Background:  ART interruption invariably results in viral rebound, independent of achieved immune reconstitution. Prior immune-based strategies have failed to achieve durable viral suppression.

Methods:  Twenty-three HIV-1-infected, ART-suppressed individuals (viral load <50 copies/mL, CD4 count >450 cells/mL) were randomized to receive 180 (arm A) or 90 (arm B) μg/week of pegylated interferon-α2a (pIFN-α). After 5 weeks of overlapping treatment, ART was discontinued and pIFN-α monotherapy was maintained up to 24 weeks; subjects with virological failure (viral load ≥400 copies/mL) or serious adverse events (SAE) restarted the previous ART regimen. Primary analysis compared the observed proportion of subjects with viral load <400 copies/mL following 12 weeks of pIFN-α monotherapy to an expected proportion of 0.09 (exact binomial test, SAE treated as virologic failures). Secondary analysis: compare the proportion of subjects sustaining viral load <400 copies/mL through week 24 with historical controls (ART interruption, no pIFN-α); measure integrated peripheral blood mononuclear cell (PBMC) HIV DNA levels by Alu-PCR as correlates of viral control.

Results:  Of 23 patients enrolled, 3 were excluded from the primary analysis (1 lost to follow up, 1 incarcerated, 1 grade 2 LFT). Nine of the 20 remaining subjects (45%) maintained a viral load <400 copies/mL through week 12, which was significantly greater than the expected proportion of 0.09 (arm A, p = 0.0088; arm B, p = 0.0010, combined arms, p <0.0001). Four of these 9 subjects had viral load <48 copies/mL (p = 0.0027, combined arms). A sensitivity analysis including the 3 excluded subjects confirmed the result (39%, arm A, p = 0.0178, arm B, p = 0.0017, combined arms, p <0.0001). The 9 subjects with viral load <400 copies/mL at week 12 continued pIFN-α for up to 12 more weeks; at week 24, 6 out of 9 (67%) subjects had viral load <400 copies/mL on pIFN-α monotherapy. CD4 counts remained >300 cells/μL in all subjects. SAE included neutropenia (n = 1) and depression (n = 3). Upon ART resumption all subjects were successfully suppressed at viral load <48 copies/mL. A net decrease (p = 0.01) in integrated HIV DNA copies/CD4⁺ T cell between baseline and week 12 was observed in all subjects who remained suppressed viral load <400 copies/mL, but not in those with viral breakthrough >400 copies/mL.

Conclusions:  We provide the first proof-of-concept that pIFN-α-based immunotherapy results in durable control HIV replication following ART interruption. We also show, for the first time, that pIFN-α-mediated viral control impacts HIV.