Normothermic Machine Perfusion versus Cold Storage of Liver in Pig Model: A Meta-Analysis

Background

Liver transplantation (LT) has become the main treatment for end-stage liver disease. Advances in surgical technology, improvements in perioperative management, and the application of immunosuppressive agents have greatly improved the success rate and long-term survival rate of LT. However, the success of LT has created a new dilemma, as the supply of organs is insufficient to meet the demand. Hence, marginal donor livers, affected by severe fatty degeneration, older donors, prolonged hospitalization, and donation after circulatory death [1], are being considered to compensate for the gap between the supply and demand of livers for transplantation. However, grafts of suboptimal standard increase the risk of ischemic reperfusion injury (IRI) and post-transplant organ failure [2].

The reduction of IRI is currently one of the biggest issues in transplantation. In the past few decades, different measures have been taken to reduce IRI and improve the preservation quality of the donor liver. At present, the standard organ preservation method is static cold storage (CS). The low temperature (0–4°C) environment reduces the metabolism of organs [3], but cannot completely terminate metabolism, and so the lack of metabolic substrates and the accumulation of metabolites will still cause damage to the donor liver [4,5]. Therefore, CS cannot ensure the quality of the donor liver after preservation of marginal donor livers.

Normothermic mechanical perfusion (NMP) does not use low temperature to reduce the rate of organ metabolism, and instead aims to provide an approximately normal physiological environment for organs ex vivo [6]. NMP can continuously supply oxygen and metabolic substrates to the organ via a mechanical pump, and the perfusate can dilute the metabolites immediately instead of accumulating in the tissue. Thus, NMP can theoretically avoid cold ischemia injury during preservation and reduce IRI.

Although NMP has been used in clinical trials, the outcomes of NMP preservation remain unclear. One study verified the safety and feasibility of NMP in clinical application [7]. In addition, a retrospective control study showed that alanine aminotransferase (ALT) and aspartate aminotransferase (AST) levels at postoperative day 3 tended to be lower in the NMP group than the CS group [8], with no significant difference between NMP and CS in postoperative graft function, duration of ICU stay, and duration of postoperative hospitalization [8]. Another clinical study reported no significant difference in 30-day survival rate of NMP versus CS grafts, although the duration of ICU stay and duration of hospitalization were significantly longer in the NMP group than in the CS group [9]. However, the livers preserved by NMP in these clinical trials had been rejected clinically as suboptimal, and were not equal to livers preserved by CS. In contrast, animal studies investigating liver preservation via NMP versus CS have used equivalent quality livers for each preservation method. A number of animal experiments using NMP have been published, with the pig model being the closest to humans [10]. Hence, we collated these reports and carried out the present meta-analysis to quantitatively analyze the differences in the preservation of the liver using NMP versus CS; this will provide a theoretical basis for the selection of suitable organ preservation methods in clinical practice.

Material and Methods

LITERATURE SEARCH:

We performed a complete literature search to find articles related to NMP of the liver in the PubMed, EMBASE, and Cochrane Library databases using the following search terms: “normothermic perfusion” or “normothermic preservation” and “liver” and “transplantation”. The retrieval date ended in June 2017 and was updated in October 2017. We excluded non-English literature, unpublished research, and conference summaries. There were no restrictions on the species in the retrieval process, but only the literature on pigs was included in the final literature review.

The literature was included according to the following criteria: 1) studies evaluating ex vivo NMP of the liver, 2) CS used as a control, 3) preservation effect was evaluated by LT or simulated transplantation. The perfusion temperature was 35–38°C [11], and there were no restrictions on the NMP perfusate or CS preservation solution. In included studies, the NMP group had the same preservation time as the CS group, and all conditions were basically the same except for the preservation method. Exclusion criteria included: non-randomized or non-controlled experiments from which the data could not be obtained from the literature. We used the ARRIVE (Animal Research: Reporting of In Vivo Experiments) guidelines [12] to evaluate the quality of the included studies. Two authors independently performed the database search and selected the literature for review according to the inclusion and exclusion criteria; any differences in opinion were resolved by discussion.

The following information was collected: name of the first author, publication date, sample size, preservation time, evaluation method, observation time, and outcome. If one outcome was reported by more than 4 articles, the data was extracted and merged for analysis. If the original text did not directly provide the data, the image data was extracted using the image software GetData Graph Digitizer, version 2.20 (http://getdata-graph-digitizer.com/).

STATISTICAL ANALYSIS:

Statistical analysis was carried out in accordance with the Cochrane Handbook for Systematic Reviews of Interventions, version 5.1.0 (http://training.cochrane.org/handbook). We used the standardized mean difference (SMD, Cohen’s method) to compare the results of the different studies using a common measure [13]. We pooled the overall effect using the random effects model and used the Q test and the I2 value to quantify the heterogeneity between studies. P values less than 0.5 and I2 values more than 50% were taken to indicate severe heterogeneity. We carried out subgroup analyses based on warm ischemia time (WIT), preservation time, and observation time. Sensitivity analysis was also conducted by excluding these studies from the analysis one-by-one to verify whether the same trend was observed with the remaining trials. STATA software version 12.0 (StataCorp, TX, USA) was used to analyze the data. P<0.05 was considered to indicate statistically significant differences.

Results

ELIGIBLE STUDIES:

The PRISMA flow diagram was used in the search process (Figure 1). According to the retrieval strategy, 320 articles were preliminarily retrieved (194 from PubMed and 126 from Embase), of which 82 were repeated. We excluded 212 articles based on the titles and abstracts, as they were reviews, letters, other organ studies, studies on mechanisms of IRI, or non-controlled studies. There were 17 studies that met the inclusion criteria, including 3 on rats, 12 on pigs, and 2 in humans. The studies by Boehnert et al. 2013 [14], Brockmann et al. 2009 [15], and Schon et al. 2001 [6] each contained several groups of experiments. After discussion amongst the authors, these groups of experiments within a single study that met the inclusion criteria were selected separately and regarded as different studies. Each group of experiments within a single study were differentiated by “a”, “b”, and “c”. A final total of 16 studies from 12 articles were included, all of which were studies conducted on pigs that were published from 2001–2016, all, with 95 cases in the experimental group and 83 in the control group (Table 1). Quality scoring was carried out according to the ARRIVE guidelines, with an average of 14.9±1.5 (range: 13 to 18) (Table 2).

HEPATOCELLULAR INJURY:

The determination of liver enzyme concentrations can reflect the degree of liver damage. Nine studies reported the level of ALT, 8 of which examined the concentration of the perfusate [16–19,21–24], and one study tested the serum concentration after orthotopic liver transplantation (OLT) [14]; the SMD and 95% CI were −3.83 (−4.88, −2.79), indicating that the ALT level of the NMP preservation group was significantly lower than that of the CS preservation group (Figure 2). Ten studies reported the level of AST, 6 study examined the concentration of the perfusate [16–19,21,22], and the 4 study tested the serum concentration after OLT [6,14,20]; the SMD and 95% CI were −3.31 (−4.93, −1.68), indicating that the AST level of the NMP preservation group was significantly lower than that of the CS preservation group (Figure 3).

Subgroup analysis of ALT (Table 3) showed that regarding the WIT and observation time, most of the 95% CI between the subgroups was overlapped, suggesting that there was no significant difference between the subgroups. According to the preservation time, the 95% CI overlap between the 1–4 hour group and the 10 hour group showed no significant difference. However, compared with the 1–4 hour and the 10 hour groups, only a small part of the 95% CI of the 24 hour group overlapped; this difference was statistically significant, which indicates that the effect of NMP versus CS on the ALT level becomes more obvious with the prolongation of the preservation time. However, this trend was not seen in the subgroup analysis of AST (Table 4).

SINUSOIDAL ENDOTHELIAL CELL INJURY:

Hyaluronic acid (HA) is a matrix polysaccharide enzyme that is known to be associated with inflammation [25] and has been used as a marker of sinusoidal endothelial cell (SEC) damage [21,26]. Five studies reported the level of HA, 3 study examined the concentration of the perfusate [16,21,22], and the 2 study tested the serum concentration after OLT [6]; the SMD and 95% CI were −3.25 (−4.43, −2.06), indicating that the HA level of the NMP group was significantly lower than that of the CS group (Figure 4).

LIVER SYNTHETIC FUNCTION:

The amount of bile production is an important index for evaluating the function of liver synthesis. Ten studies reported the level of bile production during the evaluation period [14,16–19,21–24]; the SMD and 95% CI were 3.96 (2.09, 5.84), indicating that the amount of bile production in the NMP preservation group was significantly greater than that in the CS preservation group (Figure 5). Subgroup analysis showed that there was no significant difference in the effect of WIT and preservation time on bile production, but with the prolongation of observation time, the difference between the NMP group and the CS group tended to become more obvious (Table 5).

SURVIVAL RATE:

Seven studies performed orthotopic LT after liver preservation using NMP or CS [6,14,15,20], and 6 of these studies reported the survival rate after transplantation [6,15,20]. The RR and 95% CI were 2.02 (0.79, 5.15), indicating that there was no significant difference in the survival rate between the NMP and CS groups (Figure 6).

Discussion

HETEROGENEITY:

In the present meta-analysis, there was a high degree of heterogeneity for the analyses of AST level, bile production, and survival rate. Subgroup analyses were performed according to WIT, preservation time, and observation time because of the different experimental designs. Subgroup analysis showed that heterogeneity could not be eliminated by grouping, so we speculate that these 3 factors were not the main source of heterogeneity.

Because of the small number of included articles, we did not detect any publication bias. The studies included in the present meta-analysis were all published in English, and so there may have been a certain amount of language bias. In addition, unpublished literature was not included; this unpublished literature may contain negative results, as studies with positive results may be easier to publish. However, not all of the included studies had positive results. Some studies reported that there was no significant difference between the 2 preservation methods regarding the level of AST [6,20] and bile production [14]. Moreover, there was no significant difference between the 2 preservation methods regarding the overall postoperative survival rate.

There was also an inevitable degree of experimental heterogeneity. Although we strictly adhered to the inclusion criteria, the working principle and perfusion parameters of the NMP system used in the various studies were similar, but the equipment and operation methods used were not exactly the same. We consider that this was the main source of heterogeneity. However, the sensitivity analysis in which the included studies were eliminated one-by-one showed that the final results did not change significantly. Hence, we consider that the results of the present meta-analysis are reliable.

LIMITATIONS:

Our study has some limitations. First, although the NMP system is very expensive, there is no research into the cost of the NMP system. If a comparison of NMP and CS was carried out in combination with health economics, the results may have been different. Second, a relatively small number of studies were included in the present meta-analysis and the sample size was small. Third, animal experiments currently only focus on the short-term effects of the model. Postoperative long-term complications, such as biliary stricture formation, cannot be evaluated after only 5–7 days of observation, and both clinical and animal experiments have shown that biliary stricture formation is related to prolonged preservation time and WIT. Therefore, future studies with large sample sizes are needed to evaluate the long-term benefits of NMP.

Conclusions

Compared with CS, preservation of liver via NMP can protect hepatocytes and SECs, reduce the levels of ALT, AST, and HA, significantly increase bile production, and cause no change in the 5–7 days natural survival rate in the pig model. The NMP system is a novel organ preservation method that requires further investigation in clinical practice, including cost-effectiveness evaluation.