799
Improved Toxicity Profile of Recent HAART Regimens
Luuk Gras*1, C Smit1, A van Sighem1, P Reiss2, F de Wolf1,3, and F de Wolf1,3
1HIV Monitoring Fndn, Amsterdam, The Netherlands; 2Ctr for Immunity and Infection, Academic Med Ctr, Univ of Amsterdam, The Netherlands; and 3Imperial Coll Sch of Med, London, UK
Background: Adverse effects of HAART can result in poorer adherence
or discontinuation of treatment, thereby increasing the chance of emergence of
resistance. We investigated the incidence of toxicity-driven therapy changes in
the Netherlands ATHENA observational cohort and compared differences in time to
the first toxicity driven therapy changes between HAART combinations in
relation to calendar time of first starting HAART.
Methods: The incidence of toxicity-driven therapy changes
during the first 3 years after the start of HAART was analyzed using Poisson
regression and related to baseline characteristics in 6835 treatment-naïve patients
starting HAART between July 1, 1996 and December 31, 2005. Time to first toxicity-driven
therapy changes was studied using Cox models in 2345 patients starting HAART
including either efavirenz (EFV), nevirapine NVP), or lopinavir/ritonavir (LPV/r)
combined with either tenofovir/lamivudine (TDF+3TC) or zidovudine/lamivudine
(AZT+3TC).
Results: Patients were followed for 16,491 person-years of
which 14,858 person-years on HAART. The overall incidence of toxicity-driven
therapy changes was 23.6/100 person-years on HAART (95%CI 22.8 to 24.4). The
incidence of toxicity-driven therapy changes was highest (67.7/100 person-years
on HAART) during the first 3 months after starting HAART and declined to 29.5/100
person-years on HAART between 3 and 6 months and 13.1/100 person-years on HAART
between 24 and 36 months (p <0.0001).
In multivariate analyses, patients starting HAART in 2002-2005 had 0.77 (0.73
to 0.83, p <0.0001) times lower
risk compared to those starting in 1997-1999. The risk for patients starting in
2000-2001 did not differ from those starting in 1997-1999 (p = 0.36). The incidence of toxicity-driven therapy changes was
higher in patients who had high pre-HAART CD4 cell count (p = 0.01), pre-HAART HIV-RNA level ≥4 log10
copies/mL (p = 0.0003) or a prior CDC
stage C event (p = 0.0003). The
incidence was higher in female patients (p
<0.0001), patients infected through heterosexual contact compared with
homosexual (p = 0.0002) and older
patients (p = 0.0001). Patients
starting with TDF+3TC+NVP or TDF+3TC+EFV were found to be the least likely to experience
a toxicity-driven therapy changes compared to AZT+3TC+NVP; adjusted HR 0.34 (0.19
to 0.60, p = 0.0002) and 0.65 (0.47
to 0.89, p = 0.008), respectively.
The other regimens were not significantly different from each other.
Conclusions: The incidence of changing HAART as a result of
toxicity has declined over time and particularly with the introduction of novel
regimens after 2002. HAART regimens including TDF+3TC combined with EFV or NVP
compared with those including AZT+3TC were found to be better tolerated.
|
