Abstract

Introduction
The morbidity and mortality of people living with HIV (PLWH) has decreased markedly after the introduction of highly effective antiretroviral therapy (ART). The safety profile of drugs used for HIV has improved over time, and nowadays, co-formulated combinations with excellent tolerability are available. One of the most commonly used is elvitegravir/cobicistat/emtricitabine/tenofovir alafenamide (EVG/COBI/FTC/TAF) due to its effectiveness in viral suppression and good tolerability [1-2].
PLWH are at greater risk of suffering hepatobiliary complications, not only caused by opportunistic infections or AIDS cholangiopathy related to advanced immunosuppression but also due to drug-induced hepatotoxicity, use of alcohol or other substances, coinfections with hepatitis virus or liver steatosis [3].
This article describes a case of a patient who developed a severe increase in cholestasis parameters after starting EVG/COBI/FTC/TAF regimen, probably related to integrase strand transfer inhibitor (InSTI). Liver enzyme elevation during InSTI-based therapy is an unusual finding. To our knowledge, this is the first report in the literature of EVG/COBI/FTC/TAF-associated cholestasis. The article tries to generate knowledge in order to strengthen pharmacovigilance of this adverse effect and reviews the available data on cholestasis in PLWH.
Case report
A 57-year-old man with a history of injecting drug use diagnosed of HIV infection in 2004 and coinfected with hepatitis virus B (HBV) and C (HCV) with chronic stage F0–F1 liver disease was followed at our institution. The patient started ART based on efavirenz/emtricitabine/tenofovir disoproxil fumarate (EFA/FTC/TDF), achieving adequate immunovirological control of HIV infection and HBV (CD4 count of 483 cells/mm3 (23%), undetectable viral load of HIV and HBV DNA). He received HCV treatment with interferon/ribavirin in 2008, without showing viral response.
In October 2016, in order to avoid interactions in the context of possible new treatment for HCV infection with direct acting antivirals, EFA/FTC/TDF was replaced by EVG/COBI/FTC/TAF. Three months later, a marked dissociated cholestasis was diagnosed based on alkaline phosphatase (ALP) and gamma-glutamyltransferase (GGT) levels more than four times the upper limit of normal (ALP 452U/L, GGT 2519 U/L), although bilirubin values were normal (total bilirubin 0.6 mg/dL and direct bilirubin 0.2 mg/dL) and the patient was asymptomatic (Figure 1). A moderate increase in aspartate transaminase (AST) and alanine transaminase (ALT) was also detected with maximum values of 261U/L and 220U/L, respectively. The changes in liver function tests induced by EVG/COBI/FTC/TAF. ALP: alkaline phosphatase; ALT: alanine transaminase; AST: aspartate transaminase; COBI: cobicistat; EFA: efavirenz; FTC: emtricitabine; GGT: gamma-glutamyltransferase; HBV: hepatitis virus B; HCV: hepatitis virus C; RPV: rilpivirine; TAF: tenofovir alafenamide; TDF: tenofovir disoproxil fumarate; VL: viral load.
The study of cholestasis was initially focused on the evaluation of the bile duct, so an abdominal ultrasound was requested which showed intra- and extrahepatic dilatation of the bile duct. Computed tomography confirmed the dilatation and reported a doubtful image suggestive of ampulloma in papilla. To obtain anatomopathological samples, an ultrasound-guided fine needle puncture aspiration was performed, with no evidence of malignancy. Moreover, other cholestasis aetiologies such as toxicities (no antirretroviral drugs, supplements, herbals or over-the-counter medications), acute infections and autoimmune diseases were ruled out.
After two years of follow-up, the clinical, analytical and radiological abnormalities remained stable. The wide diagnostic study did not reveal lesions. Finally, taking into account the temporality sequence of the analytical alterations with the drug exposure, a possible relationship between the patient’s cholestasis and treatment with EVG/COBI/FTC/TAF was suspected in spite of the absence of related literature. Therefore, it was decided to discontinue EVG/COBI and to modify ART to rilpivirine/FTC/TAF. In the next control, two months later, the blood test showed a marked improvement in liver profile until resolution of cholestasis (ALP 108U/L, GGT 80U/L, total bilirubin 0.4 mg/dL, AST 41U/L and ALT 47U/L); the values were similar to patient’s baseline parameters. Currently, the patient persists with mild dilatation of the common bile duct, without analytical alterations.
Discussion
Hepatobiliary diseases are common in PLWV requiring extensive study to reach the diagnosis. The evaluation consists of a detailed clinical history and wide diagnostic study with blood test, cultures, imaging and biopsy, if necessary.
Even though it was not the case of our patient, the initial target in advanced immunosuppressed PLWH with hepatobiliary manifestations is to rule out acquired immune deficiency syndrome (AIDS)–related cholangiopathy [4]. Although the aetiology of this entity is not clear, suspicion is mainly focused on opportunistic infections (OIs), such as Mycobacterium avium complex, cytomegalovirus, cryptococcosis or histoplasmosis. In those cases, abdominal pain, nauseas and diarrhoea are common findings during anamnesis, and cultures and diagnostic imaging are necessary for a definitive diagnosis.
Hepatic involvement due to malignancy should be suspected in PLWH since AIDS-related (NHL [5] and Kaposi sarcoma) and non–AIDS-defining cancers sometimes debut as liver damage. In our patient, the symptoms were not suggestive of malignancy and imaging studies ruled out neoplastic aetiology.
PLWH are at higher risk than the general population for suffering liver steatosis due to diverse host and viral factors and also probably related to some antiretroviral drugs [6]. In the presented case, the abdominal ultrasound did not reveal signs suggestive of steatosis, such as hyperechoic and bright liver parenchyma.
Hepatotropic viruses generally cause cholangitic lesions by infiltrating intraepithelial lymphocytes and generating lymphoid aggregates, not directly affecting the duct [7]. The damage is usually reversible without structural destruction. Among hepatotropic viruses, HBV and especially HCV are mainly responsible, although hepatitis A and E can also lead to cholangitis [8]. In our patient, chronic stage F0–F1 liver disease related to coinfection with HBV and HCV was stable with undetectable HBV DNA and without elevation of HCV RNA at the time of cholestasis.
Autoimmune hepatitis is a chronic progressive liver parenchymal inflammation caused by regulatory T-cell dysfunction. It is a relatively rare cause of hepatitis among PLWH. The diagnosis is based on clinical and serological features which are confirmed with liver biopsy. Simplified diagnostic scoring system of autoimmune hepatitis collects most of the diagnostic criteria [9]. Blood tests characteristically detect hypergammaglobulinemia with elevated antinuclear, anti-smooth muscle and anti-liver kidney microsome antibodies. In the case described, autoimmune hepatitis was ruled out because no significant elevation of those antibodies was detected in laboratory tests.
Some antiretroviral drugs can lead to liver toxicity [10]. Among protease inhibitors, the use of indinavir, saquinavir, nelfinavir and full-dose ritonavir can produce hepatotoxicity. The majority of nucleoside analogue reverse transcriptase inhibitors can induce mitochondrial damage. However, abacavir, lamivudine and tenofovir seem to have a safer profile [10]. Liver toxicity related to nonnucleoside reverse transcriptase inhibitors has also been described, especially associated with nevirapine. InSTI-based regimens rarely lead to liver enzyme elevation. In fact, some researchers have observed protective liver effect while taking these drugs, even in HBV-/HCV-coinfected PLWH, concluding that InSTI could be a good option in patients with elevated liver parameters [11,12]. To ascribe the adverse effect of the drug, other causes must be ruled out and liver damage must coincide with the introduction of the treatment. The temporality sequence of the cholestasis with the EVG/COBI/FTC/TAF exposure and the resolution after drug discontinuation led us to think that the adverse effect could be pharmacological.
The most common adverse effects of EVG/COBI/FTC/TAF in 5–10% of patients include nausea, headache, diarrhoea, insomnia and fatigue [13]. Although transient and asymptomatic, grade 3–4 elevation of AST and ALT has been observed in patients with underlying liver disease [14], to the best of our knowledge, the presence of cholestasis has not been reported as a drug-related event in pivotal clinical trials, cohort studies or studies reporting real life [15-17].
Management of this condition should consider transient discontinuation of the drug and replacement with another safe and effective regimen to avoid invasive diagnostic studies like liver biopsy.
There are different questionnaires to help clinicians estimate the adverse drug reaction probability. The one we calculated is called Naranjo scale [18] and a score of 3 was obtained. According to the result, EVG/COBI/FTC/TAF could have been the possible cause of cholestasis. However, it was not possible to answer all the items of the scale because EVG/COBI/FTC/TAF was never reintroduced as equally effective therapeutic alternatives were available.
Intra- and extrahepatic dilatation persisted after liver enzyme normalization. However, no structural lesions or moderate fibrosis was detected in subsequent imaging tests. The absence of findings led us to think that bile dilatation could correspond to previous lithiasis in a coinfected HIV patient.
In conclusion, clinicians should be aware of EVG/COBI/FTC/TAF-induced cholestasis, which is not currently included in the drug data sheet [19] and not described in the literature [20] as this may trigger time-consuming, expensive and invasive procedures to rule out other causes. We encourage pharmacovigilance of this adverse effect in order to detect possible future cholestasis cases related to EVG/COBI/FTC/TAF.
Footnotes
Acknowledgements
We thank the colleagues from HIV Unit of Hospital Clinic of Barcelona.
Author contributions
Ainoa Ugarte is the main author of the case report; the other members are collaborator authors.
Informed consent
The patient gave us his oral informed consent to publish his case anonymously.
