Acute HIV infection (AHI) is defined as the stage of human immunodeficiency virus (HIV) infection when individuals have actively replicating virus but do not yet have detectable antibodies to the virus (6
). The Centers for Disease Control and Prevention (CDC) conducted a prospective study to investigate the utility of pooled nucleic acid amplification testing (NAAT) to identify AHI cases in targeted and routine HIV screening programs (4
). In that study, plasma specimens from individuals who received a negative result on an HIV antibody screening test were pooled using a one-stage, 1:16 pooling scheme and NAAT was conducted on the 16-member pools using the Aptima HIV-1 RNA qualitative assay (Gen-Probe Inc., San Diego, CA) (2
). The 16-member pooling scheme was chosen based on the procedure specified by a FDA-approved donor-screening assay that uses the same technology as the Aptima assay. This strategy was evaluated in a separate validation study in which plasma samples containing 1,000 to 5,000 copies/ml of a well-characterized HIV-1 virus standard were consistently identified from 16- and 32-member pools tested in multiple replicates using the Aptima assay (1
), indicating that this method was sufficiently sensitive for AHI detection.
The 1:16 plasma pooling scheme was used successfully in the CDC AHI study to identify HIV-1 RNA-reactive pools from which individual cases of AHI were verified by additional NAAT and serologic testing (4
). Larger pool sizes have been used effectively to detect AHI by other NAAT methodologies (5
), and our previous results (1
) suggested that the Aptima assay might be capable of detecting AHI using higher-level pooling schemes. Our objective was to test a 128-member pooling strategy on true clinical samples collected in real time from individuals who sought testing in the early stages of HIV-1 infection. Therefore, we conducted a retrospective study to determine if the AHIs that were identified in the CDC AHI study by NAAT screening of 16-member, seronegative plasma pools would have been detected if a 1:128 pooling scheme had been employed for the initial NAAT screening.
Remnant material from 16-member plasma pools that were originally created for the prospective CDC AHI study and stored at −20°C was used to create new 128-member pools. Each original 16-member plasma pool consisted of 100 μl of plasma from 16 different HIV antibody-negative individual patient specimens. Twenty-one original 16-member pools that were previously confirmed to be positive for HIV-1 RNA had sufficient sample volume available for further evaluation. Each reactive 16-member pool had been verified to contain plasma from one HIV-1 RNA-positive sample and 15 HIV-1 RNA-negative samples. The individual RNA-positive specimens from these pools had been quantitated and had viral loads ranging from 1,520 to greater than 500,000 copies/ml (Versant HIV-1 RNA 3.0 assay; Siemens Healthcare Diagnostics) (). To create the new 128-member pools for this study, the 16-member plasma pools were thawed and 200 μl of each of the 21 HIV-1 RNA-positive 16-member pools was manually combined with 200 μl from seven HIV-1 RNA-negative 16-member pools. Using this procedure, 21 new 128-member plasma pools were created.
Nucleic acid amplification test results for different pool sizes
The 21 new 128-member pools were tested using the Aptima HIV-1 RNA qualitative assay according to the manufacturer's instructions. Each pool was initially tested once; HIV-1 RNA was detected in 20 of the 21 pools (). The one pool in which HIV-1 RNA was not detected included an HIV-1 RNA-positive specimen with a plasma HIV-1 viral load of 1,827 copies/ml. In accounting for the dilution process, this reaction mixture contained 7.1 copies of HIV-1 RNA. Although nonreactive pools would not typically be retested during routine screening, we repeated testing of this pool in two separate assays to obtain additional data, and a reactive result occurred in each repeat test. One other specimen had a comparable viral load (1,520 copies/ml; 5.9 RNA copies/reaction). Three separate assays were also performed for the 128-member pool of this specimen, and the result was reactive for HIV-1 RNA in all three assays (). The mean signal-to-cutoff ratio (S/CO) for all replicates of the two 128-member pools containing specimens with less than 2,000 HIV-1 RNA copies/ml was 12.19 (range, 11.51 to 12.78). The viral load was at least 10-fold higher for the other 19 reactive 128-member pools, and the mean S/CO was 24.54 (range, 13.51 to 34.15).
Our results suggest that most AHI cases would be detected by screening 128-specimen pools using the Aptima HIV-1 qualitative RNA assay. However, our inability to detect RNA in one of three replicates of a pool created from a specimen with 1,827 copies/ml of HIV RNA suggests that a false-negative result may occur if the specimen is collected during the eclipse period when HIV is undetectable or early in the acute period of infection when the viral load is very low. RNA was detected in five of the six reactions that were performed on specimens with fewer than 2,000 RNA copies/ml. More research would be needed to determine if the single false-negative result occurred because of a technical error or assay limitations. Therefore, an intermediate pool size between 16 and 128 members may offer sufficient sensitivity for AHI detection. The feasibility of conducting AHI screening in the public health setting has been established (5
), and conducting pooled NAAT following routine antibody screening has been shown to increase HIV case detection (4
). However, the cost-effectiveness of conducting pooled NAAT screening depends on several factors, including the incidence of HIV in the population, the frequency of antibody testing, and the sensitivity of the screening assay (3
). Using estimates of a cost-effectiveness analysis of pooled NAAT in three public health settings, the 128-member pooling scheme would reduce reagent costs by over 90% for each of the settings in the CDC study. Targeted screening of persons in high-risk settings using pooled NAAT may be an appropriate use of the technology, and our results demonstrate that pooling strategies can be optimized by evaluating different pooling levels to reduce the cost of NAAT without compromising the capacity to detect AHI. Laboratories that are performing large-volume testing and considering AHI screening with NAAT may consider 128-member pools as an option for cost savings.