The authors have no conflict of interest to declare.
Increasing the Donor Pool: Consideration of Prehospital Cardiac Arrest in Controlled Donation After Circulatory Death for Liver Transplantation
Article first published online: 20 DEC 2013
© 2013 American Association for the Study of Liver Diseases
Volume 20, Issue 1, pages 63–71, January 2014
How to Cite
Elaffandi, A. H., Bonney, G. K., Gunson, B., Scalera, I., Mergental, H., Isaac, J. R., Bramhall, S. R., Mirza, D. F., P.R. Perera, M. T. and Muiesan, P. (2014), Increasing the Donor Pool: Consideration of Prehospital Cardiac Arrest in Controlled Donation After Circulatory Death for Liver Transplantation. Liver Transpl, 20: 63–71. doi: 10.1002/lt.23772
- Issue published online: 20 DEC 2013
- Article first published online: 20 DEC 2013
- Accepted manuscript online: 18 OCT 2013 11:35AM EST
- Manuscript Accepted: 1 OCT 2013
- Manuscript Received: 4 JUL 2013
Donor warm ischemia has implications for outcomes after liver transplantation (LT) using organs from donation after circulatory death (DCD) donors. Prehospital cardiac arrest (PHCA) before donation may generate a further ischemic insult. The aim of this single-center study of 108 consecutive DCD LT procedures was to compare the outcomes of PHCA and non-PHCA cohorts. A review of a prospectively collected database of all DCD grafts transplanted between January 2007 and October 2011 was undertaken to identify donors who had sustained PHCA. The unit policy was to consider such donors when transaminase levels were ≤4 times the normal range and had an improving trend. Twenty-six of the 108 DCD transplants were from DCD donors with PHCA, and 82 were in the non-PHCA cohort. A comparative analysis of the PHCA and non-PHCA cohorts showed better short-term results (a low incidence of acute kidney injury) for the PHCA group but satisfactory long-term results for both groups with no significant differences in graft or patient survival between them. In conclusion, a careful donor selection policy for including PHCA DCD donors with normalized liver function tests or transaminase levels ≤ 4 times the norm resulted in successful transplantation and could boost the donor pool with no adverse outcomes. Liver Transpl 20:63-71, 2014. © 2013 AASLD.
acute kidney injury
body mass index
cold ischemia time
donation after brain death
donation after circulatory death
donor warm ischemia time
ischemic-type biliary lesion
intensive treatment unit
liver function test
Model for End-Stage Liver Disease
prehospital cardiac arrest
UK Model for End-Stage Liver Disease
The United Kingdom has seen a sustained rise in the number of patients on the liver transplantation (LT) waiting list. This has been met in recent years with the increased use of donation after circulatory death (DCD) grafts. According to the currently accepted guidelines recommended by the Institute of Medicine and the British Transplantation Society, organ donation can proceed after at least 5 minutes of asystole following the withdrawal of treatment. The warm ischemia insult linked to DCD donors (Maastricht category III) is associated with specific early and late posttransplant complications in comparison with donation after brain death (DBD) grafts.
Both hemodynamic instability and hypoxemia in DCD organs contribute to the donor warm ischemia time (DWIT), which may result in an increased rate of complications after DCD LT. The careful selection of DCD donor and recipient combinations is crucial for maintaining an overall benefit and excellent results in comparison with DBD grafts after LT. Currently, a DWIT greater than 30 minutes is considered a relative contraindication to transplantation.
However, even before donation, a proportion of DCD grafts will have sustained an initial ischemic insult[8-10] because of prehospital cardiac arrest (PHCA). The duration of PHCA may have profound implications for the potential quality of a DCD organ, and it generates an additional ischemic insult to the liver graft. At several centers, PHCA is considered a relative contraindication for DCD liver donation and transplantation: 46% of the PHCA DCD grafts accepted by us were rejected by other transplant units because of PHCA before they were used by our unit. To date, the issue of PHCA in DCD donors and its implications for LT outcomes have not yet been reported. The aim of this study was to describe the process of donor selection, including donors with abnormal liver function tests (LFTs), at a single, large-volume center in the United Kingdom and the outcomes for recipients of grafts from DCD donors who had experienced PHCA; this may have a positive influence on the DCD donation pool.
PATIENTS AND METHODS
Using a prospectively maintained database from the Liver Unit of Queen Elizabeth Hospital (Birmingham, United Kingdom), we undertook a retrospective analysis of all sequential DCD LT procedures performed between January 1, 2007 and October 1, 2011. The study protocol received a priori approval by the institutional review committee. Additional data were obtained from National Health Service Blood and Transplant. The DCD cohort was divided into 2 groups: those who had sustained PHCA and those who had not.
The analyzed donor details included the following: demographic criteria, cause of death, history related to the occurrence and length of PHCA, predonation biochemical results, body mass index (BMI), cold ischemia time (CIT), and DWIT. Downtime was defined as the time between witnessed PHCA and the reported start of cardiopulmonary resuscitation. The analyzed recipient variables included demographic criteria, indications for transplantation, and Model for End-Stage Liver Disease (MELD) and UK Model for End-Stage Liver Disease (UKELD) scores.
Postoperative LFTs and coagulation studies were used as markers of early graft function. Primary nonfunction (PNF) was defined as primary graft failure within 7 days of transplantation requiring retransplantation or leading to patient death. Delayed graft function was defined as primary graft failure characterized by cholestasis within 6 months of transplantation in the absence of hepatic artery thrombosis or biliary complications also leading to retransplantation or patient death.
The incidence of posttransplant acute kidney injury (AKI) and the need for dialysis in the first 5 days were also measured as surrogate markers of delayed graft function because of the recent evidence for a significant increase in AKI with DCD LT versus DBD LT, which is related to the severity of the ischemia/reperfusion injury and peak transaminase levels. In accordance with the Acute Dialysis Quality Initiative Group, AKI was defined as an increase in the serum creatinine level greater than 2 times the baseline serum creatinine level.
The analyzed outcomes included early graft function, AKI, posttransplant dialysis requirements, biliary complications, and long-term patient and graft survival.
Donor and Recipient Considerations
The consideration of a DCD liver donor by the transplant surgeon was based on several factors, including age, BMI, cause of death, organ support, past medical history, and blood results. For donors who had sustained PHCA, donation was considered if the peak transaminase levels were no greater than 4 times the norm and the trend of liver function had been improving in the previous 6 to 12 hours. Following what was considered a relatively safe upper limit for a viable uncontrolled DCD graft after normothermic regional perfusion, we set this restriction on transaminase levels.
At no point was the transplant team involved in the withdrawal process before donation. Circulatory death was determined by a sustained asystole, and an interval of at least 5 minutes was required before the certification of death by the attending physician in accordance with national guidelines. Generally, donors who did not proceed to asystole within an hour of treatment withdrawal or who sustained a DWIT greater than 30 minutes were not considered further for liver donation. The procurement was performed with a modification of the super-rapid technique previously described.[14, 15] DWIT was defined as the interval between hypotension with a systolic blood pressure < 50 mm Hg and/or desaturation < 80% (whichever came first) and the initiation of aortic perfusion. CIT was defined as the time between aortic perfusion at organ retrieval and the start of implantation (when the liver was taken out of the ice).
In general, recipients expected to have a difficult explant (ie, a previous history of upper abdominal surgery, preoperative portal vein thrombosis, or retransplantation) were excluded from receiving a DCD liver graft in an attempt to minimize CIT, whereas those with less severe portal hypertension and hepatocellular carcinoma were more likely to receive DCD grafts.
An analysis of peak liver function and coagulation studies during the first posttransplant week was performed for PHCA recipients; the data were compared with those for the non-PHCA group. Also, the incidence of AKI and posttransplant dialysis in the first 5 days was compared between the 2 groups. The incidence of biliary complications was also compared between the groups. All patients were followed up at the posttransplant clinic on a regular basis. Liver ultrasonography was used as the first imaging technique in case of cholestatic liver function. Patients with abnormal ultrasonography findings were further investigated with magnetic resonance cholangiopancreatography accordingly. Biliary complications were classified as anastomotic strictures, nonanastomotic strictures, or ischemic-type biliary lesions (ITBLs; defined as intrahepatic lesions in the absence of arterial complications). Overall patient survival and graft survival as well as causes of death and graft loss were described. Graft survival was timed from the transplant date to the date of retransplantation or death (whichever came first), and it was censored for the date of the end of the study period.
Categorical variables were examined with a chi-square test. A Shapiro-Wilk test for normality was performed for continuous variables, and subsequently, a Mann-Whitney U test and a Student t test were performed for nonparametric and parametric data, respectively. Graft survival and patient survival between the groups were compared with Kaplan-Meier plots and log-rank tests. SPSS 17.0 (SPSS, Chicago, IL) was used for the statistical analysis. A P value < 0.05 was considered significant.
Patient and Donor Demographics
Between January 1, 2007 and October 1, 2011, livers from 95 potential DCD donors who had sustained PHCA were offered to our unit. Twenty-six of these liver offers were accepted and subsequently transplanted; the details are described later. Twelve of the 26 PHCA donors (46%) were rejected by other units before being offered to our center. For the 69 remaining potential donors, the median PHCA downtime was 42 minutes (range = 30-60 minutes) with a median intensive treatment unit (ITU) stay of 1.5 days (range = 0-5 days) before the offer. On these grounds, 13 of the remaining 69 offers were declined solely because of PHCA. The remaining 56 offers were declined because of the association of PHCA and other risk factors, so 1 offer could have been rejected because of multiple risk factors. Such factors included logistics (n = 8), past medical history (n = 17), age (n = 13), alcohol consumption (n = 5), BMI (n = 7), no suitable recipients (n = 3), abnormal LFTs (n = 14), and others [small donor size (n = 2), virology (n = 1), high lactate level (n = 1), unstable donor (n = 1), and donor sepsis (n = 2)].
In the same study period, DCD LT was performed 108 times at Queen Elizabeth Hospital. The median age of the DCD donors was 50 years (range = 11-72 years), and 52% of the donors were males. The mean DWIT and CIT values were 20 minutes (7-32 minutes, standard deviation = 8.4 minutes) and 445 minutes (184-709 minutes, standard deviation = 106 minutes), respectively. Recipients of DCD grafts had a median age of 56.1 years (range = 21-70 years) and mean MELD and UKELD scores at transplantation of 14 (range = 6-27) and 53 (range = 39-75), respectively.
Twenty-six of the potential DCD donors had previously sustained PHCA with a median downtime of 20 minutes (range = 1-50 minutes) and a median ITU stay of 2 days (range = 1-14 days) before organ procurement. At the time of donation, the mean alanine aminotransferase (ALT) level for DCD donors was 83 U/L. In a subgroup analysis, we divided the PHCA cohort into 2 groups: those who received grafts from donors with an ALT level ≥ 100 U/L (n = 19) and those who received grafts from donors with an ALT level < 100 U/L (n = 7). We found that a donor ALT level ≥ 100 U/L did not significantly affect recipient survival (P = 0.07). The recipients' aspartate aminotransferase (AST) levels on days 0 and 1 were 1002 U/L (range = 35-11,850 U/L) and 1528 U/L (range = 535-11,037 U/L), respectively. The peak LFTs after transplantation were as follows: AST, 2460 U/L (range = 694-11,850 U/L); alkaline phosphatase (ALP), 1188 U/L (range = 303-3128 U/L); bilirubin, 77.5 μmol/L (range = 23-757 μmol/L); and international normalized ratio, 1.8 (range = 1.3-5.9). In this PHCA cohort, the recipients spent an average of 3 days in the ITU after transplantation (range = 1-16 days).
Comparative Analysis of the PHCA and Non-PHCA Cohorts
|Variable||PHCA Group (n = 26)||Non-PHCA Group (n = 82)||P Value|
|Age (years-median value)||50.5||50||0.89|
|Sex: male (n)||13||43||0.81|
|Cause of death (n)|
|ITU stay (days-mean value)||1.4||1.9||0.46|
|DWIT (minutes-mean value)||19.2||20.2||0.6|
|CIT (minutes-mean value)||431.8||449.8||0.5|
|Variable||PHCA Group (n = 26)||Non-PHCA Group (n = 82)||P Value|
|Sex: male (n)||17||46||0.40|
|Age (years-median value)||54.8||57||0.94|
There were no differences in age, sex, ITU stay, or blood results before donation [AST, ALP, gamma-glutamyltransferase (GGT), sodium, and bilirubin] between the 2 groups (Table 1). Apart from the donor ALT serum levels, which were significantly higher among PHCA donors (P = 0.001), the predonation blood tests were not significantly different. It was also observed that CIT (431.8 minutes for the PHCA group versus 449.8 minutes for the non-PHCA group) and DWIT (19.2 minutes for the PHCA group versus 20.2 minutes for the non-PHCA group) had a tendency to be longer for the non-PHCA group, although this did not reach statistical significance (P = 0.5 and P = 0.6, respectively). Unsurprisingly, in the group that had sustained PHCA before donation, the proportion of donors who had anoxic brain injury was larger than that in the non-PHCA group (23% versus 1%), with trauma and stroke being the most common causes of death in the non-PHCA group (24% and 66%, respectively; P < 0.001).
Recipient variables were similarly subcategorized into the PHCA and non-PHCA groups (Table 2). There were no differences in the recipient ages or MELD/UKELD scores between the 2 groups. The difference in the indications for transplantation was not significant, with most of the patients undergoing transplantation for hepatocellular carcinoma or end-stage liver disease secondary to alcohol, cholestatic diseases, or viral hepatitis.
The peak LFT values on days 1 and 7 were compared for all recipients in the PHCA and non-PHCA groups. The peak LFT values included AST levels of 3576 and 3039 U/L, serum bilirubin levels of 149 and 93 μmol/L, and international normalized ratios of 2.05 and 1.79 for the PHCA and non-PHCA groups, respectively. Day 1 values included AST levels of 1694 and 2118 U/L and ALP levels of 163 and 201 U/L for the PHCA and non-PHCA groups, respectively. Finally, day 7 values included AST levels of 152 and 257 U/L, ALP values of 677 and 777 U/L, serum bilirubin levels of 49 and 64 μmol/L, and international normalized ratios of 1.05 and 1.15 for the PHCA and non-PHCA groups, respectively. The differences between the 2 groups were not significant for any of the studied values. Although the requirement for hemofiltration in the first 5 days was not statistically different between the 2 cohorts (10 recorded for the non-PHCA group versus none recorded for the PHCA group, P = 0.06), there was a greater trend of AKI in the non-PHCA group (38 recorded for the non-PHCA group versus 5 recorded for the PHCA group, P = 0.01).
With a median follow-up of 18 months (range = 1-59 months), the overall 30-day, 1-year and 2-year graft survival rates for all DCD recipients were 89%, 84%, and 79%, respectively. There was no significant difference in graft survival between the PHCA and non-PHCA cohorts (P = 0.44; Fig. 1A). Causes of graft loss included PNF (n = 1), portal vein thrombosis (n = 1), hepatic artery thrombosis (n = 3), and ITBL (n = 1). There were 10 biliary complications (6 anastomotic strictures, 3 ITBLs, and 1 nonanastomotic stricture) in the DCD cohort (Table 3), and all occurred in the non-PHCA group (P = 0.06). Anastomotic strictures (n = 6) were managed conservatively (3/6), with endoscopic retrograde cholangiopancreatography stenting (1/6), or with surgical biliary reconstruction (2/6). The nonanastomotic stricture (n = 1) was managed conservatively. The remaining 3 patients who developed ITBLs were treated either conservatively (2/3) or with percutaneous transhepatic cholangiography stenting (1/3). One of these patients developed multiple hepatic abscesses and died while awaiting retransplantation.
|Anastomotic stricture (n = 6)||Conservative management (3)|
|Endoscopic retrograde cholangiopancreatography stenting (1)|
|Surgical bypass (2)|
|ITBLs (n = 3)||Conservative management (2)|
|Percutaneous transhepatic cholangiography stenting (1)|
|Nonanastomotic stricture (n = 1)||Conservative management (1)|
In the same period of follow-up, the overall 30-day, 1-year, and 2-year patient survival rates for all DCD recipients were 92%, 86%, and 80%, respectively. Again, there was no significant difference in patient survival between the 2 groups (P = 0.63; Fig. 1B). Overall, there were 19 postoperative deaths in this cohort. Three of these resulted from graft loss [hepatic artery thrombosis (1), ITBL (1), and PNF (1)]. Notably, there were no significant differences in patient or graft survival between PHCA and non-PHCA recipients (Fig. 1).
Because of the insufficient number of cadaveric organs available for transplantation in the United Kingdom, DCD grafts are increasingly being used. Donor factors are carefully considered before transplantation because the long-term risks associated with these grafts have implications for the overall benefits of transplantation. Factors such as donor age, BMI, steatosis, and warm ischemia time have recently been studied in extended criteria DCD donors. PHCA has been considered a relative contraindication for the use of these grafts by some transplant centers because of the implications of the additional ischemic insult sustained by DCD livers, which are already penalized by DWIT. This is the first analysis to consider the effects of PHCA on the outcomes of the transplantation of DCD grafts.
In this article, we describe sequential DCD transplants at a single, large-volume center in the United Kingdom. The initial consideration of DCD donors is made by the transplant surgeon on the basis of donor demographics such as age, BMI, cause of death, organ support, past medical history, and blood results. It is the unit's policy to consider a DCD donor with PHCA if the downtime has been accurately recorded and if there is a trend toward an improvement of LFTs in the hours preceding donation. As we expected, predonation ALT values were significantly higher in the PHCA cohort (P = 0.001). Hence, a further subgroup analysis within the PHCA cohort was conducted, and it showed that an ALT level > 100 U/L did not negatively affect recipient survival (P = 0.07). The overall cutoff for transaminases that we considered was no greater than 4 times the normal level. This was in keeping with a recent report demonstrating that select grafts from donors with high serum transaminase levels could be used for transplantation with satisfactory results.
The donor risk index proposed by Feng et al. shows that organs from donors who have suffered trauma have better long-term outcomes after transplantation than organs from donors who have suffered anoxic brain injury. This may be due to the fact that donors who have sustained trauma are younger or donors who have suffered anoxia are more frequently associated with hypotension and hypoxemia, which may negatively affect long-term outcomes. Although the PHCA cohort in our study included significantly more donors who had suffered anoxia, the long-term outcomes were not significantly different. This may be due to the fact that the donor risk index was not specifically designed for a DCD setting. Thus, the effect of a donor's cause of death on long-term graft survival after DCD is as yet undetermined.
Potentially, PHCA may add significant additional risk to the existing risks of DWIT and CIT in the setting of DCD LT. Adverse outcomes of DCD transplantation, including high rates of PNF and ITBLs, have been linked with the duration of DWIT and CIT.[20-22] Following large-scale studies on the outcomes of DCD LT based on the Organ Procurement and Transplantation Network and United Network for Organ Sharing databases, the American Society of Transplant Surgeons[23-25] issued recommendations to select DCD liver grafts with a DWIT less than 30 minutes and to aim for a CIT less than 8 hours. Efforts to minimize CIT include efficient organization, careful recipient selection, and expedited implantation. The mean DWIT and CIT values in this study were similar for the PHCA and non-PHCA groups and were within the suggested national guidelines. These findings may be associated with our particularly strict selection of donors with PHCA, which is aimed at reducing the already greater risks for the recipient of a DCD liver.
We compared both the short- and long-term outcomes of PHCA and non-PHCA cohorts. AKI, the need for dialysis, and acute liver injury based on peak serum transaminase levels are often used as surrogate markers of early outcomes of DCD LT.[9, 26] Surprisingly, the short-term outcomes were better for the PHCA cohort because AKI was noted significantly more frequently in the non-PHCA cohort. The 1-year patient and graft survival rates of 86% and 84% were consistent with most series and better than others after DCD LT.[8, 28] Although these findings may show better short-term outcomes and satisfactory long-term outcomes for the PHCA cohort, they may in fact reflect a very conservative and selective policy for PHCA donors with limited additional risk factors.
The selection criteria for controlled DCD liver donors have been more restrictive because of the greater potential risk of PNF and biliary complications. When marginal DCD grafts are used, it is logical to argue against compounding multiple risk factors (eg, long DWIT, CIT, advanced age, obesity, long ITU stay, and absence of PHCA), which may contribute to the increased susceptibility of these grafts to AKI, PNF, and biliary complications. In this study, we have shown that our overall rate of PNF (1%) is in keeping with others' rates.[8, 20] It is also possible to hypothesize that a mechanism of ischemic preconditioning may have positively affected the initial graft function by maintaining the hepatic microcirculation and decreasing Kupffer cell activation for clinically relevant ischemic periods. The main mechanism of its protection is probably the release of nitric oxide, by which it prevents sinusoidal perfusion failure and leukocyte adherence. In addition, the short interval of ischemia during ischemic preconditioning generates mild oxidative stress, which then induces natural defense mechanisms against subsequently lethal injury.
Recipient selection is a difficult task, with the aim being to achieve the maximum benefit from this additional organ resource. Preferably, organs from PHCA donors should be used for recipients capable of withstanding a period of initial dysfunction. However, decisions were made in a case-by-case fashion and depended on the calculated risks for each PHCA donor at the time of the transplant. Our results showed no difference between the recipients of the 2 cohorts.
Biliary complications after DCD transplantation continue to be a clinical issue. It is accepted that the initial injury to the bile duct is more likely to be ischemic in nature, and this makes it a further concern for patients who have sustained PHCA. The risk of biliary complications is high, and they have been reported in up to 50% of recipients of uncontrolled DCD livers versus 30% to 40% in a controlled DCD setting.[30, 31] In this study, the ITBL rate was relatively low in comparison with rates in other reports[32-34] and was surprisingly recorded only in the non-PHCA cohort (3.7%), whereas the PHCA group showed no incidence of ITBLs in the long-term follow-up (median = 18 months). It remains our unit protocol to closely monitor liver function in DCD recipients, and even mild cholestatic derangements are promptly investigated with ultrasound and magnetic resonance cholangiopancreatography.
In this study, there were 10 biliary complications in the non-PHCA cohort (12.2%); several studies have shown similar or higher rates of biliary complications (13.7%-33%) after DCD LT.[8, 30, 35] The average time to developing ITBLs is usually 90 to 120 days. The lack of biliary complications in the PHCA cohort is difficult to explain. We can only speculate that the superselection of PHCA donors may have produced generally better quality grafts in comparison with the grafts from non-PHCA donors. A bias due to the relatively small cohort may, however, best explain this finding.
We describe here the outcomes of DCD transplantation using grafts from donors who had suffered PHCA. PHCA has historically been a relative contraindication for a consideration of DCD donation. We found no difference in overall patient or graft survival, PNF, or biliary complications in comparison with a non-PHCA cohort. Carefully selecting PHCA DCD donors, inquiring in detail about the circumstances of PHCA, assessing the trend of serum transaminases as a marker of cellular damage, and excluding donors with multiple risk factors for donation have produced outcomes comparable to those with non-PHCA DCD liver grafts. The significance of such a comparative study is limited only by the prescreening pattern used to select PHCA liver grafts suitable for transplantation. Such practices are not uncommon with DCD liver grafts because without a tight selection process, the possibilities of PNF or early graft failure would make a randomized trial unacceptably risky. Although this is a retrospective study with a preliminary analysis of a relatively small cohort, our findings show that PHCA DCD donors should not be excluded from liver donation and, if they are adequately selected, can provide additional liver grafts with satisfactory early and late outcomes. DBD grafts are well known for their superior quality in comparison with DCD grafts, so such a conservative and highly selective policy for PHCA DCDs may not only positively influence the DCD donation pool in the United Kingdom but also have a similar effect on DBD donation. Moreover, National Health Service Blood and Transplant has limited the potential erosion of DBD donors into the DCD pool, and this resulted in an increase in net DBD liver donation in 2012.
- 2Institute of Medicine. Non-Heart-Beating Organ Transplantation: Practice and Protocols. Washington, DC: National Academy Press; 2000.
- 3Guidelines Relating to Solid Organ Transplants From Non-Heart Beating Donors. London, United Kingdom: British Transplantation Society; 2004.
- 12for Acute Dialysis Quality Initiative Workgroup. Acute renal failure—definition, outcome measures, animal models, fluid therapy and information technology needs: the Second International Consensus Conference of the Acute Dialysis Quality Initiative (ADQI) Group. Crit Care 2004;8:R204-R212., , , , ;