Hepatectomy is a basic and effective treatment for primary and secondary liver malignancies that improves survival rates, particularly for patients with early and middle stage localized disease [1, 2]. With the development of modern medicine, precise hepatic segmentectomies are becoming increasingly mature, and as such higher requirements are needed for anesthesia and perioperative management [3]. The main surgical problem during segmental hepatectomies is intraoperative blood loss. During liver resections, intermittent portal vein triple clamping (Pringle maneuver) is associated with controlled low central venous pressure, which reduces intraoperative blood loss [2, 4]. However, subsequent tissue ischemia and reperfusion may lead to hepatic ischemia-reperfusion injuries (HIRIs), which usually occur when blood supply to the liver is temporarily blocked and subsequently restored [5].

The mechanisms involved in HIRIs are complex and yet to be fully understood. These include the adhesion of white blood cells to endothelial cells, the activation of Kupffer cells, the release of inflammatory cytokines, free radicals, nitric oxide and adenosine, the induction of the inflammatory cascade, and cellular apoptosis [6, 7]. Tolerance of liver tissue to ischemia depends on several factors, such as duration of ischemia, liver collateral circulation, and liver metabolic needs, among others. Therefore, it is difficult to determine the exact safe ischemic time for each surgery. On the other hand, while the restoration of blood flow is essential to prevent irreversible liver cell damage, the reperfusion itself may aggravate ischemic liver cell damage.

After an extensive hepatectomy, HIRI of the residual liver may be a serious complication, leading to postoperative liver dysfunction and increased mortality [5]. In order to protect the residual liver from HIRI, several techniques have been used, including drugs and ischemic preconditioning, or remote ischemic preconditioning (RIPC), none of which are established as standard of care. Organ protection by RIPC began with the study of cardiac muscle protection, which involves repeated temporary cessation of blood flow to the limbs [5, 8]. RIPC procedures are non-invasive and therefore a more suitable method to reduce HIRI.

Even though RIPC has been shown to have hepatoprotective effects in several animal experiments [9,10,11], patient-based studies have shown controversial results [12,13,14]. Only two systematic reviews on this topic were found in the literature [15, 16], both of which contained fewer studies and less data than this study. In addition, one mistakenly included patients with remote ischemic postconditioning (RIpostC) in the meta-analysis, and the other included a randomized controlled trial (RCT) of liver transplant recipients. Moreover, these were published within a year of each other, despite yielding conflicting conclusions. Hence, this study aimed to provide an updated systematic review of the perioperative effects of RIPC in patients undergoing hepatectomy. The study’s hypothesis is that RIPC is beneficial in reducing the effects of HIRI in patients undergoing hepatectomy.

This systematic review was prepared in concordance with PRISMA and AMSTAR2 recommendations to assess methodological quality [17, 18]. RCTs were included to compare perioperative outcomes in patients undergoing hepatectomy with or without RIPC. This systematic review has been registered with PROSPERO, under registration number CRD42022333383.

Articles published until December 2023 were searched via PubMed, OVID, Web of Science, Cochrane library clinical trial databases, Embase, and other sources without language restrictions. Search terms consisted of various combinations of ‘remote ischemic preconditioning’, ‘distant ischemic preconditioning’, ‘remote ischemic conditioning’, ‘remote ischemic adaptation’, ‘limb ischemic preconditioning’, or ‘RIPC’, and ‘hepatectomy’, ‘liver resection’, ‘liver transplantation’, ‘liver graft’, or ‘hepatic ischemia-reperfusion’. In addition, references of included studies and other existing meta-analyses were collected to obtain additional eligible studies (Supplementary Table S1).

Four researchers independently reviewed and retrieved all full-text articles simultaneously. Different views were discussed among the four researchers, and duplicated articles in databases were merged. When duplicate studies were found from the same population, the latest or most complete study was included.

The inclusion criteria were as follows: (1) subjects were patients undergoing hepatectomy or living donor hepatectomy; (2) intervention was RIPC versus control group without RIPC; (3) research type was prospective RCT; (4) outcomes were postoperative liver synthetic function.

The exclusion criteria were as follows: (1) animal experimental studies and ex vivo, in vitro or in silico model studies; (2) retrospective or single-arm studies; (3) case studies, cross-over studies, studies without a separate control group, editorials, meta-analyses and reviews; (4) abstract only studies; (5) studies without postoperative aminotransferase levels or data from review articles.

Two researchers (Chun Tian and Aihua Wang) independently extracted data from each article. Any disagreements were resolved by consensus of a third researcher (He Huang). The following information was extracted from the included articles: the first author, year of publication; country or region of study, type of study, sample size, demographic data, outcomes, among others. If this information was not available in the study’s text, study graphs were enlarged and measured using the Grab software. In instances where data were not reported or unclear, researchers were contacted via e-mail (max. 2 attempts).

Included RCTs were assessed using the recommended Cochrane Collaboration biased-risk assessment table. This assessment was carried out independently by four researchers. Any disagreements were resolved by consensus. The biased-risk assessment table included random sequence generation, allocation concealment, blinding of participants and personnel, blinding of outcome assessment, incomplete outcome data, selective reporting, and other bias. Each study was classified as high, low, or uncertain risk.

The methodological quality of the results was evaluated using the Grades of Recommendation, Assessment, Development, and Evaluation (GRADE) guidelines. Ultimately, the quality of evidence for each outcome was rated as high, moderate, low, or very low.

The primary outcomes evaluated in this study were those directly related to postoperative liver synthetic function, including postoperative alanine transaminase (ALT), aspartate transaminase (AST), total bilirubin (TBIL), and albumin (ALB) levels. The secondary outcomes assessed included duration of surgery and Pringle, length of postoperative hospital stay, intraoperative blood loss and transfusion, indocyanine green (ICG) clearance, hepatocyte apoptosis index, postoperative complications, and others.

Statistical analysis was performed using the Review Manager 5.4 software. Continuous outcomes were reported as weight mean differences (WMD) with 95% confidence intervals (CI), and dichotomous outcomes were presented as odds ratios (OR) with 95% CI. In order to quantify inconsistencies of studies included in the meta-analysis, Cochran’s Q-test and I2 statistics were used. Low heterogeneity was considered when I2 ≤ 50%, and the fixed-effect model was used for analyses. Moderate heterogeneity was considered when I2 > 50% and high heterogeneity when I2 > 75%, and the random-effects model was used for analyses. Subgroups analyses or sensitivity analyses were then performed, and a descriptive analysis was conducted if a meta-analysis was inappropriate. Publication bias was assessed using funnel plots. Results were considered statistically significant when P < 0.05.