AEGiS-Glasgow [PL8.1] Treatment interruptions: where are we?

Eighth International Congress on Drug Therapy in HIV Infection

Glasgow, UK - 12-16 November 2006

[PL8.1] TREATMENT INTERRUPTIONS: WHERE ARE WE?

Int Cong Drug Therapy HIV 2006 Nov 12-16;8:Abstract No. PL8.1

Bernard Hirschel
Geneva University Hospital, Geneva, Switzerland

PURPOSE OF THE STUDY: The SMART study randomized more than 5000 patients with CD4 counts above 350/µl to either a treatment interruption group (STI, with treatment to cease at CD4 counts above 350/µL, and to start again below 250/µl), or to continuous treatment. Results confounded expectations. Previously, acute retroviral syndrome and development of resistance were the perceived disadvantages, less toxicity the perceived advantages of STIs. In reality, neither resistance nor ARS proved to be a problem, but surprisingly, there was an excess of disease in the STI group. This excess included not only AIDS-defining conditions such as oesophageal candidiasis and Pneumocystis carinii pneumonia, but also deaths not obviously related to AIDS, and even cardiovascular, renal, and hepatic disease. The latter were thought to represent toxicity of anti-retroviral drugs, but because they were more frequent in the group which received less drugs, this notion must obviously be revisited.

Are SMART’s results in contradiction with other, smaller STI studies? Some, such as Trivacan and DART, with criteria for starting and stopping ARVs similar to SMART’s, also show an excess of clinical events in the STI groups. However, other studies, using different criteria, do not show such an excess. Some of these relative dwarfs are too small to yield useful information regarding clinical endpoints. The Staccato study accumulated 484 patient years in the STI group; had SMART’s event rates applied to Staccato, 16 AIDS-defining events or deaths would have occurred, whereas in reality only one was observed. This difference remains unexplained. Criteria for starting treatment again (a CD4 count of 350 in Staccato, 250 in SMART) may explain some, but not all of the difference.

SMART is large enough to allow sub-analyses of particular populations. One can, for instance, ask whether differences between the STI and continuous treatment groups would have persisted if treatment had started again at higher levels than 250 CD4 cells. The answer is that differences between the STI and continuous treatment groups persist however you slice the pie.

In conclusion, SMART’s results imply that STIs will always be associated with a slightly increased risk of AIDS-defining events and death. It stands to reason, however, that the shorter the STI, and the higher the CD4 count, the lower the excess risk. At some point, the cost of continuing treatment at high CD4 counts counterbalances the increased risk of interrupting it. Where is that point? Economical, political and psychological, rather than medical arguments will decide.

Plenary Session: Oral Papers

2006-11-12
PL8.1

Copyright © 2006 - Thomson ACUMED® All rights reserved. Thomson ACUMED® is an intelligent and innovative medical marketing and communications agency – a new division of The Gardiner-Caldwell Group Ltd, part of The Thomson Corporation, located in Tytherington, UK.

Reproduction of this abstract (other than one copy for personal reference) must be cleared through the authors.