Distinguishing Brain Tumor Recurrence from Radiation Necrosis: Diagnostic Limitations of Multimodal MRI and the Role of Surgical Resection

Purpose Differentiating tumor recurrence from radiation necrosis (RN) after radiotherapy in brain tumors remains a diagnostic challenge. Multimodal MRI (mMRI), combining diffusion-weighted imaging (DWI), perfusion-weighted imaging, and MR spectroscopy (MRS), is commonly used to improve diagnostic accuracy, though its reliability is still debated. This study evaluates the diagnostic performance of mMRI and the role of surgical resection in establishing a definitive diagnosis. Methods We retrospectively and prospectively included patients with brain tumors who developed new or enlarging contrast-enhancing lesions on follow-up MRI after stereotactic radiosurgery (SRS) or whole-brain radiotherapy (WBRT). All patients underwent mMRI (DWI, perfusion, and MRS) and were assessed by a multidisciplinary team. Surgical resection was performed based on clinical and radiological findings, and histopathology provided the definitive diagnosis. Diagnostic performance metrics and ROC analysis were calculated. Results Fifty-four patients were included. mMRI suggested recurrence in 40 cases (74%), RN in 12 (22%), and was inconclusive in 2 (4%). Histology confirmed pure tumor recurrence in 27 (50%), mixed recurrence and RN in 9 (17%), and pure RN in 6 (11%). Among cases with mMRI-suggested RN, only 1 (8%) was confirmed as pure RN. Both inconclusive cases had FDG-PET-confirmed recurrence with mixed pathology. mMRI showed a 15% false-positive rate for recurrence and a 92% false-negative rate for RN. The area under the ROC curve was 0.45. Conclusion mMRI has significant limitations in distinguishing recurrence from RN. Surgical resection remains the most reliable diagnostic method and should be integrated into a multidisciplinary approach for managing post-radiotherapy lesions.