The Use and Response to Second Combination ART Regimens in EuroSIDA
Alessandro Cozzi-Lepri*1, M Cunnington2, A Lazzarin3, B Clotet4, B Knysz5, P Gargalianos6, C Katlama7, A Karlsson8, A Phillips1, J Lundgren9, and EuroSIDA Study Group
1Royal Free and Univ Coll Med Sch, London, UK; 2GlaxoSmithKline R&D, UK; 3San Raffaele Vita-Salute Univ, Milan, Italy; 4irsiCaixa Fndn and Lluita contra la SIDA Fndn, Badalona, Spain; 5Wrocaw Med Univ, Poland; 61st IKA Hosp, Athens, Greece; 7Hosp Pitie-Salpetriere, Paris, France; 8Karolinska Univ Hosp, Stockholm, Sweden; and 9Copenhagen HIV Prgm, Hvidovre, Denmark
Background: Little is known about the response to
second line combined ART (cART), despite the fact that newly emerging therapies
are likely to be initially aimed at this patient group. This study aims to
describe the incidence and predictors of virological failure to second line
cART in EuroSIDA.
Methods: All patients of EuroSIDA who started cART (≥3-drug-containing
regimen) from ART-naïve, experienced confirmed virologic failure to this first
regimen and subsequently started a second cART regimen were included. The time
of starting ≥1 new drug as part of cART after a confirmed virologic
failure of the first regimen was defined as baseline; CVF was defined as the time
of the first of 2 consecutive viral loads >400 copies/mL ≥6 months
after starting the first cART. Patient follow-up ended with the confirmed virologic
failure of the second regimen or at the date of last available viral load.
Kaplan-Meier curves were used to estimate the time to confirmed virologic
failure and proportional hazards Cox regression (stratified by centre) to
assess predictors of confirmed virologic failure.
Results: Of 695 patients, 77% were male and 84%
Caucasian with a median age of 38 years at baseline. The median baseline CD4
count was 275 cells/mm3 (IQR 165 to 417) and median baseline viral
load was 4.24 log copies/mL (IQR 3.54 to 4.89). The median time from initiation
of first cART to baseline was 26 months (IQR 16 to 42). The percentage of
patients who had experienced virologic failure to 1, 2, and 3 classes before
baseline were 2%, 79%, and 19%, respectively. Patients started their second
regimen on average in 2001 with 2 nucleosides plus: a NNRTI (n = 205,
29%), a single protease inhibitor (PI) (164, 24%), a PI/ritonavir (r) (196, 28%),
abacavir (20, 3%), or other regimens containing >3 drugs (110, 16%); 31%
started a single new drug class, 1% started 2 classes, and the remaining
started ≥1 new drug within the same class. Overall, 310 patients (45%) experienced
confirmed virologic failure on their second regimen and the Kaplan-Meier median
time to failure was 39 months (95%CI:25 to 48). Factors independently associated
with the risk of confirmed virologic failure to second cART are shown in the table.
Conclusions: This study identified a high viral load
at initiation, low predicted virological activity of the new drugs started, not
achieving a viral load ≤400 on first cART and shorter time between
initiating the first and second regimen as the main predictors of confirmed virologic
failure to second cART. The latter 2 factors may reflect the fact that non-adherent
patients failed quickly both regimens, though the role of adherence and drug
resistance warrants further investigation.