Explainable artificial intelligence prediction-based model in laparoscopic liver surgery for segments 7 and 8: an international multicenter study.

BACKGROUND: Artificial intelligence (AI) is becoming more useful as a decision-making and outcomes predictor tool. We have developed AI models to predict surgical complexity and the postoperative course in laparoscopic liver surgery for segments 7 and 8. METHODS: We included patients with lesions located in segments 7 and 8 operated by minimally invasive liver surgery from an international multi-institutional database. We have employed AI models to predict surgical complexity and postoperative outcomes. Furthermore, we have applied SHapley Additive exPlanations (SHAP) to make the AI models interpretable. Finally, we analyzed the surgeries not converted to open versus those converted to open. RESULTS: Overall, 585 patients and 22 variables were included. Multi-layer Perceptron (MLP) showed the highest performance for predicting surgery complexity and Random Forest (RF) for predicting postoperative outcomes. SHAP detected that MLP and RF gave the highest relevance to the variables "resection type" and "largest tumor size" for predicting surgery complexity and postoperative outcomes. In addition, we explored between surgeries converted to open and non-converted, finding statistically significant differences in the variables "tumor location," "blood loss," "complications," and "operation time." CONCLUSION: We have observed how the application of SHAP allows us to understand the predictions of AI models in surgical complexity and the postoperative outcomes of laparoscopic liver surgery in segments 7 and 8.

Complications after Hepatectomy for Hepatocellular Carcinoma Independently Shorten Survival: A Western, Single-Center Audit.

BACKGROUND: The impact of postoperative complications (POCs) on long-term outcomes following hepatocellular carcinoma (HCC) resection remains to be ascertained. METHODS: All consecutive HCC resected at a single center were analyzed. Patients with POCs, classified according to Clavien classification, were compared to those without in terms of demographics, pathology, management, overall survival (OS), and disease-free survival (DFS). Independent prognostic factors of POCs were identified using multivariable regression models. RESULTS: Among 341 patients, overall POCs rate was 34% (n = 116) and grade III-IV POCs rate was 14.4% (n = 49). POCs were an independent negative factor for OS [hazard ratio (HR) 1.40, 95% confidence interval (CI) 1.12-2.26, p = 0.009] with BCLC stage, the need for combined procedure, intraoperative transfusion, and the METAVIR score of the underlying parenchyma. Similarly, occurrence of POCs was associated independently with DFS (HR 1.59, 95% CI 1.18-2.15, p = 0.002), together with the presence of portal hypertension, BCLC stage, the need for combined procedure, intraoperative transfusion, and the presence of satellite nodules. After stratification, the negative impact of morbidity on OS and DFS reached statistical significance in the BCLC stage A subset only (p = 0.026, and p < 0.001, respectively). Open resection, intraoperative transfusion, and the existence of underlying liver injury were independent predictors of POCs. CONCLUSIONS: POCs should be considered as a long-term prognostic factor. Careful patient selection requiring underlying liver assessment and appropriate strategy, such as mini-invasive surgery and restricted transfusion policy, might be promoted to prevent POCs.

Robotic-Assisted Versus Laparoscopic Left Lateral Sectionectomy: Analysis of Surgical Outcomes and Costs by a Propensity Score Matched Cohort Study.

BACKGROUND: After comparing with open approach, left lateral sectionectomy (LLS) has become standard in terms of short-term outcomes without jeopardizing long-term survival when performed for malignancy. The aim of this study was to compare the short-term and economic outcomes of laparoscopic (L-LLS) and robotic (R-LLS) LLS. METHODS: All consecutive patients who underwent L-LLS or R-LLS from 1997 to 2014 were analyzed. Short-term and economic outcomes were compared between the two groups using a propensity score matching (PSM). RESULTS: Ninety-six consecutive cases of LLS were performed using the laparoscopic (80 cases; 83 %) or robotic (16 cases; 17 %) approach. The two groups were similar for operative and surgical outcomes. Operation time was similar in the R-LLS compared to the L-LLS group (190 vs. 162 min; p = 0.10). Perioperative costs were higher (1457 € vs. 576 €; p < 0.0001) in the R-LLS group than in the L-LLS group; however, postoperative costs were similar between the two groups (4065 € in the R-LLS group vs. 5459 € in the L-LLS group; p = 0.30). Total costs were similar between the two groups (5522 € in the R-LLS group vs. 6035€ in the L-LLS group; p = 0.70). The PSM included 14 patients for each group. Surgical and economic outcomes remained similar after PSM, except for total operating time which was significantly longer in the R-LLS group than in the L-LLS group. CONCLUSIONS: Even if feasible and safe, the robotic approach does not seem so far to offer additional benefit in terms of intra- and postoperative outcomes over the laparoscopic approach in patients requiring LLS. Total costs associated with the R-LLS group are not greater than that associated with the L-LLS group, which is the standard of care so far.

Sleeve Gastrectomy After Liver Transplantation: Feasibility and Outcomes.

BACKGROUND: Knowledge regarding the feasibility and safety of sleeve gastrectomy (SG) in obese liver transplant recipients is scarce. We report our experience of sleeve gastrectomy following liver transplantation (LT). METHODS: All patients who had undergone LT and subsequently underwent SG at our institution were retrospectively reviewed. Surgical outcomes, liver and kidney function tests, outcomes of obesity-related comorbidities, and excess weight loss were analyzed. RESULTS: Between May 2008 and February 2015, six consecutive patients underwent SG after LT. Three procedures (50%) were performed totally by laparoscopy, and three by upfront laparotomy for concomitant incisional hernia complex repair. Within the first 30 days, one complication occurred: early gastric fistula that required multiple endoscopic procedures and re-intervention, followed by death 19 months after SG due to multi-organ failure. Another patient had one late complication: chronic infection on a parietal mesh successfully controlled by mesh removal. Excess weight loss averaged 76% at 2 years with a median BMI of 28 (21-39) kg/m2. Median follow-up was 37.2 months (range 13-101 months). Median length of stay was 9 days (range: 6-81 days). CONCLUSIONS: SG is technically feasible after LT and resulted in weight loss without adversely affecting graft function and immunosuppression. However, morbidity and mortality are high.