Improved Lipid Profiles and Maintenance of Virologic Control in Heavily Pretreated HIV-Infected Patients Who Switched from Stavudine to Tenofovir Treatment

Previous studies have suggested that antiretroviral regimens that include tenofovir may be associated with beneficial effects on blood lipids in HIV-1— infected individuals, compared with regimens that include stavudine [1–5]. To investigate the impact of replacing stavudine with tenofovir in combination antiretroviral regimens, we conducted a retrospective, computerized chart analysis of patients from 2 HIV treatment centers caring for 1860 patients in Hamburg, Germany (Praxis St. Georg), and Pavia, Italy (IRCCS Policlinico San Matteo Hospital).

We identified 66 HIV-1— infected patients whose treatment had been switched from stavudine to tenofovir without any additional changes to either their antiretroviral regimen or their concomitant lipid-lowering therapy. No change in lipid-lowering treatment was allowed within 1 month prior to the switch from stavudine to tenofovir, and patients had to have been receiving stable stavudine treatment for at least 1 month. The first switch was performed in August 2001, at which time tenofovir became available within an expanded access program in Germany. Patients were observed until any change in their combination antiretroviral regimen occurred. They were monitored at the time of treatment switch (baseline) and every 3 months thereafter. The time point of the last observation included in the current analysis was 1 August 2004. Plasma HIV-1 RNA load was measured using the Roche Cobas Amplicor HIV1 Monitor 1.5 (lower limit of detection, 50 copies/mL), and CD4⁺ cell count was measured using flow cell cytometry. Cholesterol (total, high-density lipoprotein and low-density lipoprotein) and triglyceride serum levels were determined using standard laboratory methods. Changes in quantitative variables were assessed using Student' t test for paired observations. Pairwise McNemar tests with respect to baseline values were performed to assess the change in viral load (<50 copies/mL vs. ⩾ 50 copies/mL and <400 copies/mL vs. ⩾ 400 copies/mL). Two-sided P-values were calculated.

The 58 men and 8 women had a mean age of 46 years (range, 28–62 years). Forty patients (60%) had advanced HIV-1 disease, defined as Centers for Disease Control and Prevention stage C or a CD4⁺ cell count nadir of <200 cells/µ L. The median duration of prior antiretroviral therapy was 8.7 years (range, 1–13.5 years). The patients were treated with stavudine for a median duration of 54 months (range, 1–95 months; mean, 48.2 months). Reasons for switching therapy from stavudine (>1 reason may apply per patient) included lipodystrophy (32 patients), polyneuropathy (21 patients), hyperlipidemia (11 patients), elevated liver function test values (10 patients), and an elevated plasma viral load (10 patients). Some degree of physician-defined lipodystrophy was present in 79% of patients. Ten patients received concomitant lipid-lowering therapy for a median duration of 18 months (range, 4–47 months).

At the time of treatment switch, 49 patients received protease inhibitor— sparing regimens (comprising tenofovir, 1 or 2 nucleoside reverse-transcriptase inhibitors, and 1 nonnucleoside reverse-transcriptase inhibitor; 27 patients received efavirenz, and 22 patients received nevirapine), 17 patients received regimens containing a protease inhibitor (14 received lopinavir-ritonavir), and 2 patients received regimens containing 3 classes of antiretroviral drugs. The median duration of follow-up after the treatment switch from stavudine to tenofovir was 18 months (range, 4–36 months; mean, 21 months). At the time of the switch, the mean CD4⁺ T cell count was 597 cells/µ L (range, 86–1237 cells/µ L), and CD4⁺ T cell counts remained stable throughout the 18 months following the switch to tenofovir, as shown in figure 1A. The median plasma viral load was <50 copies/mL at the time of switch, with 51 patients (77%) having a load of <50 copies/mL. The proportion of patients with a plasma viral load of <50 copies/mL increased to 91% six months after switching to tenofovir (P =.039). The proportion remained >90% until 18 months after the switch, although these data did not achieve statistical significance because of the smaller patient numbers at later time points ( figure 1B).

View larger version:

• In this window
• In a new window

• Download as PowerPoint Slide

Figure 1

A, Mean CD4⁺ cell counts, with 95% CIs, for the 18 months following the switch from stavudine to tenofovir. B, Proportion of patients, with 95% CIs, with an HIV-1 load (VL) less than the limit of detection (<50 copies/mL).

At the time of the treatment switch to tenofovir, the mean total cholesterol value was 227 mg/dL, and 67% of patients had total cholesterol values greater than the normal level (200 mg/dL). Mean total cholesterol values decreased significantly (mean decrease, 18 mg/dL) within 3 months after substituting tenofovir for stavudine (P =.003) and remained significantly lower than baseline values during the 18 months of follow-up (mean decrease, 36 mg/dL; P =.002), as shown in figure 2A.

View larger version:

• In this window
• In a new window

• Download as PowerPoint Slide

Figure 2

Mean total cholesterol (A) and triglyceride (B) values, with 95% CIs, for the 18 months following the switch from stavudine to tenofovir.

Two of 10 patients were able to discontinue lipid-lowering therapy. Low-density lipoprotein cholesterol values decreased to a lesser degree but did not reach statistical significance. This may be because of the smaller number of patients for whom these measurements were available. The mean triglyceride value was 289 mg/dL at the time of treatment switch, with 56% of patients having elevated values. There was a rapid and significant decrease (mean decrease, 84 mg/dL) in serum triglyceride concentrations within the first 3 months after starting tenofovir (P <.001) ( figure 2B). Triglyceride values subsequently increased and were not significantly different from baseline values at later time points.

This retrospective study supports long-term improvement in blood cholesterol concentrations in patients with long-standing HIV-1 infection whose only treatment change was the substitution of tenofovir for stavudine. These data confirm the findings of 2 previous cohort studies [2, 3]. In these studies, changes in triglyceride values from baseline were not statistically significant. This finding may reflect the difficulty in obtaining fasting blood samples from patients in an observational study. In the prospective GS-903 trial [1], cholesterol and triglyceride values were significantly lower in the tenofovir arm, compared with values in the stavudine arm, after 48, 96, and 144 weeks. After the termination of the 903 trial, 85 patients from the stavudine arm were switched to tenofovir, which resulted in a significant decrease in total cholesterol and triglyceride levels after 48 weeks [5].

A high proportion of patients in our study had signs of lipodystrophy and insulin resistance, which may also affect triglyceride metabolism. Although a few patients reported a subjective decrease of lipoatrophy, most patients and doctors felt that the morphologic changes of fat redistribution syndrome remained stable. As has been shown in other studies (e.g., the Mitox-Study [6] and the RAVE-Study [7]), switching from a thymidine analogue (stavudine or zidovudine) to a nonthymidine analogue (abacavir or tenofovir) is associated with decreased mitochondrial toxicity and insulin resistance, as well as a significant increase in subcutaneous fat [8, 9]. However, the absolute increase was too small to be subjectively felt by the patients. Serial evaluation of insulin sensitivity or objective measurements of fat redistribution (dual-energy X-ray absorptiometry [DEXA] scans) are not performed on a regular basis in our clinical practice and, therefore, such data were not available for this analysis. Thus, our retrospective study cannot answer the important question of changes in insulin sensitivity and/or improvement in fat distribution after a switch from stavudine to tenofovir.

Finally, this study shows that regimens containing tenofovir provide effective control of HIV-1 infection. CD4⁺ cell counts remained stable following the switch, and the proportion of patients with plasma viral loads <50 copies/mL increased following the substitution of tenofovir for stavudine. Although 67% of patients in this study had received previous suboptimal antiretroviral therapy or had documented resistance to nucleoside analogues, no cases of viral rebound occurred following the switch to tenofovir.

• Received June 12, 2005.
• Accepted August 30, 2005.

• © 2005 by the Infectious Diseases Society of America