icon star paper   Articles  
Back grey_arrow_rt.gif
Does HAART harm the Liver? NNRTI-related or -unrelated hepatotoxicity?
  Hepatology Aug 2002 Vol 36 Numb 2
As you know a number of studies by HIV researchers have found HCV/HIV coinfected patients on HAART can experience elevated liver enzymes (AST/ALT) exascerbated by HCV or HBV infection. They call this hepatoxicity. In this letter in Hepatology researchers question whether or not this hepatoxicity actually translates into liver damage. They do not feel such increases translate into liver damage but say we cannot conclude such elevations equal liver damage and we better need studies to explore this.
Letter to the Editor:
We read with interest the article by Sulkowski et al. describing the incidence of severe hepatotoxicity related to antiretroviral therapies including efavirenz and nevirapine. The investigators conclude that increases in transaminase levels are more frequent in patients coinfected with hepatitis B virus (HBV) or hepatitis C virus and in those treated concurrently with protease inhibitors. Although these findings represent useful clinical information, several points warrant further discussion.
First, the toxicity grading scale used in the study to define severe hepatotoxicity is limited because the magnitude of the increase of serum transaminase levels has no prognostic value. The occurrence of liver failure, as assessed by impaired coagulation, defines severe hepatic injury and is associated with an unfavorable outcome. In the article by Sulkowski et al., no information is provided concerning the incidence of liver failure. Moreover, the use of a single alanine transaminase determination as a baseline for the definition of severe hepatotoxicity is inadequate because spontaneous fluctuations are frequent in patients with chronic viral hepatitis.
Second, the causal relationship between non-nucleoside analog reverse transcriptase inhibitor (NNRTI) therapy and hepatotoxicity is unclear from the current data. As conceded by the investigators, the increases in transaminase levels in these patients were likely multifactorial. Biochemical abnormalities in patients receiving highly active antiretroviral therapy (HAART) may also reflect hepatotoxicity caused by the hepatotropic viruses, fatty liver disease, or the effects of alcohol and other drugs; the precise diagnosis of biochemical liver abnormalities in this setting is very complex. Obviously, consensus criteria for the assessment of causality in cases of drug-induced liver damage cannot be used in this population because of a high prevalence of coinfection with hepatotropic viruses and concomitant therapy with several potentially hepatotoxic medications. In this setting, liver histology may substantially narrow the differential diagnosis. Unfortunately, in the study by Sulkowski et al., pathologic information was not reported. Moreover, the favorable outcome described in 40% of the patients continuing antiretroviral therapy despite severe hepatotoxicity is surprising, and argues against NNRTI as the cause of hypersensitivity or idiosyncratic reactions.
Third, in our opinion, an increased risk for NNRTI-related liver injury in patients with chronic viral hepatitis remains to be definitively shown. In the absence of an untreated control group, a change in transaminase levels from baseline is a suboptimal measure of hepatotoxicity in HBV- or hepatitis C virus-coinfected patients. Such changes may represent spontaneous fluctuations, or result from the clearance of infected hepatocytes owing to immune reconstitution or antiviral therapies. In patients with chronic hepatitis B, acute flares may indicate subsequent viral clearance4 and have been described in human immunodeficiency virus-HBV-coinfected patients receiving protease inhibitors and adefovir dipivoxil. However, no information was presented describing the use of therapies against HBV during or preceding the study, and the evolution of HBV serology was not reported.
Finally, though Sulkowski et al. emphasize the potential of NNRTIs to cause severe hepatic injury, this phenomenon is likely uncommon. Because HAART-associated immune restoration may have a positive impact on the course of liver disease in coinfected patients, the assessment of the risk-benefit ratio of NNRTIs, especially in this population, requires further study.
Pascal Lebray, M.D.1
Yves Benhamou, M.D., Ph.D.2
Stanislas Pol, M.D., Ph.D.1
Robert P. Myers, M.D.2
Thierry Poynard, M.D., Ph.D.2
Vincent Di Martino, M.D., Ph.D.2
1Unite d'Hepatologie et INSERM U370
H˘pital Necker-Enfants malades
Paris, France
2 Groupe Hospitalier
Paris, France
  icon paper stack View Older Articles   Back to Top   www.natap.org