Program Schedule

The Performance of Quantitative HIV-1 RNA Testing (Viral Load) to Diagnose Acute HIV Infection in the New HIV Laboratory Testing Algorithm

Session: Oral Abstract Session: HIV Testing and Cascade of Care
Thursday, October 9, 2014: 8:30 AM
Room: The Pennsylvania Convention Center: 107-AB

Background: In the new HIV testing algorithm, specimens reactive on an HIV antigen/antibody combination immunoassay (IA) but negative on a supplemental confirmatory test are tested for HIV-1 RNA to distinguish acute HIV infection (AHI) from false positive IA results. We evaluated the diagnostic performance of quantitative HIV-1 RNA viral load (VL) and IA signal-to-cutoff ratios (S/COs) to diagnose AHI in this setting.

Methods: The STOP study is a prospective study evaluating methods to detect AHI. Participants (age > 12 years) from 12 HIV testing sites in sexually transmitted infection clinics and community-based programs were screened for HIV with an IA (Architect, Abbott). Repeatedly reactive IA results (S/CO ≥1.00) were confirmed with an HIV-1/HIV-2 antibody differentiation assay (Multispot, Bio-Rad). Discordant IA (reactive) and Multispot (negative) results were tested with a FDA-approved qualitative HIV-1 RNA assay (Aptima; qual-RNA) and/or VL (Abbott m2000 or Roche Amplicor v1.5).

Results: Among 86,929 participants who received HIV testing from September 2011 to October 2013, 191 (0.22%) had discordant IA (reactive) and Multispot (negative) results of whom 101 (52.9%) were diagnosed with AHI and 90 (47.1%) were determined to be HIV negative. Among 70 participants who had both qual-RNA and VL performed, 69 (98.6%, 95% Confidence Interval [CI] 90.6 99.8%) had concordant results (49 both reactive, 20 both negative). The one discordant result (non-reactive qual-RNA but detectable VL) was determined to be consistent with AHI on repeat testing. Among participants with AHI who had a VL test (n=100), the estimated geometric mean VL was 1,408,316 (95% CI 915,166 2,167,704) copies/ml and only two of the VLs were less than 10,000 copies/ml (Figure). Among participants determined to be HIV negative who had a VL test (n=50), all VL results were undetectable. IA S/COs among participants with AHI (median 23.68, Interquartile Range [IQR] 7.44 60.43) were also significantly higher than S/COs among those determined to be HIV negative (median 1.91, IQR 1.37 2.93) (p <0.0001).

Conclusion: VL, which is more widely available and clinically useful than qual-RNA, consistently distinguished AHI from false positive IA results in the context of the new HIV testing algorithm.


Hsiu Wu, MD1, Stephanie Cohen, MD, MPH2, Emily Westheimer, MS3, Cindy Gay, MD, MPH4, Laura Hall, MPH1,5, Charles E. Rose, PhD1, Lisa B. Hightow-Weidman, MD, MPH4, Jie Fu, PhD6 and Philip J. Peters, MD1, (1)Division of HIV/AIDS Prevention, CDC, Atlanta, GA, (2)San Francisco Department of Public Health, San Francisco, CA, (3)New York City Department of Health and Mental Hygiene, New York, NY, (4)University of North Carolina, Chapel Hill, NC, (5)ICF International, Atlanta, GA, (6)Bureau of Public Health Laboratory, New York City Department of Health and Mental Hygiene, New York, NY


H. Wu, None

S. Cohen, None

E. Westheimer, None

C. Gay, None

L. Hall, None

C. E. Rose, None

L. B. Hightow-Weidman, None

J. Fu, None

P. J. Peters, None

Previous Abstract | Next Abstract >>

Findings in the abstracts are embargoed until 12:01 a.m. EDT, Oct. 8th with the exception of research findings presented at the IDWeek press conferences.

Sponsoring Societies:

© 2014, All Rights Reserved.

Follow IDWeek