Rapid Communications

Use of an enzyme immunoassay for detection of antibody to human immunodeficiency virus in low risk populations

DR. FRANS PEETOOM AND THE COOPERATIVE STUDY GROUP*

Pacific NW Region. Portland, Oregon Received for publication June 21, 1986; revision received and accepted Septder 24, 1986.

ABSTRACT: Systems LAV EIA) in low risk populations from five geographically distinct sites. grown in the T-cell line CEM. initially reactive and 0 . 2 percent were repeatably reactive. repeatably reactive specimens were confirmed by immuneprecipitation (RIP). percent specificity compared to RIP and had a positive predictive value of 67 percent. this test generates a low proportion of false positive EIA results and is highly suitable to screen blood donors.

Antibody to Human Immunodeficiency Virus (HIV) was detected by an enzyme immunoassay (Genetic The assay used virus (LAV)

In blood and plasma donor populations, 0.4 percent of the samples were Fourteen of the twenty-one (67 percent)

The assay demonstrated 99.9 The data suggest that

The etiologic agent of acquired immuno- deficiency syndrome (AIDS) is a retrovirus recently renamed as Human immunodeficiency virus (HIV), and previously called Lymphadenopathy-associated virus (LAV). Human T-cell Lymphotropic Virus Type I11 (HTLV-III), and AIDS-associated retrovirus (ARV).(1-4) been shown to be indicative of infection since virus can be isolated from 60 to 80 percent of seropositive individuals.(5) In order to prevent transmission of infection by blood and blood products. enzyme immunoassays have been developed for the detection of antibody in blood donors.(6-7)

trials of a new EIA for detection of antibody to HIV. This test uses the LAV virus grown in a modified T-cell line, CEM, in contrast to a majority of tests which use the HTLV-111 virus grown in the H9 cell line. Results obtained in tests of almost 10,000 blood donors indicated that the LAV EIA was 99.8 percent specific by the criterion of Food and Drug Administration (FDA) and 99.9 percent specific when compared to the reference method. immuneprecipitation (RIP), which detects antibody to specific viral proteins.

The antibody response to the virus has

We report here the results obtained in clinical

Materials and Methods

Study population

specimens obtained from subjects from five geographically distributed testing sites, including one plasma and four blood screening centers. Specimens from three testing sites for voluntary blood donors were obtained prospectively under informed consent while specimens from one testing site were anonymous. plasma donors were obtained without informed consent as salvage specimens without any identifying information.

Enzyme inmunoassay

(Genetic Systems Corporation, Seattle, WA) in the five study centers according to the manufacturer's instructions. The test was performed using microtest plates with wells coated with purified disrupted LAV produced in CEM, a modified T-cell line. In summary, a 100 p1 aliquot of a 1:401 dilution of the test serum or plasma was added to each well. Following incubation at 37OC for 1

The study was conducted on serum and plasma

Plasma samples from 3110 paid

All specimens were tested by the LAV EIA

hour, the wells were washed, 100 p 1 of horseradish peroxidase-labeled goat anti-human immunoglobulin was added to each well, and the plates were reincubated at 37°C for 1 hour. After repeated washing, 100 ~1 of a tetramethylbenzidine chromogen and H202 solution was added to each well, and after 30 minutes incubation in the dark at room temperature, 100 111 of 3N H2S04 was added to stop the chromogenic reaction. Absorbance was then measured at 450 nm with a 630 nm reference filter in an automatic microplate reader.

determined by adding 0.225 optical density (OD) to the mean of three negative controls assayed on the same plate. considered negative. Any specimens with OD readings at or above the cutoff were rediluted and retested in duplicate. Any specimens for which one or both repeat tests were reactive were considered positive for anti-HIV.

Reference method A subset of the specimens was tested by the

reference method, immuneprecipitation (RIP). RIP assays were performed on all repeatably reactive specimens and on a computer-selected random sample of at least 100 EIA negative specimens from each test site (total 514 specimens). RIP was performed on [35S]methionine labeled viral extracts according to a previously described method.(8) Specimens with antibody to viral antigens, p25, gp41, alone or in combination with evidence of antibody to p18, p34. p55, p68, gpll0. or gp150, were considered positive. considered negative.

For each microtest plate, a cutoff value was

Values below the cutoff were

All other patterns were

Results

Specimens from 6593 voluntary blood donors and 3110 paid plasma donors were tested. The results are presented in Table 1. As shown, 99.6 percent of the random donor population did not react initially, 0.4 percent were initially reactive and 0 . 2 percent were repeatably reactive. (62 percent) of 34 initially reactive specimens were repeatably reactive upon retesting.

ficity of the LAV EIA. a reference method, immuneprecipitation (RIP) was performed on all

Twenty-one

In order to assess the sensitivity and speci-

TRANSFUSION 1986-Vol. 26. No. 6 DETECTION OF H I V 537

Table 1. DETECTION OF ANTIBODY TO HIV IN RANDOM DONORS

~ _ _ _ ~

Reactivity Number Initially Repeatably

Site Tested Non-Reactive Reactive Reactive (%) ( X ) (%) (%)

1 1851 1848 3 3 (100) (99.8) (0.2) ( 0 . 2 )

2 793 786 7 3 (100) (99.1) (0.9) (0.4)

3 2000 1994 6 5

4 1949 1946 3 0

5 3110 3095 15 10

(100) (99.7) (0.3) (0 .2)

(100) (99.8) (0.2) (0.0)

(100) (99.5) (0.5) (0.3)

TOTALS 9703 9669 34 21 (100) (99.6) (0.4) (0 .2)

repeatably reactive specimens and on 514 specimens that did not react chosen randomly from each test site. The results of RIP studies of the 21 repeatably reactive specimens and the 514 EIA-negative donor samples selected at random are shown in Table 2. samples, 14 were found to have anti-LAV by RIP, indicating a positive predictive value of a repeatably reactive LAV EIA in blood donors of 67 percent. blood donors in this study (14 in 9703) was too low to permit a reliable estimate of sensitivity based on negative RIP findings in only 514 donors. However, the failure to find a false-negative specimen among these donors correlated with the high sensitivity found in the clinical trials that supported licensure (9).

Of the 21 repeatably reactive

The 0.15 prevalence of anti-HIV among

Table 2 . COMPARISON OF LAV EIA lU REFERENCE ~ T K J D (n = 535)

Imuneprecipitation (RIP) + -

GSC LAV EIA + 14 7

- 0 514

Discussion

We report here the results of a clinical trial to evaluate an EIA designed to detect antibody to HIV in samples from blood and plasma donors. The criterion specified by the FDA was one method used to estimate specificity. That is. the percent of blood and plasma donors found to be repeatably reactive subtracted from 100 percent defined specificity. In this study, 0.2 percent of donors Were repeatably reactive which indicated a

specificity of 99.8 percent.

here, sensitivity of the LAV EIA previously was determined in the clinical trials supporting licensure (9). The FDA criterion defined sensitivity as the percentage of patients with AIDS, as determined by the Centers for Disease Control criteria, found to have a repeatably reactive LAV EIA. In 181 AIDS patients studied, each had a repeatably reactive LAV EIA. indicating a sensitivity of 100 percent. An addition. all 181 AIDS patients in that study had antibody to HIV on RIP analysis.

The EIA was tested on 9703 plasma and serum samples from blood donation centers representing geographically distinct sites. These centers were considered to be both high and low risk locations. In this population, 34 specimens were initially reactive and 2 1 were repeatably reactive. Thus. 62 percent of the initially reactive specimens vere repeatable. Based on the assumption that 100 percent of blood donors would have negative EU, the test showed 99.8 percent specificity in the blood donor population. However, of the 0.2 percent repeatably reactive samples, 14 of 21 or 67 percent were positive by the reference method. None of the 514 randomly selected samples that were EIA negative were positive by RIP. of the samples that were negative in the RIP and positive in the EIA showed weak reactions to single viral proteins p18 and p25 by Western blot. The significance of these reactions is unknown. Using the RIP reference method the EIA demonstrated 99.9 percent specificity and the predictive value of a positive result was 67 percent.

virus grown and purified from a modified T-cell leukemia line, CEM. The viral antigen is different from that produced from the HTLV-I11 infected H9 cell line which has been shown to express HLA antigens which copurify with the viral extract (10). Although the causes were not studied. we speculate that the 99.9 percent specificity of the LAV assay as compared to the reference method was in part to reduced contamination of the viral extract by cellular antigens such as HIA since such antigens may react with HLA antibodies in donor sera.

detection of antibody to HIV. The assay cut-off was established in order to maximize test sensitivity. The sensitivity of the assay in clinical evaluation of AIDS patients was 100 percent and the specificity in the blood donor population was 99.8 percent by FDA criteria and 99.9 percent compared to a reference method. use of this test might result in a reduction of rejection of blood and blood donors.

The Cooperative Study Group: 1) Ms. Betty Curry, Cutter Biological Special Testing Laboratory, 9499 Kearney Villa Road X108, San Diego, CA 92126. 2) Roger Dodd, Ph.D.. American Red Cross, 9312 Old Georgetown Road, Bethesda. MD 20814. 3) Chang T. Fang. M.S. . American Red Cross. 9312 Old Georgetown Road, Bethesda, MD 20814. 4) Lynn Goldstein. M.D.. Genetic Systems Corporation, 3005 First Avenue. Seattle, WA 98121. 5) Frans Peetoom. M.D. , Pacific NW Region. 4200 SW Corbett, PO Box 70. Portland, OR 97207. 6) Herbert Perkina, M.D.. Irwin Memorial Blood Center, 270 Masonic Avenue, San Francisco. CA 94118. 7) George Schneider.

Although not addressed in the studies reported

Six of seven,

The EIA studied used the LAV isolate of the

In summary, the LAV EIA studied is an assay for

The

PEETOOM ET AL. TRANSFUSION Vol. 26. No. 6- 1986

M.D., Spokane and Inland Empire Blood Bank, South 507 Washington Spokane, WA 99204. 8) Mike Wandell. Ph.D., Genetic Systems Corporation, 3005 First Avenue, Seattle, WA 98121.

1.

2.

3.

4.

5.

References

Barre-Sinoussi F, Chermann JC, Rey F, et al. 6. Isolation of T-lysphotropic retrovirus from a patient at risk for acquired immune deficiency syndrome (AIDS). Science 1983; 220:868-871. Gallo RC, Salahuddin S Z , Popovic M, et al.

retroviruses (HTLV-111) from patients with AIDS and at risk for AIDS. Science 1984; 224:500-503. Levy JA, Hoffman AD, Kramer SM, Landis JA, Shimabukuro JM. Oshiro LS. Isolation of 8. lymphocytophathis retroviruses from San Francisco patients with AIDS. Science 1984; 225:840-842. 9. Coffin J, Haase A, Levy JA, et al. Human immunodeficiency viruses. Science 1986; 232:697. 10. Salahuddin SZ, Markham PD, Popovic M. et al.

Frequent detection and isolation of cytopathic 7.

Isolation of infectious human T-cell leukemia/lymphotropic virus type 111 (HTLV-111) from patients with acquired immunodeficiency syndrome (AIDS) or AIDS-related complex (ARC) and fr.om healthy carriers: A study of risk groups and tissue sources. PNAS 1985;

Brun-Vezinet F, Barre-Sinoussi F, Saimot AG, et al. Detection of IgC antibodies to lymphadenopathy-associated virus in patients with AIDS or lymphadenopathy syndrome. Lancet 1984; 1:1253-1256. Sarngadharan MG, Popovic M, Bruch L, Schupbach J, Gallo RC. Antibodies reactive with human T-lymphotropic retroviruses (HTLV-111) in the serum of patients with AIDS. Science 1984; 224:506-508. Rohrschneider LR. Immunofluorescence on avian sarcoma virus-transformed cells: Localization of the src gene product. Cell 1979; 16:ll-24. Summary of Basis of Approval. Office of Biologics, Food and Drug Administration, Bethesda, MD. February, 1985. Sayers MH. Beatty PG, Hansen JA. HLA antibodies as a cause of false-positive reactions in screening enzyme immunoassays for antibodies to humaa T-lymphotropic virus Type 111. Transfusion 1986; 26:113-5.

82:5530-5534.