The response of previously treated HIV infection to new agents is of great concern in this era of rapidly evolving clinical practice. A recent open-label clinical study suggested that patients who were nucleoside-experienced had less suppression of HIV-1 RNA after receipt of stavudine (D4T) than patients who were nucleoside-naive . In this study, 73% of nucleoside-experienced patients had received zidovudine (ZDV). Use of a dual nucleoside analog reverse transcriptase inhibitor (NRTI) regimen of D4T and lamivudine (3TC) was associated with a decline in HIV-1 RNA of 1.32log10 copies/ml at 4 weeks and 0.66log10 copies/ml at 24 weeks in nucleoside-experienced patients compared with a decline of 1.96log10 copies/ml at 4 weeks and 1.66log10 copies/ml at 24 weeks in nucleoside-naive patients. Nucleoside-naive patients had an increase in CD4 cell count of approximately 100×106/l, compared with less than 50×106/l in ZDV-experienced patients. A possible explanation for this finding for was suggested by an in vitro study demonstrating that phosphorylation of D4T was compromised by prior exposure to ZDV .
These studies have raised concerns regarding the effectiveness of D4T in patients who have previously had ZDV therapy, and have caused some to suggest that the sequencing of NRTI may be important. This is a potentially significant concern, since the number of available NRTI is limited, D4T has been shown to be an effective NRTI , and HIV-infected patients who have taken ZDV, which has been available since 1987 , are numerous.
We were interested in determining whether we could demonstrate a difference in response to D4T therapy between HIV-infected patients who were ZDV- experienced or ZDV-naive. We are now in an era where dual NRTI regimens are no longer recommended and patients are most likely to receive a combination antiretroviral drug regimen containing a protease inhibitor or a non-NRTI (NNRTI) [5,6]. The availability of a large clinical database at Johns Hopkins Hospital allowed us to conduct this analysis.
The Johns Hopkins HIV Clinic provides longitudinal primary and subspecialty care for a large proportion of the HIV-infected patients in the Baltimore metropolitan area. The clinic is staffed by 48 primary care providers (infectious disease and general internal medicine faculty physicians, physician assistants, nurse practitioners, and postdoctoral fellows) and a broad range of specialty-trained physicians who have a clinical focus in HIV disease. In addition to providing ambulatory care through the clinic, the Johns Hopkins AIDS Service provides inpatient, chronic care and home health services. To be registered in the clinic, patients must have a previous diagnosis of HIV infection. Every new patient undergoes an initial comprehensive medical and psychosocial evaluation by a primary care provider, a nurse, and a social worker, using structured instruments to collect comprehensive demographic, clinical, laboratory, pharmaceutical, and psychosocial characteristics. Extensive clinical information is maintained in the record, which includes subsequent drug therapy, clinical, laboratory and radiology evaluations, subspecialty referral information, hospitalizations, and evaluations from facilities other than Johns Hopkins. To ensure a high rate of longitudinal follow-up, medical records from other institutions in which the patients received care is routinely sought. Management of patients in the clinic is individualized according to the needs and preferences of patients and their clinicians. A medical record review for adherence to treatment and therapeutic monitoring guidelines is conducted weekly on a random sample of patients receiving care in the clinic.
For this analysis, we identified all patients who received D4T for the first time after 1 January 1995. To be included, D4T must have been used for at least 90 days. We then stratified these patients by previous use of ZDV. To be considered ZDV-experienced, at least 90 days of prior therapy was required. We defined ZDV-naive patients as those who had never received ZDV. Patients who had received less than 90 days of ZDV were not included in this analysis.
As a first step in our analysis, we compared several demographic and clinical characteristics of ZDV- experienced and ZDV-naive patients. We compared age, race/ethnicity, sex, HIV transmission risk group, concomitant use of protease inhibitors, concomitant use of NRTI and NNRTI, and both CD4 cell count (measured by flow cytometry, Becton Dickinson, San Jose, California, USA) and HIV-1 RNA level (measured by the Amplicor assay, Roche Molecular Systems, Inc., Branchburg, New Jersey, USA), at the start of D4T therapy. The comparisons were performed using the χ2 (categorical variables) or the t-test (continuous variables).
We then compared HIV-1 RNA response and CD4 cell response between the ZDV-experienced and ZDV-naive patients at 3, 6 and 12 months after start of D4T therapy. For each time, a ±1 month window was used with the value closest to the specific timepoint chosen. ZDV-naive and experienced patients were compared with respect to HIV-1 RNA and CD4 cell count (absolute and change from baseline) as well as the percentage of patients who achieved undetectable HIV-1 RNA (<400 copies/ml). Continuous variables were compared using the Wilcoxon rank-sum test. Comparisons of the percentage of patients achieving undetectable HIV-1 RNA was performed using the χ2 test. Patients were included in the analysis at each timepoint if the patient remained on D4T and laboratory results were available.
As the final step in our analysis, we assessed the response by HIV-1 RNA and CD4 cell counts over time for each group of patients, using general estimating equations repeated measures analysis (PROC GENMOD, SAS Institute, Cary, North Carolina, USA), which uses the method of Liang and Zeger . This analysis provides a method for including patients who have some missing data, and as a repeated measure technique, adjusts the variance for multiple measures for a single individual. We performed both a univariate analysis (HIV-1 RNA or CD4 cell count by D4T group) and a multivariate analysis, adjusting for baseline CD4 cell count or HIV-1 RNA and other demographic and clinical variables.
The characteristics of the ZDV-experienced (n=130) and ZDV-naive (n=98) groups are shown in Table 1. There were no demographic differences in age, sex, race/ethnicity or HIV transmission risk group. Fifty-eight per cent of the ZDV-naive patients received a concomitant protease inhibitor, compared with 67% of the ZDV-experienced patients (P=0.18). There was no significant difference between ZDV-naive patients and ZDV-experienced patients in baseline median CD4 cell count or HIV-1 RNA. The median duration of D4T therapy was 175 days in ZDV-experienced and 194 days in ZDV-naive patients (P=0.28). The median duration of ZDV therapy was 563 days. Only 16% of the ZDV-experienced patients had a duration of ZDV use less than 180 days.
No significant differences were found in HIV-1 RNA response at 3, 6 or 12 months between the two groups (Table 2). There was also no difference in the percentage of patients who achieved an undetectable (<400 copies/ml) HIV-1 RNA level (Table 3). In total, an undetectable HIV-1 RNA level was achieved in 49% of ZDV-naive patients and in 51% of ZDV-experienced patients (P=0.79). There were also no significant differences at 3, 6 or 12 months in the median change in CD4 cell count (Table 4).
No significant differences were found between ZDV-naive and ZDV-experienced patients by repeated measures analysis in either the HIV-1 RNA or the CD4 cell response (Table 5). ZDV-naive patients had a mean difference in change in HIV-1 RNA of 0.03log10 copies/ml [95% confidence interval (CI), -0.42 to 0.48] and a mean difference in change in CD4 cell count of 16×106/l (95% CI, -13 to 46) compared with ZDV-experienced patients. By multivariate analysis there were no significant differences after adjusting for concomitant protease inhibitor use and baseline CD4 cell count and HIV-1 RNA. ZDV-naive patients had a mean difference in change in HIV-1 RNA of 0.17log10 copies/ml (95% CI, -0.32 to 0.66) and a mean difference in change in CD4 cell count of 22×106/l (95% CI, -6 to 56) compared with ZDV-experienced patients.
Our analyses of patients from the Johns Hopkins HIV Clinic suggest that the use of D4T after ZDV therapy does not result in a poorer HIV-1 RNA or CD4 cell response compared with using D4T in ZDV-naive patients. In both groups, approximately 50% of patients achieved HIV-1 RNA of <400 copies/ml and the CD4 cell count increased by a median of approximately 80×106/l. Multivariate analyses indicated that there were no differences between the groups‚ responses. Our results suggest that the earlier clinical studies demonstrating a diminished response to D4T in ZDV-experienced patients may not generalize to the current clinical practice setting. The patients in the study by Katlama, et al. received dual NRTI therapy alone . However, in contrast to the study by Katlama, et al., the majority of our patients received concomitant protease inhibitor therapy. It may be that when D4T is used as part of a protease inhibitor-containing regimen, differences in efficacy associated with prior ZDV use are not important enough to translate into differences in HIV-1 RNA or CD4 cell responses. Unfortunately, there were insufficient patients to assess the HIV-1 RNA response with dual NRTI therapy alone.
In addition to the results from clinical trials, in vitro data suggest that the phosphorylation of D4T may be compromised by prior use of ZDV . It is possible that this in vitro result is not of sufficient importance or duration to offset the clinical efficacy of D4T. Unfortunately, we did not have a sufficient sample size to examine the question of whether HIV-1 RNA response to ZDV is compromised by prior use of D4T. Given that ZDV has been licensed for several years longer than D4T, use of D4T prior to ZDV was limited in our database to fewer than 25 patients. There is reason to speculate that just as ZDV use may impair subsequent phosphorylation of D4T, D4T use might also impair ZDV phosphorylation. This can be examined as our sample size increases in the future.
It is possible that we failed to detect differences in HIV-1 RNA response rates in our cohort because of our sample sizes. We had an 80% statistical power to detect a difference of 0.3log10 copies/ml in HIV-1 RNA between the ZDV-experienced and ZDV-naive groups and a 17% difference in the rate of achieving an undetectable HIV-1 RNA. Our differences were much smaller, and no trends occurred over time that suggested emerging differences. Nevertheless, larger sample sizes would be required to detect even smaller differences between the two groups of patients. It is also possible that the use of protease inhibitors masked differences in D4T response rates. However, since the use of a protease inhibitor reflects current treatment guidelines [5,6], our results may be more relevant to current clinical practice.
These results also raise the question of why only half of the patients in this cohort achieved undetectable HIV-1 RNA, a considerably lower proportion than seen in most clinical trials. However, this is consistent with data we have reported for our entire patient cohort  and with data reported from other HIV clinics . Non-adherence to therapy may be one of the most important reasons for this difference, but other factors may also play a role, including selection of antiretroviral regimens, prior antiretroviral therapy, and baseline HIV-1 RNA and CD4 cell count.
In summary, our study suggests that the use of D4T in ZDV-experience patients should not necessarily be a cause for concern. Unlike previous trials in which patients received NRTI alone, most of our patients also received a protease inhibitor. The debate over sequencing of NRTI may be less relevant when they are combined with protease inhibitors. Longer-term follow-up data from ongoing clinical trials will help to clarify this issue further.
1. Katlama C, Valantin M, Matheron S, et al.
: Efficacy and tolerability of stavudine plus lamivudine in treatment-naive and treatment-experienced patients with HIV-1 infection. Ann Intern Med
2. Sommadossi JP, Zhou XJ, Moore J, et al.
: Impairment of stavudine (d4T) phosphorylation in patients receiving a combination of zidovudine (ZDV) and d4T (ACTG 290). Fifth Conference on Retroviruses and Opportunistic Infections.
Chicago, February 1998 [abstract 3].
3. Murray HW, Squires KE, Weiss W, et al.
: Stavudine in patients with AIDS and AIDS-related complex: ACTG 089. J Infect Dis
4. Fischl MA, Richmann DD, Grieco MH, et al.
: The efficacy of AZT in the treatment of patients with AIDS and ARC. N Engl J Med
5. Feinberg MB, Carpenter C, Fauci AS, et al.
: Report of the NIH Panel Define Principles of Therapy of HIV Infection and guidelines for the use of antiretroviral agents in HIV infected adults and adolescents. Ann Intern Med
6. Carpenter CCJ, Fischl MA, Hammer SM, et al.
: Antiretroviral therapy for HIV infection in 1998: updated recommendations of the International AIDS Society-USA Panel. JAMA
7. Liang K-Y, Zeger SL: Longitudinal data analysis using generalized linear models. Biometrika
8. Moore RD, Keruly JC, Gallant JE, Chaisson RE: Response to therapy with protease inhibitors in clinical practice. 12th World AIDS Conference.
Geneva, June-July 1998 [abstract 22347].
9. Deeks S, Beatty G, Cohen PT, Grant R, Volberding P: Viral load and CD4+ T cell changes in patients failing potent protease inhibitor therapy. Fifth Conference on Retroviruses and Opportunistic Infections.
Chicago, February 1998 [abstract 419].