LIVER TRANSPLANTATION 12:1365-1370, 2006 ORIGINAL ARTICLE Surgical Injuries of Postmortem Donor Livers: Incidence and Impact on Outcome After Adult Liver Transplantation Danielle M. Nijkamp, 1 Maarten J.H. Slooff, 2 Christian S. van der Hilst, 3 Alexander J.C. IJtsma, 2 Koert P. de Jong, 2 Paul M.J.G. Peeters, 2 and Robert J. Porte 1,2 1 Section of Transplantation and Organ Donation, 2 Section of Hepatobiliary Surgery and Liver Transplantation, Department of Surgery, and 3 Department of Medical Technology Assessment, University Medical Center Groningen, Groningen, The Netherlands The exact frequency and clinical consequences of surgical hepatic injuries during organ procurement are unknown. We analyzed the incidence, risk factors, and clinical outcome of surgical injuries in 241 adult liver grafts. Hepatic injuries were categorized as parenchymal, vascular, or biliary. Outcome variables were bleeding complications, hepatic artery thrombosis (HAT), and graft survival. In 82 livers (34%), 96 injuries were detected. Most injuries were minor, but clinically relevant injuries were detected in 6.6% (16/241) of the livers. Fifty (21%) liver grafts had some degree of parenchymal or capsular injury, 40 (17%) had vascular injury, and 6 (2%) had an injury to the bile duct. Procurement region was the only risk factor significantly associated with surgical injury. The rate of hepatic artery injury was significantly higher in livers with aberrant arterial anatomy. Bleeding complications were found in 18% of patients who received livers with a parenchymal or capsular injury in contrast to 9% without parenchymal injury (P 0.065). HAT was found in 23% of the patients who received a liver with arterial injury compared to 4% without arterial injury (P 0.001). Overall graft survival rates were not significantly different for grafts with or without anatomical injury. In conclusion, surgical injuries of donor livers are an underestimated problem in liver transplantation and can be observed in about one-third of all cases. Clinically relevant injuries are detected in 6.6% of all liver grafts. Arterial injuries are associated with an increased risk of HAT. Liver Transpl 12:1365-1370, 2006. 2006 AASLD. Received November 30, 2005; accepted March 18, 2006. The success of organ transplantation has led to an increasing demand for transplantable organs that has outgrown the availability of postmortem donor organs in many parts of the world. As a consequence, waiting lists for most types of organ transplantation have grown rapidly during the last decade. 1 It is therefore of paramount importance that organs for transplantation are retrieved without surgical injury. Very few studies have focused on the prevalence and clinical impact of liver damage caused by surgical handling during organ retrieval. 2,3 We believe that this problem is underreported and therefore maybe underestimated. Information on this subject, however, is clinically relevant. Surgical injuries to the donor liver may lead to a higher morbidity, to graft dysfunction, or even to graft loss after transplantation. In the worst case, donor livers may even become unsuitable for transplantation. In this way, surgically induced hepatic injury during organ retrieval can have a negative impact on the donor pool size. Identification and quantification of the problem can be used in the training of surgical teams responsible for organ procurement. The aim of this study was to assess the frequency and types of surgical injuries of liver grafts during organ procurement. The clinical impact of these injuries on outcome after subsequent liver transplantation was studied. In addition, we aimed to identify risk factors for surgically induced hepatic injury during organ procurement. Abbreviation: HAT, hepatic artery thrombosis. Address reprint requests to Danielle M. Nijkamp, Department of Surgery, University Medical Center Groningen, P.O. Box 30001, 9700 RB Groningen, The Netherlands. Telephone: 0031 50 361 3264; FAX: 0031 50 361 9050; E-mail: d.m.nijkamp@chir.umcg.nl DOI 10.1002/lt.20809 Published online in Wiley InterScience (www.interscience.wiley.com). 2006 American Association for the Study of Liver Diseases.
1366 NIJKAMP ET AL. MATERIALS AND METHODS Donors and Recipients Between September 1996 and September 2004, 421 orthotopic liver transplantations were performed in our center. For the current study we have excluded pediatric recipients (n 88), retransplantations (n 83), split/reduced-size liver transplants in adults (n 5), and cases with intraoperative death (n 4), resulting in a series of 241 adult full-size liver transplants. Donor-related data including age, gender, body mass index (kg/m 2 ), type of donor (heart beating vs. nonheart beating), and surgical variables related to the organ procurement were obtained from our prospective hospital database, organ reports, and donor information forms from Eurotransplant. Data on anatomical injuries were obtained from organ reports from Eurotransplant completed at the time of organ procurement, as well as from quality forms and operative notes completed by the transplant surgeon after transplantation. Recipient-related data including intraoperative and postoperative complications and patient and graft survival were obtained from our prospective hospital database and from patient files. Outcome variables analyzed in recipients after transplantation included bleeding complications, hepatic artery thrombosis (HAT), and graft survival. All recipients received immunosuppression based on calcineurin inhibitors and a rapid steroid taper. Ultrasound Doppler examination was performed routinely on postoperative days 1, 4, and 7 to check patency of the hepatic vessels. After the first week, ultrasound Doppler examination was performed only when there was a clinical suspicion for vascular complications. Operative Technique for Liver Procurement All livers were procured using the rapid cooling technique as described by Starzl et al. 4 In summary, after thoraco-laparotomy the distal aorta (with or without cannulation of the portal vein) was cannulated for flushing of the abdominal organs with ice-cold preservation solution (University of Wisconsin or histidinetryptophan-ketoglutarate solution). In addition, organs were topically cooled using ice-cold saline. Dissection of the liver and its vessels was performed either before or after the viscera were flushed, depending on the preference of the surgeon and hemodynamic stability of the donor. Definitions Surgical Injuries Surgical injuries of donor grafts were categorized as parenchymal, vascular, or bile duct injuries. Parenchymal injuries were subdivided into capsule tears, parenchymal ruptures, and subcapsular hematomas. Vascular injuries were subdivided into arterial, venous, and portal injuries. Arterial injuries were defined as intimal dissection, or erroneous partial or complete transsection of the celiac trunk, common hepatic artery, (aberrant) right hepatic artery, or (aberrant) left hepatic artery. Biliary lesions included stripping of the extrahepatic bile duct, ligation of the bile duct, and too short transection of the bile duct. The latter was defined as a very short bile duct that did not allow a normal duct-to-duct anastomosis, making a Roux-en-Y hepaticojejunostomy necessary. Surgical Retrieval Teams In the Netherlands, procurement of abdominal organs for transplantation is covered by 5 regional surgical retrieval teams, including 2 in the western part of the country and 3 in the eastern part of the country. These multiorgan retrieval teams are responsible for all multiorgan procurements within a designated geographical area, independently of whether organs are allocated to recipients within that area or in other parts of the country. Donor livers accepted from outside of the country are usually procured by a local foreign team. Distinction was made between donor surgeons who are liver transplant surgeons and donor surgeons who are not involved with liver transplantation. Retrieval Regions Based on the organization of organ procurement in the Netherlands, multiorgan retrieval teams were classified into 3 groups: national region 1, national region 2, and international. Within a multiorgan retrieval team, the responsible surgeon can be either someone with experience in liver transplantation or not. International teams included teams from all other countries within the Eurotransplant area (Belgium, Luxembourg, Germany, Austria, and Slovenia). Seven livers were procured by centers outside the Eurotransplant area (Switzerland, Italy, Poland, and Slovakia). Statistical Methods All statistical analyses were performed using SPSS version 12.0 (SPSS Inc., Chicago, IL). Data for continuous variables were expressed as mean standard deviation. Categorical variables were analyzed with chisquare test and continuous variables were analyzed with Student s t test (parametric) or Mann-Whitney U test (nonparametric), depending on the distribution. Time-to-event variables were studied with Kaplan- Meier survival analysis. A P value less than 0.05 was considered statistically significant. RESULTS Frequencies and Types of Hepatic Procurement Injury Characteristics of the donors are summarized in Table 1. There were no surgical injuries detected in 159 of the 241 donor livers (66%), whereas 96 anatomical injuries were detected in 82 donor livers (34%). In 14 livers a combined injury of the parenchyma and a vasculature structure or the bile duct was found. Most frequently occurring injuries were liver capsule tears (28/241;
SURGICAL INJURIES OF DONOR LIVERS 1367 TABLE 1. Characteristics of the 241 Donors Characteristic Number (%) Mean (range) Gender Male 115 (48%) Female 126 (52%) Status of donor patient Heart beating 239 (99%) Non-heart beating 2 (1%) Age, years 44.5 (12-72) BMI (kg/m 2 ), SD 24.3 3.2 (15-46) Abbreviation: BMI, body mass index. TABLE 2. Specification of Procurement Injuries of Donor Livers Percentage of Injured Livers Percentage of All Donor Livers Injuries Numbers (n 82) (n 241) Parenchymal* 50 61% 21% Capsule tear 28 34% 12% Parenchymal rupture 16 20% 7% Subcapsular hematoma 6 7% 2% Vascular 40 49% 17 Arterial 30 37% 12% Right hepatic artery 14 Celiac trunk 6 Left hepatic artery 5 Common hepatic artery 4 Right and left hepatic artery 1 Venous 8 10% 3% Portal 4 5% 2% Bile duct 6 7% 2% Stripped 3 4% 1% Ligated 2 2% 0.7% Too short 1 1% 0.3% *In 14 donor livers, parenchymal injury was detected in combination with another type of injury: arterial (n 7), venous (n 2), portal (n 1), arterial and venous (n 1), arterial and bile duct (n 1), and bile duct (n 2). Two livers had combined injuries (arterial-venous and arterial-portal). 12%) and arterial injuries (30/241; 12%). Of the arterial lesions, injury of the right or left hepatic artery was most frequent (20/30; 67%) (Table 2). The incidence of arterial injury in a liver with aberrant arterial anatomy was 26.8% (11/41), whereas in livers without aberrant anatomy it was 9.5% (19/200) (P 0.002). All injuries were secondary to surgical handling during procurement and were not the result of the trauma that may have caused death of the donor patient. Only 20 of the 96 (21%) injuries were reported by the procurement team. All other lesions were discovered by the transplant team. Although all surgeons responsible for liver procurement had been trained in multiorgan retrieval, the majority of them had no experience with liver transplantation itself (65%) (Table 3). Over the 8-year study period, the frequency of hepatic procurement injury did not change significantly. Risk Factors for Hepatic Procurement Injury Comparing the effects of various potential risk factors for surgical damage, only the procurement region was significantly related with surgical hepatic injuries during procurement (Table 3). Injuries were encountered in 27/116 (23%) of the liver grafts procured in national region 1, compared to 27/59 (46%) in national region 2, and to 28/66 (42%) of the liver grafts procured abroad (international) (P 0.003). Other factors such as donor age, body mass index, and whether the pancreas was also procured, did not sig-
1368 NIJKAMP ET AL. TABLE 3 Analysis of Potential Risk Factors for Surgical Hepatic Injury During Procurement Variable No injury (n 159) Injury (n 82) P Value Body mass index of the donor 25 109 (69%) 49 (31%) 25 50 (60%) 33 (40%) 0.173 Age of the donor 50 years 91 (65%) 48 (35%) 50 years 68 (67%) 34 (33%) 0.846 Donor surgeon Not involved in liver transplantation 101 (62%) 61 (38%) Liver transplant surgeon 54 (72%) 21 (28%) 0.186 Missing* 4 0 Organ procurement region National region 1 89 (77%) 27 (23%) National region 2 32 (54%) 27 (46%) International 38 (58%) 28 (42%) 0.003 Procurement of pancreas No 65 (63%) 39 (37%) Yes 92 (69%) 42 (31%) 0.320 Missing 2 1 *In 4 cases the name of the transplant surgeon was missing. In 3 cases information on whether or not the pancreas had been procured could not be retrieved. nificantly predict the occurrence of surgical injuries of the donor liver. Donor surgeons who do not transplant livers themselves were more frequently responsible for hepatic damage than surgeons who were involved in a liver transplant program (38% vs. 28%). However, this effect was not statistically significant (Table 3). Clinical Consequences of Hepatic Procurement injury Fortunately, most surgically induced liver graft injuries were minor and without clinical consequences. However, clinically relevant injuries were detected in 16/ 241 (6.6%) of the livers. Bleeding Complications Bleeding complications were observed in 9 of the 50 patients (18%) who received a liver graft with a parenchymal injury in contrast to 17 in 191 patients (9%) without a parenchymal injury (P 0.06). Capsule tears led to bleeding complications in 18% (5/28), parenchymal ruptures in 19% (3/16), and subcapsular hematomas in 17% (1/6). These bleeding complications were treated either by applying Argon coagulation or topical hemostatic agents, or by packing the liver and planning relaparotomy for removal of the gauzes. In 7 of the 9 patients with a bleeding complication caused by a parenchymal injury, 9 relaparotomies were performed to treat these complications (Table 4). Hepatic Artery Thrombosis HAT was found in 7 of the 30 patients (23%) who received a liver graft with an arterial injury, compared to 9 cases of HAT in 211 patients (4%) without an arterial injury (P 0.001). Two cases of HAT were detected intraoperatively. In 1 of these patients thrombosis of a repaired right hepatic artery was treated by thrombectomy and resection of the injured arterial segment. This was followed by anastomosing the right hepatic artery to the gastroduodenal artery of the donor. The other patient experienced intraoperative thrombosis of a repaired dominant right hepatic artery, and this patient was immediately listed for high urgency retransplantation. Two of the 7 patients with HAT underwent immediate retransplantation. In 2 other patients a relaparotomy was performed and a thrombectomy of the hepatic artery was attempted. In 1 of these patients a new reconstruction of the hepatic artery was made with a donor iliac artery conduit after thrombectomy. The other patient was immediately listed for urgent retransplantation after a similar reconstruction failed. One patient with HAT was treated conservatively, as no liver dysfunction or biliary complications developed. In total, 5 reinterventions, including 4 retransplantations, have been performed in 7 patients who developed HAT of a liver with an arterial injury (Table 4). Six of the 7 cases of HAT occurred in grafts with a repaired injury of the (aberrant) right or left hepatic artery. Bile Duct Complications All biliary injuries, including complete stripping and very short transsection of the bile duct, were treated by performing a hepaticojejunostomy at the level of the bile duct bifurcation. In none of these cases a hepaticojejunostomy would have been necessary (e.g., for primary sclerosing cholangitis) if the donor liver had not sustained a bile duct injury. None of the reconstructive procedures resulted in further complications. There were no significant differences in the 1-year
SURGICAL INJURIES OF DONOR LIVERS 1369 TABLE 4. Description of Parenchymal and Arterial Injuries Leading to a Complication Injury (n) Complication (n) Treatment (n) Parenchymal (9) Capsule tear Intraoperative bleeding without reintervention (1) Argon coagulation (1) Intraoperative bleeding with reintervention (4) Packing (4) Parenchymal rupture Intraoperative bleeding without reintervention (1) Topical hemostatic agents (1) Intraoperative bleeding with reintervention (2) Packing (2) Subcapsular hematoma Intraoperative bleeding with reintervention (1) Packing (1) Arterial (7) Right hepatic artery Intraoperative HAT (2) Thrombectomy reconstruction (1) Retransplantation (1) Postoperative HAT (2) Thrombectomy retransplantation (1) Thrombectomy reconstruction (1) Left hepatic artery Postoperative HAT (2) Retransplantation (2) Celiac Trunk Postoperative HAT (1) Expectative (1) graft survival rates of livers with or without parenchymal (78% vs. 82%, respectively; P 0.597) or livers with or without arterial injury (87% vs. 80%, respectively; P 0.398). DISCUSSION Shortages of postmortem donor livers have pushed transplant teams into exploring alternative sources of donor organs, including compromised, non-heart beating, and living donors, in the hope of increasing transplantation rates. 1,5,6 Much less attention has been paid to the influence of surgically induced hepatic injuries during procurement on graft survival and morbidity after transplantation. This study shows that surgically induced injuries can be found in 34% of all postmortem adult liver grafts transplanted in our center. Fortunately, most injuries occurred without clinical consequences. However, injury-related complications were detected in 6.6% of all liver grafts. Parenchymal and vascular injuries are the most prevalent types of injury, occurring in 21% and 17% of the livers, respectively. Most parenchymal injuries are small capsule tears or superficial lacerations at the right posterior side of the liver. Fortunately, these lesions are usually without clinical consequences. Although transplantation of livers with parenchymal injury was more frequently complicated by bleeding problems than transplantation of livers without parenchymal injury, this did not reach statistical significance. Nevertheless, these lesions can and should be avoided by careful handling and proper dissection of the liver. In contrast with the parenchymal and capsular injuries, injuries of the arterial vasculature did have clinical implications. Arterial injuries were associated with a significantly higher incidence of HAT after transplantation, compared to cases without arterial injury. Arterial injuries were significantly more frequently found in donor livers with aberrant arterial anatomy compared to livers without aberrant anatomy. Twenty out of the 30 (67%) arterial injuries were located in the (aberrant) right or left hepatic artery, and although a reconstruction of the injured vessel was performed on the back table, thrombosis occurred in 6 of these 20 cases. This could imply that the arterial injury has been more extensive than the visible lesion alone. It is our impression that arterial injuries are a sign of inexperienced and careless handling of the liver that may have resulted in more extensive damage than the visible lesions alone. The higher frequency of injuries of aberrant arteries also suggests unawareness of these anatomical anomalies by some of the procurement surgeons. This is supported by the fact that only 21% of all injuries were reported by the procurement team. Although 4 of the 7 patients with early HAT required retransplantation, we did not find a significant influence of surgical injuries on overall graft survival. This discrepancy is most likely explained by the fact that graft survival is influenced by multiple other factors and arterial injury is, fortunately, still found in only a minority of the liver grafts. Surgically induced injuries of donor organs during procurement procedures are not readily discussed or reported by surgeons. Only 2 other studies on surgical injuries of donor livers during organ procurement, 2,3 and 2 studies on kidney injuries have been reported. 7,8 In 1994, Lerut et al. 5 reported on technical problems encountered in 39 donor livers procured and shipped by another team. These investigators observed 14 major injuries in 9 of the 39 (23%) livers. Most of the injuries were similar to those found in our study, including erroneous transsection of a left or right hepatic artery or intimal dissections, as well as capsular and parenchymal damage. Severe bleeding complications occurred in 5 cases with capsular and parenchymal injury. Two grafts with serious parenchymal lesions and decapsulation had to be retransplanted in this series. In another series, described by Soliman et al., 2 parenchymal liver injury was reported in 23 of 572 (4%) liver transplant procedures. Although major bleeding
1370 NIJKAMP ET AL. problems were encountered in this series due to these lesions, all of these were treated surgically or conservatively and no patient had to be retransplanted for parenchymal liver injury. The largest series on donor organ injury during retrieval procedures is on kidney grafts and based on the United Kingdom National Transplant Database. 7 In this series, damage was reported in 1,729 of 9,014 (19%) kidneys. This percentage of organ injury is similar to the percentage found in our series. Despite the high rate of damage to kidneys at retrieval in the United Kingdom study, most lesions could be repaired, and there was no adverse effect on transplant survival. However, in 96 of the 9,014 (1%) kidneys, anatomical injuries were so severe that they were considered to be not transplantable. 7 Similar percentages have been reported by the United Network for Organ Sharing in the United States. 9 In our series of liver transplants, we did not have to discard any liver because of anatomical injuries. Remarkably, only 20 of the 96 surgical injuries were reported by the procurement teams. Although this could indicate that surgeons are frequently not aware of the injury they are causing during organ procurement, it could also reflect the reluctance to report these injuries as mistakes. A discrepancy in the reporting by procurement teams and transplant surgeons has also been reported by Lerut et al. 3 Multiorgan donations are major surgical procedures that require a high level of expertise and training, similar to any other type of major surgery. We observed a lower rate of surgical injuries by procurement teams in which surgeons participated who were also involved in liver transplant procedures, compared to those who had no experience with this type of transplantation. This difference was not statistically significant, but it is a trend that may become statistically significant in a larger study. We did find a significant differences in the rate of injuries among different regional procurement teams. It is not likely that the unexpectedly high rate of surgically induced liver graft injuries in our series is due to a local or national problem. Many livers transplanted in our center are procured by teams outside The Netherlands and the percentage of injuries found in those livers was similar to those found in parts of our country. In the United Kingdom report on kidney injuries by Wigmore et al., 7 a higher rate of kidney damage was found for retrieval of only kidneys by a renal team, compared to retrieval by a liver team when both kidneys and liver were procured. These observations collectively suggest that experience of the operating surgeon is an important factor influencing organ injury during procurement. Interestingly, we did not find a higher rate of hepatic injury when the pancreas was procured as well. Although pancreas retrieval can be technically challenging due to its retroperitoneal localization and the shared arterial vasculature with the liver, this apparently does not necessarily lead to a higher amount of surgical injuries. Although in the United Kingdom National Transplant Database analysis the relative frequency of kidney damage was found to be increased with donor age,7 we found no significant effect of donor age or body mass index on the incidence of liver graft injury. This study emphasizes the need of a low threshold for reporting all surgical injuries during organ retrieval by the procurement team. This will allow the transplant teams to better anticipate and prepare for surgical repair or reconstructions during the back table procedure prior to transplantation. In addition, this analysis emphasizes the need for adequate surgical training of procurement surgeons. A few years ago, the Netherlands Transplant Foundation initiated an annual hands-on training course for surgeons participating in the regional procurement teams. This initiative was recently adopted and transferred to a European level by the European Society for Organ Transplantation. Hopefully, these educational activities will contribute to a reduction of the rate of surgically induced injury of donor organs, thereby contributing to the overall success of transplantation. In conclusion, anatomical injuries of donor livers during organ procurement are not infrequent and may lead to increased morbidity after transplantation, especially bleeding complications and HAT. Increased awareness of this problem and more adequate training of procurement surgeons are necessary to reduce the incidence of these iatrogenic injuries and to improve outcome after transplantation. REFERENCES 1. Adam R, McMaster P, O Grady JG, Castaing D, Klempnauer JL, Jamieson N, et al. Evolution of liver transplantation in Europe: report of the European liver transplant registry. Liver Transpl 2003;9:1231-1243. 2. Soliman T, Langer F, Puhalla H, Pokorny H, Gruenberger T, Steininger R. Parenchymal liver injury in orthotopic liver transplantation. Transplantation 2000; 69: 2079-2084. 3. Lerut J, Reding R, de Ville de Goyet J, Baranski A, Barker A, Otte JB. Technical problems in shipped hepatic allografts: the UCL experience. Transplantation 1994;7:297-301. 4. Starzl TE, Miller C, Broznick B, Makowka L. An improved technique for multiple organ harvesting. Surg Gynecol Obstet 1987; 65:343-348. 5. Busuttil RW, Tanaka K. The utility of marginal donors in liver transplantation. Liver Transpl 2003;9:651-663. 6. Edmond JC, Freeman RB, Renz JF, Yersiz H, Rogiers X, Busuttil RW. Optimizing the use of donated cadaver livers: analysis and policy development to increase the application of split-liver transplantation. Liver Transpl 2002;8:863-872. 7. Wigmore SJ, Seeney FM, Pleass HCC, Praseedom RK, Forsythe JL, for the Kidney Advisory Group. Kidney damage during organ retrieval: data from UK National Transplant Database. Lancet 1999;354:1143-1146. 8. Eschwege P, Droupy S, Blanchet P, Hammoudi Y, Laassou K, El Hadj A, et al. Surgical injuries occurring during kidney procurement performed by a renal transplantation team. Transplant Proc 2002;34:844. 9. Davis C, Mar C. Reduction of organ-retrieval damage and organ-discard rates. Lancet 1999;354:1136-1137.