Optimal Scan Window for Detection of Hypervascular Hepatocellular Carcinomas During MDCT Examination

Abstract

OBJECTIVE. The purpose of this study was to define the optimal scan window for acquiring arterial phase images in the detection of hypervascular hepatocellular carcinomas (HCCs).

MATERIALS AND METHODS. Biphasic arterial phase CT examinations were performed using a 16-MDCT scanner on 198 patients (159 men and 39 women; mean age, 59 years; age range, 25-82 years) with nodular HCC. All examinations were performed after administering 120-150 mL of a nonionic contrast media (370 mg I/mL) at a rate of 3-4 mL/s. The scan delay—the interval between when the bolus-tracking program detected the threshold enhancement of 100 H in the abdominal aorta and the start of the first arterial scan—was progressively lengthened by 2-second intervals, from 10 seconds in group 1 to 20 seconds in group 6. The second arterial phase scan was started 6 seconds after the end of the early scan. A tube collimation of 1.5 mm, a table feed of 18 mm per rotation, an image thickness of 3 mm, and 3-mm increments were used. The duration of each phase scan was 4.5-8.8 seconds. Tumor-to-liver attenuation difference (TLAD) at the first (TLAD1) and second (TLAD2) arterial phase images were compared lesion by lesion. Four observers assigned subjective ratings of visual conspicuity and individual preferences for each phase in each group.

RESULTS. The mean threshold time (100 H) was 18.4 ± 3.1 seconds, and 97% of patients were within the range of 13-24 seconds. The mean TLAD1 of groups 3 to 6 and the mean TLAD2 of groups 1 to 5 were all comparable; they were also all significantly (p < 0.005) higher than the mean TLAD1 of groups 1 and 2 and the mean TLAD2 of group 6. In groups 1 and 2, the mean TLAD2 was significantly higher than the mean TLAD1 (p < 0.001); in groups 5 and 6, the mean TLAD1 was significantly higher than the mean TLAD2 (p < 0.001). In groups 3 and 4, the mean TLAD1 and TLAD2 were similar. The visual conspicuity and individual preferences were higher for the first-phase image in groups 3 to 6 and the second-phase image in groups 1 and 2.

CONCLUSION. The optimal scan window for arterial phase images in the detection of HCC seems to be approximately 14-30 seconds from the 100-H threshold.