![]() |
14th International AIDS ConferenceBarcelona, Spain - July 7-12, 2002 |
Int Conf AIDS 2002 Jul 7-12; 14:(abstract no. C10680)
Sterne JA, Grabar S, d'Arminio Monforte A, Rickenbach M, de Wolf F, Lundgren JD, Fusco J, Miller V, Raffi F, Dabis F, Hogg RS, Lampe F, Gill MJ, Salzberger B, May M, Egger M
ART Cohort Collaboration, University of Bristol, United Kingdom
BACKGROUND: Measurements of prognostic factors some time after the start of HAART may be useful in predicting progression because they reflect initial treatment response. The importance of such measurements compared to baseline values is unclear.
METHODS: The ART Cohort Collaboration includes drug naïve patients initiating HAART from 13 cohorts in Europe and North America. We used Cox regression to estimate the association of CD4 count and HIV-1 RNA, measured at baseline and after 6 months, with progression to AIDS and death. Patients with missing measurements at 6 months were excluded.
RESULTS: Measurements of CD4 count and HIV-1 RNA were made around 6 months after the start of HAART (median 6.4, interquartile range 5.5 to 7.8 months). 418 of 9696 patients followed for >6 months progressed to AIDS or death during 15130 years of follow up (185 died). CD4 cell count and HIV-1 RNA at 6 months were strongly associated with AIDS-free survival. Although there were strong associations with baseline levels in univariable analyses, these associations disappeared after controlling for levels at 6 months (see tables, which show hazard ratios and 95% CI after controlling for other prognostic factors). [table: see text] The effect of 6 month CD4 was greater in individuals with detectable than undetectable RNA at 6 months (interaction P=0.007). Age>50, history of IDU, and CDC stage C events before the start of HAART or 0-6 months after starting HAART were associated with decreased AIDS-free survival and survival. There was no evidence of associations with gender or year of starting HAART.
CONCLUSIONS: In predicting prognosis 6 months after starting HAART, baseline levels of CD4 and HIV-1 RNA are of little or no relevance once current levels of these factors are taken into account. Prognostic models based on these data have been developed (see May et al and Chene et al).
020707
C10680
Copyright © 2002 - International AIDS Society (IAS). Reproduction of this abstract (other than one copy for personal reference) must be cleared through the IAS.