Abstract
Background: This exploratory study assessed whether qualitative dynamic contrast-enhanced MRI (DCE-MRI) time–intensity curves (TICs) differentiate non-progressive treatment-related changes (NPTRC) from tumor recurrence after stereotactic radiotherapy for brain metastases. Methods: Thirty-two patients with 51 treated lesions underwent DCE-MRI. Five neuroradiologists inde-pendently delineated a volume of interest and classified the resulting TIC as Type A–D. Longitudinal clinical and MRI follow-up provided the reference outcome, which was recorded before the TIC framework was finalized. An A/B versus C/D grouping was evaluated as a secondary exploratory analysis. Results: Thirty-five lesions were classified as NPTRC and 16 as tumor recurrence. All 17 Type A lesions were NPTRC, and all 13 Type C/D lesions showed recurrence. Type B occurred in both groups (18 NPTRC; 3 recurrences). The A/B versus C/D grouping yielded 81.3% sensitivity and 100% specificity. The ordinal TIC analysis yielded an area under the curve of 0.952. Interobserver agree-ment was high for the four TIC categories (Fleiss’ κ = 0.889). Conclusions: Type A was associated with NPTRC, Types C/D with recurrence, and Type B represented an indeterminate pattern. These descriptive findings require independent validation before clinical use.