 |
|
|
| |
Study Finds Increase in AST After starting ART Increases Risk For death More Than Having HCV
|
| |
| |
AIDS 2002;16:1357-1362. July 5 2002.
Authors: Corinne Rancinana; Didier Neaub; Marianne Savèsa; Sylvie Lawson-Ayayia; Fabrice Bonnetc; Patrick Merciéd; Michel Duponb; Patrice Couzigoue; François Dabisa; Geneviève Chênea; and the Groupe d'Epidémiologie Clinique du SIDA en Aquitaine (GECSA)
The results of this study suggest HAART may contribute to accelerated liver disease more than HCV. Although these study investigators found HCV was not associated with poorer survival, this is not the same as asking if HIV accelerates HCV and follow-up in this study appears limited. Perhaps both HAART and HCV accelerate HCV progression in coinfected, but the effect of HAART has not been well studied. In addition, limitations upon the author¹s conclusion based on limited study data are discussed below by the authors. They say that they are unsure about the duration of time patients in this study had HCV and they suggest that given more time of HCV duration having HCV could emerge as a factor in poorer survival. The authors also suggest that toxicity or elevated liver enzymes could be due to elevated HIV drug blood levels because of liver impairment in processing HIV drugs. And they suggest that monitoring HIV drug blood levels IN HCV-infected patients should be considered.
STUDY OBJECTIVE: To study whether hepatitis C virus (HCV) co-infection or the severe elevation of transaminases is associated with survival after the initiation of antiretroviral combination therapy.
Design: Prospective hospital-based cohort (Aquitaine Cohort)
Methods: HIV-infected adults started on an antiretroviral combination before 30 June 1999. HCV infection was defined as antibody detection or positive HCV RNA. Severe elevation of transaminases was defined as a value of aspartate or alanine aminotransferase (AST, ALT) above five times the upper limit of normal values. Survival was studied using a Cox model, including at least baseline HCV status and transaminases as a time-dependent covariate.
Results: Overall, 995 patients were analysed, including 576 HCV-positive individuals (58%). At baseline, HCV-positive patients were younger, more often injecting drug users and women, and had more frequently elevated transaminases. A shorter survival was associated with AIDS stage [hazard ratio (HR) versus non-AIDS 1.67; 95% confidence interval (CI) 1.03; 2.68], lower CD4 cell count (HR for 50 cells/mm3 lower 1.33; CI 1.17; 1.51), lower haemoglobin (HR for 1 g/dl lower 1.20; CI 1.07; 1.35), lower platelet count (HR for 10 000 cells/mm3 lower 1.04; CI 1.01; 1.07), and AST during follow-up (HR for 200 IU/l 2.30; CI 1.32; 4.03). HCV co-infection (HR 1.20; CI 0.75; 1.92) was not statistically associated with survival.
Conclusion: The occurrence of a severe elevation of transaminases was associated with poorer survival, although HCV was not. If liver toxicity may be treatment induced, plasma drug concentrations could guide dosage adjustments of antiretroviral treatments currently prescribed to optimize their use.
Discussion by authors. In this study of HIV-infected patients treated with antiretroviral agents, patients who experienced an increase of AST above 200 IU/l after treatment initiation had a higher risk of death than other patients, whereas HCV co-infection in itself was not associated with a poorer prognosis. As in previous studies on the survival of HIV-infected patients [15], we found that a lower CD4 cell count, the presence of an AIDS-defining event, low haemoglobin and platelets levels were still associated with a poorer outcome.
Most antiretroviral agents are potentially hepatotoxic. The incidence of the elevation of transaminases in patients under an antiretroviral regimen ranges from 3 to 10% within the first 6 months. However, the comparison between studies may be difficult because the definition used for the elevation of transaminases varies: ALT measure above 200 IU/l or five times the upper limit of normal value or transaminases measure above five times the upper limit of normal taking into account a variation if transaminases were elevated at study entry. In our analysis, the elevation of transaminases, one year after treatment initiation, tended to be more frequent in patients initially treated by dual NRTI (approximately 5%), than in patients treated by combination with PI (approximately 2%). Previous studies showed a high incidence of the elevation of transaminases in patients treated with PI-containing regimens, but did not in patients treated with an NRTI regimen. However, our analysis only took into account the initial antiretroviral treatment, and an elevation of transaminases might have occurred after the intensification of an initial dual NRTI regimen.
In HIV-infected patients treated with an antiretroviral combination, one of the most important risk factors of severe hepatic toxicity is co-infection by HCV or hepatitis B virus. In our study, the status for HCV remained unknown for 806 patients, 45% of the initial cohort. Because the baseline characteristics of these latter patients were similar to those of HCV-negative patients (data not shown), we believe that patients with unknown status for HCV had negative antibodies, yielding a minimum prevalence of 32% (576/1801) for co-infection in our cohort, compared with a range of 25-52% in recent
studies. Moreover, we also found a relationship between HCV status and AST or ALT levels, consistent with previous reports.
Whether HCV co-infection has a deleterious effect on the progression of HIV infection remains controversial. Some studies set up before the era of antiretroviral combinations showed that HCV-positive patients had a more rapid evolution to AIDS or death than patients not co-infected; although another study did not show a difference. Results were not adjusted for other covariates known to be related to survival, except in one recent study.
Different mechanisms in the era of antiretroviral combinations could be involved for a differential survival according to HCV status or the elevation of transaminases. First, several years are necessary to observe consequences on morbidity and mortality related to HCV. Under an antiretroviral combination, life- threatening complications caused by HCV may occur before the progression of HIV. We believe that patients included in our cohort may still not have a long enough duration of HCV infection, although it was not possible to document precisely the time since patients acquired HCV infection. As co-infected patients survived until they were treated by antiretroviral agents, they could have been selected among less frail patients. Second, a treatment-induced severe increase in hepatic transaminase levels may directly lead to the discontinuation of an efficacious antiretroviral therapy. Nevertheless, in our study, the dates of cessation of treatments could not be documented precisely enough to test this hypothesis. Third, an elevation of liver enzymes may be a consequence of immune restoration occurring under highly active antiretroviral therapy.
Conclusion by Authors
A co-infection in patients treated with antiretroviral treatments should seriously be considered in the decisions for subsequent case management. If liver dysfunction is treatment-induced, plasma drug concentrations could guide dosage adjustments of antiretroviral treatments currently prescribed to optimize their use. In the future, the development of new antiretroviral
treatments inducing less liver toxicity should be a priority.
Results
Description of the study sample
Overall, 2776 adult patients from the Aquitaine Cohort started a combination of antiretroviral treatment before 30 June 1999. We compared 576 HCV-positive and 419 HCV-negative eligible patients, among whom 761 started treatment on dual NRTI (76%) and 234 (24%) a combination including PI.
HCV-positive patients were more frequently women than HCV-negative patients, had been infected by HIV more frequently through intravenous drug use, were younger and had higher liver enzymes at treatment initiation. The most frequent dual NRTI were zidovudine-zalcitabine (42%) and zidovudine-didanosine (38%). The most frequent triple combinations were
stavudine-lamivudine-indinavir (29%), zidovudine-lamivudine-indinavir (19%), zidovudine-lamivudine-nelfinavir (8%), stavudine-lamivudine-nelfinavir (7%) and stavudine-didanosine-nelfinavir (5%).
Survival analysis
During follow-up, 13% of HCV-positive patients (n = 78) and 10% of HCV-negative patients (n = 40) died. In HCV-positive patients, the main cause of death was directly related to hepatic pathology in eight cases: hepatic failure (n = 5) or cirrhosis (n =3), or to haemorrhage in two cases; it was HIV or AIDS-related in 31 cases, and related to an overdose in five cases or a
suicide in four cases. The causes of death in HCV-negative patients were related to hepatic failure in two cases or to hepatocarcinoma in one case; the other main causes were HIV or AIDS-related (n = 15), or cardiorespiratory failure (n = 12).
After 36 months, survival probability was 89.7% [95% confidence interval (CI) 86.8; 92.6] in HCV-positive patients and 92.0% (95% CI 88.7; 95.4) in HCV-negative patients (P = 0.09).
In the multivariate analysis, a higher risk of death was associated with AIDS, lower CD4 cell count, lower haemoglobin and platelet count. It was also associated with an increase of AST during follow-up. Survival was neither significantly associated with HCV co-infection [hazard ratio (HR) 1.20, 95% CI 0.75; 1.92], with the more recent period of March 1997 or later (HR 1.96; 95% CI 0.60; 6.37), nor with treatment intensification during follow-up (HR 0.68; 95% CI 0.39; 1.16). There was no significant interaction between HCV co-infection and the date of antiretroviral treatment initiation
(before and after 1st March 1997) (P = 0.68). In a sensitivity analysis, we considered the elevation of transaminases in terms of a mean rise since baseline. A higher risk of death was associated with a mean rise of AST 12 months after the initiation of an antiretroviral combination (HR for 50 IU/l higher 1.26, 95% CI 1.03; 1.54). Survival was neither significantly associated with HCV co-infection (HR 1.32, 95% CI 0.71; 2.45), with the most recent period of March 1997 or later, nor with treatment intensification.
|
|
| |
|
|
|
|
|
