The treatment of HIV infection with highly active antiretroviral therapy (HAART) is based on the short-term goal of maximal and durable suppression of plasma HIV RNA levels in an effort to restore and preserve immunologic function, improve quality of life, and reduce HIV-related morbidity and mortality.1,2
Several studies have recently demonstrated that baseline plasma HIV RNA levels ≥100,000 copies/mL are independently associated with mortality after the initiation of HAART.3,4 These findings have caused a great deal of uncertainty among clinicians, because the explanation for these observations remains unknown.5,6 For instance, the association between baseline plasma HIV RNA and mortality is difficult to reconcile with studies indicating that higher plasma HIV RNA levels at baseline are not associated with poorer virologic outcomes after the initiation of HAART.7,8
A limitation of previous studies that have examined the role of baseline plasma HIV RNA and CD4 cell count on virologic responses3,7,9,10 is that they have not been adjusted for patient adherence, which is known to be among the strongest determinants of virologic suppression.11-15 Therefore, we conducted the present population-based study to evaluate the role of baseline plasma HIV RNA and CD4 cell count in predicting virologic responses to HAART while taking into consideration the potentially confounding effect of patient adherence.
The HAART Observational Medical Evaluation and Research (HOMER) study run through the British Columbia Center for Excellence in HIV/AIDS Drug Treatment Program has been described in detail elsewhere.10,16 Briefly, the center is the only free source of antiretroviral medications in the province of British Columbia, and pharmaceutic sales suggest that <1% of HIV-infected patients obtain antiretrovirals outside the program.17 For all program participants, a complete prospective profile of antiretroviral therapy is maintained.
In the present study, analyses were restricted to HIV-infected men and women who were antiretroviral naïve, were first prescribed triple-drug antiretroviral therapy between August 1, 1996 and July 31, 2000, and were followed to March 31, 2002. Study subjects were initially prescribed triple-drug combination therapy with regimens that included 2 nucleoside reverse transcriptase inhibitors and a protease inhibitor or a nonnucleoside reverse transcriptase inhibitor at the discretion of the enrolling physician. Patients who initiated HAART with more than 3 antiretroviral agents were excluded from the present study, because studies that have reported an association between baseline HIV RNA ≥100,000 copies/mL and mortality have been restricted to patients starting triple therapy.3,4 For the purposes of these analyses, we followed the intent-to-treat principle, and subjects were included as they were first dispensed antiretrovirals, regardless of whether they later modified their therapeutic regimen.
HIV-1 RNA Responses
We evaluated time to HIV RNA suppression and rebound. As previously reported,7,14 suppression was defined as the first of at least 2 consecutive plasma HIV RNA measures <500 copies/mL. Event-free patients were right-censored at the date of last HIV RNA measure before March 31, 2002. Rebound was defined as the first of at least 2 consecutive plasma HIV-1 RNA levels ≥500 copies/mL after any HIV RNA measure <500 copies/mL.7 The rebound event was conservatively assumed to occur at the midpoint between the last undetectable HIV RNA measure and the first of the 2 consecutive measures ≥500 copies/mL.18 Suppressed patients who did not rebound were right-censored as of the last HIV RNA measure before March 31, 2002. To be consistent with previous studies, the rebound analysis was restricted to those patients who were suppressed ≤8 months after the initiation of therapy.7,18
For both outcomes (suppression and rebound), we evaluated the cumulative event rates using Kaplan-Meier methods, and Cox proportional hazards regression was used to calculate the univariate and adjusted relative hazards.19 The assumption of proportional hazards was validated by inspection of log (−log [survival function]) estimates against log time plots. To evaluate the impact of baseline virologic and immunologic status, we stratified patients into low (<50,000 copies/mL), medium (50-99,999 copies/mL), and high (≥100,000 copies/mL) HIV RNA strata, and low (<50 cells/μL), medium (50-199 cells/μL), and high (≥200 cells/μL) CD4 cell count strata. These categories were selected a priori on the basis of several recent studies and therapeutic guidelines that have highlighted the clinical significance of these cutoffs.1-4,10,20
We also further stratified patients into adherent and nonadherent categories using prescription refill compliance.21 The definition of adherence was based on the ratio of time that medication dispensed would last as a proportion of follow-up time during the first year on therapy. We have previously demonstrated how this estimate strongly and independently predicts virologic response and mortality and how it can adjust for the potentially confounding effect of treatment interruption.14,15,18,22 Patients were a priori defined as nonadherent if they received antiretroviral medications <95% of the time, based on previously published work.14,18 Additional variables examined in these analyses included protease inhibitor use in the initial regimen (yes vs. no), a prior diagnosis of AIDS (yes vs. no), age, gender, physician experience (≥6 patients previously enrolled in the program),17 and date of therapy initiation (before or after July 1997).23
To derive adjusted relative hazards of each event (suppression and rebound) among adherent and nonadherent patients in each HIV RNA stratum, we also developed fixed models with indicator variables for each adherence and HIV RNA stratum while adjusting for CD4 cell count and other relevant covariates. All multivariate models described were fit using the same protocol of adjusting for all variables that were statistically significant (P < 0.05) in univariate analyses.
Because patients with a lower baseline plasma HIV RNA level may be expected to achieve virologic suppression more rapidly,7 as a subanalysis, we also fit a logistic model to evaluate whether potential differences in the rate of virologic suppression translated into differences in the likelihood of ever achieving HIV RNA suppression during follow-up. In addition, in an effort to reconcile the reported association between plasma HIV RNA levels and mortality,3,4 we examined plasma HIV RNA suppression rates among those patients who died during the study period. All statistical analyses were performed using SAS software version 8.0 (SAS, Cary, NC). All tests of significance were 2-sided, with a P value <0.05 indicating that an association was statistically significant.
Between August 1, 1996 and July 31, 2000, 1583 antiretroviral-naive participants aged 18 years and older began triple-drug combination therapy. Of these, 161 (10.0%) were excluded from this analysis for not having baseline CD4 cell count and plasma HIV-1 RNA level measures available within 6 months before the start of antiretroviral therapy. Therefore, the study sample was based on 1422 (90.0%) subjects (1198 [84.3%] men, 224 [15.7%] women). No differences in sex, baseline AIDS status, and subsequent mortality were observed between the study sample and those excluded. Persons excluded from this analysis were more likely to be younger (P = 0.04) and taking protease inhibitors (P = 0.02), however. The overall median follow-up time was 40.1 months (interquartile range [IQR]: 27.7-52.9 months). At baseline, the median age of participants was 37.2 years (IQR: 32.2-43.7 years), median CD4 cell count was 270 cells/μL (IQR: 130-420 cells/μL), and median plasma HIV RNA level was 120,000 copies/mL (IQR: 38,000-300,000 copies/mL). Overall, 983 (69.1%) patients initiated therapy with a protease inhibitor, and 439 (30.9%) patients initiated therapy with a nonnucleoside reverse transcriptase inhibitor. Antiretroviral regimens used in the cohort have been recently described,10 and a detailed description of medications used by patients in the present study can be obtained from the corresponding author. Overall, there were 193 deaths during the study period, among which 39 were attributed to accident/suicide (25 [13.0%]) or to illicit drug overdoses (14 [7.3%]).
HIV-1 RNA Suppression (<500 copies/mL)
Figure 1 shows the Kaplan-Meier cumulative plasma HIV RNA suppression estimates for the overall cohort stratified by adherent (panel A) and nonadherent (panel B) patients and by HIV RNA strata. Among adherent patients, statistically significant differences in suppression rates were noted between the 3 HIV RNA strata (all log-rank comparisons, P < 0.05).
Table 1 shows the adjusted relative hazard of suppression. Baseline CD4 cell count was not statistically associated with time to suppression. In contrast, patients in the medium and high plasma HIV RNA strata took longer to suppress (both P < 0.05) than patients in the low HIV RNA stratum after adjustment for adherence, which was also highly statistically significant (P < 0.05). In this and the models presented below, there were no statistically significant interaction effects detected between baseline covariates, but fixed models with indicator variables were prepared to compare adherent patients in each baseline HIV RNA stratum. Here, the previous finding was confirmed by building a Cox model with indicator variables representing each adherence and baseline HIV RNA stratum, with adherent patients in the low HIV RNA stratum being the reference category. As shown in Table 2, adherent patients in the medium and high HIV RNA strata were slower to suppress than adherent patients in the low HIV RNA stratum (both P < 0.05).
HIV-1 RNA Rebound (>500 copies/mL)
Figure 2 shows the Kaplan-Meier cumulative plasma HIV RNA rebound estimates for the overall cohort stratified by adherent (panel A) and nonadherent (panel B) patients and by HIV RNA strata. Among adherent patients, no statistical differences were noted between the 3 HIV RNA strata (all log-rank comparisons, P > 0.1).
Table 3 shows the adjusted relative hazard of rebound for the 970 patients who were suppressed to <500 copies/mL at least once within 32 weeks after the initiation of HAART. As shown here, there were no statistical differences in the rate of HIV RNA rebound between baseline CD4 cell count or HIV RNA strata. As shown in Table 4, this was confirmed when we fit a Cox model with indicator variables and found no difference between adherent patients in the various HIV RNA strata (both P > 0.1).
We returned to the observed difference in the rate of HIV RNA suppression by HIV RNA strata that was observed in Figure 1 (panel A) and confirmed in Table 1. We then built a logistic model to evaluate whether differences in the time to suppression observed in the Cox regression analyses translated into differences in the likelihood of ever achieving HIV RNA suppression at any time during follow-up. As shown in Table 5, although there was no difference in the adjusted odds of suppression among adherent patients in the low and medium baseline HIV RNA strata (P = 0.197), adherent patients in the ≥100,000-copies/mL HIV RNA stratum had a markedly lower odds of achieving suppression at any time during follow-up (adjusted odds ratio [AOR] = 0.27 [95% confidence interval (CI): 0.13-0.54]; P < 0.001) even after adjustment for baseline CD4 cell count and other relevant covariates.
Finally, taking all these findings into consideration, we evaluated whether the previously reported association between baseline HIV RNA ≥100,000 copies/mL and mortality3,4 could possibly be the result of a relation between failure to achieve virologic suppression and baseline HIV RNA. We excluded the 39 accidental deaths because we assumed that they were potential confounders, and we examined virologic responses among the 154 nonaccidental deaths observed during follow-up. We found that 102 (66%) patients who died never achieved suppression during follow-up as defined previously. We also noted that among the 31 patients who were ≥95% adherent but did not achieve HIV RNA suppression during follow-up and later died, 27 (87.1%) had baseline HIV RNA ≥100,000 copies/mL. The nonaccidental causes of death in the cohort have been previously described10 and are overwhelmingly attributed to AIDS-related causes.
The present study demonstrates that higher HIV RNA levels were associated with slower rates of virologic suppression and that patients with HIV RNA ≥100,000 copies/mL were statistically less likely ever to achieve virologic suppression at any time during follow-up. In contrast, baseline HIV RNA and CD4 cell count were not statistically associated with virologic rebound and CD4 cell count was not associated with suppression. It is also noteworthy that most deaths observed during the study period were among patients who did not achieve HIV RNA suppression and that among adherent patients who did not suppress and later died, 87% had baseline HIV RNA ≥100,000 copies/mL.
In the present study, baseline CD4 cell count was not associated with time to virologic suppression or rebound among adherent patients. This finding contrasts with those of several recent studies that have considered this issue.8,9,24,25 Of note, such findings have been cited as reason to consider earlier initiation of antiretroviral therapy in recent guidelines.1,2 These earlier observational studies were not adjusted for patient adherence, however, and our findings suggest that prior analyses may have been at least partially confounded by characteristics of patients who present later for treatment, which may include, among others, incomplete adherence.26 Our findings are also consistent with the findings of a collaborative study that baseline CD4 cell count was not associated with virologic responses.7
Similarly, our observation that adherent patients with baseline HIV RNA ≥100,000 copies/mL were significantly less likely ever to suppress plasma HIV RNA during follow-up is in contrast with recently reported findings.7,8 Again, our findings may differ because of the fact that these earlier studies were unable to adjust for patient adherence. Our results are particularly striking, considering that the median follow-up interval was longer than 40 months. Nevertheless, in comparison to adherent patients with baseline HIV RNA <50,000 copies/mL, adherent patients with baseline HIV RNA ≥100,000 copies/mL were less likely ever to achieve HIV RNA suppression during the entire follow-up period (AOR = 0.27 [95% CI: 0.13-0.54]; P < 0.001), even after adjustment for baseline CD4 cell count and other covariates.
We also found that among adherent patients who did not suppress HIV RNA during follow-up and who later died, 87.1% had baseline HIV RNA ≥100,000 copies/mL. It is also noteworthy that there was no statistical difference in the odds of achieving HIV RNA suppression between patients in the <50,000-copies/mL and 50,000-99,999-copies/mL strata, which may explain why we have recently shown no difference in mortality among these patients.4 Together, these findings suggest that the association between baseline HIV RNA ≥100,000 copies/mL and mortality3,4 may be attributable to the inadequate potency of conventional triple-therapy regimens for many patients with baseline HIV RNA ≥100,000 copies/mL. In fact, although data have not been available to date, it should be noted that this explanation has been previously speculated on.27
One possible clinical strategy to address our findings would be to stage patients for earlier antiretroviral treatment before HIV RNA rises to greater than 100,000 copies/mL. It has been previously well articulated by others6,7 that HIV RNA only increases gradually and nonuniformly among untreated patients7,28,29 and that earlier initiation of treatment would place patients at increased risk of side effects, toxicity, and premature evolution of resistance.30,31 An alternative strategy that should ideally be explored through a prospective clinical trial would be to stage patients with higher HIV RNA levels for more potent initial antiretroviral regimens.32,33
It is important to stress that these data arose in a setting in which all HIV/AIDS care, antiretrovirals, and laboratory monitoring are available free of charge and in which previous studies have shown that virtually all patients acquire antiretrovirals through a single centralized source.17 In addition, because a complete prospective record of antiretroviral dispensation was maintained, it was possible to determine precisely each individual's level of prescription refill compliance. Although using refill compliance as a surrogate for adherence has been previously validated,14,15,21,22 a limitation of the present study is that patients were defined as adherent based on their behavior during the first year of therapy and then assigned to adherent strata for an analysis of baseline characteristics. As previously discussed,16,18,22 this approach strives to reduce the potential for reverse causation, where patients stop therapy because of morbidity caused by HIV infection. It also enables a comparison between patients who were similarly adherent but had different baseline characteristics. In addition, like other population-based studies,7,34 we were unable to use the ultrasensitive assay of <50 copies/mL, because 500 copies/mL was the lower limit of detection for most of the study period. If differences existed between the 3 baseline HIV RNA strata with regard to the degree of suppression below 500 copies/mL, however, analyses of clinical trial data have demonstrated that differences in rebound would likely have been observed.35 Finally, we should also point out that many of the patients beginning HAART in the present study would have initiated therapy with a triple-combination regimen that is less potent than the regimens currently used in contemporary practice. Although this is a limitation of the present study, it can also be viewed as a strength, because the observational studies that have demonstrated elevated mortality rates among patients initiating HAART with plasma HIV RNA ≥100,000 copies/mL were conducted among patients initiating HAART over a similar period and using similar HAART regimens. In fact, several of these studies were conducted among the same cohort the present study was drawn from.4,10
In summary, we found that patients with HIV RNA ≥100,000 copies/mL were statistically less likely ever to achieve virologic suppression at any time during follow-up. This finding suggests that the association between baseline plasma HIV RNA ≥100,000 copies/mL and mortality3,4 may be explained by a more common failure of conventional triple-therapy regimens to suppress HIV RNA among patients with higher HIV RNA levels. In contrast to several recent studies that did not adjust for patient adherence,9,24,25 we found that baseline HIV RNA and CD4 cell count were not statistically associated with virologic rebound among adherent patients and that CD4 cell count was not associated with virologic suppression. If appropriately confirmed, these findings have important implications for the development of therapeutic guidelines.
The authors thank Bonnie Devlin, Diane Campbell, Elizabeth Ferris, Nada Gataric, Kelly Hsu Myrna Reginaldo, Chandra Lips, and Peter Vann for their research and administrative assistance. They also thank Keith Chan, Mark Tyndall, Kevin Craib, and Martin Schechter for their advice on the statistical methods.
1. Dybul M, Fauci AS, Bartlett JG, et al. Guidelines for using antiretroviral agents among HIV-infected adults and adolescents. Ann Intern Med
2. Yeni PG, Hammer SM, Carpenter CC, et al. Antiretroviral treatment for adult HIV infection in 2002: updated recommendations of the International AIDS Society-USA Panel. JAMA
3. Egger M, May M, Chene G, et al. Prognosis of HIV-1-infected patients starting highly active antiretroviral therapy: a collaborative analysis of prospective studies. Lancet
4. Wood E, Hogg RS, Yip B, et al. Higher baseline plasma HIV-1 RNA levels are associated with increased mortality after the initiation of triple drug antiretroviral therapy. J Infect Dis
5. Lane HC, Neaton JD. When to start therapy for HIV infection: a swinging pendulum in search of data. Ann Intern Med
6. Phillips AN, Lepri AC, Lampe F, et al. When should antiretroviral therapy be started for HIV infection? Interpreting the evidence from observational studies. AIDS
7. Phillips AN, Staszewski S, Weber R, et al. HIV viral load response to antiretroviral therapy according to the baseline CD4 cell count and viral load. JAMA
8. Skowron G, Street JC, Obee EM. Baseline CD4+ cell count, not viral load, correlates with virologic suppression
induced by potent antiretroviral therapy. J Acquir Immune Defic Syndr
9. Palella FJ, Jr, Deloria-Knoll M, Chmiel JS, et al. Survival
benefit of initiating antiretroviral therapy in HIV-infected persons in different CD4+ cell strata. Ann Intern Med
10. Hogg RS, Yip B, Chan KJ, et al. Rates of disease progression by baseline CD4 cell count and viral load after initiating triple-drug therapy. JAMA
11. Paterson DL, Swindells S, Mohr J, et al. Adherence
to protease inhibitor therapy and outcomes in patients with HIV infection. Ann Intern Med
12. Liu H, Golin CE, Miller LG, et al. A comparison study of multiple measures of adherence
to HIV protease inhibitors. Ann Intern Med
13. Gifford AL, Bormann JE, Shively MJ, et al. Predictors of self-reported adherence
and plasma HIV concentrations in patients on multidrug antiretroviral regimens. J Acquir Immune Defic Syndr
14. Low-Beer S, Yip B, O'Shaughnessy MV, et al. Adherence
to triple therapy and viral load response. J Acquir Immune Defic Syndr
15. Palepu A, Yip B, Miller C, et al. Factors associated with the response to antiretroviral therapy among HIV-infected patients with and without a history of injection drug use. AIDS
16. Wood E, Hogg RS, Yip B, et al. Is there a baseline CD4 cell count that precludes a survival
response to modern antiretroviral therapy? AIDS
17. Strathdee SA, Palepu A, Cornelisse PG, et al. Barriers to use of free antiretroviral therapy in injection drug users. JAMA
18. Wood E, Montaner JS, Yip B, et al. Adherence
and plasma HIV RNA responses to highly active antiretroviral therapy among HIV-1 infected injection drug users. Can Med Assoc J
19. Cox DR. Regression models and life tables. J R Stat Soc B
20. Ledergerber B, Egger M, Erard V, et al. AIDS-related opportunistic illnesses occurring after initiation of potent antiretroviral therapy: the Swiss HIV Cohort Study. JAMA
21. Steiner JF, Prochazka AV. The assessment of refill compliance using pharmacy records: methods, validity, and applications. J Clin Epidemiol
22. Hogg RS, Heath KV, Bangsberg D, et al. Intermittent use of triple combination therapy is predictive of mortality at baseline and after one year of follow-up. AIDS
23. Carpenter CC, Fischl MA, Hammer SM, et al. Antiretroviral therapy for HIV infection in 1997. Updated recommendations of the International AIDS Society-USA Panel. JAMA
24. Deeks SG, Hecht FM, Swanson M, et al. HIV RNA and CD4 cell count response to protease inhibitor therapy in an urban AIDS clinic: response to both initial and salvage therapy. AIDS
25. Lucas GM, 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
26. Celentano DD, Galai N, Sethi AK, et al. Time to initiating highly active antiretroviral therapy among HIV-infected injection drug users. AIDS
27. Sterling TR, Chaisson RE, Moore RD. HIV-1 RNA, CD4 T-lymphocytes, and clinical response to highly active antiretroviral therapy. AIDS
28. O'Brien TR, Rosenberg PS, Yellin F, et al. Longitudinal HIV-1 RNA levels in a cohort of homosexual men. J Acquir Immune Defic Syndr Hum Retrovirol
29. Lyles CM, Dorrucci M, Vlahov D, et al. Longitudinal human immunodeficiency virus type 1 load in the Italian Seroconversion Study: correlates and temporal trends of virus load. J Infect Dis
30. Heath KV, Hogg RS, Chan K, et al. Lipodystrophy-associated morphological, cholesterol, and triglyceride abnormalities in a population-based HIV/AIDS treatment program. AIDS
31. Cote HC, Brumme ZL, Craib KJ, et al. Changes in mitochondrial DNA as a marker of nucleoside toxicity in HIV-infected patients. N Engl J Med
32. Weverling GJ, Lange JM, Jurriaans S, et al. Alternative multidrug regimen provides improved suppression
of HIV-1 replication over triple therapy. AIDS
33. van Praag RM, Wit FW, Jurriaans S, et al. Improved long-term suppression
of HIV-1 replication with a triple-class multidrug regimen compared with standard of care antiretroviral therapy. AIDS
34. Ledergerber B, Egger M, Opravil M, et al. Clinical progression and virological failure on highly active antiretroviral therapy in HIV-1 patients: a prospective cohort study. Swiss HIV Cohort Study. Lancet
35. Raboud JM, Montaner JS, Conway B, et al. Suppression
of plasma viral load
below 20 copies/ml is required to achieve a long-term response to therapy. AIDS
Keywords:© 2005 Lippincott Williams & Wilkins, Inc.
plasma viral load; adherence; suppression; rebound; survival