In this prospective study, magnetic resonance imaging (MRI) was compared with computed tomography (CT) in patients with non-Hodgkin's lymphoma (NHDL), and with CT and laparotomy in patients with Hodgkin's disease (HD). Among 31 patients with NHDL, there was agreement between MRI and CT findings in 21 patients; MRI findings were positive and CT findings were negative in 8 patients; and MRI findings were negative and CT findings were positive in 2 patients. The false-negative findings of MRI included a laparotomy-proven mesenteric mass and a 6.6-cm lesion in the spleen, both shown by CT. In 13 evaluable patients with HD, there was agreement between MRI and laparotomy findings in 8 patients; MRI findings were positive and laparotomy findings were negative in 4 patients; and MRI findings were negative and laparotomy findings were positive in 1 patient. CT findings agreed with laparotomy findings in nine patients; CT findings were positive and laparotomy findings were negative in one patient: and CT findings were negative and laparotomy findings were positive in three patients. This suggested that MRI, although more sensitive than CT, was less specific with more false-positive findings. The spin-lattice relaxation time (T1) for the spleen was generally higher in patients with HD who had pathologically confirmed splenic involvement, and tended to increase with increasing spleen weight. This study was performed on a prototype 0.15 Tesla (Technicare Inc., Solon, OH) resistive unit at a time when methods and reporting for MRI were still being developed. Although MRI appears to have fewer false-negative findings than CT in evaluating the abdomen of lymphoma patients, the lack of a gastrointestinal contrast agent and specificity of T1 elevations in the spleen would not suggest that MRI could replace CT or laparotomy as a staging technique.
