Clinical Significance of Updated MRI Simulation Versus Immediate Postoperative MRI for Radiation Treatment Planning in Patients With Glioblastoma: Changes in FLAIR Signal International Journal of Radiation Oncology, Biology, Physics POSTER VIEWING ABSTRACT| VOLUME 96, ISSUE 2, SUPPLEMENT , E103, OCTOBER 01, 2016
Date
10/01/2016Abstract
Purpose/Objective(s)
Standard radiation therapy (RT) planning for the treatment of glioblastoma (GBM) includes CT simulation fused with post-operative MRI. Immediate post-operative MRI (ipMRI) taken within 72 hours of resection is frequently used for this purpose. Because there is a delay between surgery and RT start date, the ipMRI may not account for acute changes that occur before the RT start date. Since the implementation of a dedicated MRI simulator (MR Sim) in our Department in 2008, all pts undergo updated MR Sim at the time of CT simulation for treatment planning when feasible. To determine the clinical significance of this updated imaging, we investigated the changes in FLAIR signal between ipMRI and MR Sim.
Materials/Methods
From September 2008 to November 2015, 454 pts underwent MR Sim, of which 131 had GBM. We excluded pts who were lost early to follow-up, did not complete treatment, or who did not have ipMRI available, yielding 88 pts. A neuro-radiologist measured the FLAIR signal changes between the ipMRI and MR Sim using the MacDonald 2D approach. The image slice with the largest amount of disease was identified and the longest orthogonal diameters were measured in 2 dimensions. Clinical significance for cross sectional changes was set at > 400 mm2 as the calculated 2D translation of the standard 2 cm volume expansion from FLAIR during treatment planning. Using chi-square and paired T test study, statistical significance for clinical variables of interest was computed to be alpha < 0.05.
Results
Of the 88 pts, 51 pts had developed significantly smaller FLAIR, 8 had significantly larger FLAIR, and 29 had no significant change. Days between ipMRI and MRI Sim was found to be a significant predictor of change after 14 days when comparing pts with smaller FLAIR to all other pts and those with no change (P = 0.018 and 0.019 respectively). Additionally, 14 days between scans was found to be significant when comparing pts with larger or smaller FLAIR with those with no change (P = 0.003). No other clinical variable including age, sex, KPS, resection extent, and use of post-operative steroids was found to be a significant predictor of FLAIR change. Median survival for all pts was 15.1 months with 1-yr OS 68% and 2-yr OS 27%. FLAIR change was not predictive of OS (P = 0.68).
Conclusion
At 14 days after post-operative MRI, a majority of pts show significant changes in FLAIR, suggesting that updated MRI should be considered in pts who plan to start RT > 14 days from surgery. Most pts have smaller FLAIR which may allow for reduced treatment volumes leading to potentially less side effects. Future directions include analyzing 3D volumetric changes in FLAIR and T1 + contrast signal between these scans as well as patterns of failure.
Author List
A. Dayal M.L. Siker S. Firat J.A. Bovi C.J. Schultz W.M. Mueller J.M. Connelly S.D. RandAuthor
Wade M. Mueller MD Professor in the Neurosurgery department at Medical College of WisconsinView Online
