hepatitis c viral infection in liver transplant recipients

Hepatitis C Viral Infection in Liver Transplant Recipients

LINDA D. FERRELL,~ TERESA L. WRIGHT,2 JOHN ROBERTS,3 NANCY h C H E R 3 AND JOHN LAKE? Departments of 'Pathology, 2Medicine and 3Surgery, the School of Medicine, University of California-San Francisco,

San Francisco, California 94143

In this study we examined multiple serial liver biopsy specimens h m liver transplant recipients to determine the pathological features of hepatitis C virus- induced hepatitis. Hepatitis C virus infections acquired after transplantation and previous infections that recurred in patients after transplantation were confbmed by the results of the polymerase chain reaction. Of 43 patients infected with the hepatitis C Vinrs, 18 had a mild form of chronic hepatitis. Four patients had hepatitis that progressed to focal bridging fibrosis or cirrhosis. There were no significant clinical or pathological differences between infections acquired after transplantation and recurrent infections (as determined by polymerase chain reaction) except that acquired infections more often developed into hepatitis. Findings indicative of hepatitis C infection included portal and parenchymal mononuclear Mil- trates of varying degrees, acidophilic necrosis and Bwollen hepatocytes. Other common findings included lymphoid aggregates, bile duct damage and fatty change. Atypical pathological conditions included extensive hepatocyte swelling or acidophilic necrosis with minimal idammation mimicking ischemia and ductal or ductular damage and proliferation with mixed portal Miltrates mimicking rejection or ob- structioa. We conclude that in transplant recipients infection by the hepatitis C virus usually produces a mild disease state, but the diagnosis of hepatitis can be difficult to make because indicators of hepatitis may mimic those of rejection, ischemia, obstruction or other hepatic infections. Serial biopsy specimens with persistent pathology and polymerase chain reaction may be necessary to define the preeence of a hepatitis C virus lesion. (HEPATOLOGY 1992; 16:865-876.)

Hepatitis C virus (HCV) infection is a leading cause of end-stage liver disease requiring liver transplantation (1) and is responsible for much of the non-A, non-B (NANB) hepatitis that follows blood transfusions (2, 3). Until recently, HCV was a diagnosis of exclusion because diagnostic tools were unavailable to definitively confirm the infection. Now with the development of molecular techniques such as the polymerase chain reaction (PCR), HCV can be identified with a high degree of confidence

Received November 29, 1991; accepted May 17, 1992. Addressreprintrequeststo: LindaD. Ferrell, M.D., DepartmentofPathology,

School of Medicine, University of California, San Francisco, CA 94143-0506. 31/1/39487

even in early stages of the disease and even in those immunosuppressed patients without adequate antibody response to infection (4, 5 ) .

In this study we examined multiple serial biopsy samples from liver transplant recipients with HCV to define the pathological features of HCV infection in liver transplant recipients. There are several reasons for such a study. Although the pathology of NANB hepatitis has been adequately described in patients who did not receive transplants (without molecular confirmation of the diagnosis by PCR) (6, 71, immunosuppression may alter the histological appearance of viral hepatitis after transplantation, as has been shown for hepatitis B (8,9). Thus, although diagnostic findings of hepatitis (e.g., mononuclear infiltrates in portal and parenchymal zones and the presence of acidophilic hepatocytic necrosis) in liver biopsy samples can indicate infection, at times a patient with viral hepatitis may initially or only have swollen hepatocytes, acidophilic bodies without a cellular mononuclear infiltrate (a cytopathic effect) or both. Therefore correct diagnosis may elude the unwary physician. Transplant recipients also may have a variety of other problems such as rejection, obstruction, cytomegalovirus ( 0 - r e l a t e d hepatitis, ischemic injury or a drug reaction (10). All of these complications can mimic some of the findings seen in HCV infection, or they can occur simultaneously with the hepatitis. Thus it may be difficult to determine to what extent certain clinical or pathological alterations are the result of HCV infection by examining only one biopsy specimen. Fi- nally, the advent of interferon therapy for HCV infection (11, 12) has made it more imperative that the patients with HCV infection be separated from those whose liver dysfunction is a result of other causes, because the side effects of interferon treatment can cause considerable morbidity (13).

MATERIALS AND METHODS In this study we examined 43 liver transplant recipients

from the University of California-San Francisco from 1988 to 1991. These patients tested positive for HCV by PCR of sera, liver samples or both gathered after transplantation, and they had at least 1 yr of follow-up. The total number of transplant recipients during this time was 236; 177 patients with adequate serum samples before and after transplantation were tested for HCV by PCR. Sera from the remaining cases were unavailable for PCR.

The patients were divided into two groups, those with

865

866 FERRELL ET AL HEPATOLOGY

TABLE 1. Clinical parameters of patients with histological hepatitis

Patient no. hepatitis after transplantation End point classification AST end point” Active phase Duration of First hepatitis diagnosis

Group Ab 3.5 mo CAH, severe 10 x normal - 1 l y r 1 mo CAH, moderate 4 x normal - 2 l y r 3 mo CAH, moderate 9 x normal - 3 2yr

4 l y r 7 mo CAH, mild 7 x normal + 5 1.5 yr 4 mo CAH, mild 5 x normal -

CAH, mild 3 x normal - 6 1.5 yr l y r 7 l y r 5 mo CAH, mild 4 x normal +

4 mo CAH, mild 2 x normal - 8 l y r CPH 3 x normal - 9 1.5 yr l y r

10 1.5 yr 5 mo CPH 1.5 x normal -

11 l y r 2 mo CAH, severe 3-6 x normal + 12 1.5 yr 4 mo CAH, mild

10 mo CAH, mild 4 x normal - 13 2yr 3 mo CAH, mild 2 x normal - 14 2yr

15 2.5 yr 3 mo CAH, mild 1.5 x normal - 16 2.5 yr 7 mo CPH 4 x normal + 17 1.5 yr 4 mo CPH 1.5 x normal - 18 l y r 19 l y r 20 l y r 5 mo CLH/CPH Within normal limits + 21 2yr 11 mo CLWCPH 2 x normal +

CLH/CPH 4 x normal - 22 1.5 yr l y r

Group R”

- 1.5 x normal

3 mo CLH Within normal limits - 4 mo CLH 4 x normal -

“Classification of hepatitis and elevation of AST by end of study; the latter is expressed as multiples of normal before interferon therapy

bTotal no. of HCV-positive patients with and without hepatitis = 14. ‘Total no. of HCV-positive patients with and without hepatitis = 29.

(patients 1, 7, 8, 9, 10, 14, and 16).

infections acquired after transplantation (group A) and those with infections recurring after transplantation (group R). Two patients who tested positive before transplantation (one was also positive for HBV) but who were negative after transplantation and seven patients with coexisting HBV infection, as determined by the detection of HBsAg in serum, PCR in serum or both, were excluded from the study. One patient who tested positive for hepatitis in one biopsy specimen but who tested negative for hepatitis in another biopsy specimen 2 wk later was not counted as definitively positive for HCV infection. This was done to ensure that the overall findings of histological hepatitis were conclusive for the tabulation of histological results and not representative of CMV or a nonspecific reaction to another systemic infection or process. In addition, all patients with CMV-related hepatitis or with positive CMV cultures were not tabulated as having HCV infection.

We also examined 21 controls chosen at random. This group consisted of transplant recipients who tested negative for HCV by PCR and who were followed for similar time periods. The control patients had the following primary diseases: two had PBC, four had primary sclerosing cholangitis, five had alco- holic cirrhosis, two had autoimmune hepatitis, five had cryptogenic cirrhosis, one had hemochromatosis, one had a,-antitrypsin deficiency and one had Budd-Chiari Syndrome. Only the final two samples from the control patients were reviewed, corresponding with a PCR result that was negative for HCV in the same approximate time period. Patients 1,2,3, 4, 7, 11 and 19 received interferon for 3 to 6 mo during the study.

Histological Review. Each patient underwent liver biopsies weekly (“protocol biopsies”) until two consecutive specimens

were normal; they had yearly biopsies thereafter. A biopsy was also performed when liver function test results (bilirubin, AST, alkaline phosphatase or all three) were abnormal on routine screening. Each specimen was fixed in neutral-buffered for- malin, embedded in paraffin, sectioned at 5 pm and stained with hematoxylin and eosin. At least three level sections of each specimen (HCV-positive patients and controls) were reviewed by the pathologist (L. D. F.), who was blinded to the patient’s serological results and PCR status for the presence or absence of possible hepatitis. The number of biopsy samples reviewed for each patient ranged from 3 to 20 (mean = 9). The average length of each specimen was 1.5 to 2.5 cm. The corresponding liver function test results were recorded at the time of each biopsy.

The diagnosis of histological hepatitis required the presence of some degree of portal and parenchymal mononuclear infiltration and isolated hepatocyte necrosis as evidenced by acidophilic bodies. When present, hepatitis was classified as chronic after it had been present for 6 mo, as determined clinically or histologically. The chronic lesions were further classified as CAH, chronic persistent hepatitis (CPH), chronic lobular hepatitis (CLH) or CPH and CLH combined. For a diagnosis of CAH, a portal mononuclear infiltrate with piecemeal necrosis and disruption of the limiting plate by inflammatory cells was necessary. Fibrosis was not necessary for CAH. In addition, CAH was graded as mild if only focal, spotty, piecemeal necrosis was present with or without minimal fibrosis; as moderate if there was moderate piecemeal necrosis with fibrosis, with or without significant cell swelling; and as severe if there was moderate to severe piecemeal necrosis with bridging fibrosis (possibly cirrhosis), with or

Vol. 16, No. 4, 1992 HCV IN LrVER TRANSPLANT PATIENTS 867

FIG. 1. Expanded portal zone with intense infiltrate, predominantly mononuclear, in a patient with severe CAH. Hepatocytes in the periportal zone are swollen. Inset: Higher magnification shows the infiltrate to be predominantly lymphocytic. Probable lymphocyte is present in the duct wall.

without cell swelling. A prominent portal mononuclear infiltrate (without piecemeal necrosis, portal fibrosis or disruption of the limiting plate) and a scant parenchymal infiltrate with acidophilic necrosis (1 + ) were labeled CPH. A more prominent lobular hepatitis with 2 + to 3+ acidophilic necrosis and no piecemeal necrosis or fibrosis was labeled CLH.

The predominance of portal vs. parenchymal inflammation and the location of swollen hepatocytes (diffuse vs. centrilobular) were noted. The major histological features that we evaluated were aggregates of lymphocytes (so-called lymphoid aggregates), ductal damage and fatty change, because these have been previously associated with NANB hepatitis in patients who did not receive transplants (6, 7). We also determined whether the proliferating cholangioles or the interlobular bile ducts were damaged (14). Other parameters such a8 rejection, duct damage with or without concurrent rejection, fibrosis with or without bridging or cirrhosis and cholestasis were also investigated. A diagnosis of rejection or consistent with rejection was made if the portal tracts contained a mixed inflammatory infiltrate (lymphocytes, eosinophils and neutrophils) with inflammation of the ducts or d u d e s and ductal or ductular damage as evidenced by damage or loss of their lining epithelium.

H C V A n t i M y and PCR Methods. Serum from each patient, which had been frozen at the time of transplantation, and one sample obtained more than 3 mo after liver transplantation were stored at - 70" C and tested for HCV antibodies by ELISA (Ortho Diagnostic Systems, Raritan, NJ), as previously described (1). Samples for PCR from these specimens were placed in 0.1% diethyl pyrocarbonate. Nucleic acid was extracted from

50 p1 of serum or 20 to 30 mg of liver by guanidinium isothiocyanate and phenol-chloroform-isoamyl alcohol, as previously described (4, 5).

RESULTS Clinical Groups and Parameters. At some time after

their liver transplantations, 51% of the patients (22 of 43) with HCV infection detected by PCR had histological evidence of hepatitis other than CMV or HBV infection (Table 1). In group A 10 of 14 patients (71%) had hepatitis. In group R only 12 of 29 (41%) had hepatitis. Four of the patients (three in group A and one in group R) progressed to a moderate or severe form of CAH (Fig. 1) with fibrosis, end-stage cirrhosis or both. Two of these patients received new transplants, one died before retransplantation and one had stable cirrhosis. The other patients exhibited mild CAH (Fig. 2), CPH (Fig. 3) or CLH (Fig. 4) on histological sections. Evidence of definite hepatitis in more than one biopsy specimen, as defined by the concurrent findings of necrosis and inflammation described above, could not be found in 21 patients. One patient did show hepatitis in one biopsy sample, but in a biopsy sample taken 2 wk later, hepatitis could not be detected. Whether this represented a self-limiting HCV lesion or changes caused by other infections or drug therapy is unknown; therefore this patient was not tabulated as having definite hepatitis.

868 FERRELL ET AL. HEPMOLOGY

FIG. 2. Focal lymphoid aggregate and minimal inflammatory activity at the limiting plate in a patient with mild CAH. Note acidophilic body (arrow).

Some of the hepatitis-negative patients did show occasional mild portal or parenchymal mononuclear infiltrates (Table 21, or they had normal biopsy specimens. The control patients, who did not have HCV infection detectable by PCR, showed similar mild degrees of portal and parenchymal infiltration by lymphocytes; thus the controls did not differ significantly from the HCV- positive patients without hepatitis.

The total number of biopsies performed in each patient with hepatitis ranged from 5 to 19 (mean = 9). The number of biopsy samples per patient that actually revealed histological evidence of hepatitis ranged from 2 to 14 (mean = 4). Each patient without hepatitis had between 3 and 13 biopsies (mean = 6). Hepatitis developed in patients 1 mo to 1 yr after transplantation (mean = 4 mo). The hepatitis-positive patients were followed for 6 mo to 2 yr after the diagnosis was made (mean = 1.5 yr). The hepatitis-negative group was followed for 1 to 3 yr (mean = 2 yr). The control group was followed for 1 to 3.5 yr (mean = 2 yr).

Histological Study of HCV Patients Mer Transplan- tation. Some of the pathological features seen in liver biopsy samples of hepatitis C patients are summarized in Tables 2 and 3. All patients with hepatitis showed the histological features of portal and parenchymal inflammation at some point during their course. Acidophilic necrosis of hepatocytes was invariably present re- gardless of the degree of inflammation. The histological changes were separated into groups according to the

natural history and temporal development of HCV infection in these patients (Tables 2 and 3). The earliest stages of hepatitis in all patients routinely consisted of spotty hepatocyte necrosis, often with no or only mild cell swelling. Inflammation in the portal zones and parenchyma was present but at variable degrees. Al- though in seven cases acidophilic necrosis was present in biopsy samples taken before the definitive diagnosis of hepatitis, the diagnosis of hepatitis was withheld because of the absence of a fully developed triad of portal and parenchymal inflammation associated with the necrosis. In these cases the lesions progressed to fully developed hepatitis within 2 to 4 wk. In six patients prominent histological changes associated with markedly elevated liver function tests were present for 2 to 6 mo (mean = 3 mo). After 3 to 6 mo each of these six patients’ diseases subsided into milder forms of chronic hepatitis, and the patients had uniform and consistent histological findings in subsequent biopsy specimens up to the end of this study. The remainder of the patients had consistent histological findings on each subsequent biopsy once the diagnosis of hepatitis was made. The early active phase seen in the six patients above demonstrated more acidophilic necrosis, cell swelling and ductular damage with rejection than was seen both in their later, mild hepatitis and in that of the patients without this active phase. The patients with severe CAH throughout their course, however, often showed more ductular damage also.

Vol. 16, No. 4, 1992 HCV IN LIVER TRANSPLANT PATIENTS 869

FIG. 3. CPH with portal mononuclear infiltrate confined to portal zone. Note lymphocyte in the wall of the interlobular duct furrow).

Diffuse hepatocyte swelling and zonal swelling around the central vein were frequently seen in all of the hepatitis patients, and centrilobular swelling was uniformly present in the six patients with the active phase (Table 2 and Fig. 5). Swelling and acidophilic bodies were often not associated with an inflammatory reaction around the affected cells or in the nearby parenchyma (Figs. 5 and 6). This minimal inflammatory response was seen in at least one of the serial biopsy specimens, with hepatitis in 36% of the patients (8 of 22). The enlarged cells at times would maintain their eosinophilic cytoplasm (Fig. 7) but at other times (even on different biopsy samples from the same patient) would demon- strate a paler cytoplasm more typical of ballooned cells (Fig. 5) . In the zones adjacent to areas of swollen hepatocytes, smaller hepatocytes were often seen, some- times with double cell plates (6 of 22 cases or 27%). These smaller cells may represent regenerating hepatocytes. (Fig. 7).

Duct damage associated with concurrent definitive rejection (confirmed by biopsy reassessment after therapy) was seen in 50% of the patients (11 of 22) (Figs. 8 and 9); 50% (1 1 of 22) showed minimal duct or ductular damage at some point in time without significant rejection (confirmed by biopsy and clinical follow-up). Both interlobular ducts and proliferating ductules showed damage, although the proliferating ductules were significantly more involved. Ductal or ductular damage without concurrent rejection was not as extensive as that caused by rejection, and it usually

consisted of a periductal infiltrate of lymphocytes, occasional mononuclear cells adjacent to the basement membrane or within the duct epithelial wall (Figs. 1 and 3) and possible slight foci of duct epithelial loss. Neutrophils and eosinophils were not usually present.

Lymphoid aggregates (Figs. 2 and 3) were seen in at least one biopsy sample in 17 patients (77%) (Tables 2 and 3). Patients with the early active phase showed no lymphoid aggregates during that phase, but three (50%) had them later. Four patients had only one aggregate in only one biopsy specimen. In two patients lymphoid aggregates first appeared in biopsy samples after 1 yr of hepatitis.

Fatty change was seen in 64% of the cases (14 of 22). Cholestasis was limited to the centrilobular zones in the patients with active phases of hepatitis and it occurred late in the course of the severe cases of CAH.

Interferon therapy in seven patients (Table 1) did not significantly interfere with our tabulations. We cited only histological features present before interferon therapy. Four of the patients had moderate to severe lesions that remained unchanged during therapy, and the other three had mild lesions that did not progress. In these latter three patients, the end point may have been altered, and progression to cirrhosis for the other four may have been slowed.

DISCUSSION This study helps to establish several clinical and

histological features of HCV infection in liver transplant


FIG. 4. CLH with spotty acidophilic necrosis of hepatocytes in the lobule. Note the mononuclear cells associated with the necrotic cells almost surrounding the cell on the left.

TABLE 2. Pathological conditions of HCV-infected patients vs. those of controls HCV-positive with hepatitis (n = 22)

HCV-positive Consistent activity HCV-negative controls without hepatitis Early active phase thughout

(n = 16) Morphological features (n = 21) (n = 21) (n = 6)

Mononuclear infiltrate Portal 11 (52)" Parenchymal 12 (57)

Acidophilic necrosis 1 (5) Lymphocyte aggregates 5 (24) Hepatocyte swelling 0 (0 )

14 (67) 6 (100) 16 (100) 9 (43) 6 16 (100) 0 (0) 6 (100Ib 16 (100) 2 (9) 0 early (OF 14 (87) 2 (9)d 6 10 (62)

"Data expressed as no. of patients showing the features in at least one biopsy specimen (percentage of the whole). bIncreased degree of change in active phase as compared with later phase. cLymphoid aggregates developed in three patients during later stages (after 6 mo). dPatients had focal, mild changes only.

recipients. From a clinical standpoint, it is important to note that only 51% of the patients had histological hepatitis even in their immunocompromised state. Of these patients 41% had a relatively mild form of the disease, and only four patients (9%) had significant fibrosis or cirrhosis, which has resulted in death or retransplantation in three of these patients to date. Overall, the predominant effects of HCV infections on this population appear to be rather mild and would most likely not result in loss of the allograft in 2 to 3 yr. Longer follow-up will show whether HCV infection in these patients becomes more destructive and results in

loss of the graft. These findings suggest that hepatitis C, whether acquired or recurrent, is a less aggressive form of hepatitis than hepatitis B in liver transplant recipients (8, 9).

From a histological standpoint, HCV infection usually begins with spotty necrosis and variable degrees of mononuclear inflammation and cell swelling, followed either by an active phase that subsides into a chronic hepatitis (6 of 22 or 27%) or by a consistent histological picture of chronic hepatitis that remains at a stable level of activity throughout the course (16 of 22 or 73%). As previously described (6,7) in patients not receiving


FIG. 5. Centrilobular ballooning of hepatocytes and scant inflammation mimicking ischemia. Note acidophilic body more typical of hepatitis (arrow).

TABLE 3. Temporal sequence of histological findinga in HCV infection Biopey 611ding~

Before diagnosis Dilyploeis-2 wk 2-8 wk 2-4 mo 4-6 mo 6-12 mo >12 mo Morphological feat- (n = 7) (n = 22) (n = 9) (n = 11) (n = 10) (n = 14) (n = 10)

Mononuclear infiltrate Portal Parenchymal

Acidophilic necrosis Lymphoid aggregates Hepatocyte swelling Fatty change Dud damage Possible rejection

22 (100) 22 (100) 22 (100) 8 (36)

13 (59) 13 (59) 14 (64) 7 (32)

9 (100) 9 (100) 9 (100) 3 (33) 6 (67) 7 (78) 4 (44) 3 (33)

11 (100) 11 (100) 9 (82) 5 (45) 7 (64) 6 (54) 8 (73) 4 (36)

10 (100) 10 (100) 10 (100) 3 (30) 7 (70) 7 (70) 6 (60) 2 (20)

14 (100) 14 (100) 14 (100) 5 (36) 5 (36) 6 (43) 5 (36) 4 (28)

~~ ~

Before diagnosis = specimens taken within 1 mo before the definitive dllagnosis of hepatitis; diagnosis-% wk = specimens taken at the time of diagnosis and for the following 2 wk; 2-8 wk = specimens taken 2 wk after diagnosis and through the following 6 wk; 2-4 mo = specimens taken 2 mo after diagnosis and through the following 2 mo; 4-6 mo = specimens taken 4 mo after diagnosis and through the following 2 mo; 6-12 mo = specimens taken 6 mo after diagnosis and through the following 6 mo; > 12 mo = specimens taken 12 mo after diagnosis.

"Data expressed as no. of patients (percentage of whole).

transplants, typical histological features of viral hepatitis in general, and of NANB specifically, were also seen in this group of immunocompromised patients. The classical findings of viral hepatitis included portal mononuclear infiltrates, parenchymal mononuclear infiltrates or both and spotty acidophilic necrosis as seen in immunological injuries. Other typical features frequently present included swollen hepatocytes and regeneration, centrilobular injury (inflammation and

swelling), lymphoid aggregation, fatty change and duct damage. The latter three findings were more frequently induced by HCV infection than by other causes of hepatitis (6, 71, but in our population some difEculties arose in determining whether rejection or hepatitis was responsible for the duct damage and whether drug therapy (such as prednisone) or nutritional factors caused the fatty change. The lymphoid aggregates were not seen in all patients with hepatitis, as has been

872 FERRELLETAL HEPATOLOGY

FIG. 6. CAH with bridging necrosis in active stage. Portal zone is at left, surviving parenchyma at right. Note acidophilic necrosis (arrows) and relative paucity of inflammation.

reported in patients with NANB hepatitis who did not receive transplants (6).

Unfortunately, not all biopsy samples showed the typical features described above (Table 4); 36% of our patients had at least one biopsy specimen with atypical findings. The most common findings were centrilobular or diffuse cell swelling and parenchymal or piecemeal necrosis without significant inflammation of the portal tracts and parenchyma (Figs. 5 and 61, suggesting a cytopathic or ischemic form of injury. The swollen hepatocytes, especially when isolated to the centrilobular zone, mimicked ischemic injury (15-17) and histologically showed ballooned cells (Fig. 5), cholestasis and even congestion. The centrilobular ballooned cells were thought at first to represent superimposed ischemia, but we were never able to document arterial disease in these patients. At times, the centrilobular hepatocytes were enlarged and maintained the eosinophilic appearance of the cytoplasm (Fig. 71, but they did not have the pale appearance of the ballooned cells more typical of ischemia or severe hepatocyte damage of any cause. It is possible that these large eosinophilic cells were infected, mildly injured hepatocytes with impaired regenerative activities that became enlarged. In these cases groups of smaller than normal hepatocytes ar- ranged in double-cell plates were often prevalent in the zones adjacent to the enlarged cells; these small hepatocytes might represent regenerating hepatocytes. The precise reason for the presence of the centrilobular,

enlarged, eosinophilic hepatocytes is unclear, and we recognize that they could represent a nonspecific reaction caused by a variety of factors such as drug toxicity, ischemia or infection by HCV.

Other atypical features of hepatitis that were present in our cases included portal-based lesions such as marked ductular proliferation that mimicked obstruction (present in one of the severe CAH cases) and ductal inflammation and damage that mimicked or represented superimposed rejection (Figs. 8 and 9). The lesion that mimicked obstruction showed an unusual amount of ductal proliferation with apparent edema of the portal zone and a more prominent neutrophilic infiltrate and less prominent lymphoid infiltrate than is usually seen in hepatitis.

The lesions that had concurrent rejection or that mimicked rejection were usually a diagnostic problem earlier in the course of the hepatitis (in the first 4 mo) in patients who had experienced previous episodes of rejection and in those patients who had more active or more severe hepatitis. Rejection alone was typically observed as a portal inflammatory infiltrate composed of neutrophils (predominantly around ducts or ductules), a few eosinophils and numerous lymphocytes (some in the duct wall). Ductal and ductular damage was present, but parenchymal injury was usually minimal except in more severe cases of rejection (Table 5 ) . In contrast, hepatitis predominantly showed lymphocytic infiltrates of portal tracts and parenchyma with hepatocyte necrosis and


F?G. 7. Eosinophilic enlargement of hepatocytes and scant mononuclear inflammation in centrilobular zone. Regeneration is present in the more peripheral zones.

FIG. 8. Portal tract showing significant interlobular duct damage farrow) with somewhat scant mixed inflammatory infiltrate suggestive of rejection. The patient was not treated for rejection, and the ductal lesions were not preeent in the subsequent biopsy sample taken 6 wk later. The periportal hepatocytes are swollen.

874 FERRELLETAL HEPATOLOGY

FIG. 9. (A) Explant of HCV-positive patient showing prominent ductular proliferation without sigdicant ductular damage. (B) Same patient (transplant at time of retransplantation) showing ductular proliferation but with more prominent damage.

TABLE 4. Histological features of HCV infection that created diagnostic difficulties

______ ~~

Histological feature FrequencgP

Hepatocyte swelling with minimal inflammation Centrilobular ballooning mimicking ischemia Florid ductal proliferation with acute peri-

8 (36) 3 (14) 3 (14)

cholangitis mimicking obstruction Hepatitis mixed with obstruction 1(4) Hepatitis mixed with rejection Duct or ductule damage without rejection

11 (50) 11 (50)

Patients with hepatitis caused by CMV, HBV or drug reaction were

"No. of patients in this study exhibiting feature (percentage of the excluded from the study.

whole that this no. represents).

minimal ductal damage. However, in any one biopsy specimen, a great variety of cell types in the portal inflammatory infiltrate of a hepatitis could be present, especially in more severe cases, and ductal or ductular proliferation could be prominent in both hepatitis and rejection (Table 5 and Figs. 8 and 9).

In cases where the cause of the duct or ductular damage could not be determined, follow-up biopsies after therapy for rejection helped to clarify whether the damage was caused by HCV or rejection. The result was that at some point in time many patients had mild rejection and hepatitis concurrently (50%), and the hepatitis persisted after therapy for rejection. Any

residual ductal or ductular damage subsequently seen did not result in loss of bile ducts, although ductules often showed extensive attenuation and loss of uni- formity of epithelium, mimicking the vanishing duct syndrome of rejection (18) (Fig. 9). This histological change could also be associated with a rise in serum bilirubin levels. For example, one of the patients requiring retransplantation showed a marked rise in the serum bilirubin level to 30 mg./dl in the last 6 wks of his course. During this time his biopsy specimens showed increased ductal damage, which was similar to that seen in his transplant at removal (Fig. 91, suggesting concurrent rejection. When duct damage was seen without concurrent rejection, the smaller proliferating ductules were more involved than the interlobular bile ducts. Because some of these ductules most likely represented diverticuli arising from the hepatic cords rather than primary ductal structures themselves (141, damage or loss would not cause cholestasis or a clinical syndrome of vanishing ducts as seen in chronic rejection or PBC. Thus, on the basis of this study, significant interlobular duct damage associated with a mixed portal infiltrate was considered the most important definitive parameter for determining the presence of rejection in a patient with HCV infection, although intense damage to proliferating ductules could also suggest a rejection process.

In our experience HCV infection can also mimic other forms of hepatitis caused by HBV infection, CMV infection or drugs. At our institution we have found


TABLE 6. Comparison of HCV infection with other entities HCV infection HCV infection Ductal Vaseulat. CMV HBV

Morphological feat- in normal patients" in transplant patients involvement involvement infection" infection"

Mononuclear infiltrates Portal Parenchymal

Acidophilic necrosis Disruption of limiting plate Lymphoid aggregates Hepatocyte swelling with in-

flammation Hepatocyte swelling without

inflammation Duct or ductule proliferation Piecemeal necrosis

+ + + + -1- +

+

+ +

"Data are theoretical constructs partly derived from references 1,8,9,19 and 20 and from personal observations of patients at the University of CaliforniaSan Francisco.

immunoperoxidase staining for HBcAg useful for iden- tifying HBV during biopsy, because ground-glass cells containing HBsAg often are not present and are not easily found early in recurrent HBV infection, and the remaining histological features can be similar to those of HCV infection (Table 5).

With typical hepatitis caused by CMV, obvious inclusions or more confluent zones ( > 1 cell involved) of spotty, nonzonal necrosis are usually present (19, 20). These necrotic zones are associated with an inflammatory infiltrate composed predominantly of Kupffer cells, neutrophils or both rather than lymphocytes. Occasional granulomas or microgranulomas can also be seen. We have seen some cases of CMV infection with portal and parenchymal mononuclear infiltrates associated with single-cell, spotty, acidophilic necrosis of hepatocytes, which can also be seen in HCV infection (Table 5). Immunoperoxidase staining for early CMV antigen has been reported to be helpful in cases of hepatitis possibly caused by CMV where no cytomegalic cells or inclusions could be found in biopsy samples (2 1); however, empiric therapy with CMV antibody therapy and clinical follow-up with biopsy also distinguishes HCV from CMV in problem cases.

Drug-induced hepatitis may be difficult to discern from hepatitis C, but in our experience most of the drugs routinely used for immunosuppression do not seem to cause typical hepatitis; thus we have not had much experience in this area. However, with the advent of new immunosuppressive drugs, hepatitis-like lesions (a necroinflammatory type of liver injury) may become more prevalent. Cyclosporine has been implicated in centrilobular ballooning (lo), but this reaction has not been uniformly established (22). Other drugs used for treating clinical problems such as infection (isoniazid or antibiotics) or psychological disorders (phenothiazines or antidepressants) could be more of a problem, but cessation of the drug exposure could help clarify the issue.

In summary, diagnostic problems arise in the interpretation of persistent pathological conditions in HCV-

positive patients. Routine biochemical studies do not differentiate among the various possible lesions, and one biopsy specimen may not reveal the typical pattern of injury seen in HCV infection but may have features that mimic or represent superimposed rejection, obstruction, ischemia (especially vascular rejection) or other forms of hepatitis. Serial biopsies are often needed to better define the pattern and evolution of a persistent lesion as a true hepatitis. In addition, PCR positivity for HCV adds another level of certainty as to the cause of the hepatitis, which may be especially helpful when deciding whether interferon therapy is needed.

Acknowledgment: We thank Mr. David Geller for his editorial assistance on this manuscript.

REFERENCES 1. Read A, Donegan E, Lake J , Ferrell L, Galbraith C, Kuramoto IK,

Zeldis JB. Hepatitis C in patients undergoing liver transplantation. AM Intern Med 1991;114:282-284.

2. Alter M, Gerty R, Smallwood L, Sampliner RE, Tabor E, Deinhardt F, Frosner G. Sporadic non-A, non-B hepatitis: fre- quency and epidemiology in an urban US. population. J Infect Dis

3. Knodell R, Conrad M, Dienstag J, Bell C. Etiological spectrum of post-transfusion hepatitis. Gastroenterology 1975;69: 1278- 1285.

4. Cristiano K, Di Bisceglie A, Hoofnagle J , Feinstone S. Hepatitis C viral RNA in serum of patients with chronic non-A, non-B hepatitis: detection by polymerase chain reaction using multiple primer sets. HEPATOLOGY 1991;14:51-55.

5. Wright T, Donegan M, Hsu H, Ferrell L, Lake JR, Read A, Kim M, et al. Hepatitis C viral infection in liver transplant recipients: the importance of detection of viral RNA using the polymerase chain reaction. Gastroenterology 1992;103:317-322.

6. Lefkowitch J, Apfelbaum T. Non-A, non-B hepatitis: character- ization of liver biopsy pathology. J Clin Gastroenterol 1989;ll:

7. Scheuer PJ. Non-A, non-B hepatitis. Virchows Arch A Pathol Anat Histopathol 1989;415:301-303.

8. Demetris A, Jaffe R, Sheahan D, Burham J , Spero J, Iwatsuki S, VanThiel DH. Recurrent hepatitis B in liver allograft recipients: differentiation between viral hepatitis B and rejection. Am J Path01 1986;135:161-172.

9. Demetris A, Todo S, Thiel D, Fung JJ, Iwaki Y, Sysyn G, Ming W. Evolution of hepatitis B virus liver disease after hepatic re-

1982; 145:886-893.

225-232.


placement: practical and theoretical considerations. Am J Pathol

10. Ray R, Lewin K, Colonna J, Goldstein L, Busuttil R. The role of liver biopsy in evaluating acute allograft dysfunction following liver transplantation: a clinical histologic correlation of 34 liver transplants. Hum Path01 1988;19:835-848.

11. Di Bisceglie A, Martin P, Kassianides C, et al. Recombinant interferon alpha therapy for chronic hepatitis C. A randomized, double-blind, placebo-controlled trial. N Engl J Med 1989;321:

12. Hoofnagle J , Di Bisceglie A. Treatment of chronic type C hepatitis with alpha interferon. Semin Liver Dis 1989;9:259-263.

13. Renault P, Hoofnagle J. Side effects of alpha interferon. Semin Liver Dis 1989;9:273-277.

14. Yamada S, Howe S, Scheuer PJ. Three-dimensional recon- struction of biliary pathways in primary biliary cirrhosis: a computer-assisted study. J Path01 1987;152:317-23.

15. Goldstein N, Hart J , Lewin K. Diffuse hepatocyte ballooning in liver biopsies from orthotopic liver transplant patients. Histopa-

16. Ng I, Burroughs A, Fblles K, Belli L, Scheuer P. Hepatocellular ballooning after liver transplantation: a light and electronmicro-

1990;137:667-676.

1506-1510.

tholorn 1991;18:331-338.

scopic study with clinicopathological correlation. Histopathology

17. Snover D. Problems in the interpretation of liver biopsies after liver transplantation. Am J Surg Path01 1989;13(Suppl 1):31-38.

18. Ludwig J, Wiesner R, Batts K, Perkins J , Krom R. The acute vanishing bile duct syndrome (acute irreversible rejection) after orthotopic liver transplantation. HEPATOLOGY 1987;7:476-483.

19. Paya CV, Hermans PE, Wiesner RH, Ludwig J , Smith TF, W a J, Krom RA. Cytomegalovirus hepatitis in liver transplantation: prospective analysis of 93 consecutive orthotopic liver transplantations. J Infect Dis 1989;160:752-8.

20. Snover DC, Hutton S, Balfour HH Jr , Bloomer JR. Cytomegalo- virus infection of the liver transplant recipients. J Clin Gastroen- terol 1987;9:659-65.

21. Paya CV, Holley KE, Wiesner RH, Balasubramaniam K, Smith TF, Espy MJ, Ludwig J , et al. Early diagnosis of cytomegalovirus hepatitis in liver transplant recipients: role of immunostaining, DNA hybridization, and culture of hepatic tissue. HEPATOLOGY

22. Williams JW, Vera S, Peters TG, VanHoorst S, Britt LG, Dean PJ, Haggitt R, et al. Cholestatic jaundice after hepatic transplantation. Am J Surg 1986;151:65-70.

199 1; 18:323-330.

1990;12:119-26.

BOUND VOLUMES AVAILABLE TO SUBSCRIBERS Bound volumes of HEPATOLOGY for 1992 are available to subscribers only. They may

be purchased from the publisher at a cost of $74.00 for domestic, $102.18 for Canadian and $97.00 international for Vol. 15 (January to June) and Vol. 16 (July to December). Price includes shipping charges. Each bound volume contains a subject and author index, and all advertising is removed. Copies are shipped within 60 days after publication of the last issue of the volume. The binding is durable buckram with the journal name, volume number, and year stamped in gold on the spine. Payment must accompany all orders. Contact Mosby-Year Book, Inc., Subscription Services, 11830 Westline Industrial Drive, St. Louis, MO 63146-3318, USA; phone (800)325-4177, ext. 4351, or (314)453-4351.

Subscriptions must be in force to qualify. Bound volumes are not available in place of a regular HEPATOLOGY subscription.

I

hepatitis c viral infection in liver transplant recipients

Documents