AIM: To determine the most frequent etiologies of hepatic epithelioid granulomas, and whether there was an association with chronic hepatitis C virus (HCV).

METHODS: Both a retrospective review of the pathology database of liver biopsies at our institution from 1996 through 2006 as well as data from a prospective study of hepatic fibrosis markers and liver biopsies from 2003 to 2006 were reviewed to identify cases of hepatic epithelioid granulomas. Appropriate charts, liver biopsy slides, and laboratory data were reviewed to determine all possible associations. The diagnosis of HCV was based on a positive HCV RNA.

RESULTS: There were 4578 liver biopsies and 36 (0.79%) had at least one epithelioid granuloma. HCV was the most common association. Fourteen patients had HCV, and in nine, there were no concurrent conditions known to be associated with hepatic granulomas. Prior interferon therapy and crystalloid substances from illicit intravenous injections did not account for the finding. There were hepatic epithelioid granulomas in 3 of 241 patients (1.24%) with known chronic HCV enrolled in the prospective study of hepatic fibrosis markers.

CONCLUSION: Although uncommon, hepatic granulomas may be part of the histological spectrum of chronic HCV. When epithelioid granulomas are found on the liver biopsy of someone with HCV, other clinically appropriate studies should be done, but if nothing else is found, the clinician can be comfortable with an HCV association.

Open in New Tab   Full Size Figure   Download Figure

Figure 1 Large granuloma in the hepatic lobule noted in an asymptomatic patient with chronic HCV and mildly elevated ALT (HE, × 40).

Epithelioid granulomas are found in 1%-10% of liver biopsy specimens[1-3]. They tend to fall into three broad categories which are systemic granulomatous disease, primary liver disease, or miscellaneous conditions that fit neither category. Hepatic granulomas are sometimes a surprise finding that may trigger an extensive search for an etiology. In some cases the granulomas are associated with a clinical picture which includes a high alkaline phosphatase and hepatomegaly; but these features are frequently absent. A significant percentage of liver biopsies performed today are for the staging of chronic hepatitis C. Granulomas have been reported with possible increased frequency in patients with hepatitis C with both mild disease as well as in explants[4-8]. It has been postulated that hepatitis C virus (HCV) may have a role in granuloma formation[9]. We decided to undertake a retrospective review of cases of hepatic granulomas at our institution during the last decade in order to determine the most frequent etiologies, and also to determine if there was an association with HCV.

There were a total of 4578 liver biopsies performed at our institution during this time. There were 36 (0.79%) patients identified from the surgical pathology review that had at least one hepatic epithelioid granuloma (Table 1). Special stains were performed in 26 f and tissue cultures in five. The associated diagnoses are listed in Table 1. The most common association was HCV. Fourteen (36.1%) had HCV, and nine of these had no other clinical associations to explain the granulomas. Five of the HCV patients had other confounding associations including chronic hepatitis B, sarcoidosis, histoplasmosis, and HIV (2 patients). The etiologies with more than one case in the non HCV patients were sarcoidosis (3), histoplasmosis (4), primary biliary cirrhosis (3), HIV only (2), mycobacterium avis complex (2), and primary biliary cirrhosis (2), and unidentifiable (2).

There were 241 patients with chronic HCV that had liver biopsies while enrolled in the prospective hepatic fibrosis study during 2003-2006. Three (1.24%) had granulomas, and all of these had been identified in the above database search.

We found granulomas in less than 1% of our liver biopsies. Other reports have found granulomas in up to 10% of liver biopsies[1-3]. The frequency of detection as well as the frequency of various diagnoses can depend upon the geographic location of the center and the diseases endemic in that area as well as the diligence of the histopathology laboratory and the pathologists that prepare and examine the slides. Previous studies of hepatic granulomas have primarily been from eras before HIV and widespread organ transplantation, and before HCV became prominent. Although the percentages varied, sarcoidosis and tuberculosis tended to be most common associated diseases reported. A study from our institution from another era found that tuberculosis accounted for 53% and sarcoidosis 12% of the cases of hepatic epithelioid granulomas while in 20% no etiology could be determined[1]. In a large series of 565 cases of epithelioid hepatic granulomas in over 6000 biopsies accumulated over 4 decades, Klatskin[2] found that 36% had sarcoidosis, 12% had tuberculosis and 7% were undiagnosed. While our study indeed found 3 patients with sarcoidosis and another with M. tuberculosis, most cases were associated with HCV or HIV with associated infections.

There have been several reports of granulomas in the liver biopsies of patients with HCV and no other known systemic or hepatic diseases. This may not be unexpected since the most common reason for liver biopsies at most institutions today is staging of chronic HCV, and some unexplained granulomas have long been found on liver biopsies. On the other hand, the concurrence of HCV and granulomas may be frequent enough to not be simply explained by chance. A large recent retrospective review found epithelioid granulomas in 63 of 1662 liver biopsies[5]. While primary biliary cirrhosis (23.8%), sarcoidosis (11.1%), and unknown (11.1%) were the most common associations, HCV was associated with 9.5% of the cases. Emile et al[9] found that the explants of 5 of 52 patients undergoing liver transplantation for cirrhosis from HCV had epithelioid granulomas, and none of these patients had evidence of any other diseases such as tuberculosis or sarcoidosis. Goldin et al[4] found in a retrospective study that there were granulomas in the liver biopsy specimens of 14/155 patients with HCV compared with only 3/151 with HBV and 3/129 with alcoholic liver disease. Yamamoto et al[6] found unexplained granulomas in 2/273 (0.73%) liver biopsies of patients with chronic HCV. In a more recent study from Turkey, 8/605 (1.3%) of patients with chronic hepatitis C had unexplained granulomas on liver biopsy[7]. This is the same percentage that we found among chronic HCV patients in our prospective study of hepatic fibrosis markers that underwent liver biopsies over a three year period. Consequently, it has been postulated that hepatic granulomas may themselves be part of the histologic picture of chronic HCV, albeit uncommonly. Our finding that 9 of 36 patients with hepatic epithelioid granulomas had HCV as their only disease association would support this. We cannot rule out that some of these patients could have had other undiagnosed disorders such as sarcoidosis; but in follow up, none of them have manifested pulmonary or systemic findings to support another diagnosis.

Alpha interferon in regular or pegylated form has been used in the therapy of chronic HCV since the discovery of the virus[13]. It has been speculated that interferon itself can stimulate granuloma production, and there are several reported cases of the development of sarcoidosis in patients receiving interferon therapy[15-19]. This is thought to be related to the ability of interferon to stimulate a Th 1 immune response which is felt to be responsible for the granulomas in sarcoidosis[20]. Both the sarcoidosis and granulomas have usually regressed when interferon was stopped. It is unlikely that interferon usage played a major factor in our patients since only one subject with chronic HCV and granulomas had received interferon prior to liver biopsy.

Intravenous drug use is the most common route of transmission of chronic hepatitis C[21]. Since granulomas can be induced by foreign substances that might be used in a “street prepared” illicit drug mixture, we examined all the slides of patients with HCV and granulomas with polarized light, and a crystalloid substance was found in only one patient. Therefore, foreign body induced granulomas do not appear to be an explanation for most of the granulomas found in our patients with HCV.

We realize that there are problems interpreting the results of a retrospective study such as ours. The majority of the liver biopsies that we perform are on patients that are being staged prior to proposed treatment for chronic HCV. Therefore, the finding of otherwise unexplained granulomas in multiple patients with chronic HCV may not be that unusual or surprising. Nevertheless, in this study the most common association of hepatic granulomas was chronic HCV.

We would recommend if a patient with chronic HCV should have granulomas on liver biopsy that a search for another disease should be made by utilizing special stains and other clinically appropriate tests such as a chest X-ray. However, if no other abnormality is found, the clinician should be comfortable associating the granuloma(s) with the chronic HCV.