NI-21CONTRAST ENHANCED T1 SUBTRACTION MAPS IMPROVE DETECTION OF RESIDUAL TUMOR FOLLOWING RESECTION REDUCES MEASUREMENT VARIABILITY AND PREDICTS SURVIVAL AFTER ANTI-ANGIOGENIC THERAPY IN GLIOBLASTOMA

摘要

BACKGROUND: Estimation of contrast enhancing tumor burden on contrast-enhanced post-contrast T1-weighted images (CE-MRI) has significant limitations due to T1 shortening from blood products or gliosis and T1 lengthening from changes in extracellular fluid concentration. As a consequence, estimation of residual tumor burden after surgical resection is often inaccurate, measurements of tumor volume vary significantly across readers, and tumor burden after anti-angiogenic therapies is often undetected. We hypothesize that contrast-enhanced T1 subtraction maps can be used to improve detection of residual tumor following surgery or anti-angiogenic therapy, reduce measurement variability, and predict response and survival. METHODS: The extent of resection was evaluated in 80 GBM patients using traditional CE-MRI and T1 subtraction after surgical resection by radiologists, imaging scientists, and neuro-oncologists. To assess measurement variability, enhancing tumor volumes were evaluated in 20 randomly-selected GBM patients before and after anti-angiogenic therapy by 3 imaging scientists in 3 multicenter trials. Lastly, response evaluation was performed using CE-MRI and T1 subtraction in a single center cohort : The extent of resection was changed from GTR to STR for almost all patients and all reviewers when comparing CE-MRI to T1 subtraction maps. Additionally, the confidence in this determination was significantly improved (t-test, P < 0.01). The coefficient of variability (COV) was significantly reduced when using T1 subtraction to define lesion volumes in all settings (paired t-test, P < 0.01), with the largest improvement after anti-angiogenic therapy. Volumetric response on T1 subtraction maps after bevacizumab predicted PFS and OS in both single-center and multicenter trials (Cox regression, P < 0.05). CONCLUSION: T1 subtraction maps improve measurement accuracy and confidence of contrast enhancing tumor burden and should be used for determining extent of resection and response to therapy.