TRAM MRI shows 0.89 AUC for differentiating treatment response from radionecrosis in CNS tumors

BACKGROUND: Differentiating treatment response (TR) from radionecrosis (RN) after radiotherapy in patients with CNS tumors remains a diagnostic challenge. Treatment response assessment maps (TRAMs), derived from delayed contrast-enhanced MRI, have been proposed as a promising tool to improve diagnostic accuracy in this setting. PURPOSE: The purpose of this study is to systematically evaluate the diagnostic performance of TRAM MRI in differentiating TR from RN through a diagnostic meta-analysis. DATA SOURCES: A comprehensive literature search was conducted in PubMed, EMBASE, and Lilacs, supplemented by citation tracking and gray literature, up to September 19, 2024. STUDY SELECTION: Studies were included if they evaluated TRAM MRI in differentiating TR from RN, reported sufficient data to construct 2 × 2 contingency tables, and used histology or clinical-radiologic follow-up as a reference standard. DATA ANALYSIS: A total of 7 studies involving 286 patients and 340 lesions were included. A bivariate random-effects model was used to calculate pooled sensitivity and specificity. Subgroup and meta-regression analyses were performed to explore potential sources of heterogeneity, and the Quality Assessment of Diagnostic Accuracy Studies 2 tool was applied for risk-of-bias assessment. DATA SYNTHESIS: TRAM demonstrated high pooled sensitivity (0.88; 95% CI, 0.72-0.95) and moderate specificity (0.74; 95% CI, 0.47-0.90), with an area under the summary receiver operating characteristic curve of 0.89. Substantial heterogeneity was observed across studies. Subgroup analyses identified the software used as a major source of heterogeneity. No meaningful impact was observed for tumor type, study design, reference standard, or time from radiotherapy. LIMITATIONS: Main limitations included the small number of studies, methodologic heterogeneity, lack of standardized interpretation criteria for TRAM, and limited availability of individual patient data. CONCLUSIONS: TRAM MRI demonstrates high diagnostic accuracy for differentiating TR from RN in patients with CNS tumors treated with radiotherapy. Despite heterogeneity and methodologic limitations, the results support the clinical potential of TRAM as a complementary tool in neuro-oncologic imaging, warranting further validation in prospective and standardized studies.