Performance of a modified 3-level classification in stratifying open liver resection procedures in terms of complexity and postoperative morbidity

Abstract

Background: Open liver resection (OLR) procedures have traditionally been categorized by the minor/major classification based on 3 contiguous Couinaud segments or number of resected segments (wedge resection, segmentectomy, sectionectomy, or hemihepatectomy). However, the minor/major and segment-based classifications are not always associated with surgical complexity and postoperative morbidity. We tested the hypothesis that a 3-level liver resection classification based on surgical and postoperative outcomes and originally devised for laparoscopic liver resection (LLR) is superior to classifications based on a minor/major nomenclature, and number of resected segments for stratifying OLR procedures by surgical complexity. Methods: Patients undergoing first OLR without simultaneous procedures in Houston (Houston cohort) during 1998-2016 and Tokyo (Tokyo cohort) during 1994–2014 were evaluated. The 3-level classification used in this report is a modified version of the previously reported classification, which classifies 11 LLR procedures as grade I (low complexity), grade II (intermediate complexity), or grade III (high complexity): grade I, wedge resection for anterolateral (AL)/posterosuperior (PS) segments <3 cm and left lateral sectionectomy; grade II, AL segmentectomy and left hepatectomy; grade III, PS segmentectomy, right posterior sectionectomy, right hepatectomy, central hepatectomy, and extended left/right hepatectomy. We compared outcomes between procedures in the 3 levels of our classification and the areas under the receiver-operating characteristics curves (AUCs) of our 3-level classification with the minor/major and segment-based classifications. Results: The most common indication was colorectal liver metastasis (56.6%) in Houston cohort (n = 2147) and hepatocellular carcinoma (71.2%) in Tokyo cohort (n = 1202). In both cohorts, operative time and estimated blood loss (EBL) were significantly higher for grade III than for grade I, for grade III than for grade II, and for grade II than for grade I (Figure 1A–D). In both cohorts, comprehensive complication index (CCI) was significantly higher for grade III than for grade I and for grade III than for grade II, but CCI did not differ significantly between grades I and II (Houston cohort, P = 0.096; Tokyo cohort, P = 0.426) (Figure 1E–F). For both operative time and EBL, AUCs was significantly higher for our 3-level classification than for the minor/major (operative time, 0.624 vs. 0.541, P < 0.001; EBL, 0.706 vs. 0.679, P = 0.002) or segment-based classification (operative time, 0.624 vs. 0.546, P < 0.001; EBL, 0.706 vs. 0.681, P = 0.002) in Houston cohort. Conclusion: Our 3-level classification effectively stratified 11 OLR procedures with respect to surgical and postoperative outcomes. Operative time and EBL, markers of surgical complexity, differed significantly between the grades in our classification and increased in stepwise fashion from grade I to grade III. Our 3-level classification predicted surgical complexity better than the minor/major and the segment-based classifications. These findings indicate that our 3-level classification may be a useful replacement for the minor/major and segment-based classifications in future studies analyzing open and laparoscopic liver resections at Western and Eastern centers.