Impact of donor warm ischemia time on outcomes after donation after cardiac death liver transplantation


  • David P. Foley

    Corresponding author
    1. Department of Surgery, University of Wisconsin School of Medicine and Public Health, Madison, WI
    • Address reprint requests to David P. Foley, M.D., Department of Surgery, University of Wisconsin School of Medicine and Public Health, 600 Highland Avenue, CSC H4/780, Madison, WI 53717. Telephone: 608-263-9903; FAX: 608-263-9903; E-mail:

    Search for more papers by this author

  • See Article on Page 165

  • The project described here was supported by the Clinical and Translational Science Award program through the National Institutes of Health National Center for Advancing Translational Sciences (grant UL1TR000427). The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.


area under the systolic blood pressure curve


cold ischemia time


donation after brain death


donation after cardiac death


donor warm ischemia time


median value of the slope of the systolic blood pressure based on values during the first 10 minutes


systolic blood pressure


slope of the systolic blood pressure based on values during the first 10 minutes

Livers that are recovered from donation after cardiac death (DCD) donors continue to be a valuable source of allografts for patients on the liver transplant waiting list. Although most national and single-center studies have reported inferior transplant outcomes with livers recovered from DCD donors versus those recovered from donation after brain death (DBD) donors,[1, 2] there are more recent single-center reports demonstrating similar outcomes for DCD and DBD liver transplants.[3, 4] As national and international experience grows, it is becoming clearer that the careful selection of DCD donor livers for transplantation can lead to excellent patient and graft survival rates as well as low rates of biliary complications. The challenge for all transplant surgeons and centers is identifying which DCD donor livers are suitable for transplantation.

Several studies have investigated risk factors that predict either allograft loss or biliary complications after DCD liver transplantation. Large database analyses have revealed that the selection of donor livers with a short donor warm ischemia time (DWIT), a short cold ischemia time (CIT) and a younger donor age can result in graft survival rates similar to those associated with DBD liver transplantation.[1, 5] A more recent single-center analysis suggests that a prolonged time interval between asystole and aortic cross-clamping in the donor significantly increases the risk of graft loss.[6] Other single-center studies have determined that an older donor age and prolonged CIT[7] or a prolonged total ischemia time, increased donor weight, and older donor age[8] increase the risk of ischemic cholangiopathy. In these single-center studies, the duration of DWIT was not an independent predictor of either graft loss or ischemic cholangiopathy.

The unpredictability of DWIT as a risk factor for graft failure or ischemic cholangiopathy may be due to the insufficient power of single-center studies, a selection bias from preferentially selecting donor livers with a short DWIT, or variability in the definition of DWIT. DWIT has been routinely defined as the time interval between the withdrawal of support and organ flushing in the donor. However, because a donor may maintain normal blood pressure and oxygen saturation for a period of time after withdrawal and before physiological deterioration, many believe that the total duration of DWIT is an inaccurate assessment of true end organ warm ischemia and poor allograft outcomes. If true hepatic ischemia is occurring for a shorter time period within the DWIT interval, then identifying a more specific interval may lead to lower discard rates and improved liver transplant outcomes.

Although most would agree that a prolonged DWIT can lead to significant ischemic damage to the liver and inferior graft survival, several reports have indicated that the duration of donor hypotension or hypoxemia may be a better predictor of poor outcomes after DCD liver transplantation. One multicenter study from a single organ procurement organization demonstrated that the total DWIT did not affect a composite endpoint representing a failed DCD liver allograft. However, the time interval between donor postextubation hypotension [defined as a systolic blood pressure (SBP) < 50 mm Hg] and organ flushing was a significant risk factor for inferior transplant outcomes.[9] Others have reported that if the duration of donor hypotension (defined as a mean arterial pressure < 60 mm Hg) is >20 minutes and CIT is >6 hours, there is an increased risk of graft loss.[10] In that analysis, the total DWIT was not a risk factor for allograft loss. When the DWIT, defined as the interval between the time of hypotension (SBP < 50 mm Hg) or hypoxia (oxygen saturation < 70%) and organ flushing, remains <30 minutes and CIT is minimized to <8 hours, DCD liver transplant outcomes are similar to those seen after DBD liver transplantation.[3]

In a recent issue of Liver Transplantation, Abt et al.[11] took the analysis of DWIT one step further. The authors reported a retrospective study investigating whether specific donor hemodynamic variables influenced graft survival after DCD liver transplantation. They studied the hemodynamic profile during warm ischemia for 110 DCD donors from a single organ procurement organization. In an attempt to identify more accurate predictors of graft loss, the authors divided the DWIT into specific hemodynamic variables that occurred during the withdrawal phase. These variables included (1) the area under the systolic blood pressure curve (AUCSBP), (2) the slope of the SBP over time from extubation to cross-clamping, and (3) the slope of the systolic blood pressure based on values during the first 10 minutes (SBP10). Multivariate regression models were created to study the association of donor and recipient variables with graft survival. In the model using AUCSBP, CIT, recipient age and AUCSBP were associated with graft survival. In the model using SBP10, CIT, recipient age and SBP10 were associated with graft survival. However, in each of the models, the total duration of DWIT, 23.6 ± 8.5 minutes, was not associated with graft survival. The authors then dichotomized SBP10 and generated a new variable, the median value of the slope of the systolic blood pressure based on values during the first 10 minutes (MEDSBP10). This was categorized according to the location of the slope of the blood pressure above or below the median of −7.2 mm Hg/minute. When the new model was repeated, MEDSBP10 was the only variable associated with graft survival.

This is another study investigating donor and recipient variables in which the total DWIT did not affect graft survival after DCD liver transplantation. This may have been due to a selection bias for livers with shorter DWITs. However, it could also have been due to the limitations of the multivariate model. It is likely that patients with a more rapid SBP decline in the first 10 minutes will have shorter DWITs. Therefore, the presence of factors that are potentially highly correlated (DWIT with AUCSBP, SBP10, and MEDSBP10) is problematic in that these factors may lead to unreliable estimates of hazard ratios and P values. However, the authors recognized the possibility of collinearities between covariates and reduced the number of parameters with a stepwise regression analysis. The multivariate analyses were rerun with both donor and recipient variables and included AUCSBP, SBP10, or MEDSBP10. Each of these variables had a significant impact on allograft survival. Using the Akaike information criterion, the authors concluded that MEDSBP10 was the best description of the model. Kaplan-Meier analyses revealed that patients who received allografts from donors who had a more rapid decline in their SBP10 values with steeper slopes (<−7.2) had an estimated 5-year graft survival rate of 76%, and patients with donors with less steep slopes (>−7.2) had a 5-year survival rate of 45%.

The authors should be commended for identifying other hemodynamic variables during the DWIT in the DCD donor that affect allograft survival. It makes sense that donors with a more rapid decline in SBP10 would likely have a shorter overall DWIT and decreased hepatic ischemia during the agonal phase. In addition, donors with a more gradual SBP decline in the first 10 minutes may have prolonged hepatic ischemia, and this possibly could lead to increased allograft loss. Based on this analysis the reader should not conclude that total DWIT is not a risk factor for liver allograft survival after DCD liver transplantation. It is likely that this lack of impact is due to the specific limitations of the multivariate model. It is also possible that when the total DWIT is ≤30 minutes, other dynamic factors during the DWIT are more predictive of graft loss as described in this study.

It is important to note that one should not limit the decision algorithm for selecting a DCD donor liver to only 1 or 2 variables. All donor, recipient, and transplant variables need to be considered when that critical decision is being made. This includes the Model for End-Stage Liver Disease score, which could not be studied in this analysis. However, this novel finding, that the rate of the SBP decline in the first 10 minutes is a strong predictor of allograft survival, can be added to data from other analyses reporting the impact of donor postextubation hypotension on transplant outcomes. Future multi-center studies are needed to determine whether these variables are predictive of graft loss and biliary complications when a greater number of DCD liver transplant recipients are studied.