COMPARISON OF 3 FAT-SUPPRESSION SEQUENCES FOR DETECTING VERTEBRAL METASTASES - TURBOSTIR, PHASE-CONTRAST GRADIENT-ECHO, AND MISTEC-CHOPPER AFTER GADOLINIUM INJECTION

Objectives: Assess three fat suppression sequences used to search for spinal metastases: TurboSTIR, phase contrast gradient-echo, and MISTEC -Chopper after gadolinium injection. Material and methods: A prospecti ve study was conducted in 10 patients with primary neoplasia. MIR sequ ences acquired (1 Tesla) were TurboSTIR, T1 spin-echo with and without gadilinium injection, phase contrast gradient-echo and M-Chop after g adolinium injection. Signal intensity in normal bone marrow, metastati c tissue, and subcutaneous fat as well as background noise was measure d. Signal-to-noise (S/N) ratio was determined. Lesion borders, artefac ts, and extent of detected lesions were determined quantitatively. Bon e marrow signal intensity was also recorded. Results: SIN ratio was be st with gradient-echo which identified well the borders of lesions wit hin the hemopoietic marrow. For lesions located in high-fat marrow (as in post-radiation marrow), the high intensity signal of the lesion co nfounded with the fat signal. TurboSTIR gave effective fat signal supp ression and was particularly useful for yellow marrow, less so for red marrow. This technique confounded cell proliferation with perilesiona l edema (enlarging lesion extention). In one case, this sequence did n ot detect a small lesion visible with the two other sequences. This se quence was sensitive to artefacts (especially vascular artefacts) whic h can produce false nodular images. M-Chop gave good suppression of ve rtebral fat tissue (better for yellow marrow) but subjective detection of lesions was more difficult. Conclusion : The phase contrast gradie nt-echo sequence after gadlinium injection appeared to be the best seq uence excepting cases of post-trauma (radiotherapy or chemotherapy) fa t transformation of the marrow where the TurboSTIR sequence could be p referred.