Chronic rejection (CR) in liver allografts shares risk factors and morphological characteristics, such as obliterative arteriopathy, with other solid organs. In other respects, however, there are substantial differences. In contrast with kidney and lung allografts, CR in liver grafts has a relatively rapid onset and progressive course. Retransplantation is usually required within the first year after transplantation. A significant decline in incidence from 10–15% to 3–5% has been described in recent reports of several transplant centers, although in pediatric liver graft recipients the incidence is reported to be higher (1,2). Despite its progressive course, reversibility of hepatic CR has been reported, probably due to a timely rescue therapy (3). Recovery from ductopenia (DP), which is the second major histological hallmark of CR, and return to normal histological features has been described (3–5). DP is usually defined as absence of intrahepatic bile ducts in at least 50% of the portal tracts (2,5,6).
Because most cases of CR develop from acute cellular rejection (AR), which has a substantially higher incidence of 30–70%(7), early recognition is highly relevant. A timely modification of immunosuppressive management is vital to prevent the transition to CR. However, histological features predictive of the transition from AR to CR are not well documented. Biopsy diagnosis of CR has relied mainly on DP because the other histological hallmark of CR, obliterative arteriopathy, is found predominantly in medium-sized tributaries of the hepatic artery, which are usually absent in graft biopsy specimens. Recently, a working formulation for histopathological staging of CR has been published by an international panel (2). In this new staging system, bile duct damage has been reappraised. Atrophy and pyknosis of bile duct epithelium has been included in the diagnostic criteria of CR (2). However, when DP has developed, portal inflammation (PI) usually has abated. It is conceivable that modification of antirejection treatment at this stage will have only limited or no effect. Therefore, recognition of earlier stages of CR is necessary. The diagnostic criteria of AR are mainly based on portal features. In this study we investigated the value of several parenchymal features within the context of AR to predict the development of CR. The features included central vein endothelialitis (CVE), centrilobular necrosis (CLN), central vein fibrosis (CVF), and lobular inflammation (LBI).
MATERIALS AND METHODS
Histological features of liver grafts were compared in two patient groups. One group consisted of 12 patients who all underwent retransplantation because of graft failure due to chronic rejection (CR group). In these patients the diagnosis of CR was established on histological examination of the transplantectomy specimen. The median time of retransplantation was 7 months after the first transplantation (range 2–19 months). All 12 grafts showed the 2 basic hallmarks of CR: obliterative arteriopathy and DP (2,6).
The selection of a control group of 17 patients was based on the presence of biopsy-proven AR in the 1-week graft biopsy specimen, the availability of a follow-up biopsy specimen at 1 month, a minimum graft survival of 1 year, and normal histological features in the 1-year biopsy specimen. Only primary grafts were included. The histological diagnosis and grading of AR were based on the Banff criteria (8). Demographic data of the two patient groups are summarized in Table 1.
Basic immunosuppression consisted of triple-drug immunotherapy with prednisolone, cyclosporine, and azathioprine. Conversion to FK506 was performed in five patients with CR and three patients of the control group. Histologically proven episodes of AR were treated with boluses of methylprednisolone. Patients with steroid-resistant rejection were given antithymocyte globulin.
The study consisted of a semiquantitative comparison of histological features of graft biopsy specimens of patients in the CR versus control groups. Graft biopsy specimens taken at 1 week and 1 month after orthotopic liver transplantation were studied. Of the CR group, 11 one-week biopsy specimens and 9 one-month biopsy specimens were available. Of the control group, 17 one-week and one-month biopsy specimens were included.
The progression of early histological changes to full-blown characteristics of CR was further evaluated on 11 biopsy specimens taken between 1 and 2 months after transplantation (median: 48 days, range 42–68 days) and 8 biopsy specimens taken between 3 and 6 months after transplantation (median: 103 days, range 74–149 days). Only one biopsy specimen of each period per patient was considered.
Graft biopsy was performed using a 1.6-mm Menghini needle. Formalin-fixed and paraffin-embedded tissue specimens were routinely stained with hematoxylin-eosin, periodic acid Schiff’s reagent after diastase digestion, Masson trichrome, Gomori’s reticulin, Perls’ iron staining, and rhodanine for copper.
The histological features and criteria for the semiquantitative assessment are listed in Table 2.
CVE consisted of attachment of lymphocytes to the central vein endothelium both at the luminal and subendothelial side. CVF was diagnosed in patients with thickening of the wall of the central vein, as could be assessed in the trichrome staining. Drop out of hepatocytes with or without hemorrhage was scored as CLN.
Features of AR were scored according to the Banff criteria (8). PI was assessed separately as a counterpart of LBI. Grading of LBI was performed in a similar way as the lobular hepatitis component in chronic hepatitis according to the methods of Batts and Ludwig (9). DP was scored positive when 50% or more of the portal tracts did not contain a bile duct.
Statistical analysis of the data was performed using the SPSS 9.0 package. The chi-square test or Fisher exact test (where appropriate) was performed in the comparison of incidences of the studied features between the CR and the control group. The Mann-Whitney test was used for the comparison of the grade of severity of the features between the two groups. In the longitudinal study within the groups, the Wilcoxon rank-sum test was utilized. A P value of 0.05 was considered statistically significant.
CR versus control.
In the 1-week biopsy specimens of the CR patients, AR according to the Banff criteria was diagnosed in 10 of 12 patients and in all controls, because this was one of the inclusion criteria. In the CR group, three patients had a Banff grade III AR, two patients had a grade II, five patients had a grade I, and one patient showed mild inflammatory activity in portal tracts compatible with AR but not diagnostic. One patient had no 1-week biopsy specimen. In the control group, 2 patients had a grade III rejection, 5 patients had a grade II, and 10 patients had a grade I AR.
At 1 week, the incidences of parenchymal changes were equally high in both groups as shown in Table 3. No significant differences were observed between the CR patients and the control group. Analysis of the severity of the histological features showed no significant differences at this early stage. At 1 month, the incidence of CVE was significantly higher in the CR group (P =0.04), whereas the other features showed no differences. The median grade of severity of CVE and PI was significantly higher in the CR group than in the control group (P <0.001 and P =0.014, respectively), whereas LBI reached a borderline significant difference (P =0.068;Fig. 1).
Follow-up within the groups.
A longitudinal comparison within the groups from 1 week to 1 month showed an increased incidence of CVE in the CR group, whereas the control group showed a decreased incidence (Table 3). In the control group, the median grade of PI and CVE at 1 month was significantly lower than at 1 week (P =0.046 and P =0.034;Fig. 2). In the CR group, the median grade of the parenchymal changes showed an increase between 1 week and 1 month, but only CVF reached a statistically significant difference (P =0.03;Fig. 3).
Within the CR group, all lobular features were already present in the 1-week biopsy specimen. At 1 month, the incidence of these features was beyond 75%, whereas DP was only present in less than 50% of the patients.
In the second month after transplantation, CR could be diagnosed in the graft biopsy specimens of 5 of 11 CR patients, based on the presence of DP. Four of the remaining six patients had AR (two with grade I and two with grade II according to the Banff criteria) apart from the persistent lobular changes. One patient had only parenchymal changes in his biopsy specimen consisting of severe CLN, mild LBI, and CVF. Another patient showed features of obstructive cholangiopathy. CLN, LBI, and CVF were observed in 10 of 11 patients, whereas CVE was still present in 7 patients.
Similar frequencies were observed in the graft biopsy specimens beyond the third month, in which CVE was present in five of eight patients. CLN was seen in all patients, and in seven of eight patients CVF and LBI were still present. The diagnosis of DP and consequently CR was established in three of the eight biopsy specimens. All three patients were already recognized as having CR in their previous biopsy, illustrating the progression of DP in these patients. Three of the five other biopsy specimens showed features of grade II AR, and two biopsy specimens showed grade I AR, apart from the parenchymal changes. In Figure 4 the histological features of the studied parenchymal changes are presented.
Figure 5 depicts the progression of the histological changes during the development of CR. Apart from the persistence of the parenchymal features as shown by their high incidence, all features showed an increased severity. A temporary decline at 2 months was observed for CVE and LBI. The progression of changes was not statistically significant apart from the increase of CVF between 1 week and 1 month.
The histological criteria for the diagnosis of CR in liver grafts have been firmly established and consist of obliterative foam cell arteriopathy and DP. In transplantectomy specimens, these features are readily identifiable, although the severity of both histological changes might be variable even within one specimen. On needle biopsies, the diagnosis of CR is mainly based on DP and/or severe damage of bile ducts, because the arterial changes are rarely encountered in the smaller hepatic artery tributaries. A recently published report by an international panel has dealt with the problem of biopsy diagnosis of CR and the grading of bile duct damage preceding obvious DP (2). In this report distinction of early and late features of CR has been summarized in which perivenular events have been recognized as early features (2). Our results support this statement. The present study showed that within the CR group, the lobular features were discernible at an earlier stage than DP, as shown by the incidences of these features. At 1 month, DP was present in only 4 of 9 biopsy specimens of the CR patients; at 2 months, it was present in 5 of 11 biopsy specimens. This finding is compatible with the reported median time of onset of DP, which was beyond the first month after orthotopic liver transplantation (10). At these stages all lobular features showed a higher incidence than DP.
In comparison with the control group, both the incidence and severity of CVE was significantly higher in the CR group than in the control group at 1 month. Conversely, in the control group there was not only a decrease in incidence from 1 week to 1 month but also a significant decline in the severity of CVE. These findings indicated the predictive role of CVE for the transition of AR to CR. The concomitant presence of LBI and CLN strengthened the possibility of a transition to CR, because these were also persistent features in the CR group, with a higher incidence than in the control group. Venous endothelialitis is recognized as a key feature of AR, but in establishing the diagnosis of AR, portal and central vein endothelialitis are not assessed separately. Based on our results, we recommend assessing CVE separately, because it is conceivable that CVE might represent part of a lobular alloreactive process and react differently to antirejection treatment than its portal counterpart. Venous endothelialitis is reported to be sensitive to antirejection therapy, because it is the first feature to disappear after treatment (11). This might explain the slightly declining frequency of CVE observed in our patients after the first month. However, CVE persisted in more than 60% of the CR patients after the first month, which probably represented those patients who were insensitive to antirejection treatment. In contrast with CVE, the other parenchymal features showed a persistent increase as CR progressed. CVF might result from prolonged CVE that showed its decline only after the first month, because a substantial degree of CVF has been reported in 71% of 17 CR transplantectomy specimens (12). A recent study reported a significantly greater incidence of CR in patients showing posttransplantation venous occlusive disease (VOD) (13). Their results were suggestive for the development of VOD through a pathway of CVE, CLN, and VOD (13). CVF and VOD in CR might therefore represent scar formation in the central vein and perivenular areas after prolonged inflammatory damage by CVE and LBI.
In the Banff criteria of AR, CLN was classified as a feature of severe AR when observed in the presence of moderate to severe PI (8). When assessed separately, our results showed that CLN represents another predictive feature of the transition from ACR to CR, by its higher incidence in the CR group, reaching 100% in the late biopsy specimens of the CR group. Its early presence, its persistence, and its increase in incidence during the progression of CR confirmed the presumption that CLN heralded CR. Persistence and increase in frequency of CLN was matched by LBI. The significance of CLN and LBI in the development of CR has been alluded to in earlier studies (14–16). We have also observed a “hepatitic” like phase preceding DP in two of our patients, and a similar finding has been reported recently (17). The persistence of CVE, CLN, and LBI as described above is suggestive that active damage of the centrilobular area in CR is partly mediated by a lobular cellular alloreactivity. Immunophenotyping of intraparenchymally located inflammatory cells during rejection has shown the presence of CD8+, CD57+ natural killer cells and CD68+ macrophages (18). Significant increase of lobular CD8+ cells in CR has been reported before with a selective localization in centrizonal areas (19). Moreover, another study has also reported increased expression of Fas, Fas ligand, and CD40 in zone 3 hepatocytes associated with CLN in CR (20).
The results of the present study showed that a shift from mainly portal-based graft damage in AR toward a lobular-based process represented by CVE, CLN, and LBI might indicate the transition from AR to CR. Therefore, an additional assessment of lobular features is recommended to establish an early diagnosis of CR.
The authors thank Dr. V. Fidler (Department of Epidemiology and Statistics, University of Groningn, NL) for statistical advice.
1. Wiesner R, Batts KP, Krom RA. Evolving concepts in the diagnosis, pathogenesis, and treatment of chronic hepatic allograft rejection. Liver Transpl Surg 1999; 5: 388.
2. International Panel. Update of the International Banff Schema for liver allograft rejection: working recommendations for the histopathologic staging and reporting of chronic rejection. Hepatology 2000; 31: 792.
3. Blakolmer K, Seaberg EC, Batts K, et al. Analysis of the reversibility of chronic liver allograft rejection implications for a staging schema. Am J Surg Pathol 1999; 23: 1328.
4. Freese DK, Snover DC, Sharp HL, Gross CR, Savick SK, Payne WD. Chronic rejection after liver transplantation: a study of clinical, histopathological and immunological features. Hepatology 1999; 13: 882.
5. Hubscher SG, Buckels JAC, Elias E, McMaster P, Neuberger J. Vanishing bile duct syndrome following liver transplantation: is it reversible? Transplantation 1991; 51: 1004.
6. International Working Party. Terminology for hepatic allograft rejection. Hepatology 1995; 22: 648.
7. Neuberger J. Incidence, timing and risk factors for acute and chronic rejection. Liver Transpl Surg 1999; 5: S30.
8. Anonymous. Banff schema for grading liver allograft rejection: an international consensus document. Hepatology 1997; 25: 658.
9. Batts KP, Ludwig J. Chronic hepatitis: an update on terminology and reporting. Am J Surg Pathol 1995; 19: 1409.
10. Van Hoek B, Wiesner RH, Krom RA, Ludwig J, Moore SB. Severe ductopenic rejection following liver transplantation: incidence, time of onset, risk factors, treatment and outcome. Semin Liver Dis 1992; 12: 41.
11. Snover DC, Sibley RK, Freese DK. Orthotopic liver transplantation: a pathologic study of 63 serial liver biopsies from 17 patients with specific reference to the diagnostic features and natural history of rejection. Hepatology 1984; 4: 1212.
12. Nakazawa Y, Jonsson JR, Walker NSL, et al. Fibrous obliterative lesions of veins contribute to progressive fibrosis in chronic liver allograft rejection. Hepatology 2000; 32: 1240.
13. Sebagh M, Debette M, Samuel D, et al. “Silent”presentation of veno-occlusive disease after liver transplantation as part of the process of cellular rejection with endothelial predilection. Hepatology 1999; 30: 1144.
14. Allen KJ, Rand EB, Hart J, Whitington PF. Prognostic implications of centrilobular necrosis in pediatric liver transplant recipients. Transplantation 1998; 65: 692.
15. Turlin B, Slapak GI, Hayllar KM, Heaton N, Williams R, Portmann B. Centrilobular necrosis after orthotopic liver transplantation: a longitudinal clinicopathologic study in 71 patients. Liver Transpl Surg 1995; 1: 285.
16. Nakazawa Y, Walker NI, Kerlin P, et al. Clinicopathological analysis of liver allograft biopsies with late centrilobular necrosis: a comparative study in 54 patients. Transplantation 2000; 69: 1599.
17. Quaglia A, Del Vecchio Blanco G, Burroughs A, Dhillon A. Development of ductopaenic liver allograft rejection includes a “hepatitic” phase prior to duct loss. J Hepatol 1999; 30 (suppl 1): 76.
18. Dollinger MM, Howie SE, Plevris JN, Graham AM, Hayes PC, Harrison DJ. Intrahepatic proliferation of “naïve” and “memory” T cells during liver allograft rejection: primary immune response within the allograft. FASEB J 1998; 12: 939.
19. McCaughan GW, Davies JS, Waugh JA, et al. A quantitative analysis of T lymphocyte populations in human liver allografts undergoing rejection: the use of monoclonal antibodies and double immunolabeling. Hepatology 1990; 12: 1305.
© 2002 Lippincott Williams & Wilkins, Inc.
20. Afford SC, Randhawa S, Eliopoulos AG, Hubscher SG, Young LS, Adams DH. CD40 activation induces apoptosis in cultured human hepatocytes via induction of cell surface Fas ligand expression and amplifies Fas-mediated hepatocyte death during allograft rejection. J Exp Med 1999; 189: 441.