Organs procured from donors declared dead based on irreversible cessation of cardiac function (DCD) accounted for more than 6% of kidneys and nearly 4% of livers transplanted in the United States in 2006 (1). Controlled (Maastrict type III) DCD transplants have increased virtually every year, already constitute an important source of organs, and could increase the supply by as much as 20% to 25% (2, 3). Although these organs provide life-saving therapy, concerns about outcomes temper even more widespread use. In renal transplant patients, rates of delayed graft function (DGF) and primary nonfunction are higher with DCD organs compared with standard criteria organs, although long-term graft function is generally considered to be equivalent (4–9). On the other hand, DCD livers demonstrate increased rates of biliary complications, primary nonfunction and worse patient survival compared with standard criteria livers (10–15).
Preoperative identification of factors associated with poor outcomes remains an important challenge. Recent reports address this issue. After renal transplant, time between donor extubation and asystole does not correlate with function, though recipient hemodynamics does (6, 16). Among liver transplant patients, time from donor extubation and asystole predicts ischemic liver strictures (11).
Characteristics of the period between extubation and asystole may help predict graft function. Less organ damage may occur if the donor progresses quickly to cardiac death as opposed to maintaining a heartbeat in the presence of significant hypoxia or hypotension. Consistent with this, restrictions on recovery are imposed at certain centers based on donor time to death and duration of hypoxia or hypotension. Furthermore, the 2006 Report of a National Conference on Donation after Cardiac Death recommended collecting minute-by-minute donor hemodynamic data during the period between extubation and the initiation of cold perfusion so this question can be addressed (17).
Our purpose was to determine whether DCD donor time to death or duration and severity of postextubation hypoxia or hypotension correlated with subsequent kidney or liver function, and to begin to set guidelines for these times that might be clinically useful. To do this, we performed a multicenter, retrospective analysis and attempted to establish break points such that relevant outcomes could be predicted.
PATIENTS AND METHODS
This is a multicenter, retrospective review of donor postextubation characteristics and outcomes in the respective recipients. All organs were procured by the New England Organ Bank between October 1, 1999 and April 25, 2006. There were 143 consecutive adult DCD kidney transplants and 39 DCD liver transplants at the participating centers during the study period; data on matched donor and recipient pairs were available from 134 kidneys and 37 livers, though not all data were available for all patients. Organs were excluded from this study only when the particular parameter of interest was not obtainable.
Definitions and Endpoints
Time to death was minutes from extubation until declaration of death in the donor. Donor total warm ischemia time was minutes elapsed from extubation to initiation of cold perfusion, which includes phase I (withdrawal) and phase II (acirculatory) (17). Cold ischemia time was hours elapsed from aortic perfusion in the donor to revascularization in the recipient. DGF in kidney recipients was the requirement for dialysis in the first week after transplantation. Duration of DGF was days elapsed until the last day of dialysis after transplantation. Estimated glomerular filtration rate was calculated from the Modification of Diet in Renal Disease equation (18, 19). Acute rejection was any rejection treatment within the first year after transplantation. Kidney graft loss was the date of return to chronic dialysis, retransplantation, or death. All biliary strictures, leaks, and diffuse intrahepatic biliary abnormalities were diagnosed on cholangiography. Liver graft loss was the date of retransplantation or death.
Kidney-specific endpoints were primary nonfunction, DGF, and graft function after 6 months and 1 year. Because of the small number of liver transplants, we chose a composite endpoint to represent a failed graft; this included any transplant-related death, primary nonfunction, graft loss within 1 year, or diffuse intrahepatic biliary abnormalities, a lesion that often requires retransplantation.
New England Organ Bank maintains a prospectively collected database of all consented DCD organ donors. The project was approved by the Medical Director and the Clinical Policy Board in accordance with the New England Organ Bank's research policy.
All DCD recoveries occurred in patients extubated under controlled conditions after the decision was made to withdraw support (Maastricht classification III) (20). Withdrawal of support occurred in accordance with the individual hospital policy in the intensive care unit or operating room. DCD kidney and liver donor age, gender, comorbidities (diabetes mellitus, coronary artery disease, hypertension, chronic obstructive pulmonary disease, cerebrovascular disease, hypercholesterolemia), terminal creatinine, terminal liver function tests, cause of death, and biopsy results were collected, as were vital signs in the period between consent and asystole. This included time to death, oxygen saturations, blood pressure, and heart rate.
After withdrawal of care, patients were monitored by continuous electrocardiogram, blood pressure cuff, finger pulse oximetry, and physical examination. Typically, vital parameters were recorded every 1 to 5 min. In all cases, death was defined as irreversible cessation of cardiopulmonary activity as evidenced by 5 min of asystole and absence of spontaneous respirations according to the Uniform Declaration of Death Act (21). A physician not associated with the transplant team made pronouncement of death. Initial perfusion in the first few donors was performed by means of femoral cannulation before changing to direct aortic cannulation. Livers were preserved in University of Wisconsin solution, and all kidneys were stored with continuous pulsatile perfusion (“pumped”) unless technically impossible.
Information on recipient outcomes was gathered after obtaining permission from the appropriate institutional review boards. All recipient data were collected from prospectively maintained databases and included age, gender, race, indication for transplantation, comorbidities, and posttransplantation variables as outlined above. Patients were managed according to the standard practices at each center. DGF was treated and its resolution determined according to the individual center. Kidney immunosuppression protocols differed slightly among the different centers, but in general involved induction therapy with antithymocyte globulin and maintenance therapy with cyclosporine or tacrolimus or sirolimus and prednisone with azathioprine or mycophenolate mofetil. Liver immunosuppression was based on three drugs, tacrolimus or cyclosporine, mycophenolate mofetil, and steroids.
Continuous variables are presented as mean±standard deviation; categorical values are presented as percentages. Comparisons between categorical variables were made with the chi-square test. Comparisons between continuous variables were made with the two-tailed Student t test. Allograft survival was assessed by the life table method with standard errors calculated by the Greenwood method. P-values less than 0.05 were considered statistically significant. Receiver operating characteristic curves were calculated according to standard methods using a fitted curve available at http://www.jrocfit.org.
Kidney Donor and Recipient Demographics and Donor Perideath Parameters
Donor and recipient data were available from 134 kidneys. Characteristics are presented in Table 1. The mean follow-up was 580±415 days. Donors tended to have few major medical problems. They were young (41.8±14.4 years), with a relatively low body mass index (BMI) (27.3±7.4). Only 11.2% were extended criteria donors (ECD) and DCD. Recipients were older (51.8±25.2 years) with multiple comorbidities (37% with diabetes and 30.8% with coronary artery disease).
Most kidneys were transplanted from local DCD donors, with cold ischemia times averaging 13.0±5.0 hr. The rate of DGF was 45%, with a mean duration of 12.5±8.2 days. The incidence of kidney primary nonfunction was 2.9%, 1-year graft survival was 94.6% and patient survival was 97% with a mean estimated glomerular filtration rate at 1 year of 53.4 mL/min/1.73 m2. Acute rejection occurred in 12 (9%) of recipients. All patients with graft failure returned to dialysis before retransplantation.
Time from donor extubation to asystole averaged 19.1±12.1 min, and the average time from extubation to cannulation was 28.8±11.5 min. Systolic hypotension less than 50 mm Hg lasted 18.0±7.4 min, and hypoxia with oxygen saturation less than 60% lasted 24.4±11.6 min, before cannulation was achieved.
Duration of Severe Hypotension is a Better Predictor of DGF Than Time From Extubation to Asystole
Prolonged hemodynamic instability produced a trend toward higher rates of DGF. There was no significant association between DGF and time from asystole to flush, prolonged hypoxia, cold ischemic time, or time from extubation to asystole (Table 2). DGF among pumped and nonpumped kidneys was 39.8% and 52.3%, respectively. To produce meaningful clinical guidelines, we further analyzed the parameters with the lowest p values. Systolic blood pressure (SBP) less than 60 or 70 mm Hg for greater than 20 min correlated with a trend toward increased DGF (SBP <60: 35.2% vs. 58.8%, P=0.067; SBP <70: 35.9% vs. 54.5%, P=0.075) (Table 3). There was no difference in renal function, primary nonfunction, 1-year graft survival, or 1-year patient survival among these groups. In both cases, the groups with lower DGF rates had higher numbers of ECD kidneys.
Other Factors Affecting DGF Rates
Kidneys with or without DGF did not vary with regard to donor ECD status, hypertension, stroke as cause of death, or recipient hypertension.
Data on 37 transplanted livers demonstrated that donors tended to be young (average age 34.3±13.2 years) and thin (average BMI 25.1±5.8) (Table 4). Average time from extubation to asystole was 17.6±8.7 min and average time from extubation to flush was 28.6±8.8 min. Recipient MELD scores averaged 22.8±7.5. Mean follow-up time was 416.3± 328.5 days. One-year patient and graft survival rates were 81.1% and 78.9%, respectively. A total of 14 (37.8%) of recipients reached the composite endpoint of death, graft loss within 1 year, or diffuse intrahepatic biliary abnormalities.
Liver Donor Factors That Predict the Primary Endpoint
Recipients who reached the primary endpoint tended to have livers from donors who were older, exhibited longer profound hypotension in the postextubation period, and had slightly longer cold times (Tables 5 and 6). To determine clinically relevant parameters for predicting the primary endpoint, we performed a receiver operating characteristic analysis on donor and recipient parameters including donor age, cold ischemic time, time from extubation to asystole, time from extubation to flush, time with oxygen saturation less than 50% to flush, and time from systolic blood pressure of less than 50 mm Hg to flush (some representative curves are shown in Fig. 1). The best predictive test was time from systolic blood pressure less than 50 mm Hg to flush (area under curve=0.652), and the next best was donor age (area under curve=0.589). Time from extubation to flush was not predictive of the primary endpoint (area under curve=0.467). Dichotomization of time from systolic blood pressure of less than 50 mm Hg to flush revealed if this time was less than 15 min, only 20% reached the primary endpoint, whereas if this time was greater than 15 min, 55% reached the primary endpoint (P=0.046). A similar analysis revealed that in donors younger than 35, the primary endpoint was reached in 18.8%, compared with 52.4% for donors older than 35 (P=0.048). Donor BMI, time from extubation to asystole, time from extubation to flush, or duration of profound hypoxia did not correlate with the primary endpoint.
In the liver cohort, complications included eight patients with diffuse biliary strictures (21.6%), four with minor biliary strictures managed with dilation or stenting (10.8%), and one with acute humoral rejection (2.7%). Of the eight patients with diffuse intrahepatic biliary ischemia (follow-up 520±374 days), 3 (37.5%) were relisted, 3 (37.5%) died without relisting, 1 (12.5%) was relisted and died, and 1 (12.5%) survived without relisting.
This study identified donor factors in the postextubation period that correlate with poor outcomes after DCD transplantation. There is very little data on this question, leading to wide variation in organ acceptance criteria. Certain centers set limits on total time from extubation to death, others on duration and severity of hemodynamic instability, and others on both. The 2006 Report of the National Conference on Donation after Cardiac Death recommends recovering kidneys if the time from extubation to cold perfusion is less than 1 hr, and to recover livers if this time is less than 30 min (17). These recommendations are based on animal models and show that increasing this time from extubation to organ recovery decreases survival (22–28).
Among renal transplant recipients in our cohort, the high rate of DGF, 45%, is consistent with previously reported results (6, 29), though there is wide variation in reported rates. Data from the Scientific Registry of Transplant Recipients (2000–2004) report DGF rates of 40.2% and 42.3% in DCD kidneys that are pumped and not pumped, respectively (17). The use of pulsatile perfusion in our series is therefore comparable with the SRTR data.
Time from extubation to asystole did not correlate with DGF. Profound hypotension demonstrated a trend toward predicting DGF, with a P value less than 0.10. (Table 2). Dichotomization of donors based on duration of SBP less than 60 or less than 70 mm Hg, the two most suggestive parameters, revealed much higher rates of DGF in patients with prolonged hypotension, though these results did not quite meet statistical significance (Table 3). It is likely that our small sample size accounts for this lack of significance and we expect that larger studies will find significant correlation. Furthermore, donor ECD status, cause of death, or hypertension did not predict DGF, nor did recipient hypertension. No postextubation parameter predicted long-term renal function, suggesting that the injury associated with DGF is transient and recoverable.
Our data suggest that donors with prolonged postextubation hypotension will exhibit increased DGF, but that long-term function is acceptable. Since nearly all of the kidneys in this analysis were recovered within a 1-hr time period, we are unable to comment on whether it is safe to wait longer. In the absence of hemodynamic instability waiting longer might result in organs of acceptable quality. Vital signs after extubation may be a better parameter than time from extubation to asystole in determining when to abandon an organ recovery. Ultimately, the question of how long to wait for donor death must take into account the best usage of scarce resources, including the burden on the donor hospital and procurement team.
To maximize our statistical power in the DCD liver transplant recipients, we devised a composite primary endpoint consisting of diffuse intrahepatic biliary abnormalities, primary nonfunction, or death. We felt that this was the best representation of a transplant that did not achieve its intended goal. In our experience, diffuse intrahepatic biliary abnormalities often require retransplantation. Overall, 37.8% of patients in this study reached the primary endpoint, and 1-year patient and graft survival was only 81.1% and 78.9%, respectively. These outcomes are consistent with other reports on DCD liver transplantation: D'Alessandro et al. (30) reported 3-year graft survival of 59%, Foley et al. (10) reported a 67% 1-year graft survival rate with a 33% incidence of biliary strictures, and Abt et al. (11) reported a 72% 1-year graft survival rate with 33% incidence of major biliary complications. The United Network for Organ Sharing database reports 1-year graft survival of 71% in DCD liver transplant recipients (13).
Receiver operating characteristic curves were used to determine which variable might most accurately predict outcome. Based on this analysis, time from systolic blood pressure of less than 50 mm Hg to flush was the best predictor of subsequent graft function. Importantly, this parameter was superior than time from extubation to asystole, the parameter on which the National Conference based its recommendation. We did not find increasing risk of graft failure with increasing time between extubation and asystole (31). Since longer times with profound hypotension add to the time from extubation and death, these parameters are dependent, and it may be that the profound hypotensive time is the more important variable and we lack the power to demonstrate the less robust decrease in survival as time from extubation to asystole increases.
To translate our findings into clinically useful guidelines, we dichotomized donor variables to find a threshold to guide clinical decisions. Among patients who received livers from donors in which SBP was less than 50 mm Hg for more than 15 min, 55% reached the primary endpoint, compared with only 20% if this interval lasted less than or equal to 15 min (P=0.046). Similarly, if donors were older than 35 years, 52.4% of their respective recipients reached the primary endpoint, compared with 18.8% if donors were younger than 35 years (P=0.048). Importantly, these data identify diffuse intrahepatic biliary abnormalities at a very high rate, suggesting that this lesion is a particular feature of DCD liver transplantation.
Statistical analysis is complicated by the retrospective attempt to establish clinically significant breakpoints. This statistical problem should not be underestimated in the interpretation of the data. On the other hand, the very high rates of major problems predicted by seemingly relevant clinical parameters should not be discounted.
The analysis is limited by recording variables only every 1 to 5 min, compromising our ability to establish precise thresholds for predicting adverse outcomes. Furthermore, we did not examine, or did not have adequate power to address, the contribution of a number of other variables that may influence the development of our endpoints, such as donor fat content and hepatitis C status.
This study suggests that rather than using time from extubation to asystole to determine whether an organ should be recovered, time of profound hemodynamic instability may be a better predictor of subsequent function. Until these results are prospectively validated, they must be interpreted with caution. They highlight the need for a multicenter, prospective study with enough power to enable multivariable statistical analysis of the influence of vital signs after extubation on outcomes in DCD transplantation, specifically DGF after kidney transplant, and diffuse intrahepatic biliary abnormalities, primary nonfunction, and death, after liver transplantation. Confirmation of the results of our analysis may provide criteria to dramatically improve outcomes after DCD liver transplantation.
This work was supported by the Julie Henry fund for outcomes research at Beth Israel Deaconess Hospital. We gratefully appreciate the participation of the New England Organ Bank in assembling the data for this report.
2. Aoun S, Ramos E. Expanding the donor pool: Effect on graft outcome. Transplant Proc
1999; 31: 3379.
3. Sanchez-Fructuoso AI, Prats D, Torrente J, et al. Renal transplantation from non-heart beating donors: A promising alternative to enlarge the donor pool. J Am Soc Nephrol
2000; 11: 350.
4. D'Alessandro AM, Fernandez LA, Chin LT, et al. Donation after cardiac death: The University of Wisconsin experience. Ann Transplant
2004; 9: 68.
5. Cooper JT, Chin LT, Krieger NR, et al. Donation after cardiac death: The University of Wisconsin experience with renal transplantation. Am J Transplant
2004; 4: 1490.
6. Sohrabi S, Navarro A, Wilson C, et al. Renal graft function after prolonged agonal time in non-heart-beating donors. Transplant Proc
2006; 38: 3400.
7. Sanni AO, Wilson CH, Wyrley-Birch H, et al. Non-heart-beating kidney
transplantation: 6-year outcomes. Transplant Proc
2006; 38: 3396.
8. Rengel M, Kanter J, Puerta M, et al. Kidney
transplantation with grafts from non-heart-beating donors. Transplant Proc
2006; 38: 890.
9. Alonso A, Fernandez-Rivera C, Villaverde P, et al. Renal transplantation from non-heart-beating donors: A single-center 10-year experience. Transplant Proc
2005; 37: 3658.
10. Foley DP, Fernandez LA, Leverson G, et al. Donation after cardiac death: The University of Wisconsin experience with liver
transplantation. Ann Surg
2005; 242: 724.
11. Abt P, Crawford M, Desai N, et al. Liver
transplantation from controlled non-heart-beating donors: An increased incidence of biliary complications. Transplantation
2003; 75: 1659.
12. Manzarbeitia CY, Ortiz JA, Jeon H, et al. Long-term outcome of controlled, non-heart-beating donor liver
2004; 78: 211.
13. Mateo R, Cho Y, Singh G, et al. Risk factors for graft survival after liver
transplantation from donation after cardiac death donors: An analysis of OPTN/UNOS data. Am J Transplant
2006; 6: 791.
14. Feng S, Goodrich NP, Bragg-Gresham JL, et al. Characteristics associated with liver
graft failure: The concept of a donor risk index. Am J Transplant
2006; 6: 783.
15. Doshi MD, Hunsicker LG. Short- and long-term outcomes with the use of kidneys and livers donated after cardiac death. Am J Transplant
2007; 7: 122.
16. Snoeijs MG, Wiermans B, Christiaans MH, et al. Recipient hemodynamics during non-heart-beating donor kidney
transplantation are major predictors of primary nonfunction. Am J Transplant
2007; 7: 1158.
17. Bernat JL, D'Alessandro AM, Port FK, et al. Report of a national conference on donation after cardiac death. Am J Transplant
2006; 6: 281.
18. Poggio ED, Nef PC, Wang X, et al. Performance of the Cockcroft-Gault and modification of diet in renal disease equations in estimating GFR in ill hospitalized patients. Am J Kidney Dis
2005; 46: 242.
19. Levey AS, Bosch JP, Lewis JB, et al. A more accurate method to estimate glomerular filtration rate from serum creatinine: A new prediction equation. Modification of Diet in Renal Disease Study Group. Ann Intern Med
1999; 130: 461.
20. Kootstra G, Daemen JH, Oomen AP. Categories of non-heart-beating donors. Transplant Proc
1995; 27: 2893.
21. Guidelines for the determination of death. Report of the medical consultants on the diagnosis of death to the President's Commission for the Study of Ethical Problems in Medicine and Biomedical and Behavioral Research. JAMA
1981; 246: 2184.
22. Monbaliu D, van Pelt J, De Vos R, et al. Primary graft nonfunction and Kupffer cell activation after liver
transplantation from non-heart-beating donors in pigs. Liver Transpl
2007; 13: 239.
23. Takada Y, Taniguchi H, Fukunaga K, et al. Hepatic allograft procurement from non-heart-beating donors: Limits of warm ischemia in porcine liver
1997; 63: 369.
24. Reich DJ, Munoz SJ, Rothstein KD, et al. Controlled non-heart-beating donor liver
transplantation: A successful single center experience, with topic update. Transplantation
2000; 70: 1159.
25. Garcia-Valdecasas JC, Tabet J, Valero R, et al. Evaluation of ischemic injury during liver
procurement from non-heart-beating donors. Eur Surg Res
1999; 31: 447.
26. Gruenberger T, Steininger R, Sautner T, et al. Influence of donor criteria on postoperative graft function after orthotopic liver
transplantation. Transpl Int
1994; 7(suppl 1): S672.
27. Casavilla A, Ramirez C, Shapiro R, et al. Experience with liver
allografts from non-heart-beating donors. Transplant Proc
1995; 27: 2898.
28. Lee YJ, Lee SG, Kwon TW, et al. Donor characteristics for liver
transplantation and risk factors for early poor graft function and survival. Transplant Proc
1996; 28: 1663.
29. Mueller AR, Platz KP, Krause P, et al. Perioperative factors influencing patient outcome after liver
transplantation. Transpl Int
2000; 13(suppl 1): S158.
30. D'Alessandro AM, Hoffmann RM, Knechtle SJ, et al. Liver
transplantation from controlled non-heart-beating donors. Surgery
2000; 128: 579.
31. Pomfret EA, Fryer JP, Sima CS, et al. Liver
and intestine transplantation in the United States, 1996–2005. Am J Transplant
2007; 7: 1376.
Keywords:© 2008 Lippincott Williams & Wilkins, Inc.
Liver; Kidney; DCD; Hemodynamic instability