Secondary Logo

Journal Logo

Liver Biopsy Findings in 501 Patients Infected with Human Immunodeficiency Virus (HIV)

Poles, Michael A.; Dieterich, Douglas T.; Schwarz, Elliot D.; Weinshel, Elizabeth H.; Lew, Edward A.; Lew, Ron; Scholes, John V.

Journal of Acquired Immune Deficiency Syndromes and Human Retrovirology: February 1st, 1996 - Volume 11 - Issue 2 - p 170-177
Clinical Science
Free

Summary: Patients infected with human immunodeficiency virus (HIV) are at risk for a variety of liver diseases. We undertook a retrospective study of 501 HIV-seropositive patients to assess the yield of percutaneous liver biopsy. The most common indications for liver biopsy were liver test abnormalities (89.5%), fever for 2 weeks (71.9%), and hepatomegaly (52.0%). The most common biopsy-derived diagnosis was Mycobacterium avium complex (MAC), seen in 87 (17.4%) biopsies. Mycobacterium tuberculosis was found in 13 biopsies (2.6%). In 28 biopsies (5.6%) mycobacteria was seen, but speciation of the organism was not possible. Chronic active viral hepatitis was seen in 60 biopsies (12.0%). Opportunistic hepatic infection from other organisms was found in 14 biopsies (2.8%). The most common neoplasm was lymphoma, which was seen in 12 biopsies (2.4%). MAC infection of the liver was associated with elevated alkaline phosphatase (p = 0.01). Among patients with fever for 2 weeks after an extensive negative workup including bone marrow biopsy, 58.2% had a diagnosis by liver biopsy. Overall, 64.3% of liver biopsies yielded a histopathological diagnosis, 45.7% of which were potentially treatable. We could not evaluate whether liver biopsy had a positive effect on patient outcome and survival, nor did we attempt to prove that liver biopsy resulted in a change in treatment or a change in preprocedure clinical diagnosis. Thus, questions about the efficacy of liver biopsy cannot be answered. Liver biopsy may be a helpful diagnostic tool in HIV-positive patients with fever, liver test abnormalities, or hepatomegaly.

New York University Medical Center, New York, New York, U.S.A.

Address correspondence and reprint requests to Dr. Douglas T. Dieterich, New York University Medical Center, 345 E. 37th St. Suite 207, New York, NY 10016, U.S.A.

Manuscript received December 6, 1994; accepted October 9, 1995.

Liver disease is a frequent cause of morbidity and mortality in patients with AIDS. The liver plays a vital role in patients infected with HIV, serving as a reservoir of prior infection with hepatitis viruses and opportunistic infections and possibly as a site of human immunodeficiency virus (HIV) replication. The liver receives a large percentage of the cardiac output and is thus a window to the diagnosis of disseminated, blood-borne opportunistic infections. Clinicians are often required to evaluate patients with HIV disease who have fever, hepatomegaly, or abnormalities of their liver tests and require a systematic method of evaluating these problems. Clinicians can treat lymphoma, Mycobacterium avium complex (MAC), cytomegalovirus (CMV), and other disorders; therefore, it is important to find a sensitive and specific means to diagnose the liver disease in patients with AIDS. The usefulness of percutaneous liver biopsy in making these diagnoses is controversial. Because the liver is commonly involved in systemic disease that has disseminated lymphohematogenously, a diagnosis may often be established through less invasive means. Yet, despite an extensive evaluation, many patients with disseminated infection remain undiagnosed. Nearly all liver biopsies from patients with AIDS are found to have some abnormal histological features, many due to opportunistic infections and neoplasms. We undertook a retrospective review of 501 liver biopsies to evaluate the yield of liver biopsy in patients with HIV.

Back to Top | Article Outline

METHODS

From December 1984 to May 1992, 501 patients at Tisch Hospital (41%), Bellevue Hospital Medical Center (34%), and the Department of Veteran's Affairs Medical Center, New York (25%) who were seropositive for HIV underwent percutaneous liver biopsy. These patients represent all known patients with HIV who underwent percutaneous liver biopsy at these hospitals. The histology and patient data were systematically examined retrospectively by one of the authors (M.A.P.). Autopsy data were not included in the analysis, nor were biopsies performed during laparotomy.

Each patient's clinical record was assiduously reviewed for demographic information and risk factors for contracting HIV. Liver tests included alanine transaminase (ALT), aspartate transaminase (AST), gamma glutamyl transpeptidase (GGTP), alkaline phosphatase, and total bilirubin level. Laboratory values used for analysis were those prior to and closest to the day of biopsy. Hepatitis virus serologies, including those for hepatitis B virus (HBV) and hepatitis C virus (HCV), were recorded when available. Also noted were indications for biopsy and concurrent diagnostic workup. Histologic results were determined both through the clinical record and by examination of the final reports in the departments of pathology. Confirmation of culture results obtained from the chart was made by examination of records in the microbiology laboratories of each institution. The information abstracted from the patient charts did not include preoperative diagnosis or postbiopsy treatment changes, as these data were largely incomplete.

All percutaneous biopsies were performed using standard techniques by members of the department of gastroenterology in the three hospitals. Most biopsies were done with a 17-gauge Jamshidi needle. Unfixed tissue was cultured for mycobacterial, fungal, and viral pathogens. Specimens were also fixed in 10% formalin. The samples were stained with hematoxylin and eosin, trichrome, periodic acid-Schiff, Gomori-methamine silver, and Ziehl-Neelson stains. When available, Orcein stain for hepatitis B surface antigen (HBsAg) was also performed.

Fever was defined as body temperature above 38°C more than twice in a week for a total of 2 weeks despite a workup that failed to yield a diagnosis. Although not established protocol in patients with AIDS, patients with fever underwent a bone marrow biospy before liver biopsy to rule out acid-fast bacilli in addition to a noninvasive workup to determine the cause of fever, including culture of all body fluids that were clinically indicated as well as a chest radiograph. Febrile patients whose bone marrow biopsy failed to reveal a diagnosis underwent percutaneous liver biopsy. Abnormalities of liver tests were defined as an elevation of either transaminases (ALT, AST), alkaline phosphatase, or total bilirubin greater than ×1.5 the normal values for the hospital from where the patient was enrolled. These abnormalities were present on repeat testing over a 1-month period before biopsy was undertaken. As no significant difference in values were noted during the 1-month period, the values chosen for analysis were those closest to biopsy. Hepatomegaly was defined clinically by the physician under whose care the patient was hospitalized.

After dividing patients into groups based on diagnosis, the laboratory data of each group were compared to a reference group comprising HIV-seropositive patients with normal or nonspecific histologic findings. These values were statistically compared using a two-tailed paired student's t test. A p value <0.05 was considered statistically significant.

Back to Top | Article Outline

RESULTS

Of the 501 patients in our study, 91.2% (457 of 501) were males, and 8.8% (44 of 501) were females. The median age of the patients studied was 38 years (range, 0.5-67). Of the 454 patients for whom race could be ascertained, 193 (42.5%) were white, 140 (30.8%) were Hispanic, 120 (26.4%) were black, and one (0.2%) was Asian. Data on risk factor for seroconversion were available for 446 patients; 212 were homosexual (47.5%), 179 were i.v. drug users (40.1%), 26 were heterosexual with a known seropositive partner (5.8%), and 17 patients were both homosexual and i.v. drug users (3.7%). Six patients (1.3%) had transfusion-associated seroconversion, and six (1.3%) patients were children who acquired the disease through vertical transmission. The median duration of the diagnosis of HIV seropositivity before performance of the biopsy was 12.0 months (range, 0-93). The median CD4 lymphocyte count, available for 349 patients, was 150.5 cells/mm3. Sixty-two percent of these patients (216 of 349) had CD4 cell counts <100 cells/mm3. Information regarding past opportunistic infections was available in 372 patients. Of these patients, 147 (39.5%) of these patients reported previous infections with Pneumocystis carinii, 66 (17.7%) reported past episodes of Kaposi's sarcoma, 56 (15.1%) reported candidal esophagitis, 27 (7.3%) reported CMV infection, 16 (4.3%) reported MAC infection, and seven (1.9%) reported extrapulmonary Mycobacterium tuberculosis (MTB). Of these 372 patients, 86 (23%) were without prior opportunistic infection. There did not appear to be a distinct correlation between the presence of a particular past opportunistic infection and finding on liver biopsy.

Of 501 biopsies reviewed, 392 had clear indications outlined in the patient records. Of 392 liver biopsies, 351 were performed to evaluate abnormalities in liver chemistries (89.5%), 282 to evaluate fever (71.9%), and 243 (62.0%) for both abnormalities in liver chemistries and fever. On physical exam or diagnostic imaging, 204 (52%) had hepatomegaly, and 41 (10.5%) had right upper quadrant pain. Abdominal computed tomography (CT) scan or ultrasound led to discovery of a mass or nodule in 31 patients (7.9%) who were biopsied under CT guidance. No clear indication for biopsy was documented in 109 patient records.

The diagnostic findings made by liver biopsy are shown in Table 1. Granulomatous hepatitis was found in 191 patients. Most (n = 87, or 17.4%) of the entire study population were found by culture of liver tissue to be due to MAC. MTB, seen in 14 biopsies (2.8%), was a less common finding. In 28 patients (5.6%), mycobacteria were noted on Ziehl-Neelson stain, but speciation of the organism could not be made by other means. The biopsies of 62 patients (12.4%) showed granulomatous changes, but no acid-fast bacilli or other etiologic factors were discovered. The granulomata of these 62 biopsies were typically small, non-necrotizing, and poorly formed, which many consider characteristic of MAC. In our series, patients with these findings were often treated with anti-MAC therapy; the results of such therapy are unknown.

Other organisms noted on biopsy included CMV in five patients, Pneumocystis carinii in two patients, Candida albicans in two, schistosoma in two, Histoplasma capsulatum in one, adenovirus in one, and salmonella in one patient. The most common neoplasm found on liver biopsy was large cell lymphoma, seen in 12 patients (2.4%). Other tumors included hepatoma in three patients, Kaposi's sarcoma in two, metastatic melanoma in one and metastatic adenocarcinoma in one. Although the lymphomas, hepatomas, and metastatic adenocarcinoma were radiographically identified as masses, the other neoplasms were not apparent on imaging.

Chart review of 372 patients yielded data on hepatitis virus serologies; of these, 267 had evidence of past or present hepatitis B, and 25 patients had serologic evidence of hepatitis C. Of these 267 patients, 144 (50.2%) showed serologic evidence of past exposure to hepatitis B, and 31 patients (11.6%) were surface antigen positive, indicating a chronic carrier state. Twenty-five of 69 patients tested (36.2%) had serologic evidence of hepatitis C. Most of our liver biopsies were performed before 1990, when testing for HCV became available. Of the 190 patients with serologic evidence of past or present viral hepatitis, only 60 (12.0% of total) were diagnosed as having chronic active viral hepatitis on liver biopsy. Of these, eight (13.3%) were hepatitis B surface antibody positive, 14 (23.3%) were HBsAg positive, and 17 (28.3%) were hepatitis C positive. In 21 patients with chronic active hepatitis diagnosed by biopsy, serologies were not available (20.0%). Thirty-one biopsies (6.2%) showed the liver to be cirrhotic; 10 patients had evidence of prior hepatitis B infection, six (19.4%) had ongoing hepatitis B infection, and two (6.5%) had evidence of hepatitis C. Eleven (35.5%) of these patients had no serologies available, and three (9.7%) had negative serologies for hepatitis B and C. Eighty-nine of 367 patients (24.3%) admitted to a history of heavy alcohol use, and 15 (3.0%) had histologic evidence of acute steatohepatitis, indicating damage by alcohol.

Nonspecific liver biopsy findings included steatosis, Kupffer cell hyperplasia, hemosiderosis, sinusoidal dilatation, nonspecific reactive hepatitis, and cholestasis. These were the only findings in 165 biopsies (32.9%), and 14 biopsies (2.8%) were completely normal. Thus, 322 (64.3%) of the liver biopsies in our series yielded a histopathological diagnosis. If granulomatous hepatitis without organisms is considered nondiagnostic, the yield from liver biopsies among the patients studied falls to 48.1%. In each instance, biopsy of the liver yielded a diagnosis that was not previously recognized at another site. Five biopsies revealed more than one diagnosis.

Stratification of the biopsy results according to CD4 cell count was performed (Table 2) but failed to reveal a large difference in the yield of biopsy between the strata. Biopsies in patients with a CD4 cell count of 0 to 50/mm3 yielded a diagnosis in 69.4% (65.3% excluding granulomatous hepatitis without organisms); in patients with a CD4 cell count between 50 and 100/mm3, 66.9% (50.8%) were diagnostic; in patients with a CD4 cell count of 100 to 200/mm3 biopsies yielded a diagnosis in 46% (34%); in patients with CD4 cell count 200 to 300/mm3 biopsies yielded a diagnosis in 70% (45%); and in patients with >300 CD4 cells/mm3, the yield was 61.5% (46.2%). As expected, MAC was diagnosed almost exclusively in patients with <100 CD4 cells. Additionally, 13 of 18 biopsies that were diagnostic of granulomas with unspeciable mycobacteria were also found in patients with <100 CD4 cells, but only 23 of 42 (54.8%) biopsies with granulomas with organisms were from patients with such severe immunosuppression. Other infections were also exclusively found in patients with severe immunosuppression.

There did not appear to be a noticeable difference in yield from liver biopsies when the data were grouped according to indication for biopsy. Among the 351 patients who had biopsies to evaluate abnormalities in liver chemistries, 218 yielded a histopathological diagnosis (62.1%). Of the 282 patients who underwent biopsy to evaluate fever, 183 (64.9%) yielded a histopathologic diagnosis; of the 243 who had biopsies for both indications, 156 (64.2%) were diagnostic. Of patients undergoing biopsy, 219 did not meet the criteria for fever, and 138 of these (63%) yielded a diagnosis, 13 of which were mycobacterial (MAC eight cases; unspeciated mycobacteria, four cases; and MTB, one case). Despite a lack of clear-cut indication for biopsy of 109 patients, 66 (60.6%) yielded a histopathologic diagnosis. Data analysis did not yield significant differences when these patients were segregated out.

Information regarding laboratory analysis is listed in Table 4. The median ALT of patients with viral hepatitis (192.0 mg/dl) was significantly elevated compared with the reference group (54 mg/dl, p = 0.037). Comparison of AST and GGTP yielded no significant differences. Alkaline phosphatase levels in patients with MAC (median, 436.0 mg/dl) were significantly elevated (p = 0.01) compared with the reference group (median = 322.5 mg/dl). Because of the high number of false-positive results, the positive predictive value of an alkaline phosphatase >200 IU/L was only 42% for diagnosis of MAC and 58% for any diagnosis.

Back to Top | Article Outline

DISCUSSION

The HIV-seropositive population is increasing rapidly. In one study 90% of HIV-seropositive i.v. drug users had evidence of liver disease when they first presented with illness (1). This is a problem all clinicians must face. This study of 501 HIV-seropositive patients shows that percutaneous liver biopsy may be a helpful adjunct to the diagnosis of liver disease in these patients.

In our opinion, diagnostic biopsies include those with granulomatous hepatitis, other infections, viral and alcoholic hepatitis, cirrhosis, and neoplasia. Sixty-four percent of all percutaneous liver biopsies revealed such findings. If granulomatous hepatitis is considered nondiagnostic because of its nonspecific nature, 48.1% of biopsies may be considered diagnostic. We believe that viral hepatitis, MAC, MTB, other opportunistic hepatic infections, and neoplasms represent potentially treatable conditions; 229 or 501 (45.7%) of our biopsies yielded a potentially treatable condition. Further, 58.2% of patients with fever who had already undergone nondiagnostic bone marrow biopsy were diagnosed by liver biopsy. Given that the cause of fever, liver test abnormality, and hepatomegaly were unknown in all patients before they underwent biopsy, this procedure was of value in establishing diagnosis in a large number of patients. It is standard practice in our medical center to defer liver biopsy when patients are diagnosed with an entity that may be responsible for fever, elevated liver chemistries, or hepatomegaly. Thus, each biopsy that was undertaken was diagnostic of an abnormality that was not previously recognized at another site. We were interested to note that only 1.5% of biopsies yielded more than one diagnosis. Such singularity of disease is atypical in patients with HIV, but we are not able to offer an explanation for this finding.

Liver biopsy is a controversial procedure in the evaluation of liver disease in patients with AIDS. Because the liver is commonly involved in systemic disease that has disseminated lymphohematogenously, a diagnosis may often be established through less invasive means. These include biopsy and culture of peripheral sites, such as the blood, skin, gastrointestinal mucosa, lymph nodes, and bone marrow. Many clinicians argue against performance of liver biopsy. Lebovics et al. failed to find a correlation between indications for liver biopsy and histopathological findings (2). Schneiderman et al., in a study of 85 patients with AIDS who underwent liver biopsy (26 biopsy, 59 autopsy), reported that liver biopsy rarely influenced therapy or led to improved survival (3). They found AIDS-related opportunistic infections and neoplasms in 42% of patients, but in most of these cases, the hepatic involvement was part of a disseminated process that could have been diagnosed by other means. The impressions gained from our study appear to contradict their conclusions. Before undergoing liver biopsy for fever, as in most cases, patients underwent an extensive workup, which included bone marrow biopsy and culture of body fluids for bacteria, fungi, and acid-fast organisms. Without liver biopsy, many patients in this study would have gone undiagnosed. The conclusions from the study by Schneiderman were based on the assumption that we lack effective treatment for CMV and MAC, which were found in 38.5% of their biopsies. Given the improved armamentarium for treatment of MAC and the success of ganciclovir and foscarnet in CMV disease, such therapeutic nihilism is probably not warranted. Proponents for liver biopsy argue that it may reveal a specific diagnosis in 30 to 80% of patients with AIDS and that the finding of opportunistic infections or neoplasms on liver biopsy may also yield important prognostic information (2,3,4-9). An elevated alkaline phosphate above 300 IU/L at the time of biopsy improves yield, with a sensitivity and specificity of 75% as a screening test for diagnostic liver biopsy (7). Thus Cappel et al. suggest that biopsy be performed only in patients with fever and an alkaline phosphatase level that is greater than twice normal. As our ability to treat opportunistic infections of the liver increases, so too will the necessity and importance of liver biopsy in the diagnosis of patients with AIDS and liver disease.

As reported by others, the most common histologic finding in our study was granulomatous hepatitis (2,4,5). The granulomas were small, histiocyte-predominant, and poorly formed. Almost half of the cases of granulomatous hepatitis were associated with MAC infection, the most common intrahepatic AIDS-specific diagnosis found on liver biopsy (3). It has been described in 10 to 55% of autopsies in patients with AIDS and on liver biopsy in 10 to 30% of patients in other studies (2,4,5,10,11). This organism is commonly found on biopsy or culture of more peripheral sites. Bonacini concluded that the highest yield for diagnosis of MAC is blood cultures, followed by bone marrow biopsy and culture, and lastly liver biopsy and culture (12). Minguez et al. report that bone marrow culture is more sensitive in the diagnosis of disseminated mycobacteriosis compared with liver biopsy culture (yield 74 vs 21%) (13). We found that bone marrow biopsy may fail to diagnose MAC infection in a significant proportion of patients. After biopsy and culture of blood and bone marrow, 17% of our patients with fever were found to have hepatic MAC infection on liver biopsy. Other studies have concluded that liver biopsy is more sensitive than bone marrow or lymph node aspirate for detecting mycobacterial infection, revealing acid-fast bacilli in a higher percentage of patients than bone marrow biopsy (75 vs 25%) (7,14). Histology is more sensitive than culture in hepatic MAC infection and is also a more rapid means of diagnosis, which may be made in a few hours or up to 3 days, compared with 4 to 6 weeks for culture positivity. Nonetheless biopsy samples do require culture to differentiate MAC from other mycobacterial species and to establish the antibiotic sensitivity profile. Species identification time can be improved by sequencing amplified DNA from liver biopsies (15). Although granulomas without organisms do not represent a clinically useful diagnosis, they can often explain elevations of liver tests and unidentified causes of fever.

The predominance of MAC versus MTB cases is striking in light of the resurgence of MTB in the large indigent population served by Bellevue Hospital and the Department of Veteran's Affairs Medial Centers. Patients with AIDS are at high risk of developing tuberculosis, usually from reactivation of a latent infection. One study reported that extrapulmonary MTB is seen in 60% of patients with lung disease, 7.5% with hepatic infection (16); this was not obvious from our findings. Many of the liver biopsies with organism-negative granulomatous hepatitis were considered to be due to MAC and were treated as such. Information about the outcome of this treatment is unavailable and requires further analysis. It is possible that other infectious agents or drug-induced hepatitis may have been responsible for the changes.

A high coincidence of viral hepatitis and HIV is expected given their shared epidemiologic risks. Most patients in our series underwent liver biopsy before testing for hepatitis C became available, but among patients for whom information was available, exposure to hepatitis C was seen in only 36.2% of patients, which is much lower than expected from this population. The incidence of hepatitis C among HIV-positive users of i.v. drugs is estimated to be 50 to 90%, far greater than seen in this study (17). Evidence of serologic exposure to HBV was seen in only 50.2%. Markers of prior hepatitis B infection have been found in up to 95% of patients with AIDS, and about 10% are HBsAg positive (2,4). The true incidence of viral hepatitis diagnosed by biopsy may have been greatly underestimated because most patients with liver abnormalities and positive hepatitis virus serologies would not have had biopsies. These patients would have been given a diagnosis of viral hepatitis, if it were clinically appropriate, and biopsy would probably not have been performed. Liver biopsy is not clinically useful in the diagnosis of viral hepatitis, which can be diagnosed serologically. Liver biopsy, on the other hand, can give useful prognostic information and can allow for the quantification of HCV RNA in liver tissue.

HIV can significantly alter the development, course, and treatment of chronic hepatitis B and C. Cirrhosis was only seen in 6.2% of patients, which is surprising given the high incidence of exposure to hepatotropic viruses. Patients with HIV appear to develop less severe liver disease during chronic HBV infection, possibly because of the attenuated immune response or an increased T suppressor cell activity in response to viral hepatitis in patients with AIDS. Conversely, as the number of T cells decreases, the concentration of HBV viral particles in the serum increases, which might result in an increased propensity to liver damage (18). Some studies show that once exposed to hepatitis B, HIV-positive persons are more likely to develop chronic infection and subsequent chronic active hepatitis and cirrhosis (19,20). Immunologic changes associated with AIDS may also permit increased replication of hepatitis C. Eyster et al. found that in HIV and HCV coinfected hemophiliacs, the HCV load increases over time and is enhanced by HIV and that HCV RNA may increase as immune deficiency progresses. The higher levels of HCV viremia found in these patients was associated with an increased incidence of hepatic failure (21,22). Other studies refute a potentiating effect of HIV on hepatitis C (23,24). New assays under study can quantify the amount of hepatitis C in blood and liver tissue from biopsy and may provide further information in this regard. Discovery and treatment of hepatitis B and C with interferon may prevent the progression to liver failure and cirrhosis. Hepatitis B may have a role in increasing replication of HIV in vitro, and possibly in vivo; thus, treatment may also affect the natural history of HIV disease (25,26). Still, in one study, survival was not decreased in patients with AIDS and persistent hepatitis B infection, although it appeared that patients with more severe immunosuppression were less likely to clear HBV infection after exposure and more likely to reactivate latent HBV infection (27).

The low incidence of CMV hepatitis in this population with severe immunosuppression is also surprising. Liver involvement of CMV found at autopsy has been reported elsewhere to occur in 33 to 44% of AIDS patients with CMV (28,29), Schneiderman et al. found that 7.7% of biopsies were diagnosed with CMV by liver biopsy (3). Whereas treatment of CMV hepatitis has not been formally studied, both ganciclovir and foscarnet have beneficial effects on CMV infection of other organ systems and can prolong life in these patients (30). The cancers most commonly found in the livers of AIDS patients are non-Hodgkin's lymphoma and Kaposi's Sarcoma (KS). Up to one third of patients with cutaneous KS will have liver involvement, which usually remains unidentified until autopsy (31). The gastrointestinal tract and liver are the most common extranodal sites of involvement by lymphoma in patients with AIDS. The greater incidence of lymphoma compared with KS in this study may have been due to the improved ability to detect this cancer radiographically rather than KS resulting in directed biopsy.

A total of 32.9% of biopsies showed nonspecific findings, in many cases perhaps because of yet unidentified organisms or HIV itself. HIV can directly infect hepatocytes and may have an indirect effect on the liver through associated T-cell abnormalities (32). Hepatic infection with HIV may be via the CD4 receptor, which has been discovered on Kupffer cells, and endothelial sinusoid cells (33). Other possible causes of nonspecific liver pathology include malnutrition with or without the use of total parenteral nutrition and exposure to hepatotoxic drugs. Up to 90% of patients with AIDS are treated with a least one drug associated with hepatotoxicity, which is a likely contributor to hepatic abnormalities (1). Macrovesicular steatosis, the most common nonspecific finding, is likely due to the effects of chronic disease and is possibly associated with the increased levels of tumor necrosis factor reported in these patients. In addition, some of the granulomatous hepatitis and/or nonspecific findings detected before 1990 could represent sequelae of undetected HCV.

Neither radiologic testing nor liver chemistries are predictive of specific liver abnormalities. Whereas patients infected with MAC had significantly elevated alkaline phosphatase levels (p = 0.01) and patients with viral hepatitis had significantly elevated AST levels (p = 0.037), all diagnostic groups, including those with nonspecific or normal liver biopsy findings, had elevated liver chemistries. The association of granulomatous disease of the liver with significantly elevated alkaline phosphatase has been reported elsewhere (5) and is presumably due to its infiltrative nature causing obstruction of terminal branches of the biliary tree.

Past studies have shown inpatient liver biopsy to be a safe procedure with morbidity of 0.1 to 0.6% and mortality of 0 to 0.12% (34-36). Although there may be slightly increased bleeding risk in patients with AIDS (37), liver biopsy is the most specific diagnostic method for liver disease in patients seropositive for HIV. We did not attempt to evaluate the complication rate of biopsies in this large population, but a subset of these biopsies performed at Bellevue Hospital were examined (37). Of 100 random biopsies, five patients (5%) experienced bleeding complications defined as a hematocrit drop of more than 10 points. One of these events was fatal. It is not possible to establish safety criteria for liver biopsy in HIV-seropositive patients based on these data, but we avoid biopsy in patients with a prothrombin time greater than 3 s over control or a platelet count below 50,000. If biopsy is emergent in patients with a higher prothrombin time or lower platelet count, we perform the procedure after administration of fresh-frozen plasma or platelet supplementation. Future studies should help to determine when it is unsafe to perform liver biopsy in this population, and guidelines need to be established.

This study is limited by its retrospective design and by reliance on data from three different hospitals. With this in mind, we could not evaluate whether liver biopsy had a positive effect on patient outcome and survival; nor did we attempt to prove that liver biopsy resulted in a change in treatment or a change in preprocedure clinical diagnoses or survival. Thus, questions about the efficacy of liver biopsy cannot be answered. The study is unique because of its large size and its epidemiological characteristics. The population studied is representative of the scope of the AIDS epidemic, and the evolving course of the epidemic is captured by the large time period this study spans. Due to the large number of patients examined, the retrospective design, and number of hospitals involved, liver biopsies were examined by numerous pathologists, which is unavoidable in a study of its kind. A further study reexamining these biopsies with a focus on using newer techniques for identifying CMV and HCV will be undertaken. Further studies should also determine how many biopsies must be deferred in HIV-seropositive patients due to coagulopathy, right-sided pleural effusions, or other contraindications.

This study was not designed to determine whether liver biopsy should be attempted in all HIV-seropositive patients with fever, abnormalities of liver tests, or hepatomegaly of uncertain etiology. What it does reveal is that when biopsies were performed in patients with the above characteristics, 64.3% were diagnostic, many with treatable disorders. The utility of liver biopsy among these patients will still be controversial. Some argue that most illnesses detected by liver biopsy lack definitive treatment; so it is unreasonable to subject the patient to an invasive procedure associated with a significant complication rate. Furthermore, many of the disorders are systemic, and diagnoses can be made by noninvasive means or by a less invasive bone marrow biopsy.

Acknowledgment: We dedicate this paper in memory of Elliot Schwarz, whose untimely death occurred before the paper was published.

Back to Top | Article Outline

REFERENCES

1. Dworkin BM, Stahl RE, Giardina MA, et al. The liver in acquired immune deficiency syndrome: emphasis on patients with intravenous drug abuse. Am J Gastroenterol 1987;82:231-6.
2. Lebovics E, Thung SN, Schaffner F, et al. The liver in acquired immunodeficiency syndrome: a clinical and histologic study. Hepatology 1985;5:293-8.
3. Schneiderman DJ, Arenson DM, Cello JP, et al. Hepatic disease in patients with the acquired immune deficiency syndrome (AIDS). Hepatology 1987;7:925-30.
4. Gordon SC, Reddy R, Gould EE, et al. The spectrum of liver disease in the acquired immunodeficiency syndrome. J Hepatol 1986;2:475-84.
5. Kahn SA, Saltzman BR, Klein RS. Hepatic disorders in the acquired immune deficiency syndrome: a clinical and pathological study. Am J Gastroenterol 1986;81:1145-8.
6. Comer GM, Mukherjee S, Scholes JV, et al. Liver biopsies in the acquired immune deficiency syndrome: influence of endemic disease and drug abuse. Am J Gastroenterol 1989;84:1525-31.
7. Cappell MS, Schwartz MS, Biempica L. Clinical utility of liver biopsy in patients with serum antibodies to the human immunodeficiency virus. Am J Med 1990;88:123-30.
8. Orenstein MS, Tavitian A, Yonk B, et al. Granulomatous involvement of the liver in patients with AIDS. Gut 1985;26:1220-5.
9. Boylston AW, Cook HT, Francis ND, et al. Biopsy pathology of acquired immunodeficiency syndrome (AIDS). J Clin Pathol 1987;40:1-8.
10. Waisman J, Rotterdam H, Niedt GN, et al. AIDS: an overview of the pathology. Pathol Res Pract 1987;182:729-54.
11. Hawkins CC, Gold JWM, Whimbey E, et al. Mycobacterium avium complex infections in patients with acquired immunodeficiency syndrome. Ann Intern Med 1986;105:184-8.
12. Bonacini M. Hepatobiliary complications in patients with human immunodeficiency virus infection. Am J Med 1992;92:404-11.
13. Minguez A, Bernard E, Biron F, et al. Yield of bone marrow (BM) cultures in the diagnosis of tuberculosis (TB) and Mycobacterium avium complex (MAC) infection in HIV infected patients. Presented at the VIIth International Conference of AIDS, 1991;268.
14. Prego V, Glatt AE, Roy V, et al. Comparative yield of blood culture for fungi and mycobacteria, liver biopsy and bone marrow biopsy in the diagnosis of fever of undetermined origin in human immunodeficiency virus-infected patients. Arch Intern Med 1990;150:333-6.
15. Elmer S, Broers B, Boettger E, et al. Rapid diagnosis of M. avium infection by sequencing of amplified DNA from a liver biopsy [Abstract PO-B07-1144]. IXth International Conference on AIDS, Berlin, 1993.
16. Chaisson RE, Schecter GF, Theuer CP, et al. Tuberculosis in patients with the acquired immunodeficiency syndrome. Clinical features, response of therapy and survival. Am Rev Respir Dis 1987;136:570-4.
17. Stevens CE, Taylor PE, Pindyck J, et al. Epidemiology of hepatitis C virus. A preliminary study in volunteer blood donors. JAMA 1990;263:49-53.
18. Aguilera A, Carnero A, Scoane R, et al. Influence of HIV infection on the viral replication markers of chronic hepatitis B [Abstract PO-B08-1319]. Ninth International Conference on AIDS, Berlin, 1993.
19. Botti P, Pistelli A, Gambassi F, et al. HBV and HCV infection in IV drug addicts; coinfection with HIV. Arch Virol 1992;4:S329-32.
20. Eskild A, Magnus P, Petersen G, et al. Hepatitis B antibodies in HIV-infected homosexual men are associated with more rapid progression to AIDS. AIDS 1992;6:571-4.
21. Eyster ME, Diamondstone LS Lien J-M, et al. Natural history of hepatitis C virus infection in multitransfused hemophiliacs: effect of coinfection with human immunodeficiency virus. J Acquir Immune Defic Syndr 1993;6:602-10.
22. Eyster ME, Fried MW, DiBisceglie AM, et al. HCV RNA in hemophiliacs with antibody to hepatitis C virus (HCV): relationship to HIV infection and liver failure [Abstract PO-B02-0924]. IXth International Conference on AIDS, Berlin, 1993.
23. Quan CM, Krajden M, Grigoriew GA, et al. Hepatitis C virus infection in patients infected with the human immunodeficiency virus. Clin Infect Dis 1993;17:117-9.
24. Llibre JM, Garcia E, Aloy A, et al. Hepatitis C virus infection and progression of infection due to human immunodeficiency virus. Clin Infect Dis 1993;16:182.
25. Clouse KA, Robbins PB, Fernie B, et al. Viral antigen stimulation of the production of human monokines capable of regulating HIV 1 expression. J Immunol 1989;143:470-5.
26. Twu JS, Robinson WS. Hepatitis virus X gene can transactivate heterologous viral sequences. Proc Natl Acad Sci USA 1989;86:2046-50.
27. Scharschmidt BF, Held MJ, Hollander HH, et al. Hepatitis B in patients with HIV infection: relationship to AIDS and patient survival. Ann Intern Med 1992;10:837-8.
28. Reichert CM, O'Leary TJ, Levens DL, et al. Autopsy pathology in the acquired immune deficiency syndrome. Am J Pathol 1983;12:357-82.
29. Guardia LA, Luna MA, Smith JL Jr, et al. Acquired immune deficiency postmortem findings. Am J Clin Pathol 1984;81:549-57.
30. Mortality in patients with the acquired immunodeficiency syndrome treated with either foscarnet or ganciclovir for cytomegalovirus retinitis. Studies of the ocular complications of AIDS research group, in collaboration with the AIDS Clinical Trials Group. N Engl J Med 1992;326:213-20.
31. Bach N, Theise ND, Schaffner F. Hepatic histopathology in the acquired immunodeficiency syndrome. Semin Liver Dis 1992;12:205-12.
32. Cao YZ, Dieterich D, Thomas PA, et al. Identification and quantification of HIV-1 in the liver of patients with AIDS. AIDS 1992;6:65-70.
33. Scoazec JY, Feldmann G. Both macrophages and endothelial cells of the human hepatic sinusoid express the CD4 molecule, a receptor for the human immunodeficiency virus. Hepatology 1990;12:505-10.
34. Terry R. Risks of needle biopsy of the liver. Br Med J 1952;1:1102-5.
35. Sherlock S, Dick R, Van Leeuwen DJ. Liver biopsy today. The Royal Free Hospital experience. J Hepatol 1984;1:75-85.
36. Piccinino F, Sagnelli E, Pasquale G, et al. Complications following percutaneous liver biopsy. A multicentre retrospective study on 68,276 biopsies. J Hepatol 1986;2:165-73.
37. Schwarz E, Lew R, Dicker MA, et al. Complications of liver biopsies in patients seropositive for human immunodeficiency virus (HIV) [Abstract]. Gastroenterology 1993;4: A777.
Keywords:

Liver; Biopsy; HIV; Histology; Diagnosis

© Lippincott-Raven Publishers.