Highly active antiretroviral therapy (HAART) has dramatically changed the life expectancy of HIV-positive individuals, reducing the risk of opportunistic infections and deaths.1,2 However, other illnesses such as liver cirrhosis have emerged as important causes of morbidity and mortality in this population.3-5 Besides hepatitis B (HBV) or C (HCV) viruses, other multiple factors may cause liver damage in HIV-positive individuals, including opportunistic pathogens, tumors, autoimmune diseases, alcohol abuse, and prescription of potentially hepatotoxic medications. However, persistent liver enzyme elevation in the absence of any recognizable cause is seen only rarely in HIV-positive patients. The role of HIV itself, a new unknown hepatotropic agent, and/or some antiretrovirals has been postulated for these cases.6,7
Herein, we describe a group of HIV-positive individuals with liver disease of unknown origin, in which extensive studies were conducted to rule out potential known causes of liver damage. The frequency of this entity and its clinical presentation and possible causes were assessed.
PATIENTS AND METHODS
Cryptogenic liver disease (CLD) was defined as persistently elevated aminotransferase levels in the absence of replicative HCV and HBV infections or other common causes of liver disease (alcohol abuse, hepatotoxic medications, etc).8 Normal values of liver enzymes were 10 to 55 IU/L for ALT and 10 to 40 IU/L for AST. We identified subjects initially meeting these criteria while examining HIV-positive patients attending 2 large HIV clinics in Spain-one in Madrid and another in Seville-in the year 2004. The clinical charts of these individuals were retrospectively reviewed, and their assessment completed when needed to rule out other less frequent causes of liver disease, including autoimmune hepatitis, nonalcoholic steato-hepatitis (NASH), hemochromatosis, Wilson disease, porfiria, á1-antitrypsin deficiency and syphilis. Moreover, given the nature of this study, further studies were undertaken to exclude any possible known cause of liver damage.
The presence of "occult" HBV and HCV infections was excluded after testing plasma and peripheral blood mononuclear cells for the presence of HBV DNA and/or HCV RNA using sensitive real-time polymerase chain reaction assays, as described elsewhere, which have a detection limit of 6 and 10 IU/mL, respectively.9,10 The presence of genetic sequences of cytomegalovirus and transfusion transmitted virus were excluded in plasma using specific polymerase chain reaction tests.11,12 Patients with significant obesity, grade 3 to 4 dyslipidemias, hyperglycemia, insulin resistance, and/or showing ultrasonographic findings compatible with fatty liver disease were also excluded because these conditions have typically been associated with NASH.
The stage of liver fibrosis was assessed by liver biopsy and/or elastography, the latest using FibroScan (Echosens, Paris, France), a new noninvasive tool recently validated to assess liver fibrosis.13,14 In patients from whom a liver biopsy could be obtained, an experienced pathologist examined the tissue in detail searching for any recognizable cause of liver damage. To investigate which factors could be associated with the development of CLD, a group of HIV-positive controls with normal liver function tests was selected and matched by age, gender, and CD4 status.
A case report form was built in which information on demographics (age, gender, risk category), date of HIV diagnosis, duration of antiretroviral treatment, exposure to different antiretrovirals, and use of other medications was recorded. Information regarding the nadir CD4 count and baseline plasma HIV RNA and the current values for CD4 counts, plasma HIV RNA, transaminase levels, total bilirubin, prothrombin time, albumin, and most recent ultrasonographic findings were also recorded.
Once the CLD cases were identified, we chose a control for each who met the following inclusion criteria: HIV infection, absence of chronic HBV and/or HCV infections, and repeatedly normal transaminases levels. Patients were matched by age, gender, and CD4 counts.
χ2, Student t, and Mann-Whitney U tests were used to compare categorical and continuous variables between groups, respectively. Univariate and multivariate logistic regression analyses were performed to identify factors associated with CLD. For all analyses, a 2-sided P value of 0.05 was considered significant. All analyses were performed using the statistical software package SPSS, version 10.0 (SPSS Inc, Chicago, IL).
A total of 17 (0.5%) out of 3200 HIV-positive patients met the CLD criteria. The main demographic characteristics are summarized in Table 1. The majority (14/17, 82.4%) were male and had acquired HIV infection through homosexual relationships (13/17, 76%). Their median age was 42 years (range, 27-57). The median time since HIV diagnosis was over 15 years (range, 5-21).
The median time from the initial diagnosis of CLD was 5 years (range, 1-15). All subjects had been exposed to antiretroviral therapy, and the majority (88.2%) were taking HAART and had undetectable viral load at the time the study was conducted in the year 2004. All except 1 patient had CD4 counts >200 cells/μL (Table 2). The median time of exposure to the most potential hepatotoxic antiretrovirals was 49 months (range, 22-98) for didanosine (ddI), 21 (range, 0-59) for nevirapine (NVP), 36 (range, 8-74) for stavudine (d4T), and 0 (0-5) for ritonavir (RTV). None of these individuals was taking any other potentially hepatotoxic drug.
The main liver-related features are summarized in Table 3. Interestingly, parameters of hepatic function were within normal limits in most patients. The histologic findings based on either liver biopsy and/or elastography were as follows: F0-F1 in 7 (41%), F3 in 5 (29%), and F4 in 5 (29%) patients. A liver biopsy could be obtained from 5 patients and unfortunately did not reveal findings other than different unspecific degrees of liver fibrosis, microvesicular steatosis, and inflammation. For the remaining subjects with or without clinically manifest liver cirrhosis, patients refused a liver biopsy. In them, liver fibrosis was assessed through elastography, a noninvasive procedure.13,14
Nine (52.9%) out of 17 patients developed portal hypertension. Episodes of ascites were recorded in 8 (47%), portal thrombosis in 6 (35.2%), bleeding from grade III-IV esophageal varices in 5 (29.4%), and recurrent episodes of hepatic encephalopathy in 2 (11.7%). None of the patients had died since the time of first diagnosis of CLD until the last control in December 2005. Moreover, removal of ddI and/or d4T was followed by normalization of liver enzymes and steady recovery with no further episodes of liver decompensation in 5 out 7 patients.
Table 4 summarises the main features of the CLD cases and controls. There was a trend toward higher CD4 counts and higher plasma HIV RNA in controls compared with CLD cases. Similarly, CLD cases were more often homosexual men than controls, but the difference did not reach statistical significance. Despite a more extensive exposure to antiretroviral therapy in controls than CLD cases, the length of exposure to ddI was the only independently and significantly associated predictor of CLD. Overall, patients with CLD had received ddI for a median of 49 months compared with 29 months in controls (100% of cases and 94% of controls, respectively). In contrast, exposure to NVP and/or d4T tended to be less frequent and/or shorten in CLD cases than that in controls. Multivariate logistic regression analysis identified longer ddI exposure as the only factor associated with the development of CLD (Table 5).
We have reported a series of cases of liver disease of unknown origin in HIV-infected individuals. Hepatic complications developed in a substantial proportion (53%) of these patients. Occult chronic hepatitis B and C and other uncommon liver conditions were ruled out reasonably in all patients. To our knowledge, this is the first report of a relatively large group of CLD in HIV-positive individuals. The current prevalence of CLD seems to be low, around 0.5%, but it could be on the rise because exposure to some antiretroviral drugs seems to be involved in its pathogenesis. It will be of interest to know whether similar cases have been identified by other groups.
In the general population, CLD is a rare entity with rates below 0.01%15-17 and NASH accounts for a large proportion of cases. In this way, obesity, hyperglycemia, and hyperlipidemia are frequently associated and predisposing conditions. Although we could not exclude definitively NASH in our patients because a liver biopsy could not be obtained in all cases, we excluded all individuals with signs of fat infiltration in the liver by ultrasonograhy, or having dyslipidemias, increased body mass index, hyperglycemia, and/or insulin resistance. However, it should be highlighted that the hepatic tissue taken from the 5 patients who underwent a liver biopsy uniformly showed microvesicular steatosis, which could reflect mitochondrial damage.7,18
Although HIV itself has been associated to specific organ damage in infected persons, causing HIV-associated dementia,19 nephropathy,20 or enteropathy,21 our findings do no support a direct role of HIV as cause of CLD. First, the duration of HIV infection was longer in controls than CLD cases. Second, there was no evidence of higher plasma viremia and/or lower CD4 counts in CLD cases than controls, even when considering nadir values before initiating antiretroviral therapy. Finally and most importantly, no liver enzyme elevations were noted in CLD cases before initiating antiretroviral drugs.
Although antiretroviral therapy is clearly of benefit in most HIV-infected persons because it provides complete virus suppression and significant immune recovery, the hepatotoxicity of antiretroviral agents is a matter of growing concern.22 Aside from acute episodes of liver toxicity associated with some antiretroviral drugs (ie, NVP and efavirenz),23 long-term consequences of iatrogenic liver damage need to be assessed more carefully in HIV-infected persons. For example, recent reports have highlighted that prolonged NVP exposure could be associated with liver enzymes elevations and/or enhanced liver fibrosis, particularly in subjects with concomitant chronic hepatitis C.24-27 On the other hand, potent antiretroviral therapy may halt the deleterious impact of HIV-related immunodeficiency on the faster liver fibrosis progression seen in HIV/HCV-coinfected patients.28-30 In agreement with those observations, recent clinical findings from large cohorts of HIV-infected individuals have highlighted that although HAART may be generally of benefit to reduce the incidence of liver-related complications in patients with underlying HBV and/or HCV infections,31-33 hepatic morbidity might be increased as a result of antiretroviral exposure.34,35
In the case-control study, the only single predictor of developing CLD in our population was long-term exposure to ddI. This nucleoside analogue is known to produce hepatic mitochondrial damage,36,37 and patients with HCV-related liver cirrhosis may be prone to liver decompensation.38 Moreover, ddI treatment has occasionally been associated with hyperlactatemia and lactic acidosis in which liver damage was a prominent part of the syndrome.39,40 However, in our knowledge, our study is the first to demonstrate that prolonged exposure to ddI could be associated specifically with chronic liver disease and severe hepatic impairment. Of note, 3 of our patients had received ddI along with hydroxyurea, which is known to enhance the intracellular metabolites of ddI and eventually their toxic effects.41
Direct histologic findings could only be collected from 5 of our subjects in whom a liver biopsy was performed. Two showed liver cirrhosis and the others showed mild fibrosis. Besides fibrosis and inflammation, different degrees of microvesicular steatosis were present in all cases. Moreover, most CLD cases showed elevated GGT and alkaline phosphatase, and progressive cholestasis has already been reported in several cases of mitochondrial dysfunction in HIV/HCV-coinfected patients taking ddI along with ribavirin, which enhances the intracellular phosphorylated forms of the former.42-44 Although our patients had not been exposed to ribavirin, 3 of them had taken hydroxyurea along with ddI, as mentioned earlier. No other relevant features could be recognized in these biopsies, including features of severe NASH such as Mallory bodies, zone 3 pericellular fibrosis, hepatocyte ballooning, and/or lobular inflammation. Prior reports have underlined that there are no characteristic histopathologic findings in patients experiencing ddI-related hepatotoxicity.45 Therefore, we are confident that our findings exclude other potential causes of liver damage and may be compatible with ddI-related hepatotoxicity.
One additional striking finding in this series of CLD cases was the frequency of complications related to the portal venous system. Portal thrombosis was recognized in one third of patients. Inasmuch as screening for coagulation disorders yielded negative results in these patients, an alternative explanation is warranted. Some of the CLD cases had normal liver function tests but showed portal thrombosis. Clearly, decompensated liver disease (ascites, bleeding, etc) in these patients could not be explained by either mild fibrosis or microvesicular steatosis. Presinusoidal portal hypertension was the main feature in these patients. Nodular regenerative hyperplasia of the liver is an uncommon hepatic lesion often associated with noncirrhotic portal hypertension.46 It is believed to be associated with some drugs (ie, azathioprine and 6-thioguanine),47,48 immunologic disorders or gastrointestinal pathogens.46 Because autoimmune diseases were excluded in all our CLD patients, a putative drug-related and/or microbial etiology was the most likely cause. This condition is difficult to diagnose,49 often missed in the liver biopsy, and, in some cases, it may evolve to liver cirrhosis. Moreover, the recognition of cholestasis in all but one of the CLD cases is in agreement with such a diagnosis.46
It is noteworthy that most HIV-infected subjects with CLD in our series were men who had sex with men (MSM), whereas this group only represents 41% of our whole HIV-infected study population. Hypothetically, repeated episodes of pylephlebitis caused by gastrointestinal pathogens could be more frequent among MSM. Moreover, septic microtrombophlebitis could be favored by the massive disruption of the intestinal immunologic barrier caused by HIV infection,50 increasing the risk of portal thrombosis in subjects having a reduced portal flow secondary to iatrogenic liver damage.
In summary, we have described a series of HIV-infected individuals with liver damage of unknown etiology, some of whom developed serious liver complications, often derived from portal hypertension. Our preliminary findings suggest that antiretroviral therapy, and in particular prolonged ddI exposure, might be involved in the pathogenesis of this condition. Thus, although antiretroviral therapy may overall ameliorate liver disease progression in most patients coinfected with HCV or HBV, prolonged exposure to some antiretrovirals might be detrimental for the liver in a different subset of individuals.
We would like to thank Xavier Forns, Mike Youle, Massimo Puoti, Stanislas Pol, and Yves Benhamou for reviewing the manuscript.
1. Palella F, Delaney K, Moorman A, et al. Declining morbidity and mortality among patients with advanced HIV
Outpatient Study Investigators. N Engl J Med
2. The CASCADE Collaboration. Survival after introduction of HAART in people with known duration of HIV
-1 infection. Lancet
3. Martin-Carbonero L, Soriano V, Valencia E, et al. Increasing impact of chronic viral hepatitis
on hospital admission and mortality among HIV
-infected patients. AIDS Res Hum Retroviruses
4. Bica I, McGovern B, Dhaz R, et al. Increasing mortality due to end-stage liver
disease in patients with HIV
infection. Clin Infect Dis
5. Rosenthal E, Poiree M, Pradier C, et al. Mortality due to hepatitis
disease in HIV
-infected patients in France (Mortavic 2001 study). AIDS
6. Duffy L, Daum F, Kahn E, et al. Hepatitis
in children with AIDS: histopathologic and immunocytologic features. Gastroenterology
7. Walker U, Bauerle J, Laguno M, et al. Depletion of mitochondrial DNA in liver
under antiretroviral therapy
, stavudine or zalcitabine. Hepatology
8. Skelly M, James P, Ryder S. Findings on liver
biopsy to investigate abnormal liver
function tests in the absence of diagnostic serology. J Hepatol
9. Weiss J, Wu H, Farrenkopf B, et al. Real time TaqMan PCR detection and quantification of HBV genotypes A-G with the use of an internal quantitation standard. J Clin Virol
10. Konnick E, Williams S, Ashwood E, et al. Evaluation of the COBAS hepatitis
C virus TaqMan specific reagent assay and comparison to the COBAS Amplicor HCV Monitor v2.0 and Versant HCV bDNA v3.0 assays. J Clin Microbiol
11. Gartner B, Fischinger J, Litwicki A, et al. Evaluation of a new automated, standardized generic nucleic acid extraction method used in combination with cytomegalovirus DNA quantification by COBAS Amplicor CMV Monitor. J Clin Microbiol
12. Koidl C, Michael B, Berg J, et al. Detection of transfusion-transmitted virus DNA by real-time PCR. J Clin Virol
13. Sandrin L, Fourquet B, Hasquenoph J, et al. Transient elastography: a new non-invasive method for assessment of hepatic fibrosis. Ultrasound Med Biol
14. Castera L, Vergniol J, Foucher J, et al. Prospective comparison of transient elastography, Fibrotest, APRI, and liver
biopsy for the assessment of fibrosis in chronic hepatitis
15. Greeve M, Ferrell L, Kim M, et al. Cirrhosis of undefined pathogenesis: absence of evidence for unknown viruses or autoimmune processes. Hepatology
16. Caldwell S, Oelsner D, Iezzoni J, et al. Cryptogenic cirrhosis: clinical characterization and risk factors for underlying disease. Hepatology
17. Caldwell S, Crespo D. The spectrum expanded: cryptogenic cirrhosis and the natural history of non-alcoholic fatty liver
disease. J Hepatol
18. De Mendoza C, Sanchez-Conde M, Timmermans E, et al. Mitochondrial DNA depletion in HIV
-infected patients is more pronounced with chronic hepatitis
C and enhanced following treatment with pegylated interferon plus ribavirin. Antivir Ther
19. Gonzalez-Scarano F, Martin-Garcia J. The neuropathogenesis of AIDS. Nat Rev Immunol
20. Lu T, Ross M. HIV
-associated nephropathy. A brief review. Mt Sinai J Med
21. Giovanni B, Calabrese C, Manfredi R, et al. HIV
enteropathy: undescribed ultrastructural changes of duodenal mucosa and their regression after triple antiretroviral therapy
. Dig Dis Sci
22. Nuñez M, Soriano V. Hepatotoxicity of antiretrovirals: incidence, mechanisms and management. Drug Saf
23. Sanne I, Mommeja-Marin H, Hinkle J, et al. Severe hepatotoxicity associated with nevirapine use in HIV
-infected subjects. J Infect Dis
24. Gonzalez de Requena D, Núñez M, Jiménez-Nacher I, et al. Liver
toxicity caused by nevirapine. AIDS
25. Martinez E, Blanco JL, Arnaiz J, et al. Hepatotoxicity in HIV
-1 infected patients receiving nevirapine containing antiretroviral therapy
26. Núñez M, Gonzalez de Requena D, Gonzalez-Lahoz J, et al. Interactions between nevirapine plasma levels, chronic hepatitis
C, and the development of liver
toxicity in HIV
-infected patients. AIDS Res Hum Retroviruses
27. Macías J, Castellano V, Merchante N, et al. Effect of antiretroviral drugs on liver
fibrosis in HIV
-infected patients with chronic hepatitis
C: harmful impact of nevirapine. AIDS
28. Benhamou Y, De Martino V, Bochet M. Factors affecting liver
fibrosis in HIV
C virus coinfected patients: impact of protease inhibitor therapy. Hepatology
29. Fuster D, Planas R, Muga R, et al. Advanced liver
fibrosis in HIV
/HCV-coinfected patients on antiretroviral therapy
. AIDS Res Hum Retroviruses
30. Mariné-Barjoan E, Saint-Paul M, Pradier C, et al. Impact of antiretroviral treatment on progression of hepatic fibrosis in HIV
C virus co-infected patients. AIDS
31. Qurishi N, Kreuzberg C, Lichters G, et al. Effect of antiretroviral therapy
-related mortality in patients with HIV
C virus coinfection. Lancet
32. Bonacini M, Bzowej N, Louie S, Wohl A. Survival in patients with HIV
infection and viral hepatitis
B or C. AIDS
33. Nuñez M, Soriano V. How effective is HAART in HCV and HIV
34. Clark S, Creighton S, Portmann B, et al. Acute liver
failure associated with antiretroviral treatment for HIV
: a report of six cases. J Hepatol
35. Nuñez M, Soriano V. Risks and benefits of antiretroviral therapy
-HCV co-infected patients. J Hepatol
36. Setzer B, Schlesier M, Walker U. Effects of didanosine
-related depletion of mtDNA in human T lymphocytes. J Infect Dis
37. Lee H, Hanes J, Johnson K. Toxicity of nucleoside analogues used to treat AIDS and the selectivity of the mitochondrial DNA polymerase. Biochemistry
38. Mauss S, Valenti W, de Pamphilis J, et al. Risk factors for hepatic decompensation in patients with HIV
-HCV coinfection and liver
cirrhosis during interferon-based therapy. AIDS
39. Lai K, Gang D, Zawacki J, et al. Fulminant hepatic failure associated with 2,3-dideoxyinosine (ddI). Ann Intern Med
40. Bissuel F, Bruneel F, Habersetzer F, et al. Fulminant hepatitis
with severe lactate acidosis in HIV
-infected patients on didanosine
therapy. J Intern Med
41. Foli A, Lori F, Maserati R, et al. Hydroxyurea and didanosine
is a more potent combination than hydroxyurea and zidovudine. Antivir Ther
42. Lafeuillade A, Hittinger G, Chadapaud S. Increased mitochondrial toxicity with ribavirin in HIV
/HCV coinfection. Lancet
43. Guyader D, Poinsignon Y, Cano Y, Saout L. Fatal lactic acidosis in a HIV
-positive patient treated with interferon and ribavirin for chronic hepatitis
C. J Hepatol
44. Bani-Sadr F, Carrat F, Pol S, et al. Risk factors for symptomatic mitochondrial toxicity in HIV
/HCV-coinfected patients during interferon plus ribavirin-based therapy. J Acquir Immune Defic Syndr
45. Lacaille F, Fournet J, Blanche S. Clinical utility of liver
biopsy in children with acquired immunodeficiency syndrome. Pediatr Infect Dis J
46. Arvanitaki M, Adler M. Nodular regenerative hyperplasia of the liver
-a review of 14 cases. Hepatogastroenterology
47. Daniel F, Cadranel J, Seksik P, et al. Azathioprine induced nodular regenerative hyperplasia in IBD patients. Gastroenterol Clin Biol
48. De Bruyne R, Portmann B, Samyn M, et al. Chronic liver
disease related to 6-thioguanine in children with acute lymphoblastic leukaemia. J Hepatol
49. Ibarrola C, Colina F. Clinico-pathological features of nine cases of non-cirrhotic portal hypertension: current definitions and criteria are inadequate. Histopathology
50. Veazey R, Lackner A. Getting to the guts of HIV
pathogenesis. J Exp Med
Keywords:Copyright © 2006 Wolters Kluwer Health, Inc. All rights reserved.
HIV; liver; didanosine; hepatitis; antiretroviral therapy