A.J. Demetris, M.D., Jorge Rakela, M.D. and Forrest Dodson, M.D.
Chronic hepatitis C virus-induced cirrhosis is one of the most common indications for liver transplantation throughout the world(1, 2). At most large centers, it accounts for about 30% of all patients undergoing hepatic replacement(2-5). Unfortunately, there is rapid re-infection of the liver allograft from reservoirs of the C virus remaining in the body. In the majority of recipients, this leads to recurrent hepatitis, and recapitulation of the same sequence of events that originally led to liver transplantation. However, similar to the original infection, recurrent disease in the allograft usually evolves slowly and thus hepatic replacement can significantly prolong survival(3). In addition, with the development of new and more effective anti-viral drugs, the potential now exists to break this disheartening cycle of infection, hepatitis, fibrosis and eventually, cirrhosis.
Even if newer anti-viral agents can effectively cure HCV infection, HCV-induced cirrhosis requiring liver transplantation will continue to be a major problem for the foreseeable future. The incidence of infection in the general population ranges from 1-3 % in low incidence areas to 25-30 % in high risk countries or regions. Thus, understanding important issues related to recurrence of HCV infection/disease in the liver allograft will be essential for optimal patient management.
Incidence and Risk Factors for Recurrent and De novo Infection and Disease, and Viral Titers after Transplantation
Re-infection of the liver allograft with the hepatitis C virus after transplantation and subsequent systemic viremia is practically a universal occurrence(6-8). Fortunately, the screening of blood products for HCV, has led to a low incidence of de novo infection, which is on the order of 0.84% in some studies(8, 9), but as high as 10-20% in others(10, 11).
As might be expected with the need for immunosuppression, HCV RNA levels increase markedly following liver transplantation and patients with high pretransplant levels have higher mean posttransplant levels than those with low pretransplant levels(12, 13). Posttransplant levels are similar in patients with recurrent and acquired infection. The use of additional immunosuppressive therapy(discussed below) to treat acute rejection, such as corticosteroids or anti-lymphocyte preparations further enhances viral replication. However, the significance of the actual level as measured by the bDNA assay in an individual patient, is uncertain.
Hepatitis develops in a majority of patients after transplantation (Table 1), and manifests primarily by persistent elevation of the liver injury tests(esp. ALT and AST) to 1.5 - 2 times the normal levels. The liver injury tests may further increase when the first histological signs of hepatitis appear, usually between 3-6 weeks after transplantation, but an earlier onset within 10-14 days has also been observed. By 3-5 years, the incidence of biopsy-proven chronic hepatitis of varying severity, ranges from 13 to > 90% (Table 1). Although the majority of studies report chronic low grade necro-inflammatory activity with HCV-induced disease, a small(about 5%(14, 15)) percentage of patients develop a rapid downhill course, and are very difficult to manage.
Increased immunosuppression to control episodes of acute rejection has been associated with an increased incidence, earlier onset and/or more severe recurrent acute and/or chronic HCV-induced disease in the liver allograft(14, 16-19). In addition, Charlton et al(19) showed that higher viral titers before transplantation are predictive of more aggressive recurrent HCV disease after transplantation, with subsequent lower patient and graft survival(19). The use of immunosuppressive agents associated with a higher incidence of HCV includes interleukin 2 (IL2) receptor antibodies(20), corticosteroids(14, 16, 18), azathioprine, OKT3 and "total" immunosuppression(14, 16). For example, treatment of rejection with corticosteroids has been shown to increase serum HCV RNA from 4-100-fold(12), although again, the significance of viral titers in an individual patient is uncertain.
Studies examining the influence of the primary immunosuppressant on the post-transplant course of HCV have yielded mixed results. Some show that the baseline immunosuppressive agent did not significantly influence recurrent HCV disease activity(18, 19, 21, 22), whereas others suggest that more powerful baseline drugs, such as Tacrolimus, foster the development of more aggressive disease after transplantation(23). Mueller et al(24) suggested that Tacrolimus use in HCV+ recipients was associated with a higher incidence of acute rejection, which in turn, might influence the hepatitis activity, since more treatment with corticosteroids might be required. In a small series of renal transplant recipients with HCV where the confounding influence of rejection and a liver allograft could be avoided, the baseline immunosuppressive agent did not affect the hepatitis C disease activity in the native liver(25). This important area of concern is currently the subject of clinical trials.
Influence of HCV Genotypes
Since most HCV infections in liver allograft recipients represent recurrent or persistent disease, the distribution of HCV genotypes after transplantation generally reflects that seen in the recipient population of that particular center before the operation. In several European centers(26, 27) and in a North American site(28), type 1b is the most prevalent, accounting for 25% to 60%, or more of the patients. In another large American study, type 1a is the predominant genotype(29).
Some studies show that different HCV genotypes are not associated with specific clinical courses of recurrent HCV infection after OLT(8, 19, 30), but many of them report an earlier onset and/or increased incidence of disease, higher viremia, and/or severity of hepatitis in patients infected with the type 1b virus(12, 22, 26-28, 31, 32). Even in the studies where the disease appears to be more severe in those with type 1b virus, there are no significant differences in patient or graft survival by genotype. In addition, there are no specific qualitative histopathological features that can be used to distinguish between the several different viral genotypes on the basis of routine histopathology alone. Thus, more long term studies are needed to evaluate the effect of viral genotype on HCV disease after transplantation.
The Influence of the Donor HCV Status
Since recurrent infection is universal after liver transplantation and in general, HCV has been described as an indolent disease, many centers utilize HCV+ donors. At the University of Pittsburgh, we routinely screen such donors by frozen section histology at the time of harvesting and only use those donor livers that show minimal to mild inflammation and minimal to no fibrosis. Formal evaluation of the policy to use HCV+ donors shows that the rates of recurrent hepatitis and serious disease after transplantation is not affected by the HCV status of the donor(33). Thus, minimally diseased HCV+ donor organs that would have otherwise been discarded, can be used to prolong the life of a recipient with endstage HCV-induced liver failure(33).
When an HCV+ donor liver is implanted into a HCV+ recipient, Laskus et al(34) showed that within a few months after transplantation either the donor(57%) or the recipient(43%) strain predominated. Subtype 1b and type 1(1a + 1b) appeared to have a replicative advantage because they became the predominant strains in all recipient/donor pairs in which they were present. This is consistent with several studies that show a more aggressive course associated with genotype 1b and higher levels of viremia(12, 35). Interestingly, patients retaining their own strain were found to have significantly more active liver disease than those infected by the donor strain(34). This may be related to the existence of primed lymphocytes within the body of HCV+ recipients.
General Statistics on Graft and Patient Survival after Liver Transplantation for HCV
Unless complicated by hepatocellular carcinoma(20), patients undergoing liver replacement for HCV-induced disease at most large centers show patient and graft survival rates that are similar(2, 11, 20, 22, 23) or slightly less(31, 36), than the survival rates for other non-neoplastic indications, except for hepatitis B virus-related disease, which is lower. Figures for patient survival after liver transplantation for HCV at 3 to 5 years are in the range of 63 - 70 %(2, 3, 22, 31), with projected 10 year survival rate of 62%(20).
Even though the patient survival figures are good at 3-5 years, it is clear that recurrent or de novo HCV can cause significant liver allograft damage (Table 1), which is very difficult to manage. Thus, longer follow-up will likely show the presence of HCV infection to have an increasingly negative influence on graft and patient survival(3, 22, 37), including a higher incidence of cirrhosis(22). Moreover, re-transplantation for recurrent HCV disease at some centers is plagued by lower survival rates because of sepsis, other infectious complications and renal failure(15, 23, 37). Therefore, the impetus exists to control the severity of HCV disease in the allograft and avoid retransplantation altogether. However, if retransplantation becomes necessary, it should be carried out before these complications develop(15).
Histopathology and Pathophysiology of Recurrent Disease
The histopathological recognition of recurrent or de novo viral hepatitis, and features used to distinguish it from rejection were first described in detail for hepatitis B(38, 39)(LINK). In general, the same set of histopathological criteria apply to hepatitis C. In fact, these features were used to accurately document the recurrence of non-A, non-B viral hepatitis, even before HCV was discovered(40). However, with the discovery of HCV, and the ability to determine the genotype and measure viral titers, more detailed knowledge has increased our understanding of viral pathogenesis.
In general, the histopathological appearance of HCV in the liver allograft is very similar to that seen in the general population. The virus likely infects the allograft within minutes after transplantation and the first reliable histopathological features that can be used to detect disease activity usually appear within 3-6 weeks, although in some cases, recurrence can be detected as early as 10 - 14 days after transplantation. Once infection occurs, development of the disease follows a predictable evolution through several well-recognized clinical and histopathological phases(3, 40-42).
The acute or initial phase is largely based in the lobule. It is characterized by "lobular hepatitis", which includes lobular disarray, Kupffer's cell hypertrophy and spotty acidophilic necrosis of hepatocytes, a mild increase in sinusoidal mononuclear cells and variable degrees of mononuclear portal inflammation. Macrovesicular steatosis of periportal and midzonal hepatocytes is also often seen. Inflammatory bile duct damage, such as lymphocytes invading the basement membrane with reactive changes in the biliary epithelium, can be focally seen, but it is neither severe, nor widespread. Some studies showed that acute lobular hepatitis on biopsy is accompanied by a steep increase in HCV RNA levels and the appearance of core and NS4 antigens in the graft(12), while others show no strong relationship between level of viremia and degree of hepatic damage(13).
During the transitional phase, the lobular changes usually start to wane, while portal inflammation usually increases, making it an established feature. Thus, there is usually less lobular disarray, inflammation and necrosis and more portal inflammation, which often becomes arranged into nodular aggregates, some of which can develop germinal centers. Again, focal bile duct inflammation and damage can be seen, but it is neither severe, nor widespread. In addition, interface activity(piecemeal necrosis) with an accompanying ductular reaction begin to distort the normal lobular architecture.
The chronic phase usually begins between 4 and 12 months after transplantation and the predominant features are portal inflammation and periportal hepatitis, with varying degrees of lobular disarray and lobular necro-inflammatory activity. Again, inflammatory bile duct damage can be seen, but it is usually not severe, nor widespread and there is no evidence of significant ductopenia. Inflammation beneath the endothelium and/or around the connective tissue sheath of the central vein, known in the liver as central venulitis, can also be seen in an occasional vessel, but similar to the duct damage, it is neither severe, nor widespread. In our experience, patients with an overlapping HCV/autoimmune disorder, can also show a significant plasmacytic component to the inflammation.
Several studies have specifically examined the effect of viral genotype and titers on quantitative and qualitative aspects of the histopathology of recurrent HCV. Some studies showed no correlation between viral titers and the severity of liver damage(43), while others showed that HCV RNA was higher during the lobular phase of the infection and that progression to the chronic phase with interface activity was associated with a highly significant decrease of liver HCV RNA(discussed above). This suggests that progression to chronic active disease is associated with a host immune response to the virus, which is marked by aberrant intrahepatic expression of molecules involved in antigen recognition, intercellular and vascular adhesion, and recruitment and activation of cytotoxic T lymphocytes(43). In fact, higher levels of viral replication during the lobular phase(35) might precipitate a stronger response because they are an independent predictor of progression to chronic active disease in some studies(35). Other studies have shown that ballooning degeneration and cholestasis at initial presentation(44) are associated with the subsequent development of allograft cirrhosis(44). In one study there was no significant difference in the histopathology of recurrent versus de novo disease, whereas in another, de novo infection more often lead to significant disease(41).
Histopathological Differential Diagnosis
The relative difficulty in distinguishing between recurrent or de novo viral hepatitis C and other causes of liver allograft dysfunction depends on the time after transplantation. During the first several weeks, HCV is an uncommon cause of allograft dysfunction, although in isolated cases the typical sequence of changes described above, can begin in 10 - 14 days. However, the majority of cases of allograft dysfunction attributable to recurrent hepatitis C usually have a slightly later onset, between 3 and 7 weeks after transplantation. In contrast, the majority of acute rejection episodes occur within the first 30 days, with a median of 8 days(45).
The most useful histopathological feature in distinguishing between hepatitis and other causes of allograft dysfunction is the primary focus of the necro-inflammatory activity. In the lobular or acute phase of hepatitis, the most conspicuous changes are in the lobules, as described above. Portal inflammation is usually not prominent during this time, but can be seen, and in some cases, matches the severity seen in acute rejection. The key to distinguishing between the two cause of dysfunction in such cases is the comparative paucity of panlobular necro-inflammatory activity in acute rejection, although inflammation in and around the central vein can be seen; the prevalence of inflammatory bile duct damage, which is greater in acute rejection than in hepatitis; the zonal regularity of changes associated with acute rejection in contrast to the irregular findings in hepatitis; and the clinical history, including the time after transplantation. The difficulty lies in cases of graft dysfunction occurring between 3 and 4 weeks after transplantation.
Established chronic hepatitis is usually not difficult to distinguish from acute or chronic rejection. However, acute and/or chronic rejection and recurrent HCV can occur together, making it difficult in some cases to determine which is the more important cause of allograft injury and dysfunction. In addition, any other cause of allograft dysfunction that is associated with chronic portal inflammation will also enter into the differential diagnosis. This includes recurrent or de novo HBV, autoimmune hepatitis, PBC, drug-induced chronic hepatitis, chronic obstructive cholangiopathy and non-A-G hepatitis.
In distinguishing between the above possibilities, the pathologist should first determine whether or not acute rejection is present. The key features used to identify rejection are a) the prevalence and severity of mononuclear inflammatory bile duct damage; and b) prevalence and severity of central vein inflammation. Both of these features are more prevalent (i.e., involve a greater percentage of the structures present) and are more severe, in acute rejection than in non-rejection related causes allograft dysfunction. Conversely, the key feature used to identify non-rejection causes of chronic portal inflammation, including HCV, are a) the level of activity at the interface zone and b) the presence of a ductular response, both of which are minimal or absent in rejection. Thus, histological evidence of type I and II ductular reactions(46), and creation of a periportal epithelial-mesenchymal wedge, formed by biliary epithelial cells and the surrounding stroma that progressively distorts the lobular architecture, is strong evidence of a non-rejection related cause of allograft dysfunction.
Once acute and chronic rejection have been confidently excluded, distinguishing HCV from HBV, autoimmune hepatitis, drug induced hepatitis, PBC and obstructive cholangiopathy follows the same set of guidelines as those used in the general population. These are based on a combination of clinical and histopathological findings, and therefore are briefly discussed here. For example, the distinction between recurrent or de novo autoimmune hepatitis and HCV is largely based on the clinical profile and the results of laboratory tests, other than the liver biopsy, such as auto-antibodies(e.g. ANA, ASMA, LKM, etc.), and the clinical profile(e.g. original disease is autoimmune or has autoimmune component). There are no specific histopathological features that can reliably distinguish between autoimmune and HCV-induced hepatitis, although it has been our experience that autoimmune disease shows more plasma-cell rich inflammation, and less steatosis and portal lymphoid nodules.
The distinction between HCV and recurrent PBC can be quite difficult in some cases because granulomatous duct destructive lesions can be seen in allograft recipients with HCV(47) and patients with PBC before transplantation, can also harbor HCV or acquire it during, or after the operation. Fortunately, since de novo PBC after liver transplantation is rare, simply knowing the original disease greatly facilitates sorting through the various histopathological diagnostic possibilities. In cases where difficulties are encountered, the key histopathological clue in distinguishing between the two disorders turns on recognition of the "biliary" gestalt that typically evolves in biopsies with recurrent PBC, but not in HCV. This includes patchy non-suppurative destructive cholangitis (florid duct lesion); a prominent ductular reaction at the interface zone, which is combined with small bile duct loss and periportal edema (halo sign); periportal lysosomal pigment and copper/protein deposition; and only mild lobular changes.
Obstructive cholangiopathy is usually not difficult to distinguish from chronic HCV, although on occasion, incomplete biliary tract obstruction or stricturing can be associated a predominantly mononuclear portal tract infiltrate and interface activity that vaguely resembles that seen in chronic hepatitis. More often however, the major difficulty is distinguishing between acute rejection and obstructive cholangiopathy, which is discussed elsewhere.
Fibrosing Cholestatic Hepatitis
A particularly devastating presentation of HCV in liver allograft recipients deserves special mention, even though it is uncommon. It was first described with hepatitis B virus infection in liver allograft recipients(39, 48), and termed "fibrosing cholestatic hepatitis" (FCH), which is a descriptive phrase that highlights the most pertinent histopathological features. Shortly after its discovery, FCH was also soon recognized as a complication of HCV in liver allograft recipients as well(41, 49-52).
HCV-associated FCH is characterized clinically by malaise, jaundice, and a marked increase in the cholestatic liver injury tests, such as bilirubin, alkaline phosphatase and gamma glutamyltranspeptidase. The disease often evolves subacutely over a period of weeks to months with the development of functional liver failure with encephalopathy and/or cirrhosis. The key to early recognition of this disorder relies on a suspicion by both the clinical physician and the pathologist. The clinical findings have already been discussed. On histopathologic examination, the typical case shows extensive centrilobular hepatocyte swelling and degeneration, cholestasis, spotty acidophilic hepatocyte necrosis, and Kupffer's cell hypertrophy, combined with portal tracts expansion because of ductular proliferation, fibrosis and a mild mixed or neutrophilic predominant portal infiltrate(41, 51). In such cases, the combination of predominantly neutrophilic portal inflammation, ductular proliferation, acute cholangiolitis and cholestasis, mimics obstructive cholangiopathy and/or sepsis.
The key to distinguishing FCH from other causes of dysfunction that show similar histopathological findings lies in the clinical history; the results of other diagnostic tests; and close histopathological examination of the biopsy slide. Information in the clinical history which should arouse suspicion of FCH includes HCV/HBV positive serology/PCR; recent and/or intensive immunosuppressive treatment; relatively rapid onset of jaundice without evidence of acute rejection on biopsy; and lack of evidence of either biliary tract obstruction or stricturing, sepsis or an intra-abdominal infection. Features on the biopsy that are helpful in distinguishing FCH from obstructive cholangiopathy include the lack of acute cholangitis and periductal edema in the true bile ducts(as opposed to proliferated cholangioles), spotty acidophilic necrosis of hepatocytes and lobular disarray, all of which are typical of FCH, but either not seen or not as prominent in obstructive cholangiopathy.
In contrast to the typical chronic HCV with prominent mononuclear inflammation, the liver allograft damage observed in FCH does not appear to be related to an immune response by the host, at least on the basis of routine histopathological findings. In contrast, there is some evidence that the liver damage in HCV FCH may actually be related to persistently high HCV RNA levels(52, 53), as with HBV(39). Thus, there may be two different mechanisms of liver allograft injury related to HCV, which in turn, might depend on the immune status of the recipient. When the recipient is capable of mounting a relatively strong immune response, the liver damage is likely mediated by the immune system and the histopathology shows prominent mononuclear portal inflammation, similar to that seen in the general population. In contrast, when the host is severely immune compromised, unchecked viral replication could conceivably cause the liver injury, which takes on the histopathological appearance of FCH.
Co-Infection with Hepatitis B and G Viruses
Huang et al(54) described the evolution of hepatitis in liver allograft recipients with HBV and HCV co-infections. They found two major histopathological patterns: one with predominant features of HCV infection, and the other with those of HBV infection. Although the study was limited by the number of patients, it appeared that the presence of HCV might actually improve the clinical outcome of HBV patients, as compared with the expected outcome of persistent HBV infection alone(54). In contrast, others have found that HBV-related disease more frequently dominates the post-transplant course and by dominating the course, it adversely affects patient and allograft survival(55).
Co-infection with the hepatitis G virus does not appear to influence either the clinical or pathological course of recurrent or de novo hepatitis C in liver allograft recipients(56). This is likely related to the fact that the hepatitis G virus does not replicate in the liver(57), at least in patients with hepatitis C and G co-infections.
Chronic Rejection Complicating Hepatitis C Virus Infection
Several groups have reported a higher incidence of chronic(ductopenic) rejection in HCV+ recipients(5, 58, 59). There are several possible non-exclusionary explanations for this association. First, it is well known that viral infections can cause an inflammatory microenvironment within an allograft, which in turn, can upregulate adhesion, co-stimulatory and MHC antigens(60-62). All of these factors, in concert, can precipitate a rejection response. Partial MHC class I compatibility between the donor and recipient might permit MHC-restricted T cell-mediated response to viral infection at the same time as an allogenic response(60-62). In addition, the ability to properly respond to infectious and environmental antigenic challenges may be compromised in allografts, leading to a vicious cycle, alternating between a rejection and infectious immune reaction(63). These scenarios are made even more likely if there is a concomitant reduction in immunosuppression or the addition of an immune stimulator like alpha-interferon(64, 65), either of which can be used to treat the viral infection.
Finally, as discussed above, there exists a significant problem in the misdiagnosis of acute rejection as hepatitis and vice versa, particularly during early onset HCV associated with mononuclear or mixed portal inflammation. Close attention to the histopathological findings however, can avoid this potential problem in many, but not all, cases. Reference to the above discussion about the histopathological differential diagnosis has already covered this topic in detail.
It is important for both the clinician and the pathologist to be aware of protocols used for HCV treatment after transplantation and how these might influence the histopathological changes seen in biopsy and HCV-induced disease activity. As discussed above, the universal recurrence, high incidence of chronic hepatitis, and relatively poor outcome of recipients undergoing hepatic retransplantation because of HCV, has necessitated the treatment of HCV after transplantation with the same drugs used to treat it before hepatic replacement. Several major problems exist however: 1) the immune activating properties of alpha-interferon-2B(IFN) can potentially trigger a rejection reaction; 2) the presence of immunosuppression theoretically decreases the likelihood that the infection will be cleared since this also requires an immune response. Both of these theoretical limitations have manifested to a variable extent in the patient care arena, but their impact has not led clinical physicians to abandon the treatment effort.
For example, in the general population, IFN treatment alone results in clearance of serum HCV RNA in 15-20% of HCV+ patients. This percentage can be increased to 50% or more when IFN is combined with Ribavirin. In contrast, IFN treatment alone in HCV+ liver allograft recipients is relatively ineffective, regardless of whether it is used as therapy after the hepatitis is diagnosed, or prophylactically after transplantation in all HCV+ recipients(66-70). In allograft recipients, IFN rarely, if ever, leads to clearance of HCV RNA from the serum. In addition, treatment with IFN increases the risk of rejection in some(68), but not most studies(67, 69, 70). Fortunately, the combination of Ribavirin and IFN in liver allograft recipients has led to some encouraging preliminary results, in this otherwise disappointing area of patient care.
The typical treatment protocol consists of 3 million units of alpha-IFN, given 3 X per week, intramuscularly, over a period of 6 months to a year, which is combined with 1 - 1.2 grams Ribavirin/day, given orally. With this combination protocol, Bizollon et al(66, 67) described a return to normal levels of serum ALT levels in 90 % or more of HCV+ liver allograft recipients; about a 50% rate of clearing of serum HCV RNA by RT-PCR during therapy, with others showing a significant decrease in HCV RNA; and histopathological improvement in a majority of patients. Unfortunately, after the treatment protocol was completed, they also describe reappearance of HCV RNA in 50% of the HCV+ recipients who initially cleared the virus(67).
Gane et al(69), compared treatment with IFN alone, to Ribavirin alone, and found that Ribavirin more often led to a return to normal of liver injury tests and decreased the lobular inflammation, but did not improve total liver necro-inflammatory activity, nor did it decrease the level of viremia. In contrast, IFN alone was less effective in normalizing the liver injury tests and decreasing the lobular inflammation, but it significantly decreased viremia. Thus, the evidence to date strongly suggests that in liver allograft recipients, combination therapy with IFN and Ribavirin is superior to therapy with either agent alone. Importantly, those treated with regimens containing IFN should be monitored more closely than others for the development of rejection. However, the threat of rejection should not preclude IFN use, particularly when it has the potential for a significant therapeutic benefit when combined with Ribavirin.
Currently, a multi-center trial is underway to test the efficacy of Ribavirin/IFN when given empirically beginning 6 weeks after transplantation. Additionally, investigators have attempted to use this combination of drugs immediately after transplantation in an attempt to prevent subsequent infection of the allograft. The final results of both of these trials have not been reported. However, since the five year survival rate does not appear to be significantly affected by the presence of HCV infection, any empirical or prophylactic treatment regimens must be safe, as well as effective. Hopefully, pharmacological treatment will be developed that can clear HCV infection prior to transplantation, so clinicians will not have to contend with it after transplantation, when adequate control is much more difficult.