Shortly after highly active antiretroviral therapy (HAART) was introduced, studies from US, Canadian, and European observational cohorts suggested that, on average, only approximately 50% to 60% of patients receiving HAART in clinical practice achieved virologic suppression by 6 months.1-4 This contrasts with higher rates of virologic suppression seen in randomized clinical trials of HAART, which is a potential concern, because early suppression of viral load is associated with better outcomes.5 Possible reasons for these early differences included the suboptimal addition of new antiretroviral (ARV) drugs to failing regimens, difficulty in tolerating certain drugs because of adverse effects, high medication burden, complexity of the regimen, and an inadequate understanding of the importance of high levels of adherence. We sought to determine how virologic response to HAART has changed in clinical practice since 1996. The Johns Hopkins HIV Cohort provided us with the opportunity to examine these trends over time.
The Johns Hopkins AIDS Service provides care for a large proportion of HIV-infected patients in the Baltimore metropolitan area. Longitudinal primary and subspecialty care is integrated in a single hospital-based HIV clinic. An observational, longitudinal, clinical database has been maintained on patients receiving primary HIV care in our HIV/AIDS clinic since 1990. In this longitudinal database, data are updated regularly using clinic and inpatient clinical documentation (from the Johns Hopkins AIDS Service and elsewhere), laboratory testing, pharmacy, and social services. Race and/or ethnicity and HIV transmission risk group are based on patient self-report. Prescription of ARV therapy (drug, dates of use, and dose) is well documented by the medical provider and support staff. Trained abstractors record all this information onto standardized forms for processing. Details of the methodology have been previously described.6
We analyzed data from those patients who started their initial HAART regimen (defined by Department of Health and Human Services [DHHS] guidelines7) from 1996 through 2002. We did not assess a 3-nucleoside reverse transcriptase inhibitor (NRTI) regimen as a HAART regimen, because the use of this combination occurred in only 1999 through 2002 and this regimen is no longer considered to be an adequately potent regimen.8 A 3-NRTI combination alone was used as the initial HAART regimen in only 13 patients from 1999 through 2002. We assessed suppression of HIV-1 RNA to <400 copies/mL by 6 and 12 months after starting HAART. If the patient suppressed to <400 copies/mL before each of these time points, he or she was counted as having suppressed even if he or she subsequently had an HIV-1 RNA measurement >400 copies/mL. An additional analysis was done of patients who remained suppressed at the time threshold of interest (eg, 6 or 12 months). In a subset analysis, we also assessed viral suppression to <50 copies/mL from 1999 through 2002, when the more sensitive assay was available in our clinic. Another subset analysis was done on just those patients who were ARV naive at baseline. All HIV-1 RNA levels from 1996 through 2002 were done using the Roche Amplicor reverse transcriptase polymerase chain reaction (RT-PCR) assay.
In addition to viral load, other end points assessed included the following: (1) change in CD4 count from baseline by 6 months and 12 months, in which the CD4 count obtained closest to the time point was used; (2) change from baseline HAART regimen to an alternative regimen (eg, PI to nonnucleoside reverse transcriptase inhibitor [NNRTI], NNRTI to PI), in which boosting of a nonboosted PI was not included; and (3) development of a new AIDS-defining opportunistic infection based on US Public Health Service (USPHS) guidelines.9
Analyses are presented for 1996, 1997 through 1998, 1999 through 2000, and 2001 through 2002. A comparison of the percentage of patients over time who had viral suppression was done by means of the Mantel-Haenszel test. The change in CD4 cell count over time was assessed using the Kruskal-Wallis test. Multivariable analysis of HIV-1 RNA suppression at 6 and 12 months was done by means of logistic regression. Logistic regression computes relative odds rather than relative risk. Conceptually, the relative odds relate to the odds of having the response (eg, HIV-1 RNA suppression) versus not having the response for those with the risk factor to these same odds for those without the risk factor.
Two multivariable models were assessed. In the first (model 1), the associations of multiple variables with viral load suppression to <400 copies/mL were delineated. These variables included the patient being ARV naive, baseline CD4 cell count, HIV-1 RNA level, HAART regimen (single PI, boosted PI, or NNRTI), and demographic variables (sex, race, and HIV risk group). In the second (model 2), calendar year was also included as a surrogate for temporal factors that were otherwise unmeasured. Examples of such factors are provider experience, adherence interventions, changes over time in medication side effects, and ease of taking the medication.
Finally, if a patient died or was censored because of loss to follow-up before 12 months, the last HIV-1 RNA load or CD4 cell count was carried forward and used. Censoring averaged only 2.1% by 12 months, with no significant differences by year in censoring (range: 1.4%-4.1%).
The characteristics of our patient population in 1996 and subsequent years (grouped in 2-year increments) are shown in Table 1. Over time, there was a significant increase (P < 0.05) in the proportion of women and African Americans as well as in HIV transmission by heterosexual sex comparing 1996 with later years. There were no significant differences in CD4 cell count at the start of therapy, but there were significantly fewer patients with a prior history of nucleoside-only ARV therapy use and fewer patients receiving ARV agents to which they had previously been exposed over calendar time. There was a difference over time in the initial HAART regimen, with a significant decline in the use of an unboosted PI and an increase in the use of a ritonavir-boosted PI or an NNRTI.
We next examined the virologic response by year in which the initial HAART regimen was used (Table 2). HIV-1 RNA suppression ranged from 43.8% (1996) to 72.4% (2001-2002) in patients by 6 months and from 60.1% (1996) to 79.9% (2001-2002) by 12 months after the initial HAART regimen was started. HIV-1 RNA suppression was sustained (ie, no rebound) at each of these time points in 1% to 5% fewer patients at 6 months and in 5% to 7% fewer patients at 12 months (data in parentheses in Table 2). Of note, a change in the initially selected HAART regimen occurred in 3.4% (1996) to 7.3% (2001-2002) of patients by 6 months and in 8.6% (1996) to 11.8% (2001-2002) of patients by 12 months.
In the subset analysis of ARV-naive patients, HIV-1 RNA suppression ranged from 58.6% (1996) to 74.3% (2001-2002) in patients at 6 months and from 72.4% (1996) to 78.1% (2001-2002) by 12 months. In subset analyses, we found that the percentage of patients who suppressed to <50 copies/mL was 61.6% at 6 months and 72.6% at 12 months in 1999 through 2000 and 58.2% at 6 months and 72.8% at 12 months in 2001 through 2002.
The increase in CD4 count from baseline ranged from 67 cells/mm3 (1996) to 99 cells/mm3 (2001-2002) at 6 months and from 116 cells/mm3 (1996) to 142 cells/mm3 (2001-2002) at 12 months. Opportunistic illness rates declined from 13.6% (1996) to 9.8% (2001-2002) by 12 months of follow-up (not shown in Table 2).
Our multivariable analysis of HIV-1 RNA suppression by 6 months (model 1) showed significant associations with the use of an NNRTI and boosted PIs compared with unboosted PIs, being ARV naive, a lower baseline HIV-1 RNA level, and higher baseline CD4 count (Table 3). Adjusting for these variables in model 2, we found that calendar time was also associated with HIV-1 RNA suppression. Multivariable analyses of HIV-1 RNA suppression at 12 months were similar (data not shown).
In our cohort of HIV-infected patients under care in Baltimore, there has been an improvement from 1996 to 2002 in viral suppression and in CD4 response associated with initial HAART. Use of an NNRTI or boosted PI was more likely than use of an unboosted PI to be associated with HIV-1 RNA suppression. HIV-1 RNA suppression was also associated with being ARV naive as demonstrated in our subset analysis and our multivariable analysis. Adjusting for the HAART regimen, being ARV naive, and other clinical and demographic variables, there was still an association between calendar time and HIV-1 RNA suppression. This association of calendar time suggests that other unmeasured factors for which calendar time is a surrogate, such as adherence support, are also likely to be contributing to improved HIV-1 RNA suppression. Unfortunately, we do not have objective measures of adherence from 1996 through 2002 to assess this hypothesis more directly. Other unmeasured factors that have a temporal trend such as decisions by the provider or patient to withhold therapy until the patient was well prepared may also be reflected in the significant association of calendar time with HIV suppression. These could include factors such as active drug use, something that was not measured in our patients across all calendar years.
We believe that these results reflect improvements in therapy and expertise in using that therapy. Given that these results are from a single center, it is possible that they may not generalize to other HIV clinical practice settings. The relatively low rate of achieving HIV suppression in our cohort was similar to the rate of suppression found in other clinical cohorts with early use of HAART, however.2,3 Our results may similarly generalize to current HIV clinical practice, where ARV therapy is similarly available.
1. Lucas G, Chaisson RE, Moore RD. Highly active antiretroviral therapy
in a large urban clinic: risk factors for virologic failure and adverse drug reactions. Ann Intern Med
2. Mocroft A, Gill MJ, Davidson W, et al. Predictors of a viral response and subsequent virological treatment failure in patients with HIV starting a protease inhibitor. AIDS
3. Paredes R, Mocroft A, Kirk O, et al. Predictors of virological success and ensuing failure in HIV-positive patients starting highly active antiretroviral therapy
in Europe: results from the EuroSIDA study. Arch Intern Med
4. Kirk O, Mocroft A, Pradier C, et al. Clinical outcome among HIV-infected patients starting saquinavir hard gel compared to ritonavir or indinavir. AIDS
5. Yamashita TE, Phair JP, Munoz A, et al. Immunologic and virologic response to highly active antiretroviral therapy
in the Multicenter AIDS Cohort Study. AIDS
6. Moore RD. Understanding the clinical and economic outcomes of HIV therapy: the Johns Hopkins HIV clinical practice cohort. J Acquir Immune Defic Syndr
7. Department of Heath and Human Services. Guidelines for the use of antiretroviral agents in HIV-infected adolescents and adults. Available at: http://www.aidsinfo.nih.gov
. Accessed February 25, 2005.
8. Gulick R, Ribaudo HJ, Shikuma CM, et al. Triple-nucleoside regimens versus efavirenz-containing regimens for the initial treatment of HIV-1 infection. N Engl J Med
9. Centers for Disease Control and Prevention. Guidelines for preventing opportunistic infections among HIV-infected persons, 2002. Recommendations of the US Public Health Service and the Infectious Diseases Society of America. MMWR Recomm Rep