Telephone: 90 312 212 29 12; Fax: 90 312 215 42 16
Value of an individual liver biopsy in the preoperative evaluation of apparently healthy potential liver donors†
Article first published online: 26 MAR 2008
Copyright © 2008 American Association for the Study of Liver Diseases
Volume 14, Issue 4, pages 541–546, April 2008
How to Cite
Savas, N., Coskun, M., Bilezikci, B., Uruc, I., Karakayali, H., Yilmaz, U. and Haberal, M. (2008), Value of an individual liver biopsy in the preoperative evaluation of apparently healthy potential liver donors. Liver Transpl, 14: 541–546. doi: 10.1002/lt.21410
See Editorial on Page 420
- Issue published online: 26 MAR 2008
- Article first published online: 26 MAR 2008
- Manuscript Accepted: 29 OCT 2007
- Manuscript Received: 15 FEB 2007
Living donor liver transplantation has acquired widespread acceptance. A thorough workup of the potential living donor is guided by 2 objectives: the first is ensuring the safety of the surgical procedure for the donor, and the second is identifying donor grafts that pose potential risks for the recipient. Of 305 living-related liver donors, liver biopsy was performed in 201. The results of those patients' liver function tests and serologic tests were within normal limits. Forty-one of the 201 patients had steatosis on abdominal ultrasonography or tomography. Of 201 liver biopsies, 94 (46.8%) demonstrated normal findings, whereas 107 (53.2%) showed abnormal findings on pathology. Of 107 patients with abnormal pathological findings, 32 (29.9%) had fatty changes, 4 (3.7%) had steatohepatitis, and 71 (66.4%) had nonsteatotic histologic findings including fibrosis, hepatitis, and granulomatous reactions. In conclusion, our data show that half of the living-related liver donors, appearing absolutely healthy, had biopsies with abnormal pathology results. As it is well known that increasing percentages of steatosis may reduce the functional mass of the graft and occult liver diseases may not be detected without liver biopsy, we recommend that liver biopsies be part of the routine evaluation protocol for living-related liver donors in all transplantation institutions. Liver Transpl 14:541–546, 2008. © 2008 AASLD.
The growing discrepancy between the number of patients listed for liver transplant and the availability of deceased-donor organs has led to the acceptance by many persons of living-related donor liver transplant (LDLT) as a life-saving surgical innovation.
One unresolved issue in the living-donor evaluation process is the need for liver biopsies. A survey of 42 transplant centers that perform LDLT showed that only 14% performed routine liver biopsies, 60% performed liver biopsies only in selected cases, and 26% did not biopsy any donor candidates.1 Several methods, including liver function testing, abdominal ultrasonography (USG), computed tomography (CT) scanning, and magnetic resonance imaging (MRI), have been used to evaluate liver steatosis and other liver abnormalities. However, liver biopsy is the gold standard for detecting pathological liver abnormalities.
The presence of any kind of hepatitis, fibrosis, or moderate or severe steatosis is an absolute contraindication for living donation at most centers.2–4 However, whether a potential donor with mild steatosis (≤ 30%) should not be allowed to donate remains controversial.
Our aim was to evaluate the importance and weigh the risks and benefits of preoperative donor liver biopsy in selecting candidates for LDLT and to identify the histopathological abnormalities of liver biopsies of apparently healthy donors as a representative figure of our population.
MATERIALS AND METHODS
We retrospectively analyzed the data of 305 living-related liver donor candidates for 96 liver transplants performed between 2001 and 2006 at Baskent University in Ankara, Turkey.
Our evaluation protocol consists of 4 steps. The first step includes a thorough history and physical examination; blood group typing; hematologic, biochemical, and serologic testing for viral hepatitis; abdominal USG; and a psychological evaluation. The second step consists of evaluating cardiologic and pulmonary diseases. The third step is an abdominal CT scan or abdominal MRI and liver volume measurement. The fourth step is performing a liver biopsy and magnetic resonance cholangiopancreatography (Table 1). All donors were first-degree or second-degree relatives of the recipients. No donor organs were obtained from executed prisoners or other institutionalized persons.
|Step 1||Clinical evaluation with history and physical examination; blood group typing; hematologic, biochemical, and serologic tests for viral hepatitis; abdominal USG; and psychological evaluation|
|Step 2||Evaluation for cardiologic and pulmonary diseases|
|Step 3||Abdominal CT or abdominal MRI and liver volume measurement|
|Step 4||Liver biopsy and magnetic resonance cholangiopancreatography|
Body Mass Index (BMI)
We assessed the use of BMI, defined as the donor weight (kg) divided by the height (m2), in predicting the degree of histologically determined steatosis.
All histologic specimens were obtained during LDLT preoperative evaluation. Specimens were obtained from a single biopsy of the right lobe with an 18-gauge biopsy needle that provided adequate hepatic specimens (at least 10 different portal triads for each specimen). An experienced liver pathologist, blinded to the radiologic and surgical findings, retrospectively reviewed the histologic findings of the selected donor livers. All liver biopsy specimens were fixed in formalin, embedded in paraffin, and cut into serial sections 3 to 4 μm thick. They were then prepared with hematoxylin and eosin, Masson's trichrome, and Gomori's reticulin stains to identify collagen fibers and with Perls stain (which reveals iron). Liver pathologies such as fatty change of the liver, steatohepatitis, granulomatous reactions, fibrosis, portal tract mononuclear inflammation, iron deposition, and focal hepatocellular necrosis were all reported. The degree of macrovesicular steatosis was quantified on a percentage scale, which estimated the amount of liver parenchyma that had been replaced by macrovesicular steatotic droplets. Macrovesicular steatosis was graded as mild (0%-30%), moderate (30%-60%), or severe (>60%).
All statistical analyses were performed with SPSS software (Statistical Package for the Social Sciences, version 11.5, SSPS, Inc., Chicago, IL). Results are expressed as means ± standard deviations and percentages unless otherwise stated. The chi-square and Mann-Whitney U tests were used for statistical analyses. Values for P less than 0.05 were considered statistically significant.
Of 305 living-related liver donors, 201 [108 men (53.7%) and 93 women (46.3%)] reached step 4 for liver biopsy. The mean age of the donors was 38.3 ± 10.9 years (range, 18-61 years), and the results of liver function and serologic tests were within normal limits.
Forty-one of the 201 donors (20.3%) had steatosis on abdominal USG or CT. Thirty-three of 41 donors had mild steatosis, 2 had moderate steatosis, and 6 had severe steatosis on abdominal USG or CT. Of 201 liver biopsy specimens, the results of 94 (46.8%) were normal, and 107 (53.2%) demonstrated pathological findings. Of 107 donors with pathological findings on biopsies, 32 (29.9%) had fatty changes of differing degrees, and 4 (3.7%) had steatohepatitis. In 71 (66.4%) of the donor candidates, histologic findings other than steatosis were observed. In this group, 6 (5.6%) had fibrosis, 3 (2.8%) had hepatitis of unknown etiology, 2 (1.8%) had granulomatous reactions on biopsy, and 60 (56%) had other nonsteatotic histologic findings such as minimal portal tract inflammation, hepatocellular swelling, focal spotty necrosis, sinusoidal dilatation, lipofuscin pigment deposition, and hepatocyte apoptosis (Table 2). For those patients with nonsteatotic histologic findings, no etiology was found.
|Portal tract inflammation||13|
|Focal spotty necrosis||36|
|Lipofuscin pigment deposition||18|
Of 160 donors with normal results on abdominal USG or CT, 18 (11.2%) had fatty changes, 9 (5.6%) had granulomatous reactions, fibrosis, or hepatitis, 49 (30.6%) had other nonsteatotic liver pathologies such as portal tract inflammation, hepatocellular swelling, or focal spotty necrosis, and 84 (52.5%) had normal findings on liver biopsy.
Of 41 donors with steatosis on abdominal USG or CT, pathological steatosis (fatty change) was found only in 16 donors (39%); in the remaining 25 donors, pathological steatosis was not found, and this was statistically significant (P < 0.0001). The sensitivity of abdominal USG or CT in predicting real steatosis was 47%, the specificity was 84.5%, the positive predictive value was 39%, and the negative predictive value was 88% (Table 3).
|Steatosis on Pathology|
|Steatosis on USG or CT||Absent||142||18||160|
The BMI was calculated in 196 of 201 donors. The BMI of donors with steatosis on pathology was greater than the BMI of donors with normal findings on biopsy (29.6 ± 3.4 versus 24.9 ± 4.1), and this result also was statistically significant (P < 0.0001; Fig. 1).
The mean age of donors with steatosis on pathology was greater than the mean age of donors with normal findings on biopsy (44.7 ± 10.3 years versus 37.0 ± 10.5 years); this result was statistically significant (P < 0.0001; Fig. 2).
When we grouped donors according to BMI > 25 and BMI < 25, we found that 26.7% of donors with BMI > 25 had steatosis on pathology, whereas 1.3% of donors with BMI < 25 had steatosis on pathology. There was a statistically significant difference between the BMI and steatosis on pathology (P < 0.0001, odds ratio = 28.8). The sensitivity of BMI in showing steatosis on pathology was 96.9%, the specificity was 48.2%, the positive predictive value was 27.7%, and the negative predictive value was 98.8% (Table 4).
|Steatosis on Pathology|
When we examined the relationship between sex and steatosis on pathology, we found that 20.4% of the male donors and 12.8% of the female donors had steatosis on pathology, but no statistically significant difference existed between sex and steatosis on pathology.
When we examined the relationship between degree of steatosis on USG or CT with the presence of steatosis on pathology, we found that 29.4% of patients with a mild degree, 100% of patients with a moderate degree, and 66.7% of patients with a high degree of steatosis on USG or CT had fatty changes on pathology, and this also was statistically significant (Table 5; P < 0.05).
|Steatosis on Pathology|
|Degree of steatosis on USG or CT||Mild||23||10||33|
None of the patients experienced any complications related to the liver biopsy procedure.
LDLT is guided by 2 main principles: first, donor morbidity and mortality must be kept to a minimum, and second, the graft and recipient survival should be acceptably high, as in conventional deceased-donor liver transplant.5 Therefore, the ideal evaluation process should thoroughly assess potential graft function while minimizing risks to the donor candidate. Liver steatosis, particularly the macrovesicular type, in donor biopsy specimens has been associated with an increased risk of primary graft nonfunction.6, 7 Each percentage of steatosis might reduce the functional mass of the graft by 1%.8 Therefore, accurate preoperative assessment of the degree of graft steatosis is important for transplantation, as it has been shown that in deceased-donor grafts, steatosis > 30% increases the risk of primary nonfunction of the grafts to 13% versus 2.5% for nonsteatotic grafts.9 A large, retrospective, single-center study evaluating the effect of increasing grades of deceased-donor liver steatosis on recipient outcome showed increased initial poor graft function, increased graft loss rates, and higher 3-month patient mortality rates among the recipients of deceased-donor livers that had moderate and severe steatosis.10 In another study, severe macrovesicular steatosis in the living-related liver donor was associated with a greater than 60% risk of primary nonfunction after transplant. It was suggested that moderate degrees of macrovesicular steatosis in the donor livers might result in decreased hepatocyte regeneration and higher rates of graft dysfunction, nonfunction, and ischemic injury.11 Therefore, both deceased and living-related liver donors with greater than 30% steatosis are considered by some to be a relative risk factor for liver dysfunction and an absolute risk factor when it is greater than 60%.2–4
Although not all centers require a preoperative donor liver biopsy, it is a mandatory part of our donor evaluation protocol.12 At our center, we accept only donors with normal results on liver biopsy or those with only a mild degree of hepatosteatosis. Because of this, noninvasive tests are performed on liver donors initially, and only if the noninvasive tests are within our parameters of acceptance do we perform a liver biopsy on potential donors. In donors in whom steatosis is detected before liver biopsy, we perform a liver biopsy (1) to exclude steatohepatitis or other pathologies and (2) to detect the degree of steatosis and treat the donor's own health. On follow-up, we had no donor mortality; however, in 7 donors we had complications including incisional hernia, vascular and biliary complications, and intra-abdominal abscess formation. There were no complications related to the graft being too small or functioning poorly when transplanted.
Despite normal results on their liver function tests and normal radiologic findings, nearly half of our liver donors had some degree of liver pathology, Also, we found that 5.6% of liver donors had nonsteatotic liver pathologies that contraindicate transplant, such as granulomatous reactions, fibrosis, and hepatitis of unknown origin, on liver biopsy regardless of normal results on their liver function tests and normal radiologic findings of the liver. This is an important issue for recipient and donor health that cannot be predicted without liver biopsy. This emphasizes the necessity of liver biopsy in all liver donors and especially in those with low graft weight.
To transplant grafts of adequate size, some transplant centers use donor liver acceptability guidelines including either graft weight as a percentage of standard liver mass or graft to recipient body weight ratios. Both values show a good linear correlation and are acceptable.13 A liver with a graft to recipient body weight ratio less than 0.8% increases the risk of graft loss, with hyperperfusion proposed as the mechanism of injury.14 Aside from actual mass reduction, it has been proposed that significant steatosis causes graft dysfunction by disrupting the microcirculation or altering cell membrane fluidity.15, 16
Liver biopsy is generally safe and currently the gold standard, but noninvasive methods such as BMI, body topography, serum aminotransferase and lipid levels, and waist/hip ratio are other methods that also can determine liver steatosis.17–19 USG, CT, and MRI are common noninvasive means of estimating hepatic fat content, but their sensitivities for identifying hepatic fat differ from study to study.2, 20, 21 One study found that a combination of USG and CT had an overall accuracy rate of 65% for hepatic steatosis ≥ 10%, however, when the hepatic content was increased to ≥30%, the accuracy rate of USG and CT increased to 80%.20 In another study, it was found that in 14% of liver donors, radiologically occult conditions were detected only on histologic analysis, and liver biopsy was suggested for liver donors to exclude clinically and radiologically occult liver diseases.2 In our study, we estimated the sensitivity and specificity of abdominal USG or CT to predict liver steatosis and found that the sensitivity was as low as 47%, whereas the specificity was as high as 84.5%. This is important and suggests that radiologic tests including abdominal USG or CT are not enough to demonstrate hepatic steatosis. Interestingly, in our study, among donors with a high grade of steatosis on USG or CT, only 66.7% had steatosis on biopsy, and 33.3% did not have steatosis on biopsy; this was surprising for us. This could be related to the heterogenicity of liver steatosis, which begs the question of how many biopsy specimens should be taken to correctly diagnose liver pathology.
Frankel and associates22 investigated the optimal number of deceased-donor biopsy sites needed to evaluate liver histology for transplant and found that no single site predicted pathological findings better than another. They also found that although 1 liver biopsy is a good representative of histologic characteristics in the liver, when a biopsy shows mild macrovesicular steatosis, there is a 32% chance of finding more significant macrovesicular steatosis and a 5% chance of finding >40% macrovesicular steatosis, which may be of sufficient severity to preclude transplant. The authors therefore recommend that, when mild to moderate pathological steatosis is identified, 2 biopsies be taken to best predict the overall histologic characteristics of donor livers. According to that study, if we had more than 1 biopsy specimen from the donor candidates, we would have had more proportional results with which to correlate the degree of steatosis by radiologic imaging and liver biopsy.
There are several studies estimating the predictive value of BMI on hepatic steatosis.17, 20, 23 Rinella and coworkers17 reported BMI as a reliable predictor of hepatic steatosis with a positive correlation between increasing BMI and steatosis grade on biopsy. It has been suggested that liver biopsy can be avoided in subjects with normal BMI but that living donors with high BMI should undergo liver biopsy because biochemical and imaging data are not reliable enough to accurately diagnose the degree of steatosis. Conversely, Ryan and colleagues20 showed that normal BMI does not consistently predict the absence of a fatty liver. In their study, they found that 4 of 45 donors (8.8%) with normal BMI had 10% to 20% steatosis on biopsy. Similarly, Tran and colleagues23 suggested that BMI was not a predictive factor for steatosis.
In our study, we found that patients with steatosis on pathology had a mean BMI of 29.6, which was greater than that of patients with normal liver pathology, who had a mean BMI of 24.9. Only 1 patient with a BMI < 25 had steatosis on pathology; the remaining patients with steatosis on pathology had BMIs > 25. We also demonstrated that patients with BMIs > 25 had a 29-fold greater chance of having steatosis on liver biopsy than patients with BMIs < 25. Our results demonstrate that BMI may suggest the presence of liver steatosis but not other liver pathologies including fibrosis, hepatitis, and granulomatous reactions.
Similarly to Ryan and coworkers20 and Nadalin and colleagues,5 we observed a high rate of nonsteatotic liver pathologies. We found nonsteatotic liver pathologies in 35.4% of the donors in our LDLT population in Turkey, which means that in every 3 liver donors, 1 might have a nonsteatotic liver pathology that could preclude liver transplant. Although most of the nonsteatotic liver pathologies in our donors were mild, most of our liver transplant recipients were adults, and in cases in which the liver had a graft to recipient body weight ratio of only 0.8%, even milder changes become important with respect to risk of graft loss or a small-for-size liver. Because of this, we accept only donors with normal liver biopsy results or those with only a mild degree of hepatosteatosis (<5%). This approach may seem conservative, but we believe that it minimizes surgical risk for the donor and reduces the possibility of a poor outcome related to a steatotic graft for the recipient.
We conclude that morbidity and mortality of the donor and recipient in subsequent stages could be prevented by pretransplant liver biopsy of the donor. This is especially true in cases of extended liver resection (critical remnant liver volume to body weight ratio or graft to recipient body weight ratio) in which good liver quality is essential to guarantee the safety of the donor and the recipient. We believe that less invasive investigations with a high rate of donor exclusion should precede liver biopsy and that liver biopsy should be performed in every healthy-appearing living-liver donor at all transplant institutions.