Anand K. H. Singh MD, Presenter: Nothing to Disclose

Edwin Choy MD, Abstract Co-Author: Nothing to Disclose

Yen-Lin Evelyn Chen MD, Abstract Co-Author: Nothing to Disclose

Gordon J. Harris PhD, Abstract Co-Author: Medical Advisory Board, Fovia, Inc

Wenli Cai PhD, Abstract Co-Author: Nothing to Disclose

David C. Harmon MD, Abstract Co-Author: Nothing to Disclose

To assess the feasibility and accuracy of MRI-based non-contrast enhancing tumor fraction quantification in treated soft tissue sarcomas by comparing with pathologically derived-necrosis fraction estimates based on partial tissue staining method.

In this on-going retrospective study, post-gadolinium T1 weighted fat saturated (FS) sequence from MRI data sets of 39 patients treated for pathologically proven soft tissue sarcomas with chemotherapy (n=9/39), radiation (n=7/39) and both (n=23/39) were retrieved from the database. MRI was obtained within 1-15 days prior to surgical excision. Semi-automated mapping for estimation of necrotic and total tumor volume was performed using thresholding tool and blow-up tool on a 3D workstation respectively on the same post contrast T1-weighted FS sequence. Post excision partial-tissue stained pathology percentage estimate for non-viable tissue was taken as standard of reference for percentage of non-contrast enhancing tumor volumes obtained by MRI quantification and compared using by correlation and Bland Altman (BA) analysis. The time for performing segmentations was recorded.

In majority of the cases (37/39 cases), the necrotic volume percentage derived from MRI volumetry was within 1.96 standard deviation (SD) [-82.0 to 115.7] of the BA plot indicating similarity to pathological necrosis with an excellent correlation (r = 0.8852; p<0.0001; 95% confidence interval= 0.7905 to 0.9386). There were 2 outliers (2 cases) in the BA plot where the mean difference between necrotic percentage derived from MRI volumetry and pathology was 14.38%. Average time taken for performing segmentation and percentage volume generations of non enhancing tumor portions on MRI was 13.517 minutes.

With the use of existing semi-automated 3D segmentation techniques, the non-contrast enhancing tumor fraction of treated sarcoma on MRI correlates well with the partial tissue stained pathological scores estimated after treated tumor excision. These findings necessitate further investigation for such correlation on prospective basis with standardization of MRI protocol and pathology reporting methods.

Considering the present limitations posed by imaging response assessment criteria, this study provides a reliable objective biomarker measure for response assessment in sarcomas for future validation.

Singh, A, Choy, E, Chen, Y, Harris, G, Cai, W, Harmon, D, MR Imaging-based Necrosis Quantification in Treated Soft Tissue Sarcoma with Existing Quantification Methods: Comparison with Partial Tissue-stained Pathological Scores.  Radiological Society of North America 2010 Scientific Assembly and Annual Meeting, November 28 - December 3, 2010 ,Chicago IL. http://archive.rsna.org/2010/9011961.html