Prognostic features of ewing's sarcoma on plain radiograph and computed tomography scan after initial treatment

From the *Department of Radiology, Jewish Hospital, and the tMusculoskeleta1 Section, Mallinckrodt Institute of Radiology, St. Louis, Missouri; the $Department of Radiation Oncology, Stanford University Medical Center, Stanford, California; the SPediatric Oncology Group Statistical Office, Gainesville, Florida; the 11 Department of Radiation Oncology, Roger Williams General Hospital, Providence, Rhode Island; and the #Pediatric Oncology Group Operations Office, St. Louis, Missouri. Supported in part by grants from the National Cancer Institute and the National Institutes of Health (CA-30969, CA-29139, CA05587, and CA-33603). Principal investigators and their participating parent institutions are as follows (with grant numbers in parentheses): J. Akabutu, Alberta Pediatric Oncology Consortium, Edmonton, Alberta, Canada; Y. Ravindranath, Children’s Hospital of Michigan, Detroit, Michigan (CA-29691); R. Ducos, Children’s Hospital of New Orleans, New Orleans, Louisiana (C-35272); J. Falletta, Duke University Medical Center, Durham, North Carolina (CA-15525); A. Ragab, Emory University Medical School, Atlanta, Georgia (CA-20549); J. Talbert, Florida Community Clinical Oncology, Gainesville, Florida (CA-35 157); B. Leventhal, Johns Hopkins School of Medicine, Baltimore, Maryland (CA-28476); V. M. Whitehead, McGill University, Montreal, Canada (CA-33587); H. Maurer, Medical College of Virginia, Richmond, Virginia (CA-28530); J. Laver, Medical University of South Carolina, Charleston, South Carolina; B. Camitta, Midwest Children’s Cancer Center, Milwaukee, Wisconsin (CA-32053); E. Forman, New England Pediatric Consortium, Providence, Rhode Island (CA-29293); R. Nitschke, Oklahoma University, Oklahoma City, Oklahoma (CA-I 1233); D. Maybee, Uniformed Services Oncology Consortium, Washington, DC; E. Douglass, St. Christopher’s Hospital, Philadelphia, Pennsylvania; M. Link, Stanford University Medical Center, Palo Alto, California (CA-33603); H. Wagner, Swiss Pediatric Oncology Group, Bern, Switzerland; R. Dubowy, State University of New York, Syracuse, New York (CA-41721); R. Castleberry, University of Alabama, Birmingham, Alabama (CA-25408); D. Berry, University of Arkansas, Little Rock, Arkansas (CA-41188); F. Kung, University of California at San Diego, San Diego, California (CA-28439); T. Vats, University of Kansas, Kansas City, Kansas (CA-28841); J. Pullen, University of Mississippi Medical Center, Jackson, Mississippi (CA015989); S. Toledano, University of Miami, Miami, Florida (CAphy (CT) appearance of Ewing sarcoma for indicators of decreased survival or future disease progression. Methods. The authors evaluated CT scans and plain radiographs of the primary tumor site from 105 patients with Ewing sarcoma at diagnosis (prebiopsy), after induction chemotherapy (13 weeks), and after radiation therapy (20 weeks). Results. Data suggest an association between postinduction CT findings of medullary involvement, cortical destruction, lysis, permeation, and unhealed pathologic fracture and decreased survival. On the postradiation scans, only medullary involvement was associated with worsened survival. No plain radiographic features were significant at any time. Absolute greatest tumor dimension was not significantly related to survival or tumor progression. The Cox model suggested that fractional change in greatest tumor dimension on CT at the time points studied relative to the prebiopsy CT was correlated to survival. Log-rank testing did not corroborate this finding. All significant associations appeared to result from adverse outcomes in small subgroups. Conclusions. Our data suggest that CT obtained immediately after induction chemotherapy and radiation may have some limited use in predicting the long-term prognosis of patients with Ewing sarcoma. Cancer 1993; 72: 2503-10.