Background:  Low levels of free plasma virus are common in patients successfully treated with HAART. The effect of repeated treatment interruptions on this residual viremia has not been reported. We here describe changes in HIV RNA <50 copies and in HIV DNA in patients enrolled in a substudy of the ISS PART clinical trial. 

Methods:  The ISS PART was a randomized, controlled trial comparing 24 months of continuous (arm A) with 24 months of intermittent HAART (arm B, 5 structured treatment interruptions of 1- to 3-month duration, each followed by 3 months therapy). In this substudy, 33 arm A and 25 arm B patients participated where residual viremia (assay cut-off:  2.5 copies HIV RNA/mL) and HIV DNA (assay cut-off: 10 copies/106 peripheral blood mononuclear cells [PBMC]) were studied at baseline and at month 24 (3 months after the end of the last treatment interruption in arm B). Non-parametric tests were used to compare frequency distributions.

Results:  At baseline, arm A and B patients had similar demographic, clinical, and immunovirologic characteristics. All of them were on first line HAART and had HIV RNA <50 copies/mL. Median CD4 cell count/mm³ was 727 (range 339 to 1459), with no difference between arms. The proportion of subjects with HIV RNA ≤2.5 copies was 51.5% in A and 60% in B (p = 0.52). HIV DNA values ≤10 copies/10⁶ PBMC were found in 45.5% of arm A and in 56% of arm B subjects (p = 0.42). In arm B median time off-therapy was 274 days (range 237 to 298). After 24 months, plasma HIV RNA values were <50 copies/mL in all patients, with 54.5% subjects in arm A and 32% in arm B having ≤2.5 copies/mL (difference arm A – arm B = 22.5; 95%CI –2.4; 47.1, χ² test, p = 0.09). The analysis of percentile distributions confirmed a different trend in the 2 arms, with an increase of ≥14.4 copies in 25% arm B subjects and a reduction of ≥10.2 copies in 25% arm A subjects (Mann-Whitney test, p = 0 .024). No significant differences from baseline were seen in the proportion of subjects with ≤10 copies HIV DNA/10⁶ PBMC. No correlation was found between HIV RNA and HIV DNA values.

Conclusions:  In our population, repeated interruptions of HAART over 24 months induced a moderate increase in residual viremia, without affecting  HIV-1 proviral DNA load in PBMC. This may reflect increased viral release from reservoirs during off-treatment periods or ongoing replication. The implications of this finding in terms of viral evolution and long-term virologic control deserve further investigation.