Dolutegravir-induced liver injury leading to sub-acute liver failure requiring transplantation: a case report and review of literature

Case report

A 28-year-old Black African woman with a five-year history of human immunodeficiency virus (HIV)-1 infection presented with a three-month duration of progressive liver dysfunction, after starting a dolutegravir-containing antiretroviral therapy (ART) regimen. At diagnosis of HIV her nadir CD4 cell count was 303 cells/mm³. She also had serological evidence of resolved hepatitis B virus (HBV) infection, with positive anti-core and anti-surface antibody, but no other significant medical, social or travel history. She had previously been on an abacavir-containing regimen for more than six months after testing negative for HLA-B5701 and then was well controlled on Truvada (tenofovir disoproxil fumarate [TDF]/emtricitabine) and raltegravir. However, due to the patient’s desire to be on a single-pill ART regimen, she was switched to Triumeq (dolutegravir/abacavir/lamivudine). Having had a normal baseline liver enzyme profile, a mild abnormality in alanine aminotransferase (ALT) was first detected four weeks after starting Triumeq, which peaked at 303 IU/mL after ten weeks of therapy. She was switched back to Truvada and raltegravir, but despite this, liver enzymes continued to deteriorate with predominant transaminitis and ALT rising to 1636 IU/mL two weeks later. All ART was stopped, but the patient was admitted to hospital with nausea and jaundice; blood tests showed a total bilirubin of 124 µmol/L, alkaline phosphatase 131 IU/L and aspartate aminotransferase 1198 IU/L. International normalized ratio (INR) was also elevated at 2.14.

A systematic screen for potential causes was unremarkable; extended viral serological testing was negative, including HBV surface antigen and HBV DNA, hepatitis E IgM and RNA, hepatitis C total antibody and RNA and cytomegalovirus, Epstein–Barr virus and herpes simplex virus DNA. Her autoimmune profile was also unremarkable, with negative autoantibodies and non-specifically elevated serum immunoglobulins. Cross-sectional imaging with triphasic contrast-enhanced computed tomography showed normal liver architecture with patent porto-hepatic vasculature and no features of chronic liver disease. Histology obtained via trans-jugular liver biopsy showed hepatocellular necrosis and ongoing lobular activity, with portal lymphocytic infiltrates and scattered eosinophils, features consistent with drug induced liver injury (DILI) in the absence of serological markers of viral or autoimmune disease.

The patient was managed with n-acetylcysteine infusion at 6.25 mg/kg/h, alongside supportive care, including empirical broad-spectrum antibiotics and antifungals. Despite this, she continued to deteriorate clinically with the development of hepatic encephalopathy on day 14 of admission, with total bilirubin rising to 354 µmol/L and INR 4.82. She fulfilled UK listing criteria for super-urgent liver transplantation for sub-acute liver failure, after multi-disciplinary team review and confirmation that HIV infection was well-controlled (CD4 cell count >700 cells/mm³ and HIV-1 RNA <40 cps/mL). Three days after listing, the patient was transplanted with a whole cadaveric liver donated after brain death. The explanted liver showed histological features compatible with the clinical diagnosis of DILI (Figures 1 and 2). In addition, there were granulomas containing Schistosoma ova which were incidental and not considered relevant to the clinical presentation (Figure 2(b)). ¹

OPEN IN VIEWER

Figure 1.

(a) Representative macroscopic section of the explanted liver, with gallbladder. Most of the parenchyma is collapsed (dark areas), with scant map-like areas of residual parenchyma (yellowish areas). The inset (c) shows an approximation of a map-like area and the interface with the collapsed liver. (b) Area of residual parenchyma with vague nodular appearance (H&E). The nodules are composed of regenerative hepatocytes. The areas in between represent confluent and bridging necrosis around central hepatic venules (arrows). Portal areas are within the regenerative nodules (arrowheads). (d) Representative histologic findings at the collapsed areas. CT: central hepatic venule; PT:portal tract.

OPEN IN VIEWER

Figure 2.

(a) Portal tracts contain mild lymphocytic inflammatory component with occasional plasma cells. Ongoing activity is present at the parenchyma with hepatocellular disarray, some hepatocyte ballooning, multiple necroinflammatory foci and apoptotic bodies (H&E). (b) Portal granuloma with the remains of Schistosoma ova (arrowhead), surrounded by macrophages, occasional multinucleated cells, lymphocytes and numerous eosinophils (arrows). (c) Portal areas do not show evidence of established fibrosis, as represented in this image (Sirius Red). A portal area is seen at the centre of a regenerative nodule. Two central hepatic venules are seen at the left side of the picture surrounded by collapsed parenchyma and absence of residual hepatocytes. (d) Same image with reticulin stain. Irregular periportal liver plates (arrows) and bridging collapse of the reticulin framework at the perivenular areas (arrowheads). CT: central hepatic venule; PT:portal tract; PV: portal vein.

The patient had a standard post-operative course with a single episode of moderate rejection, which was managed with augmentation of the immunosuppression regimen to include tacrolimus and mycophenolate mofetil. She was restarted on Truvada and raltegravir and successfully discharged 14 days post-transplant. She remains well one year post-transplantation, with normal liver graft function and undetectable HIV RNA.

Discussion

The efficacy and safety of dolutegravir has been established in several phase III clinical trials. ² The SINGLE study compared Triumeq (dolutegravir/abacavir/lamivudine) and Atripla (TDF/emtricitabine/efavirenz) and demonstrated a low risk of hepatotoxicity with significantly fewer patients in the Triumeq arm, developing elevated ALT requiring drug discontinuation. ³ There were no reports of severe hepatotoxicity. Only 16% of the participants, however, were women as is often the case in ART clinical trials. This potentially leads to under-reporting of adverse events that are more prevalent in women, such as DILI. The ARIA study, specifically comparing Triumeq in women, has shown good 48-week safety data and more studies of this nature are needed. ⁴

DILI is a clinico-pathological diagnosis with a range of presentations, from asymptomatic liver enzyme elevation to acute liver failure. Idiosyncratic DILI can result from many pharmacological agents, but is commonly seen with ART; up to 18% of patients treated with ART experience some form of DILI. ⁵ Risk factors for idiosyncratic DILI include age, gender and host genetic factors. Women have a higher risk of DILI and a more aggressive hepatitis with progression to acute liver failure. ⁶ US transplant data showed that African-American women accounted for 78% of all transplants performed for idiosyncratic DILI. ⁷ Genetic associations with DILI have also been identified; HLA-B*5701 carriage is associated with abacavir hypersensitivity with a specificity of 100% and also confers an 80-fold risk increase of flucloxacillin-related liver injury.^8,9 The risk of ART-related DILI is also influenced by underlying liver disease, with chronic HBV and HCV co-infection being important factors. ¹⁰

The diagnosis of DILI relies on a temporal relationship between a causative drug and the onset of liver injury; however, this latency period can vary from a few days to several months. Exclusion of other causes of liver injury is important, including serological and radiological investigation. Although there are no hallmark histological features, liver biopsy remains an important investigation to define the pattern of liver injury and to assess for an underlying chronic hepatopathy. Prompt diagnosis and withdrawal of the offending drug is paramount but in significant liver dysfunction, n-acetylcysteine can be considered; it has a clear benefit in paracetamol overdose but has been shown to improve transplant-free survival in non-paracetamol causes of acute liver failure. ¹¹ Ultimately, cases that progress to acute or subacute liver failure may require liver transplantation. A robust multi-disciplinary review and exclusion of contra-indicating factors are important to ensure that the patient achieves an optimal outcome.