Liver volume measurement plays a critical role in the clinical success of living donor liver transplantation (LDLT). CT liver volumetry is used for volumetric assessment of the donor hepatectomy. However, the degree of accuracy of the preoperative predicted volumetric measurements remains unclear.

This systematic review will assess the accuracy of CT liver volumetry.

A systematic literature search was conducted to evaluate the accuracy of the CT based liver volumetry. To assess the difference between the left and right liver lobe the ratio between preoperative estimated graft volume (EGV) and actual graft weight (AGW) was calculated. A meta-analysis was performed to compare the EGV to the AGW of the left and right lobe with automated and manual CT volume prediction.

Thirty-one studies met the inclusion criteria and 1336 patients were included in the meta-analysis. The EGV of the right liver is overestimated with a mean of 4.01% and 2.99% for automated and manual volumetry respectively. The left lobe is overestimated with 6.28% and 14.41% for automated and manual volumetry respectively. For a right lobe liver graft automated volume prediction showed a mean difference of 43.9 g (95% confidence interval (CI): 21.21–66.54, p<0.001) between EGV and AGW. For manual volume prediction this was 34.0 g (95% CI: 11.85–56.11, p = 0.003). For a left lobe with automated volume prediction the mean difference was 46.0 g (95% CI: 20.91–71.09, p<0.001) between EGV and AGW. For manual volume prediction this was 39.6 g (95% CI: 8.40–70.74, p = 0.01).

The volumes of the right and left liver lobe seem to be overestimated with automated and manual CT volume prediction. Considering the larger estimation error in the prediction of the left liver lobe, special attention should be paid to the volume when performing a LDLT with a left liver lobe to prevent small for size syndrome.