B era (annualized CR 1.16; p < 0.001) followed by significant attenuation at the Model C transition (annualized CR 0.90; p = 0.01). The primary segmented model showed no significant incremental slope change at Perfexion or Icon; sensitivity analyses indicated that the estimated Icon-era association was model-dependent.

Conclusions

Long-term Gamma Knife utilization demonstrated sustained institutional growth, with the clearest adjusted temporal trends in metastatic disease and repeat treatment. Because calendar time and sequential platform adoption were collinear, these observational findings describe evolving utilization and practice patterns but do not establish that platform changes caused the observed trends.