A cytostatic drug improves control of HIV-1 replication during structured treatment interruptions: a randomized study

OBJECTIVE: To study the effect of highly active antiretroviral therapy (HAART) with and without hydroxyurea (HU) on changes in plasma viral load (VL) set‐point, and on HIV‐1‐specific responses, after five cycles of structured treatment interruptions (STI). METHODS: A group of 20 patients taking HAART for chronic HIV infection with VL < 20 copies/ml were randomized to continue HAART or HAART plus HU for 24 weeks followed by five STI cycles. HU was also stopped in cycles 1‐3 but continued in cycles 4 and 5. The number of individuals maintaining a VL set‐point < 5000 copies/ml during the fifth interruption were determined. RESULTS: VL remained < 5000 copies/ml in eight out of nine patients in the HU group and in four out of ten patients in the HAART group after a median 48 weeks of follow‐up after the fifth interruption ( P=0.039). By STI cycle 5, there was a significant increase in the neutralizing activity (NA), in both magnitude and breadth of the total cytotoxic T lymphocyte (CTL) response and in lymphoproliferative response (LPR) from baseline. No significant differences were observed between HAART and HU groups in NA, CTL and LPR at any time‐point. There were no differences in the NA titers at any time‐point between responder and non‐responder patients. There was a trend for higher CTL and LPR levels in responder patients (P= 0.10). CONCLUSIONS: In this randomized, controlled study of STI with cycles of HAART or HAART plus HU, a lower peak VL rebound and a lower VL set‐point was achieved in patients continuing HU while other drugs were discontinued. HU did not blunt anti‐HIV‐1‐specific responses; however, control of VL did not correlate with anti‐HIV‐1‐specific cellular immune responses.