1401. A Prediction Model of Pre-treatment HIV RNA Levels in Naïve Thai HIV-infected Patients: An Application for Resource-limited Settings
Session: Poster Abstract Session: HIV: Antiretroviral Therapy
Friday, October 6, 2017
Room: Poster Hall CD
Posters
  • posterIDWeek.pdf (643.2 kB)
  • Background: Use of abacavir (ABC) and rilpivirine (RPV) in the first-line regimen for naïve HIV-infected patients with pre-treatment HIV RNA >100,000 copies/mL is not recommended due to a high rate of treatment failure. If a model could accurately predict pre-treatment HIV RNA levels, it would be a useful tool for the selection ABC or RPV in the first-line regimen.

    Methods: Thai HIV-infected adults enrolled in the TREAT Asia HIV Observational Database (TAHOD) and additional patients of Ramathibodi Hospital were eligible if they had an HIV RNA result at the time of antiretroviral therapy initiation. Factors associated with pre-treatment HIV RNA <100,000 copies/mL were determined by logistic regression. Based on the results of the final model, a prediction model was created.

    Results: A total of 1,223 patients were included in the analysis. Among those in the derivation data set, median [interquartile range (IQR)] age was 36.3 (30.5-42.9) years, median (IQR) CD4 count was 122 (39-216) cells/mm3, and pre-treatment HIV RNA was 100,000 (32,449-229,777) copies/mL. Factors associated with pre-treatment HIV RNA <100,000 copies/mL were anemia [odds ratio (OR) 2.05 vs. no anemia; 95% confidence interval (CI) 1.28-3.27], CD4 count >200 cells/mm3 (OR 3.00 vs. CD4 count <200 cells/mm3; 95%CI 2.08-4.33), and non-heterosexual HIV exposure (OR 1.61 vs. heterosexual HIV exposure; 95%CI 1.07-2.43). No AIDS-defining illness (11.5), no anemia (18.5), age <35 years (11), CD4 count >200 cells/mm3 (27), duration of HIV infection >1 year (9), and weight >50 years (11) were included in the clinical prediction tool scores. A score ≥45 yielded a sensitivity of 45.3%, specificity of 76.7%, positive predictive value of 68.1%, and negative predictive value of 56.1% among patients in the derivation. The area under the receiver-operator characteristic curve was 0.655 (95% CI 0.614- 0.696) and 0.600 (95% CI 0.533-0.667) in the derivation and validation patients, respectively.

    Conclusion: Our final prediction model had poor sensitivity and specificity for predicting HIV RNA <100,000 copies/mL. Further study on a larger population with a greater diversity of data variables available is necessary to improve the model. Pre-treatment HIV RNA remains necessary before ABC or RPV initiation for naïve Thai HIV-infected patients.

    Nisha Pongpech, MD, Faculty of Medicine Ramathibodi Hospital, Mahidol University, Department of Medicine, Bangkok, Thailand, Anchalee Avihingsanon, MD, PhD, HIV-NAT, Thai Red Cross AIDS Research Center, Bangkok, Thailand, Romanee Chaiwarith, MD, MHS, Medicine, Faculty of Medicine Chiang Mai University, Chiang Mai, Thailand, Pacharee Kantipong, MD, Chiangrai Prachnukroh Hospital, Chiang Rai, Thailand, David Boettiger, PhD, The Kirby Institute, UNSW Australia, Sydney, Australia, Jeremy Ross, MD, TREAT Asia, amfAR - The Foundation for AIDS Research, Bangkok, Thailand and Sasisopin Kiertiburanakul, MD, MHS, Department of Medicine, Faculty of Medicine Ramathibodi Hospital, Mahidol University, Bangkok, Thailand

    Disclosures:

    N. Pongpech, None

    A. Avihingsanon, None

    R. Chaiwarith, None

    P. Kantipong, None

    D. Boettiger, None

    J. Ross, None

    S. Kiertiburanakul, None

    << Previous Abstract | Next Abstract

    Findings in the abstracts are embargoed until 12:01 a.m. PDT, Wednesday Oct. 4th with the exception of research findings presented at the IDWeek press conferences.