Important note: Information in this article was accurate in January 2004. The state of the art may have changed since the publication date.
 |
Since its availability in 1996, the use of highly active antiretroviral therapy (HAART) to treat HIV-infection has led to declines in U.S. AIDS mortality rates and increases in the number of persons living with HIV. For HIV-infected individuals, drug resistance is a major concern, but it is unknown whether resistance is an inevitable consequence of long-term HAART use. HAART can be inherently complex if it involves taking many pills at specific times or with food restrictions, or if it is associated with side effects.
Non-adherence is recognized as one of the main causes of treatment failure. The association between non-adherence to antiretroviral therapy (ART) and HIV drug resistance was first shown in patients receiving protease inhibitor (PI) monotherapy [Vanhove, et al. JAMA. 1996 Dec 25;276(24):1955-6]. Studies have also demonstrated that incomplete adherence to HAART is associated with drug resistance. In 1998, Montaner and colleagues showed in a clinical trial that incomplete adherence to HAART was also associated with HIV drug resistance [JAMA. 1998 Mar 25;279(12):930-7]. In 2000, two separate studies demonstrated that patients who experienced virologic failure did not necessarily have resistance to all drugs in their regimen [Descamps, et al. JAMA. 2000 Jan 12;283(2):205-11; Havlir, et al. JAMA. 2000 Jan 12;283(2):229-34]. This finding provided a rationale for resistance testing in patients failing therapy, which was subsequently recommended by an International AIDS Society-USA Panel [Hirsch, et al. JAMA. 2000 May 10;283(18):2417-26] and by a panel from the Department of Health and Human Services [January 28, 2000 version; http://aidsinfo.nih.gov].
Paterson and colleagues showed that at least 95% adherence was needed to maximally suppress viral replication in patients receiving HAART [Paterson, et al. Ann Intern Med. 2000 Jul 4;133(1):21-301]. However, while the relationship between adherence and risk of virologic failure appears to be more or less linear, the relationship between adherence and the risk of drug resistance is not. Friedland and Williams proposed that there are two clinical scenarios resulting in minimal development of resistance: 1) extremely high levels of adherence leading to maximal suppression of viral replication; and 2) very poor adherence, resulting in no inhibition of viral replication [AIDS. 1999 Sep;13 Suppl 1:S61-72]. In the first scenario, resistance is unlikely to develop because there is little or no viral replication. In the second scenario, viral replication takes place in the absence of antiretroviral drug exposure sufficient to result in selective pressure favoring resistant virus. Between these two extremes lies a middle zone in which viral replication occurs in the face of drug pressure, leading to a risk of drug resistance (see Figure 1A). Several studies have attempted to confirm this hypothetical “bellshaped” relationship between adherence and resistance and to determine the level of adherence to HAART associated with the maximum risk of development of HIV drug resistance.
In 2001, Gallego and colleagues examined patterns of drug resistance mutations among HIV-infected patients who were failing their first HAART regimen that included indinavir at 3 different hospitals in Spain [AIDS. 2001 Sep 7;15(13):1701-6]. Genotypic resistance was examined at the time of first virologic failure (>50 c/mL) among 51 patients reporting at least 90% adherence to their regimen and 14 patients reporting <90% adherence. They found that among patients taking >90% of their medication, 51% had resistance to nucleosides and 27% had resistance to protease inhibitors. None of the subjects with <90% adherence were found to have mutations associated with HIV drug resistance. Since patients had to have maintained viral suppression for 6 months to be eligible for the study, the researchers could be assured that resistance detected at the time of failure was incident rather than prevalent. However, low sample size prevented the investigators from examining the relationship between adherence and resistance more thoroughly. The crosssectional nature of the study was also a limitation, as the effect of long-term non-adherence on resistance development could not be assessed.
In 2002, Walsh and colleagues also showed that lower adherence was associated with lower detection of HIV drug resistance [J Acquir Immune Defic Syndr. 2002 Jul 1;30(3):278-87]. The investigators examined adherence assessed by various methods among 68 HIV-infected adults receiving their first PI-containing HAART regimen for at least 6 months at a London HIV clinic. Among the 32 patients who experienced virologic failure and the 36 who did not, mean adherence was 84% and 95% of pills taken in the previous 3 days, respectively. Subjects who had experienced virologic failure were tested for resistance, which revealed that subjects whose virus had a sensitive phenotype had significantly lower average adherence (69%) than those with either intermediate (93%) or resistant phenotype (99%). Since they demonstrated a significant positive linear relationship between adherence and HIV resistance, the researchers suggested that if there is a "bell-shaped" relationship between adherence and resistance, and that the association is probably skewed to the left, such that the greatest risk of drug resistance results from only marginally sub-optimal adherence. The investigators acknowledged that since their study was cross-sectional, it was not possible to determine if adherence measured during the study was reflective of past adherence and if resistance detected during the study was due to past non-adherence.
In a recently published prospective study of HIV-infected urban poor individuals in the Research on Access to Care in the Homeless (REACH) cohort in San Francisco, Bangsberg and colleagues examined the relationship between adherence, viral suppression, and HIV drug resistance among 148 participants who received at least one month of HAART [AIDS. 2003 Sep 5;17(13):1925-32]. Adherence was assessed by unannounced pill counts at the participants’ place of residence and measured at every 3 to 6 weeks for one year. Higher adherence was correlated with longer time on treatment (p<0.0001) and viral suppression defined as viral load <50 c/mL. Over the 12-month period, mean adherence was 82% for those with a mean viral load <50 c/mL and 58% for those with a mean viral load >50 c/mL. Based on model estimations from 22 (15%) participants who developed new drug resistance mutations and 37 (25%) participants who maintained viral suppression during the period of observation, the investigators suggested that 23% of drug resistance mutations occur in individuals with 92-100% adherence, 30% with 79-91% adherence, 15% with 58-78% adherence, 20% with 42-57% adherence, and 12% with 0-41% adherence (see Figure 1B). The authors concluded that high levels of resistance correlate with increasing rates of viral suppression, but also that high levels of adherence correlate with drug resistance.
Another recently published prospective study conducted employed a rigorous study design to determine the level of non-adherence to HAART that was associated with the greatest risk of HIV drug resistance [Sethi, et al. Clin Infect Dis. 2003 Oct 15;37(8):1112-8]. The researchers followed 195 patients who had maintained viral suppression while receiving HAART at the Johns Hopkins Moore Clinic for one year. Adherence in the past 3-days was assessed by questionnaire and collected each time patients had an appointment with their HIV clinician. The investigators found that cumulative adherence of 70 to 89% prior to the development of resistance was associated with more than a 3- fold greater risk of resistance compared to those with <70% and >90% adherence. Overall incidence of resistance was 14.5 per 100 personyears and was 9.6, 13.6, 36.6, 44.9, and 12.3 per 100 person-years for patients whose cumulative adherence was 100%, 90-99%, 80-89%, 70- 79%, and <70%, respectively (see Figure 1C). It was also found that missing a scheduled clinic visit in the past month was independently associated with more than a 2-fold increase in the risk of developing drug resistance. The researchers concluded that high-level adherence was needed to avoid the development of resistance. The limitation of this study was the use of self-reported adherence, which is generally higher than objective adherence measures.
The relationship between adherence and resistance also depends on the regimen potency. In a sub-analysis of the Abbott 863 trial, a phase 3 study of antiretroviral-naïve patients receiving d4T, 3TC, and either lopinavir/ritonivir (N=326) or nelfinavir (N=327), King and colleagues examined adherence as measured by pill count and the development of resistance [IAS Conf HIV Pathog Treat 2003 Jul 13-16;2nd: Abstract No. 798]. Overall, a bell-shaped relationship between adherence and resistance was observed, with highest probability of resistance observed among those with 80-85% adherence. When examined by regimen type, individuals with perfect adherence to the nelfinavir-based regimen were over 3 times more likely to develop resistance than those with perfect adherence to the more potent lopinavir/ritonavir-based regimen.
In summary, there is growing evidence to support the hypothesis that the relationship between adherence and drug resistance is in fact bell-shaped. This becomes a dilemma for HIV clinicians and patients, who must decide whether their adherence is high enough to prevent resistance. For patients who are poorly adherent, adherence-promoting interventions may actually increase the risk of resistance if they promote adherence that falls short of the high levels necessary to maximally suppress viral replication.
Although very poor adherence appears to be associated with a decreased risk of resistance, there is evidence that even minimal adherence can lead to resistance since drug-resistant HIV has been detected after just a single-dose of nevirapine to prevent mother-to-child transmission [Jackson, et al. AIDS. 2000 Jul 28;14(11):F111-5]. Moreover, non-adherence is associated with virologic failure, immunologic failure, and clinical disease progression. In some cases, complete interruption of therapy might be preferable to continued therapy with poor adherence, with reinitiation after the reasons for non-adherence are addressed. This reinforces the need for HIV clinicians to counsel patients about the risks and consequences of non-adherence prior to beginning HAART and on an ongoing basis once HAART has been initiated.
Unfortunately, even perfect adherence may not prevent resistance, especially in those treated with less potent regimens and those who have acquired or were infected with drug-resistant virus. High-level adherence to regimens containing agents with a low genetic barrier to resistance (one mutation causes resistance) is also necessary. Routine monitoring and documentation of patient adherence is critical, along with ongoing assessment of potential barriers to maintaining long-term adherence. Adherence to clinic visits is also important, as it is associated with medication adherence. Clinic staff must continually assess the underlying reasons for missed appointments and attempt to improve show rates.
HIV-infected individuals who are nonadherent to HAART will likely develop drug resistance. To avoid the development of resistance, regimens must be highly potent and near-perfect adherence appears to be necessary. Unfortunately, for many patients, adherence can be difficult. A collective effort on the parts of HIV clinicians, clinic staff, and patients is required to assure longterm high-level adherence.
20040101
JH20040104
©1997-2004. The Johns Hopkins University AIDS Service, Division of Infectious Diseases. Permission to use and reproduce portions of this newsletter is hereby granted provided that author and publication are fully credited and both copyright and permission notice appear with reprinted material. Inquiries may be directed to Sharon McAvinue, Managing Editor. Website: Johns Hopkins AIDS Service.
AEGiS is made possible through unrestricted grants from Boehringer Ingelheim, Elton John AIDS Foundation, Elton John AIDS Foundation, the National Library of Medicine, and donations from users like you. Always watch for outdated information. This article first appeared in 2004. This material is designed to support, not replace, the relationship that exists between you and your doctor.
AEGiS presents published material, reprinted with permission and neither endorses nor opposes any material. All information contained on this website, including information relating to health conditions, products, and treatments, is for informational purposes only. It is often presented in summary or aggregate form. It is not meant to be a substitute for the advice provided by your own physician or other medical professionals. Always discuss treatment options with a doctor who specializes in treating HIV.
Copyright ©1980, 2004. AEGiS. All materials appearing on AEGiS are protected by copyright as a collective work or compilation under U.S. copyright and other laws and are the property of AEGiS, or the party credited as the provider of the content.
