Quantitative analysis of dynamic hepatobiliary scintigraphy for differentiating complications following liver transplantation

Objectives We investigated whether quantitative analysis of dynamic hepatobiliary scintigraphy (HBS) can differentiate rejection and biliary obstruction following liver transplantation (LT).

Methods Among 259 patients who underwent dynamic Tc-99m DISIDA HBS for abnormal liver function following LT, those with normal HBS, visible leakage, or recurred hepatitis were excluded. Thus, images of a final of 72 patients were visually graded (0-3) for uptake and excretion, and quantitatively analyzed to obtain the portal perfusion index [PPI, (90 - portal angle)/{(arterial angle + (90 - portal angle)}] using first pass flow data, and extraction (k₂₁) and excretion rate constants (k₃), and mean residence time (MRT) using a 3 compartment model. The diagnostic accuracy of the parameters was compared with the final diagnosis based on biopsy, radiology, and clinical course.

Results The final diagnosis was rejection in 24 and biliary obstruction in 48 patients. The rejection group had significantly higher PPI (0.83 ± 0.06 vs. 0.69 ± 0.05, p<0.001), k₃ (0.18 vs. 0.05, p=0.01), and excretion grade (1.71 vs. 0.90, p<0.001). Receiver operating characteristic analysis demonstrated the highest accuracy for PPI (AUC=0.93, p≤0.01 with all others). This was followed by excretion grade (AUC=0.75), k₃ (AUC=0.74), and MRT (AUC=0.68). Using the PPI of 0.72 as threshold, the sensitivity was 95.8% and specificity 79.2%.

Conclusions PPI is a useful index for differentiating rejection from biliary obstruction after LT.