Human Immunodeficiency Virus and the Liver: Lessons Learned and Still to Be Learned

The liver may be the most important organ in the world of human immunodeficiency virus (HIV) today. It is involved with so many aspects of HIV pathogenesis and treatment that its importance in HIV care mirrors the liver's importance in the body. This spring marks the 22nd anniversary of the first report of the beginning of the HIV era. In this issue of Seminars in Liver Disease, we offer a few primers on HIV for the practicing gastroenterologist or hepatologist. We also review the effects of hepatitis viruses B (HBV) and C (HCV) and GB virus type C on the liver in the course of HIV and consider the liver toxicity of antiretroviral treatment and how to best manage it.

What can the hepatologist learn from the study of HIV? There are many vital lessons, but the most obvious come from virology. There is, first of all, a basic appreciation of the HIV virus, which may now be the best-understood virus in the pantheon of science. It has a few things in common with hepatitis viruses, but understanding the basic immunology and virology of HIV will put you in good stead in comprehending the same issues with HBV and HCV. The pharmacologic treatment of HIV has advanced exponentially in the past 22 years. There are now 16 drugs approved for the treatment of HIV in the United States. They are always used in combination to avoid viral resistance and to avoid overlapping toxicities. The use of resistance testing in HIV has become standard practice. The HBV and HCV treatment worlds can use the resistance principles from HIV and apply them easily to their particular disciplines. Appreciation of these principles will be of particular importance as development of small molecule candidates for HCV treatment advances. The use of therapeutic drug monitoring for efficacy and perhaps toxicity avoidance is standard in some countries in Europe and probably should be so in the United States. It is now possible, for example, to measure ribavirin levels,[1] and measuring levels may improve the efficacy and safety of HCV treatment in both HCV-infected and HCV/HIV-coinfected individuals. The early studies of ribavirin for HIV clearly pointed out its ability to increase didanosine (ddI) levels both in vitro and in vivo.[2] This has become particularly important in treating HIV/HCV-coinfected patients. Several reports of hepatic decompensation, lactic acidosis, and death have been attributed to using ribavirin together with ddI, and perhaps with stavudine (d4T) as well.[3] [4] It would therefore be our recommendation not to use these pairs together to treat HCV. It is relatively easy now to find a nucleoside backbone for HIV therapy that does not contain "d" drugs (ddI, D4T, dideoxycytidine [ddC]) while a patient is receiving ribavirin for HCV therapy.

One very clear lesson learned from the treatment of HIV/HCV coinfected patients is that there is a very real sustained response rate to therapy. HCV can be eradicated in HIV-infected patients as well as in HIV-seronegative patients. Treating the HCV may offer protection from liver disease, better toleration of HIV medication, and therefore an even better prognosis for HIV. Patients with HIV and HCV should in general be treated just like patients who do not have HIV.

Another lesson learned from HIV that can be applied to the world of liver disease is the use of erythropoietin (EPO) to treat ribavirin-induced anemia. EPO has long been used to treat HIV and zidovudine (AZT)-induced anemia in HIV-infected patients, so the obvious next step was to use it in the treatment of HCV. EPO clearly works to increase hemoglobin levels, the amount of ribavirin that can be administered, and quality of life in anemic HCV patients taking ribavirin. Whether that translates into a better-sustained virological response for HCV therapy remains to be seen.[5] Using EPO in HIV/ HCV-infected patients being treated with interferon and ribavirin is an especially easy leap of faith because of the prior experience in HIV-related anemia.[6]

In the world of hepatitis B, the lessons learned from HIV are obvious but are not yet being applied, except in some rare clinical trials. The most obvious lesson is that nucleoside analogue monotherapy sooner or later leads to resistance in almost every patient. Because hepatitis B has a reverse transcription phase of replication, some of the nucleoside and nucleotide reverse transcriptase inhibitors used to treat HIV, such as lamivudine (3TC) and tenofovir, are active against HBV.[7] The problem of HBV resistance to 3TC is becoming common in the treatment of HBV alone and is nearly universal in the treatment of HBV in the presence of HIV.[8] Fortunately, tenofovir is effective against both 3TC-resistant HIV and HBV, and no HBV resistance has yet been demonstrated against tenofovir or adefovir, its nucleotide cousin.

The pathogen named hepatitis GB virus type C is an interesting virus that may not be a problem after all. It is a common infection that does not appear to be associated with any disease state. Current data suggest that it grows in lymphocytes and not in hepatocytes, which is probably why it does not cause liver disease. In addition, in HIV-infected patients, it may actually prolong survival! GBV-C in vitro leads to decreased HIV replication.[9]

Because the likelihood of more severe liver disease occurring in individuals coinfected with both hepatitis B and C has been amply demonstrated, what is the likely outcome for patients failing treatment for these agents? For hepatitis C, there is the possibility of maintenance pegylated interferon. Both pegylated interferons have been used for this: alfa 2b at 0.5 μg/kg and alfa 2a at 90 μg weekly. For hepatitis B, continued treatment is the only hope. There are some data suggesting that the resistant virus is slightly less fit. There is little else to offer these patients, because transplantation is not a viable option for most.

That brings us to the subject of liver transplantation in HIV-infected individuals. At a recent global "think tank" organized by the American Foundation for AIDS Research, eight centers doing liver transplants in HIV-infected individuals presented their experience to date. The range of experience was from 3 to 19 transplants in the post-highly active antiretroviral therapy (HAART) era. Overall, in about 60 transplants, the 1-year survival was about the same as United Network for Organ Sharing (UNOS) data: 85%.[10] The National Institutes of Health (NIH) is to be congratulated for sponsoring a large multicenter transplant trial. Within this trial, the inclusion criteria for liver transplant in HIV-infected patients are basically the same as for any liver transplant, with the following proviso. Each center gets to choose from two options. Either the patient has to have his or her HIV viral load at < 50 copies for 3 months or be suppressible to that level after transplant by taking antiretroviral drugs. The other option would allow more discretion to the HIV physician taking care of the patient and tolerate slightly higher viral loads at transplant if the patient is considered to be suppressible after transplant. So far it appears that there is little downside risk in transplanting the livers of carefully selected HIV-infected patients. A controlled multicenter protocol is about to begin, the results of which will be extremely important to both the HIV and the transplant communities. The shortage of organs in this country is an increasing problem for transplant centers, but the great strides in HIV treatment have led to nearly normal life expectancies for patients with HIV.[11] This argues strongly that, within some obvious parameters, patients with HIV should not be treated any differently in the transplant world than people with diabetes or hypertension are treated. HIV is a chronic, controllable disease.

Hans Popper once said that drug-induced liver toxicity is the penalty for progress. Nowhere is that more apparent than in the field of HIV. There have been 16 drugs approved since the epidemic began in 1981. The potential benefit of therapy for HIV is enormous. Survival times after diagnosis with HIV have changed from weeks in the early 1980s to virtually normal life expectancy today. This makes establishing an accurate diagnosis of drug-induced liver toxicity all the more important. A simple but highly useful way to approach this is to employ a checklist of five key questions[12]: (1) Has this drug or class of drugs been previously described as hepatotoxic? (2) Is the timing appropriate in this case, given what is known about the particular liver toxicity of that drug? (3) Did liver tests improve when the drug was discontinued? (4) Did test abnormalities recur if the drug was restarted? (5) Were other causes of hepatitis ruled out? The situation can be more complicated in HIV patients, of course, but the process should be the same. Figures 1 through 4, "Algorithms for Hepatotoxicity Management," are currently a useful way to look at this problem. However, because things are always changing in the HIV field, new pathways may need to be added to these algorithms in the future.

The article by Montessori et al in this issue[13] points out a phenomenon that overlaps the fields of HIV and hepatitis, that of the mitochondrial toxicity of nucleoside analogues. The first report in the literature was that of fialuridine.[14] It caused severe mitochondrial toxicity, liver failure, lactic acidosis, pancreatitis, and the whole spectrum of symptoms that we now readily identify with nucleoside analogues. Almost all nucleoside analogues are capable of producing this phenomenon. Accordingly, the clinician should always be vigilant for the patient who complains of nausea, vomiting, abdominal pain, and fatigue and who is on nucleosides. This is especially true with ddI, d4T, and ddC, although all of the nucleosides active against HIV are capable of producing this toxicity. The easiest, although not most accurate, way to diagnose this is to measure the serum lactate. If it is elevated, it is essential to stop all medications and confirm the diagnosis.[15] In HIV care, it is dogma that you must stop all HIV medications, not just the one you suspect is causing the toxicity, because of the likelihood of developing resistance.

The protease inhibitors present a complex problem. Many of them inhibit p450 cyp3a, and many are hepatotoxic, as described in the Sulkowski article in this issue.[16] However, the protease inhibitors also seem to be implicated in the pathogenesis of the lipodystrophy syndrome.[17] They can cause insulin resistance and hyperlipidemia,[18] both of which can have secondary effects on the liver. A recent study by Sutinen et al[19] actually measures liver fat in patients with HIV and the lipodystrophy syndrome. They found that liver fat was significantly increased (by 53%) in the lipodystrophy patients compared with the HIV-infected patients on HAART without lipodystrophy or HIV-seronegative controls. Liver fat in this study correlated with insulin resistance but not with intra-abdominal fat. That is the case of course, in nonalcoholic steato hepatitis (NASH) in the HIV-seronegative population.[20] As in NASH patients with insulin resistance, agents that increase insulin sensitivity such as rosiglitazone are useful and in this case may even help with body fat redistribution.[21] Of course some drugs in that family, such as troglitazone, have been implicated in liver toxicity as well.[22] This should remind us to be vigilant to the possibility that any drug a patient is taking, whether designed for HIV or not, is a potential cause of liver toxicity.

This should give us a lot to think about for our patients who have HIV, are taking nucleoside analogues and protease inhibitors, and perhaps have hepatitis C as well. The possibilities for additive or synergistic toxicities are enormous. As reviewed in articles within this issue, the management of the metabolic complications of HIV medications to mitigate toxicity should constantly be considered, and is well-characterized in another recent publication.[23]

The non-nucleoside reverse transcriptase inhibitors (NNRTI) present less of a conundrum but also have some issues that need to be fleshed out. There has been much confusion in this area because of lack of accurate data and of prospective randomized trials comparing the two major players in the field, nevirapine and efavirenz. Varying definitions of liver toxicity in the literature have also clouded the field. All investigators in the future should use the National Institute of Allergy and Infectious Disease grading scale of toxicity, so that studies reporting liver toxicity may be compared with each other. It appears that, in the earliest reports, both nevirapine and efavirenz demonstrated about 3% grade 3 (aspartate aminotransferase [AST]/alanine aminotransferase [ALT] 5 times the upper limit of normal) or 4 (AST/ALT 10 times the upper limits of normal) liver toxicity.[24] Then there were reports of higher toxicity with nevirapine, up to about 15% in the Atlantic trial.[25] All patients in that trial were also taking ddI, which is the most likely of the nucleosides to cause mitochondrial toxicity and hepatitis.[26] After that report, there were other reports of acute hepatitis in patients and in health-care workers taking nevirapine for HIV prophylaxis.[27] Finally, the data from the Triangle Trial, FTC 302, was released showing again a 15% rate of liver toxicity with nevirapine compared with none for efavirenz.[28] The AIDS Clinical Trials Group (ACTG) database of more than 10,000 patients, on the other hand, suggested that both efavirenz and nevirapine were about equal in their grade 3 and 4 liver toxicities.[27] Boehringer Ingelheim, the manufacturer of nevirapine, undertook an exhaustive review of its own database along with many other databases to try to shed some light on this problem and came up with some real answers to help explain these conflicting findings. The results point out that, in a review of more than 4000 patients, virtually all of the liver toxicity of nevirapine occured in the first 6 weeks and was idiosyncratic and often accompanied by fever, rash, and eosinophila. After the first 6 weeks, there is no difference between nevirapine and placebo in liver toxicity. The important risk factors for this problem are female sex and CD4 count of > 250 and male sex and CD4 count of > 400.[29] The level of CD4 cells at which treatment is recommended for HIV is < 350.[30] This is a good example of ethical pharmaceutical behavior to shed some light on a real problem for patients and clinicians. A randomized comparative trial of efavirenz versus nevirapine was presented in February 2003, allowing a fair comparison of antiviral efficacy and toxicity. The results showed comparable efficacy, but some showed increased liver toxicity with nevirapine given once daily (13%) versus 4.5% for efavirenz. The nevirapine twice daily arm showed 7.8% grade 3/4 liver toxicity, which was not significantly different than the efavirenz arm or the combined nevirapine and efavirenz arm.[31] This seems unusual to me. If the toxicity is early and idiosyncratic, why would giving the drug once daily make any difference in toxicity? Clearly this question needs further exploration.

When considering liver toxicity in the HIV patient, many factors must be considered. There are new areas to discuss, such as pharmacogenomics, drug disposition, and drug targets, about which very little is really known.[32] Drug metabolism may be an entirely individual process. All drugs are capable of causing liver toxicity. Some diseases, such as chronic hepatitis B and C predispose to liver toxicity, and treating them may lessen the chances of it occurring. When using drugs in HIV patients, a careful analysis of the risk to benefit ratio must be done. A thorough knowledge of the risks as outlined here will help the clinician make the best decision for individual patients.

Where does all this information lead? The conclusions to be drawn from these data suggest that thoughtful consideration of all factors involved in each and every case is the duty of the hepatologist. Treatment of HIV-related liver problems is complex and difficult but really should be approached in the same rational way that hepatologists face any liver problem today. Careful collection of the data, evaluation of the patient, and treatment when appropriate should be given to patients regardless of their HIV status. That HIV status, and the multitude of problems that it can bring, adds complexity to the treatment of liver disease but should not be an excuse for therapeutic nihilism or second-class treatment of any kind. We can all hope for an exponential increase in knowledge-such as that which has occurred in HIV-to happen in the world of liver research in the years ahead. Seeing the depth and breadth of our authors' writing in this issue, and using the lessons learned from the study of HIV, it will certainly occur.

REFERENCES

• 1 Larrat S, Stanke-Labesque F, Plage A. Ribavirin quantification in combination treatment of chronic hepatitis C.  Antimicrob Agents Chemother . 2003;  47 124-129
• 2 Vogt M W. Interaction between didanosine and ribavirin.  Science . 1987;  235 1376-1379
• 3 Mauss S. Risk factors for hepatic decompensation in cirrhotic patients with HIV-HCV coinfection treated with peg-interferon alfa or interferon alfa and ribavirin or placebo. Presented at the European Congress on AIDS. November 3, 2002; Glasgow, Scotland
• 4 Esteban R. Treatment of chronic hepatitis in HCV/HIV coinfected patients. Presented at the XIV International AIDS Conference. July 8, 2002; Barcelona, Spain. MoOrB113
• 5 Dieterich D T. Epoetin alfa treatment of anemic HCV patients allows for maintenance of ribavirin dose, increases hemoglobin levels and improves quality of life (in press).  Am J Gastroenterol 2003.
• 6 Dieterich D T. Treatment of hepatitis C and anemia in human immunodeficiency virus-infected patients.  J Infect Dis . 2002;  185 S128-S137
• 7 Nelson M, Portsmouth S, Stebbing J. An open label study of tenofovir in HIV-1 and hepatitis B virus co-infected individuals.  AIDS . 2003;  17 F7-F10
• 8 Benhamou Y, Bochet M, Thibault V. Long-term incidence of hepatitis B virus resistance to lamivudine in human immunodeficiency virus-infected patients.  Hepatology . 1999;  30 1302-1306
• 9 Stapleton J T. GB virus type C/hepatitis G virus.  Semin Liver Dis . 2003;  23 137-148
• 10 Roland M E, Carlson L, Ragni M. Solid organ transplantation in HIV-infected recipients: forty-seven cases in the HAART era. Presented at the XIV International AIDS Conference. July 8, 2002; Barcelona, Spain. Abstract MoOr B1060
• 11 Pallella F J, Delaney K M, Moorman A C. Declining morbidity and mortality among patients with advanced human immunodeficiency virus infection.  N Engl J Med . 1998;  338 853-860
• 12 Kaplowitz N, Lewis J H, Watkins P B. Did this drug cause my patient's hepatitis?.  Ann Intern Med . 2002;  136 480-483
• 13 Montessori V, Harris J, Montaner J SG. Nucleoside analogue liver toxicity.  Semin Liver Dis . 2003;  23 167-172
• 14 Lewis W, Dalakis M C. Mitochondrial toxicity of antiviral drugs.  Nat Med . 1995;  1 417-421
• 15 Delgado J, Harris M, Tesiorowski A, Montaner J SG. Symptomatic elevations of lactic acid and their response to treatment manipulation in HIV infected individuals-a case series.  Clin Infect Dis . 2001;  33 2072-2074
• 16 Sulkowski M S. Protease inhibitor toxicity.  Semin Liver Dis . 2003;  23 183-194
• 17 Carr A, Samaraas K, Burton S. A syndrome of peripheral lipodystrophy, hyperlipidemia and insulin resistance in patients receiving HIV protease inhibitors.  AIDS . 1998;  12 F51-58
• 18 Behrens G, Dejam A, Schmidt H. Impaired glucose tolerance, beta cell function and lipid metabolism in HIV patients under treatment with protease inhibitors.  AIDS . 1999;  13 F63-F70
• 19 Sutinen J, Hakkinen A, Westerbacka J. Increased fat accumulation in the liver in HIV-infected patients with antiretroviral therapy-associated lipodystrophy.  AIDS . 2002;  16 2183-2193
• 20 Brunt E M. NASH: definition and pathology.  Semin Liver Dis . 2001;  21 3-16
• 21 Gelato M C. Improved insulin sensitivity with body fat distribution in HIV-infected patients treated with rosiglitazone: a pilot study.  J Acquir Immune Defic Syndr . 2002;  31 163-170
• 22 Graham D J, Drinkard C R, Shatin D. Incidence of idiopathic acute liver failure and hospitalized liver injury in patients treated with troglitazone.  Am J Gastroenterol . 2003;  98 175-179
• 23 Schambelan M, Benson C, Carr A. Management of metabolic complications associated with antiretroviral therapy for HIV-1 infection: recommendations of an IAS-USA Panel.  J Acquir Immune Defic Syndr . 2002;  31 257-275
• 24 Cheeseman S H, Havlir D, McLauglin M M. Phase I/II evaluation of nevirapine alone and in combination with zidovudine for infection with human immunodeficiency virus.  J Acquir Immune Defic Syndr Hum Retrovirol . 1995;  8 141-145
• 25 D'Aquila R T, Hughes M D, Johnson V A. Nevirapine, ziodovudine, and didanosine compared with zidovudine and didanosine in patients with HIV-1 infection.  Ann Intern Med . 1996;  124 1019-1030
• 26 Riesler R, Servoss J, Sherman K. Incidence of hepatotoxicity and mortality in 21 adult antiretroviral treatment trials ACTG Liver Disease Focus Group. Presented at the IAS 2001. July 8, 2001; Buenos Aires, Argentina. Abstract 43
• 27 Centers for Disease Control and Prevention. Serious adverse events attributed to nevirapine regimens for post-exposure prophylaxis after HIV exposures-worldwide, 1997-2000.  MMWR Morb Mortal Wkly Rept . 2001;  49 1153-1156
• 28 Van der  Horst C. et al .FTC 302 results of a randomized trial. Presented at the Conference on Retroviruses and Opportunistic Infections. February 3, 2001; Chicago, IL. Abstract 18. 
• 29 Stern J O. Nevirapine hepatic safety project. Presented at the XIV International AIDS Conference. July 8, 2002; Barcelona, Spain
• 30 Yeni P G, Hammer S M, Carpenter C CJ. Antiretroviral treatment for adult HIV infection in 2002 updated recommendations of the IAS-USA panel.  JAMA . 2002;  288 222-235
• 31 van Leth F, Hassink E, Phanphk P. et al .Results of the 2NN study: a randomized comparative trial of first line antitretiviral therapy with regimens containing either nevirapine alone, efavirenz alone, or both drugs combined together with stavudine and lamivudine. Presented at the 10th Conference on Retroviruses and Opportunistic Infections. February 14, 2003; Boston. Abstract 176. 
• 32 Evans W E, McLeod H L. Pharmacogenomics drug disposition, drug targets and side effects.  N Engl J Med . 2003;  348 538-549