Vaccines for preventing or treating genital herpes would be a major public health advance. Several observations support the protective effect of preexisting HSV-specific immune responses in limiting the acquisition of HSV-2, the predominant virus type isolated from the genital tract: 1) Reinfection of persons with a second strain of the same viral subtype in the same anatomical area rarely occurs [12-15]; 2) acquisition of neonatal HSV-2 is uncommon among infants exposed to the virus at the time of delivery if the mother is HSV-2-seropositive [16, 17]; and 3) previous HSV-1 infection of the oral-labial region reduces the acquisition and severity of subsequent HSV-2 infection by 40% to 60% [7, 8, 18-22].

Neutralizing antibodies to HSV are predominantly directed to the viral surface glycoproteins, especially to the two essential and abundant glycoproteins D (gD) and B (gB) [23-25]. Monoclonal antibodies to gB and gD protect mice from experimental challenge with HSV [26, 27]. Herpes simplex virus type 2 glycoprotein D (gD2) and B (gB2) have been produced by the use of recombinant DNA technology in Chinese hamster ovary cells as carboxyl terminal-truncated derivatives [28, 29]. These recombinant proteins have been shown to protect guinea pigs from experimental genital challenge with HSV [30-33]. The extent of protection primarily depends on maintaining conformational epitopes of the protein and the potency of the adjuvant used [33, 34]. Immunization with formulations of the two recombinant proteins in the MF59 adjuvant emulsion, which contains squalene, polysorbate 80 (Tween 80, ICI America, Wilmington, Delaware), and sorbitan trioleate (Span 85, ICI America), increases humoral and cellular immune responses and induces protection from experimental challenge in animals [33-35]. Immunization with vaccines containing the recombinant proteins gB and gD combined with the MF59 adjuvant also reduces the rate of subsequent recurrences of genital herpes among previously infected guinea pigs [36, 37].

Our investigations were the initial phase I and II clinical trials to evaluate an HSV-2 subunit vaccine that contained derivatives of gB2 and gD2 in the MF59 adjuvant emulsion and was administered to persons seronegative for HSV or seropositive only for HSV-1 antibodies. These serologic groups represent the target population for a vaccine that prevents the acquisition of HSV-2, the major pathogen of genital herpes.

Methods

Top Methods Results Discussion Author & Article Info References

Patients

We analyzed 137 patients who had no evidence of HSV-2 antibodies in their enrollment serum samples. Patients were enrolled in two phase I and two phase II clinical trials (Table 1); 65 patients lacked antibodies to both HSV-1 and HSV-2 at entry (HSV-seronegative), and 72 had only HSV-1 antibodies at entry (HSV-1-seropositive), as determined by HSV Western blot assay [7, 38]. The two initial phase I, open-label trials were done in a dose-escalating format: The first patients received 10-µgrams doses of gD2 and gB2, the second cohort received 30 µg of each protein, and the last cohort received 100 µg of each protein Table 1. In the subsequent two phase II trials, patients were randomly assigned to receive one of these three doses. Because the demographic, side-effect, and immunogenicity profiles of the 23 patients enrolled in the dose-escalating trials and the 114 patients enrolled in the randomized trials were similar, the data have been combined and analyzed together. All patients were recruited from three metropolitan areas, were in excellent health without chronic disease or evidence of genital HSV-2 infection, and were not receiving long-term immunosuppressive therapy or acyclovir. The 114 patients enrolled in the phase II trials were selected from 313 volunteers originally screened for the trial, of whom 83 were rejected because they had HSV-2 antibodies. Fifty of the 67 patients seronegative for HSV and 64 of the 163 patients seropositive for HSV-1 were enrolled. Vaccine was administered intramuscularly into the deltoid at 0, 1, and 6 months. Serum samples for HSV-2 antibody assays were obtained before enrollment and at study days 0 (before immunization), 28, 42, 180, 194, 270, and 360. All patients were examined in the clinic for 30 minutes after each injection and again 24 hours after injection. In addition, patients were given diary cards to record local symptoms (pain, erythema, and induration), constitutional symptoms (headache, fever, malaise, and myalgia), and any disruption of activities on a daily basis for the 7 days after each injection.

Side effects were classified as mild, moderate, or severe according to the following definitions: mild = transient, with no limitation in activity; moderate = moderate effect on the patient's daily activity; and severe = medical intervention required, with a marked effect on patient's daily activity. Occurrences of erythema or induration were recorded as 0 to 10 mm, greater than 10 to 30 mm, greater than 30 to 50 mm, and greater than 50 mm on the basis of the maximal diameter measured during the visits or from diary cards for the 7-day observation period after each injection. The maximum severity of each symptom recorded during this 7-day period is similarly presented. Blood was drawn on days 0, 1, and 14 for the assessment of serum chemistries, liver functions, leukocyte counts, and erythrocyte sedimentation rates.

A single lot of vaccine was used for each antigen dose for all vaccinations. The protocol was approved by the institutional review board at each institution, and each patient signed an informed consent form for the screening and vaccination protocol regimen. Patients were monetarily compensated for their time and were instructed to avoid sexual exposure to genital lesions and persons with HSV-2 infection. To compare the HSV gB2, gD2, and neutralizing antibody titers in vaccine recipients with those found in naturally infected persons, we analyzed these HSV antibody responses in unimmunized persons whose HSV Western blot results indicated the presence of only HSV-1 antibodies (n = 249) or only HSV-2 antibodies (n = 133). These persons were being screened for participation in other vaccine trials. Ninety-eight (74%) of the 133 persons seropositive for HSV-2 had frequently recurring (4 to 14 episodes per year), culture-proven, symptomatic genital HSV-2 infection [39]. We did not collect information on the symptoms of the remaining 35 patients seropositive for HSV-2 and the 249 patients seropositive for HSV-1. Thus, the naturally infected comparison group represented a cross-section of persons with subclinical and clinical HSV-1 and HSV-2 infection. Because neutralizing antibodies to HSV-2 have not been found to vary with symptomatic reactivations and because the HSV-2 titers of the 98 symptomatic and the 35 other HSV-2-seropositive patients did not differ, we used the initial serum sample that was collected for assaying the gB2, gD2, and HSV neutralizing antibodies among the naturally infected cohort [39, 40].

Laboratory Studies

Antibodies to HSV-2 gB and gD were measured by an enzyme-linked immunosorbent assay (ELISA) as previously described [41]. In brief, 50 µL of recombinant proteins diluted to 6 µg/mL were separately dispensed into 96-well microtiter plates and incubated for 1 hour at 37 °C. Unbound antigen was removed by washing. Serum samples were initially diluted 1:27, the lowest dilution that gave no background value in seronegative patients. Serial dilutions that increased threefold from 1:27 to 1:59 049 were made in the plate, and the plates were incubated for 1 hour at 37 °C. Unbound antibodies were removed by washing, and bound antibody-antigen complexes were detected by goat antihuman IgG conjugated to horseradish peroxidase, with subsequent colorimetric development with o-phenylenediamine dihydrochloride (Sigma, St. Louis, Missouri) as the substrate. Optical densities were read at 490 nm with a Titertek spectrophotometer (Flow Laboratories, McClean, Virginia), and titers were determined as the reciprocal of the dilution at which an optical density of 1.0 was reached. A standard human HSV-positive serum sample (Boston Biomedica, Inc., West Bridgewater, Massachusetts) was included on each plate, and the final titer was adjusted so that it matched the standard curve. All serum samples were run in duplicate against both antigens, and reported titers represent the average of the two assay points. In addition, a second internal standard was run on each plate to confirm the internal consistency of the assay. The coefficients of variation for the gB2 and gD2 assays were approximately 34% and 30%, respectively.

Neutralizing HSV-2 antibody titers were measured with a complement-dependent microneutralization assay. Serum samples were diluted with a twofold serial dilution and then combined with 1.7 x 10² plaque-forming units of HSV-2 strain 333 [42] and an equal volume of guinea pig complement at a 1:125 dilution (Gibco BRL, Gaithersburg, Maryland) for 2 hours at 37 °C. The mixture was subsequently transferred to a 96-well microtiter plate that contained confluent Vero cell monolayers and was incubated at 37 °C in a CO₂ incubator for 65 to 72 hours. The medium was aspirated, and cells were fixed and stained with crystal violet in 4% formaldehyde. The neutralization titer was reported as the reciprocal of the serum dilution that inhibited cytolysis of the cell monolayer by 50%. All plates contained an HSV-positive human serum of a known neutralization titer as a standard for normalization between assays. All laboratory assays were done without knowledge of vaccination status or dose.

Estimation of Herpes Simplex Virus Glycoprotein D2-Specific T Cells

To evaluate the ability of the vaccine to elicit T-cell responses to HSV glycoproteins, we measured the frequency of gD2- and gB2-specific T cells in peripheral blood mononuclear cells (PBC) before and after vaccination among all 15 HSV-seronegative patients in one phase I trial Table 1, 7 of 8 HSV-1-seropositive patients in another phase I trial, and 10 HSV-2-seropositive patients with culture-proven recurrent genital HSV-2 infection. Peripheral blood mononuclear cells were isolated from heparinized blood that was obtained before and after vaccination and were stored in vapor-phase liquid nitrogen ( –195°C) until all study specimens were obtained; the cells were then thawed and assayed in parallel. To estimate the precursor frequency of memory T cells to HSV gB2 and gD2, we used an adaptation of the limiting dilution analysis [43-45]. Thawed cells were resuspended in RPMI-1640 (Roswell Park Memorial Institute medium, JRH Biosciences, Leneka, Kansas); were supplemented with 2 mM of glutamine, 1 mM of sodium pyruvate, 5 mM of HEPES (n-2-hydroxyethylpiperazine-n'-2-ethane sulfonic acid) with a pH of 7.2, 50 µg/mL of gentamicin, and 10% pooled human sera; and were counted in a hemocytometer. Viable cells were seeded into 96-well microtiter plates at a concentration at which 37% of the wells were expected to contain no HSV-specific precursor cells. For assays in which a low number of responders was expected, 70 000 to 100 000 PBCs were seeded per well. For samples in which preliminary experiments suggested a high number of antigen-specific responder T cells, 12 000 cells per well were used. For each assay, 48 wells containing the patients' PBCs were incubated with 1 µg of gD2, 48 wells with 1 µg of gB2, 36 or 48 wells with media alone, 4 wells with 1 µg of phytohemagglutinin, and 9 to 44 wells with 2 Lf/mL (flocculation units) of tetanus toxoid antigen. After 4 days of culture (2 days with phytohemagglutinin), the cells were pulse-labeled with Hydrogen-3-thymidine, lysed onto glass fiber filters, and counted in a scintillation counter. Antigen-reactive wells were defined as those in which counts per minute were 3 standard deviations above the mean cutoff for media control wells. Frequency estimates were calculated as the number of T cells per 10⁶PBCs using a specialized computer program written for this application [46, 47]. When other aliquots of several samples were retrieved and tested at a different time, the intra-assay variability for the estimated number of HSV gD2- and gB2-specific T cells was less than 0.3 log₁₀.

Statistical Analysis

Geometric mean titers were calculated for gB2 and gD2 ELISA antibody titers, and HSV neutralizing antibody titers by vaccine group and study day. A 95% CI for the geometric mean titer was calculated by exponentiating the lower and upper limits of the 95% CI for the mean of the log_e titers for each vaccine group and study day in the following manner: (mean log_etiters) ±{t (0.975, n-1) X (standard deviation log_e titers)/n}. All CIs are 95% and are reported in the following format: (CI, GMT X/÷ multiplication factor). For example, to illustrate this format, a CI of 7474 X/÷ 3.077 corresponds to a 95% CI of 2429 to 22 997. In addition, for each antigen and vaccine group, we calculated the geometric mean ratios of the titer values 28 and 194 days after immunization to the titer values before immunization (day 0) [48]. Differences among the three vaccine groups in mean log_e titers (gB2 or gD2 ELISA titers and HSV neutralizing antibody titers) or mean log_e (titer on day 28 [or day 194]/titer on day 0) ratios of titers were assessed using one-way analysis of variance with a single factor for vaccine group. If overall vaccine group differences with analysis of variance were statistically significant (P 0.05), pairwise t-tests of the means were done to assess pairwise differences between vaccine groups. Differences between individual pairs of vaccine groups were reported as statistically significant only if the overall analysis-of-variance group comparison and the individual pairwise comparison were significant at the 0.05 level or less. Within vaccine groups, changes in gB2 and gD2 ELISA titers and HSV neutralizing antibody titers from before immunization (day 0) to after immunization (days 28 and 194) were assessed using the paired t-test and Wilcoxon signed-rank test. All statistical tests were done at the two-sided 5% significance level [49].

Results

Top Methods Results Discussion Author & Article Info References

Between May 1991 and August 1992, 65 HSV-seronegative and 72 HSV-1-seropositive patients were enrolled (Table 1). The median age of the patients was 35 years (range, 18 to 76 years); 93% were white, 3% were black, and 4% were Asian or Hispanic. Twenty-nine patients (5 HSV-seronegative and 24 HSV-1-seropositive) received the 10-µgrams dose, 52 (29 HSV-seronegative and 23 HSV-1-seropositive) received the 30- µgrams dose, and 56 (31 HSV-seronegative and 25 HSV-1-seropositive) received the 100-µgrams dose. Of the 137 patients enrolled, 32 (23%) did not complete the respective protocols. Twenty-two withdrew from the study because of administrative reasons: Eighteen patients moved from the area or did not return at the appointed times for vaccination (without a report of adverse events), and 4 experienced conditions that precluded (by protocol specification) continuation in the study, such as pregnancy, diagnosis of diabetes mellitus at study entry, baseline chronic Bell palsy and chronic sinus infection, and chronic recurrent pancreatitis. Ten patients (7.3%) withdrew before the trials were completed because of side effects associated with vaccine (see below). We used all available patient information in the safety analyses.

Antibody Responses to Glycoproteins B and D

All immunized HSV-1-seropositive patients showed at least a threefold increase from baseline titers of gB2- and gD2-specific antibody after the first vaccine dose (Figure 1). The geometric mean titer for gB2 among 133 persons with naturally acquired genital HSV-2 infection was 3559 (range, 410 to 25 362); the titer for gD2 was 1209 (range, 76 to 6716). Among the HSV-1-seropositive patients, one dose of the 30-µgrams vaccine elevated the ELISA geometric mean titer for gB2 to 18 559 (CI, 18 559 x/÷ 1.45) and elevated the titer for gD2 to 3980 (CI, 3980 x/÷ 1.71). These titers were five-and threefold greater than those seen in persons with naturally acquired genital HSV-2 infection. After three doses, all HSV-1-seropositive patients who received either 30 µg (n = 15) or 100 µg (n = 14) of vaccine had gB2 titers greater than 9100 and gD2 titers greater than 3600, that is, in excess of the geometric mean titer seen in patients with naturally acquired HSV-2 infection. For each dose (10, 30, or 100 µg), there were statistically significant increases in gD2 and gB2 ELISA antibody titers from before immunization to after the first immunization (P = 0.01 and 0.001, respectively, for the increases in gD2 and gB2 titers at day 28 for those who received the 30-µgrams dose) and after the third immunization (P = 0.0002 and 0.0009, respectively, for the increases in gD2 and gB2 titers at day 194 for those who received the 30-µgrams dose). There were no significant differences among the three dosage levels in ELISA titers at day 28 or day 194.

View larger version (31K): [in this window] [in a new window]   Figure 1. Geometric mean enzyme-linked immunosorbent assay(ELISA) titers to herpes simplex virus type 2 (HSV-2) glycoproteins D2 and B2 before and after immunization with recombinant subunit HSV-2 vaccine in MF59. Arrows denote days of vaccination (days 0, 28, and 180). The standard error and 95% CIs as compared with baseline values are given in the text. The dose of each glycoprotein at each injection and the number of patients receiving each dose are shown. The geometric mean titer (GMT) to glycoproteins B2 and D2 for persons with naturally acquired HSV-2 infection are 3560 and 1209, respectively (see text). HSV = herpes simplex virus.

In the HSV-seronegative population, the 30-µgrams and 100- µgrams doses were more immunogenic than the 10-µgrams dose (P = 0.025 at day 194) (Figure 1). For gB2, statistically significant differences in antibody titers were noted after the second immunization (P < 0.02 for days 42 through 360). For gD2, statistically significant differences in antibody titers were observed after the third immunization (P < 0.05 for days 194 to 310). For example, at day 194, 2 weeks after the third immunization, the ELISA geometric mean titers for gB2 were 7474 for the 10-µgrams dose (CI, 7474 x/÷ 3.08), 14 158 for the 30-µgrams dose (CI, 14 158 x/÷ 1.26), and 21 066 for the 100-µgrams dose (CI, 21 066 x/÷ 1.31). Similarly, the ELISA geometric mean titers for gD2 at day 194 were 2514 for the 10- µgrams dose (CI, 2514 x/÷ 4.68), 8094 for the 30- µgrams dose (CI, 8094 x/÷ 1.47), and 6410 for the 100- µgrams dose (CI, 6410 x/÷ 1.38). Specifically, gB2 titers of at least 3600 and gD2 titers greater than 1200 were measured in 21 of 23 (91%) patients receiving the 30-µgrams dose after the second immunization and in 22 of 23 patients after the third immunization.

Neutralizing Antibody Responses

The median HSV-2 neutralizing titer in the 133 patients with naturally acquired HSV-2 antibodies was 136, whereas the titer for the 249 patients with naturally acquired HSV-1 infection was 68. Only 9% of these HSV-1-seropositive patients had HSV-2 neutralizing titers greater than 136 (Figure 2, top). Two weeks after the third immunization with the 30-µgrams vaccine, both HSV-seronegative patients and HSV-1-seropositive patients developed HSV neutralizing antibody responses similar to responses in persons with naturally acquired infection (Figure 2, bottom).

View larger version (28K): [in this window] [in a new window]   Figure 2. Top. Herpes simplex virus type 2 (HSV-2) neutralizing antibodies in persons with naturally acquired herpes simplex virus type 1 (HSV-1) and HSV-2 infection. One hundred thirty-three patients were seropositive for only HSV-2 by Western blot, and 249 patients were HSV-1-seropositive only. Bottom. Herpes simplex virus type 2 neutralizing antibodies in patients 2 weeks after the third immunization with the 30- µgrams vaccine. Twenty-three patients were HSV-seronegative, and 15 were HSV-1-seropositive. Asterisk indicates the dilution at which there is 50% inhibition of cytolysis of the cell monolayer.

View larger version (25K): [in this window] [in a new window]   Figure 3. Geometric mean herpes simplex virus neutralizing antibody titers for patients who had all three immunizations. Seropositive and seronegative status refers to status at study entry; arrows denote days of vaccination (days 0, 28, and 180). The dose of each glycoprotein and the number of patients receiving each dose are shown. The geometric mean titer neutralizing titer ±SE for persons with naturally acquired herpes simplex virus type 2 (HSV-2) was 141 (CI, 141 x/÷ 1.14) (see text and Figure 2, top). HSV-1 = herpes simplex virus type 1.

Neutralizing titers took longer to develop in the HSV-seronegative population and were higher in recipients of the 30- µgrams and 100-µgrams doses. None of the HSV-seronegative patients developed neutralizing antibody titers of 136 or more after a single immunization, whereas 6 (11%) had titers of 136 or more after two immunizations and 38 (72%) reached that level after three immunizations. Herpes simplex virus type 2 neutralizing titers of 136 or more were elicited by three immunizations in 2 of 5 patients (40%) who received the 10-µgrams dose, 18 of 25 patients (72%) who received the 30- µgrams dose, and 18 of 23 patients (78%) who received the 100- µgrams dose. The 30-µgrams and 100-µgrams doses generated significantly higher geometric mean titers than did the 10-µgrams dose (P = 0.025). At day 360, 20 of 23 patients (87%) who received the 30- µgrams dose and 23 of 25 patients (92%) who received the 100- µgrams dose had measurable neutralization titers (>1:12) (Figure 3). However, after 6 months, titers waned in the HSV-seronegative patients, with only 9% and 8%, respectively, having titers of 136 or more at day 360.

Frequency of Glycoprotein gD2-Specific T Cells

At enrollment, HSV-seronegative patients rarely had detectable memory T-cell responses to gB2 and gD2 (Table 2). By frequency analysis, the median precursor frequency for gB2- and gD2-specific T cells was 0 and 0.4 cells per million PBCs, and only one HSV-seronegative patient had more than 5 antigen-responsive cells per million PBCs at baseline. In contrast, 8 of 10 patients with recurrent genital HSV-2 had detectable baseline responses to both gD2 and gB2, whereas the other two patients responded to gD2 alone. The median number of HSV-2 gD2-reactive T cells was 34.1/million PBCs compared with 12.9 million gB2-reactive T cells/million PBCs.

Vaccination with the gD2-gB2-MF59 vaccine was associated with a prompt and sustained memory T-cell response to gB2 and gD2 (Table 2). After the second dose of vaccine at day 42, gD2- and gB2-specific T cells were detected in 13 of the 15 (87%) and 14 of the 15 (93%) HSV-seronegative patients who were tested; among patients who received the 30-µgrams dose, antigen-reactive T-cell responses increased a median of 34-fold for gD2 and 40-fold for gB2. Compared with the sample obtained before immunization, the samples of 14 of 15 (93%) HSV-seronegative patients contained 10-fold more gD2- and gB2-specific memory T cells. T-cell responses tended to remain elevated throughout the vaccination period, with median responses similar at days 42 and 180. The median peak T-cell precursor frequencies after vaccination were 68 for gD2 and 32 for gB2.

Tetanus toxoid responses Table 2 and phytohemagglutinin responses did not significantly change during immunization. The tetanus toxoid response did appear to increase after immunization (103.1 compared with 41.2 responders per million PBCs on day 42 compared with day 0).

Side Effects

Vaccination side effects were limited in severity and duration. Transient pain and erythema at the injection site were common, especially in HSV-1-seropositive patients. Overall, 123 of 137 patients (90%) reported pain at the deltoid injection site with at least one of the three injections. The pain was rated as mild in 28 of 68 (41%) HSV-1-seropositive patients and in 28 of 55 (51%) HSV-seronegative patients (Table 3). Pain requiring an alteration in activity or the use of over-the-counter analgesic medication (acetaminophen or nonsteroidal anti-inflammatory drugs) occurred in 3 of 65 (4.6%) HSV-seronegative patients and 5 of 72 (6.9%) HSV-1-seropositive patients. The median duration of pain was 2 days. Injection-site erythema or induration of more than 10 mm was noted in 20 of 65 (31%) HSV-seronegative patients and 32 of 72 (44%) HSV-1-seropositive patients. A greater frequency of moderate to severe pain and a larger area of local erythema were associated with vaccination in HSV-1-seropositive patients who received three 100- µgrams doses compared with those receiving the 30-µgrams or 10- µgrams dose.

Headache, the most frequent systemic side effect, was reported in 14 of 65 (22%) HSV-seronegative patients compared with 29 of 72 (40%) HSV-1-seropositive patients. Fever (temperature >38.3 °C) developed in none of the 65 seronegative patients and in 7 of 72 (10%) HSV-1-seropositive patients. The median duration of fever for all patients reporting fever was 24 hours or less; all patients received antipyretic agents, and all promptly responded. Among the 7 persons with a temperature greater than 38.3 °C, the maximal temperature was 39.9 °C. The frequency and severity of local pain, induration, headache, and fever did not vary with the number of injections (Table 3). In more recent studies in which the MF59 adjuvant was given alone, the frequency of fever (temperature >37.7 °C), headache (mild or more severe), and local pain (moderate or more severe) were 0%, 0%, and 5%, respectively, among vaccinees receiving a 0-, 1-, and 6-month regimen (Corey L. Unpublished data).

Two (3%) of the 65 HSV-seronegative patients and 8 (11%) of the 72 HSV-1-seropositive patients withdrew from the trial because discomfort associated with vaccination caused them to lose time from work or school. This occurred in 6 patients after the first dose and in 4 patients after the second dose. The reasons for premature withdrawal were urticaria after vaccination in 2 patients, mild zoster 8 days after immunization in 1 patient, moderate to severe systemic symptoms (fever, chills, headache, myalgia, arthralgia, fatigue, malaise, or nausea) in 5 patients, and moderate to severe local pain and redness in 2 patients.

Discussion

Top Methods Results Discussion Author & Article Info References

This recombinant subunit vaccine containing the truncated derivatives of HSV-2 gB2 and gD2 combined with a novel oil-emulsion adjuvant appears to be highly immunogenic and well tolerated. In these two phase I and two phase II trials involving immunocompetent HSV-2 seronegative adults, the vaccine elicited antibody and memory T-cell responses that equaled or exceeded those measured in persons with naturally acquired HSV-2 infection. The gB2- and gD2-specific antibody ELISA titers and the HSV-2 neutralizing antibody titers in all HSV-1-seropositive patients and in 72% of HSV-seronegative patients after three immunizations were equal to or greater than those seen with naturally occurring HSV-2 infection. The estimated frequency of memory T cells reactive to HSV gD2 and gB2 increased immediately after immunization and remained at levels similar to those detected in persons with HSV-2 infection for at least 5 months after the second immunization. Although local side effects frequently occurred with this vaccine, most reactions were mild.

Using the same laboratory assays, we showed that in all HSV-1-seropositive patients, three immunizations with any of the doses tested induced neutralizing antibody titers for gB2 and gD2 that exceeded those found in patients with naturally acquired HSV-2 infection. Among patients who had HSV-1 antibodies, the response to vaccination was prompt, remained stable between the second and third immunizations, and remained elevated for the 1-year follow-up. Patients seronegative for HSV needed the three-immunization regimen to achieve neutralizing antibody titers equaling those of natural HSV-2 infection. However, with this regimen, 72% of the HSV-seronegative patients who entered the trial achieved this level of neutralizing antibody, as well as gB2 and gD2 antibody levels that equaled or exceeded those seen in natural infection. In the seronegative patients, the HSV-2-specific neutralizing antibodies rapidly decreased in the 12-month period, 6 months after the third immunization. The effect of an additional booster dose is under investigation. Preliminary data suggest that an additional 30-µgrams dose of each protein is associated with an additional boost in antibody titers; peak neutralizing antibodies seem to be nearly twofold higher than those after the third dose, with 100% of HSV-seronegative patients having titers of 136 or more after the fourth 30-µgrams vaccine dose.

Immunization produced a marked increase in our ability to detect circulating memory T-cell response to HSV gD2 and gB2. Of interest, the T-cell responses increased more rapidly than the antibody response and appeared more stable in HSV-seronegative patients than did the neutralizing antibodies. The median precursor frequency elicited by immunization among seronegative patients was similar to that in persons with naturally acquired genital herpes and did not wane substantially by 5 months after the second immunization. Although phytohemagglutinin responses remained stable, we did see a slight twofold increase in memory T-cell responses to tetanus toxoid. The cause and consistency of this "bystander" effect is being studied.

The role of HSV-specific T-cell responses and neutralizing antibodies in limiting the acquisition of genital herpes remains unclear. Recent studies have shown that HSV-2 is transmitted in fewer than 1% of infants exposed to infectious HSV-2 at the time of delivery if the mother is HSV-2-seropositive. In contrast, acquisition rates of neonatal herpes approach 50% in infants exposed to HSV-2- if the mother is still HSV-2-seronegative (recent infection) and approach 25% to 40% if the mother is HSV-1-seropositive only [18, 19]. These data suggest that antibody is important in preventing vertical transmission. In addition, both low antibody-dependent cellular cytotoxicity and low neutralizing antibody levels appear to be associated with high rates of acquiring neonatal herpes [50]. The antibody titers generated in these trials with gB2-gD2 subunit vaccine in MF59 are 100 times greater than those achieved in a previous study of a subunit mixed glycoprotein vaccine in an alum adjuvant (gB and gD ELISA titers were tested in the same assay). This latter glycoprotein-alum vaccine was subsequently proved to be ineffective in protecting HSV-2-seronegative persons from acquiring genital HSV-2 infection [51].

The MF59 adjuvant in this vaccine formulation was an important factor in the increased antibody and T-cell responses that we achieved. In two separate preliminary studies done before initiation of this trial, 30 µg of the gD2 protein used in this vaccine was formulated in alum and MF59 [39]. None of the seronegative patients who received the gD2-alum vaccine developed a neutralizing titer of 136 or more after three immunizations. The peak neutralizing response among the four patients receiving the alum adjuvant was 19, whereas the response among the three patients receiving the gD2-MF59 product was 216. Similarly, peak gD2-specific T-cell responses after four doses of the gD2-alum vaccine averaged just 11 cells per million PBCs compared with 260.5 cells per million PBCs, or 20-fold higher, in the patients who received the gD2-MF59 vaccine.

Side effects associated with vaccination were common but transient, mostly mild, and responsive to over-the-counter doses of either acetaminophen or nonsteroidal anti-inflammatory medications. The frequency and severity of side effects were sporadic and did not increase with the number of injections, nor did the occurrence of systemic side effects after an earlier injection predict recurrence of side effects after a subsequent injection. Patients seropositive for HSV-1 had a greater frequency of side effects than HSV-seronegative patients, which suggests that HSV-specific T or B cells may be involved in the genesis of these side effects, perhaps through local cytokine production. Side effects after vaccination also appear to be related to the concentration of the antigen and to the addition of MF59 adjuvant. In earlier studies in which the same gD2 protein derivative as that used in this study was formulated with alum, systemic reactions such as fever, headache, and myalgias occurred less frequently [39]. These data suggest that the combination of previous HSV serologic status and the antigen-adjuvant mixture influences the reactogenicity to the vaccine. On the basis of the results of these studies, phase III efficacy trials for preventing genital herpes are being done using a 0-, 1-, and 6-month regimen with 30 µg of gB2 and 30 µg of gD2 in each dose.

In summary, this recombinant HSV-2 specific glycoprotein vaccine appears to be highly immunogenic and well tolerated. Persons negative for HSV-2 who receive three 30-µgrams immunizations each of recombinant glycoprotein gD2 and gB2 in the MF59 adjuvant develop antibodies to the major neutralizing glycoproteins of HSV-2 that are 100 times greater than that achieved by previous glycoprotein preparations tested in the same assay [51]. Moreover, 88% of HSV-1-seropositive patients and 72% of HSV-seronegative patients developed neutralizing antibody responses to HSV-2 that equaled or exceeded those in an unselected cohort of persons with naturally acquired HSV-2 infection. The epidemiologic studies showing that HSV-2-specific immune responses appear to protect against reinfection, transmission to the neonate, and transmission between sexual partners all support optimism about the potential utility of such a vaccine. Phase III efficacy trials to evaluate this hypothesis are in progress.