357   SSITT: A Prospective Trial of Strategic Treatment Interruptions in 128 Patients.

C. Fagard¹, M. Lebraz¹, H. Gunthard⁴, C. Tortajada⁸, F. Garcia⁸, Battegay³, H. J. Furrer⁵, P. Vernazza⁶, E. Bernasconi², L. Ruiz⁸, A. Telenti⁷, A. Oxenius⁹, R. Phillips⁹, S. Yerly¹, J. Gatell⁸, R. Weber⁴, T. Perneger¹, P. Erb³, L. Perrin¹, and B. Hirschel*¹for the Swiss HIV Cohort Study.
¹Geneva, Switzerland;²Lugano, Switzerland;³Basel, Switzerland;⁴Zürich, Switzerland;⁵Bern, Switzerland;⁶St. Gall, Switzerland;⁷Lausanne, Switzerland;⁸Barcelona, Spain; and⁹Oxford, UK.

Background:HIV-specific immune responses (IR) decrease during HAART. Rebounds during treatment (Rx) interruption may stimulate IR and eventually permit discontinuation of HAART.

Methods:The Swiss-Spanish Intermittent Treatment Trial (SSITT) enrolled patients (pts) who were ART-naive before HAART, without treatment failure during HAART, with VL <50 for >6 months, and NNRTI-naive. Rx is stopped for 2 wks and started again for 8 wks, for four cycles. Pts who fail to reach VL <50 upon re-treatment are excluded. At wk 40, Rx is definitely suspended unless VL rebounds >5000. Endpoints: amplitude of rebounds and N of pts with VL <5000 at wk 52. VL were measured by Roche HIV Monitor.

Results:128 patients were enrolled between April 1999 and March 2000. Before HAART: median CD4 count 388, median VL 4.5 logs. Median duration of HAART before SSITT was 26 months, and median CD4 at start of SITT was 727. 80 pts have had four Rx stops, and 55 have reached week 52. Viral rebound to >50 occurred in 76% of the first treatment stops and in 79% of the fourth treatment stops (P = NS). The median VL of later rebounds was similar to the median of the first rebound. There was no significant change in CD4 counts. 24/128 pts (19% so far) have not lowered their VL to <50 after re-treatment and were excluded; one of these developed resistance and required alternative therapy. High pre-ART VL and low CD4 counts, as well as high rebounds, predicted such failure (RH of 3.5 per log increase in rebound). In ITT analysis, 9/54 (17%, 95% CI 8—30%) had a VL <5000 after 12 weeks off therapy at wk 52, 3/55 (6%) <50. Many of these responders seem to have started HAART early after diagnosis/infection.

Conclusions:After four structured treatment interruptions, viremia remained <5000 copies after 12 weeks without therapy in about 1 in 6 patients. Immunologic and virologic correlates, as well as durability, of such responses will be presented.