Comparison between CT volumetry and extracellular volume fraction using liver dynamic CT for the predictive ability of liver fibrosis in patients with hepatocellular carcinoma

Objectives To compare the predictive ability of liver fibrosis (LF) by CT-volumetry (CTV) for liver and spleen and extracellular volume fraction (ECV) for liver in patients undergoing liver resection. Methods We retrospectively analysed 90 consecutive patients who underwent CTV and ECV. Manually placed region-of-interest ECV (manual-ECV), rigid-registration ECV (rigid-ECV), and nonrigid-registration ECV (nonrigid-ECV) were calculated as ECV(%) = (1-haematocrit) × (ΔHU liver /ΔHU aorta ), where ΔHU = subtraction of unenhanced phase from equilibrium phase (240 s). Manual-ECV was compared with CTV for the estimation of LF. The total liver volume to body surface area (TLV/BSA), splenic volume to BSA (SV/BSA), ratio of TLV to SV (TLV/SV), ratio of right liver volume to SV (RV/SV), and liver segmental volume ratio (LSVR) were measured. ROC analyses were performed for ECV and CTV. Results After excluding 10 patients, seventy-eight (97.5%) out of 80 patients had a Child-Pugh score of 5 points, and two (2.5%) patients had a Child-Pugh score of 6 points. AUC of ECV showed no significant difference among manual-ECV, rigid-ECV, and nonrigid-ECV. TLV/BSA, SV/BSA, TLV/SV, and RV/SV had a higher correlation with LF grades than manual-ECV. AUC of SV/BSA was significantly higher than that of manual-ECV in F0-1 vs F2-4 and F0-2 vs F3-4. AUC of SV/BSA (0.76–0.83) was higher than that of manual-ECV (0.61–0.75) for all LF grades, although manual-ECV could differentiate between F0-3 and F4 at high AUC (0.75). Conclusions In patients undergoing liver resection, SV/BSA is a better method for estimating severe LF grades, although manual-ECV has the ability to estimate cirrhosis (≥ F4). Key Points The splenic volume is a better method for estimating liver fibrosis grades. The extracellular volume fraction is also a candidate for the estimation of severe liver fibrosis.