OBJECTIVE: We compared how well T1-weighted and short inversion time inversion recovery (STIR) MR images obtained before and after preoperative chemotherapy reveal the extent of longitudinal intraosseous tumor involvement in osteosarcoma of children. MATERIALS AND METHODS: MR images were obtained at diagnosis and after preoperative multiagent chemotherapy in 20 children with osteosarcoma arising in the long bones. Images were reviewed to determine the length of the abnormal intraosseous signal intensity on paired longitudinal T1-weighted and STIR images taken at diagnosis and after chemotherapy. These measurements were compared with those made during a review of similarly oriented pathologic sections of the resected bone. Median differences were calculated and analyzed using Wilcoxon's signed-rank test. RESULTS: Abnormalities detected on T1-weighted images corresponded more closely to pathologic findings than did abnormalities detected on STIR images. On STIR images, readers overestimated tumor extent in 73% of both pre- and postchemotherapy studies. Readers overestimated tumor length on 29 of the 40 STIR images and on 13 of the 40 T1-weighted images. Readers underestimated tumor length on five STIR images and 11 T1-weighted images. Median differences between measurements made at the pathologic examination and on STIR images were statistically significant, both for imaging at diagnosis (p = .001) and for imaging after chemotherapy (p = .005); however, no significant differences were found between measurements made at the pathologic examination and measurements made on T1-weighted images. Measurements of tumor length on either type of imaging did not change significantly after chemotherapy. MR imaging showed 100% sensitivity for epiphyseal tumor spread but poor specificity. False-positive readings occurred in seven of 13 patients with abnormal signal intensity that extended into the adjacent epiphyses. CONCLUSION: Readers estimated intraosseous tumor extent more accurately on T1-weighted images than on STIR images. Readers overestimated tumor length on STIR images. T1-weighted longitudinal images taken before chemotherapy can be used for early planning of surgical approaches to osteosarcoma in children.
OBJECTIVE: We compared how well T1-weighted and short inversion time inversion recovery (STIR) MR images obtained before and after preoperative chemotherapy reveal the extent of longitudinal intraosseous tumor involvement in osteosarcoma of children. MATERIALS AND METHODS: MR images were obtained at diagnosis and after preoperative multiagent chemotherapy in 20 children with osteosarcoma arising in the long bones. Images were reviewed to determine the length of the abnormal intraosseous signal intensity on paired longitudinal T1-weighted and STIR images taken at diagnosis and after chemotherapy. These measurements were compared with those made during a review of similarly oriented pathologic sections of the resected bone. Median differences were calculated and analyzed using Wilcoxon's signed-rank test. RESULTS: Abnormalities detected on T1-weighted images corresponded more closely to pathologic findings than did abnormalities detected on STIR images. On STIR images, readers overestimated tumor extent in 73% of both pre- and postchemotherapy studies. Readers overestimated tumor length on 29 of the 40 STIR images and on 13 of the 40 T1-weighted images. Readers underestimated tumor length on five STIR images and 11 T1-weighted images. Median differences between measurements made at the pathologic examination and on STIR images were statistically significant, both for imaging at diagnosis (p = .001) and for imaging after chemotherapy (p = .005); however, no significant differences were found between measurements made at the pathologic examination and measurements made on T1-weighted images. Measurements of tumor length on either type of imaging did not change significantly after chemotherapy. MR imaging showed 100% sensitivity for epiphyseal tumor spread but poor specificity. False-positive readings occurred in seven of 13 patients with abnormal signal intensity that extended into the adjacent epiphyses. CONCLUSION: Readers estimated intraosseous tumor extent more accurately on T1-weighted images than on STIR images. Readers overestimated tumor length on STIR images. T1-weighted longitudinal images taken before chemotherapy can be used for early planning of surgical approaches to osteosarcoma in children.
Authors: Hans-J Mentzel; K Kentouche; D Sauner; C Fleischmann; S Vogt; D Gottschild; F Zintl; W A Kaiser Journal: Eur Radiol Date: 2004-07-09 Impact factor: 5.315
Authors: Fazel Khan; Andrew Pearle; Christopher Lightcap; Patrick J Boland; John H Healey Journal: Clin Orthop Relat Res Date: 2013-03 Impact factor: 4.176