Question | Discussion | References | CME Credit

Case 6: Discussion


Isolated unconjugated hyperbilirubinemia is the most common laboratory abnormality associated with the use of the protease inhibitor atazanavir (Reyataz)[1]. The development of isolated hyperbilirubinemia can also occur with the protease inhibitor indinavir (Crixivan)[2], although less frequently than with atazanavir. With both drugs, this abnormality reverses with discontinuation and is not associated with hepatocellular injury. Although not considered a serious adverse effect, higher levels of unconjugated hyperbilirubinemia associated with these drugs can manifest as jaundice and thus affect the patient's quality of life, and potentially lead to unnecessary laboratory investigation and treatment interruption. Considering the widespread use of atazanavir, clinicians caring for HIV-infected patients should have familiarity with the entity of protease inhibitor-associated hyperbilirubinemia.

Incidence and Clinical Features

In early clinical trials involving indinavir[3,4], investigators observed an elevation in serum bilirubin levels (both total and indirect) in 6 to 25% patients, with an estimated 14% of patients developing grade 3 to 4 elevations in total bilirubin (greater than 2.5 mg/dL or greater than 2.5 times the upper limit of normal [ULN]). In trials involving atazanavir, grade 3 to 4 hyperbilirubinemia occurred in 33 to 41% of patients receiving the standard unboosted atazanavir dose of 400 mg daily[5,6] and in 40 to 49% of those on a regimen that contained 300 mg of atazanavir boosted with 100 mg of ritonavir (Norvir)[7]. In the atazanavir trials, however, only 7 to 11% patients developed clinical jaundice or scleral icterus[5,6] Moreover, study participants rarely (less than 1%) discontinued therapy because of this side effect. In an observational cohort of 218 patients prescribed atazanavir, nearly 70% had mild subclinical hyperbilirubinemia (less than 1.4 x ULN) when tested 1 to 4 months after initiation of therapy, but few patients had increases greater than 5 x ULN (Figure 1)[8]. Hyperbilirubinemia associated with atazanavir or indinavir appears to be dose-dependent, with some studies suggesting increased incidence and severity of hyperbilirubinemia with higher plasma levels in the context of pharmacologic boosting or drug interactions[9,10,11]. One study reported a significant increase in the development of grade 3 or 4 hyperbilirubinemia in patients taking atazanavir after they initiated therapy with pegylated interferon and ribavirin for the treatment for chronic hepatitis C virus infection; this change was directly related to ribavirin-induced hemolysis[12]. The use of protease inhibitors other than atazanavir or indinavir does not significantly alter bilirubin levels (Figure 2)[13]. In addition, the development of isolated hyperbilirubinemia does not occur with drugs from other antiretroviral classes and should not be confused with jaundice that may result from severe hepatotoxicity caused by nevirapine (Viramune).

Timing and Natural History

The onset of atazanavir- and indinavir-associated hyperbilirubinemia typically occurs within several months, and bilirubin levels generally peak within 4 months (range 1 to 8 months); the subsequent natural history on therapy is notable for a non-progressive course, with bilirubin levels remaining generally stable in patients on further follow-up[8,14]. In one case series, the estimated average increase in bilirubin level was 0.87 mg/dL for all patients on an atazanavir-based regimen and 0.46 mg/dL for those on indinavir[13].

Mechanism of Hyperbilirubinemia and Genetic Predisposition

In humans, glucuronidation of bilirubin is performed primarily by UDP-glucuronosyltransferase (UGT) enzymes, primarily the hepatic and gastrointestinal enzyme UGT1A1[15,16]. Both atazanavir and indinavir cause hyperbilirubinemia by competitively inhibiting the UGT1A1 enzyme[15,16] Persons with Gilbert's syndrome, an inherited condition characterized by transient hyperbilirubinemia, have a mutation in the promoter region of the gene that encodes the UGT1A1 enzyme and this mutation leads to decreased production of the UGT1A1 enzyme; the alleles bearing this variant promoter are termed UGT1A1*28 and this genetic polymorphism is present in 3-10% of the general population[17]. As might be expected, the atazanavir and indinavir-associated hyperbilirubinemia is more pronounced in individuals with Gilbert's syndrome, since these protease inhibitors inhibit the same enzyme that is quantitatively reduced in persons with Gilbert's syndrome. Indeed, one study of the Swiss HIV cohort reported persons homozygous for the UGT1A1*28 allele were more likely to develop jaundice and severe hyperbilirubinemia when exposed to atazanavir or indinavir (Figure 3)[13]. Other variants in the UGT1A gene locus have also been associated with atazanavir-related hyperbilirubinemia[18] In addition, genetic variability in the activity of efflux pump systems that affect plasma and intracellular concentrations of atazanavir may affect the severity of hyperbilirubinemia[19].

Prevention and Patient Education

At this time, there are no formal recommendations regarding routine genetic screening before use of atazanavir- or indinavir-associated hyperbilirubinemia. The cost-effectiveness and utility of such genetic testing remains a subject of debate[20]. In general, clinicians should inform all patients who plan to initiate either atazanavir or indinavir about the possibility of developing hyperbilirubinemia and possibly jaundice, particularly since the development of jaundice can be alarming to patients. On the other hand, patients should be reassured that hyperbilirubinemia and jaundice in this setting are not associated with liver damage and will reverse after discontinuing the antiretroviral medication. Patients coinfected with chronic hepatitis B or C do not appear to have increased risk of developing hyperbilirubinemia with administration of these medications. The use of atazanavir or indinavir should be avoided if possible in patients with known Gilbert's syndrome, as these individuals are more likely to develop overt jaundice. In addition, clinicians should avoid the concomitant use of atazanavir and indinavir because of potential additive effects and risk of severe hyperbilirubinemia[21] A baseline chemistry panel that includes a serum bilirubin level should be performed on all patients prior to initiating either atazanavir or indinavir. Patients with previously undiagnosed Gilbert's, or underlying liver disease, may be identified in this manner.


For patients who start on an antiretroviral regimen that includes atazanavir or indinavir, a follow-up chemistry panel is generally recommended at 3 months after starting therapy and every 6 months thereafter. An isolated elevation in total bilirubin should be confirmed as predominantly unconjugated by testing the indirect fraction of bilirubin. Additional work-up is unnecessary if liver enzymes are otherwise consistent with baseline values. The presence of elevated conjugated bilirubin or changes in serum hepatic aminotransferases or alkaline phosphatase warrants further investigation for other causes of hyperbilirubinemia, such as drug hepatotoxicity, viral hepatitis, or cholestasis. It is important to recognize that patients with acute hemolysis will also develop increased indirect bilirubin levels. Indinavir (along with commonly prescribed medications such as dapsone or ribavirin) can cause a hemolytic anemia; in these cases, a complete blood count should be performed to screen for anemia. Other findings consistent with acute hemolysis include reticulocytosis, increased lactate dehydrogenase (LDH) levels, and decreased haptoglobin levels.


For patients who develop clinically-evident jaundice, the decision of whether to discontinue the offending protease inhibitor usually depends on how severe and noticeable the jaundice is, and whether the patient is willing to tolerate it. Although discontinuation of ritonavir (with appropriate dose adjustment of atazanavir or indinavir) may lead to resolution of jaundice in some patients, it is not appropriate in patients with protease inhibitor resistance or in patients who are taking atazanavir with efavirenz (Sustiva), tenofovir (Viread), or any co-formulated product that contains either or both of these drugs. Dose reduction of these protease inhibitors is not recommended in this setting. In most cases, a switch to an alternative regimen is necessary for patients who develop an unacceptable level of jaundice.

[Back to Case 6 Question | See Case 6 References]

CME Credit | Back to Top

  • The following link will open in a new window.
    Figure 1. Incidence of Hyperbilirubinemia among Patients Taking Atazanavir
      in the HIV Outpatient Study (HOPS)

    From: Ward D, Curtin J, Richardson J, et al. Hyperbilirubinemia among patients in the HIV Outpatient Study (HOPS) receiving atazanavir. In: The 3rd IAS Conference on HIV Pathogenesis and Treatment. Rio de Janeiro, Brazil, 2005. Abstract TuPe2.1B02. <

    Figure 1
  • The following link will open in a new window.
    Figure 2. Relationship of Protease-Inhibitor Component of Antiretroviral
      Therapy and Total Bilirubin Levels

    This graph is derived from a multivariate analysis of the contribution of protease-inhibitor components in antiretroviral therapy on patient total bilirubin levels (adjusted for age, sex, chronic hepatitis B or C, longitudinal relationship between hemoglobin and bilirubin level, and presence of UGT1A1 allelic variants associated with Gilbert's syndrome). The p value for atazanavir and indinavir was <0.0001.

    From: Rotger M, Taffe P, Bleiber G, et al. Gilbert syndrome and the development of antiretroviral therapy-associated hyperbilirubinemia. J Infect Dis. 2005;192:1381-6.

    Figure 2
  • The following link will open in a new window.
    Figure 3. Relationship of <em>UGT1A1</em> Allelic Variants on Total Bilirubin Levels

    Abbreviations: PI = protease inhibitor; LPV = lopinavir; NFV = nelfinavir; RTV = ritonavir; ATV = atazanavir; IDV = indinavir

    This graph shows the proportion of patients who had jaundice (defined as a total bilirubin level greater than 2.5 mg/dl [43 umol/L] on at least 2 occasions), by presence and type of protease inhibitor (PI) and UGT1A1 allelic variant. This study involved 96 participants and 1386 measurements of total bilirubin levels during a median follow-up of 6 years.

    From: Rotger M, Taffe P, Bleiber G, et al. Gilbert syndrome and the development of antiretroviral therapy-associated hyperbilirubinemia. J Infect Dis. 2005;192:1381-6.

    Figure 3