Department of Digestive Diseases, Institut Mutualiste Montsouris, Université Paris Descartes, Paris, France.
Hepato-Biliary-Pancreatic Surgery Division, Department of Surgery, Graduate School of Medicine, The University of Tokyo, 7-3-1 Hongo, Bunkyo-ku, Tokyo, 113-8655, Japan.
Surg Endosc. 2020 May;34(5):2056-2066. doi: 10.1007/s00464-019-06986-6. Epub 2019 Jul 23.
A procedure-based laparoscopic liver resection (LLR) classification (IMM classification) stratified 11 different LLR procedures into 3 grades. IMM classification assessed the difficulty of LLR differently than an index-based LLR classification (IWATE criteria), which scored each procedure on an index scale of 12. We validated the difference of 3 IMM grades using an external cohort, evaluated the IMM classification using the scores of the IWATE criteria, and compared the performance of IMM classification with the IWATE criteria and the minor/major classification.
Patients undergoing LLR without simultaneous procedures were selected from a prospectively maintained database at the Institut Mutualiste Montsouris (IMM cohort) and from the database of 43 Japanese institutions (JMI cohort). Surgical and postoperative outcomes were evaluated according to the 3 IMM grades using the JMI cohort. The 11 LLR procedures included in the IMM classification were scored according to the IWATE criteria. The area under the curves (AUCs) for surgical and postoperative outcomes were compared.
In the JMI (n = 1867) cohort, operative time, blood loss, conversion rate, and major complication rate were significantly associated with a stepwise increase in grades from I to III (all, P < 0.001). In the IMM (n = 433) and JMI cohorts, IMM grades I, II, and III corresponded to three low-scoring, two intermediate-scoring, and six high-scoring LLR procedures as per the IWATE criteria, respectively. Mean ± standard deviation among the IMM grades were significantly different: 3.7 ± 1.4 (grade I) versus 7.5 ± 1.7 (grade II) versus 10.2 ± 1.0 (grade III) (P < 0.001) in the IMM cohort and 3.6 ± 1.4 (grade I) versus 6.7 ± 1.5 (grade II) versus 9.3 ± 1.4 (grade III) (P < 0.001) in the JMI cohort. The AUCs for surgical and postoperative outcomes are higher for the 3-level IMM classification than for the minor/major classification.
The difference of 3 IMM grades with respect to surgical and postoperative outcomes was validated using an external cohort. The 3-level procedure-based IMM classification was in accordance with the index-based IWATE criteria. The IMM classification performed better than the minor/major classification for stratifying LLR procedures.
基于手术的腹腔镜肝切除术（LLR）分类（IMM 分类）将 11 种不同的 LLR 手术分为 3 个等级。与基于指数的 LLR 分类（IWATE 标准）不同，IMM 分类评估 LLR 的难度，后者根据 12 个指数对每个手术进行评分。我们使用外部队列验证了 3 个 IMM 等级的差异，使用 IWATE 标准评估了 IMM 分类，并比较了 IMM 分类与 IWATE 标准和小/大分类的性能。
从 Institut Mutualiste Montsouris（IMM 队列）的前瞻性维护数据库和 43 个日本机构的数据库中选择未同时进行手术的 LLR 患者（JMI 队列）。使用 JMI 队列根据 3 个 IMM 等级评估手术和术后结果。根据 IMM 分类包含的 11 种 LLR 手术，根据 IWATE 标准进行评分。比较手术和术后结果的曲线下面积（AUC）。
在 JMI（n=1867）队列中，手术时间、出血量、转化率和主要并发症率与等级从 I 到 III 的逐步增加显著相关（均 P<0.001）。在 IMM（n=433）和 JMI 队列中，根据 IWATE 标准，IMM 等级 I、II 和 III 分别对应三个低评分、两个中评分和六个高评分的 LLR 手术。IMM 等级之间的平均值±标准差差异显著：IMM 队列中为 3.7±1.4（等级 I）与 7.5±1.7（等级 II）与 10.2±1.0（等级 III）（P<0.001），JMI 队列中为 3.6±1.4（等级 I）与 6.7±1.5（等级 II）与 9.3±1.4（等级 III）（P<0.001）。手术和术后结果的 AUC 更高的是 3 级 IMM 分类，而不是小/大分类。
使用外部队列验证了 3 个 IMM 等级在手术和术后结果方面的差异。基于手术的 3 级 IMM 分类与基于指数的 IWATE 标准一致。与小/大分类相比，IMM 分类在 LLR 手术分层方面表现更好。
