Revisiting the Safety of Living Liver Donors by Reassessing 441 Donor Hepatectomies: Is a Larger Hepatectomy Complication-Prone?

Authors


Abstract

Donor safety is of paramount importance in performing living donor liver transplantation (LDLT). We retrospectively reviewed donor medical records to confirm whether larger donor hepatectomy is absolutely complication-prone. A total of 441 living donor hepatectomies were performed between October 1996 and July 2012 in our institute, which were divided into three eras (Era I, October 1996 to March 2004; Era II, April 2004 to March 2008; Era III, April 2008 to July 2012) and the incidences of postoperative complications were compared among the three types of hepatectomy—right hepatectomy (RH), left hepatectomy (LH) and left lateral segmentectomy (LLS). Although severe complications (Clavien's grade 3 or more) frequently occurred in RH in Eras I and II (15.4% and 10.7%, respectively), the incidence in Era III decreased to the comparable level observed in LH and LLS (5.4% in RH, 2.3% in LH and 5.3% in LLS). The incidence of postoperative complications did not relate to the type of hepatectomy selected in the latest era. Since most complications after hepatectomy were considered preventable, step-by-step meticulous surgical procedures are a prerequisite for further assuring donor safety irrespective of the type of hepatectomy selected.

Abbreviations
LDLT

living donor liver transplantation

LH

left hepatectomy

LLS

left lateral segmentectomy

PT-INR

prothrombin time-international normalized ratio

RH

right hepatectomy

TB

total bilirubin

Introduction

Since the initial cases of living donor liver transplantation (LDLT) were performed [1, 2], more than 20 years have already passed and LDLT has become one of the standard therapeutic modalities for patients with end stage liver disease [3-6] as well as hepatocellular carcinoma [7-9]. The safety of living liver donors is of paramount importance and this issue has been aggressively discussed so far [10-16]. Although the liver has a capacity to regenerate, it is generally believed that the larger the resected liver parenchyma is, the more frequently postoperative complications will occur [13, 14, 17, 18]. We implemented the LDLT program in 1996 and have performed more than 400 LDLTs so far. In our experience, as observed in other transplantation centers, the incidence of postoperative complications after right hepatic donation was higher than that of left-sided hepatic donation [13, 18]. That is, a larger hepatectomy was considered complication-prone. Moreover, the recipient survival after left hepatic grafting was comparable to that after right hepatic grafting [13, 18]. We had strongly advocated the use of left-sided hepatic grafts for the sake of donor safety. Perioperative managements and operative procedures of living liver donors seemed to be established with cumulative experience. In this paper, we revisited the safety of living liver donors from the experience of more than 400 LDLTs focusing on postoperative complications after donor hepatectomy.

Methods

The medical records of 441 living liver donors who underwent a donor hepatectomy in our institute between October 1996 and July 2012 were retrospectively reviewed. All LDLTs were performed after obtaining approval from the Ethics and Indications Committee for LDLT. The donors were divided into the following three eras mainly according to Japanese year terms which begins on April 1: Era I, between October 1996 and March 2004; Era II, between April 2004 and March 2008; Era III, between April 2008 and July 2012. Donors who underwent a posterior sectionectomy were excluded from the study because of the small number (n = 6). All the donors were followed for at least 1 year. The incidences of postoperative complications, intraoperative blood losses of the three types of hepatectomy—right hepatectomy (RH), left hepatectomy (LH) and left lateral segmentectomy (LLS)—were compared in each era. The levels of total bilirubin (TB) and prothrombin time-international normalized ratio (PT-INR), the lengths of postoperative hospital stays were compared among the three types of hepatectomy in Era III. Postoperative complications were graded according to the Clavien's classification [19].

Psychiatric complications, such as delirium, depression or alopecia areata, were not discussed in the current study because of the difficulty in grading these complications. All superficial wound problems, such as wound dehiscence, wound infection or keloid, were regarded as grade I, even though some donors with such problems underwent repair surgery under local anesthesia later. Transient hyperbilirubinemia and prolongation of prothrombin time were not considered morbidity. Bile leaks were graded according to the international study group of liver surgery [20].

The hepatectomy procedures and perioperative managements were the same ways as previously described [12, 13, 21]. Our strict criterion for performing donor RH was that preoperative volumetric analyses confirmed the remnant liver volume was estimated to be no less than 35% of the total liver volume. Donors had to have normal transaminase levels in preoperative evaluations. Candidate donors whose transaminases were beyond the normal ranges were rejected. Most of these abnormalities were caused by steatosis of the liver. We offered the rejected candidate donors a diet program in order to normalize transaminase levels.

Prevention of venous thromboembolism was provided through the use of intermittent pneumatic compression and compression stockings and through encouraging early ambulation.

NCSS 2007 software [22] was used for statistical analyses. Proportions were compared using chi-square test, while numerical measures were compared by one-way analysis of variance (ANOVA) or t-test. A p-value of less than 0.05 was considered to be statistically significant.

Results

Postoperative complications of the three types of hepatectomy in each era

Postoperative complications frequently occurred after RH in Eras I and II (Figure 1). However, the incidence dramatically dropped in Era III. The incidences of postoperative complications in Era III were comparable among the three types of hepatectomy. We did not encounter any Clavien's grade IV or V complications throughout our entire experience. When we defined Clavien's grade III or more as a severe morbidity, the incidence of postoperative severe morbidity after RH gradually decreased (15.4% in Era I, 10.7% in Era II, 5.4% in Era III). The details of postoperative complications are summarized in Table 1. Clavien's grade III complications consisted mainly of bile leak and biliary stricture. Complications related to upper gastrointestinal tract, such as gastric stasis and duodenal ulcer were frequently observed after LH, while pleural cavity-related complications, such as pneumothorax and pleural effusion, occurred only after RH.

Figure 1.

The incidences of postoperative complications of each era in each graft type. Postoperative complications were graded according to Clavien's classification system. When a donor had two or more complications, only the complication with the highest grade was counted. LH, left hepatectomy; LLS, left lateral segmentectomy; RH, right hepatectomy.

Table 1. Details of postoperative complications of each type of hepatectomy in each era
GradeComplicationRHLHLLS
Era I (n = 26)Era II (n = 56)Era III (n = 56)Era I (n = 104)Era II (n = 89)Era III (n = 86)Era I (n = 12)Era II (n = 7)Era III (n = 19)
  • Complications were graded according to the Clavien classification [19]. Some donors had two or more complications.
  • Figures before the parentheses are the number of donors in each complication. Figures in the parentheses are the incidence of each complication expressed as percentage in each graft type in each era.
  • LH, left hepatectomy; LLS, left lateral segmentectomy; RH, right hepatectomy.
  • 1Bile leaks are graded according to the International Study Group of Liver Surgery [20].
  • 2Gastric stasis is caused by narrowing of the gastric outlet resulted from upward distortion of the stomach into a dead space created by LH.
ISuperficial wound problem4 (15.4)15 (26.8)4 (7.1)17 (16.3)14 (15.7)9 (10.5)2 (16.7)2 (10.5)
 Bile leak—grade A13 (11.5)1 (1.8)1 (1.1)1 (8.3)
 Pleural effusion1 (3.8)
 Edema1 (1.1)
 Refractory abdominal pain1 (14.3)
 Intra-abdominal abscess1 (5.3)
 Hoarseness1 (1.8)1 (1.8)1 (1.0)1 (1.1)1 (1.2)
 Numbness in the upper extremities1 (3.8)1 (1.0)2 (2.2)1 (1.2)
IIIntra-abdominal abscess1 (3.8)1 (5.3)
 Duodenal ulcer2 (1.9)1 (1.1)2 (2.3)
 Gastric stasis21 (1)1 (1.1)11 (12.8)1 (8.3)1 (5.3)
 Ileus1 (1)1 (1.2)
 Gastroenteritis1 (1.8)
IIIaPleural effusion requiring thoracentesis1 (3.8)1 (1.8)
 Bile leak—grade B12 (7.7)1 (1.0)3 (3.4)1 (14.3)
 Biliary stricture requiring percutaneous transhepatic cholangioplasty1 (3.8)
 Pneumothorax2 (7.7)1 (1.8)
 Iatrogenic hydrothorax1 (3.8)
 Intra-abdominal abscess1 (3.8)1 (1.8)
IIIbBiliary stricture requiring surgical repair1 (1.0)
 Biliary stricture requiring endoscopic stenting under general anesthesia1 (1.8)
 Abdominal incisional hernia1 (1.8)2 (2.2)1 (1.2)
 Intra-abdominal bleeding requiring relaparotomy1 (1.1) 
 Bile leak—grade C11 (1.1)1 (1.2)
 Refractory abdominal pain requiring adhesiolysis1 (5.3)

Biliary complications and treatments

Biliary complications and treatments are summarized in Table 2. A total of 19 donors suffered a biliary complication. There were 16 bile leaks and 3 biliary strictures. Grade A bile leaks were treated only by postponing drain removal. Grade B bile leaks needed percutaneous drainage. Grade C bile leaks were treated by surgical repair under general anesthesia. There were three biliary strictures in the current study. The incidence dramatically decreased in Era III.

Table 2. Summary of biliary complications and treatments
EraType of hepatectomyType of complicationTreatment
  1. Bile leaks are graded according to the International Study Group of Liver Surgery [20].
  2. LH, left hepatectomy; LLS, left lateral segmentectomy; RH, right hepatectomy.
IRHBile leak—grade APostponement of drain removal
  Bile leak—grade APostponement of drain removal
  Bile leak—grade APostponement of drain removal
  Bile leak—grade BPercutaneous drainage
  Bile leak—grade BPercutaneous drainage
  Biliary stricturePercutaneous transhepatic cholangioplasty
 LHBile leak—grade BPercutaneous drainage
  Biliary strictureSurgical repair
 LLSBile leak—grade APostponement of drain removal
IIRHBile leak—grade APostponement of drain removal
  Biliary strictureEndoscopic stenting
 LHBile leak—grade APostponement of drain removal
  Bile leak—grade BPercutaneous drainage
  Bile leak—grade BPercutaneous drainage
  Bile leak—grade BPercutaneous drainage
  Bile leak—grade CSurgical repair
 LLSBile leak—grade APostponement of drain removal
  Bile leak—grade BPercutaneous drainage
IIILHBile leak—grade CSurgical repair

Intraoperative blood losses of the three types of hepatectomy in each era

As shown in Figure 2, the intraoperative blood losses had decreased with the cumulative experience. The decrease in Era III was notable. The decrease was mainly attributed to the use of intermittent Pringle maneuver during liver transection as previously reported [12, 23].

Figure 2.

Intraoperative blood losses of each graft type in each era.

Kinetics of total bilirubin and prothrombin time-international normalized ratio after three types of hepatectomy

There were transient increases of TB after RH compared to LH and LLS (at days 3, 5 and 7; p < 0.001) (Figure 3). The peak level was observed at postoperative day 3. The levels of TB gradually decreased and returned to normal until 1 month after operation.

Figure 3.

Kinetics of total bilirubin after the three types of hepatectomy. Pre, preoperation.

The increases of PT-INR were observed not only after RH, but also after LH and LLS. The levels of PT-INR at days 1, 3, 5 after RH were significantly higher than those after LH or LLS at each time point (p < 0.001; Figure 4). The levels of PT-INR after LH were comparable to those after LLS throughout the observed period.

Figure 4.

Kinetics of prothrombin time-international normalized ratio (PT-INR) after the three types of hepatectomy. Pre, preoperation.

Comparisons among the three types of hepatectomy in Era III

The donor characteristics and operative results were compared in the latest era (Era III). The data are summarized in Table 3. The mean age of donors who underwent LLS was the youngest among the three types of hepatectomy. Most of the LLS donors were the father or the mother of a small recipient. Donors who underwent LH were male-dominant. LH resulted in the most increased blood loss among the three types of hepatectomy. The mean blood loss in Era III was 390 mL. We evaluated the risk factors for blood loss of more than 400 mL regarding all 161 donors of Era III, which revealed that “male donor” was a statistically significant risk factor for blood loss of more than 400 mL (χ2 8.29, p < 0.01) while “LH” was not (χ2 1.44, p = 0.23). There were no statistical differences among the three types of hepatectomy with regard to operation times. The mean lengths of postoperative hospital stays of the three types of hepatectomy were almost equal. Furthermore, there were no statistically significant differences in the incidences of postoperative complications among the three types of hepatectomy. There was no relationship between the postoperative complications and the intraoperative blood losses.

Table 3. Donor characteristics and surgical results of the three types of hepatectomy in Era III
FactorsType of hepatectomy
RHLHLLS
  • LH, left hepatectomy; LLS, left lateral segmentectomy; RH, right hepatectomy.
  • Values are expressed as a mean ± SEM.
  • 1Statistically significant by chi-square test (p < 0.001).
  • 2Statistically significant by one-way ANOVA (p < 0.001).
  • 3Complication was graded according to Clavien's classification [19]. When a donor had two or more complications, only the complication with the highest grade was recorded.
Number568619
Male/female24/32162/2417/121
Age38.4 ± 1.436.2 ± 1.131.7 ± 2.4
Estimated remnant liver volume (%)39.3 ± 0.6262.7 ± 0.5275.4 ± 1.02
Intraoperative blood loss (mL)366 ± 322431 ± 262276 ± 542
Operation time (min)413 ± 11421 ± 9347 ± 19
Postoperative hospital stay (day)12.7 ± 0.512.5 ± 0.512.6 ± 0.9
Complication rate (%)3
Grade I8.910.510.5
Grade II1.816.310.5
Grade IIIa5.400
Grade IIIb02.35.3
Total16.129.126.3

Discussion

The current study clearly showed that the incidence of postoperative complications after RH had decreased to the same level as observed in LH and LLS donors in the latest era probably because of cumulative experiences of surgical procedures and perioperative managements. Although the blood tests showed the burden of losing large liver parenchyma in RH was larger compared to LH or LLS as shown in Figures 3 and 4, the test results rapidly returned to the normal levels. Now that RH has become a safe procedure with regard to the incidence of postoperative complications, a smaller hepatectomy (LH and LLS) should not be considered safer in living liver donation. In reviewing the medical records, most complications were preventable by step-by-step meticulous surgical procedures.

“Gastric stasis” occurs when the stomach is distorted into the dead space created by LH. Donors with gastric stasis complain of abdominal discomfort and abdominal fullness caused by partial obstruction of the gastric outlet. For ulcer prevention, lansoprazole or omeprazole was routinely used during perioperative periods. These drugs were administered to the donors approximately for 1 month after operation. The five donors suffered from duodenal ulcers after cessation of these drugs. The exact reason why duodenal ulcers occur predominantly after LH has not yet been elucidated. One presumed reason is that after LH the stomach is distorted upward and attached to the raw cutting surface of the liver, which impairs blood circulation around the pylorus causing the development of duodenal ulcer. Hepatobiliary surgeons occasionally encounter such phenomena after LH for a malignant hepatic tumor as well. We applied Seprafilm™ (Kaken Pharmaceutical Co., Ltd., Tokyo, Japan) over the stomach in order to prevent the adhesions. In addition, the dead spaces were filled with the greater omentum in order to prevent the distortion of the stomach. However, these means have not yet proven to be effective. If we can lessen the incidence of these complications, the safety of living donors will further increase.

Pleural cavity-related complications such as pneumothorax and pleural effusion occurred only after RH, probably resulting from rotating the right lobe by dissecting the right triangular and coronary ligament. Pleural effusion might have resulted from deprivation of large hepatic parenchyma after RH.

Severe complications (Clavien's grade III) consisted mainly of biliary stricture and bile leak. Three donors suffered biliary stricture in the current study. One donor required reoperation to release the stricture caused by excessive suturing near the right hepatic duct after LH. One biliary stricture was treated only by percutaneous transhepatic cholangioplasty. One donor required repeat endoscopic cholangioplasty under general anesthesia. The biliary stricture of this donor was considered to be caused by excessive skeletonization of the bile duct during the hepatic artery isolation, leading to biliary ischemia. Biliary stricture caused by ischemia is considered refractory and may lead to secondary cholestatic liver disease. Extensive skeletonization of the common hepatic or bile duct should be avoided in order to prevent devastating biliary complications [12]. We introduced real-time cholangiography using a C-arm in dividing the right or left hepatic duct in the middle of Era II [12]. Real-time cholangiography had enabled surgeons to keep the remnant biliary system intact by cutting the hepatic duct under direct vision, which resulted in no biliary stricture in Era III. Meticulous closure of small Glisson branches during transection of liver parenchyma is a prerequisite for preventing bile leakage. Also, excessive use of electrocautery around bile ducts on the cutting surface in an attempt to stop bleeding should be avoided. In order to prevent exposure of relatively large intrahepatic bile ducts on the cutting surface, which may increase the possibility of major bile leaks, it is far more important to transect the liver along the precise border between the left and right hemilivers.

Several surgeons documented the accepted lower safety margin of donor remnant liver volume might be 30% of the total liver volume in LDLT [24-27]. Transplant surgeons have to set strict limitation for the safety margin of remnant liver volumes. Otherwise, a tragedy caused by an extremely small remnant liver would occur. As shown in Figures 3 and 4, metabolic and productive burdens of remnant livers proved by blood tests after RH were large compared to those after LH or LLS, even though we set the limit on remnant liver volume at 35%. There have been devastating consequences after living liver donation around the world, most of which occurred after right hepatic donation [28-32]. Enthusiasm for the use of extensive RH grafts has been recently dampened in many transplant centers since those devastating reports were published. When a small remnant liver cannot sustain metabolic or productive demand of the body, the consequence will be donor mortality. Nobody can exactly know how much remnant liver is absolutely safe. However, setting the limit of remnant liver volume at 35% of the total liver volume is absolutely safer than setting the limit at 30%. Furthermore, our policy in performing RH is that an RH graft does not include the middle hepatic vein. Losing the middle hepatic vein from the remnant liver will result in congestion of the segment IV, which may further worsen the function of the remnant liver and increase the possibility of donor mortality. Although there is concern about blood congestion of the anterior sector of RH grafts without the middle hepatic vein after reperfusion in recipients, technical advancements have enabled surgeons to safely reconstruct the drainage veins of the anterior sector in such grafts [33, 34]. The major drawback of setting the limitation at 35% is the fact that the more the limitation is raised, the more candidate donors are rejected. In fact, approximately one of every 10 candidate donors for RH was rejected in our institute because his/her preoperative volumetric analysis revealed the remnant liver volume was less than 35% of the total liver volume. Volumetric analyses were done by students in our graduate school who were given only minimum clinical information [35, 36]. Even a candidate donor whose estimated remnant volume was 34.9% was rejected. Although our strict criterion for donor RH may result in death of a candidate recipient who eventually cannot undergo LDLT, the safety of healthy living liver donors is the top priority in our LDLT program.

The donors were allowed to leave the hospital when they felt confident about living their homes themselves. Their decisions for discharge were completely voluntary. The lengths of postoperative hospital stays were comparable among the three types of hepatectomy in Era III, which indicated that systemic recovery was almost equal among the three types of hepatectomy.

Confirmation of true safety of living liver donors needs long-term evaluation. Living liver donation is one of the major abdominal operations and considerable adhesions and large operation scars in the upper abdomen are inevitable. Donors may suffer a malignant disease in the stomach, in the biliary tree, or in the liver itself in the future. Undergoing donor operation may decrease curability of such future diseases. Adhesions in the abdomen may increase the incidence of miscarriage among young female donors. Large operation scars may impair activity of daily life. These long-term impacts of living liver donation have yet to be elucidated. Personnel who participate in LDLT programs have to take responsibility for following living donors long after liver donation in order to confirm the true safety.

The management styles for living liver donors have not changed since the implementation of our LDLT program. However, there was a considerable replacement of staff surgeons at the beginning of Era II, which might have related to the ironical increase of donor complications in LH in Era II. However, the most important result in this paper is that the incidence of postoperative complications after RH had steadily decreased with the cumulative experience. Although the replacement of surgical staffs might have led to the increase of complications in LH, the complication rates in RH had decreased nonetheless. At least two donor surgeons who had sufficient experience in hepatectomy attend donor operations in our department because surgical skills in donor operations critically affect the incidence of surgical complications after donor hepatectomy. The most important thing is that staff surgeons for donor operations have to hand down their surgical skills to younger surgeons. Trainees for living donor hepatectomy, who have already had the experience of sufficient hepatobiliary surgeries, first attend donor operations as the second assistant. Then, they become the first assistant with the aid of an experienced second assistant. Finally, they become an operator with the aid of experienced first and second assistants.

In conclusion, the incidence of postoperative complications after donor RH was comparable to those after donor LH and LLS in the latest series. We believed that this was probably because of our strict criterion for RH in which a donor hepatectomy was performed only when the estimated remnant liver volume is no less than 35% of the total liver volume. Although the safety of RH was confirmed by the current study, we will continue to advocate the use of LH grafts whenever possible for the sake of donor safety. Meticulous surgical procedures are needed in order to decrease donor morbidity, because most of the complications we encountered were not related to the small volume of a remnant liver but to the step-by-step surgical procedures.

Disclosure

The authors of this manuscript have no conflicts of interest to disclose as described by the American Journal of Transplantation.

Ancillary