Impact of the learning-curve with parenchyma sparing hepatectomy for deep tumors: time vs complexity?

Introduction: Parenchyma sparing hepatectomy (PSH) assures adequate future liver remnant (FLR) for deep located tumors, however, PSH requires complex surgical procedures with a learning curve that has never been analysed. Methods: Data from 127 consecutive patients treated with a PSH during the first 9 years of practice were registered in a prospective database. Two patients with preoperative total bilirubin >5 mg/dl were excluded. The initial period (4y) of low (< 14/y) and the later period (5y) of high (>14/y) case volume were compared. Results: Low (n=24) and high (n=101) case volume groups were comparable (P>0.05) for patient's baseline characteristics, PSH requiring HV exposure/resection (92% vs 78%), PSH with vascular reconstruction (25% vs 25%), extrahepatic surgical procedures (25% vs 28%), median blood loss (150 vs 150 ml), blood transfusion (37% vs 27%), overall (50% vs 38%) and severe (8% vs 8%) postoperative complications, and median length of stay (8.5 vs 9 days). Overall 90-day mortality was 1.6%. In high volume case period there was an increasing attitude in first/second order glissonian pedicles exposure (33% vs 68%; P=0.001) and the mean largest lesion size was greater (41±24 cm vs 55±37 cm, P=0.034). Furthermore the operative time was shorter (714±316 min vs 595±242 min, P=0.048) and the resected liver surface per minute was larger (0.93±0.35 cm2/min vs 1.44±0.69 cm2/min, P=0.001). Conclusions: The learning curve had significant impact on operative time and on liver-cut surface area per minute, in spite of an increased attitude in first/second order glissonian pedicles exposure and increased lesion size.