A Comparative Efficacy Trial in Germany in Infants Who Received Either the Lederle/Takeda Acellular Pertussis Component DTP (DTaP) Vaccine, the Lederle Whole-Cell Component DTP Vaccine, or DT Vaccine
Background. The goal of the trial was to determine the efficacy of a multicomponent acellular pertussis vaccine against Bordetella illnesses in comparison with a whole-cell product and DT.
Design. In a randomized, double-blind fashion, 2- to 4-month-old infants received 4 doses of either DTP or DTaP vaccine at 3, 4.5, 6, and 15 to 18 months of age. The controls received 3 doses (3, 4.5, 15 to 18 months of age) of DT vaccine. The DTP vaccine was Lederle adsorbed vaccine (licensed in the United States) and DTaP was Lederle/Takeda adsorbed vaccine. Follow-up for vaccine efficacy started 2 weeks after the third dose (DTP/DTaP) and at the same age (6.5 months) in DT recipients. Reactogenicity of all doses of all three vaccines was documented by standardized parent diary cards. In addition, all subjects were monitored for respiratory illnesses and serious adverse events by biweekly phone calls.
Results. From May 1991 to January 1993, a total of 10 271 infants were enrolled: 8532 received either DTP or DTaP and 1739 received DT. Specific efficacy against B pertussisinfections with cough ≥7 days duration was 83% (95% confidence interval [CI]: 76–88) and 72% (95% CI: 62–79) for DTP and DTaP, respectively; results for DTP and DTaP based on ≥21 days of cough with either paroxysms, whoop or posttussive vomiting (PWV) were 93% (95% CI: 89–96) and 83% (95% CI: 76–88), respectively. For DTaP vaccine, efficacy was higher after the fourth dose as compared with its efficacy after the third dose (78% vs 62% for cough ≥7 days and 85% vs 76% for cough ≥21 days with PWV). For DTP vaccine, efficacy was less varied after the third and fourth dose (78% vs 85% for cough ≥7 days and 93% vs 93% for cough ≥21 days with PWV). In contrast with DTP, the DTaP vaccine had some efficacy against B parapertussisinfection (point estimate for cough ≥7 days: 31% [95% CI: −10–56]). All vaccines were generally well-tolerated. However, side reactions were significantly less after DTaP compared with DTP.
Conclusions. Like other multicomponent acellular pertussis vaccines, the Lederle/Takeda DTaP vaccine demonstrated good efficacy against mild and typical pertussis due to B pertussisinfections. Interestingly, it also may have some efficacy againstB parapertussis. Based on the results of this trial, the vaccine was licensed in the United States in December 1996 for all 5 doses of the currently recommended immunization schedule in this country.
- DTaP =
- diphtheria-tetanus toxoids, acellular pertussis vaccine, adsorbed •
- DTP =
- diphtheria-tetanus toxoids, whole cell pertussis vaccine, adsorbed •
- DT =
- diphtheria-tetanus vaccine •
- FHA =
- filamentous hemagglutinin •
- PT =
- pertussis toxin •
- PCR =
- polymerase chain reaction •
- IgG =
- immunoglobulin G •
- IgA =
- immunoglobulin A •
- ELISA =
- enzyme-linked immunosorbent assay •
- EU/mL; ELISA units per mL; WHO =
- World Health Organization •
- PWV =
- paroxysms, whoop, or posttussive vomiting •
- MAL =
- minimal acute level •
- CI =
- confidence interval
The Lederle/Takeda acellular pertussis component diphtheria-tetanus toxoids, pertussis vaccine, adsorbed (DTaP) has been extensively evaluated in the United States and Germany, and DTaP vaccines containing the same acellular pertussis component (Takeda and Wyeth/Takeda) have been studied in Japan and the United States.1-22 Immunogenicity of vaccine components and low reactogenicity have been demonstrated in infants, older children, and adults.1-21 Moreover, a high degree of efficacy was shown in an unblinded household contact study in Japanese children who were 2 years old or older.22 Based on these data this vaccine was licensed in the United States for use as the fourth and fifth doses in children ≥15 months of age in 1991.23 More recently, as the result of data in this communication, the vaccine was also licensed for infant use in December 1996.24
In 1959, after pertussis immunization had been introduced in West Germany, the yearly incidence of the illness based on notified cases was 79/100 000.25 For the next 15 years the whole-cell vaccine was commonly used with approximately 50% to 60% of children being immunized.26 However, the influence of pertussis vaccination on the epidemiology of the illness could not be reliably established because required notification of pertussis to health authorities had been abolished in 1961. Although it was generally felt that the incidence decreased over the years, pertussis still remained endemic and epidemic. In 1974, concerns about events thought to be serious side effects of the vaccine, led to a limited recommendation of pertussis vaccine and subsequently the immunization rate dropped to approximately 15%.26 There was regional variability from virtually no immunizations in northern areas of West Germany to vaccination rates of 30% in parts of Bavaria. As a consequence, an increase in cases of pertussis occurred over the years with an estimated yearly incidence of 180/100 000 in the 1980s.27This allowed us to perform the present efficacy trial.
This was a longitudinal cohort study in which infants were vaccinated with the Lederle/Takeda acellular pertussis component DTP vaccine (DTaP), or Lederle whole-cell component DTP vaccine (DTP) or any diphtheria and tetanus toxoids vaccine (DT) licensed in Germany. DTaP and DTP vaccines were administered in a double-blind, randomized manner whereas DT vaccine was administered in an open arm of the study based on parent preference. Study vaccines (DTaP and DTP) were supplied in single-dose vials in groups of 10. Each vial contained a subject number and the vaccines were assigned in numerical sequence to enrollees. Subject numbers for potential DT recipients were supplied to the investigators and were used in numerical sequence in DT recipients. The planned enrollment was 6000 infants in the DTaP/DTP blinded group and 2000 infants in the DT group.
The trial was carried out primarily in the private practices of German pediatricians; a small number of general practice physicians and hospital-based pediatricians also participated.
The Lederle/Takeda DTaP vaccine (Wyeth-Lederle Vaccines and Pediatrics, Pearl River, NY) is similar to the vaccine licensed in the United States in 1991 for the fourth and fifth vaccine doses and is the same as the presently licensed vaccine for infant use. Its composition per dose (0.5 mL) is: 300 hemagglutinating units of the acellular pertussis component that contains 40 μg pertussis protein with approximately 86% filamentous hemagglutinin (FHA), 8% toxoided pertussis toxin (PT), 4% pertactin, and 2% fimbriae-2. The vaccine is formulated to contain 9.0 Lf diphtheria toxoid, 5.0 Lf tetanus toxoid, aluminum 0.23 mg, and thimerosal 1:10 000. Five lots of DTaP vaccine were used.
The DTP vaccine (Wyeth-Lederle Vaccines and Pediatrics, Pearl River, NY) is the same vaccine licensed in the United States. It is formulated to contain ≥4 protective pertussis units, 0.1 mg aluminum, 12.5 Lf diphtheria toxoid, 5.0 Lf tetanus toxoid, and thimerosal 1:10 000 per dose (0.5 mL). Four lots of DTP vaccine were used. The DT vaccine was decided by study physician preference and included any product licensed in Germany.
Participating physicians enrolled healthy 2-to 4-month-old (60 to 123 days) infants after obtaining informed consent from the parents. Parents of potential subjects were first asked if they would like to have their child vaccinated against pertussis. If the answer was “yes,” they were asked if they would agree to have their child enrolled in the blinded arm of the study (DTaP/DTP). If they declined pertussis vaccine and their preference was DT vaccine they were asked to have their child enrolled in the DT open study arm.
Infants participating in the blinded group received 4 doses of vaccine with the first dose given at 2 to 4 months of age, the second and third doses given at least 6 weeks after the preceding dose, and the fourth dose at 15 to 18 months of age. DT recipients received their first dose at 2 to 4 months of age, a second dose at least 6 weeks later, and a third dose at 15 to 18 months of age.
Sera were collected from DTaP and DTP vaccinees 1 month after the third and fourth doses and from DT recipients at the equivalent time points (3 months after the second dose and 1 month after the third dose). In addition sera were collected from a randomized preselected sample of subjects in each vaccine group at approximately 3-month intervals. These sera were used to construct antibody kinetic curves for each vaccination group so that serologic diagnosis of infection could on occasions be made by antibody values in a single serum at the time of an illness in a study subject.
Adverse Event Monitoring
The parents kept a reaction diary for 3 days after each dose of vaccine and long-term follow-up was monitored by telephone every 2 weeks.
Monitoring Of Cough Illnesses
Parents of study subjects were instructed to report all cough illnesses in the vaccinee or other family members to their study physician. In addition all families of study subjects were monitored by biweekly phone calls. If a cough illness in a vaccinee or a family member of ≥7 days' duration without improvement was reported, the study physician saw the ill vaccinee or family member and obtained a nasopharyngeal swab and an acute blood sample. If the cough illness persisted for ≥14 days, the vaccinee or family member was seen by one of three Central Investigators from the Study Center. The Central Investigators (co-authors S.S-G., M.Ü., and U.H.) performed a standardized history and physical examination and made a clinical diagnosis of 1 of 5 categories: definite pertussis, definite pertussis with a complication, probable pertussis, possible pertussis, or not pertussis. The Central Investigators were blinded as to whether the child had received a pertussis containing vaccine or DT; this blinding was occasionally broken inadvertently by the parent or the primary physician.
Identified cough illnesses were monitored by the study physician for illness duration by a family follow-up book. In instances of cough of ≥7 days, the study physician completed a suspected pertussis case report form. Convalescent-phase blood samples were obtained by the study physicians 6 to 8 weeks after the illness onset.
The method of specimen collection, transport, and culture has been previously presented.28
Polymerase Chain Reaction (PCR)
During the final year of the case observation period of the trial PCR was also used for diagnosis. The studies were performed by Dr Gabriela Schmidt-Schläpfer in Basel, Switzerland, and the methods have been previously reported.29 Specimens for PCR were collected using a second Dacron (Mast Diagnostica, Reinfeld, Germany) nasopharyngeal swab.
Serum antibodies both immunoglobulin G (IgG) and immunoglobulin A (IgA), to PT, FHA, pertactin, and fimbriae-2 were determined by enzyme-linked immunosorbent assay (ELISA) at Wyeth-Lederle Vaccines and Pediatrics by a modification of the parallel line method30 as previously described.8 9 16Results were standardized by the use of US human reference antisera lot 3 and lot 4. All ELISA values are expressed as ELISA units per mL (EU/mL).
Agglutinins were measured by the microagglutination assay in the Erlangen laboratory using B pertussis strain 460 as the antigen.30 Titers are expressed as the reciprocal of the highest serum dilution with agglutination.
After an extensive lecture campaign in 1990 and early 1991 to inform German physicians regarding the safety and need for pertussis immunization, the study got underway in May 1991. It was recognized early by our group that acceptance of pertussis immunization by parents was generally more physician-directed than parent-directed. The initial physicians who joined the study were mainly those who recommended pertussis immunization. Gradually, over 12 months, after visits by the study coordinator to physician's offices, new physicians entered into the trial. In general, the physicians who joined later had initially been reluctant to use DTP vaccine. Finally, physicians in 227 sites actively participated in the trial. At these sites 93% of the physicians were pediatricians in private practice and 7% were general practitioners or hospital-based pediatricians. The study initially started in Bavaria and West Berlin. Later, the Saar region was added as well as a substantial number of sites in the 2 states (Baden-Württemberg and Hessen) on Bavaria's western border.
Overall Study Monitoring
Quintiles (Germany) GmbH was the principal monitor of the trial. Personnel from Quintiles enrolled study physicians, monitored all records, managed, and computerized all physician subject information and laboratory data.
The Erlangen study group included the director and co-director of the project (K.S. and J.D.C.), the study coordinator (U.H.), and the Central Investigators. All case data generated by the Central Investigators were added to the Quintiles data base.
The views of the manufacturer and the principal investigators of the project were, on occasion, different. The manufacturer carried out an independent analysis of efficacy which used slightly different criteria from that presented in this report.
The primary case definition determined by the manufacturer, as presented in the original study protocol, was to be consistent with the criteria established by an ad hoc committee of the World Health Organization (WHO).31 By this definition a case was defined as the presence of at least 21 consecutive days of paroxysmal cough and a positive culture for B pertussis or a significant increase in IgA or IgG antibody to PT, FHA, or fimbriae or a culture-confirmed household contact. However, because of a previous study by our group as well as further analysis of a previous efficacy trial in Sweden, we did not plan to use this as our primary case definition because cases of pertussis would be removed from the analysis and efficacy would be artificially inflated.28 32 33
After the establishment of a modified WHO case definition, the blind was broken and a preliminary analysis of efficacy by the manufacturer was presented at the International Pertussis Vaccine Meeting in Rome34 in November 1995. The case definitions used in our present analysis were established after the unblinding by the manufacturer and, except for information on culture, our criteria were determined before the availability of specific clinical and serological data.
Based on data related to cases that were culture positive for B parapertussis we suspected that the two study vaccines might have some efficacy against B parapertussis as well as B pertussis. An analysis of acute and convalescent phase sera from subjects who were culture-positive for B parapertussisindicated that most had an antibody rise to FHA and some to pertactin, and fimbriae-2 by ELISA and some also had modest agglutinin titer rises.
From the above, three primary case definitions were formulated based on the following serologic and culture results. The first definition included positive culture for either B pertussis or B parapertussis, or a household contact to a culture proven case due to either Bordetella species or significant IgG or IgA antibody responses by ELISA to any of the 4 B pertussisantigens (PT, FHA, pertactin, fimbriae-2) or a significant agglutinin titer rise. The second definition was specific for illness due toB pertussis and includes positive B pertussisculture, or a household contact to a culture proven B pertussis illness or significant IgG or IgA antibody responses by ELISA to PT. The third primary definition included only B parapertussis culture in the vaccinee or a family member or a significant IgG or IgA ELISA antibody response to FHA, pertactin, or fimbriae-2 in the absence of a significant antibody response to PT. All three of these primary definitions required a cough of ≥7 days' duration.
Multiple secondary case definitions were established that used different durations of cough illness and severity with different laboratory criteria as well as efficacy calculated on the basis of clinical criteria alone.
Because of an ambiguous translation of the word paroxysm from English to German, it was noted that paroxysmal cough was underreported by primary physicians. Because of this our clinical case definition representative of the WHO case definition is modified. This definition is ≥21 days of cough with the presence of either paroxysms, whoop, or posttussive vomiting (PWV).
Cases were determined serologically by the demonstration of significant rises (between acute phase and convalescent-phase samples) in IgA or IgG antibody to the four B pertussis antigens by ELISA or a fourfold increase in agglutinin titer. In subjects with only a convalescent-phase serum sample or in those in whom collection of the acute-phase specimen was late, significant rises were determined by the use of the respective postimmunization value and the calculation of the acute-phase value from an antibody kinetic curve.
Specifically, post-third and post-fourth dose kinetic curves were developed for DTaP, DTP, and DT vaccine groups. A total of 1058 sera from randomly preselected DTaP and DTP recipients were available. They were separated into the following time intervals after immunization: 14–63 days, 64–138 days, 139–236 days, and >236 days after the third but before the fourth dose and 14–74 days, 75–149 days and >149 days after the fourth dose. There was a mean of 66 (range, 34–87) serum samples per time interval per vaccine group. For each time interval IgG and IgA antibody values against PT, FHA, pertactin, and fimbriae 2 were plotted and percentiles were determined. For each cough episode a child's post-third dose value percentile (if onset of cough was before the fourth dose) or post-fourth dose value percentile (if onset of cough was after the fourth dose) was determined from the kinetic curve. The antibody value of the determined percentile, in the time interval when the cough started, was the calculated acute value. If an appropriate post-third or post-fourth dose serum was not available then the acute-phase value was estimated from the 95th percentile value for the respective time interval.
Sera were also randomly preselected from 620 DT subjects throughout the study period for kinetic curve construction. Most of these values were below the minimal acute level (MAL) (see below) so that percentiles could not be assigned. Therefore in DT subjects the MAL is used in all time periods as the calculated acute value.
For significant ELISA antibody rises the 95th, 99th, and 99.9th percentile fold change limits were determined from assay quantitation data supplied by Wyeth-Lederle Vaccines and Pediatrics. The percentile fold change limits varied among the different case definitions. ForB pertussis-specific case definitions in which only a single antibody to PT (IgG or IgA) established the diagnosis the 99th percentile fold limit was used; if two values (both PT IgG and IgA) were used then the 95th percentile fold limit was acceptable. A single positive antibody rise to PT based on the 95th percentile fold change limit was accepted if at least one supporting rise to FHA, pertactin, or fimbriae-2 was noted using the 99th percentile fold change limit. For other broader definitions with multiple antigens, the 99.9th percentile fold change limit was used for diagnosis by a single test; the 99th and 95th percentile fold change limits were acceptable if the diagnosis was made on the basis of 2 or ≥3 positives, respectively.
The MALs (IgG and IgA) for each antigen were determined by the visual examination of the frequency distribution of values in randomly preselected sera from DT vaccines collected 2, 3, and 5 months after the second dose (approximately 6, 7, and 9 months of age). These MAL values were similar to the 95th percentile except for FHA IgG and PT IgG; for these two antibodies the visual breakpoint determined a higher value. The MALs were as follows: IgG, PT 5.7 EU/mL; FHA, 7.0 EU/mL; pertactin, 8.0 EU/mL; fimbriae-2, 1.5 EU/mL; IgA, PT 3.7 EU/mL; FHA 2.7 EU/mL, pertactin, 5.0 EU/mL; and fimbriae-2, 12.5 EU/mL. The agglutinin MAL was 16.
Vaccine efficacy was calculated using Cox proportional-hazards regression35 with SAS (SAS, Cary, NC) software for personal computers. Differences between proportions were determined by the χ2 test or Fisher's exact test.
Enrollment began in May 1991 and at the end of the first year, 7352 children had been enrolled. However, only 15% were DT recipients. Therefore, enrollment was continued until January 31, 1993 at which time 10 271 children had been enrolled; 4273 were entered in the DTaP group, 4259 in the DTP group and 1739 (17%) in the DT group. Of those enrolled 97% completed the initial primary series (DTaP/DTP 3 doses and DT 2 doses) and 92% completed the reinforcing dose (DTaP/DTP fourth dose and DT third dose). The follow-up period ended December 15, 1994. Eight hundred sixteen subjects (8%) did not complete the full immunization series; 283 were in the DTaP group, 335 in the DTP group and 198 were in the DT group. The reasons for leaving the study were: change of physician or relocation (25%), parent's refusal to continue (25%), adverse events (16%), received marketed DTP (9%), lost to follow-up (9%), pertussis (8%) and protocol violation, unknown, and miscellaneous reasons (7%). A higher percentage of DTP subjects discontinued because of adverse experiences (DTP: 2.2%, DTaP: 0.8%, and DT: 0.2%) and the overall dropout rate due to multiple reasons was higher in DT subjects (11%) then in DTaP (7%) or DTP (8%) group participants.
A total of 38 045 doses of the study vaccines (DTaP, DTP or DT) were administered during the course of the study. Of these doses 61% were administered in the anterior lateral thigh, 33% in the buttock, 2.7% in the deltoid and in 3.3% the site of immunization was not reported. All doses were administered intramuscularly.
The efficacy per protocol analysis included 9597 total subjects with 1615 (93%) DT subjects, 4025 (94%) DTaP subjects, and 3957 (93%) DTP subjects. In the efficacy analysis the follow-up period began 14 days after the third dose in DTaP/DTP vaccinees and 61 days after the second dose in DT recipients. Follow-up years were: DT 2916, DTaP 8581, and DTP 8416. During the follow-up period there were 13.1 cough reports per 100 follow-up years in DT subjects and 8.4 and 8.5 reports per 100 follow-up years in DTaP and DTP subjects, respectively. Of the 1824 children with reported cough illnesses during the follow-up period 81% had one or more serum specimens collected in relation to the illness. In 566 instances (31%) paired sera were available.
The Central Investigators evaluated 1745 cough episodes in study children of which 1245 occurred within the follow-up period; there were 9.5 evaluations per 100 follow-up years in DT subjects and 5.7 evaluations per 100 follow-up years in both DTaP and DTP recipients.
There were 95 cases from whom B pertussis was isolated or who were household contacts of culture-confirmed B pertussisillness cases. In this group serologic data were available in 84 (88%). Of these, 10 (12%) did not have a significant IgG or IgA antibody response to any of the four B pertussis antigens. Of the 74 with an antibody response 64 (86%) had a significant response to PT. The percent responses to the other antigens were: FHA, 85%; pertactin, 66%; and fimbriae-2, 24%. Of the 10 patients without antibody rise to PT, 4 had rises to FHA, 4 to pertactin, 1 to fimbriae-2, and 1 to both fimbriae-2 and pertactin.
There were 28 cases from whom B parapertussis was isolated or who were household contacts of culture-confirmed B parapertussis cases. In this group serologic data were available in 23 (82%). Of these 4 (17%) did not have a significant IgG or IgA antibody response to any of the four B pertussis antigens. All 19 of those with an antibody response had a significant rise to FHA. Three of the 19 responders (16%) had a significant titer rise to PT. The percent responses to pertactin and fimbriae-2 were 79% and 42%, respectively.
Overall there were 401 laboratory-confirmed cases of pertussis due to either B pertussis or B parapertussis (Table1). Of this group 123 (31%) had a positive culture or household contact with a culture proven case and the remainder (278) were diagnosed by antibody titer rises.
There were 238 cases satisfying our B pertussis-specific primary case definition (Table 1). Of this group, 208 (87%) could be diagnosed serologically; but 65 (31%) had other laboratory evidence of infection, as well. Of the 30 cases without serologic evidence of infection 14 were diagnosed by culture and 16 by household contact with a culture-confirmed case. Ninety-five (40%) of the cases could be diagnosed by culture or by household contact. There were 44 (18%) positive cultures in study subjects.
The median duration of cough in DT recipients was 47 days as compared with 32 days in DTP vaccinees (P = .0005) and 33 days in DTaP vaccinees (P = .001). The percentages of DT, DTP, and DTaP patients with PWV were 92, 46, and 62, respectively (P < .0001 for DT vs DTP and DT vs DTaP; P = .07 for DTP vs DTaP). In contrast, characteristics of B parapertussis infection were similar among the 3 groups. The median duration of cough was 29, 23, and 24 days for DT, DTP, and DTaP recipients (all P > .38), respectively and 59%, 52%, and 56% experienced PWV (allP > .66).
Primary Case Definitions (Table 1)
The attack rate of laboratory-confirmed pertussis due to B pertussis or B parapertussis during the observation period was 4.4 per 100 person-years in DT recipients. Forty percent of the cases had laboratory evidence of B parapertussisinfection. In DT recipients 21% of the cases were due to B parapertussis infection.
The efficacy of DTaP vaccine against mild and typical pertussis (cough ≥7 days' duration) using laboratory criteria specific for B pertussis infection was 72% (95% confidence interval [CI] = 62–79); using the same criteria the DTP vaccine had an efficacy of 83% (95% CI = 76–88). When the case definition includes all laboratory-confirmed cases of mild and typical pertussis due to eitherB pertussis or B parapertussis the efficacy of both vaccines decreases significantly; DTaP vaccine efficacy was 63% (95% CI = 53–71) and DTP vaccine efficacy was 64% (95% CI = 55–72). The decrease is greatest in DTP vaccine recipients and this may be explained by its lack of specific efficacy against B parapertussis (VE = −6% [95% CI = −64–31]). In contrast, DTaP vaccine had modest efficacy against B parapertussis (VE = 31% [95% CI = −10–56]).
There were 28 subjects from whom B parapertussis was isolated or who had a household contact with a B parapertussis case. If these cases are removed from the data set but the broad serologic criteria retained, the efficacy against mild and typical pertussis is minimally affected; DTaP efficacy remains 63% (95% CI = 53–71) and DTP efficacy is 66% (95% CI = 56–74).
Typical Pertussis (Modified WHO Clinical Case Definition) (Table2)
DTaP and DTP vaccine efficacy against typical pertussis for three different laboratory criteria are presented in Table 2. When compared with Table 1 it can be noted that the use of the modified WHO case definition leads to the removal of a significant number of laboratory-confirmed cases in each vaccine group; in the B pertussis-specific group, 47%, 64%, and 12% of the cases were removed from the DTaP, DTP, and DT vaccine groups, respectively.
The efficacies of DTaP and DTP vaccines using B pertussis-specific laboratory criteria and the modified WHO case definition are 83% (95% CI = 76–88) and 93% (95% CI = 89–96), respectively. With this case definition the whole-cell vaccine has significantly greater efficacy than the acellular product. When the laboratory criteria are expanded to include all serology, the efficacy of DTaP falls slightly to 79% (95% CI = 72–85) whereas the DTP efficacy decreases significantly to 84% (95% CI = 77–89).
Efficacy Analyzed by Duration and Severity of Cough Illness
In Table 3, DTaP and DTP vaccine efficacy based on laboratory-confirmed B pertussis infection is presented by duration of cough with and without PWV. The point estimate of efficacy for cases with and without PWV increased with increasing days of cough; however, there was little change after 21 days. In contrast, when only cases with PWV are included, duration had only a minimal effect on efficacy with either vaccine.
Efficacy After Three and Four Vaccine Doses
In Table 4, the efficacy of DTaP and DTP vaccines after three and four vaccine doses is presented. The observation period between the third vaccine dose and the fourth dose was relatively short (7684 person-years); therefore, the efficacy point estimates have relatively large 95% CIs. Against mild and typical pertussis and typical pertussis, the efficacy of DTaP vaccine was greater after the fourth dose than after the third dose (76% and 85% vs 62% and 78%, respectively). Similarly, the efficacy of DTP vaccine against mild and typical disease after 4 doses was better than after 3 doses (85% vs 78%). However, DTP vaccine efficacy against typical pertussis was equally good after the third and fourth vaccine doses (93%).
Efficacy Based on Clinical Diagnosis
Our study was unique in that almost all possible cases with a cough duration of ≥14 days were evaluated by one of three pediatricians (Central Investigators). Efficacy against typical clinical disease is presented and compared with laboratory-confirmed pertussis (B pertussis and B parapertussis) in Table 5. The efficacy against typical pertussis due to either B pertussis or B parapertussis, when determined clinically, is almost identical to that determined by clinical and laboratory study. However, even though the calculated efficacies by the case definitions were similar, the case make-up by category varied considerably. If the clinical-laboratory definition is used as the standard, the sensitivity was 55% and the specificity was 99%. When only cases in DT vaccines are considered, the sensitivity was 74%.
In contrast with the general reliability of the Central Investigators' ability to accurately diagnose pertussis due to B pertussisand B parapertussis, the reliability of duration of cough as reported by study physicians is analyzed in Table6. The percentage of cough illnesses that are due to Bordetella infections increases in linear fashion with increasing duration intervals. However, only 46% of cough illnesses and 57% of cough illnesses with PWV of ≥21 days' duration had laboratory evidence of Bordetella infection.
The Addition of PCR for Diagnosis
The addition of PCR for diagnosis during the last year of the study resulted in the detection of an additional 27 cases with cough of ≥7 days' duration. The additional cases occurred mainly in the DTaP group (10 cases) and the DTP group (13 cases) that led to a slight reduction in calculated efficacy. For specific B pertussisinfections, efficacy rates for DTaP and DTP decreased from 72% and 83% to 69% and 79%, respectively.
Possible Effect of Unblinded DT Group
Since the DT group in our trial was unblinded the manufacturer did an extensive prospective analysis to identify factors that differed between the randomized and non-randomized strata. Variables examined included demographic data on household size, history of pertussis illness, and immunizations in all household members and many more. Variables, which by logistic regression were found to be different between strata, were further analyzed in a stepwise fashion using Poisson regression to see if they were related to developing pertussis. Two variables were found to be significantly different between strata and were related to pertussis: single adult households and siblings of a study child in a household who were not vaccinated against pertussis. Based on these two variables, vaccine efficacy against typical pertussis as calculated by the manufacturer, was adjusted from 83.9% to 81.0% for DTaP and from 92.6% to 91.1% for DTP, respectively.
This procedure, however, implies a theoretically increased risk of acquiring pertussis. Our approach was different. Only 2% of all study families were single adult households. However, in 43% of DT recipients, all siblings were not vaccinated against pertussis, as compared with 28% and 26% in DTaP and DTP recipients, respectively. Assuming, that more unvaccinated siblings in DT vaccinees could be a bias resulting in higher vaccine efficacy, we identified all cases of pertussis in study subjects that were secondary to pertussis in an unvaccinated sibling and excluded them from efficacy calculations. In our calculations we used four different definitions for the primary case in the household. To our surprise, for cough illness of ≥7 days and laboratory criteria specific for B pertussis in cases, the removal of cases resulted in an increase rather than a decrease in efficacy for both DTaP and DTP vaccines. In situations in which all serology contributed to the diagnosis, the removal of household contact cases resulted in minimal or no change in efficacy. Based on these findings, no alterations have been made in our efficacy results.
Reactogenicity data have been presented elsewhere, and therefore, will only be summarized in this report.36 37 During the study 10 271 children received a total of 16 644 doses of DTaP, 16 424 doses of DTP and 4977 doses of DT. Both pertussis-containing vaccines were well-tolerated. Local reactions (erythema and induration at the injection site) and minor systemic events (fever, fussiness, drowsiness, and anorexia) were more common after all doses of DTP vaccine than after DTaP and DT vaccines.36 Antipyretics were more commonly given to DTP recipients than to DTaP recipients.
Severe adverse events were rare.37 Persistent inconsolable crying was more common in DTP recipients (1 per 113 doses) than in DTaP vaccinees (1 per 497 doses). High fever (≥40.5°C) was more common in DTP recipients (1 per 5359 doses) than in DTaP vaccinees (1 per 16 239 doses). One hypotonic-hyporesponsive episode occurred in a DTP recipient after the second dose. Convulsions within 3 days of immunization occurred in one DTaP recipient and 3 DTP recipients.
In the last few years 7 efficacy trials in which 8 DTaP vaccines were evaluated, have been completed and the results of 4 of the trials have been published and all have been presented.36-43Including our trial, there were four longitudinal cohort studies with a DT control group and in two of the trials, in addition to ours, there was a whole-cell pertussis component DTP vaccine control. A major important difference in our trial from the other three was that the DT group was blinded in the other trials whereas it was nonblinded in our trial.
Our trial was unique in other aspects as well. It was carried out in 227 private medical practices rather than in public clinics. Therefore, all study children were examined and monitored by physicians and not paramedical personnel. In addition, almost all possible cases with significant cough illnesses were evaluated by three physicians (Central Investigators) who followed an extensive structured protocol.
Our trial also differed from the other three cohort studies in that pertussis vaccine recipients received four doses of vaccine rather than three doses.
Another unique aspect of our trial was the collection of serum specimens from randomly preselected DT, DTP, and DTaP vaccine recipients at approximately 3-month intervals. The availability of these sera allowed us to construct IgA and IgG ELISA antibody kinetic curves for the four relevant B pertussis antigens (PT, FHA, pertactin, and fimbriae-2). The availability of these curves enabled us to calculate preexposure antibody values so that serologic evidence of infection could be determined in cases without acute-phase sera or in cases in which acute-phase sera were collected late.
In this trial we used three primary case definitions and examined over 60 different secondary case definitions. Because pertussis is an illness caused by both B pertussis and B parapertussis, three primary case definitions were considered in this trial.44 For comparison with other trials secondary case definitions that used the modified WHO clinical case definition were necessary.
For all case definitions involving laboratory criteria the vast majority of cases had serologic evidence of infection. For example with our B pertussis-specific primary case definition 208 of the total 238 cases had significant antibody titer rises to PT. Only 44 subjects had positive cultures and only 40% of the cases could be diagnosed by positive culture or by contact with a culture-positive household member.
In addition to absolute efficacy, vaccination altered the course of disease in children with vaccine failure. The median duration of cough in DT recipients was significantly longer (47 days) than the median duration in DTaP and DTP (33 and 32 days, respectively) vaccine failures. In addition, PWV occurred significantly more frequently in children who did not receive a pertussis vaccine than in DTaP or DTP vaccine failures (92% vs 62% and 46%, respectively).
During the period of this study B parapertussis as well asB pertussis was a common cause of illness. Of the total culture positives 24% were B parapertussis. Overall, when serology and household data are included there were 401 cases of laboratory-confirmed pertussis. Of this group 40% had laboratory data more consistent with B parapertussis infection than B pertussis infection.
Of the 28 children from whom B parapertussis was isolated or who had a household contact with a B parapertussis case, 3 had an antibody response to PT. Because B parapertussis does not liberate PT, this suggests dual infection with both bacterial species.44 45 Evidence of dual infection has been noted by others.46 47
As noted in Table 1 the efficacy of DTaP and DTP vaccines against specific B pertussis infections with cough of ≥7 days' duration was 72% (95% CI = 62–79) and 83% (95% CI = 76–88), respectively. The efficacy against pertussis caused by eitherB pertussis or B parapertussis of both vaccines is considerably less (63% and 64% for DTaP and DTP, respectively). This difference is most marked for DTP. DTaP appears to have modest efficacy against B parapertussis, but because the lower bound of the 95% CI is <1%, this observation is not confirmed.
The finding of some efficacy of DTaP vaccine against B parapertussis is surprising because it was shown by Lautrop48 four decades ago that children who had infection due to B pertussis were not necessarily subsequently protected against B parapertussis illness. In addition in a murine respiratory model Khelef at al49 found that antibody to B pertussis FHA or pertactin did not protect againstB parapertussis infection. The difference in our findings as compared with the murine model studies may be due to variations in bacterial and antibody concentrations in our subjects.
The efficacy of both DTaP and DTP against typical pertussis is presented in Table 2. Using the B pertussis-specific case definition, the efficacy of DTaP and DTP vaccines were 83% (95% CI = 76–88) and 93% (95% CI = 89–96), respectively. The manufacturer carried out an independent analysis of efficacy using a similar clinical and household contact case definition but with different B pertussis specific serologic criteria (significant rise in PT IgG between acute convalescent-phase samples or PT IgA values significantly above the background values) and noted similar efficacy point estimates. Their efficacy values were 81% (95% CI = 73–87) for DTaP and 91% (95% CI = 85–95) for DTP.24
The point estimate of efficacy against typical pertussis is 79% (95% CI = 72–85) when serology to FHA, pertactin and fimbriae-2 are included. Calculated efficacy in this trial is based on an observation period that includes the period of time between the third and fourth dose, as well as, the time after the fourth dose. As noted in Table 4, the point estimate of efficacy of DTaP vaccine appears to be considerably less in the interval after the third dose than after the fourth dose, especially against mild illness.
The comparison of different vaccines between trials is difficult. In the Italian and Swedish double-blind trials the efficacy of the same lot of Connaught DTP vaccine (Connaught Laboratories, Swiftwater, PA) varied substantially.38 41 For example, when both mild and typical pertussis were included in the case definition the efficacy in Sweden was 41% (95% CI = 30–51) and it was 23% (95% CI = 1–40) in Italy—an 18% difference.
In the Swedish and Italian trials antibody to FHA was also considered evidence of B pertussis infection. As noted above, our data indicate that if illnesses due to B parapertussis infections are prevalent, calculated efficacy decreases substantially. In the Swedish and Italian trials the vaccinees only received three doses of vaccine and not the conventional reinforcing fourth dose that we used in our trial. Considering these various facets of the Swedish trial, the Italian trial, and our trial, it appears that the Lederle/Takeda DTaP vaccine after a four-dose schedule has an efficacy profile similar to the other three multicomponent vaccines after a three-dose schedule.
In Table 3 efficacy is analyzed by duration and severity of cough. As can be seen when any cough is considered, there is a direct correlation between duration of cough and vaccine efficacy. However, when markers of cough severity (PWV) are included in the clinical diagnosis efficacy is minimally affected by total duration of cough. These data indicate that the most precise estimate of efficacy is determined by any cough regardless of duration; the WHO case definition leads to an overestimation of vaccine efficacy.
Our data compared with that noted in the Swedish and Italian trials suggest that the Lederle DTP vaccine is more efficacious than the Connaught DTP vaccine.38 41 This is supported by the fact that the Lederle vaccine is more immunogenic.50 Both vaccines elicit a low-level response to FHA; however, the responses to PT, pertactin, and fimbriae were eightfold, twofold and fourfold greater, respectively, in Lederle vaccine recipients compared with responses in Connaught vaccine recipients.
Our Central Investigator data indicate that trained physicians can identify typical pertussis in unvaccinated subjects with a high degree of sensitivity and specificity. In vaccinees, the specificity of diagnosis remains high whereas the sensitivity decreases considerably.
During the last year of observation PCR was added for diagnosis. This resulted in a modest increase in the detection of cases; a greater proportion of these cases were noted in vaccinees and therefore efficacy decreased slightly.
The Lederle/Takeda DTaP vaccine is less reactogenic than Lederle DTP vaccine.16 36 37 Persistent inconsolable crying was 4 times more common in DTP recipients than in DTaP recipients. High fever (≥40.5°C) was three times more common in DTP vaccinees than in DTaP vaccinees. Only one DTaP recipient had a convulsion in temporal relationship with immunization; no DTaP vaccinee had a hypotonic-hyporesponsive episode.
In the National Institutes of Health National Institute of Allergy and Infectious Diseases phase 2 trial it was noted that the two DTaP vaccines that contained a lower concentration of diphtheria toxoid had, in general, less local reactions.21 In a pilot study we noted that erythema and induration after the third dose of DT (containing an estimated 25 Lf of diphtheria toxoid) was significantly more common than after the fourth dose of Lederle/Takeda DTaP vaccine.51
In summary, the Lederle/Takeda DTaP vaccine is less reactogenic than Lederle DTP vaccine and its efficacy is similar to that of other multicomponent DTaP vaccines when the US primary immunization schedule is used.
This study was sponsored by Wyeth-Lederle Vaccines and Pediatrics, Pearl River, NY.
We thank the staff of Wyeth-Lederle Praxis Vaccines and Pediatrics and Quintiles Ltd and to the Study Advisory Board for their contributions to this trial. We thank by Carmen Lorenz and Regina Rost for laboratory assistance and Ilse Reiff, Ingeborg Boatey, Gerlind Baierlacher, Monika Reissinger, Jutta Heinrich and Sheila Walton for secretarial assistance.
The Pertussis Vaccine Study Group are: W. Müller, Augsburg; A. Neugebauer, Neusäβ; K. Sailer, Augsburg; H. Keller, Aschaffenburg; U. Kircher, Gersfeld/Rhön; B. Netzel, München; H. Sachsenhauser-Kratzer, Mering; M. Thelen, Planegg; B. und K.E. Buck, Ottobrunn; G. Nath, Krumbach; E. Clapier, Landsberg/Lech; P. Gelius, Marktredwitz; B. Graf zu Castell, Kaufering; H.-J. Hess, Freyung; E. Maas-Doyle, Erlangen-Tennenlohe; H.P.R. Mayer, Bamberg; K. Renner, Marktoberdorf; I. Seltsam, Gemünden; G. Seuwen, Kempten/Allgäu; N. Totzauer, Münchberg; A.-M. Zange, München; I. Bernsau, Stadtbergen; W. Knipping, Ottobrunn; Ch. Neumayer, Dachau; G. Salzer, Regensburg; R.E. Ullner, Dorfen; G. Bergen, Haβfurt; J. Haselhuber, Landshut; N. Herrmann, Ansbach; U. John-Grafe, Steinbach/W.; M. Kaiser, Augsburg; H. Lehn, Dachau; M. Mayer, Würzburg; G. Pitz, Kempten/Allgäu; H. Preidel, Olching; P. Schäffler, Baldham/München; H. Schweppe-Nickl, Postbauer-Heng; H.-Ch. Sengespeik, München; Th. Hangen und W.Geltinger, Landshut; A. Angst, Kaufbeuren; L. Boctor, Erlangen; H. Lichtenstern, Pocking; M. Pieringer, Regensburg; H. Reploh, Bad Tölz; A. Vahle, Landshut; G. Börzsönyi, Freising; R. Haas, Rosenheim; K.-H. Leppik und W.Eberhardt, Erlangen; D. Woiczechowski, Tirschenreuth; E. Berz, München; G. Dorn, C. Hager, Ingolstadt; G. Grundherr, Wasserburg/Inn; U. Janssen, Höchberg; Chr. Brückmann, Brannenburg; J. Lussem-Spanel, Dillingen; K. Müller, Bamberg; F. Engelhardt, Nürnberg; W. Pritsch, Kötzting; A. Rudolph, Lichtenfels; I. Tichmann-Schumann und R.Wörnle, München; C. Förster, Neustadt/Cbg.; J. Gaisbauer, Simbach/Inn; I. Hiller, Fürth; U. Lindlbauer-Eisenach, München; T. Kandler und A.Rühl, Nürnberg; R. Benckendorff, Augsburg; J. Helming und H.Singer, Ingolstadt; E. Judex, Regensburg; H. Kollaschinski, Marktredwitz; P. Lautenbach, Herzogenaurach; J. Lehmann, Immenstadt/Allgäu; G. Lysy, Herzogenaurach; A. Neudecker, Mühldorf; R. Schalkhäuser, Ochsenfurt; H. Schilling, Freystadt; H.A. Schmitz, Obing; F. Scholz, Deggendorf; V. Treu, Karlsfeld; G.B. Vit, Schweinfurt; U. Werner-Jung, Kemnath-Stadt; D. Derbacher, Zirndorf; L. Distel, Neustadt/Aisch; U. Klein, Oberaudorf; Ch. Miller, Frauenau; H. Reiniger, Kirchheim-Heimstetten; K.-H. Weigand, R.Müller und U.v.Eichhorn, Deggendorf; A. Busse, Tegernsee; W. Hiemeyer, Kempten/Allgäu; M. Miedaner, Elisabethszell; A. Schneider, Schweinfurt; W. Schuck, Laufach; D. Schweingel, Bayreuth; B. Simon, München; Chr. Sturm, Memmingen; C. Wittermann, Weilheim; W. Stahl, Nürnberg-Eibach; P. Tcherepnine, Roding; W. Müller, Hof; D. Müller-Bühl und R.Ringert, Alzenau; Chr. Dittmann, Mering; S. und I. Habash, Cham; W. Kunz und K.Skrodzki, Forchheim; E. Lippoldmüller, München; J. Mugler, Fürth; M. Schimmer, Hauzenberg; L. Tenderich, Augsburg; U. Heininger, Erlangen; W.-D. Klaiber, Lindau; E. Preissler und W.Theil, Gersthofen; U. Zimmer, Rothenburg o.d.T.; P. Peller, Rosenheim; W. Winter, Dentlein; G. Akinlaja, Schweinfurt; H.-L. Eschenbacher und R.Ulmer, Lauf/Peg.; S. Jobst und H.G.Schatz, Bayreuth; A. und A. Biebl, Barsbüttel; M. Hummel, Irchenrieth; N. Pauly und L.Zimmermann, Aichach; H.-Chr. Kuderna, Aichach; U. Schamberger, Coburg; U. Schreiner, Oberaurach-Unterschleichach; G. Fuhrmann, München; U. Goering, Pegnitz; K.H. Walther, Frankfurt; S.und P. Duttler, Biberbach; R.-P. Garus, Schwabmünchen; W. von Gloeden, Erlangen; R. Schipper, Monheim; J. Mücke, St.Ingbert; M. Schmidt, Nürnberg; F. Abid, Saarbrücken; H. und H.G. Holzner und W.Harr, Königsbronn; A. Schaaff, Eckental-Eschenau; M. Steiner, Saulgau; F.J. Breyer, St.Ingbert; W. Weidner, H.P. Fritz, Weingarten; W. Heffungs, Meckenbeuren; E. Heitz, Bad Waldsee; B. Höhmann, Aalen; K. Jessenberger, Berlin; K.-E. Mai, Tettnang; A. Olischläger, Biberach an der Riβ; B. Seitz, Traben-Trarbach; K.T. Weber, Berlin; A. Ziegler, Augsburg; F. Beer, Markdorf; R. Besser, Kempten; I. Göpfert-Geyer und M.Plieth, Berlin; R. Kratzsch, Nürnberg; F. Maechler, Berlin; Th. Rautenstrauch, Haar/München; K.H. Staudacher, Weingarten; U. Bernsau, Augsburg; P. Brommer, Tann; D. Bulle und W.Raff, Ravensburg; W. Daffner, Nürnberg-Langwasser; G. Grötzinger, Pfaffenhofen; H. Keudel, Unterhaching; R.J. Kühnelt, Berlin; H. Muth, Berlin; W. Pintgen, Geretsried; U.-B. Rupf, Uhlendingen; Ch. Scheurle, Ravensburg; R. Spallek, Bad Wurzach; G. Thomas, München; K.-J. Taube, Berlin; G. Behr-Heinz und S.Heiland, Friedrichshafen; W. Deigendesch, Metzingen; E. Gimpl, Schweinfurt; H. Hilber und A.Freilinger, Au/Hallertau; Th. Morandini, Schönenberg-Kübelberg; D. Schlegel, Welzheim; F.C. Sitzmann, Homburg/Saar; R. Till, Amberg; E. Reiser, Isny/Allgäu; N. Schmidt, Schongau; P. Seidl, Waldkirchen; J. Zimmer, Ehingen; R. Freund, Berlin; H. Litzenbörger, Berlin; D. Lasius, Berlin; E. Dietmair, W.Wagner und P.Wörle, Bobingen; E. Göhre, Berlin; D. Grunert, Nördlingen; M. Rottmann, Stuttgart; E. Schubert, Roth; P. Seyyedi, Lampertheim; A. Steigenberger, Vilsbiburg; J. Gromball und R.Reif, Nürnberg; K. Kirsten, Ettenheim; Chr. Schaefer, Göppingen; L. und J. Viethen, Berchtesgaden; E. Zöller, Idar–Oderstein; D. Kahn, F.Schulze und S.Schartl, Berlin; E. Eidelloth, Lindenberg; A. Renz, Bermatingen; P. Wolff, Pfullendorf; F. Puls, Friedrichshafen; R. Faul, Stuttgart; B. Koch, Ebersberg; W. Conzelmann, Urbach; M. Müller, Ulm; H.-W. Klopp und U.Lorenz, Winterbach; J. Brunnberg, Würzburg; F. Ditlmann, Augsburg; J. Altmann, Immenstadt; K.-W. Weigel, Karlstadt; K.-P. Groβe, Höchstadt/Aisch; P. Treitz, Püttlingen; U. Behre und A.Burgert, Kehl; F. Ladwein, Saarlouis; J. Lauenstein, Lebach; W. Wahlen, M.Büttner und J.Richter, Homburg/Saar; K.Y. Tjhen, Rottenburg; H. Treib, Bous; J. Disselhoff, Offenburg; H.-P. Niedermeier, Erding; G. Stiepani und W.Gröner, Ravensburg; R. Krämer, Illingen; G. Lück-Coerper, Saarlouis; H.-D. Hüwer, Langenselbold; G. Ulbricht und U.Fegeler, Berlin; E. Schrickel, Murnau; M. Walther-Richters, Meitingen; R. Wörrlein und G.Kottsieper, Ansbach; B. Kostadinowa und Th.Unger, Berlin; K.B. Karsten, Dillingen; Pilot study contributors: P. Jakob, Erlangen; W. Kasper, Erlangen.
- Received July 16, 1997.
- Accepted September 18, 1997.
Reprint requests to (J.D.C.) Division of Pediatric Infectious Diseases, UCLA School of Medicine, 10833 Le Conte Ave, Los Angeles, CA 90095-1752.
- ↵Lewis K, Cherry JD, Baraff LJ, et al. An open study evaluating the reactogenicity and immunogenicity of a DTP vaccine containing an acellular pertussis component in four to six year old children. Proceedings of the Fourth International Symposium on Pertussis; Joint IABS/SHO Meeting; Geneva, Switzerland, 1984. 1985;61:563–569
- Pichichero ME,
- Badgett JT,
- Rodgers GC,
- et al.
- Blumberg DA,
- Mink CM,
- Cherry JD,
- et al.
- Cherry JD,
- Mortimer EA,
- Hackell JG,
- and the Multicenter APDT Vaccine Study Groups
- Stehr K,
- Lugauer S,
- Heininger U,
- Uhlenbusch R,
- Cherry JD
- Watson B,
- Cawein A,
- McKee BL,
- and Hackell JG
- Edwards KM,
- Meade BD,
- Decker MD,
- et al.
- Decker MD,
- Edwards KM,
- Steinhoff MC,
- et al.
- ↵Centers for Disease Control. Pertussis vaccination: use of acellular pertussis vaccines among infants and young children: recommendations of the Advisory Committee on Immunization Practices (ACIP). MMWR 1997;46(RR-7):1–25
- ↵Stehr K. Pertussisschutzimpfung. In: Spiess H, ed. Impfkompendium. 4th ed. New York, NY: Georg Thieme Verlag Stuttgart; 1994:143–159
- Stehr K,
- Heininger U
- Finger H,
- Wirsing von König CH,
- Tacken A,
- Wassilak SGF
- ↵Manclark CR, Meade BD, Burstyn DG. Serologic response to Bordetella pertussis. In: Rose NR, Friedman H, Fahey JL, eds. Manual of Clinical Laboratory Immunology. 3rd ed. Washington, DC: American Society for Microbiology; 1986:388–394
- ↵World Health Organization. WHO Meeting on Case Definition of Pertussis: Geneva. January 10–11, 1991. Geneva, Switzerland: World Health Organization; 1991:4–5 (Issue no. MIN/EPI/PERT/91.1)
- Stehr K,
- Cherry JD
- ↵Kalbfleisch JD, Pren KL. The Statisical Analysis of Failure Time Data. New York, NY: John Wiley; 1980
- Schmitt-Grohé S,
- Stehr K,
- Cherry JD,
- et al.
- Überall MA,
- Stehr K,
- Cherry JD,
- et al.
- Hausman S,
- Cherry JD,
- Heininger U,
- Wirsing von König CH,
- Burns DL
- Iwata S,
- Aoyama T,
- Goto A,
- et al.
- Khelef N,
- Danve B,
- Quentin-Millet MJ,
- Guiso N
- Edwards KM,
- Decker MD,
- Halsey NA,
- Koblin BA,
- Townsend T,
- Auerbach B,
- Karzon DT
- ↵Cherry JD. Strategies for diphtheria, tetanus, and pertussis (DTP) immunization. Report of the 104th Ross Conference on Pediatric Research, Strategies for Pediatric Vaccines: Conventional and Molecular Approaches, 13; Columbus, OH; Ross Products Division, Abbott Laboratories; 1994:218–225
- Copyright © 1998 American Academy of Pediatrics