Evaluation of SMARTube to Detect HIV Infection Before Seroconversion Using Standard Methods

Abstract

The acute phase of HIV infection carries substantial risk of transmission; identification of acute-phase infections may offer opportunities to reduce that risk. SMARTube incubation of blood specimens is designed to stimulate in vivo-primed HIV-specific lymphocytes to produce HIV antibodies in vitro. The resulting supernatant (S-plasma) can be tested to identify acute infections with commercially available HIV assays. We assessed the performance of the SMARTube to identify acute HIV infections in studies at three developing country sites. We conducted HIV incidence studies in Ho Chi Minh City, Vietnam, and Bloemfontein and Rustenburg, South Africa. We estimated HIV incidence in cross-sectional samples and measured prospective incidence in uninfected women followed for up to 12 months. We incorporated SMARTube into the HIV testing algorithm at cross-sectional screening and monthly follow-up visits. We tested 1,384 persons in Vietnam, 1,145 women in Bloemfontein, and 538 persons in Rustenburg. Cross-sectional samples from 11 participants that tested positive with SMARTube after an initial unincubated negative test result (11 of 2,472; 0.4% of all specimens) were considered “potential acute” infections. Matching samples from 3 of the 11 (27.3%) were confirmed by polymerase chain reaction (PCR) testing. In follow-up of 355, 401, and 223 uninfected women in Vietnam, Bloemfontein, and Rustenburg, respectively, 11 seroconversions occurred in Bloemfontein and Rustenburg. In four of these incident infections (36.4%), SMARTube testing had resulted in earlier detection of HIV infection than the eventual seroconversion visits. In our field studies, pretreatment with SMARTube allowed the identification of acute HIV-1 infection in some new infections, but with a positive predictive value of 27%. Larger studies are needed to evaluate SMARTube as an alternative to technically challenging and costly enzyme immunoassay and PCR testing to detect acute HIV infection.

Introduction

UNAIDS estimates that 35 million people worldwide are living with HIV, two thirds of them in sub-Saharan Africa.¹ The period after infection but before seroconversion, known as acute infection, carries an unusually high risk of transmission, making this a key focus for prevention.²

The period from initial HIV infection until seroconversion on current HIV antibody assays is known as acute HIV infection and characterized by an initial eclipse phase, followed by rising viral RNA levels and then p24 antigenemia.³ High HIV levels during acute infection, before immune control, contribute to increased transmission of the virus during the period.^4

–7 Detection methods for acute HIV infection include polymerase chain reaction (PCR) and the p24 assay, which directly measures viral antigen levels. Both tests demand skilled personnel using costly equipment and are expensive to run on a per-test basis. A cost-effective and less technically challenging method would enable the rapid detection of acute HIV infections in resource-limited settings. Identification of individuals in the acute phase of infection could afford an opportunity for preventive and therapeutic interventions when the individual may be highly infectious.² Recent evidence indicates benefits associated with antiretroviral treatment during the acute infection period.⁸

As part of the U.S. Agency for International Development (USAID)–sponsored Site Identification and Development Initiative (SIDI), we evaluated the performance of a new technology, the SMARTube™ (Rehovot, Israel), to detect acute HIV infections in specimens collected in cross-sectional and longitudinal testing.

The SMARTube is used to pretreat human blood samples with a proprietary cocktail of stimulatory agents, resulting in activation, differentiation, and proliferation of lymphocytes and subsequent amplification of specific HIV antibodies in the cultured sample.⁹ Incubation of whole blood in SMARTube allows for in vivo-primed HIV-specific lymphocytes to be stimulated in vitro to produce HIV antibodies. The resulting supernatant can then be tested with commercially available HIV diagnostic tests. The technology has been evaluated in field tests^10

–13 and is registered in the European Union (CE Mark), Israel, Russia, South Africa, Nigeria, Romania, Hungary, and Turkey.

We report in this study on the performance of the SMARTube in HIV incidence studies that were conducted at three developing country sites.

Materials and Methods

Study design

The SIDI project conducted HIV incidence studies using similar protocols in several countries, including Vietnam and South Africa. The primary objectives of these studies were to estimate HIV incidence in cross-sectional samples using the BED-CEIA assay¹⁴ and compare those estimates with a prospective incidence measurement in a cohort followed for up to 12 months. The main incidence results of each study are reported elsewhere.^15–16

A secondary objective of the studies was to evaluate the performance of SMARTube to detect acute HIV infection. The studies incorporated the SMARTube into the HIV testing algorithm for both the cross-sectional screening visit and each monthly follow-up visit in the prospective phase.

Study sites and cohorts

The incidence study in Ho Chi Minh City, Vietnam, was conducted at three sites located in three different districts: the Binh Thanh, district 4, and district 8 Community Counseling and Support Centers (CCSCs). Each of these CCSCs houses a Voluntary Counseling and Testing (VCT) and an HIV Outpatient Clinic (OPC) unit. These sites are part of the VCT/OPC clinic network overseen by the HCMC Provincial AIDS Committee. Cross-sectional screening comprised women (predominantly commercial sex workers) and their male partners recruited by local peer educators or other study participants using modified respondent-driven sampling methods. The prospective cohort included only HIV-negative, opiate-negative women identified in cross-sectional screening.

In South Africa, studies were conducted at two sites: JOSHA Research Center in Bloemfontein and The Aurum Institute's Rustenburg Research Centre in downtown Rustenburg. In Bloemfontein, participants in cross-sectional screening were women presumed at higher risk of HIV infection based on behavioral risk factors. These women were recruited using methods developed based on feedback from community engagement activities. The prospective cohort recruited HIV-negative women directly from cross-sectional screening. In Rustenburg, cross-sectional screening recruited both men and women through a variety of community-centered activities, including modified respondent-driven sampling.¹⁷ Similar to Bloemfontein, the prospective cohort comprised HIV-negative women recruited directly from cross-sectional screening.

Ethical considerations

All study protocols, informed consent forms, participant education and recruitment materials, and case report forms were reviewed and approved by the FHI 360 ethics committee and the local ethics committees for each site.

In all studies, informed consent was obtained from each participant before enrollment and data collection. Written consent was also obtained for long-term specimen storage and possible future testing, although this was not a requirement for study participation. All consent forms were developed in English and translated into local languages to enhance comprehension.

Laboratory methods

HIV diagnostic tests

The HIV test algorithm used in the studies varied according to the national standards. In Vietnam, the HIV testing algorithm included two enzyme immunoassays (EIAs), followed by one rapid test for confirmation; a positive result in all three assays was required to classify an individual as HIV positive. Plasma samples from each study phase were tested using (1) MUREX Ag/Ab Combination EIA, (2) Genscreen HIV-1/2 Enzyme Immunoassay version 2.0, and (3) Abbott Determine™ HIV-1/2 Rapid Test.

In the South African studies, we performed two rapid tests in parallel (Abbott Determine™ and Uni-Gold™ from Trinity Biotech) using whole blood from finger-prick sampling. For discordant results, the Bioline test (Standard Diagnostics, Inc.) was used as a tiebreaker. Two positive results were required to classify the participant as HIV positive.

All tests were run according to the manufacturer's kit instructions. Participants were informed of their HIV status upon confirmation of serostatus per national standards.

SMARTube

A 2 ml whole blood sample (sodium heparin tube) was collected from all cross-sectional participants and at each follow-up visit made by prospective participants. Within several hours (per the package insert), 1 ml whole blood was added to the SMARTube, and the sample was incubated at 37°C with 5% CO₂ for 3 days. The samples were then centrifuged, and the supernatant (termed “S-plasma”) was transferred to a cryotube and stored at −70°C before use. S-plasma was tested using the Determine HIV-1/2 Rapid Test at all sites and an EIA that varied by site (Genscreen HIV-1/2 version 2.0 in Vietnam, the AxSYM HIV-1/2 gO in Bloemfontein, and Vironostika HIV Uni-Form II plus O in Rustenburg). As incubation in the SMARTube dilutes the sample, the dilution was compensated for in EIAs by increasing the sample volume in the assay. Volume compensation was not possible for rapid testing of the S-plasma.

We compared S-plasma results with corresponding nonincubated samples from the same visits using the same HIV diagnostic tests. Samples that scored positive on the diagnostic test after incubation in SMARTube, but negative before incubation, were termed “potential acute HIV infection.” Confirmation of acute infection status was determined by testing a time-matched nonincubated EDTA plasma sample for the presence of HIV RNA using AMPLICOR HIV-1 MONITOR Test, version 1.5 (Roche) in Vietnam and the Abbott RealTime HIV-1 assay for both South African sites. Samples with a detectable viral load were considered “confirmed acute HIV infection.” If PCR results were negative, additional testing with EIA and western blot was performed on nonincubated plasma to assess potential false-positive SMARTube or false-negative PCR results.

When possible, potential acutely infected participants were recalled to the study clinic for an interim visit (if not already part of the prospective cohort) to draw additional blood samples for HIV testing. Participants were not informed of their SMARTube results as the testing was conducted for research purposes only.

Statistical methods

We did not assess HIV PCR status for most cross-sectional participants; hence, we could not calculate true sensitivity and specificity results for SMARTube-incubated samples. Rather, we compared results for the initial test with and without SMARTube incubation and calculated concordance rates for initial positive test results and initial negative test results. We also report the number and proportion of potential acute infections confirmed as acute and the positive predictive percentage and its exact 95% confidence interval (CI) of confirmed acute infections during cross-sectional screening. Among those PCR-confirmed acute infections in the prospective cohorts, we report the sensitivity results of SMARTube incubation on two downstream diagnostic HIV tests (the Determine HIV-1/2 Rapid Test and an EIA, depending on study site) to identify acute infections earlier than ordinary testing in longitudinal observation.

Results

HIV diagnostic test and SMARTube results on 1,384 men and women in Vietnam, 1,145 women in Bloemfontein, and 538 persons in Rustenburg (Table 1) were included in the cross-sectional analysis. Seronegative women who entered the prospective study at the three sites totaled 355, 401, and 223 in Vietnam, Bloemfontein, and Rustenburg, respectively.

Table 1.

Concordance Between Cross-Sectional HIV Screening Tests With and Without SMARTube Incubation

Site	Initial test positive (N)	S-plasma + initial test positive (N)	Concordance, % (95% CI)	Initial test negative (N)	S-plasma + initial test negative (N)	Concordance, % (95% CI)
Vietnam^a	233	228	97.9 (95.1–99.3)	1,151	1,147	99.7 (99.1–99.9)
Vietnam^b	233	233	100 (98.4–100)	1,151	1,151	100 (99.7–100)
Bloemfontein, RSA^b	236	236	100 (98.5–100)	909	903	99.3 (98.6–99.8)
Rustenburg, RSA^b	126	126	100 (97.1–100)	412	411	99.8 (98.7–100)

Initial test: Genscreen HIV-1/2 EIA.

Initial test: Determine HIV-1/2 Rapid Test.

CI, confidence interval; EIA, enzyme immunoassay; RSA, Republic of South Africa.

Test result concordance with and without SMARTube incubation

We compared results for the initial screening test with and without SMARTube incubation. In South Africa and Vietnam, all HIV-positive specimens tested with Determine HIV-1/2 also tested positive using Determine after the SMARTube incubation (100% concordant; left side, Table 1). In Vietnam, where we used Genscreen EIA as the diagnostic test, 5 of 233 HIV-positive plasma samples (2.1%) tested negative after a matching sample was incubated in SMARTube (97.9% concordant). Of note, four of the five discordant tests also tested negative using the other diagnostic tests (Determine and Murex) in the testing algorithm there. All four had low OD values near the assay cutoff, and retesting of the same S-plasma sample yielded a negative result. For the one sample that retested positive, the participant was recalled 6 weeks after the initial visit, and both plasma and S-plasma samples tested negative using Genscreen EIA. Concordance for initial negative results using Genscreen or Determine ranged from 99.3% to 99.8% (right side, Table 1).

Detection of acute infection

S-plasma samples that tested positive after an initial unincubated negative test result were considered “potential acute” infections. In cross-sectional screening, 11 participants (11 of 2,472; 0.4% of all specimens) were HIV negative per national algorithm but were positive on HIV diagnostic tests using S-plasma: four in Vietnam using Genscreen EIA (three of four with low optical densities and negative results upon retest), six in Bloemfontein, and one in Rustenburg, both of the latter sites using Determine. In Vietnam, no potential acute infections were identified after SMARTube incubation using the Determine test (all negative plasma samples were also S-plasma negative). Specimens from 3 of these 11 potential acute participants (27.3%) were confirmed by PCR testing using specimens from that cross-sectional visit (Table 2). Thus, the overall percentage of confirmed acute infections among initial test negative specimens was 0.1% (3 of 2,472; 95% CI 0.0–0.4).

Table 2.

Detection of Acute HIV Infections After SMARTube Incubation

	Cross-sectional screening				Prospective phase
Site	Initial test negative ^a (N)	Potential acute	Confirmed acute	Positive predictive, % (95% CI)	Seroconverted	PCR-confirmed acute	S-plasma + initial test positive ^b (N)	Sensitivity, % (95% CI)
Vietnam	1,151	4	0	0.0 (1.1–60.2)	0	0	0	NA
Bloemfontein, RSA	909	6	2	33.3 (4.3–77.7)	9	9	4	44.4 (13.7–78.3)
Rustenburg, RSA	412	1	1	100 (2.5–100)	2	1	0	0 (0.0–97.5)
Total		11	3	27.3 (6.0–61.0)	11	10	4	40.0 (12.2–73.8)

Initial testing by Genscreen HIV-1/2 EIA in Vietnam and Determine HIV-1/2 Rapid Test in South African sites. In Vietnam, S-plasma was also tested using Determine HIV-1/2, and no potential acute infections were identified.

In an earlier study visit than the seroconversion visit.

PCR, polymerase chain reaction.

The prospective cohorts offered the opportunity to assess whether SMARTube incubation could detect infection earlier than with ordinary testing under the national algorithm. Seroconversions occurred in two of the three prospective studies: Rustenburg and Bloemfontein. In Rustenburg, two participants seroconverted (Table 2): one at her 5-month follow-up visit (the seroconversion visit according to the national algorithm) and one at her 3-month visit according to standard testing. While the SMARTube-incubated sample from the seroconversion visits also tested positive, SMARTube incubation of specimens from earlier visits had failed to detect HIV infection (Table 2). One of these two participants had seroconversion confirmed by PCR in a stored specimen from an earlier visit than the seroconversion visit; the other participant did not have a PCR result available.

The Bloemfontein prospective cohort had nine seroconversions during the course of monthly follow-up (Table 2). All nine had acute infection confirmed by PCR using stored specimens from earlier visits than the seroconversion visits (under the national algorithm). In four of the nine confirmed incident cases, SMARTube incubation of specimens from earlier visits had yielded a positive test result before the seroconversion visit (Table 2). In the remaining five confirmed seroconversion events in Bloemfontein, SMARTube incubation did not detect HIV infection earlier than the national algorithm.

Discussion

In our cross-sectional surveys, SMARTube detected 11 persons who were potentially acutely infected (11 of 2,472 seronegative persons, 0.4%). Only 3 of those 11 potential acute infections (27.3%) were confirmed by PCR testing, a fairly low positive predictive value. If we omitted the three participants in Vietnam for whom the initial S-plasma positive results had low optical densities and subsequently negative results upon retest of the same sample, the positive predictive value would rise to 37.5% (three of eight). In longitudinal testing, 4 of 11 seroconversions overall at the two South African sites (36.4%) were detected in stored specimens from an earlier visit by means of SMARTube incubation. In the seroconversions confirmed by PCR testing, sensitivity was 40% (4 of 10).

A prior study was conducted among two cohorts of Ethiopian Jews following their immigration to Israel in 1992 and 1998.¹² In these cohorts, SMARTube-incubated and nonincubated samples were tested in parallel using Recombigen HIV-1/HIV-2 EIA (Cambridge Biotech Ltd.) and Genetic Systems™ HIV-1/HIV-2 Peptide EIA (Genetic Systems Corp.). Samples testing EIA positive following incubation in SMARTube, but seronegative by EIA on nonincubated samples, were further analyzed by western blot and PCR. Of the 285 participants in the 1992 cohort, 7 were seropositive; 8 samples were EIA positive following SMARTube incubation, and original plasma from 3 of those 8 samples was PCR positive. Of the 537 participants in the 1998 cohort, 26 tested seropositive; 2 samples were EIA positive following SMARTube incubation, and original plasma samples for both were PCR positive.

Another study was done among blood donors in Kenya.¹³ A total of 513 adults aged 20–45 years and 332 high school age donors were enrolled, and parallel testing of SMARTube-incubated and nonincubated samples was conducted using EIA-based assays (Genetic Systems Corp. and/or Sanofi Pasteur). Further western blot and PCR analyses were conducted on seronegative samples that tested positive after incubation. Forty-five of 513 adults were seropositive; 17 seronegative samples were EIA positive following SMARTube incubation, and PCR results for original plasma samples were positive for 8 of 15 tested. SMARTube incubation increased the observed HIV prevalence rate in the cohort from 8.8% to 12.1%. In the high school cohort, 12 samples were seropositive, and SMARTube incubation identified additional 10 samples as antibody positive; the observed HIV prevalence increased from 3.6% to 6.6%. A separate follow-up study there enrolled 20 pregnant women at higher risk for HIV: 7 women were seropositive, 8 women were HIV-negative pre- and postincubation in SMARTube, and 5 women were antibody positive following SMARTube incubation. During follow-up, four of five women seroconverted within 4–6 months (the fifth woman was lost to follow-up at 2 months).

Our HIV incidence studies differed from previous SMARTube studies in that we used commercially available PCR kits with detection thresholds of 40–50 viral copies/ml, while the earlier studies used a more sensitive noncommercial kit capable of detecting 1–5 copies/ml. Our study sizes and the infrequency of acute infection also weaken our findings. More important, perhaps, is the fact that we could not calculate sensitivity and specificity rates for SMARTube-incubated samples. We lacked the true HIV infection status because confirmatory PCR testing was done only for the small number of potential acute infections (pooled cross-sectional samples from the Vietnam study were tested by PCR, and no HIV-positive samples were identified; data not shown). In the setting of our studies, all we can say with confidence is that the screening tests yielded high concordance with and without SMARTube incubation.

In our hands, pretreatment of whole blood with SMARTube enabled detection of acute HIV infection before seroconversion in a minority (27.3% positive predictive percentage) of potential acute infection cases in cross-sectional screening and 40% sensitivity in detecting acute infections in prospective follow-up earlier than ordinary screening under the national algorithm. Further studies of larger size with cost-effectiveness evaluation, and in different settings, such as blood donation and antenatal care, are needed to assess the utility of SMARTube for the detection of acute HIV-1 infection as an alternative to technically challenging and costly EIA and PCR testing in resource-constrained settings.

Footnotes

Acknowledgments

The authors thank the developer of the SMARTube technology, Tamar Jehuda-Cohen, for helpful discussions and training on the use of the SMARTube. We also thank our FHI 360 colleague Timothy Mastro for valuable discussions and review of the article. The HIV incidence studies were funded by the U.S. Agency for International Development (USAID) under Cooperative Agreement No. GPO-A-00-05-00022-00, the Contraceptive and Reproductive Health Technologies Research and Utilization (CRTU) Program, and Cooperative Agreement No. GHO-A-00-05-00022-00, the Preventive Technology Agreement. Additional funding was provided by the Centers for Disease Control and Prevention (CDC) under contract no. 200-20074-05314, Task Order #7. The contents are the responsibility of FHI 360 and do not necessarily reflect the views of USAID or the CDC. We are grateful to the study participants. In Vietnam, we thank the Provincial AIDS Committee of Ho Chi Minh City, the People's Committee, and the Center for Preventive Medicine laboratory. In South Africa, we thank the The Aurum Institute and JOSHA, staff at PathCare Laboratories and the Mangaung University Community Partnership Program in Bloemfontein, and the Departments of Health in North West and Free State provinces. We thank Nguyen Thi Hoang Lan and Le Truong Giang in Vietnam; John Lombaard, Sharne Foulkes, Ilse Reblin, and Gustav Venter in Bloemfontein; and Mary Latka and Candice Chetty in Rustenburg. We are also grateful to numerous current and former FHI 360 colleagues in the North Carolina, Vietnam, and Bangkok offices.

Author Disclosure Statement

No competing financial interests exist.

References

Joint United Nations Programme on HIV/AIDS and World Health Organization: The Gap Report. Geneva. Available at www.unaids.org/en/resources/campaigns/2014/2014gapreport/gapreport (2014) accessed 11 April, 2016 .

Pilcher

, Louie

, Facente

, et al.: Performance of rapid point-of-care and laboratory tests for acute and established HIV infection in San Francisco. PLoS One, 2013; 8:e80629.

Fiebig

, Wright

, Rawal

, et al.: Dynamics of HIV viremia and antibody seroconversion in plasma donors: Implications for diagnosis and staging of primary HIV infection. AIDS, 2003; 17:1871–1879.

Wawer

, Gray

, Sewankambo

, et al.: Rates of HIV-1 transmission per coital act, by stage of HIV-1 infection, in Rakai, Uganda. J Infect Dis, 2005; 191:1403–1409.

Pilcher

, Tien

, Eron

, et al. Brief but efficient: Acute HIV infection and the sexual transmission of HIV. J Infect Dis, 2004; 189:1785–1792.

Pilcher

, Easton

, Kalichman

, et al.: Approaching “HIV elimination”: Interventions for acute HIV infection. Curr HIV/AIDS Rep, 2006; 3:150–158.

Cohen

: Amplified transmisison of HIV-1: Missing link in the HIV pandemic. Adv Stud Med, 2006; 6:157–162.

Grinsztejn

, Hosseinipour

, Ribaudo

, et al.: Effects of early versus delayed initiation of antiretroviral treatment on clinical outcomes of HIV-1 infection: Results from the phase 3 HPTN 052 randomised controlled trial. Lancet Infect Dis, 2014; 14:281–290.

Jehuda-Cohen

, Slade

, Powell

, et al.: Polyclonal B-cell activation reveals antibodies against human immunodeficiency virus type 1 (HIV-1) in HIV-1 seronegative individuals. Proc Natl Acad Sci (PNAS), 1990; 87:3972–3976.

10.

Jehuda-Cohen

, Mumo

, Bwayo

, Pezella

: Detection of HIV infection during window period using polyclonal B-cell activation test. AIDS, 1997; 11:124–125.

11.

Bar-Yehuda

, Weisman

, Kalinkovich

, et al.: High prevalence of HIV-specific immunity in seronegative Ethiopian immigrants in Israel. AIDS, 1997; 11:117–118.

12.

Novikov

, Jehuda-Cohen

: HIV type 1 infection among Ethiopian immigrants to Israel: Enhanced in vitro antibody stimulation for estimating the length of the window period. AIDS Res Hum Retrovir, 2009; 25:165–174.

13.

Mumo

, Vansover

, Jehuda-Cohen

: Detecting seronegative-early HIV infections among adult versus student Kenyan blood donors by using Stimmunology. Exp Biol Med, 2009; 234:931–939.

14.

Parekh

, Kennedy

, Dobbs

, et al.: Quantitative detection of increasing HIV type 1 antibodies after seroconversion: A simple assay for detecting recent HIV infection and estimating incidence. AIDS Res Hum Retrovir, 2002; 18:295–307.

15.

Sexton

, Costenbader

, Vinh

, et al.: Good correlation of prospective and cross-sectional measures of HIV-1 incidence in a higher-risk cohort in Ho Chi Minh City, Vietnam. AIDS Res Hum Retrovir, 2012; 28:866–873.

16.

Feldblum

, Latka

, Lombaard

, et al.: HIV incidence and prevalence among cohorts of women with higher-risk behavior in Bloemfontein and Rustenburg, South Africa: A prospective study. BMJ Open, 2012; 2:e000626.

17.

Malekinejad

, Johnston

, Kendall

, Kerr

, Rifkin

, Rutherford

: Using respondent driven sampling methodology for HIV biological surveillance in international settings: A systematic review. AIDS Behav, 2008; 12(Suppl 4):S105–S130.