Joshua T Lee BS, Presenter: Nothing to Disclose

Rakhee Sameer Gawande MBBS, Abstract Co-Author: Nothing to Disclose

Steven G DuBois MD, Abstract Co-Author: Nothing to Disclose

Jeremy Sharib BS, Abstract Co-Author: Nothing to Disclose

Neyssa Marina MD, Abstract Co-Author: Nothing to Disclose

Heike E. Daldrup-Link MD, Abstract Co-Author: Nothing to Disclose

The optimal method for assessing treatment response in patients with primary bone tumors is not clear. We sought to compare radiographic response of Ewing sarcoma family of tumors (ESFT) to determine whether response classification differs between the RECIST 1.1 (Response Evaluation Criteria in Solid Tumors), WHO (World Health Organization) and COG (Children’s Oncology Group) response criteria.

We retrospectively analyzed MR imaging scans of fifty-six patients with EFST, who were treated at Stanford and UCSF Medical Centers. Tumor size was assessed on T2-weighted sequences and postcontrast T1-weighted sequences before, during and after therapy. Tumor measurements were obtained according to the RECIST 1.1 (longest single diameter), WHO (byproduct of the longest perpendicular diameters) and COG criteria [tumor volume (V) calculated using the maximum diameters from sagittal (S), coronal (C) and axial (A) planes: V = S x C x A x F (where F = π/6 for ellipsoid tumors; F = π/4 for cylindrical tumors)]. All three guidelines share the same four response categories: progressive disease (PD) (>20% increase in RECIST/COG; > 25% in WHO), stable disease (SD) (neither PR nor PD), partial response (PR) (< 30% decrease in RECIST/COG; <50% decrease in WHO) and complete response (CR) (100% decrease). Numerical values were assigned to each response category: 1 = PD, 2 = SD, 3 = PR and 4 = CR. Concordance between the three response classification systems was assessed using Cohen’s kappa (κ) coefficient and percentage of disagreement per response category.

The κ statistic for concordance in RECIST/WHO, RECIST/COG and WHO/COG were 0.443, 0.215 and 0.349 respectively. Disagreement rates for RECIST/WHO, RECIST/COG and WHO/COG were 30.36, 48.21, and 32.14% respectively. Twenty-seven patients coded as SD by RECIST were reclassified as PR by COG. Similarly, eighteen patients originally coded as SD by WHO were re-categorized as PR by COG.

This study demonstrates a fair level of agreement, by the κ statistic, between the RECIST 1.1, WHO and COG response criteria in ESFT.

Given the degree of discordance between WHO, RECIST and COG response criteria in ESFT, evaluation of the prognostic impact may guide selection of the optimal system for future use in this disease.

Lee, J, Gawande, R, DuBois, S, Sharib, J, Marina, N, Daldrup-Link, H, Comparison of RECIST 1.1, WHO, and COG Response Criteria in Patients with Ewing Sarcoma Family of Tumors.  Radiological Society of North America 2011 Scientific Assembly and Annual Meeting, November 26 - December 2, 2011 ,Chicago IL. http://archive.rsna.org/2011/11009160.html