Incidences and risk factors of first-line HAART discontinuation: a limitation to the success of the “seek, test, treat, and retain” strategy?

Abstract

To evaluate the incidence and risk factors of first-highly active antiretroviral therapy (HAART) modifications/interruptions and their causes in a cohort of newly-treated patients by using a competing risk model. In nine centers of the French cohort Dat'AIDS, in 1 year and 2 years of censorship, a competing risk analysis was implemented in HIV1 patients aged 18 years or older first-treated between September 2002 and March 2012. In 4669 patients, 3628 modifications (77.7%) were observed (median: 13.5 months). Cumulative incidence in 1 year: 46.8% [45.4–48.3]; in 2 years: 65.3% [63.8–66.8]. Intolerance (n = 1167; 32.3%): in 1 year, except first-treated from 2002 to 2005, modifications were not different: 2002–2003 (24.6%) 2004–2005 (26.1%), 2006–2007 (19.4%), 2008–2009 (18.8%) and 2010-2011 (15.7%). Women, AIDS patients, and those aged 50 years and older had an excess risk. Therapeutic simplification (n = 1037; 28.6%): in 1 year, except first-treated from 2002 to 2003, modifications were not different: 2002–2003 (9.0%), 2004–2005 (16.0%), 2006–2007 (11.0%), 2008–2009 (15.7%) and 2010–2011 (10.0%). Conversely to injecting-drug-users and AIDS patients, women and first-treated with non-nucleosides had an excess risk. Therapeutic failure (n = 189; 5.2%): contrary to first-treated between 2002 and 2003 or 2008 and 2009, in 1 year as in 2 years, modifications were not different. In 1 year, 1.9% for 2004–2005, 1.6% for 2006–2007 and 1.2% for 2010–2011. Maximum viral load ≥5.0 log₁₀ copies/ml and CD4 <200 cells/mm³ had a high probability. The study of first-HAART modifications suggests that in 1-year follow-up, intolerance incidence in the recent calendar year is still as frequent as the previous period which may constitute a limitation to the success of the seek, test, treat, and retain.

Keywords:

• first-line HAART modification
• “seek, test, treat, and retain” strategy
• competing risk model
• risk factors
• year of first-line HAART

Acknowledgements

The authors wish to thank all the clinicians, data managers, research nurses, participants and clinical research workers who assisted in data collection. This work is one of three carried out as part of a thesis in clinical epidemiology. The opinions expressed in this paper are those of the researchers and are not necessarily those of the Dat'AIDS scientific board.

Funding

The article was funded by the Dat'AIDS scientific committee (2010 grant).

Additional information

Funding

Funding: The article was funded by the Dat'AIDS scientific committee (2010 grant).