Detecting Seronegative-Early HIV Infections Among Adult Versus Student Kenyan Blood Donors,by Using Stimmunology

A follow-up study of those detected during the “window period”, i.e., Stimmunology-positive/regular ELISA-negative cases, was performed in a separate population of pregnant women at high risk for HIV. Seven of the 20 women tested at the antenatal clinic were seropositive (both before and after Stimmunology). Among the 13 seronegative women, five were antibody positive after Stimmunology (Table 5). Serology testing in follow-up visits showed that 4/5 of the post-ST positive (pre-ST negative) seroconverted (as defined by standard plasma-based ELISA) during the following 4–6 months. The fifth one was lost to follow-up after 2 months, at which time it was still seronegative. All eight post-ST negative women remained seronegative through out the follow-up (9–11 months).

This study stresses the importance of detecting those infected individuals who are still in the window period and the value of the Stimmunology technique in detecting seronegative HIV-infected individuals, typically missed by standard serological testing methods (16–18, 24) due to potential immune-suppression within the window period.

The results of this study are further supported by those determined on samples from another African blood transfusion service, where an additional 2.2% of tested samples were ST-detected from conventional seronegative HIV carriers (unpublished results). A later study, carried out on samples from the same blood bank, led to similar conclusions, where 1.8% of the seronegative samples were determined to be positive, following ST pretreatment, and later confirmed to be infected with HIV using standard serology (unpublished results). Additionally, routine testing of Ethiopian immigrants arriving in Israel in 1992 and 1998 demonstrated an ST-enhanced identification of 0.3% and 2.7%, respectively, as seronegative HIV carriers (manuscript in preparation). Thus, the ST culture step offers a means of increasing blood safety, vital in countries where HIV prevalence and incidence are high.

As set by World Health Organization guidelines, any blood sample yielding positive recordings on an HIV-Ab ELISA must be regarded as contaminated and should be discarded. These internationally accepted regulations do not require further confirmatory testing of any sort. Nevertheless, the surprisingly high rate of missed infections and the wish to confirm the HIV-positive diagnosis of the blood donors prompted us to perform additional confirmatory testing. Aside from the repeat testing performed for each post-ST–positive sample in Kenya, these samples were also tested on an additional ELISA kit by the technical staff of a different laboratory in Israel. At that time, kits detecting only anti-HIV IgG antibodies were also used, showing that a few samples had only HIV-specific IgM, while most had both IgM and IgG. WB analysis, run both in Kenya and in Israel, offered additional information regarding the antibody repertoire of ST-treated samples. While the WB analyses of the pre-ST samples of all seronegative carriers were negative, 28.5–50% of post-ST ELISA-positive samples gave rise to bands on WB consistent with the presence of core and envelope protein-specific antibodies of the IgG isotype, as that is the only Ig isotype that the WB detects. As demonstrated in these results, WB analysis can hardly be viewed as a confirmatory assay, as such wide variations exist between the kits used. These variations often stem from differences in the choice of epitopes used to represent a specific protein of interest. Such variances can be of great significance when testing the narrow, epitope-limited spectrum of antibodies expressed by the early-stage HIV carrier. The IgG-specific WB kits will remain insensitive to the many IgM antibodies expressed throughout the early window period, especially after ST pretreatment, which is assumed to propagate antibody expression by cells silenced during the initial stages of infection. Furthermore, while the IgG-specific WB kits may provide additional information regarding the immunological status of the HIV carriers at early stages of infection, the IgM+IgG sensitive ELISA kits available today display much more sensitivity in early detection of HIV. Due to the lack of information concerning the precise point of viral contraction, any attempt at correlating appearance of HIV-specific IgG antibodies with more advanced stages within the window period would be completely speculative. Furthermore, the presence of silenced, primed B cells that have already undergone class switching (16) may reflect the course of initial infection and its entry to latency without attesting to the point along the window period at which the carrier stands. In addition, ELISA basically measures the presence of both conformational and non-conformational antibodies against the antigen in question. On the other hand, the Western blot primarily detects antibodies directed against denatured antigen and thus does not normally involve detection of conformational antibodies. Thus, differences do exist between Western Blot and ELISA, especially in antibody responses during early primary immune responses which appear to be directed more at conformational epitopes than antibodies against linear epitopes of processed antigens. It has previously been shown that early primary immune responses are directed at env and gag and later on, multiple bands against the other viral proteins are detected, which is likely due to levels of the appropriate antigen. Thus, there is a gradual maturation of the humoral response.

Viral particle presence was assessed via RT-PCR analysis using highly sensitive assays (1–5 copies/ml), and HIV was detectable in 53.3% of the seronegative but HIV-1–infected carriers. It has been reported that RT-PCR or nucleotide amplification tests, with a sensitivity threshold of 50–200 copies/ml, can detect HIV in the blood 10–12 days prior to seroconversion, and sometimes even earlier (25). However, following exposure, the virus primarily appears to target gastro-intestinal lymphoid tissues (GALT) where it replicates and causes pathology. During this initial acute infection period, the frequency of HIV-1–infected cells in the peripheral blood has been estimated to be less than 1 in a million CD4+ T cells, according to Dr. Robert Siliciano’s data and others who have studied the SIV-infected rhesus macaques. Thus, performing RT-PCR assays for HIV-1 can provide reliable information regarding the presence of infection only when the levels reach 100–500 viral copies.

In the studies reported herein, there were 7 (+ 1 weak) RT-PCR positive among the seronegative HIV-1–infected donors [using very sensitive, not commercial, RT-PCR, and no pooling] and those 7 plus 10 additional donors were detected as HIV antibody positive using the Stimmunology pre-treatment step for the blood samples prior to testing on approved diagnostic kits. The difference stems from the fact that Stimmunology has been proven to enable detection of HIV carriers at the initial lymphocyte priming stages (16, 20), which is believed to occur within days of immunogen exposure. It must be noted that while ST enhances antibody generation in vitro, it bears no effect on the sample’s viral count, allowing the diagnostician to analyze early-stage antibodies despite viral latency. As the results clearly demonstrate, some post-ST–positive samples proved to be RT-PCR negative. This phenomenon further widens the gap between viral content–dependent methods of HIV detection compared to antibody-dependent methods.

In addition to the fact that Stimmunology enables earlier detection, independent of viral latency or its levels, there are also issues of cost effectiveness. It is readily acknowledged by the WHO and other international organizations that RT-PCR–based assays are too expensive to set up and maintain in resource-poor countries and especially in countries where the HIV incidence and prevalence are high. There is a continued effort to enable detection of infected yet seronegative donors within the realm of the antibody detection assays (e.g., 4th generation ELISA-based assays that combine antigen + antibody detection). We are extremely sensitive and aware of this issue and have thus chosen to focus on attempts to identify a method that in fact can be relatively easy to perform and is cost effective.

When comparing the results of two different assays used to detect the infection in a given population, the issue of “sensitivity” should be considered. Since Stimmunology is not an assay, one would be comparing the positive antibody testing using post-ST samples to those of PCR or NAT results (DNA or RNA) in the pre-ST sample. Since the comparison is between two different assays that detect two different things, analytical sensitivity and/or diagnostic sensitivity are not comparable. However, statistical/ diagnostic sensitivity (population wise) can be addressed by looking at two parameters: 1. How many of those detected by antibodies in post-ST samples were also detected by PCR (pre-ST). 2. On the time line, which signal appears first in a newly infected person.

Table 2 addresses the first parameter. Using very sensitive PCR (1–5 copies/ml), only 7/15 of the window period samples (i.e., antibody-positive post-ST only), had detectable virus. Thus, the Stimmunology process enables the antibody testing to become more sensitive than super sensitive RT-PCR. We humbly submit that less analytically sensitive PCR/NAT (or other systems for the detection of viral genome) would be even less statistically sensitive. As to the second parameter, Table 5 addresses this issue in a follow-up study in a Kenyan population, where antibodies in post-ST samples were detectable several months prior to seroconversion. The PCR/NAT tests have been shown to detect virus in the blood 10–12 days prior to seroconversion. Following an infectious needle stick injury, the health professional remained seronegative for the 5 months’ duration of the follow-up, while virus was detected intermittently after 3 months (with borderline positivity at 2 months), with the signal lost at 5 months. However, antibodies were detectable form the second week onwards when using Stimmunology as a pre-treatment of the blood prior to the routine antibody testing (data not shown). Thus, to conclude, our findings, as described in this manuscript, demonstrate a higher sensitivity of Stimmunology + antibody testing than ultrasensitive PCR.

Calculation of the incidence, solely based on the number of ST-detected carriers, is impossible, as the serological window period length of the population remains unknown. Several models have been proposed for estimating the length of the window period, but the majority is based on large-scale studies in low-risk blood donor populations. Alternatively, some models are based on calculations of seroconversion in individuals transfused with infected blood, where the time of exposure and seroconversion are known (26). However, these models do not reflect African or other populations at high risk for HIV infections, neither in their mode and dose of infection nor in their genetic and environmental background. The increase in seroprevalence reported by Kenya’s National AIDS and STDs Control Programme (27) and other international agencies is lower than that which is implied by the percentage of window period individuals detected in this study. However, the incidence cannot be calculated based on the rise in seroprevalence of a given population, as the number of deaths over the period of interest must also be considered. Furthermore, a longer window period than that which is typically calculated for non-African populations can explain the significantly higher “incidence” reported herein. Such a model, describing a longer window period, is supported by results seen in an additional African population (28). This model also infers that the HIV replication rate seen in the active viremia (29) stage prior to seroconversion is not constant, where long periods of low-level viremia and even longer periods of latent infection (13–15) can exist. Those in the window period are those recently infected, but to actually calculate the incidence one would have to have a good estimate of the length of the window period and the length of the asymptomatic period in this population. Another key factor would be whether the donors represent the general population (e.g., general risk–associated life style and recent change in life style) or rather a skewed population representing a higher risk of recent exposures. Using Stimmunology-based results and comparing pre-ST and post-ST antibody results, both qualitative and quantitative, has enabled the calculation of the window period length in the Ethiopian cohort, indicating it to be much longer (mean of 5 months) than in the Western population (29). Since this population had a recent marked change in the level of risk associated with their “life style” and exposure to sexual workers—this population would resemble more the high school donors. The difference could be accounted to loss of virginity and frequency of sexual contacts in the younger population.

Studies are currently being done to assess the incidence rates in given populations based not only on the rate of window period samples, but also on the “Stimmunology index,” i.e., the increase in antibody levels in the sample following the Stimmunology step. This ratio of OD reading post-ST/pre-ST is higher in recent seroconvertors versus those that have been seropositive for a longer time and antibody production in vivo is fully expressed.

A follow-up study of such Stimmunology-positive/ regular plasma ELISA-negative women has shown that 4/5 of them became positive, using regular plasma on the ELISA, within 4–6 months. These data provide further support both for the validity of the Stimmunology-based assay for early detection of HIV infection and an additional indication regarding the length of the seronegative window period in the study population. Studies in other high-incidence, high-risk populations in Sub-Sahara countries also indicate the existence of a window period longer than the 3 months (for 95% seroconversions) calculated based on other populations.

The Stimmunology blood pre-treatment step is a novel and new approach to earlier and more complete detection of HIV-infected individuals. Much work has been done to try and understand the changes in the cell populations’ profile when suppression is overcome that leads to production of suppressed antibody. Due to the limitations of clinical trials in resource-poor settings, especially with unlinked samples, this was further studied using the monkey model of AIDS—the SIV infection. The frequency of various major subsets pre- and post-Stimmunology using FACS analysis was compared. While the frequency of total CD4+ cells did not change (34.5–39.4, the frequency of CD4+ T cells that expressed CD25+ rose from 3.5 to 19.4. The CD8+ cells were reduced from 37.5 to 13.4, with most of them being CD25+ . CD20+ cells increased three-fold (from 11.9 to 37.4), with CD14+ reduced from 6.3 to 3.2. Most B cells were large blasts.

The avidity of IgG antibody is well known to increase progressively with time after exposure to an immunogen (1). The avidity of the antibodies produced in ST culture of cells from SIV-infected monkeys was measured using an indirect ELISA and running each sample in two wells, one of which was treated with 8M urea. The results were expressed as an Avidity Index (AI), which was determined as the ratio of the OD value obtained following urea treatment divided by the OD obtained without urea treatment, expressed as a percentage. The post-ST samples from monkeys (n = 12) 7 days after infection were clearly positive, but the avidity was much lower (44.9 ± 9.4%) than that of chronically infected monkeys (n = 21, AI = 99.8 ± 6.5%).

In summary, results of this study reveal a higher-than-expected percentage of window period HIV carriers in young Kenyans living on the outskirts of Nairobi. These results suggest that even among young donors, more caution must be taken when screening potential blood donors. Addition of the Stimmunology technique to standard antibody testing could bridge the detection gap which is due to the window period of seronegative HIV carriers. The addition of the Stimmunology-based blood pre-treatment step and the analysis of the results does not require sophisticated automated machines nor ultragrade pure reagents to perform. However, since the five-day culture process is too long in a blood bank routine setting, it has been reduced to 3 days following a large-scale comparative study showing high (albeit not optimal) detection of the window period donors after 3 days.

As Stimmunology enables a more comprehensive detection of HIV carriers, it can serve as a means of significantly limiting the risk of HIV transmission via blood banks in areas where HIV is epidemic. Furthermore, early detection can allow for immediate enrollment of HIV carriers into prevention and education programs, hopefully stemming the incidence and prevalence of HIV infection.