Haemophilus influenzae infections and whooping cough

Bordetella Pertussis Toxin does not induce the release of pro-inflammatory cytokines in human whole blood

Pertussis Toxin (PTx) is one of the most important virulence factors of Bordetella pertussis, the cause of whooping cough. Therefore, the inactivated toxin is an obligatory constituent of acellular pertussis vaccines. It is described in the literature that both native PTx and recombinant Pertussis Toxin (PTg) activate human monocytes whereas others report an inhibition of mammalian monocytes during pertussis infection. B. pertussis, as a Gram-negative bacterium, harbours naturally lipopolysaccharide (LPS, also known as endotoxin), one of the strongest stimulators of monocytes. The latter is triggered via the interaction of endotoxin with inter alia the surface receptor CD14. Consequently, it is necessary to consider a potential contamination of Pertussis Toxin preparations with LPS. First, we determined the LPS content in different preparations of PTx and PTg. All preparations examined were contaminated with LPS; therefore, possible PTx- and PTg-driven monocyte activation independently of LPS was investigated. To meet these aims, we examined monocyte response to PTx and PTg while blocking the LPS receptor CD14 with a specific monoclonal antibody (anti-CD14 mAb). In addition, all toxin preparations examined underwent an LPS depletion. Our results show that it is contaminating LPS, not Pertussis Toxin, which activates human monocytes. Blocking the CD14 receptor prevents Pertussis Toxin-mediated induction of pro-inflammatory cytokines in human monocytes. The depletion of LPS from Pertussis Toxin leads to the same effect. Additionally, the PTx toxicity after LPS depletion procedure was confirmed by animal tests. In contrast, the original Pertussis Toxin preparations not treated as mentioned above generate strong monocyte activation. The results in this publication allow the conclusion that purified Pertussis Toxin preparations do not induce the release of pro-inflammatory cytokines in human whole blood.

The role of the adsorption process for production and control combined adsorbed vaccines

For the production and control of adsorbed vaccines it is necessary to know and respect some basic features of adsorption. Among others these are the surface charges of adsorbents and antigens, the pH of the medium, the buffer components and the maturation process. They are important for the adsorption behaviour and have influence on the stability of the adsorbate and potency of the vaccine. Individually different conditions for optimal adsorption of antigens can be respected with a production strategy, where every antigen is adsorbed separately in a controlled manner, before it is mixed with other antigens to the final bulk vaccine. This strategy is favoured over strategies with sequential or competitive adsorption of the antigens.

Active immunization in the United States: developments over the past decade

The Centers for Disease Control and Prevention has identified immunization as the most important public health advance of the 20th century. The purpose of this article is to review the changes that have taken place in active immunization in the United States over the past decade. Since 1990, new vaccines have become available to prevent five infectious diseases: varicella, rotavirus, hepatitis A, Lyme disease, and Japanese encephalitis virus infection. Improved vaccines have been developed to prevent Haemophilus influenzae type b, pneumococcus, pertussis, rabies, and typhoid infections. Immunization strategies for the prevention of hepatitis B, measles, meningococcal infections, and poliomyelitis have changed as a result of the changing epidemiology of these diseases. Combination vaccines are being developed to facilitate the delivery of multiple antigens, and improved vaccines are under development for cholera, influenza, and meningococcal disease. Major advances in molecular biology have enabled scientists to devise new approaches to the development of vaccines against diseases ranging from respiratory viral to enteric bacterial infections that continue to plague the world's population.

[Non-antibiotic treatments for community-acquired pneumonia in children]

Besides antibiotic therapy (when warranted), treatment of acute pneumonia in children usually requires only simple measures, including: fractionated feeding, analgesic and antipyretic medications and detection of hypoxemia. There is no evidence that physical chest therapy, bronchial mucus modificators and corticosteroid therapy are really useful. The best prevention relies upon vaccinations, suppression of passive smoking, and hygiene and nutritional improvements.

Novel molecular biology approaches to acellular vaccines

Bacterial toxins are commonly detoxified by chemical treatment in order to use them in human vaccines. We have used site-directed mutagenesis of toxin genes to obtain bacteria that produce naturally nontoxic mutants of bacterial toxins, such as pertussis toxin (PT), cholera toxin (CT) and Escherichia coli heat-labile enterotoxin (LT). Genetically detoxified PT showed a superior safety and immunogenicity in animal models, phase I and phase II clinical trials, and a superior protective efficacy in the early and late stage of a phase III efficacy trial, proving in a definitive and extensive way that genetic detoxification of bacterial toxins can, and should, replace chemical treatment. The results obtained with genetically inactivated LT and CT indicate that genetic detoxification of bacterial toxins can be used not only to produce vaccines for systemic immunization that are superior to the ones produced by conventional technologies, but suggest that these type of molecules may be the prototype molecules for the design and construction of innovative vaccines with a totally new design, such as mucosally delivered preventive and therapeutic vaccines.

A controlled trial of a two-component acellular, a five-component acellular, and a whole-cell pertussis vaccine

BACKGROUND

Because of concern about safety and efficacy, no pertussis vaccine has been included in the vaccination program in Sweden since 1979. To provide data that might permit the reintroduction of a pertussis vaccine, we conducted a placebo-controlled trial of two acellular and one whole-cell pertussis vaccines.

METHODS

After informed consent was obtained, 9829 children born in 1992 were randomly assigned to receive one of four vaccines: a two-component acellular diphtheria-tetanus-pertussis (DTP) vaccine (2566 children), a five-component acellular DTP vaccine (2587 children), a whole-cell DTP vaccine licensed in the United States (2102 children), or (as a control) a vaccine containing diphtheria and tetanus toxoids (DT) alone (2574 children). The vaccines were given at 2, 4, and 6 months of age, and the children were then followed for signs of pertussis for an additional 2 years (to a mean age of 21/2 years).

RESULTS

The whole-cell vaccine was associated with significantly higher rates of protracted crying, cyanosis, fever, and local reactions than the other three vaccines. The rates of adverse events were similar for the acellular vaccines and the control DT vaccine. After three doses, the efficacy of the vaccines with respect to pertussis linked to a laboratory-confirmed case of pertussis or contact with an infected household member with paroxysmal cough for > or = 21 days was 58.9 percent for the two-component vaccine (95 percent confidence interval, 50.9 to 65.9 percent), 85.2 percent for the five-component vaccine (95 percent confidence interval, 80.6 to 88.8 percent), and 48.3 percent for the whole-cell vaccine (95 percent confidence interval, 37.0 to 57.6 percent).

CONCLUSIONS

The five-component acellular pertussis vaccine we evaluated can be recommended for general use, since it has a favorable safety profile and confers sustained protection against pertussis. The two-component acellular vaccine and the whole-cell vaccine were less efficacious.

Safety and immunogenicity of two acellular pertussis vaccines with different pertussis toxoid and filamentous hemagglutinin content in infants 2-6 months old

The optimal composition and antigen content of acellular pertussis vaccines is not known. Two vaccines with different quantities of pertussis toxoid (10 and 20 micrograms) and filamentous hemagglutinin (5 and 20 micrograms) and identical 69 kD protein (3 micrograms) and fimbriae 2 and 3 (5 micrograms) combined with diphtheria and tetanus toxoids were compared in a randomized, double-blind study in 2,050 infants undergoing their primary immunization series at 8 centers in the US and Canada. A 6:1 increased antigen to lower antigen allocation was used; 96% of infants received 3 doses and completed the study. A 'clinically significant' local reaction was reported in 3-6% of participants after each dose. Erythema was the most common reaction occurring in 3-5% of infants after the second or third dose. A clinically significant systemic adverse reaction was reported in 28-34% of vaccinees (or vaccinated children) after each dose; fever (7-18%) and fussiness (12-17%) were most common. There were no differences in adverse events between the 2 vaccine formulations. Antibody responses were measured in 292 infants at 1 center. At 7 months, geometric mean anti-filamentous hemagglutinin antibody titers were higher in recipients of the higher antigen content vaccine (p < 0.001) whereas recipients of the lower antigen content formulation had higher anti-fimbriae antibody (p < 0.001) and agglutinin titers (p < 0.05). No differences were detected in anti-pertussis toxin or other antibody responses between the formulations. We conclude that increasing the antigen content of the acellular pertussis vaccine had a variable effect on antibody response but was not associated with increased adverse reactions.

Modulation of hepatic mRNA levels after administration of lipopolysaccharide and diphtheria and tetanus toxoids and pertussis vaccine adsorbed (DTP vaccine) to mice

Administration of whole-cell diphtheria and tetanus toxoids and pertussis vaccine adsorbed (DTP vaccine) caused marked depression in the expression of mRNA for isozymes of cytochrome P-450 in the livers of endotoxin-responsive and nonresponsive mice. The levels of expression of mRNA for a polycyclic aromatic hydrocarbon-inducible (CYP1A2) and an ethanol-inducible (CYP2E1) form of P-450 were reduced by 70% to 80% 8 to 12 hr after vaccination or Bordetella pertussis endotoxin administration. These effects are preceded by marked increases (threefold to sixfold) in mRNA expression for interleukin-6, interleukin-1 and tumor necrosis factor in both strains of mice, with maximal increases 1 to 2 hr after injection. This is the first demonstration that levels of cytokine mRNA are altered in the liver in response to DTP vaccine administration. The finding of increased cytokine mRNA in the livers of mice injected with vaccine supports a role for cytokines as mediators of the decreased levels of cytochrome P-450. In addition, inducible nitric oxide synthase mRNA expression is also increased after vaccine administration, with a peak at 4 hr. The temporal relationship of the increased cytokine mRNA expression, increased nitric oxide synthase and decreased expression of P-450 mRNAs suggests a mechanism by which cytokines mediate the induction of nitric oxide synthase, which increases nitric oxide and decreases the activities of some cytochromes P-450.

Comparative study of a whole-cell pertussis vaccine and a recombinant acellular pertussis vaccine. The Italian Multicenter Group for the Study of Recombinant Acellular Pertussis Vaccine

The safety and immunogenicity of an acellular pertussis vaccine containing the genetically detoxified pertussis toxin PT-9K/129G, filamentous hemagglutinin, and pertactin, together with diphtheria and tetanus toxoids, were compared with those of a whole-cell pertussis component-diphtheria-tetanus vaccine. Four hundred eighty infants were enrolled into this prospective, multicenter, double-blind study. Each infant was randomly given three doses of one of the two vaccines at 2, 4, and 6 months of age. Both local and systemic adverse reactions, reported within 48 hours and 7 days of each injection, were less frequent after the acellular vaccine than after the whole-cell vaccine. The enzyme-linked immunosorbent assay titers to pertussis toxin, filamentous hemagglutinin, and pertactin, as well as the pertussis toxin-neutralizing titer measured by the Chinese hamster ovary cell assay, were significantly higher after the acellular vaccine was given. Both vaccines induced adequate levels of anti-diphtheria and anti-tetanus antibodies. We conclude that the recombinant acellular pertussis vaccine produces fewer reactions than the whole-cell vaccine and provides a high antibody response against the antigens of Bordetella pertussis involved in bacterial adhesion and systemic toxic effects.

Successful intravenous immunoglobulin therapy for recurrent pneumococcal otitis media in young children

Serum immunoglobulin levels and naturally occurring antibody titres against Streptococcus pneumoniae were measured in seven children aged 1-1.9 years with recurrent pneumococcal acute otitis media (AOM). Three of them had low IgG2 levels. Mean antibody levels of anti-pneumococcal IgG1 and anti-pneumococcal IgG2 were significantly lower in patients when compared to those of healthy controls and children who had less frequent episodes of AOM. Following treatment with intravenous immunoglobulin (IVIG) for 6 months, anti-pneumococcal IgG1 and IgG2 antibody levels increased and the number of episodes of AOM decreased in all patients. Following the discontinuation of IVIG therapy, no AOM episode occurred. Serum levels of anti-pneumococcal IgG1 and IgG2 were normal, which were measured in three subjects at 5, 6, and 12 months after the cessation of IVIG therapy. These results suggested that delayed maturation of anti-pneumococcal antibody production caused recurrent AOM and this condition was corrected by IVIG therapy.

Engineered human vaccines

The limitations of human vaccines in use at present and the design requirements for a new generation of human vaccines are discussed. The progress in engineering of human vaccines for bacteria, viruses, parasites, and cancer is reviewed, and the data from human studies with the engineered vaccines are discussed, especially for cancer and AIDS vaccines. The final section of the review deals with the possible future developments in the field of engineered human vaccines and the requirement for effective new human adjuvants.

Epidemiology of invasive Haemophilus influenzae infections in Cape Town, South Africa

The full spectrum of invasive Haemophilus influenzae disease has not been documented previously in Africa. This 1-year prospective study was designed to determine the epidemiology of invasive Haemophilus influenzae disease in Cape Town children. During this period, 142 children with invasive disease were hospitalized; 85 (59.9%) presented with meningitis, 35 (24.6%) with pneumonia and 22 (15.5%) with other diseases. No cases of epiglottitis were seen. Sixty per cent of cases were male and 40% female. The median age of the children was 9 months, with a range of 1-144 months, and 65.5% were aged < 12 months. Neurological dysfunction was noted in 40% and 18% of children with meningitis on admission and discharge, respectively. The overall case fatality rate (95% confidence intervals) was 9.2% (4.9-15.7), and for meningitis, pneumonia and septicaemia it was 4.7% (1.2-16.4), 14.3% (4.6-31.8) and 40% (8-78.1), respectively. Serotype b accounted for 86.5% of all cases, 97.3% of cases of meningitis, 71.4% of cases of pneumonia, 50% of cases of septicaemia, all cases of arthritis and cellulitis and none of mastoiditis. The incidence rates (95% confidence intervals) for all invasive type b infections were 169 (122-198) and 47 (39-57) per 100,000 population for children < 1 and < 5 years, respectively. For meningitis the rates were 112 (84-148) and 34 (25-40) per 100,000, respectively. Rates for mixed race and white children were similar, but those for black children were more than double those rates.(ABSTRACT TRUNCATED AT 250 WORDS)

Pertussis toxin-induced alterations of murine hepatic drug metabolism following administration of diphtheria and tetanus toxoids and pertussis vaccine adsorbed

Administration of pertussis toxin (PT) in combination with diphtheria and tetanus toxoids adsorbed (DT vaccine) or with acellular pertussis vaccine adsorbed and diphtheria and tetanus toxoids (APDT) elicits dose- and time-dependent alterations in hepatic drug metabolism in mice. Cytochrome P-450 (P-450) levels were inhibited more than 50% at 7 days following a single injection of PT mixed with either vaccine. When combined with DT vaccine, 125 ng of PT was required to produce this effect, while as little as 16 ng of PT combined with APDT vaccine inhibited P-450 levels. The inhibition of P-450 levels is similar to that observed after a single injection of diphtheria and tetanus toxoids and pertussis vaccine adsorbed (DTP). Alterations of P-450 levels were accompanied by increased activities of quinone reductase but not with changes in plasma interleukin-6 or tumor necrosis factor levels. Other Bordetella pertussis virulence factors, such as filamentous hemagglutinin, fimbriae and pertactin, were also tested but had no significant effect on hepatic drug metabolism. Endotoxin or preparations containing endotoxin caused alterations in hepatic drug metabolism within 24 h, concomitant with increased interleukin-6 and tumor necrosis factor levels, but these effects had resolved by 1 week. DTP vaccine and preparations containing PT caused a marked induction of gamma interferon coincident with the maximal inhibition of P-450 levels. This effect was not present with DT or APDT vaccine alone, nor with endotoxin or any combination of factors that did not contain PT. These results demonstrate that PT is a necessary component for the sustained effects of DTP vaccine on hepatic drug metabolism and suggest a role for gamma interferon in this process.

Environmental regulation of virulence factors in Bordetella species

Many bacteria respond in a coordinate manner to environmental changes. External stimuli, sensed by receptors, are transduced to regulatory proteins which participate in well defined pathways of gene expression by varying their structure and mode of action. The network of environmental signal transduction is responsible for a fine and continuous communication between the host and the pathogenic bacteria. As a result, the gene expression machinery of the pathogen is modified continuously, in order to establish the optimal conditions for bacterial survival and multiplication.

Recombinant acellular pertussis vaccine--from the laboratory to the clinic: improving the quality of the immune response

Vaccination is the most effective way to prevent infectious diseases. Recombinant DNA technologies have provided powerful new tools to develop vaccines that were previously impossible or difficult to make, and to improve the vaccines that were already available but had been developed using old technology. In the case of whooping cough, an effective vaccine (composed of killed bacterial cells) is available, but its use is controversial because of the many side effects that have been associated with it. An improved vaccine against this disease should contain pertussis toxin, a molecule that needs to be detoxified in order to be included in the vaccine. Classical methods of detoxification, such as formaldehyde treatment have been used to inactivate this toxin. We have used recombinant DNA technologies to clone the pertussis toxin gene, express it in bacteria, map the B and T cell epitopes of the molecule, and to identify the amino acids that are important for enzymatic activity and toxicity. Finally, we have used this information to mutate the gene in the chromosome of Bordetella pertussis in order to obtain a strain that produces a molecule that is already non-toxic. This genetically inactivated pertussis toxin was tested extensively in animal models and clinical trials and was found to induce an immune response that is superior in quality and quantity to that induced by the vaccines produced by conventional technologies.

Acellular pertussis vaccine composed of genetically inactivated pertussis toxin: safety and immunogenicity in 12- to 24- and 2- to 4-month-old children

To determine whether a nontoxic derivative of pertussis toxin obtained by recombinant DNA technology, PT-9K/129G, is a good candidate for a new pertussis vaccine, we examined the safety and the immunogenicity in children of a vaccine containing 15 micrograms of PT-9K/129G protein and 0.5 mg of aluminum hydroxide per dose. Fifty-three children 12 to 24 months of age and 21 infants aged 2 to 4 months were injected with two and three doses, respectively. The vaccine did not induce significant local or systemic reactions and elicited an increase of antibody titer in more than 98% of the children. The geometric mean of the toxin-neutralizing titers increased after each dose and was 85 units in children given two doses and 196 units in those given three doses. Two children who had detectable antibody levels before the first immunization had a high response (greater than 320 units) to the first vaccine dose. The findings suggest that PT-9K/129G is a promising antigen to be included in the development of acellular pertussis vaccines.

Development and clinical testing of an acellular pertussis vaccine containing genetically detoxified pertussis toxin

In 1924 Ramon described the inactivation of diphtheria toxin by formaldehyde treatment. This method allowed the introduction of mass vaccination against diphtheria and tetanus and opened the way to the inactivation of viruses by chemical treatment. In this review we describe the use of genetic manipulations for the inactivation of pertussis toxin. The toxin inactivated by this new method is an antigen superior to those obtained by chemical treatment and has been used to develop a new vaccine against whooping cough.

Progress towards the development of new vaccines against whooping cough

Acellular vaccines against whooping cough are in the final stage of clinical testing and are likely to become available for mass immunization in the near future. Over a dozen vaccines of similar composition have been developed by vaccine companies and research laboratories; all of them contain a detoxified form of pertussis toxin (PT) that may be present alone or combined with one or more other non-toxic proteins, such as filamentous haemagglutinin (FHA), pertactin (69 kDa), and the agglutinogens (AGG). Most of the vaccines contain a PT that has been inactivated by chemical treatment, a process that reduces the immunogenicity of the molecule and may not completely eliminate the risk of reversion to toxicity. To avoid these problems, we have constructed by genetic manipulation a mutant of Bordetella pertussis that produces a non-toxic form of PT. This molecule (PT-9K/129G) contains two amino acid substitutions in the S1 subunit (Arg9-->Lys and Glu129-->Gly) which abolish the enzymatic activity of the S1 subunit and all the toxic properties of PT, without changing the immunological properties of the wild-type toxin. Following extensive preclinical studies, which have shown that PT-9K/129G is safe and more antigenic than the toxin treated with chemical agents, this molecule was tested for safety and immunogenicity in adult volunteers, 18-month-old children and 2-month-old infants. The molecule has been tested alone, combined with FHA and pertactin and also combined with diphtheria and tetanus toxoids.(ABSTRACT TRUNCATED AT 250 WORDS)

An idiotypic marker associated with a germ-line encoded kappa light chain variable region that predominates the vaccine-induced human antibody response to the Haemophilus influenzae b polysaccharide

Human antibodies specific for the Haemophilus influenzae b polysaccharide (Hib PS) frequently express a cross-reactive idiotype (CRI), and commonly utilize a VL region that is the product of the V kappa II gene A2. To examine further anti-Hib PS V region expression and to determine whether CRI expression is correlated with the V kappa IIA2 chain, we isolated a monoclonal antibody (MAb) reactive with an idiotypic determinant of anti-Hib PS antibodies. This MAb inhibited Hib PS binding but did not react with Ig isotypic determinants. The CRI recognized by this MAb, designated HibId-1, was associated with the Hib PS-combining site since the reactivity of the MAb with anti-Hib PS antibodies could be inhibited by Hib PS. HibId-1 was expressed by 17 of 17 clonally purified and sequence-defined anti-Hib PS antibodies having V kappa IIA2 L chains. In contrast, 0 of 10 anti-Hib PS antibodies having either V lambda, V kappa I, or V kappa III chains expressed HibId-1. Western blot analysis showed that the MAb anti-CRI reacted with isolated anti-Hib PS V kappa IIA2 L chains but not with H chains or other L chains, indicating that the HibId-1 determinant is localized to the V kappa IIA2 chain, and does not require pairing with H chain for expression. Anti-Hib PS antibodies bearing HibId-1 were present in at least 85% of subjects immunized with either free Hib PS or Hib PS coupled to diphtheria toxoid (Hib PS-DT), and comprised on the average 60% of the total vaccine-induced serum anti-Hib PS. HibId-1 expression was not related to age at vaccination inasmuch as infants, children, and adults had similar distributions of HibId-1-positive anti-Hib PS after vaccination with Hib PS-DT. HibId-1 was expressed at a lower frequency and comprised a smaller fraction of the total anti-Hib PS antibody in adult preimmunization sera as compared to post-Hib PS immunization sera, suggesting that immunization preferentially stimulates HibId-1-positive B cells. These data demonstrate that antibodies bearing HibId-1/V kappa IIA2 comprise a predominant component of the anti-Hib PS response induced by immunization, and that this pattern of VL expression is established early in ontogeny.

Lymphokine secretion and cytotoxic activity of human CD4+ T-cell clones against Bordetella pertussis

Human CD4+ T-cell clones specific for pertussis toxin and other Bordetella pertussis antigens have been tested for their cytotoxic activity, lymphokine production, and capacity to induce immunoglobulin synthesis. Clones specific for the S1 subunit of pertussis toxin were cytotoxic for autologous Epstein-Barr virus-transformed B cells, which had been pulsed with the native antigen, the recombinant S1 subunit of pertussis toxin, or synthetic peptides derived from the S1 amino acid sequence. The killing of antigen-pulsed target cells was class II restricted. All of the T-cell clones produced mostly interleukin-2 and gamma interferon and assisted allogeneic B cells in the production of immunoglobulins M and G but not immunoglobulin E. The potential in vivo role of the cytotoxic activity of these clones is discussed.

Phase I clinical trial of an acellular pertussis vaccine composed of genetically detoxified pertussis toxin combined with FHA and 69 kDa

An acellular pertussis vaccine composed of genetically detoxified pertussis toxin (PT-9K/129G), filamentous haemagglutinin (FHA) and pertactin (69 kDa protein) was evaluated in adult volunteers, in double blind, versus placebo. No fever was reported in either group. Mild local reactions were reported after injection of both vaccine and placebo. After the first dose a marked increase in antibodies to PT, FHA and 69 kDa protein was seen in vaccinated subjects with the exception of one who responded well to PT and FHA but did not show a humoral response to the 69 kDa protein. All vaccinees acquired cellular immunity against the three antigens. No significant variation was observed in the humoral or cellular responses after the second dose.

Identification of subregions of Bordetella pertussis filamentous hemagglutinin that stimulate human T-cell responses

Filamentous hemagglutinin (FHA), a 220-kDa protein that mediates the adhesion of Bordetella pertussis to eukaryotic cells, is a component of acellular vaccines against whooping cough. To identify the subregions of FHA that are immunogenic for T cells, 16 human T-cell clones were raised against purified FHA and tested for the recognition of recombinant and proteolytic fragments. The clones were found to map either in the carboxy-terminal or the amino-terminal part of the FHA molecule, but none of them recognized the central region, which contains a sequence that is homologous to that of the eukaryotic protein fibronectin. These data suggest that subregions of FHA that do not contain sequences that are potentially cross-reactive with self proteins may be sufficient to induce an immune response against the whole protein.

Phase II trial of whole-cell pertussis vaccine vs an acellular vaccine containing agglutinogens

An acellular pertussis vaccine containing agglutinogens 2 and 3, pertussis toxin, and filamentous haemagglutinin was developed by the Centre for Applied Microbiology and Research in the UK. 188 infants were entered into a randomised blind trial and received either the acellular or a whole-cell vaccine, combined with diphtheria and tetanus toxoids, in a 3, 5, and 8-10 month schedule. Local reactions were similar in the two groups but significantly fewer infants had systemic symptoms after the acellular vaccine. Mean log-antibody titres to the agglutinogen and toxin components were higher with the acellular than with the whole-cell vaccine. Persistence of antibodies one year after the third dose was also better in the acellular group.