Original Article
Limited Efficacy of a Haemophilus influenzae Type b Conjugate Vaccine in Alaska Native Infants
N Engl J Med 1990; 323:1393-1401November 15, 1990
- Abstract
- Abstract
Background.
The prevention of invasive Haemophilus influenzae type b disease requires a vaccine that is effective when administered during the first six months of life. The infants of Alaska Natives are at particularly high risk of invasive H. influenzae type b disease.
Methods.
To evaluate the protective efficacy of a H. influenzae type b polysaccharide—diphtheria toxoid conjugate vaccine (polyribosylribitol phosphate—diphtheria toxoid [PRP-D]), we enrolled 2102 Alaska Native infants in a randomized, double-blind, placebo-controlled trial in which either the vaccine or a saline placebo was administered at approximately two, four, and six months of age.
Results.
After 3969 subject-years of follow-up and 32 episodes of H. influenzae type b disease, the overall incidence of invasive disease was not reduced significantly in the vaccinated subjects (6.0 cases per 1000 patient-years), as compared with the placebo controls (9.6) or with other Alaska Native infants (6.0). After one, two, or three doses there was no significant protective efficacy with the vaccine; after three doses the efficacy was only 35 percent (95 percent confidence interval, -57 to 73). The lack of efficacy was not related to the age at onset of disease, age at immunization, type of disease, degree of Alaska Native heritage, time after immunization, or year of the study. Levels of H. influenzae type b anticapsular antibody in recipients of the vaccine became significantly higher than levels in those who received placebo only after the second and third doses. Even after the third dose, only 48 percent of the vaccinated infants had antibody levels of more than 0.1 μg per milliliter (geometric mean titer, 0.18). Antibody responses did not vary with the level of maternally acquired antibody, degree of Alaska Native ancestry, or age at time of the first or second immunizations, but they increased with increasing age at time of the third dose (P<0.001).
Conclusions.
We found no evidence that the PRP-D vaccine provides significant protection, at least for Alaska Native infants, against invasive diseases caused by H. influenzae type b. The ineffectiveness of the vaccine paralleled its limited immunogenicity. (N Engl J Med 1990; 323:1393–401.)
Media in This Article
Table 1
H. influenzae Type b Disease in the Study Subjects.*Table 2
Efficacy of PRP-D Vaccine in Alaska Native Infants Immunized at Two, Four, and Six Months of Age.Article Activity
- Article
INVASIVE Haemophilus influenzae type b disease —meningitis, bacteremia, septic arthritis, pneumonia, cellulitis, and epiglottitis — remains a leading worldwide cause of morbidity and mortality among young children. In the United States there are at least 20,000 cases of invasive H. influenzae type b disease annually, resulting in 1000 deaths.1 In the United States and many other countries, H. influenzae type b is the leading cause of bacterial meningitis. Several groups have been identified with an increased incidence of H. influenzae type b disease; among them, Alaska Natives (Indians, Eskimos, and Aleuts) have the highest documented risk.1 2 3 4 The reasons for this risk have been speculated to include differences in H. influenzae type b strains, the type or degree of exposure, and differences in immune response or susceptibility.5 Since antibiotic therapy cannot be expected to reduce the morbidity or mortality of H. influenzae type b disease and because the majority of disease in all populations occurs during the first year of life, the control of H. influenzae type b disease requires effective immunization of infants during their first six months.
Over the past 60 years, considerable evidence has accumulated to show that antibody to the capsule of H. influenzae type b (polyribosylribitol phosphate, or PRP) has biologic activities against the organism1 and is effective in both the treatment6 , 7 and prevention8 , 9 of disease. Unfortunately, the purified type b capsular polysaccharide vaccine, licensed for older children in 1985, is only weakly immunogenic8 , 10 and thus induces only limited protection.11 12 13 14 15 16 Furthermore, the levels of PRP antibodies needed for protection have not been adequately defined, with estimates ranging from 0.125 to 1.0 μg per milliliter.17 In attempts to enhance the immunogenicity of this T-cell—independent immunogen, several PRP conjugate vaccines have been developed that covalently link PRP to an immunogenic protein carrier. Although there are major biochemical, structural, and immunologic differences among PRP conjugate vaccines, all induce improved immunologic responses, particularly in infants.1 Developed from a modification of the scheme originally proposed by Schneerson et al., PRP—diphtheria toxoid (PRP-D) was the first PRP conjugate vaccine and the first licensed (1987) and recommended for routine use in children 15 to 18 months of age or older (ProHibit, Connaught Laboratories, Swiftwater, Pa.).18 19 20 21 PRP-D vaccine was selected for this trial because it was the first conjugate vaccine available for clinical testing in infants and because initial data suggested that it was safe and immunogenic.22 , 23
Despite the results of a Finnish trial suggesting that PRP-D vaccine prevents H. influenzae type b disease in young infants,24 concern remains about its immunogenicity and protective efficacy.22 23 24 25 We report the results of a prospective, randomized, placebo-controlled trial designed to evaluate the immunogenicity and protective efficacy of PRP-D vaccine in Alaska Native infants immunized at approximately two, four, and six months of age.
Methods
Study Population
We have previously described Alaska Native infants and their increased incidence of H. influenzae type b disease.2 , 3 , 5 , 26 For this study, we recruited only infants of Eskimo, Aleut, or Indian heritage who resided in six of eight of the most populous Alaska Native Health Board service units. Seventy-nine percent of all Alaska Native births and 88 percent of all H. influenzae type b disease in Alaska Natives occur in these six regions.3
Subjects were recruited for the trial between October 17, 1984, and January 8, 1988, and the occurrence of H. influenzae type b disease was monitored in blinded fashion through June 30, 1988. Alaska Native infants were recruited if they were four months of age or younger and had not yet received diphtheria–pertussis–tetanus (DPT) vaccine; they were not enrolled in the trial if they lived in a remote area or had a suspected immunologic disorder, seizure disorder, contraindication to receiving DPT vaccine, or major medical illness. All Alaskan infants, especially Alaska Native infants, were carefully monitored for H. influenzae type b disease, and their incidence of disease compared with that in the study subjects.
Study Design
The study was a prospective, randomized, double-blind trial comparing PRP-D vaccine and a saline placebo. Infants were recruited and assigned previously randomized vials in blocks of 10 (5 vaccine and 5 placebo). Immunizations were scheduled at two, four, and six months of age, but except for the first dose (which was given at under five months), there were no age restrictions and no restrictions on the coincidental administration of other vaccines. With a true efficacy of more than 75 percent, an estimated sample of 2000 subjects was needed for approximately 33 cases of H. influenzae type b disease to occur, providing 80 percent power to detect significant protection. There were no violations of the blinding or randomization process during the trial. The study protocol was reviewed by all participating agencies, including the National Institutes of Health, Food and Drug Administration, Centers for Disease Control (CDC), UCLA, the Indian Health Service, the Alaska Area Native Health Board, and the Native Health Boards in each participating region. Informed signed consent for participation was required from the parents in English or the appropriate native language.
Vaccine
The first production lot of PRP-D vaccine (lot 3814) was a clear aqueous solution prepared in sets of three single-subject vials, and a 0.5-ml dose (20 μg of ribose per dose) was administered intramuscularly. PRP-D vaccine or saline placebo was administered in the right thigh, and all other vaccines in the left thigh. The study preparations were indistinguishable, and at the end of the trial a 25 percent random sample of all used vials was analyzed for PRP content to test the accuracy of actual randomization. All the vials had been appropriately assigned and administered.
Case Definitions
The primary outcome considered for protection was invasive H. influenzae type b disease — disease with documented bacteremia or invasion of tissue. Minor infections, such as otitis media, sinusitis, bronchitis, or pharyngitis, were not considered. All diagnoses of H. influenzae type b disease were made by pediatricians and investigators who had no knowledge of the infants' vaccine status, and they were independently reviewed by an oversight committee and three independent experts. Standardized case definitions that had clinical and laboratory criteria were used. For the primary analyses of efficacy, only disease proved by culture of blood, cerebrospinal fluid, joint fluid, or body fluid or positive results on H. influenzae type b latex-agglutination testing of cerebrospinal fluid or serum was included. With the use of standard definitions, all other illnesses in the study subjects were categorized and used for secondary analyses of efficacy and safety.
Surveillance
Surveillance for invasive H. influenzae type b disease was maintained for the entire population of Alaska. The methods of finding cases included regular reports from all microbiology laboratories, infection-control nurses, pediatricians, and public health nurses; information from the state health department on reported episodes of bacterial meningitis (a reportable disease in Alaska); review of state death records; and periodic reviews of charts and laboratory records at selected hospitals. In all the study areas, every physician, nurse practitioner, and licensed laboratory technician was contacted at least once a month. Finally, all the parents of study participants were interviewed about intercurrent illnesses in their children at the ages of 2, 4, 6, and 18 to 24 months, and at the end of the study all hospital, clinic, and village health records were reviewed for every study subject. Throughout the trial, physicians were encouraged to obtain blood cultures and perform H. influenzae type b antigen testing in febrile children. Potential adverse events were systematically assessed for 48 hours after each immunization, and more serious potential adverse effects were detected by surveillance.
Laboratory Studies
Cultures of blood, cerebrospinal fluid, joint fluid, and other body fluids were processed by standard bacteriologic techniques.27 H. influenzae organisms were identified by the morphologic features of the colonies and by growth requirements for X and V factors. They were serotyped by slide agglutination with specific CDC antiserum and by DNA probe. Blood, urine, and cerebrospinal fluid were tested for the presence of H. influenzae type b antigen by latex-agglutination tests (Wampole Laboratories, Cranbury, N.J.), and most assays were confirmed at a CDC reference laboratory in Anchorage or at UCLA.28
Blood samples were obtained before injection from a subgroup of subjects at the first immunization at 2 months of age; at the second at 4 months; at the third at 6 months; and at 8 to 11 months. We measured antibody to the capsular polysaccharide with tritiated PRP in a modified Farr assay, using an FDA protocol and standards.29 , 30 The lower limit of reproducible sensitivity was 0.025 μg per milliliter, and to calculate a geometric mean specimens with values below this limit were assigned a value of 0.0125 μg per milliliter.
Statistical Analysis
The pairwise comparisons of rates between study groups were performed with the two-tailed Fisher's exact test, chi-square test, or MantelHaenszel test for stratified comparisons.31 Corrections for multiple analyses were made with the Bonferroni method.32 Efficacy was defined as the percent reduction in the risk of disease among vaccinated subjects, and the confidence limits were calculated by Taylor series approximations with log-transformed rates.33 Antibody levels were log transformed and compared by analysis of variance, with adjustment for age. Comparisons of continuous factors between groups were performed by t-test, or Wilcoxon nonparametric test when the data were not normally distributed.31 The association of antibody levels with several continuous time variables was assessed with Kendall's rank correlations.31
Results
Study Population
Over the 45-month study period 5982 Alaska Natives were born in the six study areas. From this cohort, 2102 subjects were recruited and immunized. Statewide surveillance for H. influenzae type b disease was maintained throughout the trial for the three subgroups of this cohort: the vaccine group (n = 1054), the placebo group (n = 1048), and Alaska Natives who were not study subjects (n = 3880). The follow-up periods were virtually identical for all groups.
There were no significant differences between the randomized groups in sex distribution, ethnicity, region or type of residence, size of household, birth weight, breast-feeding history, age at which vaccine or placebo was received, and number of doses. The same factors were then compared in the study population and the nonstudy population, revealing only one significant difference.* There were significantly more village residents and fewer city residents in the nonstudy group (P<0.0001). The overall incidence of H. influenzae type b disease, however, was not significantly different between those who were not study subjects (6.0 cases per 1000 patient-years) and those who received placebo (9.6 cases per 1000 patient-years, P = 0.08). The lower incidence of H. influenzae type b disease in those who were not study subjects may be due to less surveillance in this group.
Age at Immunization
The age at immunization did not differ between the vaccine and placebo groups. For the first dose, the ages ranged from 4 to 17 weeks, but 85 percent of the subjects were between 5 and 12 weeks of age (median, 6.8). The young age at the first immunization was due to a postpartum visit when the subjects were six weeks old, at which immunizations were often initiated. For the second dose, the subjects ranged in age from 3 to 11 months, but 63 percent were between 3 and 5 months old (median, 4.2). For the third dose, the subjects ranged in age from 4 1/2 to 15 months, but 72 percent were between 5 and 8 months old (median, 6.7).
Evaluation of Safety
No major or minor reactions occurred disproportionately in any of the three study groups, except that there were more episodes of aseptic meningitis and cases of sudden infant death syndrome in the vaccine group than among the controls, although the difference was not statistically significant.
Episodes of H. influenzae Type b Disease
During the study period, there were 32 episodes of definite or probable H. influenzae type b disease in 31 subjects in the study population (Table 1Table 1
H. influenzae Type b Disease in the Study Subjects.*) and 42 episodes in the nonstudy population (all serotype b). One patient had recurrent H. influenzae type b disease, with one episode 48 days after the first dose of vaccine and a second 47 days after the third dose. Four cases were not included in our analyses of efficacy. Two were suspected episodes of H. influenzae type b disease, one in the vaccine group and the other in the placebo group, that did not meet the case definitions established before the study began. Furthermore, in the 19-month period after the vaccine codes were broken (unblinded), there were two more cases in study subjects. There was an episode of H. influenzae type b meningitis in a recipient of vaccine at 12 months of age, 6 months after the third immunization; and there was an episode of H. influenzae type b cellulitis in a control subject at 9 months of age, 5 months after the second dose of placebo. The inclusion of these four cases, however, would not have changed any of the conclusions in the following analyses.The patients ranged in age from 1.6 to 27 months. Seven cases of disease (3 in the vaccine group and 4 in the placebo group) occurred before the second immunization, 5 more (2 vaccine and 3 placebo) before the third immunization, and 20 cases (8 vaccine, including the recurrent case, and 12 placebo) after all three immunizations (Table 1). The type of invasive H. influenzae type b disease was similar in the vaccine and control groups, in the study and nonstudy groups, and as compared with earlier data on Alaska Natives.3 For example, 50 percent of the vaccine recipients with disease had meningitis, as did 37 percent of those given placebo.
Five episodes of invasive disease occurred within 30 days of immunization. Three occurred after immunization with vaccine: one 3 days after the first dose, the second 4 days after the second dose, and the third 21 days after the third dose. Two occurred after immunization with saline placebo: one 12 and the other 19 days after the third dose. Given the previous incidence of disease between two and six months of age3 and assuming no vaccine effect, one can estimate the number of cases expected to occur by chance alone after any of the immunizations. One would thus expect 0.16, 1.11, and 4.9 cases within 1, 7, and 31 days, respectively, of any immunization. The actual number of cases did not differ significantly from these expected values (each P>0.40, by chi-square test).
Vaccine Efficacy
During the study period the overall age-adjusted incidence of H. influenzae type b disease was not significantly different between the vaccine and placebo groups (6.0 vs. 9.6 cases, respectively, per 1000 patient-years of risk), or between the vaccinated subjects and those not in the study. The estimates of vaccine efficacy for proved and probable H. influenzae type b disease after each dose of vaccine are shown in Table 2Table 2
Efficacy of PRP-D Vaccine in Alaska Native Infants Immunized at Two, Four, and Six Months of Age.. Overall, 13 episodes of invasive H. influenzae type b disease occurred in 12 infants given PRP-D vaccine, and 19 episodes occurred in control subjects (41 percent of the cases of H. influenzae type b disease involved infants who had received PRP-D vaccine). The result in the one child with recurrent disease can be considered in the analysis as either a failure after the first or a failure after the third dose, or can be excluded from analysis as a failure after the third dose. With any of the three alternatives, there is little effect on the results of efficacy analysis. Although there were cases of disease after only one and two doses of vaccine, there was no apparent trend toward protective efficacy. After the third dose of vaccine, the point estimate of vaccine efficacy was either 35 or 43 percent, depending on how the recurrent case was considered. Regardless of the case definition or the number of doses of vaccine, no point estimate of vaccine efficacy achieved statistical significance. For example, vaccine efficacy for cases of culture-proved meningitis after three doses was −29 percent (95 percent confidence interval, −473 to 71). Additional analyses of vaccine efficacy were performed with stratification of the results according to the following criteria: the number of doses and the age at the time of each immunization, dose and age at the onset of disease, dose and the interval between the last dose of vaccine and the onset of disease, year of the study (to assess efficacy over time), type of disease, and culture-proved or antigen-confirmed disease. None of these analyses indicated any trend toward improved vaccine efficacy. Specifically, the age at the first, second, and third immunizations did not affect efficacy.The incidence of 17 categories of clinically diagnosed infectious disease was also evaluated, and no incidence was reduced among vaccinated subjects. The diseases included bacterial meningitis, probable meningitis, aseptic meningitis, bacterial sepsis (proved, probable, or suspected), pneumonia (radiologically proved or probable, or other lower respiratory tract infection), bone or joint infections (definite, probable, or suspected), cellulitis (definite, probable, or suspected), and other infectious diseases or hospitalizations for infectious disease (proved or suspected).
Vaccine Efficacy According to Ethnic and Racial Background
To assess the potential role of genetically determined factors, we analyzed the relative risk of disease in strata of study subjects based on their degree of Alaska Native heterogeneity (Table 3Table 3
Influence of Racial and Ethnic Ancestry on Risk of Disease and Level of H. influenzae Type b Anticapsular Antibody.). Genetic pedigrees were obtained for nearly all study subjects by recording the number of great-grandparents who were Alaska Natives (Eskimos, Indians, or Aleuts). We have previously shown a nearly exact relation between this measure of ethnic and racial heterogeneity and results of genetic-marker studies.34 , 35 Intermarriage is almost always between Alaska Natives and whites, and is much less common between Eskimo, Indian, and other racial or ethnic groups. Importantly, at least in the control group, the risk of H. influenzae type b disease was significantly greater for those with predominantly Alaska Native ancestry (>50 percent) than for those with predominantly non-Native ancestry (≤50 percent, P = 0.001). There was no trend, however, toward improved vaccine efficacy in subjects with predominantly non-Native pedigrees. In fact, in this study the only significant estimate of vaccine efficacy was among subjects who were >75 percent Alaska Native (P<0.05). Antibody levels induced in the vaccinated subjects were also not different between ethnic and racial strata. We therefore find little if any evidence that racial or ethnic factors influenced protective efficacy or immunogenicity.Analysis of Immunogenicity
A total of 1199 serum samples (595 from the vaccine group and 604 from the placebo group) were obtained from 694 subjects (341 vaccine and 353 placebo) for H. influenzae type b anticapsular antibody testing. Table 4Table 4
H. influenzae Type b Anticapsular Antibody Levels in Alaska Native Infants, According to Age and Vaccine Group.* shows the geometric mean antibody levels when a level of 0.0125 μg per milliliter was assigned to samples with undetectable antibody and, to compare our findings with those from Finland, a level of 0.06 μg per milliliter was assigned to samples with values at or below this level.22 Antibody levels before immunization (l 1/2 to 4 months of age), after the first immunization (3 to 6 months), after the second immunization (5 to 8 months), and after the third immunization (8 to 11 months) are summarized. Only results in children of the specified ages were included in Table 4, to minimize age-related variability. Antibody levels before immunization did not correlate strongly with age (r = −0.13, P = 0.35), and they were not significantly different between the vaccine and control groups in either the sample taken before immunization or that taken after the first immunization. Significant differences were seen, however, after the second (P<0.01) and third (P<0.001) doses. When the subjects in the vaccine group were 8 to 11 months of age and had received three doses of vaccine, their antibody levels were no higher than they had been before immunization. Table 4 also shows the proportions of subjects with specified levels of antibody, and the distribution of antibody levels in samples taken before immunization is similar to the distribution in the samples taken after the third immunization. After one and two doses of PRP-D vaccine most antibody levels were lower than 0.1 μg per milliliter. After three doses of vaccine only 48 percent of the children had antibody levels higher than 0.1 μg per milliliter, and only 15 percent had levels ≥1.0 μg per milliliter. There was no significant correlation between levels of preexisting antibody (of maternal origin) and subsequent immune response (r = 0.17, P = 0.1).There was no correlation between age at the first or second immunization and subsequent levels of antibody by either categorical or continuous analyses. There was, however, a significant increase in antibody levels with increasing age at the third immunization (P = 0.001). The correlation coefficient was only 0.19, however, and the degree of increase in immune response was therefore slight over the observed age range at the third immunization, given between 21 and 54 weeks of age. The intervals between the first and second doses, the first and third, and the second and third, and the period between the third immunization and the post-immunization assessment were also evaluated, but no significant relations were found.
We also collected blood samples from 11 subjects vaccinated with PRP-D and 18 control subjects in whom H. influenzae type b disease developed. Among the 9 subjects whose samples were drawn before the onset of disease, the mean antibody titer was 0.03 μg per milliliter (range, 0.013 to 0.145), and among the 20 whose samples were drawn within five days of the onset of illness, the mean titer was 0.04 μg per milliliter (range, 0.013 to 0.225). There were no differences in antibody levels between the vaccinated subjects and those given placebo, and there were no differences in the number of doses of vaccine administered before the onset of disease. The titers may not reflect actual levels of antibody just before the time of H. influenzae type b exposure, because a subsequent rise or fall in titer could have occurred between the time of blood collection and the onset of disease, or because in children with H. influenzae type b disease the complexing of free antibody by circulating antigens may have occurred. Nevertheless, none of the titers were greater than 0.225 μg per milliliter, suggesting that these levels of antibody were not protective.
Discussion
We conclude from this randomized controlled study among Alaska Native infants that three doses of PRP-D conjugate vaccine given in early infancy induce insufficient immunity to provide significant protection against invasive H. influenzae type b disease. The maximal estimate of protective efficacy was only 43 percent, and the 95 percent confidence interval was −43 to 78 (P = 0.16). All other estimates were lower, and the incidence of H. influenzae type b disease was not lower in the vaccine group than in the placebo group, the nonstudy population, or historical controls. The study had sufficient statistical power to detect a true vaccine efficacy of at least 75 percent; the size of the sample needed to ascertain a significant lower efficacy would have been excessive.11 We considered it unlikely that a vaccine with a lower, albeit significant, protective efficacy would be generally accepted, especially given the availability of more promising H. influenzae type b vaccines1 , 36 37 38 39 40 and hyperimmune globulin.9
The poor efficacy paralleled the limited immune responses induced by PRP-D vaccine. After the first two doses of vaccine, there was little if any detectable antibody response. After the third dose, only about half the subjects had antibody levels higher than 0.1 μg per milliliter, and neither the geometric mean titer nor the distribution of antibody levels was significantly different from the values before immunization. We attempted to identify factors that might have contributed to poor immunogenicity. First, in this population the high levels of maternally acquired antibody before immunization did not influence subsequent immune responses. This is in contrast to earlier observations.40 Second, at least within the range of ages of the infants in this trial, neither the age at the first two immunizations nor the intervals between any of the doses had a detectable effect on the immune response. In this trial, the age at the first immunization was young (median, 1.7 months), but this affected neither the ultimate antibody response nor the protective efficacy. The age at the third immunization, however, did affect the levels of antibody achieved, but this effect was slight, at least within the age range evaluated. With immunization at 15 months of age, as has been recommended in the United States, PRP-D vaccine is much more immunogenic41 and protective.42 , 43
We considered two possible explanations for the limited immunogenicity and protective efficacy of this vaccine in young Alaska Native infants. First, the immunogenicity of the vaccine lot used in Alaska (lot 3814, the first production lot) may have been defective and not representative of other licensed PRP-D lots. Second, the immune responses in Alaska Natives may have been uniquely deficient as compared with those in other populations, resulting in significantly reduced vaccine efficacy. To address the first issue, another commercial lot of PRP-D vaccine was subsequently evaluated under identical conditions in Alaska Native infants, and essentially identical immune responses were observed.44 Interestingly, in addition to an assessment of antibody after the third dose at 8 to 11 months of age, a sample was also obtained in this later study 1 month after the third immunization (at 7 months of age), and its geometric mean titer was 0.45 μg per milliliter, which declined to 0.15 μg per milliliter over the next 1 to 4 months. Also, the Alaska lot (lot 3814) was given to Finnish infants, and after the slightly different schedules and statistical methods had been corrected for, there were no apparent differences in immunogenicity between Finnish and Alaska Native infants (unpublished data). Furthermore, lot 3814 was given under identical conditions to infants in Albany, New York, and no significant differences in immunogenicity were seen, as compared with Alaska Native infants.45 Finally, there were no differences in the immune responses of Alaska Native infants whose ancestry was less than one-fourth Alaska Native (the infants were predominantly white), as compared with the responses of those whose ancestry was completely Alaska Native. We conclude that antibody responses to the PRP-D lot used in this trial were similar to the responses to other PRP-D lots. Interestingly, earlier research lots of PRP-D vaccine may have been more immunogenic.22 23 24 25
The results of our study appear to differ substantially from those of a similar trial of PRP-D vaccine in Finnish infants, which showed a 94 percent protective efficacy (95 percent confidence interval, 83 to 98).24 , 46 Given the disparity in results, it is unlikely that differences in the design and conduct of the Finnish and Alaskan trials could account for the difference. A major issue in the Finnish trial was the apparent discrepancy between the immunogenicity and protective efficacy of the vaccine. A large proportion of Finnish children failed to respond with high levels of antibody. Even after three doses, about 30 percent of the subjects had titers lower than 0.15 μg per milliliter, and only 40 percent had titers higher than 1.0 μg per milliliter, but the observed efficacy was high. These titers are higher than those seen in our trial. However, antibody levels were assessed one month after immunization in Finland, whereas in Alaska they were measured two to five months after immunization. We conclude that the immunogenicity of PRP-D vaccine did not differ between Alaska Native and Finnish or other infants after correction for assay methods, statistical methods, and most important, the ages at testing44 , 45 (and unpublished data). Although the ages at administration of each of the doses differed slightly in the two studies, this is not likely to explain the difference in results. In our study, the ages at immunization did not have a striking effect on ultimate immune response or protective efficacy.
Other factors that could account for the difference in observed efficacy include differences between the populations in the degree or type of H. influenzae type b exposure or differences in susceptibility to disease. One can postulate that a higher level of immunity would be required to protect a child who is exposed more frequently or more intensely, particularly at a younger age. One can also postulate that there is less exposure in Finland, where there is a national home care program for infants during the first year of life, a shift in the incidence of H. influenzae type b disease to older ages, and a low overall incidence of disease (approximately one-third that in the United States). In Alaska there is some evidence that intense and early exposure to H. influenzae type b contributes to the increased risk of H. influenzae type b disease.5 The risk of disease in Alaska Natives is 10 times greater than in other U.S. populations and is disproportionately concentrated early in the first year of life.2 , 3
Differences in vaccine efficacy might also result from microbiologic differences in H. influenzae type b strains in the two populations. Although the invasive H. influenzae type b strains in both Finland and Alaska are predominantly serotype b, there are other microbiologic differences in outer-membrane proteins and enzyme electromorphs.47 48 49 Conceivably, strain-specific differences in virulence or infectivity could affect determinants of the protective efficacy of vaccine. It is reasonable, however, to expect that an effective vaccine should induce sufficient protective immunity for any natural exposure to H. influenzae type b.
We think human genetic factors influencing immune response or susceptibility to disease are less likely to have determined the low vaccine efficacy in Alaska. First, we have studied genetic markers in Alaska Native infants with H. influenzae type b disease and in controls and have failed to identify genetic differences that influence naturally acquired levels of H. influenzae type b antibody or explain most of the excess burden of disease.34 , 35 Second, there have been several comparative studies to detect differences in vaccine immunogenicity between Alaska Natives and other children using different H. influenzae type b vaccines, including comparisons of PRP-D vaccine in Alaska Native, Finnish, and New York infants.44 , 45 Such studies have failed to detect major differences in immune response between the populations. Even if major differences had been observed, as has been reported in another Native American population,50 one would still have to postulate that the differences in immune response were sufficiently great or prevalent to account for the differences in the incidence of disease and in vaccine efficacy. Finally, and perhaps most important, we failed to identify differences in immune response or protective efficacy among Alaska Natives on the basis of an analysis of the degree of Alaska Native heritage.
In conclusion, the findings of this study indicate that three doses of PRP-D vaccine induce marginal immune responses in young infants, and in infants at high risk three doses fail to provide significant protection. The most likely reason for these results is an intrinsic limitation in the immunogenicity of the PRP-D vaccine, at least in young infants. Other H. influenzae type b conjugates appear to induce higher levels of antibody.36 37 38 39 40 , 44 Presumably, the composition, structure, or formulation of this H. influenzae type b conjugate vaccine is not optimal for maximizing T helper—cell responsiveness. Whereas low levels of immunity may have been sufficient to protect infants, such as those in Finland, who are infrequently or less intensely exposed and who tend to have disease at older ages, we believe that this level of immunity is insufficient for Alaska Native infants. We think it imprudent to speculate about whether the results from Finland or Alaska are more applicable to the general U.S. population, since there are undoubtedly segments of the U.S. population at increased risk (like Alaska Natives) and segments at decreased risk (like Finnish infants). One must expect that vaccines will work in all groups, especially when they are being considered for routine use in a heterogeneous population like that of the United States.
Supported by a contract (NO1-AI-32512) from the National Institute for Allergy and Infectious Diseases, National Institutes of Health.
The use of trade names is for purposes of identification only and does not imply endorsement by the U.S. Public Health Service.
*Members of the study group are as follows: Kelly Burkart, Chung-Yin Chiu, Peter Christenson, Ph.D., Vickie-Marie Colaciccio, R.N., Mary Anne Fitzgerald. Wilma Goodwin, R.N., Marie Hanley, R.N., Annette Harpster, R.N., Robin James, R.N., Anne Lanier, M.D., Martina Lauterback, R.N., Penny Lehman, R.N., Jennifer McCaffrey, Katherine Mertens, R.N., Sherry Neth, R.N., Nancy Norgeot, R.N., Helen Peters, R.N., Ken Peterson, M.D., Barbara Riley, M.D., Ellen Valleroy, R.N., Robert Wainwright, M.D., and Jill Wettlaufer. R.N.
We are indebted to the Research and Education Institute of Harbor–UCLA Medical Center, the Centers for Disease Control, the Indian Health Service, the National Institutes of Health, the State of Alaska, and especially the participating Alaska Native Health Boards representing the Alaska Natives of Bristol Bay (Kanakanak), Norton Sound, Yukon—Kuskokwim Delta, Interior (Tanana Chiefs), Kotzebue (Maniilaq), and Anchorage for unwavering support; to Lisa Knutson, David Hall, Deborah Parks, Florence Watts, Mary Norton, Mary Magee, Joe Posid, and Ned Rassmussen; to our Data Monitoring and Safety Committee (Drs. David Klein, Frank Tyeryar, Paul Meier, William Blackwelder, David Fraser, Kenneth Fleshman, Walter Dowdle, and Bascom Anthony); and to the more than 2100 Alaskan Eskimo, Indian, and Aleut families who with their infants participated in this trial because of their commitment to the prevention of invasive H. influenzae type b disease.
Source Information
From the UCLA Center for Vaccine Research, Department of Pediatrics, Harbor–UCLA Medical Center, University of California, Los Angeles, Torrance (J.W., G.W.L.); and the Alaska Area Indian Health Service (G.B.) and the Arctic Investigations Laboratory, Centers for Disease Control (W.L.H.), both in Anchorage, Alaska.
- References
References
*
See NAPS document no. 04814 for two pages of supplementary material. Order from NAPS c/o Microfiche Publications, P.O. Box 3513, Grand Central Station, New York, NY 10163–3513. Remit in advance (in U.S. funds only) $7.75 for photocopies or $4 for microfiche. Outside the U.S. and Canada add postage of $4.50 (1.50 for microfiche postage).
1
Ward J, Cochi S. Haemophilus influenzae vaccines. In: Plotkin SA, Mortimer EA Jr, eds. Vaccines. Philadelphia: W.B. Saunders, 1988:300–32.
2
Ward
CrossRef | Web of Science | Medline3
Ward
CrossRef | Web of Science | Medline4
Coulehan
Web of Science | Medline5
Hall
Web of Science | Medline6
Alexander
Web of Science7
Alexander
Web of Science | Medline8
Peltola
Full Text | Web of Science | Medline9
Santosham
Full Text | Web of Science | Medline10
Greenberg
CrossRef | Web of Science | Medline11
Ward
Medline12
Osterholm
CrossRef | Web of Science | Medline13
Harrison
Web of Science | Medline14
Harrison
CrossRef | Web of Science | Medline15
Black
CrossRef | Web of Science | Medline16
Shapiro
CrossRef | Web of Science | Medline17
Kayhty
CrossRef | Web of Science | Medline18
Schneerson
CrossRef | Web of Science | Medline19
Chu
Web of Science | Medline20
American Academy of Pediatrics Committee on Infectious Diseases
Web of Science | Medline21
Update: prevention of Haemophilus influenzae type b disease . MMWR 1988; 37:13–6.
Medline22
Eskola
CrossRef | Web of Science | Medline23
Zahradnik
CrossRef24
Eskola
Full Text | Web of Science | Medline25
Ward 26
Heyward
CrossRef | Web of Science | Medline27
Kilian M. Haemophilus. In: Lennette EH, Balows A, Hausler WJ Jr, Shadomy HJ, eds. Manual of clinical microbiology. 4th ed. Washington, D.C.: American Society for Microbiology, 1985:387–93.
28
Ward
CrossRef | Web of Science | Medline29
Anderson
Web of Science | Medline30
Ward
Web of Science | Medline31
Fleiss JL. Statistical methods for rates and proportions. 2nd ed. New York: John Wiley, 1981.
32
Kleinbaum DG, Kupper LL, Morgenstern H. Epidemiologic research: principles and quantitative methods. Belmont, Calif.: Lifetime Learning Publications, 1982.
33
Miller RG Jr. Simultaneous statistical inference. 2nd ed. New York: Springer-Verlag, 1981:67–75.
34
Petersen
CrossRef | Web of Science | Medline35
Petersen
CrossRef | Web of Science | Medline36
Madore
Web of Science | Medline37
Anderson
CrossRef | Web of Science | Medline38
Weinberg
CrossRef | Web of Science | Medline39
Weinberg
CrossRef | Web of Science | Medline40
Claesson
CrossRef | Web of Science | Medline41
Berkowitz
CrossRef | Web of Science | Medline42
Greenberg
Web of Science43
Osterholm MT. Efficacy of Haemophilus b plain polysaccharide (PRP) vaccine and conjugate vaccine (PRP-D) in Minnesota. In: Program and abstracts of the 29th Interscience Conference on Antimicrobial Agents and Chemotherapy, Houston, September 17–20, 1989. Washington, D.C.: American Society for Microbiology, 1989. abstract.
44
Wainwright
Web of Science45
Ward
CrossRef | Web of Science | Medline46
Eskola
Full Text | Web of Science | Medline47
Takala
CrossRef | Web of Science | Medline48
Ward JI, Lum MK, Barenkamp SJ, et al. Unusual outer membrane protein profiles of Haemophilus influenzae type b: epidemiologic implications for Alaskan Eskimos. Presented at the 22nd Interscience Conference on Antimicrobial Agents and Chemotherapy, Miami, October 4–6, 1982.
49
Musser
CrossRef | Medline50
Siber GR, Santosham M, Priehs CM, et al. Impaired antibody (Ab) response to H. influenzae b (Hib) capsular polysaccharide (CP) in Native American children. In: Program and abstracts of the 27th Interscience Conference on Antimicrobial Agents and Chemotherapy, New York, October 4–7, 1987. Washington, D.C.: American Society for Microbiology, 1987. abstract.
- Citing Articles (48)
Citing Articles
1
Shamez N. Ladhani. (2012) Two Decades of Experience With the Haemophilus influenzae Serotype b Conjugate Vaccine in the United Kingdom. Clinical Therapeutics
CrossRef2
Gregory A. Poland. (2011) Methodologic reporting requirements for clinical trials: Advancing the science today and tomorrow. Vaccine 29:32, 5087-5089
CrossRef3
Markus Knuf, Frank Kowalzik, Dorothee Kieninger. (2011) Comparative effects of carrier proteins on vaccine-induced immune response. Vaccine 29:31, 4881-4890
CrossRef4
J. Eskola. (2010) Foresight in medicine: current challenges with Haemophilus influenzae type b conjugate vaccines. Journal of Internal Medicine 267:3, 241-250
CrossRef5
George H Swingler, Desiree Michaels, Gregory GD Hussey, George H Swingler. 2009. Conjugate vaccines for preventing Haemophilus influenzae type B infections. .
CrossRef6
Michael Bröker, Peter M. Dull, Rino Rappuoli, Paolo Costantino. (2009) Chemistry of a new investigational quadrivalent meningococcal conjugate vaccine that is immunogenic at all ages. Vaccine 27:41, 5574-5580
CrossRef7
Renuka Verma, Margaret C. Fisher. (2009) Bacterial meningitis vaccines: Not just for kids. Current Infectious Disease Reports 11:4, 302-308
CrossRef8
Gregory A. Poland, Inna G. Ovsyannikova, Robert M. Jacobson. (2008) Vaccine immunogenetics: Bedside to bench to population. Vaccine 26:49, 6183-6188
CrossRef9
Shaun K Morris, William J Moss, Neal Halsey. (2008) Haemophilus influenzae type b conjugate vaccine use and effectiveness. The Lancet Infectious Diseases 8:7, 435-443
CrossRef10
George H Swingler, Desiree Michaels, Gregory GD Hussey, George H Swingler. 2007. Conjugate vaccines for preventing Haemophilus influenzae type B infections. .
CrossRef11
Hyunju Lee, So Eun Park, Soo Young Lim, Kyong Min Choi, Hoan Jong Lee, Kyung Hyo Kim. (2007) Antibody prsistence after Haemophilus influenzae type b (Hib) primary vaccination and response to boosters in Korean children. Korean Journal of Pediatrics 50:5, 449
CrossRef12
Alan P. Knutsen. (2006) Spectrum of Antibody Deficiency Disorders with Normal or Near-Normal Immunoglobulin Levels. Pediatric Asthma, Allergy & Immunology 19:1, 51-62
CrossRef13
Kyung Hyo Kim. (2006) Present status and prospects of Haemophilus influenzae type b(Hib) immunization. Korean Journal of Pediatrics 49:3, 242
CrossRef14
Aruna Chandran, James P Watt, Mathuram Santosham. (2005) Prevention of Haemophilus influenzae type b disease: past success and future challenges. Expert Review of Vaccines 4:6, 819-827
CrossRef15
Thomas W. Hennessy, Rosalyn J. Singleton, Lisa R. Bulkow, Dana L. Bruden, Debby A. Hurlburt, Debra Parks, Matthew Moore, Alan J. Parkinson, Anne Schuchat, Jay C. Butler. (2005) Impact of heptavalent pneumococcal conjugate vaccine on invasive disease, antimicrobial resistance and colonization in Alaska Natives: progress towards elimination of a health disparity. Vaccine 23:48-49, 5464-5473
CrossRef16
TANELI PUUMALAINEN, RON DAGAN, TOMI WUORIMAA, ROSE ZETA-CAPEDING, MARILLA LUCERO, JUKKA OLLGREN, HELENA KÄYHTY, HANNA NOHYNEK. (2003) Greater antibody responses to an eleven valent mixed carrier diphtheria- or tetanus-conjugated pneumococcal vaccine in Filipino than in Finnish or Israeli infants. The Pediatric Infectious Disease Journal 22:2, 141-149
CrossRef17
P Helena Mäkelä, Helena Käyhty. (2002) Evolution of conjugate vaccines. Expert Review of Vaccines 1:3, 399-410
CrossRef18
R. Gylca, V. Gylca, O. Benes, A. Melnic, V. Chicu, C. Weisbecker, P. Willems, A. Kaufhold. (2000) A new DTPa-HBV-IPV vaccine co-administered with Hib, compared to a commercially available DTPw-IPV/Hib vaccine co-administered with HBV, given at 6, 10 and 14 weeks following HBV at birth. Vaccine 19:7-8, 825-833
CrossRef19
S. Guthridge, P. McIntyre, D. Isaacs, M. Hanlon, M. Patel. (2000) Differing serologic responses to an Haemophilus influenzae type b polysaccharide–Neisseria meningitidis outer membrane protein conjugate (PRP–OMPC) vaccine in Australian Aboriginal and Caucasian infants — implications for disease epidemiology. Vaccine 18:23, 2584-2591
CrossRef20
Juhani Eskola, Joel Ward, Ron Dagan, David Goldblatt, Fred Zepp, Claire-Anne Siegrist. (1999) Combined vaccination of Haemophilus influenzae type b conjugate and diphtheria-tetanus-pertussis containing acellular pertussis. The Lancet 354:9195, 2063-2068
CrossRef21
Alexander H. Lucas, Donald C. Reason. (1999) Polysaccharide vaccines as probes of antibody repertoires in man. Immunological Reviews 171:1, 89-104
CrossRef22
Heikki Peltola. (1999) PROPHYLAXIS OF BACTERIAL MENINGITIS. Infectious Disease Clinics of North America 13:3, 685-710
CrossRef23
Ronald Gold. (1999) EPIDEMIOLOGY OF BACTERIAL MENINGITIS. Infectious Disease Clinics of North America 13:3, 515-525
CrossRef24
PAUL T. HEATH. (1998) Haemophilus influenzae type b conjugate vaccines: a review of efficacy data. The Pediatric Infectious Disease Journal 17:Supplement, S117-S122
CrossRef25
MICHAEL D. DECKER, KATHRYN M. EDWARDS. (1998) Haemophilus influenzae type b vaccines: history, choice and comparisons. The Pediatric Infectious Disease Journal 17:Supplement, S113-S116
CrossRef26
ORIN S. LEVINE, BENJAMIN SCHWARTZ, NATHANIEL PIERCE, MARK KANE. (1998) Development, evaluation and implementation of Haemophilus influenzae type b vaccines for young children in developing countries: current status and priority actions. The Pediatric Infectious Disease Journal 17:Supplement, S95-S113
CrossRef27
Stan Houston. (1997) Tuberculosis Risk and Prevention in Travelers—What about BCG?. Journal of Travel Medicine 4:2, 76-82
CrossRef28
Kim Mulholland, Stephen Hilton, Richard Adegbola, Stanley Usen, Anslem Oparaugo, Charles Omosigho, Martin Weber, Ayo Palmer, Gisela Schneider, Kebba Jobe, George Lahai, Shabbar Jaffar, Ousman Secka, Kimi Lin, Chantal Ethevenaux, Brian Greenwood. (1997) Randomised trial of Haemophilus influenzae type-b tetanus protein conjugate for prevention of pneumonia and meningitis in Gambian infants. The Lancet 349:9060, 1191-1197
CrossRef29
Juhani Eskola, Rose-Marie Ölander, Tapani Hovi, Leila Litmanen, Sara Peltola, Helena Käyhty. (1996) Randomised trial of the effect of co-administration with acellular pertussis DTP vaccine on immunogenicity of Haemophilus influenzae type b conjugate vaccine. The Lancet 348:9043, 1688-1692
CrossRef30
J. Eskola. (1995) Use of conjugate vaccines to prevent meningitis caused by Haemophilus influenzae type b or Streptococcus pneumoniae. Journal of Hospital Infection 30, 313-321
CrossRef31
DAN M. GRANOFF, ALEXANDER H. LUCAS. (1995) Laboratory Correlates of Protection against Haemophilus influenzae Type b Disease.. Annals of the New York Academy of Sciences 754:1 Combined Vacc, 278-288
CrossRef32
Hoan-Jong Lee, Jin-Han Kang, Jørgen Henrichsen, Helle Bossen Konradsen, Seong-Hee Jang, Hee-Young Shin, Hyo-Seop Ahn, Yong Choi, Luc Hessel, Sung-Woo Nam. (1995) Immunogenicity and safety of a 23-valent pneumococcal polysaccharide vaccine in healthy children and in children at increased risk of pneumococcal infection. Vaccine 13:16, 1533-1538
CrossRef33
N. I. Leaves, J. Z. Jordens. (1994) Development of a ribotyping scheme forHaemophilus influenzae type b. European Journal of Clinical Microbiology & Infectious Diseases 13:12, 1038-1045
CrossRef34
Paul R. Burton. (1994) Helping doctors to draw appropriate inferences from the analysis of medical studies. Statistics in Medicine 13:17, 1699-1713
CrossRef35
S. Zielen, P. Ahrens, D. Hofmann, E.Louise Teare, ChristopherK. Fairley, Joanne White, NormanT. Begg. (1994) Efficacy of Hib vaccine. The Lancet 344:8925, 828-829
CrossRef36
Rüdiger von Kries, Alfred Windfuhr, Anke Lücking, Helmut Helwig. (1994) Preventing Haemophilus influenzae meningitis: Germany's experience. The Lancet 344:8920, 469
CrossRef37
P. McINTYRE. (1994) Vaccines against invasive Haemophilus influenzae type b disease. Journal of Paediatrics and Child Health 30:1, 14-18
CrossRef38
Mathuram Santosham. (1993) Prevention of Haemophilus influenzae type b disease. Vaccine 11, S52-S57
CrossRef39
J. HANNA. (1992) The epidemiology and prevention of Haemophilus influenzae infections in Australian Aboriginal children. Journal of Paediatrics and Child Health 28:5, 354-361
CrossRef40
H. Peltola, T. Kilpi, M. Anttila, H. Peltola, T. Kilpi. (1992) Rapid disappearance of Haemophilus influenzae type b meningitis after routine childhood immunisation with conjugate vaccines. The Lancet 340:8819, 592-594
CrossRef41
Wing Yen Wong, Darleen R. Powars, Linda Chan, Alan Hiti, Cage Johnson, Gary Overturf. (1992) Polysaccharide encapsulated bacterial infection in sickle cell anemia: A thirty year epidemiologic experience. American Journal of Hematology 39:3, 176-182
CrossRef42
Gregg J. Silverman. (1992) Human Antibody Responses to Bacterial Antigens: Studies of a Model Conventional Antigen and a Proposed Model B Cell Superantigen. International Reviews of Immunology 9:1, 57-78
CrossRef43
Mitchell G. Scott, Hans G. Zachalv, Moon H. Nahm. (1992) The Human Antibody V Region Repertoire to the Type B Capsular Polysaccharide of Haemophilus influenzae. International Reviews of Immunology 9:1, 45-55
CrossRef44
Santosham, Mathuram, Wolff, Mark, Reid, Raymond, Hohenboken, Matt, Bateman, Massee, Goepp, Julius, Cortese, Margaret, Sack, David, Hill, Janné, Newcomer, Wendy, Capriotti, Linda, Smith, Jeanne, Owen, Marjorie, Gahagan, Sheila, Hu, Diana, Kling, Renee, Lukacs, Linda, Ellis, Ronald W., Vella, Philip P., Calandra, Gary, Matthews, Holly, Ahonkhai, Vincent, . (1991) The Efficacy in Navajo Infants of a Conjugate Vaccine Consisting of Haemophilus influenzae Type b Polysaccharide and Neisseria meningitidis Outer-Membrane Protein Complex. New England Journal of Medicine 324:25, 1767-1772
Full Text45
Joel Ward. (1991) Prevention of invasive Haemophilus influenzae type b disease: lessons from vaccine efficacy trials. Vaccine 9, S17-S24
CrossRef46
Ziad M. Shehab, Parvin Azimi, Basim I. Asmar, Mary Gail Dunphy, Elia M. Ayoub, Adnan S. Dajani. (1991) Immunogenicity and Reactogenicity of Haemophilus influenzae Type b – Meningococcus Group B Outer Membrane Protein Conjugate Vaccine in Children 2–60 Months of Age. Scandinavian Journal of Infectious Diseases 23:6, 763-769
CrossRef47
(1990) Vaccination against Haemophilus influenzae type b. The Lancet 336:8730-8731, 1573
CrossRef48
Robbins, John B., Schneerson, Rachel, . (1990) Evaluating the Haemophilus influenzae Type b Conjugate Vaccine PRP-D. New England Journal of Medicine 323:20, 1415-1416
Full Text
