Prognostic Importance of Cumulative Intracranial Tumor Volume in Patients with Gastrointestinal Brain Metastasis Treated with Stereotactic Radiosurgery

BACKGROUND: The disease-specific Graded Prognostic Assessment (ds-GPA) for patients with gastrointestinal (GI) tract cancer brain metastases (BM) suggests Karnofsky Performance Status (KPS) as the only pertinent prognostic factor. We evaluated the prognostic importance of cumulative intracranial tumor volume (CITV).

METHODS: KPS, CITV, and overall survival were collected from consecutive patients with stereotactic radiosurgery-treated GI BM. Patients were grouped into 2 independent cohorts for development and validation of the model (termed "exploratory" and "validation" cohorts). Analyses were performed using logistic regression, Cox proportional hazards models, Net Reclassification Index (NRI >0), integrated discrimination improvement (IDI >0), and Akaike information criterion.

RESULTS: In univariable logistic regression models, both CITV and KPS were independently associated with patient survival. The association between CITV and overall survival remained robust after controlling for KPS (P < 0.001) in a multivariable Cox proportional hazards model. Based on NRI analysis of the exploratory cohort, we found that a CITV cutoff of 12 cm 3 best augments the prognostic accuracy of GI-ds-GPA. In this analysis, incorporation of CITV (as < or ≥12 cm 3) improved prognostication of the GI-specific GPA model by NRI >0 of 0.397 (95% confidence interval [CI], 0.165-0.630; P < 0.001) and IDI of 0.019 (95% CI, 0.004-0.033; P = 0.013). We confirmed the prognostic usefulness of the CITV-incorporated GI-ds-GPA in an independent validation cohort, in which CITV incorporation improved prognostic usefulness with an NRI >0 of 0.478 (95% CI, 0.257-0.699; P < 0.001) and IDI of 0.028 (95% CI, 0.014-0.043; P < 0.001).

CONCLUSIONS: CITV is an important prognostic variable in patients with stereotactic radiosurgery-treated GI BM and augments the prognostic accuracy of the GI-ds-GPA index.