OBJECTIVE: Our goal was to determine the incidence and pattern of red marrow reconversion on MRI of adults receiving granulocyte-stimulating factors as part of their chemotherapy regimen for primary musculoskeletal neoplasms and correlate the changes with WBC counts. MATERIALS AND METHODS: Twenty-five adults with soft-tissue sarcomas (n = 15) or primary bone tumors (n = 10) who underwent chemotherapy that included granulocyte-stimulating factors formed the study group. Two radiologists retrospectively evaluated the MRI studies by consensus before and after therapy to determine the presence of changes consistent with red marrow reconversion. Changes were categorized by appearance on T1- and T2-weighted images for location, pattern, and extent of marrow involvement. WBC counts at the time of MRI were recorded. Records were examined for evidence of marrow-infiltrating tumors or metastases. RESULTS: Ten (40%) of the 25 patients underwent bone marrow changes consistent with red marrow reconversion. In seven (70%) of 10 patients, the changes were diffuse in visualized bones and simulated diffuse marrow-infiltrative tumor. In three (30%) of 10 patients, the changes were focal, simulating metastases. T2 signal prolongation was identified in the marrow reconversion in nine patients (90%), although all had shortened T1 signal. Four patients (40%) had elevated WBC counts at the time of the MRI after therapy, but all had shortened T1 signal. Four patients (40%) had elevated WBC counts at the time of the MRI after therapy, five (50%) had normal counts, and one (10%) had a below-normal count. No osseous metastases or marrow-infiltrating tumors were found during follow-up. CONCLUSION: Forty percent of patients showed marrow changes mimicking tumor on MRI that were attributable to red marrow reconversion, which correlated moderately with WBC response.
OBJECTIVE: Our goal was to determine the incidence and pattern of red marrow reconversion on MRI of adults receiving granulocyte-stimulating factors as part of their chemotherapy regimen for primary musculoskeletal neoplasms and correlate the changes with WBC counts. MATERIALS AND METHODS: Twenty-five adults with soft-tissue sarcomas (n = 15) or primary bone tumors (n = 10) who underwent chemotherapy that included granulocyte-stimulating factors formed the study group. Two radiologists retrospectively evaluated the MRI studies by consensus before and after therapy to determine the presence of changes consistent with red marrow reconversion. Changes were categorized by appearance on T1- and T2-weighted images for location, pattern, and extent of marrow involvement. WBC counts at the time of MRI were recorded. Records were examined for evidence of marrow-infiltrating tumors or metastases. RESULTS: Ten (40%) of the 25 patients underwent bone marrow changes consistent with red marrow reconversion. In seven (70%) of 10 patients, the changes were diffuse in visualized bones and simulated diffuse marrow-infiltrative tumor. In three (30%) of 10 patients, the changes were focal, simulating metastases. T2 signal prolongation was identified in the marrow reconversion in nine patients (90%), although all had shortened T1 signal. Four patients (40%) had elevated WBC counts at the time of the MRI after therapy, but all had shortened T1 signal. Four patients (40%) had elevated WBC counts at the time of the MRI after therapy, five (50%) had normal counts, and one (10%) had a below-normal count. No osseous metastases or marrow-infiltrating tumors were found during follow-up. CONCLUSION: Forty percent of patients showed marrow changes mimicking tumor on MRI that were attributable to red marrow reconversion, which correlated moderately with WBC response.