|Home | About | Journals | Submit | Contact Us | Français|
Given the shortage of cadaveric organs, we began a study utilizing NHBD for OLTx and KTx. There were 24 NHBD between January 1989 and September 1993. These donors were divided into 2 groups: uncontrolled NHBD (G1) (n=14) were patients whose organs were recovered following a period of CPR; and controlled NHBD (G2) (n=10) were patients whose organs were procured after sustaining cardiopulmonary arrest (CA) following extubation in an operating room setting. Eight kidneys and 5 livers were discarded because of macroscopic or biopsy findings. In G1, 22/27 (81.5%) kidneys were transplanted; 14/22 (64%) developed ATN; 20/22 (95%) recipients were off dialysis at the time of discharge. With a mean follow-up of 32.7± 21.1 months, sixteen (73%) kidneys are still functioning, with a mean serum creatinine of 1.7±0.6 mg/dl. The one-year actuarial patient and graft survivals are 95% and 86%. In G2, 17/20 (85%) kidneys were transplanted; 13/17 (76%) kidneys experienced ATN. All patients were off dialysis by the time of discharge. With a mean follow-up of 17.6±15.4 months, twelve (70%) kidneys are still functioning, with a mean serum creatinine of 2.5±2.1 mg/dl. The one-year actuarial patient and graft survivals are 94% and 82%, respectively. In G1, 6/10 (60%) livers were transplanted; 3/6 (50%) livers functioned, the other 3 patients required ReOLTx in the first week postoperatively because of PNF(n=2) and inadequate portal flow (n=1). Two functioning livers were lost due to HAT (n=1) and CMV hepatitis (n=1). In G2, 6/7 (85.7%) livers were transplanted. All the livers (100%) functioned. 2 patients required ReOLTx for HAT at 0.9 and 1.0 months. Both patients eventually died. One patient with a functioning liver died 2 months post OLTx. The remaining 3 patients are alive and well at 27 months of follow-up. This study shows that the procurement of kidneys from both uncontrolled and controlled NHBD leads to acceptable graft function despite a high incidence of ATN. The function of liver allografts is adequate in the controlled NHBD but suboptimal in the uncontrolled NHBD, with a high rate of PNF.
The concept of utilizing organ allografts from non–heart-beating donors (NHBD)* for transplantation is not new. During the 1950s and 1960s, when clinical transplantation was in its infancy, living relatives (1) and NHBD (2) were the main source of human kidney allografts, and for unpaired organs—i.e., hearts and livers; NHBD were the only source of allografts (3–5).
Although the concept of brain death was first described in 1956 (6) it was not until 1968 that the guidelines for defining brain death were established and published (7). From that time until the present, brain dead HBD have been the most common source of organs for transplantation.
Refinements of surgical techniques, improvements in immunosuppression, development of effective organ preservation, and broader indications have led to increasingly successful results after transplantation. However, this success has exacerbated the shortage of donor organs, and this has resulted in an increasing number of potential allograft recipients dying on waiting lists.
As part of an effort to deal with this problem, we restarted an NHBD program and present the outcome of using such organs in KTx and OLTx.
Between January 1989 and September 1993, 978 kidneys and 430 livers were recovered from 498 cadaveric donors through our procurement agency. Of these, 24 (4.8%) were NHBD.
These donors were divided into two groups:
These patients (n=14) had either been pronounced brain dead by standard criteria, or were in the process of being pronounced brain dead, when they suddenly developed cardiac arrest. CPR was immediately begun, and the patients were transported to the operating room for recovery of the organs. CPR was employed for a mean of 37±29 min (range 10–100), which included the short period (<4 min from discontinuation of CPR) prior to initiating cold perfusion of the organs.
These patients (n=10) had irreversible neurological damage, but either did not fulfill brain death criteria or met the criteria of brain death but had families requesting life support withdrawal prior to organ procurement. In these cases, the patients were transported to the operating room where ventilation was discontinued, and recovery of the organs was performed once the patients were pronounced dead. The mean time from extubation until in-situ perfusion of the organs was begun was 23.8±11 min (range 10–92).
Table 1 shows our protocol for organ recovery in uncontrolled and controlled NHBD.
Eight uncontrolled and 2 controlled NHBD procurements were performed in 1989; in 1990–1991 and 1992–1993, 3 uncontrolled and 4 controlled donor procedures were done in each of the 2-year periods.
Thirteen of 14 donors in G1, and 6 of 10 in G2 were male. The mean ages of G1 and G2 were 23.7±14 (range 4–53) and 44.3±22 (range 4–66) years, respectively. Causes of death, mean prearrest donor serum creatinine, urine output, bilirubin, glutamyl oxalacetic transaminase (SGOT), and doses of vasopressor support are shown in Table 2.
In G1, the donors were transported to the operating room, and CPR was maintained until the recovery team arrived; at that point CPR was stopped and organ recovery begun. In G2, patients were transported to the operating room under controlled conditions of oxygenation and cardiac monitoring. Once in the operating room, the patients were extubated. The procurement teams waited until the patients became asystolic or developed electromechanical dissociation, and a physician, not a member of the recovery team, pronounced the patient dead. Following pronouncement of death, surgery was begun.
The so called “super-rapid” technique was used to recover the organs. After a midline incision from the xiphoid process to the symphysis pubis, the distal aorta was cannulated, and perfusion of the organs with cold preservation solution started (Fig. 1). Perfusion was routinely initiated less than 4 min after skin incision. Next, the sternum was split, the thoracic aorta was cross-clamped, and the intrapericardial inferior vena cava was vented to decompress the organs (Fig. 2). The inferior mesenteric vein was then cannulated to perfuse the portal system, and the abdominal cavity was filled with ice slush (Fig. 3). In adults, approximately 2 L of cold preservation solution (Viaspan) was infused into both the portal and the systemic arterial systems. Once the liver became palpably cold and free of blood, hepatectomy, followed by en bloc nephrectomies, was performed expeditiously.
Forty-eight kidneys and 17 livers were recovered from the 24 NHBD. One pediatric pair that was used en bloc was considered as a single organ. Consent for liver procurement was not obtained in 7 cases, and therefore could not be subsequently used. A senior staff surgeon was involved in all procurements, and surgeon’s opinion of organ viability was the single most important determinant of using or discarding an organ. Five of the 48 kidneys (all G1) and 1 of the 17 livers (G2) were deemed not suitable for transplantation because of gross evidence of hypoperfusion. In addition, all organs of indeterminate quality underwent back table biopsy.
A total of 18 kidneys from G1 and G2 were biopsied; 3 of the kidneys in G2 were discarded because of biopsy findings showing 15% glomerulosclerosis (1 case) and 50% glomerulosclerosis (2 cases).
Of the livers from G1 and G2, ten underwent back table biopsy, and 2 of the livers in G1 were discarded because of histological findings showing greater than 50% macrosteatosis (1 case) and significant centrilobular hepatocellular degeneration (1 case). Both livers were extremely firm, with blunt edges, and homogeneously pale. The remaining liver was discarded because the recipient died. In G2 an additional liver was discarded because of microscopic findings of greater than 75% steatosis and a grossly fatty appearance.
The mean recipient age was 41.3±16.0 years (range 18–64) in G1 and 46.8±9.0 years (range 27–60) in G2 (Table 4). Two (9%) patients in G1 and 2 (12%) patients in G2 were undergoing retransplantation. Three (14%) patients in G1 and 2 (12%) patients in G2 had a PRA greater than 40% at the time of transplantation. The mean CIT was 28.5±5.8 hr in G1 and 25.6±8.5 hr in G2. The causes of end-stage renal disease for both G1 and G2 were similar to the indications for kidney transplant of all candidates at our center, with diabetic nephropathy and glomerulonephritis being the most common.
The mean recipient age was 39.7±20.8 years (range 0.4–57) in G1 and 50.8±18.8 years (range 15–66) in G2 (Table 5). In G1, three patients were status 4 and three were status 3 by the UNOS classification. In G2, all patients were status 3 at the time of transplantation. The mean CIT was 10.7±2.8 hr in G1 and 11±2.3 hr in G2. The causes of end-stage liver disease for G1 and G2 liver recipients were representative for our adult candidate population.
Patient survival was calculated from the date of KTx or OLTx until death, and graft survival from the date of transplantation until graft failure, retransplantation, or patient death. Survival curves were generated using the Kaplan-Meier (product limit) method and were compared using the log-rank (Mantel-Cox) test. A P value less than .05 was considered statistically significant.
Of the 22 transplanted kidneys from G1 donors, 14 (64%) developed ATN, which lasted a mean of 14.2±12.7 days (range 4–39); 8 patients (36%) required hemodialysis posttransplant, and 20 (95%) patients were off dialysis by the time of discharge. The mean serum creatinine at discharge was 2.9±1.4 mg/dl. The one-year actuarial patient and graft survivals were 95% and 86%, respectively (Fig. 4). Two grafts failed, one at 0.5 months due to arterial thrombosis, and the other from accelerated rejection at two months. Five patients died, three with functioning kidneys, at one day (respiratory failure), 14 months (sepsis s/p above knee amputation), 16 months (lymphoma), 28 months (sepsis) and 38 months (sepsis and liver failure), respectively. With a mean follow-up of 32.6±21.1 months (range 14–60), 16 (73%) kidneys are still functioning with a mean serum creatinine of 1.7±0.6 mg/dl.
Of the 17 transplanted kidneys from G2 donors, 13 (76%) experienced ATN that lasted a mean of 10.7±7.5 days (range 3–30); 7 patients (41%) required hemodialysis posttransplant. All patients were off dialysis at the time of discharge. The mean serum creatinine at discharge was 3.2±1.7 mg/dl, and the most recent creatinine is 2.5±2.1 mg/dl at 17.6± 15.4 months of follow-up (range 1.5–5.6). The one-year actuarial patient and graft survivals were 94% and 82%, respectively (Fig. 4). One patient with a functioning kidney died at 1.5 months from a pulmonary embolus. Four grafts failed from histologically demonstrated chronic rejection at 5, 8, 32, and 38 months, respectively.
Of the 6 transplanted livers procured from G1 donors, 3 (50%) functioned, with a peak serum bilirubin of 4.2±0.4 mg/dl and a peak SGOT of 1135±955 IU/L during the first week post-OLTx. Two of these 3 patients required retransplantation due to HAT (n=1) and CMV hepatitis (n=1) at 1.4 and 1.0 months respectively after OLTx, One patient with a functioning liver is alive and well at 58 months posttransplant. The remaining 3 patients required retransplantation within the first week postoperatively because of PNF (n=2) and inadequate portal flow (n=1).
The back table biopsy of the 3 allografts that failed within the first week after transplantation were essentially normal, except for mild scattered spotty acidophilic necrosis and minimal microvesicular steatosis (10–15%), In contrast, the intraoperative postreperfusion biopsies showed large areas of centrilobular and/or periportal hemorrhagic necrosis.
Of the 6 transplanted livers that were procured from G2 donors, all (100%) functioned with a peak serum bilirubin of 8.7±8.3 mg/dl and a peak SGOT of 693±326 IU/L within the first week posttransplantation. Two patients required retransplantation at 0.9 and 1.0 months due to HAT; in one of these cases, a severe hepatic arterial vasculitis with a resulting lumen occlusion was found. Both patients died. One patient with a functioning liver died of a myocardial infarction at 2 months posttransplantation. The remaining 3 patients are alive and well at one at 6, one at 11 and one at 30 months.
The one year actuarial patient/graft survival was 67%/17% from G1 donors and 50%/50% from G2 donors (Fig. 5).
The rationale for reconsidering organ donation from non–heart-beating cadavers has been the need to address the disparity between the demand and the supply of human allografts for transplantation. This shortage of organs is the major limiting factor in the number of renal and extrarenal transplants that are presently being performed. Therefore, the transplant program at the University of Pittsburgh Medical Center, in collaboration with the regional organ procurement organization, Center for Organ Recovery and Education (C.O.K.E.), has reexamined the suitability of procuring organs from NHBD. While we have concentrated on the recovery of kidneys and livers from these NHBD, there is also the potential to expand this program to include other organs.
The results of transplantation of NHBD kidney allografts compared favorably with those of HBD. At the University of Pittsburgh, one-year kidney patient and graft survivals in the overall NHBD kidney experience were 90% and 85%, respectively. The one-year patient and graft survivals for kidneys obtained from HBD during the same period were 93% and 80%, respectively. This was not statistically significant (Fig. 4), On the, other hand, the results for NHBD liver allografts were poorer than those for HBD, however, given the small number of patients in both G1 and G2, clinically relevant comparisons could not be made.
Under routine conditions of brain death pronouncement, and with modern techniques of multiple organ procurement (8, 9) there is virtually no preoperative warm ischemia. Maintenance of cardiorespiratory function in the absence of brain function has been a major advance in reducing ischemic injury to organ allografts. In contrast, acquisition of organs from NHBD is not performed under ideal conditions, as no cardiac or respiratory activity is present for a variable period prior to organ recovery. Consequently, these organs are subjected to some warm ischemia time; the period of hypoxia and lack of effective circulation is even more pronounced in patients in whom CPR was started before recovery. Therefore, it would be anticipated that the impact of a period of warm ischemia would be even more pronounced than an equal period of warm ischemia seen in other types of liver surgery. There were no differences in the rewarming ischemia time during implantation between G1, G2 and HBD (data not shown).
In both controlled and uncontrolled situations, the procurement of livers and kidneys must be accomplished quickly, and this should be performed only by an experienced surgical team. The urgent time frame in which organs are procured from uncontrolled NHBD makes these the most difficult of all donor situations. The most important principle of the so-called super-rapid technique is that the organs must be rapidly cooled prior to any attempts at the dissection.
An initial assessment of the quality of the liver was done by the donor surgeon. In most of the hemodynamically unstable donors the livers were inhomogeneously cyanotic and showed signs of congestion as a result of a partial outflow blockage of the abdominal viscera caused by CPR maneuvers. This finding disappeared after cold perfusion in all livers with one exception. This liver was discarded because of gross evidence of hypoperfusion. When donor surgeons had any doubts about the viability of the livers, frozen section biopsies were performed. In all cases where the surgeons considered the liver unsuitable, the frozen section was confirmatory. Five of the liver allografts that were transplanted had both a back table and postreperfusion biopsy. The results of these pretransplant and posttransplant biopsies were then correlated with ultimate graft outcome to see if, in retrospect, a different decision would have been made about the suitability of the graft for transplantation. Two biopsied livers had a gross and histologically normal appearance before transplantation and developed PNF. This inability to predict organ function uniformly after transplantation by light microscopic examination of biopsy specimens has been widely discussed in the literature (10–13).
While all attempts were made to limit preoperative warm ischemia, the mean period of preoperative warm ischemia was 37 (range 10–100) min for uncontrolled NHBD and 24 (range 10–92) min for controlled NHBD. Renal transplantation from NHBD has been reported to have a high incidence of both ATN and vascular complications. One report shows that 10% of the kidneys developed thrombotic microangiopathy, which was attributed to a warm ischemic injury to the renal microvasculature (14). While the kidneys procured in our series did not have a high incidence of arterial complications, a similar pathophysiology in the procured livers may explain the high incidence of PNF and HAT. Diagnosis of HAT accounts for less than 10% of early graft failures in organs recovered from HBD (15). In our series, 3 of 12 patients (25%) developed HAT and required retransplantation; two of these patients eventually died. Although these complications were classified as technical errors, it is possible that hepatic preservation/reperfusion injury played a role in these events.
The warm ischemia and the inevitable cold preservation are well-known factors that cause injury to the parenchymal and nonparenchymal liver cells, in particular the Kupffer cells and sinusoidal cells of the microvasculature. The degree of this damage and that performed by the reperfusion mechanism can result in local clotting of the hepatic microvasculature (10, 11, 13).
This study attempted to delineate the circumstances in which renal and extrarenal organs can be procured from NHBD. Our data confirm that the procurement of kidneys from both uncontrolled and controlled NHBD leads to acceptable graft function, despite a high incidence of ATN. These results compare favorably with other reported attempts at using kidneys from NHBD. Yokoyama et al., using the in situ perfusion technique of kidneys in NHBD, reported a one-year patients and graft survival of 85%, with an ATN rate of 75% (16). Varty et al., also using a double-balloon catheter reported an overall 75% graft survival (17). It appears that the results of using organs from NHBD depends on factors inherent in minimizing warm ischemic injury. Phillips et al, obtained less-favorable results in NHBD with prolonged warm ischemia times (18).
In the controlled NHBD, immediate good liver allograft function can be anticipated—however, liver allografts from uncontrolled NHBD have a high incidence of PNF. The increased rate of arterial complications in livers from NHBD remains unexplained. Further studies and experience in this group of liver donors may help elucidate the mechanism of both PNF and HAT.
In summary, NHBD represent a viable source of cadaveric kidneys but must be carefully assessed as to their appropriateness for liver usage.
We thank Susan A. Stuart, R.N., B.S.N., C.P.T.C, for her help in the management of the non–heart-beating donors; Karen Toler, William Irish, M.S., and David Krakosky for their help with graph and slide preparation; and Susan Shandor for her help with typing the manuscript, tables, and slide preparation.
1Presented at the 20th Annual Meeting of the American Society of Transplant Surgeons, May 18–20, 1994, Chicago, IL.
*Abbreviations: ATN, Acute tubular necrosis; CA, Cardiopulmonary arrest; CIT, Cold ischemia time; CMV, Cytomegalovirus; CPR, Cardiopulmonary resuscitation; G1, Group 1; G2, Group 2; HAT, Hepatic artery thrombosis; HBD, Heart-beating donor; KTx, Kidney transplant; NHBD, Non–heart-beating donor; OLTx, Orthotopic liver transplant; PNF, Primary nonfunction; PRA, Panel-reactive antibody; UNOS, United Network of Organ Sharing.