The role of liver biopsy in evaluating acute allograft dysfunction following liver transplantation: a clinical histologic correlation of 34 liver transplants

One hundred six liver biopsy specimens from 34 orthotopic liver transplant (OLT) patients were examined and the histologic findings correlated with the clinical course of the patients to determine if specific morphologic patterns were associated with specific causes of acute allograft dysfunction. The principle causes of allograft injury in these patients appeared to be acute rejection and ischemic injury, with rarer cases of viral infection and biliary obstruction. Graft rejection causing transient liver dysfunction was associated with a mixed inflammatory infiltrate in the portal tracts and involving the interlobular bile ducts. Rejection resulting in severe, persistent dysfunction was associated with destruction and loss of the interlobular bile ducts or portal inflammation, followed by acute centrilobular hepatocyte necrosis. Ischemic liver injury was characterized by hepatocyte ballooning and/or hepatocyte necrosis. Ischemic injury causing transient graft dysfunction demonstrated focal, limited areas of hepatocyte necrosis or transient centrilobular hepatocyte ballooning. Severe ischemic injury resulting in persistent dysfunction caused diffuse hepatocyte necrosis or centrilobular ballooning followed by centrilobular hepatocyte loss and severe cholestasis with evidence of bile duct epithelial injury. The histologic patterns observed were not pathognomonic; radiologic studies, bile cultures, and other laboratory tests were necessary to rule out biliary or vascular obstruction and bacterial cholangitis. However, liver biopsies, especially serial biopsies, were helpful in suggesting the probable cause of liver dysfunction and in predicting subsequent allograft recovery or failure.     

The histological changes in transbronchial biopsy after treatment of acute lung rejection in heart-lung transplants

Transbronchial lung biopsies taken during episodes of acute lung rejection in heart-lung transplant patients were examined histologically. When the diagnosis was confirmed by microscopy, the patients were treated with augmented immunosuppression by high-dose corticosteroids. A second biopsy was obtained an average of 23.5 days after commencement of treatment. These biopsies were examined to determine the histological changes caused by treatment. In most cases, there were both quantitative and morphological differences between the infiltrates in acute rejection and in the remaining perivascular infiltrates after treatment. In acute rejection, 26 of 28 biopsies contained perivascular lymphocytic infiltrates, lymphocytes being large and blast-like. Although 20 of 28 follow-up biopsies still contained perivascular infiltrates, they were smaller and the lymphocytes smaller in size. Half the biopsies in rejection contained neutrophils, but less than half contained eosinophils in the perivascular infiltrates. After treatment, all these cells were less numerous. Another feature of treated rejection was the presence of haemosiderin around vessels suggesting earlier acute vascular injury. However, haemosiderin persists long after the cellular infiltrate has disappeared and cannot be considered a reliable feature of recently treated acute lung rejection. The bronchiolar infiltrates showed a similar pattern of responses to the perivascular infiltrates.     

Mast cells and c-Kit expression in liver allograft rejection

AIMS : The pathogenesis of rejection following liver transplantation is not fully understood. It has been postulated that mast cells may play a role in acute and chronic rejection of a number of other solid organ grafts. The aim of this study was to assess the possible role of mast cells and c-Kit+ cells in acute and chronic liver allograft rejection. METHODS AND RESULTS : Biopsy specimens from (i) 'time zero' grafts with a minimal degree of perfusion injury (controls), (ii) transplanted livers with different grades of acute rejection, and (iii) transplanted livers with end-stage chronic rejection, were stained immunohistochemically using monoclonal anti-mast cell tryptase and polyclonal anti-c-Kit antibodies. Tryptase- and c-Kit-positive cell densities were assessed by image analysis. Tryptase-positive mast cell densities (P<0.001) were strongly correlated with acute liver allograft rejection grades and chronic liver allograft rejection. Furthermore, a similarly strong relationship was found between c-Kit+ cell densities and increasing rejection grade (P<0.001). CONCLUSIONS : Tryptase- and c-Kit-positive mast cells form part of the inflammatory infiltrate in both acute and chronic liver allograft rejection, and may be important effector cells in these processes.     

Morphologic features resembling transplant rejection in core biopsies of native livers from patients with Hepatitis C

Morphologic differentiation of recurrent Hepatitis C from transplant rejection is a major problem in posttransplant liver biopsies. Although biopsies of the native livers from patients with Hepatitis C are known to display bile duct damage, other morphologic features similar to those seen in rejection, such as endotheliitis, portal eosinophils, and pericentral fibrosis, are not generally acknowledged. To determine the frequency with which features morphologically similar to rejection might be present, we examined 50 cases of core-needle biopsy from the native livers of patients with Hepatitis C for the presence of the following: bile duct damage, portal eosinophils, portal or central vein endotheliitis, ductopenia, vascular obliteration, pericentral fibrosis, and pericentral mononuclear cell infiltrate. Biopsy specimens with other concurrent disease processes were excluded. The frequency of each morphologic feature was as follows: bile duct damage (30%), portal eosinophils (42%), portal endotheliitis (20%), central vein endotheliitis (0%), pericentral mononuclear cell infiltrate (14%), ductopenia (2%), atrophic-looking bile ducts (2%), vascular obliteration (0%), and pericentral fibrosis (10%). Bile duct damage and portal endotheliitis were both more common with higher grade hepatitis (Fisher exact test, P = .001). None of the morphologic parameters correlated with biopsy stage, viral genotype, or liver function tests. We conclude that features similar to those found in acute rejection are common in Hepatitis C, whereas features resembling chronic rejection are less frequent. This study provides quantitative data that supports the need to interpret these features with great caution in posttransplant liver biopsies from patients with recurrent Hepatitis C who are suspected of rejection.     

Acute cellular rejection and cyclosporine nephrotoxicity monitored by biopsy in a renal allograft recipient. The differentiation of drug nephrotoxicity from rejection by phenotyping of cellular infiltrates

Serial allograft biopsies were performed on a renal transplant patient who experienced recurrent episodes of acute cellular rejection as well as cyclosporine nephrotoxicity. Five biopsies were performed after acute elevations of the serum creatinine level (15, 46, 155, 244, and 324 days after transplant). Each specimen was evaluated by routine histologic techniques as well as by immunofluorescence analysis and by monoclonal antibody labeling for determination of the cell phenotype of the mononuclear cell infiltrates within each specimen. The first and third specimens disclosed significant T-cell infiltrates with an equal number of T-cytotoxic-suppressor (Leu 2a) and T-helper-inducer (Leu 3a) cells in a diffuse cortical pattern, while the second biopsy showed a slightly milder infiltrate with a marked elevation (7:1) in the Leu 3a:Leu 2a ratio in the cortical-diffuse pattern. Clinically, the patient responded dramatically to cyclosporine dosage reduction following the second biopsy, and bolus steroid antirejection therapy following the first and third biopsies. These findings suggest that phenotypic cell marker analysis within the context of histologic pattern is a useful adjunct to the routine histologic evaluation of renal allograft biopsy specimens and may provide a means of differentiating rejection from cyclosporine nephrotoxicity.     

The clinical and pathologic implications of plasmacytic infiltrates in percutaneous renal allograft biopsies

Plasmacytic infiltrates in renal allograft biopsies are uncommon and morphologically distinctive lesions that may represent variants of acute rejection. This study sought significant clinical and pathologic determinants that might have influenced development of these lesions and assessed their prognostic significance. Renal allograft biopsies (n = 19), from 19 patients, with tubulointerstitial inflammatory infiltrates containing abundant plasma cells, composing 32 +/- 8% of the infiltrating mononuclear cells, were classified using Banff '97 criteria. Clonality of the infiltrates was determined by immunoperoxidase staining for kappa and lambda light chains and polymerase chain reaction for immunoglobulin heavy-chain gene rearrangements, using V(H) gene framework 3 and JH consensus primers. In situ hybridization for Epstein-Barr virus encoded RNA (EBER) was performed in 17 cases. The clinical features, histology, and outcome of these cases were compared with kidney allograft biopsies (n = 17) matched for time posttransplantation and type of rejection by Banff '97 criteria, with few plasma cells (7 +/- 5%). Sixteen of 19 biopsies (84%) with plasmacytic infiltrates had EBER-negative (in 14 cases tested) polyclonal plasma cell infiltrates that were classifiable as acute rejection (types 1A [4], 1B [10], and 2A [2]). These biopsies were obtained between 10 and 112 months posttransplantation. Graft loss from acute and/or chronic rejection was 50% at 1 year and 63% at 3 years, and the median time to graft failure was 4.5 months after biopsy. There was no significant difference in overall survival or time to graft failure compared with the controls. Three of 19 biopsies (16%) had EBER-negative polyclonal plasmacytic hyperplasia, mixed monoclonal and polyclonal polymorphous B cell hyperplasia, and monoclonal plasmacytoma-like posttransplantation lymphoproliferative disease (PTLD) and were obtained at 17 months, 12 weeks, and 7 years after transplantation, respectively. Graft nephrectomies were performed at 1, 19, and 5 months after biopsy, respectively. Plasmacytic infiltrates in renal allografts comprise a spectrum of lesions from acute rejection to PTLD, with a generally poor prognosis for long-term graft survival.     

Characterization of ectopic lymphoid structures in different types of acute renal allograft rejection

We hypothesize that T cells such as interleukin (IL)‐21⁺B cell lymphoma 6 (BCL6)⁺ T follicular helper cells can regulate B cell‐mediated immunity within the allograft during acute T cell‐mediated rejection; this process may feed chronic allograft rejection in the long term. To investigate this mechanism, we determined the presence and activation status of organized T and B cells in so‐called ectopic lymphoid structures (ELSs) in different types of acute renal allograft rejection. Biopsies showing the following primary diagnosis were included: acute/active antibody‐mediated rejection, C4d⁺ (a/aABMR), acute T cell‐mediated rejection grade I (aTCMRI) and acute T cell‐mediated rejection grade II (aTCMRII). Paraffin sections were stained for T cells (CD3 and CD4), B cells (CD20), follicular dendritic cells (FDCs, CD23), activated B cells (CD79A), immunoglobulin (Ig)D, cell proliferation (Ki67) and double immunofluorescent stainings for IL‐21 and BCL6 were performed. Infiltrates of T cells were detected in all biopsies. In aTCMRI, B cells formed aggregates surrounded by T cells. In these aggregates, FDCs, IgD and Ki67 were detected, suggesting the presence of ELSs. In contrast, a/aABMR and aTCMRII showed diffuse infiltrates of T and B cells but no FDCs and IgD. IL‐21 was present in all biopsies. However, co‐localization with BCL6 was observed mainly in aTCMRI biopsies. In conclusion, ELSs with an activated phenotype are found predominantly in aTCMRI where T cells co‐localize with B cells. These findings suggest a direct pathway of B cell alloactivation at the graft site during T cell mediated rejection.     

颗粒酶B和穿孔素免疫组织化学检测诊断肝移植急性排斥反应的敏感性与特异性

目的　探讨颗粒酶B和穿孔素两种免疫活化分子在肝移植急性排斥诊断中的作用,及其与Banff急性排斥反应组织学诊断标准之间的对应关系。方法　在常规组织学诊断基础上,将41份肝穿刺标本用颗粒酶B与穿孔素单克隆抗体进行免疫组织化学EnVision二步法标记,IPP图像分析软件计算阳性细胞数/mm2 作为免疫活化细胞指数(AI),以组织学诊断作为评判有无急性排斥反应的标准。结果　在41份肝穿刺标本中,组织学诊断为急性排斥反应21份,缺乏急性排斥反应组织学改变20份。急性排斥反应组颗粒酶B与穿孔素AI值显著高于无排斥反应组(P<0. 001),中重度排斥反应组AI值显著高于轻度及其非确定性排斥反应组(P<0. 001)。与组织学诊断比较,颗粒酶B的敏感性、特异性、阳性预测值、阴性预测值及诊断一致率分别达到90. 0%、95. 2%、94. 7%、90. 9%以及92. 7%;穿孔素的各指标也分别达到80. 0%以上。结论　颗粒酶B与穿孔素是急性排斥反应免疫效应细胞活化标志,在临床肝移植急性排斥反应时表达明显升高,作为组织学诊断急性排斥反应的辅助指标具有相当高的敏感性及特异性,可用于肝移植后肝穿刺标本的鉴别诊断。     

Apoptosis of bile duct epithelial cells in hepatic allograft rejection

Liver biopsy remains the 'gold standard' for monitoring rejection in liver transplant patients. Portal inflammation, bile duct damage and endothelialitis are recognized features of hepatic allograft rejection. The pathogenesis of the bile duct injury during rejection, however, remains unclear. To define the mechanism of bile duct damage, we studied the light- and electronmicroscopic appearance of hepatic tissue from selected patients in whom allograft failure was solely due to rejection. Of the 25 orthotopic liver transplant rejection cases examined, 17 were mild, seven were moderate and one was severe rejection. Light microscopy examination of the damaged bile duct epithelium revealed evidence of apoptosis which was confirmed by electronmicroscopy. Furthermore, there appeared to be a positive correlation between the grade of rejection and the number of apoptotic cells. Also included in the study were 13 cases of chronic active hepatitis and 10 normal livers which showed the least apoptotic cells. We conclude that the identification of apoptotic cells in damaged bile ducts in allograft biopsies might be helpful in the diagnosis of rejection and in assessment of the severity of rejection.     

Natural killer-cell activity in cyclosporine-treated renal allograft recipients

We prospectively studied natural killer (NK)-cell activity in 16 cyclosporine-treated renal transplant recipients. NK function remained intact in the group as a whole in the initial 6 months following transplantation. The percentage of CD16-positive cells within the peripheral blood mononuclear-cell population was highly correlated with NK activity both prior to and following transplantation in the absence of rejection. During rejection, the correlation was poor. A marked increase in NK activity occurred during 9 of 12 rejection episodes; similar increases in NK activity were rarely observed in the absence of rejection. Significant infiltrates of NK cells, as determined by expression of CD16, were not demonstrated in stained biopsy specimens obtained from rejecting allografts. Pretransplant NK activity did not predict clinical outcome of the allograft. Our results indicate that NK cells are activated during allograft rejection in cyclosporinetreated patients, but their exact role in the rejection process is unknown.     

Analysis of the different histologic lesions observed in transbronchial biopsy for the diagnosis of acute rejection. Clinicopathologic correlations during the first 6 months after lung transplantation

Acute rejection is an extremely common complication of lung transplantation. (1) To appreciate the interobserver variation in the interpretation of histologic findings and (2) to assess the efficacy of transbronchial biopsy (TBB) for acute rejection diagnosis and associated diseases, particularly infection, we performed a retrospective study including 53 consecutive patients who underwent at least one clinically indicated TBB during the first 6 months after lung transplantation. A total of 94 TBB was obtained. The following histologic features observed in TBB specimens-perivascular mononuclear infiltrates, lymphocytic bronchitis/bronchiolitis, and alveolar lesions, were reliably reproduced by 2 pathologists from the same transplant center, with kappa values ranging from 0.79 to 0.82. For identifying perivascular mononuclear infiltrates, discordance between the 2 observers was significantly associated with moderate/severe alveolar lesions. For the diagnosis of acute rejection, perivascular mononuclear infiltrates had a specificity of 96.5%, a positive predictive value of 97.5%, and a sensitivity of 67.7%, whereas lymphocytic bronchitis/bronchiolitis had a specificity of 56.3% and a sensitivity of 19.4%. Interestingly, there was a positive independent correlation between infection and moderate/severe alveolar histologic lesions ( P < .01). In conclusion, the interobserver agreement between experienced pathologists in TBB interpretation is good. Perivascular mononuclear infiltrates remain the cornerstone for acute rejection diagnosis. The presence of moderate/severe alveolar lesions should prompt to search for infection.     

The significance of eosinophils in mild and moderate acute cardiac allograft rejection

The presence of eosinophils has previously been associated with severe acute cardiac allograft rejection. This appears to be a relatively uncommon finding, judging from our experience and the paucity of information appearing in literature. We report three cases where a prominent infiltrate of eosinophils was noted on endomyocardial biopsy following cardiac transplantation. There was no evidence of severe acute rejection in any of these three patients, and one patient had only mild acute rejection without even focal myocardial necrosis. An infiltrate, which includes eosinophils, does not appear to be restricted to severe acute cardiac allograft rejection. Therefore, when eosinophils are noted in endomyocardial biopsy specimens, decisions to revise the immunosuppressive regimen of cardiac transplant recipients should continue to be based upon established conventional histologic criteria.     

Cytokeratin immunostaining for detection of biliary epithelium: its use in counting bile ducts in cases of liver allograft rejection.

AIMS--To see how useful the application of a bile duct specific cytokeratin antibody (AE1) was in identifying and counting bile ducts in liver allograft biopsy specimens. METHODS--Eighteen liver biopsy specimens showing acute rejection and 17 biopsy specimens plus six hepatectomy specimens showing chronic rejection were studied. Serial sections were cut and stained with haematoxylin and eosin and AE1 antibody. Two pathologists (RFH and KP) examined the sections with respect to a range of histological features. RESULTS--Similar numbers of bile ducts were identified on haematoxylin and eosin sections as on corresponding sections stained by AE1 in cases of acute rejection and end stage chronic rejection. Greater numbers of bile ducts were identified by AE1 during the early stages of chronic rejection, especially when dense portal inflammatory infiltrates were present. These were often incomplete structures or individual cells within portal tracts, and bile ducts subsequently disappeared in all cases. Ductular proliferation was clearly shown by AE1 in acute rejection and the extent seemed to correlate with the severity of rejection present. By contrast, no ductular proliferation was observed in chronic rejection. CONCLUSIONS--Haematoxylin and eosin stained sections are adequate for counting bile ducts in most biopsy specimens from patients with suspected chronic rejection. Immunostaining for biliary cytokeratins using AE1 is of limited use in occasional cases where bile ducts are obscured by inflammatory cells.     

Bile duct injury as a part of diagnostic criteria for liver allograft rejection

The decisive criterium of acute liver allograft rejection was found to be the presence of the diagnostic triad of acute rejection; ie, the presence of portal inflammatory mixed infiltrates, venous endothelialitis (both portal and central), and bile duct injury. On the basis of the presence of each of the components of the diagnostic triad, criteria for the diagnosis of different degrees of acute rejection were developed, particularly focusing attention on a detailed analysis of bile duct injury. Bile duct injury was shown to be an essential part of the histopathologic changes in all grades of acute rejection in the liver allograft, the grade of severity of bile duct injury correlating to a certain extent with the grade of severity of acute rejection. Our analyses have made it evident that bile duct injury, which most probably occurs earlier in the process of acute rejection than endothelialitis, is a more sensitive parameter than endothelialitis in the diagnosis of acute rejection. Furthermore, our analyses have revealed that bile duct injury in acute rejection is likely to be an irreversible process, depending on the number of episodes of acute rejection that previously occurred. On the other hand, it has become clear from our results that bile duct injury must not be considered to be an absolute histopathologic marker of acute rejection; however, it does have to be judged synoptically in connection with the other components of the diagnostic triad and the changes that the triad cause in the hepatic parenchyma. Additional analyses of the grade of severity of cholostases have shown that the cholostases are, to a certain degree, an accompanying phenomenon of the histopathologic changes characterizing acute rejection rather than a histopathologic change that is as significant as the presence of the components of the diagnostic triad.     

Arteriolitis in renal transplant biopsies is associated with poor graft outcome

AIMS: The Banff 1997 classification of renal allograft pathology identifies arteriolitis as a finding of uncertain significance. We sought to improve our understanding of arteriolitis by correlating its occurrence with histopathological and clinical parameters. METHODS AND RESULTS: Twenty allograft kidney biopsies from 19 patients, showing arteriolitis, were identified. Arterioles were defined as small vessels with: (1) wall thickness of 1-3 myocytes; (2) diameter less than one-third of an adjacent glomerulus; and (3) discontinuous or absent elastica. Arteriolitis was defined as mural infiltration by lymphocytes. Other histological findings were categorized according to the Banff 1997 working formulation. Ten biopsies (50%) showed type IIA rejection, seven (35%) showed type I rejection, and three (15%) showed borderline change. Two patients with borderline change had acute rejection in the next biopsy. None of the seven patients with type I rejection had previous or subsequent type II rejection on biopsy. A total 11/20 biopsies (10/19 patients) showing arteriolitis had type IIA rejection in the index or next biopsy. On follow-up, graft loss due to rejection occurred in 5/19 (26%) patients (median 126 days); all had shown type IIA rejection on a previous biopsy. Chronic allograft nephropathy developed in a further 4/19 (21%) patients (median 157 days), of whom three had shown only type I rejection on biopsy. CONCLUSION: Arteriolitis is associated with acute rejection, often type II rejection, and is associated with poor graft outcome. Other causes of arteriolitis were not encountered in this series.     

Serum amyloid A protein (SAA): a marker for liver allograft rejection in humans

Serum amyloid A protein (SAA) concentrations were monitored in 12 consecutive liver transplant recipients until the 70th postoperative day. Fourteen rejection episodes were identified histologically in 42 liver biopsies of the grafts. Of 12 rejections 8 (66.7%) were characterized by pronounced simultaneous increases in SAA concentrations in plasma, the mean peak value being 16.94 ± 8.82 mg/dl (range 4.58–28.55 mg/dl) compared with a mean normal value of 0.98 ±0.42 mg/dl in healthy controls. Of 42 biopsies 28 did not show histological evidence of graft rejection. Of 25 negative biopsies 24 (96.0%) were not accompanied by a parallel SAA increase in plasma. These findings demonstrate that measurements of SAA concentrations may provide a valuable non-invasive aid in identifying acute liver allograft rejection in humans.Abbreviations CRP C-reactive protein - ROC receiveroperating characteristic - SAA serum amyloid A protein Correspondence to: G. Feussner     

Significance of C4d staining in ABO-identical/compatible liver transplantation.

Complement degradation product C4d has become an important marker of humoral or antibody-mediated rejection in renal and heart allograft biopsies. Although there have been several reports on the detection of C4d in liver allografts, the significance of C4d in liver transplantation and its relationship with humoral rejection are still not clear. We investigated the frequency and pattern of C4d staining in liver allograft biopsies with reference to preoperative lymphocyte crossmatch tests, which detect donor-reactive lymphocyte antibody. Survival rates at 5 years were 77% for crossmatch-negative patients and 53% for crossmatch-positive patients (P=0.009). In crossmatch-negative patients, reproducible positive staining was obtained in 28 of 86 (33%) biopsies taken within 90 days after transplantation and 33 of 96 (34%) biopsies 90 days or after transplantation. Most C4d staining was observed in the portal areas, and no clear correlation was observed between C4d positivity and histological diagnosis. In crossmatch-positive patients, 9 of 11 (82%) biopsies showed positivity for C4d. C4d stained perivenular areas as well as portal areas. Histology of crossmatch-positive patients included acute rejection and cholangitis, but did not include periportal changes that were seen in humoral rejection in ABO-incompatible liver transplantation. In summary, focal C4d deposition was seen in various types of liver allograft injury and had little clinical impact on crossmatch-negative patients, but extensive C4d staining in crossmatch-positive patients may be associated with humoral rejection and poor graft survival.     

Ultrastructure of the Kidney in Acute Interstitial Nephritis

Fifteen percutaneous renal biopsies from patients with acute renal failure due to acute interstitial nephritis (AIN), in almost all cases due to drugs, were studied by electron microscopy. Differential counting of interstitial cells showed an average of 69% lymphocytes (small and large) and 11 % macrophages. Plasma cells and eosinophils were comparatively rare. The infiltrate resembled that of acute rejection, suggesting a cellular hypersensitivity reaction. Proximal and distal tubules were severely affected focally. Migration of lymphocytes through the tubular basement membrane of otherwise well-preserved tubules was considered to be the first phase. Other tubules showed extreme thinning of the tubular basement membrane, with still intact cellular walls. Rupture of the tubular basement membrane and necrotic disintegration of tubular epithelial cells are probably late phenomena. The non-necrotic tubules displayed severe reduction of proximal brush border and proximal as well as distal tubular basolateral infoldings. Focal tubular disintegration leading to tubular block and/or backleak as well as decrease of proximal tubular sodium resorption leading to a decreased glomerular filtration (a mechanism probably also acting in ischemic acute renal failure) may all be factors responsible for the acute renal failure in AIN.     

Molecular case report: IgVH analysis in acute humoral and cellular liver allograft rejection suggests a selected accumulation of effector B cells and plasma cells

Acute cellular (CLR) and humoral liver allograft rejection (HLR) are the most important immunological obstacles to successful liver transplantation. In HLR, serum antibodies play the central pathogenetic role. In CLR, CD3+ T lymphocytes drive the destructive immune response. Although CLR and HLR show different clinical symptoms and can be kept apart in most cases, they share histomorphological similarities. In CLR, hepatic B lymphocytes and plasma cells as well as B-cell-activating cytokines have recently been described, indicating that, in addition to T cells, antibody-mediated mechanisms might be involved. To analyze the impact of hepatic B cells in CLR and HLR, the immunoglobulin (Ig) variable (V)-region gene repertoire was determined from tissue of one case of CLR and one case of HLR. Complement deposits and lymphocytic infiltrate were determined using immunohistochemistry. T cells, B lymphocytes and plasma cells could be detected in both cases, whereas C3c and C4d deposits could only be demonstrated in the HLR case. The molecular analysis of 63 V-region genes showed that B cells in both allografts expressed selected V-gene repertoires. All sequences differed from the putative germline sequences by multiple somatic mutations. This suggests a clonal expansion of selected effector B cells in the portal tracts of liver allografts. Locally accumulated B cells and their antibodies might be involved in IgG-mediated complement activation in CLR and HLR.