Telephone: 49-201-723 1101; FAX: 49-201-723 5946
Preoperative donor liver biopsy for adult living donor liver transplantation: Risks and benefits
Article first published online: 20 JUL 2005
Copyright © 2005 American Association for the Study of Liver Diseases
Volume 11, Issue 8, pages 980–986, August 2005
How to Cite
Nadalin, S., Malagó, M., Valentin-Gamazo, C., Testa, G., Baba, H. A., Liu, C., Frühauf, N. R., Schaffer, R., Gerken, G., Frilling, A. and Broelsch, C. E. (2005), Preoperative donor liver biopsy for adult living donor liver transplantation: Risks and benefits. Liver Transpl, 11: 980–986. doi: 10.1002/lt.20462
- Issue published online: 20 JUL 2005
- Article first published online: 20 JUL 2005
- Manuscript Accepted: 29 MAR 2005
- Manuscript Received: 5 JAN 2005
- Department of General Surgery and Transplantation, University of Essen, Germany
The role of liver biopsy (LB) in donor selection for adult living donor liver transplantation remains controversial, since the procedure is associated with additional potential risks for the donor. From April 1998 to August 2004, 730 potential living donors for 337 adult recipients underwent our multistep evaluation program. In 144 candidates, LB was performed. LB was obtained in a percutaneous ultrasound-guided fashion by means of Menghini needle (32 cases) or Tru-cut needle (112 cases). The biopsy specimen was preserved in 5% formalin and processed with hematoxylin & eosin–stained sections. Thirty-one (21%) of 144 candidates who underwent an LB had a positive finding at histological examination that induced their exclusion from donation, of whom 21 had liver steatosis of varying kind and grade (10%-80%) and 10 had a nonsteatotic hepatopathy (non–A-D hepatitis in 6 cases, diffuse granulomatosis in 2, schistosomiasis in 1, fibrosis in 1). The only observed major complications related to LB were 2 intraparenchymal haematomas, both of which resolved spontaneously within a few months. In conclusion, based on these findings, we believe that preoperative LB in the donor selection for adult LDLT is necessary, once the initial donor screening and noninvasive evaluation is complete. Because other screening modalities can be unreliable, without preoperative LB a fraction of potential donors will be operated on inappropriately, risking both donor and recipient. The main objective of LB should be to ensure the donor's safety more than the preservation of the graft function. (Liver Transpl 2005;11:980–986.)
Living donor liver transplantation (LDLT) has acquired widespread acceptance. From its beginnings with pediatric recipients to its expansion into right hemilivers for adults, the field has advanced maintaining donor safety as the top priority. A thorough workup of the potential living donor is guided by 2 objectives: (1) assuring the safety of the surgical procedure for the donor, and (2) identifying donor grafts that pose potential risks for the recipient (i.e., primary nonfunction, delayed primary function, and the transmission of disease).1 To date there is still no gold standard for the evaluation of the living liver donor.
There are substantial differences in how centers in different countries evaluate potential donors, including variability in the use of invasive testing. The role of liver biopsy (LB) in donor selection remains especially controversial, since the procedure is associated with additional potential risks for the donor. A survey by Brown et al.2 found that among 42 (out of 84) U.S. transplant centers that responded, only 6 (14%) perform LB for all donors, 25 (60%) perform it only in selected cases, and 11 (26%) never perform LB.3–5. Additionally, each country has different medical-legal requirements, cultural standards, and availability of deceased donors, all of which influence the approach to the living donor evaluation.
The purpose of our study was to evaluate the importance and to weigh risks and benefits of the preoperative donor liver biopsy in the selection of candidates for right adult LDLT based on the experience at our institution.
Patients and Methods
|STEP 1||Clinical evaluation: history and physical examination|
|Lab tests: blood group, hematological tests, chemistry, coagulation profile, C-reactive protein, and pregnancy test|
|Serology: hepatitis A, B, and C; HIV, CMV, HSV, EBV|
|First informed consent|
|STEP 2||Imaging studies: all-in-one CT scan|
|First psychological evaluation|
|STEP 3||Special studies: ECG, chest x-ray, pulmonary function test, echocardiography, stress test|
|Laboratory: thyroid function tests (TSH, T3, T4), immunoglobulins IgA, IgG, IgM, iron, transferrin, ferritin, α-1-antitrypsin, ceruloplasmin, tumor markers (CEA, AFP, Ca19-9), factors V, VII and VIII, protein C and S, APCR, and urine sediment|
|First autologous blood donation|
|STEP 4||Second psychological evaluation (donor and recipient together)|
|Second autologous blood donation|
|STEP 5||HLA typing, cross-match|
|Ethics board evaluation|
|Final informed consent|
As we recently reported, our evaluation protocol underwent modifications over time.6 Prior to October 2000, liver biopsies were reserved for selected cases at late phase of the evaluation process detailed in Table 1. Such cases were mostly composed of individuals in whom history or biochemical, serological, or imaging findings provided some degree of uncertainty. Additionally, during this era, the evaluation included angiography to define the vascular anatomy. Both procedures were typically performed 2 days before the planned liver resection.
After October 2000, the donor death resulting from small-for-size syndrome in a steatotic remnant liver and a last-minute cancellation of a living donor transplant after liver fibrosis was detected in the donor prompted us to implement a mandatory liver biopsy during step 2 of the evaluation only after negative results of “all-in-one” computed tomography (CT) scan and of the first psychological evaluation.
Informed consent was obtained in all cases, and biopsies were performed on an outpatient basis. Doppler ultrasound (US) evaluation of the donor's liver was routinely obtained and each procedure was done under ultrasound guidance to minimize technical complications.
No patient received analgesics or anxiolytics prior to the liver biopsy. Biopsies were obtained in a percutaneous fashion, following an intercostal approach at the eighth space along the midaxillary line. Anesthesia was with local 1% xylocaine infiltration. Menghini needles (16-G suction needle, Hepafix, B-Braun, Melsungen, Germany) were used in 32 cases prior to February 2001, and since then Tru-cut needles (18-G spring-loaded cutting needle with triggering mechanism, Bard Magnum, C.R. Bard Inc., Cavington, UK) have been used in 112 cases. Follow-up US evaluations were performed immediately and 3 hours after the biopsy to detect any active bleeding, hematomas, or other lesions. Patients were asked to lie on their right side for approximately 3 hours. Vital signs (blood pressure, heart rate, respiratory rate) were monitored every 30 minutes over 3 hours' time. Additionally, a complete blood count was obtained at the end of the 3 hours. In uncomplicated cases, the donor was discharged home with instructions to perform only light activities for the next 24 hours and to refer back to our institution for any abnormality. Biopsies and US evaluations were performed in all cases by a liver transplant surgeon. Postbiopsy pain was treated with analgesics (i.e., Metamizol-Natrium or Tramadol-HCl). In case of significant decrease of hemoglobin values the potential donor was immediately hospitalized, a US and eventually a CT scan was performed and blood transfusion were administered according to the percentage of decrease related to the clinical situation and donor age.
The biopsy specimen (2.2-2.5–mm long) was preserved in 5% formalin and subsequently processed with haematoxylin & eosin–stained sections. Biopsies (mean portal tract number of 6) were evaluated by an experienced liver pathologist. Hepatic steatosis was graded quantitatively employing a 20× objective, according to the total percentage of hepatocytes involved over the 6 tissue levels. Zonal distribution of steatosis and presence of macro- or microvescicular steatosis as well as other nonsteatotic hepatic pathologies were specifically reported. Final readings were available within 10 to 24 hours.
If biopsy findings suggested any type of liver pathology that could lead to donor or recipient compromise, the evaluation process was terminated. Patients were notified of the findings, and further investigations or follow-up undertaken when necessary.
From April 1998 to August 2004, 730 potential living donors for 337 adult recipients underwent our multistep evaluation program. Three hundred sixty-two potential donors reached step 2 of our evaluation protocol. In 144, (40% of the group, 20% of all potential donors) an LB was performed. From April 1998 up to October 2000 we performed LB just in 26 (18%) selected cases.
Two events in our program led us to implement a mandatory liver biopsy. One was the death of a 38-year-old donor with congenital lipodystrophy not diagnosed preoperatively. The donor had a body mass index (BMI) of 25 and a remnant liver-volume-to-body-weight ratio (RLVBWR) of 0.8. The recipient was affected by hereditary seldom disease (Berardinelli Seip syndrome). To exclude the presence of the same pathology in the donor, a LB was performed. The biopsy showed 30% mixed steatosis. After 4 weeks of strict dietary modification, a second LB was performed and showed 20% microsteatosis. Despite suboptimal improvement, the liver resection was performed. During the postoperative phase the donor developed a SFSS with subsequent acute liver failure. On postoperative day 27, he underwent a deceased donor LT but died intraoperatively because of acute cardiac failure.
The second event involved a 28-year-old male donor candidate with normal liver function and an RLVBWR of 0.8. An unexpected suspicion of liver fibrosis by preoperative angiography (tortuous aspect of liver arterioles) was confirmed by histology. For this reason the candidate was excluded. The designated recipient unfortunately died 1 month later on the waiting list because of decompensated hepatitis C virus and hepatitis B virus cirrhosis. In light of these two significant events, an early LB became mandatory in our evaluation protocol.
Out of 730 potential donors, 622 were excluded at different phases of the evaluation (Table 2). Out of the 144 candidates who underwent a LB, 31 (21%) had a positive finding at histological examination: 21 had liver steatosis of varying kind and grade, and in the other 10 cases, a nonsteatotic hepatopathy was detected. These findings resulted in the exclusion of these candidates from donation.
|Step||No. Excluded||% of All Excluded||% of All Donors|
In 21 cases (14.5%), hepatic steatosis was detected. The type and grade of steatosis and characteristics of the patients are reported in Table 3. Ten of these candidates demonstrated a mixed micro- and macrosteatosis of different grades (range, 10%-50%), with 9 of 10 having ≥20%. Macroscopic steatosis was diagnosed in 10 potential donors ranging between 10% and 80% and was ≥20% in 8 of 10 cases. Pure microsteatosis of 50% was observed in 1 case. In 4 of 21 cases, nonalcoholic steatohepatitis was also observed.
|% Steatosis||Mix||Macrosteatosis||Microsteatosis||BMI||Cholesterol >200 mg/dL||Age, yrs||Gender, M:F|
|10||1||2||—||25 (23-26)||1/3 (242)||52 (51-60)||3:0|
|20||2||2||—||28 (24-31)||4/4 (255)||36 (36-48)||4:1|
|30||4||1||—||30 (25-31)||0/5||38 (35-47)||3:2|
|40||2||2||—||24 (23-31)||2/4 (249)||35 (25-51)||3:1|
|50||1||1||1||29 (24-39)||0/1*||29 (18-45)||2:1|
It should be noted that these 21 candidates were mainly young men (16 males, 5 females; median age, 39). The median BMI was relative low (26), and only 5 had a BMI between 30 and 31 consistent with grade 1 obesity. Nine had mild hypercholesterolemia (225-273 mg/dL), and none had a history of alcohol use (Table 3).
Fifteen candidates were excluded based on the histological findings alone and 6 because of the association steatosis and low RLVBWR (0.5-0.7).
Five candidates with normal BMI but significant steatosis (10%-40%) underwent strict diet modification over 4 to 6 weeks (Table 4). All became eligible donors, but only 3 donated. Two did not donate because the recipient died during the waiting time.
|Case||Gender/Age||Weight 1st||Weight 2nd||BMI 1st||BMI 2nd||Histology 1st||Histology 2nd||RLVBWR||Donation|
|1||M/38||75||70||25||23||Mix 30%||Micro 20%||0.8||Yes|
|3||F/41||63||54||26||22.5||Mix 30%||Mix 5%||0.8||Yes|
|4||M/27||76||70||23||21||Mix 40% (20% macro)||<10% Macro||0.8||No, recipient death|
|5||F/37||68||62||26||24||Macro 20%||<5% Macro||0.7||No, recipient death|
None of the imaging investigations (US, MRI/all-in one-CT) showed any suspicion of liver steatosis except for 1 case in whom US demonstrated a hyperechoic hepatic parenchyma (“white liver”); in this case the histology confirmed a 40% mixed steatosis.
In 10 cases (30% of positive findings and 7% of biopsied candidates), nonsteatotic hepatopathies were detected. In 6 cases, hepatitis of unknown origin (i.e., non–A-D hepatitis) was detected. Histology demonstrated a mild lobular hepatitis in 5 cases and a triaditis similar to the reported cases of Ryan et al.7 in 1 case. In 2 young female donors, a diffuse granulomatosis of unknown origin was detected.
In 1 Egyptian candidate, a primary diagnosis of schistosomiasis was made by liver histology. Further serological investigations,which are not routinely performed in our standard evaluation protocol, confirmed the diagnosis.
As reported above, in 1 case a mild portal and periportal fibrosis, initially suspected by means of digital subtraction angiograph (DSA)-angiography, was detected in a young patient with a critical remnant liver volume. Because of these findings, all 10 candidates were excluded from donation.
In 2 cases (1.3%) we observed major complications related to LB. In one instance, a 42-year-old woman biopsied with a Menghini needle was discharged with a normal blood count and normal US findings. According to her own report, and despite instructions against it, she performed physically demanding activities (moving furniture) immediately upon discharge. An intraparenchymal hematoma was detected 2 days later, when she presented with abdominal pain. Her hemoglobin level had dropped from 14 to 8 g/dL and she was transfused with 2 units of packed red blood cells. She was discharged home after 10 days of in-hospital observation. The hematoma was almost completely resolved by 6 months.
At the time of this LB, the intended recipient, (post–hepatitis B virus cirrhosis, Child-Turcotte-Pugh Score 8, United Network for Organ Sharing status 3), was already on the waiting list for deceased donor transplant. Five months later, he developed an acute decompensation of his chronic liver failure, became United Network for Organ Sharing status 2A, and died on the waiting list. The histological finding of the original donor LB showed a normal liver parenchyma. After this complication, we changed from Menghini (16 G) to Tru-Cut biopsy (18 G) needles.
The second complication occurred in a 33-year-old woman with a BMI of 28 who wanted to donate to her mother who had primary biliary cirrhosis (United Network for Organ Sharing 2b) and stage T2 hepatocellular carcinoma. The recipient had been on the waiting list for a deceased donor liver for 2 months. The LB was performed by means of a Tru-cut needle. The initial postinterventional course was uneventful. Four days later, the patient complained of upper abdominal pain. A blood count showed a decrease in the hemoglobin of 4 g/dL, and a CT scan confirmed the clinical and US suspicion of intrahepatic hematoma. The hemoglobin continued to decrease and when it reached 6 g/dL below the prebiopsy level, 2 units of packed red blood cells were transfused. Additionally, the bilirubin level increased up to 4 mg/dL with biochemical and US findings of intrahepatic cholestasis. An endoscopic retrograde cholangiopancreatography with rinsing of the bile duct and gallbladder was performed. Hemoglobin values and cholestatic parameters normalized within a few days. Almost 3 weeks after the LB, a spontaneous slow regression of the intrahepatic hematoma was observed. Six months later the hematoma was almost completely reabsorbed and no further complication occurred. The recipient's general conditions progressively deteriorated, and she died 4 months later on the waiting list.
LDLT is guided by 2 main principles: 1) donor morbidity and mortality must be kept to a minimum, and 2) graft and recipient survival should be acceptably high, as in conventional deceased donor liver transplantation. Therefore, a strict and thoughtful selection of donors is a major prerequisite for LDLT. Such careful donor scrutiny has the purpose of detecting any pre-existing condition that may increase the perioperative risk and, as a consequence, results in a large discrepancy between the numbers of potential and real donors.6, 8
One of many indications for liver biopsy is the evaluation of the status of the donor liver before transplantation.9 Therefore, the main goal of LB in the selection of living liver donors is to detect marginal organs10, 11 and exclude those donors with significant steatosis or other hepatopathies.
Fatty infiltration of the liver is common in the brain-dead donor population and has a strong correlation with primary nonfunction after cold preservation, a condition that is catastrophic to liver transplant recipients.12 More specifically, significant steatosis causes graft dysfunction by disrupting the microcirculation or altering cell membrane fluidity.13
Through these mechanisms, hepatic steatosis may not only adversely affect hepatic allograft function in the recipient but also may lead to a cascade of events that affect the donor operation and outcome. Donors with significant steatosis have been shown to have increased surgical time, increased perioperative blood loss during transection, increased transfusion requirement, and impaired hepatic regeneration.14, 15
Quantitative assessment of hepatic fatty infiltration is important in the donor evaluation for LDLT. Actually, there is no agreement on the maximum amount of steatosis for a safe LDLT, with values ranging between 30% and 60% in the literature.5 Clearly, many other factors contribute to the interpretation of the results of LB and the ultimate suitability of the potential graft (i.e., RLVBWR and/or graft-to-recipient body weight ratio). In fact, the main reports in the literature focus heavily on the importance of the graft-to-recipient body weight ratioand its correlation to the graft function7, 16 and give little attention to the important role of RLVBWR of the donor. We based our donor exclusion criteria more on the relationship between RLVBWR and the grade of liver steatosis than on graft-to-recipient body weight ratio. Consequently, we excluded potential donors with a low grade of steatosis and a significant, critical RLVBWR. Our actual policy is to exclude a priori all candidates with steatosis ≥20% or with liver steatosis 10% to 20% and a RLVBWR ≤0.8. In cases of RLVBWR >0.8 and liver steatosis 10%-20%, candidates may be reconsidered after dietary modification is attempted.
Rinella et al. reported that hepatic steatosis increases linearly with the BMI.17 They suggested that individuals with a BMI greater than 28 should undergo liver biopsy, whereas those with a BMI less than 25 and the absence of risk factors do not need one. Remarkably, Ryan et al. observed that 73% of potential donors with a BMI >25 had less than 10% hepatic steatosis. However, they also found that BMI and imaging correlated poorly with the degree of hepatic steatosis. Therefore, the indication for LB was extended to all patients with high BMI, permitting additional candidates to be considered as potential donors.7 In our experience, we excluded patients with BMI >30 a priori, not only because of the higher risk of liver steatosis but also mainly because BMI >30 also correlates with a high rate of perioperative complications (i.e., lung embolism, wound healing problems).18 Surprisingly, we found that some potential donors with normal BMI had a high percentage of liver steatosis.
Different imaging methods can potentially detect liver steatosis (i.e., US, CT, and MRI),5, 19–21 but all of them have considerably less predictive value than biopsy and become more sensitive only with increasing levels of biopsy-proven steatosis.7 Our findings confirm the experience of Ryan et al. In our series, US suggested the presence of high-grade steatosis in only 1 case (confirmed by biopsy to be 40% macrosteatosis). In no other case was there a positive finding of steatosis in the imaging investigations (MRI or all-in-one CT).
Similar to the recent report by Hwang et al., we also observed a reversibility of the grade of steatosis in all 6 highly motivated candidates who underwent a controlled diet after initial LB demonstrated significant steatosis. Afterward, all 6 donors became eligible candidates.22
Similar to Ryan et al., we observed a relative high rate of nonsteatotic liver pathologies.7
With regard to our potential donors with lobular and portal non–A-D hepatitis, the etiology and the clinical consequences are still unclear. In 4 of 6 cases, positivity for Epstein-Barr virus immunoglobulin G was demonstrated, but no histological or clinical correlation with Epstein-Barr virus could be found.
In the 2 cases of diffuse liver granulomatosis, no clear etiology has been found up to now.
None of these pathologies could have been detected by means of serological or noninvasive methods. LB, therefore, resulted in a higher rate of exclusion of donor candidates and opened new horizons in the world of liver pathology.
LB and Related Complications
Paradoxically, each invasive study used in the evaluation process to maximize donor selection and safety itself carries a small but definite risk. The incidence of complications after percutaneous biopsy is <1% in patients undergoing the examination for suspected pathology and is probably less in patients without liver disease,23 and the risk of death is extremely small.24
We do not agree with Rinella et al., who recommend US-guided biopsy of the left lobe in cases of right-lobe donation.1 We think that the risk of leaving a donor with a hematoma in the left lobe (often with critical remnant liver volume) for the sake of taking out an intact right graft is not justified.
The choice of biopsy technique should balance procedural risk with adequacy of the tissue sample. We started by using a Menghini needle until we experienced our first major complication (see Results). After that we changed to Tru-cut needles according to the guidelines reported by Bravo et al.9 The advantages of the Tru-cut biopsy system are the thinner needle, the ability to control exactly how deep within the liver one can go (i.e., 1.5- to 2-cm probes), and the option of performing the procedure 1-handed under US guidance.
Once the decision is made to perform an LB in the donor, the appropriate timing (from both a medical and aneconomical point of view) needs to be considered. We believe less invasive investigations with a high rate of donor exclusion (i.e., general lab values, hepatitis serologies, all-in-one CT, and psychosocial assessment) should precede LB. More expensive studies with a very low exclusion rate (i.e., cardiology evaluation, tumor markers) can follow LB.7, 6
Some centers defer LB until the donor operation, sometimes getting unexpected results that coerce the surgeon to abort the operation.3 There seems to be no justification in obtaining a liver biopsy at the time of the donor operation if its purpose is to assess liver quality. In the case of a pathological finding, the donor operation is aborted and the donor is exposed to the potential complications of anesthesia and an exploratory laparotomy. Other centers even perform an exploratory laparoscopy in cases of inconsistent results between clinic and liver biopsy.27
In our experience, as well as in that of others, morbidity and mortality in subsequent stages can be prevented with LB. This is especially true in the case of extended liver resection (critical RLVBWR or graft-to-recipient body weight ratio) where good liver quality is essential to guarantee the safety of the both donor and recipient.
It is our credo that donor safety overrides benefit for the recipient.28 As such, we firmly believe that the risks of LB (1.3% donor morbidity and 0% donor mortality) are justified when one considers that occult liver pathology in the donor carries a potential mortality and morbidity of 200% (both donor and recipient) when not recognized prior to LDLT.
The benefit of avoiding an unnecessary operation or a poor outcome for the donor or recipient justifies the low risk of LB-related complications (1.3%). Donors who are biopsied and do not donate as a result of the biopsy do benefit, because they are spared an unnecessary surgical exploration (or even a life-threatening liver resection in case of critical rest liver volume) and they are diagnosed with something (a hepatopathy) that would probably never have been diagnosed and that can now be treated or followed. The recently reported donor death in Japan29 confirms our position.
Therefore, we believe that preoperative LB in the donor selection for right adult LDLT is necessary, once the initial donor screening and noninvasive evaluation is complete. The main objective of LB should be to ensure the donor's safety more than the preservation of the graft function.