| |
Impaired Virologic Response to Highly Active Antiretroviral Therapy Associated With Ongoing Injection Drug Use
|
| |
| |
In the current issue of JAIDS (Journal of Acquired Immune Deficiency Syndromes 2003; 32(5):522-526) researchers from Vancouver including Julio Montaner reported on adherence to HIV therapy and current IVDUs.
HIV-infected injection drug users who reported that they continued to inject drugs while on HAART were less likely to suppress their HIV-1 RNA to <500 copies/mL on two consecutive determinations when compared with non-drug users. In contrast, former injection drug users were not less likely to achieve HIV-1 RNA suppression compared with non-drug users after adjusting for adherence and other prognostic factors. Furthermore, among current injection drug users, the use of pharmacy-based refill compliance was not a reliable measure of adherence because it was not significantly associated with HIV-1 RNA suppression.
Adherence to antiretroviral therapy was measured using pharmacy refill compliance data. In brief, we calculated as the ratio the number of days patients received antiretroviral therapy refills relative to the number of days of follow-up in the first year.
Injection drug users who continue to use drugs may not respond to highly active antiretroviral therapy (HAART) as well as other HIV-infected individuals, even after adjusting for a reliable measure of adherence. We therefore compared the virologic response among participants in a population-based HIV/AIDS Drug Treatment Program in British Columbia, Canada, by injection drug use activity. Participants who were HIV infected and naive to antiretroviral therapy and who
were prescribed antiretroviral treatment between August 1996 and December 2000 were eligible for this study. They were classified as current, former, or non-injection drug users. The main outcome was having two consecutive HIV-1
RNA levels less than 500 copies/mL. We used logistic regression to adjust for baseline HIV-1 RNA, type of antiretroviral regimen (2 nucleosides + nonnucleoside reverse transcriptase inhibitor versus 2 nucleosides + protease
inhibitor), duration of therapy (months), adherence (derived from refill compliance data), and age. A total of 578 participants were first prescribed HAART during the study period. Among them, 78 (13%) were current injection drug users, 96 (17%) were former injection drug users, and 404 (70%) never injected drugs. In the multivariable logistic regression, relative to non-drug users, current injection drug users were significantly less likely to suppress their HIV-1 RNA (adjusted OR [AOR] = 0.30, 95% CI: 0.13-0.67), and former injection drug users were not significantly different from non-drug users (AOR = 0.56, 95% CI: 0.24-1.34). There was a significant interaction between drug use and adherence. In the analyses stratified by drug use, the adherence of former and non-drug users was positively associated with HIV-1 RNA suppression (AOR = 1.33, 95% CI: 1.14-1.55), whereas for current drug users, it was not (AOR =
1.07, 95% CI: 0.87-1.33). Current injection drug users were less likely to achieve HIV-1 RNA suppression compared with non-drug users. Adherence as measured by pharmacy refill compliance was not a reliable measure in this
subpopulation.
In the multivariable logistic regression, the following factors were positively associated with achieving viral load suppression: adherence (adjusted OR [AOR] = 2.90, 95% CI: 1.43-5.88), time on therapy (AOR = 1.11, 95% CI: 1.08-1.15), two nucleosides plus NNRTI versus two nucleosides plus PI (AOR = 3.33, 95% CI: 1.42-7.81), and lower baseline HIV-1 RNA level (AOR = 2.13, 95% CI: 1.22-3.70). Relative to non-drug users, current injection drug users were significantly less likely to suppress their HIV-1 RNA (AOR = 0.37, 95% CI: 0.16-0.82), and former
injection drug users were not significantly different from non-drug users (AOR = 0.61, 95% CI: 0.25-1.45). When we fitted the model with the interaction term of drug use and adherence, the direction of the relationship between current
drug use and viral suppression changed. For participants who never used drugs or were former drug users, the following variables were independently associated with HIV-1 RNA suppression: adherence (AOR = 1.33, 95% CI: 1.14-1.55), lower baseline HIV-1 RNA (AOR = 2.16, 95% CI: 1.18-4.18), two nucleosides plus NNRTI versus two nucleosides plus PI (AOR = 4.67, 95% CI: 1.55-14.1), and months on therapy (AOR = 1.11, 95% CI: 1.07-1.15). For current drug users, we found that adherence was not independently associated with viral suppression (AOR = 1.07, 95% CI: 0.87-1.33), whereas the following factors were: lower baseline HIV-1 RNA (AOR = 4.85, 95% CI: 1.34-17.5), two
nucleosides plus NNRTI versus two nucleosides plus PI (AOR = 7.19, 95% CI: 1.46-35.7), and months on therapy (AOR = 1.13, 95% CI: 1.06-1.20).
A number of studies have found that active illicit drug and alcohol use is associated with lower adherence to HAART as measured by self-report and MEMS as well as by inferior virologic and immunologic outcomes. Arnsten et al. examined the relationship between self-reported adherence and MEMS with HIV-1 RNA suppression among current and former drug users and found that both measures were significantly correlated with viral suppression but that MEMS was a more sensitive measure of nonadherence. Moatti et al. found that nonadherence was independently associated with younger age, frequency of negative life events in the prior 6 months, alcohol consumption, and active drug use. Among patients at an urban HIV clinic, Lucas et al. found that continued use of illicit drugs was associated with reduced adherence and impaired virologic and immunologic responses to HAART. They also recently reported that patients who switched from non-drug use to substance abuse had worsening antiretroviral adherence and less frequent HIV-1 RNA suppression compared with patients who remained free of substance abuse. In addition, patients who switched from substance abuse to nonuse had improvements in adherence and HIV treatment outcomes compared with patients with persistent substance abuse over a 1-year period.
Another notable finding of our study relates to the fact that patients on NNRTI-based therapy had over a fourfold increased odds of achieving HIV-1 RNA suppression compared with those on PI-based therapy, adjusting for other
prognostic variables. This is consistent with recent observational studies. These findings remain controversial, however. Given the multiple subtle biases that may come into play when selecting HAART regimens for a given clinical situation, we believe that the relative efficacy of two nucleosides plus NNRTI versus two nucleosides plus PI should be addressed in randomized clinical trials. Based on our results, such trials should ensure that the complexity
of either regimen is comparable so as to minimize confounding a true regimen effect with decreased adherence related to treatment complexity.
Our results demonstrate that virologic outcomes were independently related to adherence among former and non-injection drug users. This has been repeatedly found in numerous studies. In contrast, our adherence measure of pharmacy refill compliance was not independently associated with virologic suppression for current drug users. Among this group, the behavior of obtaining the medications from the pharmacy did not translate to ingestion of the medications as it did with former and non-injection drug users.
Limitations of our study include our inability to capture whether participants actually took their dispensed medication. We attempted to control for the difference in adherence to antiretroviral therapy based on refill compliance, a powerful predictor of virologic and clinical outcome for former and non-injection drug users. Second, the participants who were active drug injectors in our study likely comprised drug users who their physicians may have
considered potentially more adherent to antiretroviral therapy relative to the larger pool of eligible HIV-infected active injection drug users who are not receiving any treatment. Third, we do not have data on non-injection drug use
such as crack and alcohol that could also influence adherence behavior and have an effect on HIV-1 RNA suppression. Finally, the definition of virologic suppression is relative to the sensitivity of the assay. The assay detection threshold of 500 copies/mL was commonly used when many of our participants started HAART, but this has recently decreased to 50 copies/mL.
In summary, we found that HIV-infected injection drug users who were on HAART and continued to inject drugs were less likely to suppress HIV-1 RNA when compared with non-injection drug users adjusting for other prognostic
variables. High level of adherence in the first year of therapy, longer time on therapy, lower baseline HIV-1 RNA level, and use of NNRTI-based therapies were also independently associated with superior virologic outcome for former and non-injection drug users. Adherence based on pharmacy refill data is an unreliable measure for current injection drug users because it does not capture whether the medication was ingested. For this subpopulation, more refined measures of adherence such as self-report, pill counts, and MEMS may be necessary.
|
|
| |
| |
|
|
|
