1045 
Decreasing HIV Incidence and Prevalence at the Johns Hopkins Emergency Department with a Concurrent Increase of Virally Suppressed HIV-infected Individuals
Oliver Laeyendecker*1,2, A Oliver2, J Neal2, J Gamiel2, C Krauss2, S Eshleman2, M Owen3, J Shahan2, G Kelen2, and T Quinn1,2
1NIAID, NIH, Baltimore, MD, US; 2Johns Hopkins Univ Sch of Med, Baltimore, MD, US; and 3CDC, Atlanta, GA, US
Background: HIV prevalence at the Johns Hopkins Hospital
Emergency Department (JHHED) has been previously reported for 2001 (12%) and
2003 (11%), along with the observation that viral suppression affects
titer-based cross-sectional HIV incidence assays.
Methods: An identity-unlinked serosurvey was
performed on 4475 adults in the JHHED (June 2007 to September 2007). Samples
were tested for HIV by ELISA and confirmed by Western blot. HIV+
samples were tested for viral load (Roche AMPLICOR v1.5) and for the presence
of antiviral drugs (ARV) by high-performance liquid chromatography (HPLC). We
used 2 incidence assays: Calypte HIV-1 BED Incidence EIA (BED), and an avidity
assay-based on the BioRad HIV 1/2+O EI`A (Avidity). To determine the frequency
and nature of false reactivity with each incidence assay, we tested the
following controls: chronic infected (CD4 200 to 400, not on ARV, n = 74),
advanced AIDS (CD4 <50, n = 140), ARV suppression (CD4 >500 and viral
load<50, n = 134), and elite suppressors (viral load <50 not on ARV, n = 16).
Results: JHHED HIV prevalence declined to 7.2% in
2007. The percentage of HIV+ individuals with viral suppression
(<400 copies/mL) increased from 21% in 2001 and 24% in 2003 to 33% in 2007. Of
HIV+ persons, 8.7% (28 of 321) were virally suppressed with no
evidence of ARV. The incidence estimate for BED (3.4%) was much higher than the
incidence estimate for Avidity (0.3%). Half (30/60) of the samples that tested
incident by BED had undetectable viral loads, while all (4 of 4) of the samples
identified as incident by Avidity had detectable viral loads. The frequency of
misclassification of control samples as newly-infected was: chronic infection
(7 of 74 by BED, 0 of 74 by Avidity), advanced AIDS (37 of 140 by BED, 2 of 140
by Avidity), ARV suppression (28 of 134 by BED, 0 of 134 by Avidity), and elite
suppressors (9 of 16 by BED, 0 of 16 by Avidity).
Conclusions: Virally suppressed individuals are
increasing in the JHHED and outnumber recently infected individuals. Viral
suppression can affect antibody concentration-based HIV incidence assays, such
as BED, leading to misclassification of samples as incident. BED significantly
overestimated HIV incidence in the JHHED, whereas the Avidity assay was
unaffected. Use of these assays for estimating HIV incidence in populations
that include virally-suppressed individuals should be used with caution and
confirmed by the Avidity assay. These data highlight the need for inclusion of
additional data (previous test history, ARV use) when calculating HIV incidence
using cross-sectional assays that are based on antibody titer.
|
