HIV virological rebounds but not blips predict liver fibrosis progression in antiretroviral-treated HIV/hepatitis C virus-coinfected patients

Cocaethylene, simultaneous alcohol and cocaine use, and liver fibrosis in people living with and without HIV

BACKGROUND

The simultaneous consumption of cocaine and alcohol results in the production of cocaethylene (CE) in the liver, a highly toxic metabolite. Prior research suggests that cocaine use contributes to liver disease and its concomitant use with alcohol may increase its hepatotoxicity, but studies in humans are lacking. We evaluated the role of cocaine, its simultaneous use with alcohol, and CE on liver fibrosis.

METHODS

We performed a cross-sectional analysis of the Miami Adult Studies on HIV (MASH) cohort. Cocaine use was determined via self-report, urine screen, and blood metabolites, using liquid chromatography with tandem mass spectrometry. Hazardous drinking was determined with the AUDIT-C and liver fibrosis with the Fibrosis-4 Index (FIB-4).

RESULTS

Out of 649 participants included in this analysis, 281 (43.3%) used cocaine; of those, 78 (27.8%) had CE in blood. Cocaine users with CE had higher concentrations of cocaine metabolites in blood and were more likely to drink hazardously than cocaine users without CE and cocaine non-users. Overall, cocaine use was associated with liver fibrosis. CE in blood was associated with 3.17 (95% CI: 1.61, 6.23; p = 0.0008) times the odds of liver fibrosis compared to cocaine non-users, adjusting for covariates including HIV and HCV infection. The effect of CE on liver fibrosis was significantly greater than that of cocaine or alcohol alone.

CONCLUSIONS

CE is a reliable marker of simultaneous use of cocaine and alcohol that may help identify individuals at risk of liver disease and aid in the prevention of its development or progression.

Risk of HIV viral rebound in HIV infected patients on direct acting antivirals (DAAs) treatment for HCV

Background

The dynamic of HIV-viral load (VL) remains poorly investigated in HIV/HCV patients under direct acting antivirals (DAAs).

Methods

We retrospectively evaluated HIV-VL at baseline (BL) during and up to 24 weeks post-DAAs in a cohort of 305 HIV-1/HCV patients, on ART and with no HIV virological failure (VF) in the 6 months before treatment with DAAs; during the period of observation VF was defined as confirmed VL≥50 copies/mL; virological blips (VB, transient, not confirmed, VL ≥50 copies/mL). Stepwise Cox regression models were fitted to estimate adjusted hazard ratios (aHR) of VF.

Results

Fifteen VF occurred in 13 patients over 187 person-years of follow-up (PYFU): incidence rate (IR) of 8.0 per 100-PYFU (95% CI = 4.0–12.1); 29 VBs were detected in 26 patients over 184 PYFU: IR = 15.8 per 100-PYFU (95% CI = 10.0–21.5). The most prominent factor associated with VF was the presence of BL HIV residual viremia (RV = HIV-RNA detectable but not precisely quantifiable) [aHR = 12.26 (95% CI = 3.74–40.17), P<0.0001]. Other factors were ≥1 VBs in the 6 months before DAAs [aHR = 6.95 (95% CI = 1.77–27.37) P = 0.006] number of ART regimens failed before DAAs initiation [aHR (per more regimen) = 1.22 (95% CI = 1.04–1.42), P = 0.012] and age [aHR (per year older) = 1.16 (95% CI = 1.04–1.29), P = 0.010].

Conclusions

Our findings underline the importance for close monitoring HIV-VL in selected patients. Whether this phenomenon is triggered by the rapid clearance of HCV remains to be established.

Clinical and demographic predictors of antiretroviral efficacy in HIV-HBV co-infected patients

BACKGROUND

The clinical and demographic characteristics that predict antiretroviral efficacy among patients co-infected with HIV and hepatitis B virus (HBV) remain poorly defined. We evaluated HIV virological suppression and rebound in a cohort of HIV-HBV co-infected patients initiated on antiretroviral therapy.

METHODS

A retrospective cohort analysis was performed with Canadian Observation Cohort Collaboration data. Cox proportional hazards models were used to determine the factors associated with time to virological suppression and time to virological rebound.

RESULTS

HBV status was available for 2,419 participants. A total of 8% were HBV co-infected, of whom 95% achieved virological suppression. After virological suppression, 29% of HIV-HBV co-infected participants experienced HIV virological rebound. HBV co-infection itself did not predict virological suppression or rebound risk. The rate of virological suppression was lower among patients with a history of injection drug use or baseline CD4 cell counts of <199 cells per cubic millimetre. Low baseline HIV RNA and men-who-have-sex-with-men status were significantly associated with a higher rate of virological suppression. Injection drug use and non-White race predicted viral rebound.

CONCLUSIONS

HBV co-infected HIV patients achieve similar antiretroviral outcomes as those living with HIV mono-infection. Equitable treatment outcomes may be approached by targeting resources to key subpopulations living with HIV-HBV co-infection.

Liver macrophage-associated inflammation correlates with SIV burden and is substantially reduced following cART

Liver disease is a leading contributor to morbidity and mortality during HIV infection, despite the use of combination antiretroviral therapy (cART). The precise mechanisms of liver disease during HIV infection are poorly understood partially due to the difficulty in obtaining human liver samples as well as the presence of confounding factors (e.g. hepatitis co-infection, alcohol use). Utilizing the simian immunodeficiency virus (SIV) macaque model, a controlled study was conducted to evaluate the factors associated with liver inflammation and the impact of cART. We observed an increase in hepatic macrophages during untreated SIV infection that was associated with a number of inflammatory and fibrosis mediators (TNFα, CCL3, TGFβ). Moreover, an upregulation in the macrophage chemoattractant factor CCL2 was detected in the livers of SIV-infected macaques that coincided with an increase in the number of activated CD16+ monocyte/macrophages and T cells expressing the cognate receptor CCR2. Expression of Mac387 on monocyte/macrophages further indicated that these cells recently migrated to the liver. The hepatic macrophage and T cell levels strongly correlated with liver SIV DNA levels, and were not associated with the levels of 16S bacterial DNA. Utilizing in situ hybridization, SIV-infected cells were found primarily within portal triads, and were identified as T cells. Microarray analysis identified a strong antiviral transcriptomic signature in the liver during SIV infection. In contrast, macaques treated with cART exhibited lower levels of liver macrophages and had a substantial, but not complete, reduction in their inflammatory profile. In addition, residual SIV DNA and bacteria 16S DNA were detected in the livers during cART, implicating the liver as a site on-going immune activation during antiretroviral therapy. These findings provide mechanistic insights regarding how SIV infection promotes liver inflammation through macrophage recruitment, with implications for in HIV-infected individuals.

Chronic hepatitis C infection and liver disease in HIV-coinfected patients in Asia

Data on markers of hepatitis C virus (HCV) disease in HIV-HCV-coinfected patients in resource-limited settings are scarce. We assessed HCV RNA, HCV genotype (GT), IL28B GT and liver fibrosis (FibroScan® ) in 480 HIV-infected patients with positive HCV antibody in four HIV treatment centres in South-East Asia. We enrolled 165 (34.4%) patients in Jakarta, 158 (32.9%) in Bangkok, 110 (22.9%) in Hanoi and 47 (9.8%) in Kuala Lumpur. Overall, 426 (88.8%) were male, the median (IQR) age was 38.1 (34.7-42.5) years, 365 (76.0%) reported HCV exposure through injecting drug use, and 453 (94.4%) were on combination antiretroviral therapy. The median (IQR) CD4 count was 446 (325-614) cells/mm3 and 208 (94.1%) of 221 patients tested had HIV-1 RNA <400 copies/mL. A total of 412 (85.8%) had detectable HCV RNA, at a median (IQR) of 6.2 (5.4-6.6) log10 IU/mL. Among 380 patients with HCV GT, 223 (58.7%) had GT1, 97 (25.5%) had GT3, 43 (11.3%) had GT6, eight (2.1%) had GT4, two (0.5%) had GT2, and seven (1.8%) had indeterminate GT. Of 222 patients with IL28B testing, 189 (85.1%) had rs12979860 CC genotype, and 199 (89.6%) had rs8099917 TT genotype. Of 380 patients with FibroScan® , 143 (37.6%) had no/mild liver fibrosis (F0-F1), 83 (21.8%) had moderate fibrosis (F2), 74 (19.5%) had severe fibrosis (F3), and 79 (20.8%) had cirrhosis (F4). One patient (0.3%) had FibroScan® failure. In conclusion, a high proportion of HIV-HCV-coinfected patients had chronic HCV infection. HCV GT1 was predominant, and 62% of patients had liver disease warranting prompt treatment (≥F2).

HCV co-infection and markers of liver injury and fibrosis among HIV-positive childbearing women in Ukraine: results from a cohort study

Background

Ukraine’s injecting drug use-driven HIV epidemic is among the most severe in Europe with high burden of HCV co-infection. HIV/HCV co-infected individuals are at elevated risk of HCV-related morbidity, but little is known about burden of liver disease and associated factors in the HIV-positive population in Ukraine, particularly among women.

Methods

Characteristics of 2050 HIV-positive women enrolled into the Ukrainian Study of HIV-infected Childbearing Women were described by HCV serostatus. Aspartate transaminase (AST) to platelet ratio (APRI) and FIB-4 scores were calculated and exact logistic regression models fitted to investigate factors associated with significant fibrosis (APRI >1.5) among 762 women with an APRI score available.

Results

Of 2050 HIV-positive women (median age 27.7 years, IQR 24.6-31.3), 33% were HCV co-infected (79% of those with a history of injecting drug use vs 23% without) and 17% HBsAg positive. A quarter were on antiretroviral therapy at postnatal cohort enrolment. 1% of the HIV/HCV co-infected group had ever received treatment for HCV. Overall, 24% had an alanine aminotransferase level >41 U/L and 34% an elevated AST (53% and 61% among HIV/HCV co-infected). Prevalence of significant fibrosis was 4.5%; 2.5% among 445 HIV mono-infected and 12.3% among 171 HIV/HCV co-infected women. 1.2% had a FIB-4 score >3.25 indicating advanced fibrosis. HCV RNA testing in a sub-group of 56 HIV/HCV co-infected women indicated a likely spontaneous clearance rate of 18% and predominance of HCV genotype 1, with one-third having genotype 3 infection. Factors associated with significant fibrosis were HCV co-infection (AOR 2.53 95%CI 1.03-6.23), history of injecting drug use (AOR 3.51 95%CI 1.39-8.89), WHO stage 3-4 HIV disease (AOR 3.47 95%CI 1.51-7.99 vs stage 1-2 HIV disease) and not being on combination antiretroviral therapy (AOR 3.08 95%CI 1.23-7.74), adjusted additionally for HBV co-infection, smoking and age.

Conclusions

Most HIV/HCV co-infected women had elevated liver enzymes and 12% had significant fibrosis according to APRI. Risk factors for liver fibrosis in this young HIV-positive population include poorly controlled HIV and high burden of HCV. Results highlight the importance of addressing modifiable risk factors and rolling out HCV treatment to improve the health outcomes of this group.

Statin drugs decrease progression to cirrhosis in HIV/hepatitis C virus coinfected individuals

INTRODUCTION

Chronic HIV/hepatitis C virus (HCV) coinfection carries increased risk of cirrhosis, hepatocellular carcinoma, and death. Due to anti-inflammatory properties, 3-hydroxy-3methylglutaryl coenzyme A (HMG-CoA) inhibitors (statins) may be useful adjunctive therapy to reduce liver disease progression.

METHODS

Clinical information was extracted from the Veterans Affairs HIV and HCV Clinical Case Registries (1999-2010). HIV-related variables included combination antiretroviral therapy era of diagnosis, CD4 cell count, and percentage time with undetectable HIV viral load. Metabolic variables included diabetes, low high-density lipoprotein (HDL), and hypertension. Statin use was measured as percentage time with active prescription (time-updated throughout the follow-up period). Cox proportional hazards analysis was used to determine risk factors for cirrhosis (International Classification of Diseases-9 or aminotransferase-to-platelet ratio index >2) overall and in groups stratified by alanine aminotransferase (ALT) level above and below 40 IU/l.

RESULTS

The cohort included 5985 HIV/HCV coinfected veterans. The majority was black race, and the mean age at index date was 45 years. Statin use was significantly protective of cirrhosis for patients with ALT 40 IU/l or less; for every 30% increase in time on statin, there was a 32% decreased risk of developing cirrhosis (hazard ratio 0.68, 95% confidence interval 0.47-0.98). Diabetes and low HDL were significantly associated with cirrhosis in patients with ALT greater than 40 IU/l (hazard ratio 1.15, P < 0.04 and hazard ratio 1.3, P < 0.0001).

CONCLUSION

Statin drug use is beneficial in mitigating the risk of liver disease progression for HIV/HCV coinfected patients without advanced liver disease. Low HDL and diabetes in coinfected patients with abnormal ALT have greater risk of cirrhosis development.

Correcting for Measurement Error in Time-Varying Covariates in Marginal Structural Models

Unbiased estimation of causal parameters from marginal structural models (MSMs) requires a fundamental assumption of no unmeasured confounding. Unfortunately, the time-varying covariates used to obtain inverse probability weights are often error-prone. Although substantial measurement error in important confounders is known to undermine control of confounders in conventional unweighted regression models, this issue has received comparatively limited attention in the MSM literature. Here we propose a novel application of the simulation-extrapolation (SIMEX) procedure to address measurement error in time-varying covariates, and we compare 2 approaches. The direct approach to SIMEX-based correction targets outcome model parameters, while the indirect approach corrects the weights estimated using the exposure model. We assess the performance of the proposed methods in simulations under different clinically plausible assumptions. The simulations demonstrate that measurement errors in time-dependent covariates may induce substantial bias in MSM estimators of causal effects of time-varying exposures, and that both proposed SIMEX approaches yield practically unbiased estimates in scenarios featuring low-to-moderate degrees of error. We illustrate the proposed approach in a simple analysis of the relationship between sustained virological response and liver fibrosis progression among persons infected with hepatitis C virus, while accounting for measurement error in γ-glutamyltransferase, using data collected in the Canadian Co-infection Cohort Study from 2003 to 2014.

Long-term Outcomes After Liver Transplantation Among Human Immunodeficiency Virus-Infected Recipients

BACKGROUND

Early outcomes after human immunodeficiency virus (HIV) + liver transplantation (LT) are encouraging, but data are lacking regarding long-term outcomes and comparisons with matched HIV- patients.

METHODS

We examined outcomes among 180 HIV+ LT, and compared outcomes to matched HIV- counterfactuals (Scientific Registry of Transplant Recipients 2002-2011). Iterative expanding radius matching (1:10) on recipient age, race, body mass index, hepatitis C virus (HCV), model for end-stage liver disease score, and acute rejection; and donor age and race, cold ischemia time, and year of transplant. Patient survival and graft survival were estimated using Kaplan-Meier methodology and compared using log-rank and Cox proportional hazards. Subgroup analyses were performed by transplant era (early: 2002-2007 vs. modern: 2008-2011) and HCV infection status.

RESULTS

Compared to matched HIV- controls, HIV+ LT recipients had a 1.68-fold increased risk for death (adjusted hazard ratio [aHR], 1.68, 95% confidence interval [95% CI], 1.28-2.20; P < 0.001), and a 1.70-fold increased risk for graft loss (aHR, 1.70; 95% CI, 1.31-2.20; P < 0.001). These differences persisted independent of HCV infection status. However, in the modern transplant era risk for death (aHR, 1.11; 95% CI, 0.52-2.35; P = 0.79) and graft loss (aHR, 0.89; 95% CI, 0.42-1.88; P = 0.77) were similar between monoinfected and uninfected LT recipients. In contrast, independent of transplant era, coinfected LT recipients had increased risk for death (aHR, 2.24; 95% CI, 1.43-3.53; P < 0.001) and graft loss (aHR, 2.07; 95% CI, 1.33-3.22; P = 0.001) compared to HCV+ alone LT recipients.

CONCLUSIONS

These results suggest that outcomes among monoinfected HIV+ LT recipients have improved over time. However, outcomes among HIV+ LT recipients coinfected with HCV remain concerning and motivate future survival benefit studies.

Liver Fibrosis Regression Measured by Transient Elastography in Human Immunodeficiency Virus (HIV)-Hepatitis B Virus (HBV)-Coinfected Individuals on Long-Term HBV-Active Combination Antiretroviral Therapy

Transient elastography (TE) data in HIV-HBV co-infection are lacking. The majority of this cohort had mild-moderate fibrosis, however over 28% of those with >1 TE showed liver fibrosis regression and the prevalence of advanced fibrosis (≥F3) decreased 12.3% (32.7 to 20.4%) over a median 31 months

Background. Advanced fibrosis occurs more commonly in human immunodeficiency virus (HIV)-hepatitis B virus (HBV) coinfected individuals; therefore, fibrosis monitoring is important in this population. However, transient elastography (TE) data in HIV-HBV coinfection are lacking. We aimed to assess liver fibrosis using TE in a cross-sectional study of HIV-HBV coinfected individuals receiving combination HBV-active (lamivudine and/or tenofovir/tenofovir-emtricitabine) antiretroviral therapy, identify factors associated with advanced fibrosis, and examine change in fibrosis in those with >1 TE assessment.

Methods. We assessed liver fibrosis in 70 HIV-HBV coinfected individuals on HBV-active combination antiretroviral therapy (cART). Change in fibrosis over time was examined in a subset with more than 1 TE result (n = 49). Clinical and laboratory variables at the time of the first TE were collected, and associations with advanced fibrosis (≥F3, Metavir scoring system) and fibrosis regression (of least 1 stage) were examined.

Results. The majority of the cohort (64%) had mild to moderate fibrosis at the time of the first TE, and we identified alanine transaminase, platelets, and detectable HIV ribonucleic acid as associated with advanced liver fibrosis. Alanine transaminase and platelets remained independently advanced in multivariate modeling. More than 28% of those with >1 TE subsequently showed liver fibrosis regression, and higher baseline HBV deoxyribonucleic acid was associated with regression. Prevalence of advanced fibrosis (≥F3) decreased 12.3% (32.7%–20.4%) over a median of 31 months.

Conclusions. The observed fibrosis regression in this group supports the beneficial effects of cART on liver stiffness. It would be important to study a larger group of individuals with more advanced fibrosis to more definitively assess factors associated with liver fibrosis regression.