Skip Navigation

Patterns, Predictors, and Consequences of Initial Regimen Type among HIV-Infected Women Receiving Highly Active Antiretroviral Therapy

  1. Elizabeth T. Golub1,
  2. Lorie Benning1,
  3. Anjali Sharma2,
  4. Monica Gandhi3,
  5. Mardge H. Cohen4,
  6. Mary Young5, and
  7. Stephen J. Gange1
  1. 1Johns Hopkins Bloomberg School of Public Health, Department of Epidemiology, Baltimore, Maryland
  2. 2Department of Infectious Diseases, Downstate Medical Center, State University of New York, Brooklyn
  3. 3Division of Infectious Diseases, Department of Medicine, University of California-San Francisco, Brooklyn
  4. 4CORE Center, Cook County Bureau of Health Services, Department of Medicine, Stroger Hospital and Rush Medical College, Chicago, Illinois
  5. 5Georgetown University Medical Center, Georgetown University, Washington, D.C.
  1. Reprints or correspondence: Dr. Elizabeth T. Golub, Dept. of Epidemiology, Johns Hopkins Bloomberg School of Public Health, 615 N. Wolfe St., E-7636, Baltimore, MD 21205 (egolub{at}jhsph.edu).
 
Next Section

Abstract

Background. It is important to elucidate differences among initial highly active antiretroviral therapy (HAART) regimen types in comparative studies of therapy effectiveness. We aimed to identify predictors of initiation with different HAART regimen types and the effect of initial regimen type on switching and immunologic response to therapy—controlling for those predictors—among human immunodeficiency virus (HIV)–infected women in the United States.

Methods. Participants in the Women's Interagency HIV Study underwent semiannual interview, venipuncture, and clinical examination. Those beginning with protease inhibitor-based, nonnucleoside reverse-transcriptase inhibitor (NNRTI)–based, or triple-nucleoside reverse-transcriptase inhibitor (NRTI)–based HAART during April 1996–March 2005 were eligible for analysis. Predictors of initial regimen type were assessed with polytomous logistic regression. Correlates of switching were assessed with discrete-time proportional hazards models, and immunologic response to therapy was assessed with linear regression.

Results. Among 1555 HAART initiators, CD4+ lymphocyte count and HIV load were significant predictors of initial regimen type during 1996–2002; only sociodemographic predictors were significant during 2002–2005. Initial regimen type was not a significant predictor of subsequent regimen switching. Compared with those whose initial treatment was protease inhibitor-based HAART, those who began with triple-NRTI–based regimens had significantly lower CD4+ cell counts at 1 year (P=.006) and 2 years (P=.004) after initiation; NNRTI initiators had lower CD4+ cell counts after 2 years (P=.05).

Conclusions. We demonstrate that predictors of initial regimen type among women in the United States have been changing over time. Protease inhibitor initiators had significantly higher CD4+ cell counts than did NNRTI or triple-NRTI initiators up to 2 years after HAART initiation. Adjustment for biological predictors of initial regimen is important to avoid confounding in the study of treatment effectiveness.

HAART has greatly improved survival and has decreased the incidence of AIDS-associated opportunistic infection among HIV-infected individuals. Despite their substantial effectiveness, however, the drugs comprising HAART are associated with a variety of adverse effects [1]. Studies have identified increased risk of insulin resistance [2, 3] or diabetes [4], cardiovascular disease [5, 6], and other metabolic complications [7–9] as a result of prolonged treatment with protease inhibitors (PIs). These effects, as well as high pill burden, stringent schedules, and food restrictions, can compromise adherence to PI-based regimens [10, 11], thereby reducing the effectiveness of some therapies. For example, in a cohort of >1200 patients initiating PI-based regimens, intolerance was the most frequently cited reason for discontinuing therapy [12].

Many studies have compared patients receiving nonnucleoside reverse-transcriptase inhibitors (NNRTIs) with those receiving nucleoside reverse-transcriptase inhibitors (NRTI) only or PI-based regimens, with consideration of such phenomena as switching, discontinuation, and response to HAART therapy. It is vital to make these assessments in both clinical trials (reflecting efficacy of therapy given under experimental conditions) and community-based settings (reflecting effectiveness of therapy as it is prescribed and used in practice).

Research suggests that there is no difference in the rate of HAART discontinuation between PI-based and NNRTI-based HAART [13, 14]. In the Women's Interagency HIV Study (WIHS), African American or Latina ethnicity and higher HIV load before initiation of HAART were significant predictors of regimen switching [15]. Other investigators have found that switching from a PI-based regimen to an NNRTI-based or an NRTI-based regimen can lead to lower rates of HAART discontinuation [16, 17], increased tolerability [18], better adherence [19], lower rates of treatment failure [16, 20], and higher rates of virologic suppression [21].

Finally, studies of treatment effectiveness have found similar immunologic responses to PI-based and NNRTI-based therapy but varying virologic responses. Research from the EuroSIDA Study Group compared antiretroviral-naive patients starting a PI- or NNRTI-based combination antiretroviral therapy (ART) regimen and found no difference with regard to CD4+ cell count increase [22], a finding replicated by other studies [23, 24]. However, research suggests that patients receiving PI-based therapy are less likely to achieve an undetectable HIV load than are those receiving NNRTIs [22, 23]. However, the majority of these studies have been conducted among study populations that were 67%–89% male [13, 14, 16, 17, 20, 21, 24], therefore without sufficient numbers to properly address these issues among women.

This study had 4 aims. First, we describe, among US women, patterns of initiation with PI-, NNRTI- and triple-NRTI-based regimens during April 1996–March 2005. Second, we evaluate participant factors associated with initial regimen type, because it is important to understand the nature of these subgroups for comparative studies of therapy effectiveness. Finally, we investigated the consequences of initial regimen type, controlling for the identified predictors, using 2 outcomes: treatment modifications (as measured by either switching to a new regimen or discontinuation of antiretroviral use) and HIV disease progression (as measured by CD4+ cell count). Altogether, we investigated therapy type, first as an outcome, characterizing women who begin each type of HAART regimen, and then as an exposure, with therapy changes and immunologic response to therapy as the outcomes.

Previous SectionNext Section

Methods

Study population. Data were collected from participants in WIHS, a prospective, multicenter, cohort study and the largest observational study sponsored by National Institutes of Health to focus exclusively on the study of HIV infection among US women. Recruitment and data collection methods were published elsewhere [25, 26]; WIHS participants are largely representative of those affected by the domestic HIV epidemic [25]. During 1994–1995, 2054 HIV-infected and 569 HIV-uninfected women were enrolled at 6 clinical consortia (Bronx and Brooklyn, NY; Chicago, IL; Los Angeles and San Francisco, CA; and Washington, DC). Another 737 HIV-infected and 406 HIV-uninfected women were enrolled during 2001–2002.

Initial HAART regimen was investigated among 1761 HIV-infected participants who began HAART (defined below) between 1 January 1996 and 31 March 2005 and had <1 year between their last report of a non-HAART regimen (or no ART) and first reported HAART use. Participants were excluded from analysis (88 patients [5%]) if the initial regimen included both a PI and an NNRTI or if it could not be determined which drug treatment was begun earlier. We investigated first switch from initial HAART regimen and 1- and 2-year immunological response among participants who began HAART after 1 October 1999 and had at least 1 follow-up visit for the given outcome.

Data collection. In brief, participants undergo semiannual visits, comprising extensive interviews, venipuncture, and clinical examinations. Study visits are calendar based, have 6-month study windows, and correspond to April–September and October–March cutoff dates. Interviewer-administered questionnaires assess detailed information about demographic characteristics, medical symptoms, and antiretroviral therapy. We defined HAART as any reported use of any of the following: (1) ⩾2 NRTIs in combination with at least 1 PI or 1 NNRTI, (2) 1 NRTI in combination with at least 1 PI and at least 1 NNRTI, (3) ritonavir and saquinavir in combination with 1 NRTI and no NNRTIs, and (4) an abacavir- or tenofovir-containing regimen of ⩾3 NRTIs in the absence of both PIs and NNRTIs, except for abacavir-tenofovir-lamivudine and didanosine-tenofovir-lamivudine.

Statistical analysis. Baseline characteristics of PI-, NNRTI-, and triple-NRTI–based HAART initiators, and also of switchers and nonswitchers, were compared with χ2 tests for categorical variables and with Wilcoxon rank sum tests, t tests, and 1-way analysis of variance for continuous variables. Our first aim was to describe patterns in HAART initiation. The date of initiation was either (1) known from medical-record abstraction, (2) reported by the participant as month and year, or (3) estimated as the midpoint between the date of the last “HAART-free visit” and the date of the first “HAART visit.” On the basis of observed trends in the relative proportion of initial regimen types, the 9-year study period was grouped into 3 distinct periods: period 1 (1 April 1996–31 March 2000), period 2 (1 April 2000–31 March 2002), and period 3 (1 April 2002–31 March 2005). Trends in initial regimen type were graphed using the midpoints of the calendar-based study periods (i.e., January as the midpoint of October–March and July as the midpoint of April–September).

Second, we identified predictors of initial regimen type in each period. Demographic, behavioral, immunologic, and virologic factors were investigated for potential associations with initial regimen type, on the basis of the current literature and biological plausibility. Generalized linear models were constructed using a logit link under a binomial distribution for period 1 and a multinomial distribution for periods 2 and 3 [27].

Our third aim was to evaluate trends in regimen-type switching subsequent to HAART initiation. We investigated predictors of regimen switching among recent (after 1 October 1999) HAART initiators, adjusting for potential confounders identified in aim 2. ART use at each visit subsequent to the first reported use of HAART was categorized as PI, NNRTI, triple-NRTI, or dual-based (i.e., regimens including both a PI and an NNRTI) HAART, or non-HAART (including all non-HAART regimens, as well as ART discontinuation). At each follow-up visit, current regimen type was compared with initial regimen type. If the types matched, the follow-up regimen was considered to be a nonswitch, regardless of whether the individual drugs within that class changed. If the types did not match, the follow-up regimen was considered to be a switch.

Each participant contributed ⩾1 observation, beginning with HAART initiation and ending with the first reported switch, the date last seen, or the date last seen before missing ⩾2 consecutive visits. Demographic, HIV-specific, and health care utilization characteristics were assessed at each follow-up visit, restricted to the first 4 years after HAART initiation. Discrete-time proportional hazards models were specified using a dichotomous outcome for switching, with a logit link under a binomial distribution [28].

Our final aim was to investigate the association between initial HAART regimen type and 1- and 2-year absolute CD4+ cell counts among recent HAART initiators, with adjustment for potential confounders identified in aim 2. To accomplish this aim, linear regression models were constructed for year 1 CD4+ lymphocyte count (measured 6–18 months after HAART initiation) and for year 2 CD4+ lymphocyte count (measured 1.5–2.5 years after initiation). For each outcome, 3 models are presented: (1) univariate (initial HAART regimen type only), (2) multivariate (including biological factors), and (3) “full” multivariate (including both biological and sociodemographic factors).

For each study aim, we specified univariate regression models, using a significance level of a=.10. Full models were constructed including all significant covariates identified in univariate analysis. In addition, for aims 3 and 4, we included all significant predictors from aim 2. Variables were eliminated using a backward stepwise process and the likelihood-ratio test. The final model included all forced covariates plus any additional factors that were significant at a=.05. All analyses were conducted with Stata, version 8 (Stata Corporation), and SAS software, version 9 (SAS Institute). Figures were constructed using Splus, version 7.0 (Mathsoft).

Previous SectionNext Section

Results

A total of 1673 women began HAART during periods 1 (n=1280), 2 (n=234), and 3 (n=159). We excluded 118 women because of missing data, which left 1555 for analysis (table 1). Of those, 1040 women began treatment with a PI-based regimen, 435 with an NNRTI-based regimen, and 80 with a triple-NRTI–based regimen. PI initiators had significantly lower nadir pre-HAART CD4+ lymphocyte counts and higher peak pre-HAART HIV RNA levels (P<.001) and were more likely to have had AIDS and to have been ART experienced (P<.001) before beginning HAART. Those who began treatment with NNRTI were significantly younger (P<.001) and were most likely to have reported using illegal drugs.

Figure 1
View larger version:
Figure 1

Calendar time trends in initial regimen type among 1673 participants in the Women's Interagency HIV Study. The x-axis indicates January and July as the midpoints of the calendar-based study-visit windows. NNRTI, nonnucleoside reverse-transcriptase inhibitor; NRTI, nucleoside reverse-transcriptase inhibitor; PI, protease inhibitor.

Figure 2
View larger version:
Figure 2

Significant predictors of initial HAART regimen type during 3 time periods. April 1996–March 2000 and April 2002–March 2005 models adjusted for calendar period.

Figure 3
View larger version:
Figure 3

Predictors of first switch after HAART initiation. Analysis adjusted for pre-HAART nadir CD4+ cell count, peak pre-HAART HIV RNA, age at HAART initiation, race, preferred interview language, medical insurance with prescription coverage, and calendar time. NNRTI, nonnucleoside reverse-transcriptase inhibitor; NRTI, nucleoside reverse-transcriptase inhibitor; PI, protease inhibitor. *Switch to a different regimen type, to a non-HAART regimen, or discontinue all ART.

Table 1
View larger version:
Table 1

Baseline characteristics of 1555 patients who began HAART in the Women's Interagency HIV Study (WIHS), 1996–2005.

Patterns and predictors of initial HAART regimen. Trends in initiation differed by calendar period (figure 1). Odds of beginning with an NNRTI-based (vs. PI-based) regimen increased during period 1. Compared with April–September 1996, the odds of starting with an NNRTI-based (vs. PI-based) regimen were 10 times higher during April–September 1998 (P<.001) and >50 times higher (P<.001) by October 1999–March 2000. During period 2, there were no significant trends in initiation over time. Compared with 2002, the odds of beginning with both PI-based and NNRTI-based regimens increased from the odds of 2003–2005 and were significantly (P<.05) higher than the odds of initiating with triple-NRTI–based regimens.

Predictors of initial regimen type are presented in figure 2. Because of trends observed in initial regimen during periods 1 and 3, those models were adjusted for calendar time. During time period 1 (1 April 1996–31 March 2000), women with a CD4+ cell count ⩾200 cells/mL were more than twice as likely as those with a count <200 cells/mL to begin with an NNRTI-based (vs. PI-based) regimen, and those with a baseline history of injection drug use were less likely (adjusted OR [aOR], 0.69; 95% CI, 0.48–1.00) to begin with an NNRTI-based (vs. PI-based) regimen. Not enough women initiated with triple-NRTI–based regimens during period 1 to facilitate meaningful comparisons.

During time period 2 (1 April 2000–31 March 2002), significant predictors of beginning with an NNRTI-based regimen were CD4+ cell count, increased peak pre-HAART HIV RNA level (aOR, 1.53 per 1-U log10 increase; 95% CI, 1.08–2.17 per 1-U log10 increase), and preference for a Spanish-language interview (aOR, 7.73; 95% CI, 1.70–35.11). Independent predictors of beginning with a triple-NRTI–based (vs. NNRTI-based) regimen were lower HIV RNA level (aOR, 0.45; 95% CI, 0.29–0.70) and higher age at initiation (aOR, 1.06 per 1-year increase; 95% CI, 1.02–1.12 per 1-year increase).

During time period 3 (1 April 2002–31 March 2005), both black (aOR, 0.08; 95% CI, 0.01–0.66) and Latina (aOR, 0.06; 95% CI, 0.01–0.48) women were significantly less likely than white/other women to begin HAART with an NNRTI-based (vs. PI-based) regimen and were more likely to begin with a triple-NRTI–based (vs. NNRTI-based) regimen (black women, aOR, 8.35, 95% CI, 1.63–42.87; Latina women, aOR, 4.93; 95% CI, 0.83–29.10) as were those whose medical insurance included prescription coverage (aOR, 4.27; 95% CI, 1.39–13.15).

Changes to initial regimen. A total of 301 participants (corresponding to 908 person-intervals) who began HAART after 1 October 1999 and contributed complete data were included in the longitudinal analysis of switching regimen type. Initial regimens were either PI based (n=93), NNRTI based (n=157), or triple-NRTI based (n=51). The cumulative hazard of switching was 50% by 12 months and was 62% by 24 months after HAART initiation. Compared with PI initiators, switching was greatest among triple-NRTI initiators (OR, 1.70; 95% CI, 1.05–2.74) and lowest among NNRTI initiators (OR, 0.84; 95% CI, 0.57–1.22).

Compared with those who did not switch from their initial regimen type, those who switched were significantly less likely to have begun with an NNRTI-based regimen (49% vs. 59%; P=.01), more likely to be African American (65% vs. 50%; P=.03), and more likely ART experienced at HAART initiation (55% vs. 33%; P<.001). Switchers were also marginally more likely than nonswitchers to have had clinical AIDS before HAART initiation (34% vs. 25%; P=.07) and were more likely to have used illegal drugs (20% vs. 13%; P=.09).

Independent predictors of switching (figure 3) included clinical AIDS after HAART initiation (aOR, 4.46; 95% CI, 1.39–14.25), pre-HAART ART experience (aOR, 1.69; 95% CI, 1.16–2.47), higher log10 HIV load (aOR, 1.40; 95% CI, 1.23–1.60), history of treatment by a mental health professional (aOR, 1.85; 95% CI, 1.21–2.82), and having had ⩾3 (versus <3) visits to a physician in the previous 6 months (aOR, 2.29; 95% CI, 1.55–3.39); these last 3 variables were allowed to vary over time. This model was adjusted for initial regimen type, the 6 independent predictors of initial regimen type (pre-HAART nadir CD4+ cell count, peak pre-HAART HIV load, age at initiation, race, preferred interview language, and medical insurance with prescription coverage), and calendar time. After adjustment for all other factors, initial regimen type did not significantly predict switching in this cohort.

Mean CD4+ lymphocyte count. Level of immunosuppression after initiation of therapy was assessed via participants' CD4+ lymphocyte count at 1 (n=309) and 2 (n=263) years after HAART initiation. For each time point, 3 models are presented (table 2): a univariate model of the effect of initial regimen type on mean CD4+ cell count, a multivariate model that adjusts for biological factors only, and the full model, including biological and sociodemographic factors. At 1 year after HAART initiation, there was no univariate effect of initial regimen type on CD4+ cell count. With adjustment for period of initiation, both nadir and last pre-HAART CD4+ cell count, both peak pre-HAART and concurrent HIV load, as well as interactions thereof, women who began with a triple-NRTI–based regimen had significantly lower CD4+ cell counts than did those initiating with PI regimens (P=.01). Further adjustment for prior ART use, AIDS, and demographic variables did not change the results; triple-NRTI initiators had, on average, CD4+ cell counts that were 90 cells lower than those of PI initiators (P=.006), with no difference in CD4+ count between NNRTI and PI initiators (P>.05). At 2 years after HAART initiation, triple-NRTI initiators still had CD4+ cell counts significantly lower than those of PI initiators (P=.004). In addition, NNRTI initiators had, on average, CD4+ cell counts 54 cells lower than those of PI initiators (P=.05), with adjustment for the same covariates as in the year 1 analysis.

Table 2
View larger version:
Table 2

Factors associated with CD4+ cell count at 1 and 2 years after initiation of HAART.

Previous SectionNext Section

Discussion

In this study, we found that, despite changes in predictors of initial HAART regimen over time, adjustment for biological predictors revealed a significant association between initial HAART regimen and subsequent CD4+ lymphocyte count. It is clear that a multitude of factors affect a patient's initial regimen, and we conclude that omitting those from investigations into the effectiveness of therapy has the potential to introduce residual confounding. In this cohort of US women, we were able to avoid such confounding by identifying, and subsequently controlling for, predictors of initial HAART regimen.

Trends in initial regimen type have been changing over time. Current HAART initiators in the WIHS tend to initiate with either PI-based or NNRTI-based regimens, with very few beginning with a triple-NRTI–based regimen. Predictors of initial regimen type are also changing. In the early HAART era, immunological and virological factors significantly predicted initial HAART regimen. This is consistent with findings from the EuroSIDA group, which showed that patients starting NNRTI-based regimens had significantly higher nadir CD4+ cell counts than did those initiating therapy with PI-based regimens [22]. Since 2002, however, only sociodemographic factors significantly predicted initial regimen type in our cohort. Analyses considering the effectiveness of initial regimen type on treatment outcomes in nonrandomized studies need to control for predictors of initial regimen as potential confounders.

Among recent HAART initiators, one-half switched regimen type within 12 months of initiation, and nearly two-thirds switched within the first 24 months. These high rates of switching underscore the complex nature of managing HIV-treatment regimens [29]. A combination of clinical parameters and health care utilization factors predicted switching in this population. Those who had more contact with the health care system were more likely to switch from their initial regimen type. More-frequent visits to one's health care provider may indicate an ineffective regimen, thereby making switching more likely. Those with pre-HAART ART exposure, higher viral loads, and AIDS diagnoses were also more likely to switch, further suggesting previous failed treatments.

Our study found no difference between NNRTI initiators and PI initiators in CD4+ lymphocyte count at 1 year after HAART initiation. This is consistent with a recent study that found no difference between NNRTIs and PIs with regard to AIDS, death, virologic failure, or CD4+ cell–count decline to <200 cells/mm3 [30]. triple-NRTI initiators had significantly poorer outcomes than did PI initiators.

In this study, we were unable to assess the individual physician's prescribing practices. Ideally, one could survey the physicians prescribing these therapies, which would enable a more comprehensive explanation about why certain patients are treated with certain regimens. Despite this limitation, our comprehensive data on dates when each drug treatment was begun and stopped make this study uniquely positioned to examine initiation and switching of therapy.

As the study of the effectiveness of HIV therapies in observational settings continues, it is important to keep in mind the myriad factors indicating which patients are prescribed which therapies and that these factors may change over time. Including these factors in consideration of differences in treatment effectiveness between regimens will reduce the potential for residual confounding and will more finely differentiate regimen types.

Previous SectionNext Section

Acknowledgments

Data in this manuscript were collected by the WIHS Collaborative Study Group with centers (principal investigators) at New York City/Bronx Consortium (Kathryn Anastos); Brooklyn, NY (Howard Minkoff); Washington DC Metropolitan Consortium (Mary Young); The Connie Wofsy Study Consortium of Northern California (Ruth Greenblatt); Los Angeles County/Southern California Consortium (Alexandra Levine); Chicago Consortium (Mardge Cohen); and Data Coordinating Center (Stephen Gange).

Financial support. The WIHS is funded by the National Institute of Allergy and Infectious Diseases, the National Cancer Institute, and the National Institute on Drug Abuse (UO1-AI-35004, UO1-AI-31834, UO1-AI-34994, UO1-AI-34989, UO1-AI-34993, and UO1-AI-42590). Funding is also provided by the National Institute of Child Health and Human Development (UO1-HD-32632) and the National Center for Research Resources (MO1-RR-00071, MO1-RR-00079, and MO1-RR-00083).

Potential conflicts of interest. All authors: no conflicts.

  • Received July 16, 2007.
  • Accepted September 25, 2007.
Previous Section
 

References

  1. ↵
    1. d'Arminio MA,
    2. Lepri AC,
    3. Rezza G,
    4. et al
    . Insights into the reasons for discontinuation of the first highly active antiretroviral therapy (HAART) regimen in a cohort of antiretroviral naive patients. AIDS 2000;14:499-507.
    CrossRefMedline
  2. ↵
    1. Yan Q,
    2. Hruz PW
    . Direct comparison of the acute in vivo effects of HIV protease inhibitors on peripheral glucose disposal. J Acquir Immune Defic Syndr 2005;40:398-403.
    CrossRefMedlineWeb of Science
  3. ↵
    1. Behrens G,
    2. Dejam A,
    3. Schmidt H,
    4. et al
    . Impaired glucose tolerance, beta cell function and lipid metabolism in HIV patients under treatment with protease inhibitors. AIDS 1999;13:63-70.
    CrossRef
  4. ↵
    1. Justman JE,
    2. Benning L,
    3. Danoff A,
    4. et al
    . Protease inhibitor use and the incidence of diabetes mellitus in a large cohort of HIV-infected women. J Acquir Immune Defic Syndr 2003;32:298-302.
    CrossRefMedlineWeb of Science
  5. ↵
    1. Friis-Moller N,
    2. Sabin CA,
    3. Weber R,
    4. et al
    . Combination antiretroviral therapy and the risk of myocardial infarction. N Engl J Med 2003;349:1993-2003.
    CrossRefMedlineWeb of Science
  6. ↵
    1. Zareba KM,
    2. Miller TL,
    3. Lipshultz SE
    . Cardiovascular disease and toxicities related to HIV infection and its therapies. Expert Opin Drug Saf 2005;4:1017-25.
    CrossRefMedline
  7. ↵
    1. Grunfeld C,
    2. Tien P
    . Difficulties in understanding the metabolic complications of acquired immune deficiency syndrome. Clin Infect Dis 2003;37:43-6.
    CrossRef
    1. Nolan D
    . Metabolic complications associated with HIV protease inhibitor therapy. Drugs 2003;63:2555-74.
    CrossRefMedlineWeb of Science
  9. ↵
    1. Wohl DA,
    2. McComsey G,
    3. Tebas P,
    4. et al
    . Current concepts in the diagnosis and management of metabolic complications of HIV infection and its therapy. Clin Infect Dis 2006;43:645-53.
    Abstract/FREE Full Text
  10. ↵
    1. Maggiolo F,
    2. Ravasio L,
    3. Ripamonti D,
    4. et al
    . Similar adherence rates favor different virologic outcomes for patients treated with nonnucleoside analogues or protease inhibitors. Clin Infect Dis 2005;40:158-63.
    Abstract/FREE Full Text
  11. ↵
    1. van Roon EN,
    2. Verzijl JM,
    3. Juttmann JR,
    4. Lenderink AW,
    5. Blans MJ,
    6. Egberts AC
    . Incidence of discontinuation of highly active antiretroviral combination therapy (HAART) and its determinants. J Acquir Immune Defic Syndr Hum Retrovirol 1999;20:290-4.
    Medline
  12. ↵
    1. Le Moing V,
    2. Chene G,
    3. Leport C,
    4. et al
    . Impact of discontinuation of initial protease inhibitor therapy on further virological response in a cohort of human immunodeficiency virus-infected patients. Clin Infect Dis 2002;34:239-47.
    CrossRefMedlineWeb of Science
  13. ↵
    1. Dorrucci M,
    2. Pezzotti P,
    3. Grisorio B,
    4. et al
    . Time to discontinuation of the first highly active antiretroviral therapy regimen: a comparison between protease inhibitor-and non-nucleoside reverse transcriptase inhibitor-containing regimens. AIDS 2001;15:1733-6.
    CrossRefMedlineWeb of Science
  14. ↵
    1. Mocroft A,
    2. Youle M,
    3. Moore A,
    4. et al
    . Reasons for modification and discontinuation of antiretrovirals: results from a single treatment centre. AIDS 2001;15:185-94.
    CrossRefMedlineWeb of Science
  15. ↵
    1. Kirstein LM,
    2. Greenblatt RM,
    3. Anastos K,
    4. et al
    . Prevalence and correlates of highly active antiretroviral therapy switching in the Women's Interagency HIV Study. J Acquir Immune Defic Syndr 2002;29:495-503.
    MedlineWeb of Science
  16. ↵
    1. Dieleman JP,
    2. Sturkenboom MC,
    3. Wit FW,
    4. et al
    . Low risk of treatment failure after substitution of nevirapine for protease inhibitors among human immunodeficiency virus-infected patients with virus suppression. J Infect Dis 2002;185:1261-8.
    Abstract/FREE Full Text
  17. ↵
    1. Bucher HC,
    2. Kofler A,
    3. Nuesch R,
    4. Young J,
    5. Battegay M,
    6. Opravil M
    . Meta-analysis of randomized controlled trials of simplified versus continued protease inhibitor-based antiretroviral therapy in HIV-1-infected patients. AIDS 2003;17:2451-9.
    CrossRefMedlineWeb of Science
  18. ↵
    1. Maggiolo F,
    2. Ripamonti D,
    3. Ravasio L,
    4. et al
    . Outcome of 2 simplification strategies for the treatment of human immunodeficiency virus type 1 infection. Clin Infect Dis 2003;37:41-9.
    Abstract/FREE Full Text
  19. ↵
    1. Maggiolo F,
    2. Ripamonti D,
    3. Suter F
    . Switch strategies in patients on effective HAART. J Antimicrob Chemother 2005;55:821-3.
    Abstract/FREE Full Text
  20. ↵
    1. Mocroft A,
    2. Phillips AN,
    3. Ledergerber B,
    4. et al
    . Relationship between antiretrovirals used as part of a cART regimen and CD4 cell count increases in patients with suppressed viremia. AIDS 2006;20:1141-50.
    CrossRefMedlineWeb of Science
  21. ↵
    1. Abgrall S,
    2. Yeni PG,
    3. Bouchaud O,
    4. Costagliola D
    . Comparative biological and clinical outcomes after a switch from a virologically unsuccessful first protease inhibitor-containing antiretroviral combination to a 3-drug regimen containing efavirenz, nevirapine, or abacavir. Clin Infect Dis 2008;44:120-7.
  22. ↵
    1. Mocroft A,
    2. Horban A,
    3. Clumeck N,
    4. et al
    . Comparison of single and boosted protease inhibitor versus nonnucleoside reverse transcriptase inhibitor-containing cART regimens in antiretroviral-naive patients starting cART after January 1, 2000. HIV Clin Trials 2006;7:271-84.
    CrossRefMedlineWeb of Science
  23. ↵
    1. Pulido F,
    2. Arribas JR,
    3. Miro JM,
    4. et al
    . Clinical, virologic, and immunologic response to efavirenz-or protease inhibitor-based highly active antiretroviral therapy in a cohort of antiretroviral-naive patients with advanced HIV infection (EfaVIP 2 study). J Acquir Immune Defic Syndr 2004;35:343-50.
    CrossRefMedlineWeb of Science
  24. ↵
    1. Waters L,
    2. Stebbing J,
    3. Jones R,
    4. et al
    . A comparison of the CD4 response to antiretroviral regimens in patients commencing therapy with low CD4 counts. J Antimicrob Chemother 2004;54:503-7.
    Abstract/FREE Full Text
  25. ↵
    1. Bacon MC,
    2. von Wyl V,
    3. Alden C,
    4. et al
    . The Women's Interagency HIV Study: an observational cohort brings clinical sciences to the bench. Clin Diagn Lab Immunol 2005;12:1013-9.
    CrossRefMedline
  26. ↵
    1. Barkan SE,
    2. Melnick SL,
    3. Preston-Martin S,
    4. et al
    . The Women's Interagency HIV Study. WIHS Collaborative Study Group. Epidemiology 1998;9:117-25.
    CrossRefMedlineWeb of Science
  27. ↵
    1. Hosmer DW,
    2. Lemeshow S
    . Applied logistic regression. New York: John Wiley. 1989.
  28. ↵
    1. Singer JD,
    2. Willett JB
    . Applied longitudinal data analysis: modeling change and event occurrence. New York: Oxford University Press. 2003.
  29. ↵
    1. Anderson AM,
    2. Bartlett JA
    . Changing antiretroviral therapy in the setting of virologic relapse: review of the current literature. Curr HIV/AIDS Rep 2006;3:79-85.
    CrossRefMedline
  30. ↵
    1. MacArthur RD,
    2. Novak RM,
    3. Peng G,
    4. et al
    . A comparison of three highly active antiretroviral treatment strategies consisting of non-nucleoside reverse transcriptase inhibitors, protease inhibitors, or both in the presence of nucleoside reverse transcriptase inhibitors as initial therapy (CPCRA 058 FIRST Study): a long-term randomised trial. Lancet 2006;368:2125-35.
    CrossRefMedlineWeb of Science
| Table of Contents

This Article

  1. Clin Infect Dis. (2008) 46 (2): 305-312. doi: 10.1086/524752
  1. AbstractFree
  2. » Full Text (HTML)Free
  3. Full Text (PDF)Free

Classifications

Share

  1. Email this article

Navigate This Article

Current Issue

  1. March 1, 2012 54 (5)
  1. Clinical Infectious Diseases
  1. Alert me to new issues

Most