Both groups underwent major resections (≥3 Couinaud segments) in roughly half of the cases, a rate far exceeding that reported in cirrhotic patients undergoing resection for HCC [4, 7, 12], a reflection of their well-preserved liver function. The likelihood of perioperative blood transfusion and the perioperative mortality rate were significantly higher for patients without underlying disease (44.3 vs. 24.4 %, 3.7 vs. 2.4 %, respectively), possibly a sequela of a greater mean tumor size (Table 1).

As expected, most patients without underlying disease had negligible hepatic fibrosis in the background liver parenchyma; by contrast, the majority of patients with underlying disease had histological evidence of either mild (F1) or moderate (F2) hepatic fibrosis according to the Scheuer staging system [28]. It should be noted that patients with bridging fibrosis (F3) are generally grouped with cirrhotic patients (F4) and are not included in this series.

The comparatively larger mean tumor diameter (9.7 vs. 7.3 cm) and greater percentage of patients with tumor diameter greater than 7 cm (61.3 vs. 39.0 %) observed in patients without underlying disease is clearly related to the mode of presentation: those tumors were generally recognized once they had reached an adequate size to cause mass-related symptoms. By contrast, most of the 81 HBV-positive patients in the underlying disease group were enrolled in HCC surveillance programs and came to diagnosis at an earlier temporal point in tumor growth, either through an abnormal liver ultrasound examination or elevated serum AFP. Despite larger tumor size, patients without underlying disease developed fewer poorly differentiated tumors (14.5 vs. 24.4 %). Table 1 summarizes additional pathological findings.