Assessment of early response to neoadjuvant chemotherapy in osteosarcoma using parallely installed PET/CT-MRI: Prospective study

Objectives The objective of this study was to determine whether PET-MR could predict treatment response in osteosarcoma after neoadjuvant chemotherapy (NAC).

Methods From Apr to Oct 2010, eleven of biopsy-proven osteosarcoma patients were enrolled in this prospective study. Two cycles of NAC using methotrexate, adriamycin and cisplatin were done every 3 weeks. Parallely installed PET/CT-MRI scan was performed three times before and after each cycle of chemotherapy. MRI consisted of T1, T2, diffusion weighted image, and dynamic Gd-DTPA enhanced MRI using 3-T scanner. Several parameters including SUVmax of before (SUV1) and after each chemotherapy(SUV2, SUV3), SUV change ratio between each scan, ADC (apparent diffusion coefficient) and its change ratio, steepest slope of the enhancement curve (SS1, SS2, SS3) and its change ratio, tumor volume (TV1, TV2, TV3) and change ratio, largest diameter of tumor and its change ratio were obtained. Treatment response was evaluated using Huvos grading system considering the degree of necrosis. Each parameter was compared between responders and nonresponders using Mann-Whitney test.

Results 6 patients responded NAC and 5 patients did not. SUV1 of responders were significantly lower than that of nonresponders (mean 7.92±3.50, 17.80±9.32, p=0.030). Steepest slope in prechemotherpy scan and tumor volume after chemotherapy were significantly different between two groups (SS1 mean 0.95±0.96, 1.45±0.43, p=0.009; TV2 mean 85.94±77.94, 277.06±154.14, p=0.017 ;TV3 mean 77.74±74.44, 274.30±156.73, p=0.017). SUV2 and SUV3 of responders tended to be lower than that of nonresponders (p=0.052, 0.052). On the contrary, other parameters showed no significant difference between two groups.

Conclusions PET-MR can predict treatment response in osteosarcoma after neoadjuvant chemotherapy using SUVmax , steepst slope, and tumor volume