Toward Current Standards of Donor Right Hepatectomy for Adult-to-Adult Live Donor Liver Transplantation Through the Experience of 200 Cases

Abstract

Objective:

To define the current standards of donor right hepatectomy, including the middle hepatic vein for adult-to-adult live donor liver transplantation.

Summary Background Data:

Donor morbidity and mortality are inevitable given the ultra-major nature of the donor operation. Results from a matured center could define the true impact of this donor procedure most accurately.

Patients and Methods:

From May 9, 1996 to April 13, 2005, 200 consecutive donors underwent donor right hepatectomy at the University of Hong Kong Medical Center. All right liver grafts except one included the middle hepatic vein. Donor characteristics, operation time, blood loss, hospital stay, laboratory results, and complications graded by Clavien's classification divided into four eras (each consecutive 50 cases) were compared.

Results:

Donor characteristics of the four eras were generally comparable. Operative outcomes improved progressively through the four eras. From era 1 to era 4, operation time decreased from 598 minutes (range, 378–932 minutes) to 391 minutes (range, 304–635 minutes). Blood loss also decreased from 500 mL (200–1600 mL) of era 1 to 251 mL (range, 95–595 mL) of era 4. Overall complication rate was 20.5% (41 of 200). Complications rates from eras 1 to 4 were 34%, 16%, 16%, and 16%, respectively. The most common complications were of grade I (24 of 41, 58.5%). A late donor death occurred in era 4 from the development of a duodenocaval fistula 10 weeks postoperation, giving a donor mortality of 0.5% (1 of 200).

Conclusions:

This study validated the estimated morbidity and mortality of donor right hepatectomy of 20% and 0.5%, respectively. The data provide reference for counseling potential donors and setting the standards of donor right hepatectomy in the current era.

In a consecutive series of 200 donor right hepatectomies including the middle hepatic vein, the learning curve started to plateau after the first 50 cases. This experience from an established live donor liver transplant center validated the estimated morbidity of 20% and mortality of 0.5% of this ultra-major donor operation.

Striking a balance between donor safety and recipient survival could be a challenge to adult-to-adult live donor liver transplantation (ALDLT). To provide the recipient with a liver graft of an adequate size, the right liver graft is often used. However, donor right hepatectomy is a formidable surgical operation that represents removing twice as much, or one third more, of the liver, as compared with donor left hepatectomy. It has been shown that donor right hepatectomy carries a higher morbidity.^1,2 Among the 10 known donor deaths, 6 were right liver donors,^3–8 2 left liver donors,^9,10 and 2 were not formally reported.^11,12 On top of this, inclusion of the middle hepatic vein (MHV) in the right liver graft has been an issue of debate.¹³ In our center, including the MHV¹⁴ while preserving the segment 4b venous drainage¹⁵ is nonetheless a radical right liver graft design. Conversance of skills in executing the procedure should improve the surgical outcomes. The real impact of this donor procedure could only be assessed by donor outcomes that are not adversely influenced by technical immaturity or errors. Donor right hepatectomy for ALDLT came into practice since mid-1996.¹⁶ The learning curve of this procedure started to plateau after 50 cases in our previous study involving 100 cases.¹⁷ Approaching 200 cases, the objective of this study was to scrutinize the inherent risks of the procedure unrelated to technical reasons. This should provide essential information for potential donors and liver transplant centers in matching the standard of care which could be achieved.

PATIENTS AND METHODS

From May 9, 1996 to April 13, 2005, all donors who underwent donor right hepatectomy for ALDLT were included. Donor right hepatectomy included the MHV in all except 1 donor. Donor evaluation and the technique have been described elsewhere.^14,15,18 In brief, “real-time” operative cholangiography under fluoroscopy was performed via cystic duct cannulation to determine the most appropriate site of the division of the right hepatic duct and avoid stenosis of the confluence of the hepatic ducts. The liver was transected at the midplane by ultrasonic dissector without inflow or outflow vascular occlusion. The midplane of the liver was determined by temporary occlusion of the right hepatic artery and right portal vein. On approaching the liver hilum, the right hepatic duct together with the surrounding hilar plate were divided at the most appropriate site as determined by operative cholangiography. The transection of the liver was performed all the way down to the root of the MHV at the junction with the left hepatic vein or the inferior vena cava. The segment 4 hepatic vein entering the MHV near the root of the MHV was preserved as far as possible. The MHV was transected proximal to the segment 4b hepatic vein. All the operations were performed by 3 senior surgeons (S.T.F., C.M.L., C.L.L.). Meticulous medical management and nursing care were delivered to expedite donor recovery. Donors stayed in the intensive care unit the day of operation and were transferred to the surgical ward on the first postoperative day. Three doses of intravenous augmentin (SmithKline Beecham Pharmaceuticals, Philadelphia, PA) were administered perioperatively. Prophylaxis against stress ulcer was made by H₂ antagonist, and since case no. 165, proton pump inhibitor. Hypophosphatemia was avoided by prompt phosphate supplements.¹⁹

Data collection was prospective. This included donor characteristics, operative data, perioperative course, and morbidity of all degrees. Operative data included duration of operation and blood loss. Perioperative course included intensive care unit and hospital stay. Recovery of liver functions, namely, serum total bilirubin, alanine aminotransferase (ALT), aspartate aminotransferase (AST), and prothrombin time was recorded. Complications in donors were recorded and graded by the Clavien's system.²⁰ This standardized classification of complications allowed shoulder-to-shoulder comparison of results of different eras. Donors with more than one complication were graded by the more serious complications.

Statistical Analysis

Data were expressed in median and ranges. Nonparametric tests were used in the data analysis. Overall comparisons with respect to continuous variables were performed by the Kruskal-Wallis test. If the result was positive, the Mann-Whitney U test was applied for binary comparisons between groups. Parameters in assessing outcomes of donor right hepatectomy were arbitrarily divided into four eras, each 50 consecutive cases, for comparison. SPSS version 12.0 (SPSS, Inc., Chicago, IL) was used in all statistical analyses. A P value <0.05 was considered statistically significant.

RESULTS

Donor Characteristics

Over this period of 8 years and 11 months, 200 consecutive donor right hepatectomies were performed in the University of Hong Kong Medical Center for right liver ALDLT. All except one right liver graft (case 85) included the MHV. In this donor, the segment 3 hepatic vein drained into the MHV at a level more caudal than the segment 8 hepatic vein. A decision was made to preserve the MHV with the donor, and only the segment 8 hepatic vein was included in the graft and anastomosed to the inferior vena cava.¹⁴

Donor characteristics are listed in Table 1. The donors were by definition healthy adults not older than 60 years (oldest in this series, 57 years). There was no significant difference in terms of gender ratio, body weight, body height, and body mass index of the four eras. Donor age of era 4 was slightly younger than that of era 1 (P = 0.044). The remnant left liver volume ratios were also slightly larger in eras 3 and 4.

TABLE 1. Donor Characteristics of the Four Eras

Operative and Early Postoperative Data

Outcome measures including operation time, blood loss, intensive care unit stay, and hospital stay are listed in Table 2. Figure 1 shows the operation time of the entire series plotted against case number. The trend toward shorter operations in the later part of the series was obvious and statistically significant (P = 0.000). Blood loss in era 4 was half of that of era 1. Only one donor who was in era 1 required blood transfusion. The median intensive care unit stay was 1 day except for era 1 (2 days). The median stay in hospital was 10 days (range, 5–38 days) of era 1 and shortened to 6.5 days (range, 4–16 days) of era 4.

TABLE 2. Donor Outcomes and Complications of the Four Eras

FIGURE 1. Operation time of donor right hepatectomy of the 200 cases.

Perioperative Liver Biochemistry and Hematology

Median serum total bilirubin, ALT, AST levels, prothrombin time, and hemoglobin prior to operation to the seventh postoperative day of the entire series of the four eras are plotted in Figures 2a, 2b, 2c, 2d, and 2e, respectively. A trend of earlier normalization of serum total bilirubin, ALT, AST levels was evident in era 4. Prothrombin time of era 4 normalized earlier when compared with era 1 (Fig. 2d). This trend was similar for hemoglobin (Fig. 2e).

FIGURE 2. A, Serum total bilirubin levels before operation and at the first operative week. B, Serum alanine aminotransferase levels before operation and at the first operative week. C, Serum aspartate aminotransferase levels before operation and at the first operative week. D, Prothrombin time before operation and at the first operative week. E, Hemoglobin before operation and at the first operative week.

Morbidity and Mortality

There was no significant difference in complication rates for all grades according to the Clavien classification (Table 3). The overall complication rate was 20.5% (41 of 200). The complication rates in various grades were 12% (grade I), 4% (grade II), 1.5% (grade IIIa), and 2.5% (grade IIIb). In the four eras, the complication rates were 34% (era 1), 16% (era 2), 16% (era 3), and 16% (era 4), and the difference was not statistically significant. The most common grade I complication was wound infection 15 of 24 (62.5%).

TABLE 3. Donor Complications Graded by Clavien System²⁰

In era 1, 6 donors had wound infections and among whom, 2 required antibiotic treatment and were thus in grade II. A donor with bleeding duodenal ulcer underwent upper endoscopy. More serious complications of grade IIIb occurred in 4 donors. They all required surgical intervention. The donor with intraoperative portal vein thrombosis had to undergo extraction of thrombus in the portal vein and was well afterward.

In eras 2 and 3, the complication rates were both 16% (8 of 50). The complication of biliary stricture in era 2 had to be tackled with endoscopic retrograde cholangiopancreatography. Complications in era 3 were restricted to grade I, which did not require therapeutic interventions.

In era 4, grade I complications were all wound infections. Avoidable complications in other eras, such as pressure sore and peroneal nerve palsy, were absent. Nonetheless, urinary tract infection, pleural effusion, and dog-ear appearance of the wound still occurred. A 50-year-old female donor without history of peptic ulcer and asymptomatic after the donor operation was readmitted with fever, chills, and hematemesis 10 weeks after donor right hepatectomy. Tachycardia, hypotension, and anemia all pointed to the diagnosis of upper gastrointestinal bleeding. She developed convulsion and cardiac arrest upon upper endoscopy. Attempts of resuscitation were futile. Postmortem revealed a duodenocaval fistula (Figs. 3), which was conducive to air embolism during upper endoscopy, resulting in rapid demise of the donor. She had received 2 weeks of H₂ antagonist after the donor operation. This was a late donor mortality (0.5%).

FIGURE 3. A, Duodenocaval fistula on the anterior wall of inferior vena cava (arrow). B, Duodenocaval fistula on the posterior wall of the first part of duodenum (arrow).

DISCUSSION

This is the largest single-center consecutive series of donor right hepatectomy including the MHV (except case 85). Purity of data was attributable to the single center with the operation done by 3 experienced surgeons. We have been upholding the policy that donors deserve treatment of the highest standard. Hence, residents had not been the chief operators in all these 200 cases. The highest possible standard achievable thus helps to define more precisely the field strength and substantiates right liver ALDLT as a treatment option for patients with terminal liver failure.

Practice makes perfect. For liver transplantation, which is not an exact science, working toward perfection may not be the immediately achievable goal. Nevertheless, we attempted to render the operation safe by upgrading the standard of the operation. The standard of a surgical operation could be due to many factors attributed to patients, surgeons, anesthesiologists, nurses, and supporting staff.

Patient factors will not vary as only healthy subjects are candidates for this operation.²¹ In contrast to the case of deceased donors, compromise is not allowed for accepting marginal live donors.¹⁰ In this regard, fatty livers not only jeopardize recipient success, but also donor safety because of less functional parenchyma in the remnant liver. We accept at most 10% fatty change. On the other hand, fatty livers are not a common condition among our local population. So far, we have excluded about 3% of the donors due to fatty liver after evaluation (unpublished data).

Because of the close relationship between the donor and the recipient, donor recovery is affected by the recipient outcome. For instance, an early discharge of the recipient may result in an early donor recovery, and vice versa. The same relationship is also evident in the recipient total hepatectomy as the right liver graft is not devascularized until the native liver is ready for explantation. In other words, the duration of donor right hepatectomy is also related to the smoothness of recipient total hepatectomy. The neatness in procuring the right liver graft is conducive to the success on the recipient, as a result of easier vascular and bile duct anastomoses and lower incidence of bile leakage.

The surgeon factor is the most complicated issue. The experience of major hepatectomy of liver tumor is indispensable in the acquisition of the necessary skill.²² The anesthesiologist is also to a large extent like the surgeon in this aspect. Nursing and supporting staff in conjunction with the surgeon look after the donor perioperatively. There is enough evidence showing that a higher hospital volume²³ and higher surgeon volume²⁴ contribute to a better outcome in complicated surgical operations.

The experience of the first 100 right liver LDLTs has been elaborated.¹⁷ It took 50 cases for the learning curve to plateau in our center.¹⁵ Major changes in the donor operation included determination of the right hepatic duct transection line by intraoperative cholangiography.²⁵ Technical modification based on segment 4 venous drainage pattern has been implemented to improve donor safety.¹⁵ Liver transection has been done by Cavitron Ultrasonic Surgical Dissector,²⁶ facilitated by low central venous pressure. Careful transection defining vessels and the minor bile duct could be a major factor for zero clinical biliary leakage in these 200 donors, in contrast to other series with leakage of 5% to 10%.²⁷ For graft implantation in the recipient, hepaticojejunostomy was changed to duct-to-duct anastomosis.²⁸ Nevertheless, with the extensive experience in hepaticojejunostomy, donor safety was not compromised in the step of severing the right hepatic ducts. In other words, we do not insist on a duct-to-duct anastomosis when closure of the donor left hepatic duct stump becomes hazardous. Abdominal drains were not used.²⁹ One may reckon that with the learning curve coming to a plateau, subsequent developments ever possible should be attributed to a change of practice. Of course, new techniques would require practice to go through the learning curve in a smaller scale.

A major criterion for assessing maturity of technique is the blood loss of the procedure. This has reduced from 500 mL (range, 200–1600) mL to 251 (range, 95–595 mL) mL from era 1 to era 4. Thus, dropping the unnecessary practice of cellsaver and autologous blood collection was justifiable. Type and screen was done to allow expeditious blood transfusion should it deem necessary.

The lower serum total bilirubin level from day 2 to day 5 could be a result of a slightly larger remnant left liver ratio in era 4 (Fig. 2a). The significantly lower serum ALT (Fig. 2b) and AST (Fig. 2c) reflects less intraoperative injury to the liver remnant. ALT is specific for liver cell damage and its serum level is a reflection of technical precision. AST is less specific for the liver and is also reflecting tissue damage in general. AST has a short half-life of 17 hours as compared with 47 hours of ALT. Thus, the lower ALT level extended to day 4 (Fig. 2b). In era 4, the prothrombin times were longer preoperatively and postoperatively. Furthermore, hemoglobin was lower perioperatively. This observation may be explained by a slightly higher proportion of female donors in era 4. The mean prothrombin time was 11.8 seconds in male donors and 12.1 seconds in female donors. The higher proportion of female donors in era 4 might then result in a slightly longer prothrombin time perioperatively.

Fluid management during operation is crucial. A low intravascular volume helps to minimize blood loss during parenchymal transection. Should the central venous pressure remain high, the Pringle maneuver becomes useful.³⁰ Application of the Pringle maneuver has been shown to be safe not only in liver tumor surgery³¹ but also in donor hepatectomy without any negative effects on graft quality.³² In one series, the donors with Pringle maneuver had less blood loss and shorter hospital stay.³³ We have not employed such procedure as blood loss has become less than 300 mL and hospital stay only 1 week after the first 100 cases (Table 2). Other means to lower the central venous pressure, includes lowering the tidal volume³⁴ and control of the infrahepatic inferior vena cava.³⁵ From our experience, rapport between the anesthesiologist and the surgeon is crucial for liver transection without inflow vascular occlusion. Confidence in control of bleeding by the surgeon reassures the anesthesiologist of not over-loading the donor with intravenous fluid.

Donor morbidity and mortality are the crux of ALDLT. Morbidity was at an acceptable level beyond the first 50 cases. Donor mortality under any circumstances is catastrophic. There are now 10 known donor deaths reported in the literature.^3–12 While an estimated donor mortality of 0.5%³⁶ is now an acceptable standard by the community,³⁷ a lower rate may not be achievable given the magnitude of the procedure. Stipulated in the international guideline that donors ought to be healthy, the younger donors are healthier than those already in their 50s. Thus, the absence of comorbidities could be regarded as a minimal requirement. This series reascertained the quoted donor mortality ranging from 0.2%⁸ to 1%.³⁸ The 0.5% donor mortality could now be viewed as an inevitable price the society should pay for saving the life of thousands of recipients. There are complications that will occur regardless of suitable precautions, such as pulmonary embolism,^9,10 gas gangrene of the stomach,⁵ and intracerebral hemorrhage.⁶ Donor mortality also occurred in a very experienced center.⁷ Undoubtedly, any errors from inexperience or substandard medical care are unacceptable. Therefore, the reporting and sharing of experience ought to be emphasized. Having learned from the donor mortality in our series, which is the 11th worldwide, we changed our practice. Any suggestion of peptic ulcer symptoms (nocturnal and hunger pain) and signs (pallor and anemia) mandates upper endoscopy examination. Donors are also covered perioperatively with proton pump inhibitors. Nonetheless, a report of the case of donor mortality in New York suggested that acid-blocking medications should probably be discontinued just prior to resuming a regular diet.⁵ There should be no hesitation to reject donors with minimal abnormality found during evaluation, irrespective of the condition of the recipient.¹⁰

CONCLUSION

A 20% morbidity and 0.5% mortality of the liver donor could be expected for donor right hepatectomy. This is a standard to be and could be achieved. Careful and thoughtful workup process provides better preparation for the donor for this ultra-major operation. Based on the experience from our series, the standards of practice of donor right hepatectomy could be as follows: 1) careful assessment of donor health, remnant left liver and right liver graft quality and adequacy; 2) avoidance of bleeding and need for blood products (including blood, platelets, fresh frozen plasma, and albumin) perioperatively; 3) precise transection line preserving maximal amount of functional liver for the donor; 4) no compromise of the biliary system of the remnant left liver; and 5) optimum perioperative care expediting donor recovery.