|
|
|
|
Biomarkers Predict Hepatic Flares and Death in Hepatitis-Coinfected People Starting ART
|
|
|
6th IAS Conference on HIV Pathogenesis, Treatment and Prevention, July 17-20, 2011, Rome
Mark Mascolini
High levels of inflammatory, coagulation, and fibrosis markers before starting antiretroviral therapy (ART) predicted hepatic flares in the months after treatment began and/or death within 4 years, according to an analysis of 333 people enrolled in the FIRST trial of antiretroviral initiation [1].
Although coinfection with HIV and hepatitis B or C virus (HBV, HCV) is common, little is known about predictors of hepatic flare soon after ART begins or predictors of death in coinfected people starting ART. Because biomarkers may pinpoint people at high risk of dangerous flares and poor survival, FIRST investigators [2] undertook this retrospective analysis of three types of markers in pre-ART samples from trial participants: (1) inflammatory markers (C-reactive protein, pro- and anti-inflammatory cytokines and chemokines), a coagulation marker (D-dimer), and a fibrosis marker (hyaluronic acid).
The primary goal was to see if these markers could predict hepatic flare within 4 months of starting ART or death between 1 month and 4 years of starting. The researchers defined hepatic flare as an alanine aminotransferase (ALT) level above 100 IU/mL at month 1 or 4 plus an ALT increase of more than 50 IU/mL from before treatment.
The 333-person study group had a median age of 41 years (interquartile range [IQR] 36 to 47). Fifty-eight participants (17.4%) were women, 194 (58.3%) were black, 163 (48.9%) had injected drugs, and 144 (43.2%) had an AIDS diagnosis. Median pretreatment values were 150 for CD4 count, about 125,000 copies for HIV load, 46 IU/L (IQR 30 to 68) for ALT, and 127 mg/dL (IQR 97 to 193) for triglycerides. Seventy people had HBV coinfection, 253 had HCV, and 10 had both.
The researchers counted hepatic flares in 53 people (15.9%), 22 at month 1, 38 at month 4, and 7 at both points. Flares affected 12.3% with HCV, 24.3% with HBV, and 50% with HCV and HBV (P < 0.001). Pretreatment ALT stood at a median 54 IU/L (IQR 38 to 88) in people who had flares, compared with 45 IU/L (IQR 28 to 67) in those who did not, but this difference was not statistically significant (P = 0.895), so ALT alone cannot predict hepatic flare.
Multivariate analysis that considered age, gender, lamivudine in the first regimen, and pretreatment AIDS, ALT, triglycerides, CD4 count, and HIV load identified three markers that independently raised the risk of hepatic flare, at the following adjusted odds ratios (AOR) (and 95% confidence intervals [CI]):
--Hyaluronic acid above 49 ng/mL: AOR 2.01 (1.00 to 4.00), P = 0.05
--Every IQR higher interleukin 13 level: AOR 1.66 (1.08 to 2.56), P = 0.02
--Every IQR higher interleukin 10 level: AOR 1.71 (1.25 to 2.33), P < 0.001
High pretreatment HBV DNA levels also predicted flares in people with HBV coinfection (HBV DNA above 10(8) copies/mL, AOR 13.87, 95% CI 2.10 to 95.56, P = 0.051).
Fifty-two coinfected people (15.6%) died after a median follow-up of 25.8 months, compared with 9.4% of study participants who did not have HBV or HCV (P = 0.002). Fifteen deaths could be attributed to sepsis or bacterial infection, 10 to AIDS, 4 to liver failure, 4 to cardiovascular disease, 4 to kidney failure, and the rest to other causes. Prevalence of pretreatment AIDS was significantly higher among people who died (57.7% versus 40.6%, P < 0.022), as was pretreatment triglyceride level (median 161 versus 122 mg/dL, P = 0.008).
Multivariate analysis accounting for the same variables described above identified four pre-ART predictors of death, at the following AORs (and 95% CIs):
--Every IQR higher interleukin 8 level: AOR 1.69 (1.12 to 2.55), P = 0.01
--Every IQR higher interleukin 6 level: AOR 2.15 (1.43 to 3.24), P < 0.001
--Every IQR higher TNF-alpha level: AOR 1.45 (1.05 to 1.99), P = 0.02
--Every IQR higher D-dimer levels: AOR 1.57 (1.03 to 2.39), P = 0.04
For people with HBV coinfection alone, six pretreatment markers independently predicted death: IL-6, IL-8, D-dimer, interferon gamma-induced protein 10 (IP-10), interferon-inducible T-cell alpha chemoattractant (ITAC), and monokine-induced by gamma interferon (MIG). For people with HCV alone, there were four independent mortality predictors: IL-6, IL-8, TNF-alpha, and hyaluronic acid.
The FIRST team concluded that "biomarkers of inflammation, fibrosis and coagulation pre-ART are associated with mortality in this patient population." Measuring these markers before ART begins in coinfected people could help determine who needs closer follow-up to prevent flares and early mortality.
References
1. Andrade B, Huppler Hullsieak K, Boulware D, et al. Biomarkers of inflammation, coagulation and liver fibrosis are associated with hepatic flares and death in HIV hepatitis co-infected persons. 6th IAS Conference on HIV Pathogenesis, Treatment and Prevention. July 17-20, 2011. Rome. Abstract WELBX01.
2. MacArthur RD, Chen L, Peng G, et al. Efficacy and safety of abacavir plus lamivudine versus didanosine plus stavudine when combined with a protease inhibitor, a nonnucleoside reverse transcriptase inhibitor, or both in HIV-1 positive antiretroviral-naive persons. HIV Clin Trials. 2004;5:361-370.
|
|
|
|
|
|
|