Decrease in Lipoatrophy in a Pilot Study Using a Short-Term Treatment Interruption Strategy for 48 Weeks in São Paulo,Brazil

E ditor: In the absence of a cure for HIV infection, patients must cope with long-term exposure to daily antiretroviral agents, which ultimately leads to pill fatigue and toxicities. Common toxicities include impaired glucose and lipid metabolism and body fat redistribution. Furthermore, for developing countries that need to provide treatment to a large number of infected individuals, the cost of antiretroviral agents is also a limitation.

Recent studies investigating treatment interruption and reinitiation guided by CD4⁺ T cell count thresholds demonstrated that the reemergence of viremia was associated with a number of events, including higher mortality rates and increased risk of serious non-AIDS diseases. ¹ Time-cycled strategies using long-term interruptions of 1 month or more increase the risk of selecting resistant HIV-1 strains. ² However, it is not clear whether short-term treatment interruptions lead to consistent deleterious viremia among patients. We therefore conducted a pilot study in which patients were treated for 7 days followed by 7 days of treatment interruption for 48 weeks among HIV-1-infected individuals with evidence of lipoatrophy. We hypothesized that toxicities would decrease with minimal loss of the benefits of antiretroviral treatment and minimal risk of the development of antiretroviral resistance.

Ten individuals using stable antiretroviral regimens with viral loads bellow 400 copies/ml for at least 6 months who visited the outpatient clinics of the Federal University of São Paulo were recruited and followed from September 2003 to September 2004. All patients were taking two nucleoside analog reverse transcriptase inhibitors (NRTIs) and five of the patients were taking efavirenz, and the remaining patients were taking protease inhibitors. Regarding NRTI use, six individuals were using d4T with 3TC (5) or ddI (1) and four individuals were using a fixed dose combination of zidovudine and 3TC. The study was approved by the institutional review board, and informed consent was obtained from all participants.

Basal and monthly viral loads were measured, and CD4⁺ T cell counts were determined at baseline, 24 weeks, and 48 weeks. Lipoatrophy and bone mineral density were evaluated at baseline and at 24 weeks and 48 weeks. Blood tests for safety were collected monthly. Six patients were male; the mean age of the patients was 40 years old (34 to 55) and the mean duration of antiretroviral use was 7.2 years (5 to 11). Six individuals exhibited viral loads above the detection limit for at least the time point, with 19 out of 120 viral load determinations during the study period being above the detection limit (15.8%). The mean and median viral loads during viremic episodes were both 3.4 log₁₀ copies/ml (2.6 to 5.4 log₁₀ copies/ml). The mean and median CD4 counts were 536.4 and 602 cells/mm³, respectively (from 333 to 696) at baseline, 622.3 and 572 cells/mm³, respectively (from 346 to 942) at 24 weeks, and 588.6 and 556.5 cells/mm³, respectively (from 347 to 1227) at 48 weeks. All patients were able to maintain viral loads below the detection limit upon resuming continuous treatment with the same regimen for an additional 48-week period of follow-up after the 48-week experimental period.

Anthropometric measures using the Durnin and Womersley equations ³ revealed an increase in the percentage of total body fat from baseline to 48 weeks from 24.1% (SD 6.6%) to 28.9% (SD 7.8%; Bonferroni Corrections for Multiple Comparison, p=0.001), although it is not clear whether the presence of peripheral lipoatrophy in these patients could modify this equation. The percentage of total body fat determined by bioimpedance increased from 20.3% (SD 8%) at baseline to 22.4% (SD 9.5%) at 48 weeks (Bonferroni Corrections for Multiple Comparison, p=0.007). The total body fat evaluated by DEXA also increased from 20.3% (SD=8) to 28.4% (SD=10.1) from baseline to week 48 (p=0.001). Additionally, the body fat measured using DEXA on the arms increased from 1.4 (SD=0.8) to 1.8 (SD=1) from baseline to 48 weeks (Bonferroni Corrections for Multiple Comparison, p=0.01) whereas the body fat in the legs increased from 3.3 (SD=2.3) to 4.0 (SD=2.7) from baseline to week 48 (p=0.02). No statistically significant differences were observed in the bone mineral density or the levels of glucose, lipids, and other markers from baseline to 48 weeks (data on file).

This pilot study demonstrated that lipoatrophy can be reduced with a 50% decrease in antiretroviral exposure over a 48-week period without reducing the benefits of the antiretroviral treatment. Although low-level viremia was detectable in 15.8% of the monthly measurements, there were no repercussions with respect to the ability to fully suppress viremia thereafter. Interestingly, viremia was not detected in the patients taking efavirenz-based regimens, which is consistent with the prolonged half-life of this drug. ⁴ This result is indicative of the forgiveness of efavirenz-based regimens, pointing out that the development of resistance to this drug in real life may be related to long-term interruptions of treatment. In fact, studies evaluating even shorter treatment interruption strategies using efavirenz-based regimens (5 days on and 2 days off) confirm the potential for maintaining viral suppression. ^5,6

We recognize that this study is limited by the small sample size and the uncontrolled design and by the fact that CD4⁺ and CD8⁺ T cell activation, which may increase during low-level or residual viremia, may be a source of harm in treatment interruption strategies. However, we believe that for a limited number of patients who experience long-term side effects of antiretrovirals and are able to fully suppress replication with lower antiretroviral exposure, similar strategies should be further evaluated as a treatment option.