361   CD4 Count Changes in People Interrupting HAART Therapy After a CD4 Count Increase.

A. Phillips*, M. Youle, M. Tyrer, F. Lampe, C. Chaloner, N. Penlington, A. Drinkwater, C. Sabin, A. Mocroft, and M. Johnson.
Royal Free Ctr. for HIV Med., London, UK.

Background:Little is known about the rate of decline in CD4 count after therapy interruption in people who have achieved a CD4 count benefit on HAART.

Methods:We identified 68 patients from the Royal Free Clinic who had started HAART (>3 drugs including a PI, NNRTI, or abacavir) with a median CD4 count of 157/mm3and achieved a >50/mm3increase (median increase 164/mm3) in CD4 count before interrupting therapy entirely for at least 4 weeks (median >24 weeks). The median viral load at HAART initiation was 5.30 log copies/mL and at interruption of therapy was 3.95 log copies/mL (IQR 2.6—4.7).

Results:The median duration of HAART before interruption was 44 weeks. There was a 24% cumulative probability (Kaplan-Meier estimate) of restarting HAART by 12 weeks, 46% by 24 weeks and 60% by 48 weeks after interruption. The median CD4 count at interruption was 350/mm3. The median off-therapy (last value before restarting or before week 24—whichever was first) CD4 count was 210/mm3(median reduction during therapy interruption 121/mm3). a value taken a median of 15.5 weeks after interruption of therapy. At the same time the viral load had increased to a median 5.40 log copies/mL (median rise 1.2 log copies/mL after interruption of therapy). The decline in CD4 count left patients a median of 39/mm3above the pre-HAART CD4

count at this time (range 374/mm3below to 656/mm3above). The Kaplan-Meier estimate of the median time for the CD4 count to decline back to the pre-HAART CD4 count was 24.6 weeks (for this analysis follow-up time was right-censored if therapy was restarted and at last available CD4 count), but therapy was more likely to be reinitiated in those with lower CD4 counts (RH of re-initiation 1.27 per 100 cell lower latest CD4 count).

Conclusions:Decline in CD4 count after interruption of HAART occurs at a highly variable rate. While very rapid declines frequently occur, the CD4 count remains above the pre-HAART baseline for over 24 weeks in a substantial proportion.

© 8th Conference on Retroviruses and Opportunistic Infections