Methods

Top Methods Results Discussion Author & Article Info References

We reviewed the database maintained by the AIDS Vaccine Clinical Trials Group for all uninfected, low-risk persons who had been in a study for at least 1 year, had received three or more doses of an HIV vaccine and not placebo, and had had sequential serologic testing at a study site. Commercially available enzyme-linked immunosorbent assays (ELISA) and Western blot tests were done to monitor the immunogenicity of the vaccine preparations as part of a larger panel of immunogenicity tests. We analyzed the results of ELISA (HIVAB, Abbott Laboratories, Chicago, Illinois) and Western blot tests (Cambridge Biotech Corporation, Worcester, Massachusetts) to assess the potential of vaccines to induce reactivity to HIV diagnostic tests. Manufacturer directions were followed in conducting the assays. We considered results of Western blot tests only if results of HIV ELISA testing were positive. In each of the studies considered, ELISA and Western blotting were done for each volunteer 10 to 16 times during the study, depending on the protocol. Four sets of published criteria for Western blot interpretation were applied to the pattern of bands that resulted from vaccination.

Most studies were done with 0-, 1-, and 6-month schedules of vaccinations. Details of the study designs have been reported elsewhere [1-6]. The study that used vaccinia-gp160 priming followed by gp160 boosting included longer intervals, and some volunteers vaccinated with rgp160 (produced by Vero tissue culture cells) received five doses at 1-month intervals. We considered only the results from the first year of the studies because some studies were extended for additional booster vaccinations; the 1-year serum was obtained and tested before boosting. This was done to allow comparison among studies for duration of seropositivity.

Investigators and clinic coordinators at each unit were queried to determine if volunteers had been incorrectly identified as HIV infected by external groups; selected cases of misinterpretation of HIV serologic results will be discussed.

Results

Top Methods Results Discussion Author & Article Info References

Volunteers were vaccinated with rgp160/IIIB produced in a baculovirus expression system (n = 41), rgp160/IIIB produced in Vero tissue culture cells (n = 70), rgp120/SF2 (n = 20) produced in yeast cells (n = 48), rgp120/IIIB (n = 20) or rgp120/MN (n = 16) produced in Chinese hamster ovary tissue culture cells, or live vaccinia vector expressing rgp160/IIIB (HIVAC-1e) (n = 71) followed by rgp160/IIIB boost (Table 1). Overall, 62% of volunteers had positive ELISA results for HIV-1 on one or more occasions shortly after vaccination. Depending on the study, 65% to 80% of volunteers receiving an rgp160 vaccine or recombinant vaccinia had positive ELISA results (Table 1). The rgp120 vaccines stimulated antibody that reacted with the commercial ELISA tests less often than did the rgp160 vaccines, but 5% to 56% of volunteers receiving these vaccines had positive test results. Antibody levels waned, but 6 months after the third vaccine dose or the 6-month booster, depending on the study, 31% to 58% of persons receiving the rgp160 vaccine remained positive by ELISA (Table 1). After 6 months, 0% to 16% of persons receiving rgp120 vaccine remained positive by ELISA.

Persons receiving vaccines had antibodies on Western blot to gp160, gp120, and gp41 (persons vaccinated with gp160) or to gp160 and gp120 but not gp41 (persons vaccinated with gp120) (Figure 1). In contrast, infected persons generally exhibit antibodies to core proteins (p55, p24, and p17) and polymerase (p66, p51, and p31), as well as to envelope glycoprotein Figure 1. Depending on the vaccine given and criteria used to interpret the Western blot results, 0% to 63% of vaccinated persons may test as "positive" at some time after vaccination (Figure 2). The different criteria resulted in large differences in interpreting the results of HIV serologic tests. Positive ELISA results and gp41 and gp120/160 envelope band Western blot reactivity fulfill the criteria for a positive HIV test result of the Association of State and Territorial Public Health Laboratory Directors/Centers for Disease Control and Prevention (CDC) [7] but not the criteria of the Western blot manufacturer as listed in the Food and Drug Administration-approved package insert Figure 2 [8].

View larger version (69K): [in this window] [in a new window]   Figure 1. Examples of vaccine-induced antibodies detected by Western blots. Reactive Western blots of serum samples from volunteers who received vaccines for HIV-1, including rgp160/IIIB produced in a baculovirus vector system (lane 1), rgp160/IIIB produced in Vero cells (lane 2), rgp120/SF2 produced in yeast cells (lane 3), rgp120/MN produced in Chinese hamster ovary cells (lane 4), and live vaccinia vector expressing rgp160/IIIB and boosted with rgp160 produced in a baculovirus vector system (lane 5). A control serum from a person with HIV infection is represented in lane 6.

View larger version (61K): [in this window] [in a new window]   Figure 2. Frequency of volunteers' serum samples fulfilling criteria for HIV seropositivity. Development of test results interpreted as positive for HIV infection among uninfected, low-risk recipients of HIV vaccine according to four sets of criteria. The criteria of the Association of State and Territorial Public Health Laboratory Directors/Centers for Disease Control and Prevention (ASTPHLD/CDC) includes a positive ELISA result and any two of p24, gp41, or gp120/160 bands. The criteria of the Consortium for Retrovirus Serology Standardization (CRSS) includes a positive ELISA result and 2 or more bands of p24 or p31 plus gp41 or gp120/160. The criteria of the American Red Cross (ARC) includes a positive ELISA result and 3 or more bands, 1 each of gag, pol, and env. The criteria of the Food and Drug Administration (FDA)-approved package insert (Cambridge Biotech) includes a positive ELISA result and the presence of p24, p31, and either gp41 or gp120/160. * indicates that no volunteers were HIV positive according to the FDA criteria. A = rgp160/IIIB virus (produced by baculovirus); B = rgp160/IIIB virus (produced by Vero tissue culture cell); C = rgp120/SF-2 virus (produced by yeast cells); D = rgp120/IIIB virus (produced by Chinese hamster ovary [CHO] cells); E = rgp120/MN virus (produced by CHO cells); F = live recombinant vaccinia expressing rgp160 (HIVAC-1e) followed by rgp160/IIIB boosting.

As part of the consent process, volunteers were informed in detail that antibodies to HIV stimulated by the vaccines might be detected by commercial diagnostic tests for HIV infection and that adverse social risks were possible. These social risks include health and life insurance denial, denial of enlistment in military service or the Peace Corps, and inability to donate blood. To protect volunteers and to reduce the social risks of participating in these studies, volunteers were given an identification card to verify enrollment in a study. The identification card included a toll-free telephone number for physicians, public health agencies, insurance companies, or other groups that conduct HIV antibody tests to call to confirm the volunteer's participation in an HIV vaccine study. The study site provides detailed information to testing agencies when release of information forms are received. We describe two representative cases to show potential adverse events after HIV vaccination and our approach to solving these problems.

Case 1

A college student gave voluntary informed consent for participation in a double-blind, placebo-controlled phase I clinical trial of an HIV vaccine being tested in healthy volunteers at low risk for HIV infection. At the same time, the student was completing Reserve Officer Training for service in a branch of the U.S. military. The volunteer anticipated the possible development of HIV seropositivity after participation in trials and made preliminary arrangements for special testing with the director of the Division of Retrovirology of the Walter Reed Army Institute of Research. This reference laboratory for the Department of Defense provides testing to ascertain with a high level of certainty if participants in clinical HIV vaccine trials have become seropositive because of vaccine or because of natural infection. At the end of the volunteer's participation in the trial, anti-HIV reactivity was apparent in ELISA and Western blot test results (strong positive bands to specific envelope glycoproteins corresponding to vaccine antigens only). The test results were explained to the volunteer.

After a summer of active duty, during which the U.S. military serologically screened all participants for HIV infection by ELISA, the volunteer was called to the Commanding Officer. A board of five officers, including medical officers and a chaplain, notified the volunteer of "HIV positivity." The volunteer presented to the assembled officers the NIAID HIV Vaccine Program participant card; the medical officers knew of HIV vaccine research and contacted the respective NIAID AIDS Vaccine Evaluation Unit. After the volunteer signed a release of information form, the Unit confirmed the volunteer's previous serologic status and post-trial seropositivity. The volunteer was permitted to remain on active duty.

Case 2

After completing a 2-year participation in an HIV-1 vaccine study, a volunteer developed the characteristic pattern of immune responses to vaccine, with positive ELISA results and envelope bands on Western blot. The volunteer applied for life insurance; the physical examination included an unspecified blood test, the result of which was positive for antibodies to HIV. The volunteer was denied coverage because of the positive test result.

Before clinical trials of candidate AIDS vaccines began 6 years ago, the Division of AIDS of the NIAID contacted health and life insurance organizations to inform them of the problems created by vaccine-induced seroconversion in vaccine recipients. In January 1990, the Division of AIDS sent letters to the Health Insurance Association of America, Blue Cross and Blue Shield Association, Group Health Association of America, Inc., and the American Council of Life Insurance to update them and ask for their continued commitment to nondiscrimination. Several hundred letters of understanding were received from insurance companies.

The insurance company to which the volunteer applied is one of the largest companies in the United States and was included in discussions of policy before the study. The Division of AIDS wrote a letter to the home office of the insurance company explaining the situation and asking the company to reconsider. The company did not respond. Medical personnel at the site where the volunteer was vaccinated had extensive discussions with the local insurance agent and resolved the problem. The volunteer could then be insured.

Discussion

Top Methods Results Discussion Author & Article Info References

Persons interpreting Western blot test results should be aware that participation by initially seronegative volunteers in clinical trials of HIV-1 vaccines may result in the development of detectable immune responses to HIV. Antibodies reactive in the Western blot assay are among the earliest detectable humoral responses. Several types of ELISA for screening serum samples are used today. The widely used whole-virus lysate-based ELISA test becomes reactive after repeated administration of HIV envelope subunit vaccines in most persons receiving the rgp160 vaccine and in fewer persons receiving the rgp120 vaccine. The duration of the antibodies is not yet known; depending on the vaccine used, 0% to 58% of volunteers (33% overall) remained positive by ELISA 6 months after the third vaccine dose. As more complex vaccines are tested and as the immunogenicity of vaccines is improved by the use of more powerful adjuvants and higher doses of vaccine, the rate and duration of ELISA positivity can be expected to increase.

Interpretation of Western blot test results may be complicated by the presence of weakly reactive bands found in serum samples from volunteers before (and therefore presumably after) immunization (Figure 3). Uninfected persons applying as potential volunteers for the phase I studies of candidate vaccines conducted by units of the AIDS Vaccine Clinical Trials Group are frequently noted to have one, or a few, Western blot bands before vaccination [9]. In addition, weak, transient bands may be noted in serum samples from the same uninfected person; these relate to slight differences in reactivities among Western blot test kit lots of the same manufacturer and among different manufacturers [10, 11]. A pattern of Western blot bands after vaccination that suggests reactivity to several components of HIV may complicate the distinction between antecedent bands and those induced by vaccine. Although gp120 vaccines theoretically would yield only antibody bands at gp160 and gp120 on Western blot (and therefore would not be positive according to the CDC criteria) 35% of volunteers given the rgp120/SF2 vaccine (produced by yeast cells) had positive ELISA results and a p24 band that led to positive results when combined with the vaccine-induced envelope band Figure 2.

View larger version (87K): [in this window] [in a new window]   Figure 3. Examples of indeterminate Western blots and subsequent reactivity after immunization. Left. Low risk, uninfected volunteers before immunization with indeterminate Western blot bands at p17 (volunteer A); p24 (volunteers B and C); p24 and P55 (volunteer D); p24, p66, and p80 (volunteer E); and p55, p66, and p80 (volunteer F). Right. Lanes 2 to 8 of serial Western blots from an uninfected person receiving a gp160 vaccine (arrows indicate vaccine days). The p24 band was preexisting and varied in intensity depending on the lot of Western blot test kit used to monitor the development of antibody after vaccination. The presence of p24 bands was common among low-risk, uninfected volunteers and complicated the interpretation of the Western blot test results.

Striking differences in rates of positivity were found by applying different sets of published criteria for Western blot interpretation. Use of criteria of the Association of State and Territorial Public Health Laboratory Directors/CDC yielded the highest rate of positive test results; overall, 44% of results were positive. On the basis of other published criteria, 0% to 8% of test results were positive. When the criteria approved by the FDA for the manufacturer test kit package insert were used, no volunteers had positive HIV test results. Testing agencies applying any of the above criteria, but particularly those using Association of State and Territorial Public Health Laboratory Directors/CDC criteria, must consider the patient's vaccine history. Published criteria featuring a disclaimer that the criteria are not valid for HIV vaccine recipients would reduce the social risks of volunteering in an HIV vaccine study.

To prevent misclassification of volunteers, a more sophisticated approach than Western blot testing alone will be required when more complex HIV vaccine candidates are tested. These investigational vaccines are derived from genetically engineered recombinant products or from whole inactivated HIV and may include envelope and nonenvelope antigens. A study with a recombinant vaccinia that expresses env, gag, and pol antigens has recently begun, and vaccines consisting of empty particles of HIV containing env, gag, and pol antigens have been proposed for study this year. These complex vaccines are expected to induce immune responses that are indistinguishable from HIV infection when standard ELISA and Western blot technology are used. Culture for HIV and polymerase chain reaction (PCR) will be done to confirm uninfected status when these volunteers are followed.

Worldwide, physicians and other health care personnel, government agencies, and public health organizations must acknowledge these vaccine studies and modify criteria for diagnosis of HIV infection for vaccine volunteers. European studies with HIV vaccines will probably involve an increasing number of volunteers, and efficacy studies of HIV vaccines are anticipated to include persons from developing nations as well as those from developed nations. To prevent misclassification of an initially seronegative, uninfected participant in an HIV vaccine trial, persons interpreting diagnostic test results must consider the volunteer's HIV vaccine immunization history. To maintain the ability to distinguish with confidence between infection and vaccine-induced immune responses, confidential access must be available to information about the particular vaccine formulation administered and the selective use of appropriate laboratory assays such as PCR [12, 13]. Participation in NIAID-sponsored trials can be verified through the trial site. With the volunteer's consent, medical information such as data on PCR results or viral culture can be released from the site. It should be noted that past participation in an HIV vaccine study does not guarantee uninfected status, and more complete testing with culture for HIV or PCR to verify uninfected status may be necessary. This is currently provided at no cost to all persons receiving NIAID-sponsored vaccines through the study site.

Appendix

The following are members of the NIAID AIDS Vaccine Clinical Trials Group: St. Louis University, St. Louis, Missouri: Geoffrey J. Gorse, MD, Donald Kennedy, MD, and Sharon Frey, MD; Johns Hopkins University, Baltimore, Maryland: John Lambert, MD, and David Schwartz, MD, PhD; University of Rochester, Rochester, New York: Raphael Dolin, MD, and Norbert Roberts, Jr., MD; University of Washington, Seattle, Washington: Julie McElrath, MD, PhD, and Robert Coombs, MD, PhD; Vanderbilt University, Nashville, Tennessee: Peter Wright, MD, and David Karzon, MD; National Institute of Allergy and Infectious Diseases, Bethesda, Maryland: Patricia Fast, MD, PhD, and Mary Clare Walker, PhD; University of Alabama at Birmingham, Birmingham, Alabama: Jiri Mestecky, MD, and Susan Jackson, PhD; the EMMES Corporation, Potomac, Maryland: Donald Stablein, PhD, and Mark Wolff, PhD.