| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
 |
ABSTRACT |
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
INTRODUCTION |
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The design of the current trial was influenced by the possibility that one or more serotypes in the tetravavent mixture could either enhance or suppress the replication, immunogenicity, or reactogenicity of other vaccine serotypes. Limited data in rhesus monkeys suggest that the dose ratio of two serotypes of simultaneously inoculated dengue viruses affects the degree of protection obtained after rechallenge.2 Another example of viral interaction is the trivalent Sabin oral polio vaccine in which serotype 2 interferes with serotypes 1 and 3.6 The current design was also influenced by previous monovalent dengue vaccine trials, in which the vaccine was injected twice to explore the possibility that two vaccinations would be more immunogenic than one.5 A dose interval of one month was chosen as a rapid, practical vaccination schedule for military personnel. In the first clinical trial of the tetravalent DEN 1–4 vaccine, one formulation was administered twice at an interval of one month with acceptable reactogenicity but with poor booster effect. Several of these volunteers later safely received a third vaccination three months after the second injection and a booster effect was noted.5 The current study was designed to provide additional clinical and serologic data on the safety, reactogenicity, and immunogenicity of 15 other tetravalent vaccine formulations. The 15 formulations consisted of either high or low doses of the following four strains of DEN virus attenuated by passage in primary dog kidney (PDK) cell culture: DEN-1 (45AZ5) PDK 20, DEN-2 (S16803) PDK 50, DEN-3 (CH53489) PDK 20, and DEN-4 (341750) PDK 20 (Table 1
). The high doses (~1 x 106 plaque-forming units [PFU] for DEN-1 and DEN-2 and ~1 x 105 PFU for DEN-3 and DEN-4) and low doses (~1 x 104.5 PFU for DEN-1 and DEN-2 and ~1 x 103.5 PFU for DEN-3 and DEN-4) were chosen by consensus of WRAIR and University of Maryland investigators based on previous experience with other attenuated multivalent vaccines, and by the titers of available vaccine candidates bottled and approved for trial by the Food and Drug Administration. As reported below, screening of all 16 tetravalent vaccine formulations in 64 flavivirus non-immune adult volunteers has permitted us to select two or three formulations for future expanded trial.
|
|
MATERIALS AND METHODS |
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
) were initially administered as monovalent vaccines in separate, initial Phase I trials.3–5 The DEN-1, -2, and -3 vaccines were manufactured at the Salk Institute (Swiftwater, PA). The DEN-4 vaccine was produced at the Pilot Bioproduction Facility, WRAIR (Forest Glen, Silver Spring, MD). The vaccines were lyophilized for stability and stored at -20°C in a controlled-access room at the Pilot Bioproduction Facility. The vaccines contain 50 µg/mL of neomycin base, 5.5% lactose, and 1.9 g% of human serum albumin. The manufacturing information for each DEN vaccine was included in a separate IND. All four INDs were combined into a single IND (7047) for testing of the tetravalent vaccine.
Preparation and inoculation of vaccines.
The 16 tetravalent DEN vaccine formulations were prepared under Good Manufacturing Practice conditions on the day of inoculation at the Pilot Bioproduction Facility. Freeze-dried monovalent vaccines were rehydrated with sterile water for injection and formulated as tetravalent vaccines in sterile, eight-dose (8.0 mL) glass vials. The high dose of each monovalent vaccine was undiluted vaccine, while the low dose was a 1:32 dilution of the same vaccine. Vaccine preparations were formulated on ice, and formulations 1–15 were transported immediately to the University of Maryland (College Park, MD) (UMCP) for inoculation. The high dose of all four serotypes (formulation 16) was tested first at WRAIR as described.5 Volunteers were inoculated subcutaneously over the deltoid muscle with 1.0 mL of vaccine. Following vaccination, unused vaccines were immediately transported back to WRAIR to determine viral titer. The viral composition, inoculum size, and number of volunteers immunized with each of the 16 formulations are summarized in Table 1. The final injected concentrations of tetravalent vaccine formulations ranged from 5.5 x 104 to 1.8 x 106 PFU/mL.
Volunteers. Fifty-four healthy male and female volunteers between the ages of 18 and 45 years were recruited at UMCP. The 10 volunteers recruited and immunized at WRAIR have been described.5 The results for all 64 volunteers are combined in the current report. Volunteers were recruited by posted and published advertisements followed by verbal presentation of the study design and protocol. Of 64 volunteers, 37 (58%) were men, 39 were Caucasian, 15 were African-American, 7 were Asian, and 3 were Hispanic. Their mean age was 23.6 years, with an age range of 19–45 years.
Their status of good health was determined by a normal medical history, vital signs, a physical examination if the medical history suggested any abnormality, and clinical laboratory screening. Screening included hemoglobin; hematocrit; white blood cell, differential, and platelet counts; alanine aminotransferase (ALT); aspartate aminotransferase (AST); glucose; creatinine; urea nitrogen; hepatitis B surface antigen; antibody to human immunodeficiency virus; antibody to hepatitis C virus; and urinalysis. Persons allergic to neomycin, streptomycin, gentamicin, or similar antibiotics were excluded. The volunteers were seronegative for DEN 1–4, Japanese encephalitis, St. Louis encephalitis, and yellow fever by hemagglutinin inhibition assay. They denied prior yellow fever and Japanese encephalitis vaccination, having had a flavivirus infection, and travel to dengue-endemic areas. Women had a negative pregnancy test result one hour before vaccination, and were instructed not to become pregnant for the duration of the study. All volunteers scored 70% or better on a written examination designed to insure they were familiar with all aspects of the clinical trial.
Clinical protocols were reviewed and approved by the Institutional Review Boards of the University of Maryland at Baltimore, the University of Maryland at College Park, and the Office of the Surgeon General of the United States Army. Written, informed consent was obtained from all volunteers.
Study design.
This single-blinded, outpatient Phase I/II trial was conducted in 54 healthy young adult volunteers at the University of Maryland at College Park, and in 10 volunteers at WRAIR.5 A similar clinical protocol was used at both study sites. Volunteers did not know which vaccine formulation was injected. The 54 volunteers at UMCP were divided into 15 groups of 3–4 volunteers (Table 1) and each group given a different formulation. The cohorts, initially immunized on 10/15/98, 10/21/98, and 1/19/99, consisted of 15, 16, and 23 volunteers, respectively. The second vaccination was administered 28 days later. Fifty-three of the volunteers at UMCP were inoculated on days 0 and 28; one was immunized on day 0 only. As previously described, the 10 volunteers at WRAIR were immunized with high dose formulation 16; nine of the 10 were immunized on days 0 and 28; four of these nine subjects received a third dose on day 120; one volunteer was immunized on days 0 and 120.5
Reactions to vaccines were assessed by a combination of daily symptom diaries maintained by each volunteer for 21 days after each vaccination, by telephone reports to the research nurse, and by regularly scheduled visits to the study physician on days 10, 15, and 28 after each vaccination. Volunteers took their oral temperature daily for 21 days in the mornings and evenings using a digital thermometer and recorded the results. After each vaccination, blood was drawn for clinical chemistries on days 0, 10, 15, and 28 and for DEN antibody assays on days 0 and 28. Serum was frozen 10 days after each vaccination for virus assay.
Clinical reactions.
The clinical reactions to a vaccine formulation were defined in two ways: as safe or unsafe and as non-reactogenic or reactogenic. The criterion unsafe was met if any volunteer experienced any one of four criteria: 1) any severe clinical illness not explained by a diagnosis unrelated to DEN vaccination; 2) an oral temperature 
38.5°C (101.3°F) for four determinations over a 24-hour period, a maximum daily oral temperature 38.5°C on three successive days, or a temperature exceeding 40°C (104.0°F) on any individual determination; (3) thrombocytopenia (fewer than 90,000 platelets/mm3 on two consecutive determinations, e.g., days 10 and 15) or neutropenia (absolute neutrophil count <1,000/mm3 on two consecutive determinations); and 4) serum ALT levels more than five times normal (normal range = 0–45 U/L) on three or more successive days that was otherwise unexplained. Each symptom was graded on a scale of 0 to defined as 0 = none; 1 = mild (did not affect normal activity, no medication required); 2 = moderate (required medication or change in activity); and 3 = severe (required bedrest, and or unrelieved by medication). Systemic reactions were analyzed for each subject by the Reactogenicity Index (RI), defined as the sum of total days of 1) feeling hot or a temperature 100°F (feverish or fever), 2) rash, 3) chills, 4) headache, 5) anorexia, 6) nausea and/or vomiting, 7) stomach ache 8) myalgia, 9) arthralgia, 10) eye symptoms (photophobia, redness or painful movement), and 11) itching at sites other than the injection site. If a symptom occurred at any time during a 24-hour period, it was assigned a duration of one day.
The RI allowed a quantitative comparison of vaccine reactions among subjects and vaccine formulations. The RI criteria used to grade formulations 1–15 included the duration of the 11 symptoms or signs (vida supra), but not their intensity. The intensity (grades 1–3) of any symptom was expressed in narrative fashion. The more intense reactions (grades 2 and 3) interfered with the daily activities of school, work, recreation, or sleep. In contrast, the RI criteria used by Sun and others to grade formulation 16 described the duration of nine major symptoms, included their intensity.5 The RI formula used to grade formulation 16 was proportional to, but yielded somewhat lower values than the RI used to grade formulations 1–15. Febrile and non-febrile symptoms were treated if clinically indicated with analgesics without antipyretic properties. The dose and duration of all medications used were recorded in the Volunteer Symptom Diary.
The identification of a vaccine formulation that was too reactogenic was reached by a consensus of study investigators. Thus, a vaccine formulation could be safe but too reactogenic, and so unsuited for expanded study.
Laboratory procedures. Routine clinical assays. Venous blood was drawn for a complete blood count, differential counts, and ALT and AST levels on days 0, 10, 15, 28, 38, 43, and 56 in a commercial laboratory accredited by the American College of Pathology.
Serology. The study end point determination was measurement of the 50% plaque reduction neutralization test (PRNT50) titer against DEN serotypes 1–4 28 days after the first and second vaccination.9 The four DEN strains used in the PRNT were the parents of the four stains used in the vaccine formulations. Neutralization was defined as 50% reduction in plaques at a minimum serum dilution of 1:5. Complement was not added to the neutralization mixture.
Virus isolation by delayed plaque assay.
Ten milliliters of venous blood obtained 10 days after each vaccination was allowed to clot at 4°C for 
2 hours. The serum was separated by low-speed centrifugation, divided into aliquots, frozen, labeled, and stored at -70°C before shipment to WRAIR for virus isolation. Frozen serum was thawed and 0.20 mL was inoculated onto the C6/36 clone of Aedes albopictus mosquito cell monolayers and incubated at 30°C for 14 days. Supernatant culture fluid was then harvested and assayed for virus by plaques that formed seven days after the supernatant was applied to a Vero cell monolayer overlayed with methyl red agar. The direct plaque method for determination of virus in serum was not used because of low yield in preliminary studies.
Virus detection by Taqman assay. We used a fluorogenic, reverse transcriptase-polymerase chain reaction (RT-PCR) system (Taqman assay) to determine the serotype of circulating dengue vaccine strains.10 This assay can detect each of the four dengue serotypes at similar low detection limits (20–50 PFU/mL of serum). Briefly, the viral RNA was routinely extracted from 0.10 mL of serum, followed by reverse transcription. The RT reactions were performed on RNA extracted from the equivalent of 1.6 µL of serum, according to the PE Taq RT reaction kit (Perkin Elmer-Applied Biosystems, Inc., Foster City, CA). Using serotype-specific, 3'-noncoding region-based primers and an fluorescent reporter dye, 6-carboxy-fluorescein (FAM)-labeled oligonucleotide probe, a total of 40 cycles of amplification was carried out using the ABI 7700 instrument (Perkin Elmer-Applied Biosystems). Positive identification was made when the two replicates of each serum specimen yielded Ct values less than 40 cycles. The viral titer (PFU/mL) for each dengue type was predicted and estimated from the Ct cycle number using a plasmid standard curve.10 We assayed serum specimens from all 54 volunteers 10 days after the first vaccination (formulations 1–15), and from 53 of these volunteers 10 days after the second vaccination (day 38). Volunteers vaccinated with formulation 16 were assayed by the Taqman assay and are presented elsewhere.5
Data analysis. This was a descriptive study. The design ensured that if three of the 3–4 volunteers per group develop neutralizing antibody after either of two vaccinations, the upper 95% confidence interval for infection exceeds 95% for that vaccine formulation. At least three of the four DEN serotypes in a vaccine formulation must induce antibody in at least three volunteers for that formulation to be considered for further study. Relationships between categorical variables were examined using either Fisher’s exact tests, or replicated goodness-of-fit G-tests, as appropriate. Inter-group comparisons of continuous variables were examined using Wilcoxon’s rank-sum tests. Comparisons of PRNT titers against pairs of DEN types were performed using Wilcoxon’s signed ranks tests. Comparison of RIs by low and high dose DEN formulations were performed using Wilcoxon’s rank-sum tests. Relationships between continuous variables were evaluated using Spearman’s correlation coefficients. Two-sided null hypotheses were evaluated at 5% throughout. Data analysis was performed using SAS software (SAS, Inc., Cary, NC).
|
RESULTS |
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Serologic responses.
Seroconversion rates.
The neutralizing antibody responses induced by each of the 16 vaccine formulations after the first and second vaccinations, together with the cumulative responses, are summarized in Tables 2–5. Seven of the first 15 formulations, (2, 5, 10, 11, 13, 14, 15) were considered superior because they met study criteria by eliciting antibody to at least three of four DEN serotypes in at least three (75%) volunteers in a group of 3–4 individuals (Table 2). Nine formulations were considered inferior because they failed to meet study criteria for immunogenicity (Table 4). The high dose formulation (16) induced a trivalent response in only three of the nine individuals immunized at 0 and 1 months, and in the one person immunized at 0 and 4 months (Table 4).5 There were no significant differences in the seroconversion rates by race or sex for all formulations.
|
|
|
). In 64 volunteers, 38 (59%) seroconverted to three DEN serotypes, and 16 (25%) seroconverted to all four serotypes.
When all 16 formulations were combined, the rates of seroconversion to DEN-1, DEN-2, and DEN-3 were similar (69%, 78%, and 69%, respectively), but significantly fewer volunteers seroconverted to DEN-4 (38%; P < 0.001) (Table 6). Although significantly more volunteers seroconverted to DEN-1, -2, or -3 than to DEN-4, these serotypes in the vaccine did not appear to interfere with DEN-4. For example, after the first vaccination (Table 7), 7 (58%) of 12 volunteers seroconverted after the 105 dose of monovalent DEN-4 vaccine5 compared with 17 (47%) of 36 volunteers administered the 105 dose of DEN-4 in the tetravalent vaccine. The rates of seroconversion after the first vaccination were also similar in the monovalent and tetravalent formulations for the three other DEN serotypes, suggesting no heterologous interference between serotypes. We did not compare the geometric mean titers (GMTs) in the monovalent and tetravalent trials because the PRNT50 assays were performed differently, with complement added to the serum assayed in the monovalent vaccine studies but not in the tetravalent vaccine studies.
|
|
). Of the four serotypes, only DEN-1 showed a relationship between concentration and response rate, with the high concentration being significantly more immunogenic than the low concentration. The high and low concentrations of DEN-2, DEN-3, and DEN-4 gave similar seroconversion rates.
|
and 5), the GMTs within serotypes were similar among the seven acceptable formulations (Table 3). Although the rate of seroconversion to DEN-1 was no greater than to other serotypes, the DEN-1 GMT was significantly higher than the DEN-2 GMT, and the DEN-3 GMT was significantly higher than the DEN-4 GMT (by post-hoc analysis, Table 6).
|
and 5). An antibody titer increase was defined as a PRNT50 titer increase from <1:5 to 1:5, or a 4-fold increase in titer to one or more serotypes; an antibody titer decrease was defined as a 4-fold decrease, or loss of detectable antibody against 1–3 of the vaccine serotypes. After 62 volunteers were re-vaccinated on day 28, 25 (40%) had an antibody titer increase and 28 (45%) had a titer decline. In addition, of 248 viral assays represented in the day 56 data set (62 volunteers x 4 serotypes), nearly equal numbers of sera showed an increase (17%) as a decrease (19%). Thus, re-vaccination at one month offered no discernable benefit. Clinical reactions to formulations 1–15. Local reactions. Twenty-four (44%) of 54 individuals first vaccinated and 14 (26%) of 53 re-vaccinated reported one or more local symptoms and signs at the injection site (pain, tenderness, or erythema) (P = 0.07). Between 0 and 2 volunteers of the 3–4 individuals administered each of the first 15 formulations had local reactions. The one exception was formulation 2, in which all four volunteers developed a reaction after the first vaccination, but none after the second vaccination. Reactions were generally mild; they began on days 0–1 and persisted for 1–3 days. One volunteer developed delayed pain, redness, swelling, and pruritis on days 12, 14, and 15 after the first vaccination, but did not develop a reaction to the second vaccination. One volunteer developed signs of local inflammation that persisted for seven days. No reactions required treatment, and none interfered with activity or sleep.
Systemic reactions.
Viremia and systemic reactions reported by volunteers are summarized in Table 9. The seven superior formulations (2, 5, 10, 11, 13, 14, 15) varied in their mean RI from a high of 36.5 (formulation 2) to a low of 2.0 (formulation 15). The RIs of individual volunteers are shown in Figure 1a for the seven superior formulations and in Figure 1b for the nine inferior formulations. As a group, the superior formulations were more reactogenic, although three formulations (13, 14, 15) appeared to provide an acceptable balance of immunogenicity and reactogenicity. The RIs were similar between African-Americans and Caucasians among the 10 formulations administered to both races (P = 0.38, by a binomial test).
|
|
, together with viremia results. Of 54 University of Maryland volunteers, 13 (24%) curtailed school, work, recreation, or sleep. Their symptoms lasted 1–3 days, and anti-pyretics/analgesics (principally acetaminophen) when administered, provided some relief. All reactions occurred after the first vaccination. Five of these volunteers were afebrile; however, their symptoms were similar in type, intensity, and duration to the seven volunteers who were febrile (Table 10). Formulation 2 was the most reactogenic, with rash persisting eight days in one volunteer and myalgia for seven days in another. No symptoms were reported by any of the 54 University of Maryland volunteers after day 19. Three of 10 volunteers immunized with formulation 16 at WRAIR experienced some loss of activity on the inpatient study ward.5
|
100.0°F. These individuals received six formulations; three of four administered formulations 2, 7, and 10 were febrile, and one of four administered formulations 5, 12, and 13 were febrile. Febrile episodes began 7–16 days after the first vaccination and lasted 1–4 days. The highest fever recorded was 102.2°F. Morbidity among febrile volunteers varied. Four febrile persons reported no loss of activity, despite the fact that in one such individual the fever lasted four days.
Association of reactogenicity and immunogenicity.
There was a positive relationship between the RI and seroconversion to the four DEN serotypes among 54 volunteers immunized with formulations 1–15 (Spearman’s r = 0.35, P = 0.01) (Figure 2).
|
). We found that high dose formulations had no greater reactogenicity. Unexpectedly, the low dose formulations of DEN-2 and DEN-4 were significantly more reactogenic than their high dose counterparts (Table 11), and these two formulations induced the highest mean RIs among all formulations. Among volunteers who received low dose DEN-2 formulations, the RIs were significantly higher in the subset who received high dose DEN-1 compared with low dose DEN-1 (P = 0.04, by Wilcoxon’s rank sum test). The same phenomenon occurred in volunteers who received low dose DEN-3 (P = 0.03) and low dose DEN-4 (P = 0.02). In contrast, this difference in reactogenicity between recipients of high and low dose DEN-1 was not seen in formulations with high dose DEN-2, DEN-3, or DEN-4 (P 0.59).
|
) who developed transient elevation of the ALT level (171 U/L on day 15, 55 U/L on day 22, and 19 U/L on day 28), no other volunteer developed an elevated ALT level (defined as >2 times normal: 45 U/L x 2).
Absolute neutrophil counts, platelet counts, and vaccine reactogenicity.
Among formulations, there was no statistically significant correlation between mean RI and maximum percent fall from baseline in the mean ANC (P > 0.2) or the mean platelet count (P > 0.3, by Spearman’s test). However, an association did exist between higher RIs and change of ANCs and platelet counts from baseline (Figure 3). For example, combining mean maximum % ANC changes from baseline for the combined first and second vaccinations (15 formulations x 2 vaccinations = 30 formulation data points), five (83%) of six formulations with a mean RI 10 developed a 32% mean decrease in the ANC, whereas of only 3 (13%) of 24 formulations with mean RI <10 developed a 32% decrease in the ANC (P = 0.002, by Fisher’s exact test) (Figure 3a). Similarly, for platelet counts, three (50%) of six formulations with mean RI 10 developed a 14% mean decrease, whereas only 1 (4%) of 24 formulations with a mean RI < 10 developed a 14% decrease (P = 0.02, by Fisher’s exact test) (Figure 3b). Because considerable variation existed among the RIs of the 54 volunteers and their decreases in ANCs and platelets counts, the RI was unreliable as a predictor of decreases in the ANC and platelet count for any single volunteer.
|
).
Taqman assay.
Altogether, 24 (44%) of 54 volunteers (formulations 1–15) were viremic by delayed plague assay on day 10 (Table 9); of these 24, only five were positive in the Taqman assay (four sera contained DEN-1, and one sera contained DEN-4) (Table 12). After the second vaccination, four (8%) of 53 vaccinees were viremic by delayed plague assay (day 38); none of these four were Taqman positive. The Taqman assay result was positive in two vaccinees on day 38 (both with DEN-2) whose sera were virus negative by the delayed plaque assay (Table 12). Volunteer 7-4 was the only person viremic by the Taqman assay after both vaccinations. The estimated viral titers by the Taqman assay were generally low in all vaccinees (range = 2–3,000 PFU/mL of serum) (Table 12).
|
). In contrast, only 13 (32%) of 41 volunteers with no or mild reactions were viremic (P = 0.001). Only four (8%) of 53 volunteers re-immunized on day 28 were viremic, and none had dengue symptoms. Association of viremia with antibody response (formulations 1–15). After the first vaccination, 44 volunteers developed antibody, of which 20 (45%) were viremic. Not all viremic volunteers (formulations 1–15) developed DEN antibody. Two volunteers after the first vaccination with formulation 3 and one volunteer after the second vaccination with formulation 4 did not develop antibody by days 28 or 56.
|
DISCUSSION |
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
). However, no formulation induced a tetravalent response in every individual when administered twice at a 28-day interval. In fact, there was no consistent pattern of antibody response after the second vaccination. Antibody was as likely to increase as to decrease. These results underscore the inconsistency and unpredictability of the tetravalent antibody response engendered by two vaccinations given 28 days apart. In contrast, five vaccinees at WRAIR re-vaccinated after a three- or four-month interval (four at 0, 1, 4 months, and one at 0 and 4 months) sustained a robust booster response.5 The poor booster response at one month could be attributed to the phenomenon of heterotypic immunity, first observed by Albert Sabin, who reported that immunity induced by a DEN-1 challenge of volunteers down-regulates subsequent disease induced by DEN-2 challenge.11 The heterotypic immunity lasts approximately three months and then is lost. In our trial, responses to the first vaccination may have interfered with replication of vaccine virus in the second dose one month later, which may explain why relatively few volunteers were viremic after the second inoculation (four individuals) compared with the first (28 individuals). This relative lack of viral replication after re-vaccination is consistent with vaccination affording cross-reactive immunity, which surely contributed to the unsatisfactory antibody booster response. The transience of heterotypic immunity observed by Sabin, together with the observation of Sun and others,5 suggest that re-vaccination after a three or more month interval would provide better immunogenicity than two vaccinations over one month.
Because the interactions of the four DEN serotype vaccines combined in a single formulation are difficult to predict based on responses to the monovalent vaccines, the current study was designed to detect any immunologic or clinical interactions among the four DEN serotypes. As described in the Introduction, the four serotypes were administered for practical reasons combined into high concentration, low concentration, and mixed high and low concentration formulations. In fact, we anticipated that one or more serotype(s) in the dengue vaccine mixture would dominate and suppress the replication and resulting immunogenicity of less dominant serotype(s). This assumption was based on results of a previous trial of the Thai tetravalent vaccine candidate being developed by Aventis Pasteur, which showed a predominant monotypic DEN-3 antibody response, thought to be caused by interference with DEN-1, DEN-2, and DEN-4 by DEN-3 in the vaccine mixture.12 Moreover, the combination of high and low doses of each of the four serotype monovalent vaccine candidates in the current trial was based on the unproven assumption that the lower dose of a dominant DEN serotype(s) may allow replication of the higher dose of the less robust serotype(s), in a manner similar to that found with the trivalent Sabin oral polio vaccine. A balanced "take" against all three serotypes of the Sabin vaccine was achieved only when the dose of the more robust serotype 2 polio virus was reduced compared with the less robust serotypes 1 and 3.6
We believe that the RIs are accurate. The college students were highly motivated, and they were schooled to record all signs and symptoms irrespective of perceived cause. They methodically recorded their temperature twice daily and their symptoms daily in a clinical diary used successfully in other DEN vaccine trials. In every case, the diary was audited by telephone on day 7, before the onset of most dengue symptoms, and by detailed, face-to-face interview with a study physician on days 10 and 15, during the period of most frequent and intense DEN symptoms. If symptoms started or persisted after day 15, the volunteer was instructed to call or visit the clinic on day 21. A final face-to-face audit was conducted on day 28 on all volunteers. For all these reasons, we are confident that the RI is accurate. An important advantage of the outpatient setting was that it permitted us to record how symptoms interfered with the daily routine of sleep, school, work and recreation in a real-world outpatient setting, unbiased by the intense scrutiny associated with hospitalization (Table 10).
The seven superior formulations (Figure 1a and Table 9) varied markedly in their mean RIs. Three superior formulations (13, 14, 15) appeared to provide an acceptable balance of immunogenicity and reactogenicity, although one of four volunteers who received formulation 14 curtailed work for one day. All 16 formulations fulfilled the clinical and laboratory criteria of being safe. Nevertheless, among all formulations, 13 (25%) of 54 Maryland volunteers were forced to curtail time at school, work, recreation, or sleep. Their reactions began usually 7–16 days after vaccination, lasted 1–3 days, and anti-pyretics/analgesics provided some relief. All reactions occurred after the first vaccination. No volunteer reported residual symptoms after 28 and 56 days of follow-up. Three of 10 volunteers immunized with formulation 16 at WRAIR experienced some curtailment of normal daily activity on the inpatient study ward.5 Although a significantly higher incidence of severe dengue occurred among white persons in Cuba compared with black persons during an epidemic,13 we found the RIs to be similar in African-American and Caucasian volunteers administered these attenuated viruses.
For the past 20 years, many monovalent, live DEN vaccine candidates representing the four serotypes have been evaluated in pre-clinical and clinical studies by United States Army and University of Maryland investigators3–5,7,8,14–18 Most candidates were either under-attenuated and made volunteers too ill or over-attenuated and failed to infect and immunize people. Thai DEN vaccine candidates appear to be immunogenic and relatively free of side effects.12,19–22 In the current study of tetravalent vaccines, reactogenicity was again correlated with immunogenicity (Figures 1 and 2). Predictably, viremia was commonly associated with curtailed activities (Table 10). The DEN-1 vaccine has been shown previously to be the most reactogenic of the four monovalent vaccines.5 In this study, no clear association existed between the dose of DEN-1 or DEN-3 and reactogenicity (Table 11). In contrast, a paradoxical inverse relationship existed between high and low dose formulations of DEN-2 and DEN-4, such that low doses were associated with significantly greater reactogenicity than the high dose formulations (Table 11). A possible explanation for this paradox may be the interaction of other serotypes with DEN-1; low doses of DEN-2, DEN-3, and DEN-4 were more reactogenic when combined with high dose DEN-1, suggesting they were unable to dampen the reactogenicity caused by high dose DEN-1. Our conclusion from this experience is that the interaction of DEN serotype, strain, and dose and is complex, making it difficult to predict clinical outcome based on experience with monovalent vaccines.
The DEN-1 strain induced significantly higher GMTs compared with other DEN serotypes (Tables 3 and 6), but the proportion of persons seroconverting to DEN-1 was no greater (Table 6). The DEN-1 strain was unique in that the high dose formulations were more immunogenic than the low dose formulations, whereas the other serotypes showed no dose effect (Table 8). The DEN-4 strain was significantly less immunogenic than the other serotypes (Table 6), but there was no evidence that the other serotypes interfered with DEN-4 in the tetravalent vaccine formulations (Table 7). Increasing the DEN-4 inoculum seemed to provide no immunizing advantage because high and low dose DEN-4 formulations provided similarly poor seroresponse rates (Table 8). The properties of the DEN-4 strain itself, or the high PDK-passage of the strain (Table 1) may have contributed to its weak immunogenicity. Preliminary results do suggest that DEN-4 seroconversions and antibody titers can be boosted considerably by re-immunizing volunteers at four months rather than at one month.5 Moreover, we have tested a lower-passaged DEN-4 strain (PDK-6 instead of PDK-20) as a component of another tetravalent vaccine formulation, and found it was associated with a more robust DEN-4 antibody response (Cunningham D and Sun W, unpublished data).
It is important to review the limits of the DEN virus neutralization (PRNT50) assay. We selected a PRNT titer 1:5 as a measure of immunogenicity and to justify further clinical testing. It remains to be tested if a neutralizing titer 1:5 will protect against natural DEN infection in field trials. In fact, the protective PRNT50 titer in humans is unclear. Moreover, although neutralizing antibody is the best available surrogate for DEN infection and is highly correlated with homologous protection against disease in those recovered from DEN infection, there is no proof that neutralizing antibody is absolutely necessary to afford protection from disease. Indeed, non-neutralizing monoclonal antibodies can protect mice against lethal DEN challenge, possibly by binding complement and by yet unknown mechanisms.23,24 In an unpublished DEN challenge trial, a volunteer vaccinated with the tetravalent vaccine, but who lacked detectable DEN-1 antibody at the time of subsequent DEN-1 challenge, was still protected. Conversely, in the same vaccine challenge trial, a DEN-3 titer of 1:16 induced by the tetravalent vaccine did not protect against illness caused by DEN-3 challenge (Sun W, unpublished data). Finally, the absence of neutralizing antibody after the first vaccination may not predict immunity after re-vaccination, particularly if the individual has been immunogically primed by the first dose. Thus, a highly attenuated but poorly immunogenic DEN vaccine may be safe and protective after two doses, if properly spaced.
Complement was not used in the PRNT assays in this trial. In the monovalent DEN vaccine trials, guinea pig complement tended to enhance cross-reactive neutralizing antibodies induced by monovalent vaccines against heterospecific DEN strains.4,5 Thus, to enhance the specificity of the serologic response against each of the four serotypes in the tetravalent vaccine formulations (and to obtain a more accurate estimate of the immunogenicity of each dengue strain), we did not use complement in the PRNT assays of serum obtained from tetravalent-vaccinated volunteers.
The results of the Taqman assay were disappointing. Only seven (7%) of the 107 post-vaccination serum samples were positive by this assay, suggesting that the level of viremia was too low for ready detection of DEN virus by extraction of viral RNA from 1.6 µL of serum (Table 12). In contrast, 23 sera (21%) were positive by delayed plaque assay but not by the Taqman assay. Two sera (2%) were positive by the Taqman assay but not by the delayed plaque assay. Five (5%) were positive by both the delayed plaque and Taqman assays. Thus, in our hands, the sensitivity of the Taqman assay was low compared with that of the delayed plaque assay, probably due to the low volume of serum tested per reaction.
The occurrence of viremia in this trial is a minimal estimate based on one time point. The selection of day 10 viremia was somewhat arbitrary and was based on our experience with monovalent DEN vaccines, in which viremia occurred between days 7 and 14. We were restricted to one time point so as not to burden the college students with additional clinic visits, and to ease the burden on the virus isolation facilities at WRAIR. The selection of days 10 and 15 for determination of platelet and neutrophil counts was also guided by our experience with monovalent DEN vaccines, where day 15 was the nadir for these counts. The values may not reflect precisely the temporal changes after vaccination with tetravalent vaccines. With this caveat in mind, none of the volunteers developed unsafe levels of thrombocytopenia or neutropenia after tetravalent vaccination.
Many important questions about selected tetravalent vaccine formulations need to be answered. First, we are vaccinating more flavivirus-negative, adult volunteers to confirm that formulations 13, 14, and another, new formulation are safe, well-tolerated, and immunogenic. Second, another study will carefully evaluate the safety and reactogenicity of selected formulations in flavivirus antibody-positive volunteers. Third, DEN-enhancing antibody needs to be sought in vaccinees.25 Finally, since infants often respond to wild DEN infection with few to no symptoms, suitable tetravalent vaccine formulations should be independently evaluated in infants and young children, the optimal target group of endemic dengue through vaccination. Such trials are planned.
Acknowledgments: We thank the participating study volunteers at the University of Maryland (College Park, MD) and at WRAIR (Washington, DC). Skilled research nursing was provided by JoAnna Becker and Helen Secrest at College Park, and clinical support was provided by Dr. Judith Perry. Expert laboratory support was provided by Dave Barvir, Stacie Baily, and Kelly Jones.
Financial support: This study was supported by Contract No. DAMD17-92-C-2058 (Task Order 98-010) to the University of Maryland by the United States Army Medical Research Acquisition Activity (Fort Detrick, Frederick, MD).
Disclosure: This manuscript was finalized after Bruce L. Innis retired from the U.S. Army and took a position with GlaxoSmithKline Biologicals. GSK Biologicals has a non-exclusive license for the WRAIR dengue vaccine candidate.
Authors’ addresses: Robert Edelman and Steven S. Wasserman, Center for Vaccine Development, University of Maryland School of Medicine, 685 West Baltimore St., Baltimore, MD 21201, Telephone: 410-706-5328, Fax: 410-706-6205. Sacared A. Bodison, University Health Center, University of Maryland, College Park, MD 20742, Telephone: 301-314-8117, Fax: 301-314-7845. Robert J. Putnak (Telephone: 301-319-9426), Niranjan Kanesa-thasan (Telephone: 301-319-9827), David W. Vaughn (Telephone: 301-319-9732), and Wellington Sun (Telephone: 301-319-9493), Department of Virus Diseases, Division of Communicable Diseases and Immunology, Walter Reed Army Institute of Research, 503 Robert Grant Avenue, Silver Spring, MD 20910, Fax: 301-319-9661. Kenneth H. Eckels, Department of Biologics Research, Division of Communicable Diseases and Immunology, Walter Reed Army Institute of Research, Building 501, Room 46, Silver Spring, MD 20910, Telephone: 301-319-9233, Fax: 301-319-9360. Douglas Tang, Department of Biometrics, Walter Reed Army Institute of Research, 503 Robert Grant Avenue, Silver Spring, MD 20910, Telephone: 301-319-9909. Bruce L. Innis, Glaxo-SmithKline, 1250 S. Collegeville Road, Mailcode UP4330, Collegeville, PA 19426-0989.
Reprint requests: Robert Edelman, Center for Vaccine Development, 685 West Baltimore Street, Room 480, Baltimore, MD 21201, E-mail: redelman{at}medicine.umaryland.edu.
|
REFERENCES |
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  S. J. Thomas, A. Nisalak, K. B. Anderson, D. H. Libraty, S. Kalayanarooj, D. W. Vaughn, R. Putnak, R. V. Gibbons, R. Jarman, and T. P. Endy Dengue Plaque Reduction Neutralization Test (PRNT) in Primary and Secondary Dengue Virus Infections: How Alterations in Assay Conditions Impact Performance Am J Trop Med Hyg, November 1, 2009; 81(5): 825 - 833. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. F. Wright, A. P. Durbin, S. S. Whitehead, M. R. Ikizler, S. Henderson, J. E. Blaney, B. Thumar, S. Ankrah, M. T. Rock, B. A. McKinney, et al. Phase 1 Trial of the Dengue Virus Type 4 Vaccine Candidate rDEN4{Delta}30-4995 in Healthy Adult Volunteers Am J Trop Med Hyg, November 1, 2009; 81(5): 834 - 841. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
B. Guy, V. Barban, N. Mantel, M. Aguirre, S. Gulia, J. Pontvianne, T.-M. Jourdier, L. Ramirez, V. Gregoire, C. Charnay, et al. Evaluation of Interferences between Dengue Vaccine Serotypes in a Monkey Model Am J Trop Med Hyg, February 1, 2009; 80(2): 302 - 311. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
B. Etemad, G. Batra, R. Raut, S. Dahiya, S. Khanam, S. Swaminathan, and N. Khanna An Envelope Domain III-based Chimeric Antigen Produced in Pichia pastoris Elicits Neutralizing Antibodies Against All Four Dengue Virus Serotypes Am J Trop Med Hyg, September 1, 2008; 79(3): 353 - 363. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. Raviprakash, D. Wang, D. Ewing, D. H. Holman, K. Block, J. Woraratanadharm, L. Chen, C. Hayes, J. Y. Dong, and K. Porter A Tetravalent Dengue Vaccine Based on a Complex Adenovirus Vector Provides Significant Protection in Rhesus Monkeys against All Four Serotypes of Dengue Virus J. Virol., July 15, 2008; 82(14): 6927 - 6934. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Simasathien, S. J. Thomas, V. Watanaveeradej, A. Nisalak, C. Barberousse, B. L. Innis, W. Sun, J. R. Putnak, K. H. Eckels, Y. Hutagalung, et al. Safety and Immunogenicity of a Tetravalent Live-attenuated Dengue Vaccine in Flavivirus Naive Children Am J Trop Med Hyg, March 1, 2008; 78(3): 426 - 433. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L. Chen, D. Ewing, H. Subramanian, K. Block, J. Rayner, K. D. Alterson, M. Sedegah, C. Hayes, K. Porter, and K. Raviprakash A Heterologous DNA Prime-Venezuelan Equine Encephalitis Virus Replicon Particle Boost Dengue Vaccine Regimen Affords Complete Protection from Virus Challenge in Cynomolgus Macaques J. Virol., November 1, 2007; 81(21): 11634 - 11639. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. Klungthong, R. V. Gibbons, B. Thaisomboonsuk, A. Nisalak, S. Kalayanarooj, V. Thirawuth, N. Nutkumhang, M. P. Mammen Jr., and R. G. Jarman Dengue Virus Detection Using Whole Blood for Reverse Transcriptase PCR and Virus Isolation J. Clin. Microbiol., August 1, 2007; 45(8): 2480 - 2485. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
N. U. RAJA, D. H. HOLMAN, D. WANG, K. RAVIPRAKASH, L. Y. JUOMPAN, S. B. DEITZ, M. LUO, J. ZHANG, K. R. PORTER, and J. Y. DONG INDUCTION OF BIVALENT IMMUNE RESPONSES BY EXPRESSION OF DENGUE VIRUS TYPE 1 AND TYPE 2 ANTIGENS FROM A SINGLE COMPLEX ADENOVIRAL VECTOR Am J Trop Med Hyg, April 1, 2007; 76(4): 743 - 751. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. H. Holman, D. Wang, K. Raviprakash, N. U. Raja, M. Luo, J. Zhang, K. R. Porter, and J. Y. Dong Two Complex, Adenovirus-Based Vaccines That Together Induce Immune Responses to All Four Dengue Virus Serotypes Clin. Vaccine Immunol., February 1, 2007; 14(2): 182 - 189. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. E. Blaney Jr., J. M. Matro, B. R. Murphy, and S. S. Whitehead Recombinant, Live-Attenuated Tetravalent Dengue Virus Vaccine Formulations Induce a Balanced, Broad, and Protective Neutralizing Antibody Response against Each of the Four Serotypes in Rhesus Monkeys J. Virol., May 1, 2005; 79(9): 5516 - 5528. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. P. Simmons, T. Dong, N. V. Chau, N. T. P. Dung, T. N. B. Chau, L. T. T. Thao, N. T. Dung, T. T. Hien, S. Rowland-Jones, and J. Farrar Early T-Cell Responses to Dengue Virus Epitopes in Vietnamese Adults with Secondary Dengue Virus Infections J. Virol., May 1, 2005; 79(9): 5665 - 5675. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. P. Goncalvez, R. Men, C. Wernly, R. H. Purcell, and C.-J. Lai Chimpanzee Fab Fragments and a Derived Humanized Immunoglobulin G1 Antibody That Efficiently Cross-Neutralize Dengue Type 1 and Type 2 Viruses J. Virol., December 1, 2004; 78(23): 12910 - 12918. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. E. BLANEY JR., C. T. HANSON, C.-Y. FIRESTONE, K. A. HANLEY, B. R. MURPHY, and S. S. WHITEHEAD GENETICALLY MODIFIED, LIVE ATTENUATED DENGUE VIRUS TYPE 3 VACCINE CANDIDATES Am J Trop Med Hyg, December 1, 2004; 71(6): 811 - 821. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
F. Guirakhoo, Z. Zhang, G. Myers, B. W. Johnson, K. Pugachev, R. Nichols, N. Brown, I. Levenbook, K. Draper, S. Cyrek, et al. A Single Amino Acid Substitution in the Envelope Protein of Chimeric Yellow Fever-Dengue 1 Vaccine Virus Reduces Neurovirulence for Suckling Mice and Viremia/Viscerotropism for Monkeys J. Virol., September 15, 2004; 78(18): 9998 - 10008. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
ERRATA Am J Trop Med Hyg, March 1, 2004; 70(3): 336 - 337. [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
and range of peak titers elicited by each formulation against each of the four DEN serotypes 28 days after vaccinations 1 and 2
).