Comparing outcomes of donation after cardiac death versus donation after brain death in liver transplant recipients with hepatitis C : A systematic review and meta-analysis

1Department of Medicine; 2Department of Surgery, Schulich School of Medicine and Dentistry, Western University, London; 3Hotel Dieu Grace Hospital, Windsor, Ontario Correspondence: Dr Natasha Chandok, Department of Medicine, Schulich School of Medicine and Dentistry, Western University, 339 Windermere Road, PO Box 5339, London, Ontario N6A 5A5. Telephone 519-663-3002, fax 519-663-3858, e-mail nchandok@uwo.ca Received for publication October 3, 2013. Accepted October 13, 2013 Liver transplantation (LT) is a life-saving modality for treating well-selected patients with acute liver failure, end-stage liver disease, certain metabolic disorders and early hepatocellular carcinoma. The current practice of LT is limited by the significant disparity between organ availability and the number of patients awaiting transplantation. Donation after cardiac death (DCD) has become a significant source of transplantable organs in an attempt to expand the donor pool and increase organ supply (1-4). While DCD allografts have the potential to help address the disparity between organ availability and the number of patients awaiting LT, their use has been associated with higher rates of graft failure and biliary complication, particularly ischemic cholangiopathy, compared with donation after brain death (DBD) allografts (5,6). Hepatitis C virus (HCV) infection is currently a leading indication for LT, constituting approximately 30% to 50% of all transplants (7-10). Previous studies have suggested that HCV-positive recipients review

of DBD LT have worse outcomes than HCV-negative recipients, largely due to a more rapid and severe manner of recurrence of their HCV-related liver disease (9,11).Recurrence of HCV after LT is universal, with 20% to 40% of patients progressing to cirrhosis within five years of LT (12).Previous studies have shown that liver allografts from extended-criteria donors, such as those with advanced age, are at an increased risk of earlier and more severe HCV recurrence; this has deleterious impact on both patient and graft outcomes (13)(14)(15)(16).It has also been suggested that organ cold/warm ischemia is a risk factor for increased severity of recurrence of HCV after LT.DCD organs experience warm ischemic injury not characteristic of DBD donors because of hypoperfusion and hypoxia during the agonal period of time of withdrawal of life support (11).It has, therefore, been theorized that HCV patients receiving DCD allografts may be at an increased risk for graft injury and accelerated HCV recurrence.Despite this theoretical risk, the literature investigating the outcomes of HCV-positive patients receiving DCD allografts is scant and conflicting reports have been published.
Tao et al (17) performed a retrospective matched control trial of 111 HCV-positive patients (37 receiving DCD LT and 74 matched controls receiving DBD LT).Although the two groups had similar donor and recipient characteristics, immunosuppression regimens, rates of acute cellular rejection and HCV profiles, the patients receiving DCD LT had a higher incidence of primary nonfunction (19% versus 3%; P=0.006) and significantly higher peak aspartate aminotransferase levels compared with DBD subjects.Although the survival rates were not significantly different, DCD LT recipients had lower one-and five-year survival rates (83% and 69% versus 84% and 78%, respectively; P=0.75) and graft survival rates (70% and 61% versus 82% and 74%, respectively; P=0.24).A total of 314 liver biopsies were performed; mixed modelling analysis showed that fibrosis progression rates were similar for the two groups (0.6 fibrosis units/year according to the Ishak modified staging system).The rates of severe HCV recurrence (retransplantation or death due to recurrent HCV and/or the development of stage 4/6 fibrosis or worse within two years) were not significantly different (three [8%] DCD patients versus 11 [15%] DBD patients; P=0.38).Cytomegalovirus infection (HR 7.9 [95% CI 2.1 to 28.9]; P=0.002) and acute cellular rejection (HR 6.2 [95% CI 2.0 to 19.7]; P=0.002) were the only independent risk factors for severe recurrence.
Taner et al (18) performed a retrospective analysis of 77 HCVpositive patients who received DCD liver grafts and 77 matched HCVpositive patients who received DBD liver grafts.There were no differences in one-, three-and five-year patient or graft survival rates among the groups.Multivariate analysis showed that the Model for End-Stage Liver Disease score (HR 1.037 [95% CI 1.006 to 1.069]; P=0.018]) and post-transplant cytomegalovirus infection (HR 3.367 [95% CI 1.493 to 7.593]; P=0.003) were significant factors for graft loss.A comparison of five-year protocol biopsy samples for fibrosis progression in HCV-positive patients post-transplant did not show a difference between DCD and DBD grafts.The authors concluded that DCD liver graft utilization does not cause untoward effects on disease progression or patient and graft survival compared with DBD liver grafts in HCV-positive patients.
Hernandez-Alejandro et al (19) performed a retrospective, matchedcontrol trial evaluating 17 recipients with HCV who received a DCD graft and a matched group of 42 HCV recipients transplanted with a DBD graft.They found a statistically significant decrease in graft survival in HCV-positive patients undergoing DCD transplant (73%) compared with DBD transplant (93%) (P=0.01).There was a statistically significant increase in HCV recurrence at three months (76% versus 16%; P=0.005) and severe HCV recurrence within the first year (47% versus 10%) in the DCD group (P=0.004).The authors concluded that HCV recurrence is more severe and progresses more rapidly in HCV recipients who receive DCD grafts compared with those who receive DBD grafts.DCD LT in HCV recipients is associated with a higher rate of graft failure compared with those who receive DBD grafts.
The objective of the present study was to perform a systematic review and meta-analysis of studies comparing clinical outcomes of DCD versus DBD orthotopic LT in patients with HCV.The primary outcomes of interest were patient survival rates, graft survival rates, recurrence of severe HCV, primary nonfunction, acute cellular rejection, biliary strictures (diffuse or localized, anastomotic or nonanastomotic), biliary leaks and vascular complications (hepatic artery stenosis, hepatic artery thrombosis, portal vein thrombosis).

Primary objectives
To compare one-year patient and graft survival rates in recipients transplanted for HCV with DCD versus DBD grafts; and to compare one-year patient and graft survival rates in recipients with versus without HCV undergoing DCD versus DBD LT.

secondary objectives
To determine whether DCD LT compared with DBD LT in HCVpositive patients increases rates of primary nonfunction, acute cellular rejection, biliary strictures (diffuse or localized, anastomotic or nonanastomotic), biliary leaks or vascular complications (hepatic artery stenosis, hepatic artery thrombosis portal vein thrombosis).

eligibility criteria
Inclusion criteria were studies that compared DCD versus DBD LT in patients with HCV, as well as DCD LT in patients with and without HCV; and studies that evaluated adult recipients (age ≥18 years) who underwent primary LT.To be included, studies had to include at least one of the prespecified outcomes.There was no limitation on randomized control trials and no restrictions on language.Results duplicated in multiple articles were included only once.

data items
The following items were abstracted from the articles: demographic data of the study population and comparison group including age and sex; DCD versus DBD liver donations; features of the study design including allocation concealment, blinding, intention-to-treat analysis, number of patients lost to follow-up, rate of premature termination and funding source; the outcome measures of patient survival rate and graft survival rate; rates of primary nonfunction, acute cellular rejection, biliary strictures (diffuse or localized, anastomotic or nonanastomotic), biliary leaks, ischemic cholangiography, vascular complications (hepatic artery stenosis, hepatic artery thrombosis, portal vein thrombosis), HCV recurrence and retransplantation/liverrelated death.

Risk of bias
The risk of bias on a study level was assessed by determining the adequacy of the method of randomization, allocation concealment, blinding of the trial participants, care providers and outcome assessors.Also assessed were whether the trial was terminated prematurely, whether the analysis was an intention-to-treat and the funding source.
The Grading of Recommendations, Assessment, Development and Evaluation approach (20) was used to characterize the risk of bias for each of the outcomes that had available data.

statistical analysis
The meta-analysis was performed using the Cochrane Collaboration and the Quality of Reporting of Meta-analyses (QUORUM) guidelines.Statistical analyses were performed using Review Manager 5 (www.cochrane.org/).The RR was used as a summary measure of efficacy for dichotomous data and the mean difference between groups for continuous data to summarize the outcomes for patients treated with duct-to-duct versus Roux-en-Y loop anastomosis.For all RRs and mean differences, a 95% CI was reported.All analyses were conducted on an intention-to-treat basis.

synthesis of results
Results were pooled using a Mantel-Haenszel random-effects model for dichotomous outcomes and mean difference for continuous outcomes.Statistical heterogeneity was evaluated using the I 2 statistic.An I 2 value of 0% to 25%, 25% to 50% and >50% were considered to be indicative of low, moderate and high heterogeneity, respectively.

Risk of publication bias
Funnel plots were used to assess the risk of publication bias across trials for all outcome measures.

ResulTs study selection
Fifty-eight citations were screened, of which 15 were selected for fulltext retrieval.Of these, three articles (17)(18)(19) fulfilled eligibility criteria and were, thus, selected (Figure 1).
There was no disagreement regarding eligibility of full-text articles (Cohen's kappa = 1.00), and consensus was reached among all authors on inclusion and exclusion of all articles.

study characteristics
A total of 324 study participants in three trials comparing DCD versus DBD liver transplantation in HCV-positive patients.Study characteristics are included in Table 1.

Risk of bias within trials
The included trials had a high risk of bias.All three trials were retrospective analyses performed at single centres.There was no blinding and concealment.

Risk of bias across trials
The funnel plots of the RR for all outcomes did not show evidence of publication bias.

Recipient survival
Compared with DBD, orthotopic LT with a DCD liver was not associated with a significantly decreased patient survival (three studies; risk ratio 0.89 [95% CI 0.37 to 2.11]; P=0.79;I 2 =51%) (Figure 2).Heterogeneity was potentially explained by differences in length of follow-up.The overall quality of evidence was low (20).
Graft survival DCD LT trended toward, but was not significantly associated with, a decrease in graft survival (n=2; RR 0.40 [95% CI 0.14 to 1.11]; P=0.08;I 2 =34%) (Figure 3).Heterogeneity was potentially explained by differences in length of follow-up.The overall quality of evidence was low (20).

Biliary complications
The risk of biliary leaks was not statistically significantly higher (n=2; risk ratio 2.22 [95% CI 0.42 to 11.88]; P=0.35;I 2 =35%) (Figure 4) in patients receiving a DCD versus a DBD LT.The level of heterogeneity was explained by length of follow-up, differences in the definition of biliary leak, and differences in postoperative imaging or investigations for leak.The overall quality of evidence was low (20).

Recurrence of HCv infection
Recurrence of HCV was not significantly different in patients receiving DCD versus DBD LT when patients from two studies were pooled, with a risk ratio of 2.74 (95% CI 0.36 to 20.92; P=0.33;I 2 =84%) (Figure 5).The overall quality of evidence was low (20).
disCussioN ANd CoNClusioNs summary of evidence Three trials evaluated DCD versus DBD LT in recipients with HCV infection.The use of DCD livers in HCV-positive recipients was associated with a significant increase in primary nonfunction, but no significant difference in biliary complications, graft survival rates and recipient mortality rates (Table 1).

limitations
The conclusions that can be drawn from the present systematic review were limited by the small numbers of patients and the retrospective nature of the trials included.Length of follow-up varied substantially among the studies.Techniques and experience also vary among institutions.In addition, there was a lack of histopathological correlation and generalizability of results, in addition to potential selection biases.There were significant differences in rates of HCV recurrence that could, at least partially, be explained by differences in rates of antiviral treatment within the first year post-LT.Despite these limitations, there is still value in the current metaanalysis because it is the first study of its kind and contributes to the knowledge base.Published experiences will likely never be prospective or randomized given the nature of transplantation, the complexities of organ allocation and important issues involving medical ethics.Therefore, high-grade evidence for this topic may never emerge.

implications for clinical practice
Although significant for limitations, our meta-analysis indicates there is a substantial increase in primary nonfunction and a trend toward a decrease in graft survival, but no significant difference in other important clinical outcomes between DCD and DBD allografts in HCVpositive LT recipients.
Controversy remains in the use of DCD LT in HCV-positive patients.Individually, DCD LT and the presence of HCV have negative impact on patient and graft survival (11).It remains unclear whether the combination of DCD LT and HCV synergistically confers a worse outcome.The two studies that demonstrated no significant difference in patient outcomes with use of DCD LT in HCV-positive patients (17,18) included younger patients than the study that found a significant decrease in graft survival in HCV-positive patients undergoing DCD versus DBD LT (19).
DCD grafts are not contraindicated in well-selected HCV recipients and this is potentially an underutilized method to expand the donor pool.Over the past 10 years, approximately 700 to 800 patients have died awaiting LT (www.unos.org);DCD LT may be one of many solutions requiring further investigation.implications for research DCD allografts have become a significant source of transplantable organs in an attempt to bridge the gap between supply and demand in LT.Our meta-analysis indicates an increased number of adverse events (namely primary nonfunction) with DCD allografts in HCV-positive patients and a trend toward decreased graft survival, but no significant decrease in patient survival.
Ideally, DBD allografts appear to be better suited to HCV-positive patients; however, organ availability necessitates the use of DCD allografts.The quality of evidence in the three included articles was low; a randomized control trial with protocolized liver biopsies would be ideal.However, this is not feasible because recipients could not be randomized because the process of listing recipients is complex and not amenable to a randomized control trial.In the absence of a trial, more observational, prospective and multicentre data are needed.disClosuRes: All authors declare no support from any organization for the submitted work; no financial relationships with any organizations that may have an interest in the submitted work in the previous three years; and no other relationships or activities that could appear to have influenced the submitted work.

Figure 1 )
Figure 1) Study selection.DBD Donation after brain death; DCD Donation after cardiac death; HCV Hepatitis C virus

AuTHoR CoNTRiBuTioNs:
All authors contributed to the concept and design, and provided intellectual content of critical importance to this work.Dr Wells, Ms Janik and Dr Chandok performed the electronic literature search and manual search for potential articles.Dr Wells and Dr Chandok retrieved citations, reviewed full text papers and abstracted the data.All authors participated in the analysis and interpretation.All authors approved the final version of this article.FuNdiNG: Supported by Western University Department of Medicine Program of Experimental Medicine (POEM) Research Award.

TaBLe 1 Characteristics of included studies Hernandez-alejandro et al (13), 2011 Taner et al (18), 2011 Tao et al (17), 2010
COD Cause of death; CVA Cerebral vascular accident; DBD Donation after brain death; DCD Donation after cardiac death; MELD Model for End-stage Liver Disease; N/A Not available