407. Predictors of Remaining on First-Line Randomized Efavirenz-Containing Regimens for up to Five Years in the AIDS Clinical Trials Group (ACTG): ACTG 384, A5095, A5142, A5202 and the ACTG Longitudinal Linked Randomized Trials (ALLRT) Long-Term Observational Cohort
Session: Poster Abstract Session: HIV - Antiretroviral Therapy
Friday, October 21, 2011
Room: Poster Hall B1

Background: Efavirenz (EFV) is a non-nucleoside reverse transcriptase inhibitor (NNRTI) which, with 2 NRTIs, is a recommended initial antiretroviral (ARV) regimen. With long-term ARV treatment, patients achieve sustained benefits. Understanding predictors for successfully remaining on EFV is clinically useful for selection of initial ARVs.

Methods: ARV-nave participants in 4 ACTG trials were randomized to initial ARV regimens that included EFV + 2/3 NRTIs (n=2,220). Logistic regression, using inverse probability of censoring weights models to reduce loss-to-follow-up bias, was used to identify pre-treatment predictors of remaining on EFV for 1 year (yr) and then, given a participant was taking EFV at yr 1, identifying predictors of continuing on EFV for 2, 3, 4 and 5 yrs. Within class NRTI-switches were not considered EFV failures.

Results: Pre-treatment demographics were: median CD4 227 cells/uL, 33% vRNA>100,000 cp/mL, median age 38 years, 82% male, 40% white, 10% history (Hx) of injection drug use (IDU), and 6% ≥1 sign/symptom ≥grade 3. At yr 1, 554 subjects (25%) were no longer taking EFV; primary reasons for stopping were toxicities, allergic reactions to an ARV and virologic failure. In a multivariable model, factors associated with successfully remaining on EFV at the end of yr 1 were: white race (OR: 1.37; p=0.02 vs. black race), no pre-treatment sign/symptom ≥grade 3 (OR: 1.72; p=0.01) and no IDU Hx (OR: 1.82; p<0.001). Predictors for successfully remaining on EFV at yrs 2-5 were; white race (range of ORs yrs 2-5: 1.47-1.68; p≤0.02); vRNA ≤200 at yr 1 (ORs: 5.05-21.82; p≤0.01) and no sign/symptom ≥grade 3 reported pre-treatment or during yr 1 (ORs: 1.39-1.78; p≤0.09). Additional factors associated (p<0.08) with continuation of EFV for yrs 2-4 included older age; for yr 5 having no interruption of ARV use during yr 1; for yr 3 having no IDU Hx.

Conclusion: In persons randomly assigned to EFV, factors that predicted 1 year and long term EFV use were similar. Whether genetic polymorphisms that impact EFV clearance, social factors, or other issues contribute to differences in EFV continuation for whites and blacks will require further study. People without pre-treatment or yr 1 signs/symptoms were more likely to remain on EFV for 5 yrs.

Subject Category: H. HIV/AIDS and other retroviruses

Marlene Smurzynski, MSPH, PhD1, Kunling Wu, MS1, Ronald J. Bosch, PhD2, Jeffrey T. Schouten, MD, JD3 and Ann C. Collier, MD4, (1)Center for Biostatistics in AIDS Reseach, Harvard School of Public Health, Boston, MA, (2)Harvard School of Public Health, Boston, MA, (3)Office of HIV/AIDS Network Coordination, Fred Hutchinson Cancer Research Center, Seattle, WA, (4)University of Washington, Seattle, WA


M. Smurzynski, None

K. Wu, None

R. J. Bosch, None

J. T. Schouten, Abbott Laboratories: Investigator, Research grant

A. C. Collier, Gilead Sciences: Investigator, Research support
Bristol-Myers-Squibb: Shareholder,
Abbott Laboratories: Shareholder,

Findings in the abstracts are embargoed until 12:01 a.m. EST Thursday, Oct. 20 with the exception of research findings presented at IDSA press conferences.