Innovative formulae for the estimation of standard liver volume in the era of widespread imaging analysis software.

Abstract

Various formulae have been used for the estimation of standard liver volume (SLV) in preparation for living donor liver transplantation (LDLT). However, these formulae have the disadvantage of being constructed using parameters that are substantially affected by the patient's condition. Here, we aimed to establish more precise formulae that are less affected by the general condition of the patient. We analyzed the liver volumes of LDLT donors and patients with normal livers (total: n=213) using the SYNAPSE VINCENT imaging analysis system, to develop new formulae. The accuracy of the new formulae were compared with those of existing formulae in a separate validation group of healthy patients (n=200). The new formulae were also validated using 81 LDLT recipients to assess their utility for graft selection in LDLT. Body surface area (BSA) and skeletal muscle mass index (SMI) independently affected total liver volume (TLV). We produced new formulae for SLV incorporating SMI: SLV = 32.2×L3-SMI-466.9 for men, with R2 .92, and 25.7×L3-SMI-55.97 for women, with R2 .79 (alongside a BSA formula with R2 .57), which provided the most accurate predictions of TLV in the validation group. A graft volume (GV)/SLV <.35, calculated using the new formulae, predicted the postoperative prognosis, including the development of small-for-size syndrome, sepsis, or acute rejection, significantly more effectively than GV/SLV using the previous formulae. The newly developed L3-SMI-based formula is more accurate for the estimation of SLV than previously reported formulae, and may thus help to improve the safety of LDLT.