A Rapid Late Enhancement MRI Protocol Improves Differentiation between Brain Tumor Recurrence and Treatment-Related Contrast Enhancement of Brain Parenchyma

SIMPLE SUMMARY: Differentiation between recurrence of malignant glioma and treatment-related changes (such as radiation necrosis) after radiochemotherapy using MRI can be challenging even for experienced neuroradiologists, as both entities may have a similar appearance regarding contrast enhancement and T2-signal. Using conventional MRI-sequences, early etiological assignment of a lesion can thus be difficult, although early and correct differentiation is mandatory in order to optimize individualized patient treatment. Late enhancement MRI up to 75 min after contrast administration has been described to improve differentiation between recurrent tumor tissue and treatment associated changes, but may complicate clinical workflow. We found that a more rapid late enhancement protocol still improves the specificity of follow-up MR imaging in patients with high-grade glioma, while reducing magnet time. ABSTRACT: Purpose: Differentiation between tumor recurrence and treatment-related contrast enhancement in MRI can be difficult. Late enhancement MRI up to 75 min after contrast agent application has been shown to improve differentiation between tumor recurrence and treatment-related changes. We investigated the diagnostic performance of late enhancement using a rapid MRI protocol optimized for clinical workflow. Methods: Twenty-three patients with 28 lesions suspected for glioma recurrence underwent MRI including T1-MPRAGE-series acquired 2 and 20 min after contrast agent administration. Early contrast series were subtracted from late contrast series using motion correction. Contrast enhancing lesions were retrospectively and independently evaluated by two readers blinded to the patients’ later clinical course and histology with or without the use of late enhancement series. Sensitivity, specificity, NPV, and PPV were calculated for both readers by comparing results of MRI with histological samples. Results: Using standard MR sequences, sensitivity, specificity, PPV, and NPV were 0.84, 0, 0.875, and 0 (reader 1) and 0.92, 0, 0.885, and 0 (reader 2), respectively. Early late enhancement increased sensitivity, specificity, PPV, and NPV to 1 for each value and for both readers. Inter-reader reliability increased from 0.632 (standard MRI sequences) to 1.0 (with early late enhancement). Conclusion: The described rapid late enhancement MRI protocol improves MRI-based discrimination between tumor tissue and treatment-related changes of the brain parenchyma.