Portal venous tumour extension and intrahepatic metastasis result in a poor prognosis following hepatectomy for hepatocellular carcinoma (HCC). Anatomical resection is, in theory, preferable for eradicating these types of invasion. Des-γ-carboxy prothrombin (DCP) has been reported to be associated with adverse pathological variables. This study investigated the significance of anatomical resection and DCP as predictive factors for postoperative recurrence of HCC.
A retrospective cohort study was carried out in 138 consecutive patients who underwent hepatectomy for HCC smaller than 5 cm using the Cox proportional hazards model.
Eight factors were univariately related to poor prognosis (in decreasing order of hazard ratio): intrahepatic metastasis, multiple tumours, α-fetoprotein 32 ng/ml or more; DCP greater than 0·1 arbitrary units (AU), tumour-exposed surgical margin, vascular invasion, non-anatomical resection and tumour 2·5 cm or more. Three variables (DCP, vascular invasion and tumour-exposed surgical margin) were excluded by a stepwise procedure in multivariate analysis. Although DCP was not an independent prognostic factor, a model replacing intrahepatic metastasis with DCP showed similar predictive accuracy in a receiver-operating characteristic curve.
Anatomical resection appeared to have a beneficial effect on recurrence-free survival after hepatectomy for HCC. DCP measurement was effective in predicting HCC recurrence and had the advantage that it can be assessed before operation. © 1999 British Journal of Surgery Society Ltd