Estimation of right hemiliver graft weight for living donor liver transplantation using portal vein diameter - a "second checkpoint"

Abstract

Background

Estimation of graft weight for live donor liver transplant using portal vein diameter has not been validated widely.

Objective

To observe the accuracy of portal vein diameter based formula in predicting graft weight.

Methods

Graft weight was estimated using standard liver volume (SLV) multiplied by the ratio of portal vein branch diameter (Lee’s formula). SLV was calculated using different formulae described in the literature. The most accurate formula was compared with three dimensional computed tomography volumetry in terms of accuracy of prediction of actual graft weight. Factors which predicted percentage error of more than 15 in computed tomography were analyzed.

Results

In 307 right hemiliver grafts, SLV calculation by Urata’s method in Lee’s formula was the most accurate (P=0.60 in Analysis of Variance) among all SLV methods. Lee’s formula with Urata and computed tomography volumetry had a good correlation with actual graft weight (r=0.77 vs. r=0.8) which was confirmed by Bland Altman analysis. On volumetry 45 patients had a percent error of more than 15. On logistic regression analysis, an estimated graft volume of greater than 800cc was a significant factor ( p=0.008, odds ratio 2.99) and, in these patients Lee’s  formula was better(mean error   9.2 ±7.8 vs. 20.2±4.5, p<0.001).

Conclusions

Lee's formula with SLV by Urata’s method was accurate and can act as the second checkpoint after three dimensional computed tomography volumetry. Computed tomography estimate of graft weight more than 800 is associated with higher inaccuracy and in those patients portal vein diameter based formulae fared better.