FAST SPIN-ECHO MR IN THE DETECTION OF VERTEBRAL METASTASES - COMPARISON OF 3 SEQUENCES

PURPOSE: To examine the relative capabilities for the detection of ver tebral metastases of three available fast spin-echo sequences: T1-weig hted fast spin-echo, short tau inversion recovery (STIR) fast spin-ech o, and T2-weighted fast spin-echo sequences with chemical shift select ive saturation pulse fat suppression. METHODS: Fourteen patients were evaluated prospectively over a 2-month period with T1-weighted fast sp in-echo (four echo train, four acquisitions, 1 min 59 sec-2 min 37 sec ), STIR fast spin-echo (16 echo train, four acquisitions, 2 min 30 sec -3 min 19 sec), and T2-weighted fast spin-echo (16 echo train, 4 acqui sitions, 2 min 27 sec-3 min 16 sec). For all three pulse sequences, me asurements were obtained of the signal intensities of normal marrow, a bnormal marrow, fat, and noise posterior to the spine. Contrast-to-noi se ratios were calculated for metastases in each case. Lesions were ev aluated by three observers and rated for size, location, and conspicui ty. RESULTS: Signal intensities of fat, normal marrow, and noise were highest for T1-weighted fast spin-echo sequences. STIR fast spin-echo and fat-suppressed T2-weighted fast spin-echo had approximately simila r fat-suppression capabilities. Though contrast-to-noise ratios were h ighest overall for STIR fast spin-echo, the finding was not statistica lly significant and lesion conspicuity was deemed better with fat-supp ressed T2-weighted fast spin-echo and T1-weighted fast spin-echo image s. Discrete lesions were well identified on all three pulse sequences. CONCLUSION: Fast spin-echo sequences appear promising for the detecti on of vertebral metastases. Further work should be directed toward com parison with conventional spin-echo to determine whether fast spin-ech o may replace conventional spin-echo sequences for evaluation of verte bral metastases.