8 ECHO T2 MEASUREMENTS OF ADRENAL MASSES - LIMITATIONS OF DIFFERENTIAL-DIAGNOSIS BY RELAXATION-TIME DETERMINATION

Purpose: Previous studies have revealed that benign and malignant adre nal masses differ in their T2-related MRI characteristics, but there i s sufficient overlap in these characteristics that very accurate diffe rential diagnosis is not possible. This ambiguity might be due to vari ations inherent in the measurement techniques or to real overlap in th e T2 relaxation times of the lesion groups. We attempted to reduce the scatter and overlap of data from adenomas and malignancies by using a n eight echo T2 assessment, which we compared with a two echo techniqu e and with reference tissue brightness ratio determinations. Method: F orty-eight adrenal masses in 44 adult patients were assessed; 30 were diagnosed as adenomas and 18 as malignancies by means other than MR. E ach lesion was subjected to a single slice eight SE (spin echo) image (TR 2,000, TE 20-160), from which a T2 relaxation time was calculated. T2 relaxation times were also calculated using two echoes (TE 20 and 100), as were lesion/liver, lesion/fat, and lesion/muscle brightness r atios (TE 100). The differential diagnostic efficacies of the techniqu es were compared by receiver operating characteristic (ROC) analysis. Results: Although the means of the malignant and benign groups differe d, significant overlap was present for each assessment technique. ROC analysis showed that the best differentiation was achieved by the lesi on/fat brightness ratio, followed very closely by the eight echo techn ique. The eight echo technique was significantly better at differentia tion than the two echo technique. Conclusion: Although eight echo T2 d etermination has less variability and permits slightly better differen tial diagnosis than most other T2-dependent MR techniques, considerabl e ambiguity persists that may be due to intrinsic over-lap in the T2 r elaxation characteristics of the lesions rather than to variability in imaging techniques.