Antiretrovirals and Liver Toxicity: How Big a Concern?
The benefits of combination therapy that have led to dramatic decreases in opportunistic infections and AIDS deaths over the past several years are accompanied by drawbacks, not the least of which is drug side effects. Some of these side effects are hepatotoxic -- the scary sounding term for something that can harm the liver.
Even when their health is failing, many people put off treatment, fearing that antiretrovirals will automatically result in irreversible liver damage. Most anti-HIV drugs are processed primarily by the liver, so the fear isn't irrational. But every drug is different, each class of antiretrovirals works differently, and the condition of each person's liver can differ depending on factors such as co-infection with hepatitis B or C, previous or current alcohol use, age, sex, genetics -- lots of things.
There are still many unknowns about the effects of antiretrovirals on the liver, but there's a lot we do know. If combination therapy makes sense for you clinically and you're worried about your liver, understand your personal risk factors, know what symptoms to watch out for, and -- most importantly -- carefully monitor liver function while on therapy to help you avoid the possibility of liver damage but still experience the benefits of treatment.
Many studies have looked at the prevalence and causes of liver toxicity in people on combination therapy. Results show that all three classes of antiretrovirals used to treat HIV infection are associated with different kinds and degrees of hepatotoxicity:
In most of the studies discussed below, severe liver toxicity is defined as an elevation of the liver enzymes ALT (alanine aminotransferase) and/or AST (aspartate aminotransferase) to levels greater than 5 times the upper limit of normal. Although useful as a way to classify reactions to medications, it's important to realize that liver enzyme levels don't necessarily correlate with the degree of liver cell damage. Enzyme levels can be normal in people with severe liver disease and abnormal in people with diseases that don't affect the liver.
The side effects of nucleoside analogues are fairly well known because this class of drugs has been around the longest. Recently, more attention has been paid to their effect on the liver. Numerous studies have implicated the NRTIs as damaging mitochondria, the "powerhouses" inside cells that convert nutrients into energy. Mitochondrial damage can affect many parts of the body, including muscle, nerves, the heart, and the liver. A disruption in mitochondrial energy supply can also result in an increase in lactic acid levels in the blood. If lactate levels get very high, a rare condition called lactic acidosis occurs, which is often accompanied by liver abnormalities -- elevated enzymes, fatty liver (hepatic steatosis) and acute inflammation.
For someone currently on or thinking about beginning therapy, one question to consider is which NRTIs are more likely to cause high lactate levels and, possibly, liver toxicity. Researchers from the University of California at San Diego looked at 2,144 patients who were taking NRTIs from July 1998 through September 2001 to see how often symptoms that could lead to lactic acidosis (symptomatic hyperlactatemia) occurred and to identify which NRTI combinations were most often responsible. Symptomatic hyperlactatemia was defined as having at least one symptom -- fatigue, nausea, vomiting, abdominal pain, loss of appetite, difficulty breathing or elevated ALTs -- and high lactate levels in the blood.
Eighty-one patients (only 4%) met the definition, and most of them were on combinations that included two NRTIs. The risk of developing symptomatic hyperlactatemia doubled with each additional NRTI used in a regimen. A substantially higher risk was seen in people whose regimens included Zerit (d4T) plus Videx (ddI) or Zerit plus Ziagen (abacavir), particularly compared to the relatively low risk for people taking combinations that included AZT plus Epivir (3TC). These results reinforce concern that Zerit -- especially when used in combination with Videx -- could lead to mitochondrial damage and potential liver injury.
Gilead Sciences, the manufacturer of the nucleotide analogue, Viread (tenofovir), conducted a test tube (in vitro) study to evaluate the effect of Viread and each available nucleoside analogue on the mitochondrial DNA of liver cells, muscle cells, and kidney cells. Hivid (ddC) caused the most damage, followed by Videx, Zerit, and, to a lesser degree, AZT, Epivir, Ziagen, and Viread. These results are generally consistent with the toxicities seen in human studies, but it's important to remember that what happens in the test tube may not reflect what happens in the body.
Many small studies and case reports suggest that mitochondrial damage is more likely to occur the longer you're on treatment, sometimes within several months, but often after a number of years. The relationship between lactic acidosis and liver damage is unclear. In some cases, liver damage seemingly caused by mitochondrial toxicity is not accompanied by lactic acidosis. The liver rarely lets us know when it's in trouble, and the symptoms of lactic acidosis are notoriously non-specific. Blood tests, including those that check lactic acid and bicarbonate levels, can help confirm the presence of lactic acidosis, and regular liver function tests are particularly important if you're on treatment.
Among the non-nucleoside reverse transcriptase inhibitors (NNRTIs) -- Rescriptor (delavirdine), Sustiva (efavirenz) and Viramune (nevirapine) -- Sustiva and, particularly, Viramune have been shown to cause liver damage in some people.
Viramune received a lot of unflattering attention early last year when a South African study was stopped prematurely at week 48 because of two deaths caused by liver failure. The trial, called FTC-302, was comparing FTC (emtricitabine), an investigational NRTI, to Epivir in combinations that included Zerit and either Viramune or Sustiva. Both deaths occurred in women who were on Viramune. Of the 468 people enrolled (59% were women), severe liver toxicity occurred in 17% of those taking Viramune and in none of those taking Sustiva. Only one-third of the people with liver toxicity also had the rash that commonly occurs with NNRTI drugs. One of the women who died had experienced a rash but wasn't advised to stop her medications until almost two weeks had passed. Other than being on Viramune, the only significant predictive factor for developing severe liver toxicity in this study was female sex. Women were twice as likely as men to experience liver toxicity. Three-quarters of the cases occurred within the first month on therapy, and the rest occurred later in the course of treatment.
A study published in AIDS in July of last year reported on 610 individuals who started combinations that included Viramune between September 1997 and May of 2000 in Barcelona and London. Almost half of the participants were co-infected with hepatitis C. In this study, liver damage was defined as an increase in ALT or AST levels at least three times higher than the patient's measurement before beginning treatment. This occurred in 76 participants (12.5%), but only thirteen (2%) had to stop Viramune because of the liver toxicity. Seven of the thirteen experienced symptoms of clinical hepatitis, which improved once they switched to a different regimen. The risk of liver damage increased with longer time on therapy, and only half of the cases of clinical hepatitis occurred during the first two months on treatment. The researchers concluded that people who had been on any antiretroviral therapy for longer periods of time, were co-infected with hepatitis C, or had elevated ALT levels before starting therapy were at higher risk of developing liver damage while on Viramune and, perhaps, on any other three-drug regimen.
Liver enzyme elevations or liver inflammation occur in as many as one out of four people who take Viramune, but it only rarely causes clinical hepatitis. If liver problems develop, they usually happen during the first three months on the drug, although at least one-third of people who experience liver problems develop them later.
Researchers in Madrid looked at plasma levels of Viramune in 33 people who developed elevated liver enzymes. They found significantly higher drug levels in this group compared to those whose enzymes weren't elevated, suggesting a correlation between higher plasma drug concentrations and liver toxicity. This might help explain the higher rates of liver toxicity in women compared to men. Weight, body mass and hormones affect drug metabolism. It's possible that differences in these areas could lead to higher drug levels in women.
A study published in Hepatology in January compared the occurrence of severe liver toxicity in 568 people on regimens that included either Viramune or Sustiva. Almost half of the study participants (43%) were co-infected with hepatitis C, and a much smaller 7.7% had hepatitis B. Severe liver toxicity occurred in almost 16% of those on Viramune and 8% of those on Sustiva. Again, most of the liver toxicities were detected after the participants had been on treatment for more than three months. Liver toxicity was much more likely to occur in people co-infected with hepatitis B or C and whose regimens included a protease inhibitor. Some good news -- although hepatitis C co-infection was a significant risk factor, the majority of co-infected participants (84%) did not experience severe liver damage.
Researchers from the San Francisco Community Health Network looked back at the effect of antiretrovirals on the livers of co-infected individuals who were seen at their clinics between mid-1996 and mid-2000. The findings were presented at the 9th Conference on Retroviruses and Opportunistic Infections in February. Of more than 3,000 patients whose results were available, 39% were hepatitis C co-infected and 9% had chronic hepatitis B. The researchers found that Viramune was the only antiretroviral associated with increased liver enzymes in individuals co-infected with viral hepatitis, particularly hepatitis B. Interestingly, people taking a protease inhibitor had decreased liver enzyme levels overall. Sadly, the researchers also found that patients with hepatitis C were prescribed antiretrovirals less often than those who weren't co-infected, even when CD4 count, viral load, liver enzyme levels and other factors between the two groups were similar.
Some studies contradict findings of NNRTI liver damage. One of the most interesting, published in JAIDS in April 2002, comes from the New York University School of Medicine. Researchers looked for liver toxicity in 272 patients -- almost all men -- who were on combinations that included an NNRTI. Only three patients experienced severe liver toxicity. There were no significant changes in liver enzymes in patients taking Viramune compared to those taking Sustiva or Rescriptor. Although co-infection rates were relatively low in this group (12% had hepatitis C and 9% had hepatitis B), co-infection was not associated with a significant increase in liver enzymes.
Much of these data might tempt you to steer clear of this class of drugs altogether, but the NNRTIs, particularly Viramune and Sustiva, are extremely useful to many people as part of combinations that successfully lower viral load, keep CD4 cells climbing, and help the immune system regain its strength. If you begin Viramune or Sustiva, have your liver enzyme levels checked before starting the drug and monitor liver function regularly. If symptoms of liver problems develop (fatigue, lack of appetite, weakness or nausea), call your doctor right away. The drug may need to be stopped and not restarted after symptoms clear.
Researchers at John Hopkins University sought to determine which antiretroviral drugs posed the greatest risk of liver toxicity and to establish the role of chronic hepatitis B or C infection in the development of liver toxicity. As part of their study, they followed 211 people who received protease inhibitor (PI)-based combinations between January 1996 and January 1998. Over half of the patients were co-infected with hepatitis C. Less than 3% had chronic hepatitis B. Severe liver toxicity occurred in about 10% of the participants. The highest incidence was observed in people on Norvir (ritonavir), which was associated with 48% of all cases of severe hepatotoxicity. There were no significant differences in liver damage incidence in the other treatment groups. Again, most co-infected participants (88%) didn't experience severe liver damage, which led the authors to conclude that, although co-infection is certainly a risk factor, PIs should not be withheld from people who are co-infected.
The higher likelihood of developing liver damage while on Norvir was confirmed in a study published in JAIDS in January. The cleverly named LIVERHAART Group determined the frequency of liver toxicity in 1,325 people in Italy, more than half of whom were co-infected with chronic hepatitis B, C or both. After six months on PI-based regimens, 11% (147 patients) experienced some degree of hepatotoxicity. The liver damage was mild in almost 8% of the patients and severe in just over 3%. The percentages were about the same even after one and two years on treatment. Only thirty patients (2.3%) had to stop treatment because of liver toxicity.
All protease inhibitors used in the study increased liver enzymes, but not in everyone and not all to the same degree. In descending order, the following PIs or dual-PI combinations caused severe liver damage: Norvir 11.7%, Norvir plus Fortovase (saquinavir) 11.6 %, Fortovase 3.7%, Crixivan (indinavir) 1.3%, Viracept (nelfinavir) 0%, and Norvir plus Crixivan 0% (only five patients were on this combination). Liver damage occurred most often in people co-infected with viral hepatitis. Of particular interest, people with hepatitis C who didn't respond to antiretroviral therapy after a year on treatment had the highest rates of liver toxicity. The LIVERHAART Group's data suggest that Norvir is more hepatotoxic than other PIs only during the first six months on treatment.
Results of other recent studies somewhat contradict the common belief that PIs cause liver damage. Researchers from Ottawa looked back at 66 HIV/HCV co-infected patients who were on single or dual-PI regimens and found that Norvir was as well tolerated by the liver as any other PI, whether used as the single PI or as one of two PIs in a combination. Eight of the 66 patients (12%) stopped treatment because of hepatotoxicity, but the rates were the same whether the patients were on Norvir or not. Another study of 692 HIV-positive individuals in Thailand who were on a variety of regimens showed that liver damage was rare in people who were on PI-based regimens, even when Norvir was used.
To get a sense of how often antiretroviral-related liver toxicity occurs and which drugs or combinations may be most responsible, researchers from the National Institutes of Health looked back at the data on 10,611 participants in twenty-one adult AIDS Clinical Trials Group (ACTG) studies from 1991 to 2000. Because of the years covered, the various drugs and treatment strategies ran the gamut from NRTI monotherapy to combinations that included an NNRTI, a PI and two NRTIs. Overall, 662 (6.2%) of the patients experienced severe liver damage. Varying and somewhat surprising rates of liver damage occurred on all of the treatment regimens: 7.4% in people taking just one NRTI; 4.9% on dual-NRTI therapy; 8.2% on NNRTI-based triple therapy, and 5% on PI-based triple therapy. Of the people who experienced liver toxicity, just over one-quarter stopped their medications. Forty-two of the 10,611 trial participants died of liver failure, although whether these deaths were directly related to antiretroviral use or other factors such as viral hepatitis is unclear.
These and other studies identify many factors that may increase the risk of antiretroviral liver toxicity: obesity, preexisting liver disease such as hepatitis B or C, age greater than 50, alcohol use, diabetes, and interactions between antiretrovirals, herbs, and over-the-counter and other prescription drugs. Several theories, possibly interrelated, have been proposed to explain the connections between antiretrovirals and liver problems:
Numerous studies have shown varying rates of liver toxicity associated with antiretrovirals -- individually, by class, and in combination. Exactly how antiretrovirals affect the liver still isn't fully understood. More data are needed from controlled clinical trials to clarify which individual drugs and risk factors can contribute to liver toxicity in people with HIV. How often the liver damage is irreversible is also unclear, although most people who stop the drug that's causing the problem see their liver enzymes return to normal levels.
There's no doubt that combination therapy has prolonged the lives of thousands of people with HIV. In a small but significant number of people, the drugs used can also cause liver damage. Learning as much as you can, monitoring liver function regularly, being aware of potential symptoms, and, possibly, managing metabolic or other viral conditions before starting HIV treatment may be the best ways to avoid antiretroviral-associated liver problems.
Bertrand Toulouse is a Treatment Advocate at AIDS Project Los Angeles and has been involved in HIV treatment issues for six years.
James Learned is ACRIA's Treatment Education Director and editor of ACRIA Update.
This article was provided by AIDS Community Research Initiative of America. It is a part of the publication ACRIA Update. Visit ACRIA's website to find out more about their activities, publications and services.