Pertussinum Anhang

[F.G.A.Versteegh]

Pertussis: new insights in diagnosis, incidence and clinical manifestations

THE DELIVERY OF MEDICAL CARE IS TO DO AS MUCH NOTHING AS POSSIBLE

Samuel Shem MD: The House of God

Mijn dokter

mijn dokter is

een goede dokter

hij heeft mij niet

zieker gemaakt

Willem Hussem

Maar allerdiepst op Curaçao

Treft mij de taal, de gang, de lach,

De ongedwongen oogopslag

Van iedere man en elke vrouw

Anton van Duinkerken: uit: Gesprek in Punda

Pertussis (whooping cough) is a highly contagious acute bacterial disease involving the respiratory tract and is caused mainly by Bordetella (B.) pertussis and to a lesser extent by B.parapertussis. It is most severe in young infants. It has a worldwide prevalence and occurs in all age groups.

History

The history of whooping cough starts, according to the literature, with the description by Guillaume de Baillou (1538-1616) of an epidemic in 1578 in France, published for the first time only in 1640 by his nephew. Was the disease not known before, or known maybe by other names and in different countries by different names?

Kohn suggests that the description of the Perinthus cough by Hippocrates (around 400 B.C.) might possibly be whooping cough or a mix with other diseases such as viral respiratory infections.

In the Oxford English Dictionary kinkehost is mentioned in Reginald’s Vita Godrici from around 1190. In the “Middelnederlandsch woordenboek” (dictionary of medieval Dutch) it is suggested that gisschen might be an early eastern Dutch word for whooping cough, used in the first half of the 14th century. In his History of Pediatrics in the Netherlands van Lieburg refers to the Miracle book of the

St Jan’s cathedral in ‘s Hertogenbosch, in the southern part of the Netherlands, in which a pilgrimage is described to the statue of the holy Mary because of the recovery of a boy from kychoest, in 1383. Nils Rosen von Rosenstein from Sweden states, not knowing when the disease came to his country, that in France it first appeared in 1414, without giving a source. In Schiller-Lübben’s “Mittelniederdeutsches Wörterbuch” (dictionary of medieval German) kinkhoste is found in a source from 1464. Dodonaeus (1517-1585) in his “Cruijde boeck” (book of herbs) already in 1554 describes cures for the kieckhoest!

De Baillou called it quinta, referring to Hippocrates. Coqueluche, the present name in French for whooping cough, was then the common name for influenza. Holmes reports that pertussis was called

chyne-cough in England as early as 1519. Cherry and Heininger say it was called the kink (in Scottish synonymous with fit or paroxysm) and kindhoest (a Teutonic word meaning child’s cough)

in the Middle Ages. Nils Rosen von Rosenstein calls

it in his book about pediatric diseases from 1798 Keichhussten. In a JAMA editorial names as tosse canina (dog’s bark, Italy), Wolfshusten (howling of wolves) and Eselshusten (braying of donkeys)

(both from Germany) and chincough (boisterous laughter, Old English) are given. In Chinese it is called “cough of 100 days”. In Dutch it is called kinkhoest, coming from old names as kinkhôste, kichhoest, keichhusten, as van Esso describes. In the Dictionary of the Dutch Language (Woordenboek der Nederlandsche Taal) the same names are given and others as kie(c)khoest, kijkhoest,

kikhoest. Also it refers to Dodonaeus (1517-1585) who called the disease kich, or kinchoest in a latter edition of his “Cruydt-Boeck” from 1608. Since in these old days there were no possibilities to prove the diagnosis we will never know whether all these diseases then were the same as our whooping cough that, as we know today, is caused by B.pertussis, or that they were pertussis-like syndromes,

caused by one or more other pathogens.

Clinical Manifestations

Clinical manifestations of whooping cough may show substantial variation depending on previous vaccination, earlier infection with B.pertussis, age or the clinical condition of the patient.

The clinical course is divided into 3 stages. After an incubation period of 5 to 10 days, with an upper limit of 21 days, illness begins with the catarrhal phase. This phase lasts 1 to 2 weeks and

is usually characterized by low-grade fever, rhinorrhea and progressive cough.

In the subsequent paroxysmal phase, lasting several weeks, B.pertussis causes severe and spasmodic cough episodes with a characteristic whoop, often with cyanosis and vomiting. The patient usually appears normal between attacks. Paroxysmal attacks occur more frequently at night with an average of 15 attacks per 24 hours. During the first 1 or 2 weeks of this stage the attacks increase in frequency, then remain at the same level for 2 to 3 weeks and then gradually decrease. The paroxysmal stage usually lasts 1 to 6 weeks but may persist for up to 10 weeks.

Young infants (under 6 months of age) may not have the strength to have a whoop but they do have paroxysms of coughing. The cough though may be absent and disease may then manifest itself with spells of apnoea.

Although pertussis may occur at any age, most cases of serious disease and the majority of fatalities are observed in early infancy. The most important complications in the USA are hospitalization

(72.2% in children younger than 6 months, 3.9% for those over 20 years of age), bronchopneumonia (17.3% vs. 3.4%), seizures (2.1% vs. 0.5%), acute encephalopathy (0.5% vs. 0.1%), the latter frequently resulting in death or lifelong brain damage and death (0.5% vs. 0). Heininger reported in proven pertussis patients in Germany an overall complication rate of 5.8%, pneumonia (29%) being the most frequent complication. In infants <6 months of age, the rate of complications was 23.8%.

At the end of the catarrhal phase, a leukocytosis with an absolute and relative lymphocytosis frequently begins reaching its peak at the height of the paroxysmal stage. At this time, the total blood

leukocyte levels may resemble those of leukemia (> 100,000/mm3), with 60 to 80 % lymphocytes.

The convalescent phase, the last stage, lasting 1 to 3 weeks, is characterized by a gradual, continuous decline of the cough before the patient returns to normal. However, paroxysms often recur with subsequent respiratory infections for many months after the onset of pertussis. Fever is generally minimal throughout the course of pertussis.

Microbiology

The genus Bordetella contains species of related bacteria with similar morphology, size and staining reactions.

To date there are 8 species known of Bordetella:

B.pertussis,

B.parapertussis 28,29,

B.bronchiseptica,

B.avium, (formerly designated Alcaligenes faecalis),

B.hinzii 32,33 (formerly designated A.faecalis type II),

B.holmesii,

B.trematum

B.petrii.

B.pertussis, B.parapertussis and B.bronchiseptica are genomically closely related.

The first four are respiratory pathogens. B.pertussis is an obligate human pathogen. B.pertussis was long considered the sole agent of whooping cough. A mild, pertussis-like disease in humans may be caused by B.parapertussis and occasionally by B.bronchiseptica. B.parapertussis appears both in humans and animals. The natural habitat of B.bronchiseptica is the respiratory tract of smaller animals (rabbits, cats and dogs). Human infections with B.bronchiseptica are rare and occur only after close contact with carrier animals, no human to human transmission occurs.

Most patients with severe disease by B.bronchiseptica have an (immuno-)compromised clinical status. B.avium and B.hinzii are important in birds. B.hinzii and B.holmesii are found in blood cultures

from immune compromised patients. B.trematum and B.petrii have been recently discovered, B.trematum in wounds in humans, B.petrii (an anaerobic species) in a bioreactor.

B.pertussis is a small (approximately 0.8 μm by 0.4 μm), rod-shaped, or coccoid, or ovoid Gramnegative bacterium that is encapsulated and does not produce spores. It is a strict aerobe. It is arranged

singly or in small groups and is not easily distinguished from Haemophilus species.

B.pertussis and B.parapertussis are nonmotile.

Bacteriological confirmation of suspected whooping cough is often missed, as culturable B.pertussis does not seem to persist far beyond the catarrhal stage and in addition requires special growth factors

to grow on artificial media.

B.pertussis, the causative agent of pertussis with affinity to the mucosal layers of the human respiratory tract, has different antigenic or biologically active components, although their exact chemical structure and location in the bacterial cell are known only in part.

Pathogenesis

Infection results in colonization and rapid multiplication of the bacteria on the mucous membranes of the respiratory tract. It produces a number of virulence factors, which comprise pertussis toxin, adenylate cyclase toxin, filamentous haemagglutinin, fimbriae, tracheal cytotoxin, pertactin and dermonecrotic toxin. The expression of these factors is regulated by the bvg locus 39,40. This system assures that the organism synthesizes components only in response to certain environmental stimuli. Bacteremia does not occur. Studies of the different B.pertussis adhesion molecules and toxins and their corresponding biologic activities have yielded plausible explanations for many of the symptoms of whooping cough. In humans, an initial local peribronchial lymphoid hyperplasia occurs accompanied

or followed by necrotizing inflammation and leukocyte infiltration in parts of the larynx, trachea and bronchi. Usually, peribronchiolitis and variable patterns of atelectasis and emphysema also develop.

To date, there is no possible explanation for the development of the characteristic paroxysmal coughing in pertussis.

Pertussis toxin is assumed to be one essential protective immunogen but numerous findings indicate that other components, such as filamentous hemagglutinin, heat-labile toxin, agglutinogens, outer membrane proteins and adenylate cyclase toxin, may also contribute to immunity after infection or vaccination 44-46. In addition, it was recently shown that antibodies to pertactin but not to pertussis toxin, fimbriae, or filamentous hemagglutinin, are crucial for phagocytosis of B.pertussis 46a. The immunogenicity of these substances may be significantly increased by the presence of pertussis toxin.

This synergism indicates that pertussis toxin could function as an adjuvant to a variety of protective antigens of B.pertussis. The defense mechanisms are both nonspecific (local inflammation, increase

in macrophage activity and production of interferon) and specific (proliferation of specific B and T cells).

The nature of immunity in whooping cough is, however, incompletely understood. A role of circulating antibody in immunity is indicated by the correlation between protection of human vaccinees and their antibody titers 44-46. However, effective immunity does not necessarily depend on the presence of protective antibodies and immunity to whooping cough may therefore be mediated essentially

by cellular mechanisms 48,49. This cell-mediated immunity may be considered the crucial carrier of long-term immunity and titers of specific humoral antibodies may diminish over the years.

Transmission

In most countries B.pertussis is endemic with superimposed epidemic cycles. These cycles occur approximately every 4 years in vaccinated populations and approximately every 2 to 3 years in non vaccinated populations, although in the Netherlands the incidence is higher now (every 2-3 years) compared to the period prior to the epidemic in 1996-1997 (every 4 years) 51,61. Most infections occur from July to October. Pertussis is very contagious. It is transmitted obligatory from human to human by direct contact with discharges from respiratory mucous membranes of infected persons primarily

via droplets by the airborne route. The mucous membranes of the human respiratory tract are the natural habitat for B.pertussis and B.parapertussis. Most infections occur after direct contact with diseased persons, specifically by inhalation of bacteria-bearing droplets expelled in cough spray. The patient is most infectious during the early catarrhal phase, when clinical symptoms are relatively mild and noncharacteristic. Subclinical cases may have similar epidemiologic significance.

Healthy transient carriers of B.pertussis or B.parapertussis are assumed to play no significant epidemiologic role. Chronic carriage by humans is not documented.

Immunity

Pertussis infection or vaccination results in a long-lasting but not necessarily lifelong protection against the typical clinical manifestations of the disease, or reinfection. The protection may not be complete, as atypical or unrecognized infection in presumably immune persons, particularly adults, may be easily overlooked. Also, newborn babies of mothers who have had pertussis are not necessarily protected. Hence, following previous infection, occasional exposure to B.pertussis strains circulating in the community may be required to sustain high level immunity. Although the level of antibodies to pertussis toxin, pertactin or filamentous hemagglutinin are sometimes used as serological indicators of protection 44,45, lack of generally accepted correlates of immunity and animal models are impediments to the evaluation of new pertussis vaccine candidates and the monitoring of the consistency of production.

Although vaccination has caused a firm decrease in incidence and mortality over the years, occasional local epidemics do occur. The disease is especially dangerous in the first 6 months of life. There seems no distinct influence of the season or climate on the morbidity rate but there may be an increase in the summer and fall. Older persons (i.e., adolescents and adults) and those partially protected by the vaccine may become infected with B.pertussis but usually have milder or asymptomatic disease. Pertussis in these persons may present as a persistent (>7 days) cough and may be indistinguishable from other upper respiratory infections. Inspiratory whoop is uncommon.

In some studies, evidence for a B.pertussis infection was found in 25% or more of adults with cough illness lasting >7 days 63,64. Even though the disease may be milder in older persons, these infected persons may transmit the disease to other susceptible persons, including unimmunized or under immunized infants. Adults often found to be the first case in a household with multiple pertussis cases 65,66.

Diagnosis

Whooping cough is a clinical diagnosis according to WHO criteria 67 as established in 2000:

a case diagnosed by a physician, or a person with a cough lasting at least 2 weeks with at least one of the following symptoms: paroxysms (i.e. fits) of coughing, inspiratory “whooping” or post-tussive vomiting (i.e. vomiting immediately after coughing) without other apparent cause. Criteria for laboratory confirmation are: isolation of B.pertussis or detection of genomic sequences by polymerase chain reaction (PCR) or positive paired serology (i.e. fourfold increase).

Only since Bordet and Gengou in 1906 cultured B.pertussis we could be sure about the diagnosis. Recovery of B.pertussis is the golden standard but culturing B.pertussis is not very easy. Bordetellae can be cultured from nasopharyngeal swabs or nasopharyngeal secretions.

The sensitivity of the culture depends mainly on the technique of taking the nasopharyngeal swabs (calcium alginate or Dacron) or secretions, direct inoculation of nasopharyngeal swab material onto special freshly prepared media (Bordet-Gengou or Regan-Lowe) for primary isolation and immediate aerobic incubation in a stove. B.pertussis grows slowly, thus it is recommended to extent incubation time from 7 to 14 days.

The newest test for detection of B.pertussis and B.parapertussis is by polymerase chain reaction (PCR), a very specific test and more sensitive than culture. Nasopharyngeal swabs are suspended, then incubated and amplificated. The final PCR product is analyzed by gel electrophoresis and hybridization. In later stages of the disease PCR testing is more often positive than culture, as in patients treated with antibiotics or in vaccinated patients.

The PCR yield is about 2.4 fold higher than culture. The poor performance of culture may be due the fastidious nature of B.pertussis but it is also possible that by PCR B.pertussis DNA is detected in samples in which the organisms have become nonviable. In patients with clinical symptoms of B.pertussis infection and positive serology sensitivity of PCR and culture is low (21% and 7% respectively) but specificity of both is 98% 69.

In adolescents and adults culture or PCR is not useful when disease duration is longer than 3-4 weeks 69. In contrast, in non vaccinated or partially vaccinated young children culture or PCR is useful

in any stage of the disease since they have an immature mucosal immune response and therefore a slower eradication of the bacteria. Where with increasing disease duration the usefulness of PCR and culture declines thus serology becomes more important. Already in 1911 Bordet and Gengou published the first serological methods, detecting agglutinating antibodies to whole B.pertussis cells.

This remained the hallmark of pertussis serology for more than 70 years.

During the paroxysmal phase of the disease, eradication of the bacteria by antimicrobial drugs, such as erythromycin, will not significantly change the clinical course, although there is some clinical

evidence that some macrolides might reduce coughing complaints.

Although it is better for susceptible children (unimmunized children without a history of whooping cough) to avoid contact with pertussis patients during the first 4 weeks of their

illness, this is often difficult to achieve. Exposed unimmunized children are given a macrolide for 10 days after contact is discontinued or after the patient ceases to be contagious. Exposed immunized children younger than 4 years are most probably protected but protection may be enhanced by macrolides or by a booster dose of acellular pertussis vaccine.

Vaccination

Currently, approximately 80% of the world’s children are vaccinated against pertussis, most of whom have received the diphtheria-tetanus-whole cell pertussis combination.

Pertussis vaccine is produced from smooth forms (phase I) of the bacteria as a killed whole cell vaccine. General vaccination was introduced in the Netherlands in 1952.

Furthermore, since 1 January 1999 the primary vaccination for pertussis has been advanced. From that time children are vaccinated at the age of 2, 3, 4 and 11 months, instead of 3, 4, 5 and 11 months. Finally, in November 2001 a booster vaccination with an acellular vaccine, comprised of pertussis toxin, pertactin and filamentous hemagglutinin, was introduced in the National Immunisation Programme at the age of 4 years.

Owing to a relatively mild course of disease and to occasional complications after vaccination, it has been argued that general vaccination with the whole-cell vaccine is no longer justified. Therefore acellular pertussis vaccines have been developed. These vaccines are composed very differently and contain various amounts of structural components from the bacteria.

Components available for vaccine production include pertussis toxin (which is detoxified), filamentous hemagglutinin, pertactin and fimbrial antigens 2 and 3. Since 2000 4 year old children in the Netherlands are given a booster with acellular vaccine. Recent data suggest that after primary vaccinations of infants these vaccines can convey similar levels of protection as the whole-cell vaccine.

Thus, acellular vaccines have also been licensed for primary vaccination. In the Netherlands this acellular vaccine will be introduced for regular vaccination in 2005.

Aim and outline of this thesis:

In 1996 there was an outbreak of pertussis in the Netherlands, both in vaccinated and in non vaccinated people of all ages. Many questions arose what the cause or causes for this sudden increase in B.pertussis infection were. Among others the question arose whether vaccination or previous natural infection with B.pertussis guaranteed lifelong protection. As stated before criteria for laboratory confirmation of B.pertussis infection are: isolation of B.pertussis or detection of genomic sequences by PCR or significant, ≥ 4-fold increase of IgG-PT in paired sera to a level of at least 20 U/ml.

Because many patients visit their physician after weeks of coughing, culture or PCR are less sensitive and serology may already show high levels of IgGPT without significant increase anymore.

Thus we wondered if one point serology could be a useful tool in the diagnosis of B.pertussis infection.

Since there are no cut off values in one point serology as proof of actual or recent B.pertussis infection it would be opportune to develop such cutoff values. Therefore it is important to establish what level of IgG-PT is proof of a recent B.pertussis infection. Consequently we wondered what the natural course of IgG-PT is after infection. Accordingly it is necessary to gain insight in the rise, peak and decline of IgGPT after natural infection with B.pertussis. Since B.pertussis infection is especially dangerous in young children not or partially vaccinated and since the main source of infection for these children are adults with often atypical clinical manifestations, it is important to gain insight in the incidence of B.pertussis infection 93. Is it possible, once knowing the natural course of IgGPT, to calculate the incidence of B. pertussis infection in the Netherlands in different age groups from available surveillance data on IgG-PT levels in the general population?

For these reasons we investigated the level of IgG-PT in patients who suffered from B.pertussis infection, in the years after their infection, every time a blood sample was taken for other reasons.

Following these patients through the years we looked out for patients with renewed clinical symptoms of B.pertussis infection and a renewed increase of IgG-PT. From other studies 17-22,24 it is

known that in some patients there may be evidence of other pathogens involved in the pertussis syndrome besides B.pertussis and in others of pertussis-like complaints without proof of B.pertussis infection. We questioned therefore whether there are many mixed infections in patients with pertussis like complaints and which pathogens are involved.

The aim of this thesis is to find an answer to the following questions:

1: Which titer of IgG-PT in one point serology is proof for recent infection?

2: May a patient suffer from B.pertussis infection more than once in a lifetime?

3: What is the natural course of antibodies against B.pertussis after infection?

4: What is the yearly incidence of B.pertussis infection?

5: What is the best way to protect newborns and not yet (fully) vaccinated babies against B.pertussis infection?

6: How often mixed infections occur in B.pertussis infection?

7: What is the role of other respiratory pathogens in the pertussis like syndrome?

Specificity and sensitivity of high levels of immunoglobulin G antibodies against pertussis toxin in a single serum sample for diagnosis of infection with Bordetella pertussis

1: Department of Infectious Diseases Epidemiology, National Institute of Public Health and the Environment, Bilthoven, The Netherlands

2: Department of Pediatrics, ‘Groene Hart’ Hospital, Gouda, The Netherlands

3: Diagnostic Laboratory for Infectious Diseases and Perinatal Screening, National Institute of Public Health and the Environment, Bilthoven, The Netherlands

4: Laboratory for Clinical Vaccine Research, National Institute of Public Health and the Environment, Bilthoven, The Netherlands

5: Laboratory of Medical Microbiology, St. Elisabeth Hospital, Tilburg, The Netherlands

Abstract

Laboratory confirmation of pertussis by culture, PCR, or detection of antibody increase in paired sera is hampered by low sensitivity in the later stages of the disease. Therefore, we investigated whether and at which level concentrations of immunoglobulin G (IgG) antibodies against pertussis toxin (PT), IgG-PT, in a single se rum sample, are indicative of active or recent pertussis. IgG-PT, as measured

by enzyme-linked immunosorbent as say in units per milliliter, was analyzed in 7,756 sera collected in a population-based study in The Netherlands, in the sera of 3,491 patients with at least a fourfold increase of IgG-PT, in paired sera of 89 patients with positive cultures and/or PCR results and in the sera of 57 patients with clinically documented pertussis with a median follow-up of 1.4 years.

We conclude that, independent of age, IgG-PT levels of at least 100 U/ml are diag nostic of recent or active infection with Bordetella pertussis. Such levels are present in less than 1% of the population

and are reached in most pertussis patients within 4 weeks after disease onset and persist only temporarily.

Abstract

Susceptibility to infection with Bordetella pertussis re -emerges several years after pertussis vaccination. However, the duration of immunity after natural infection with B. pertussis,

postulated to be lifelong, is not known. In an ongoing study, the longitudinal course of pertussis antibodies in patients who have had laboratory confirmed pertussis is being followed using sera obtained

at irregular intervals. In 4 patients a re-infection with Bordetella pertussis is described respectively 7 (patient A), 12 (patients B and C) and 3.5 (patient D) y after the first infection. It seems that the longer the interval between the infections the more severe the complaints.

Conclusion: To the authors’ knowledge, these are the first patients in whom symptomatic reinfection with Bordetella pertussis has definitely been proven by laboratory confirmation of both episodes. Bordetella pertussis infection should be considered in patients with symptoms of typical or atypical whooping cough, irrespective of their vaccination status or previous whooping cough.

Introduction

Bordetella pertussis causes a respiratory infection, which is most severe and potentially fatal in very young children. Most countries have implemented nation-wide vaccination against

B. pertussis. However, owing to waning of vaccine -induced immunity (1, 2) infections with B. pertussis persist in vaccinated populations and these may remain unrecognized because of the atypical

or mild nature of symptoms (3–6).

In contrast, it has been commonly accepted that natural infection with B. pertussis confers lifelong protection (7, 8). However, this view has been chal lenged by several reports of occurrences of pertussis

in patients who have previously had pertussis (8, 9). In these reports, definite laboratory confirmation that the first episode had indeed been pertussis was lacking.

This report describes patients who suffered labora tory -confirmed typical pertussis and subsequently con tracted a second episode of coughing which was proven to be caused by B. pertussis.

They took part in a study of the longitudinal course of pertussis antibodies after natural infection (12).

For assessment of the possible age dependence of the rate of IgG-PT decline after natural infection, we extended our studies to a larger group of pertussis patients of various ages for

whom follow-up sera were available and to a group of elderly adult patients with pertussis for whom a follow-up serum sample had been drawn 1 year after infection with B.pertussis.

For analysis we applied an adapted version of a recently described dynamic model for the inactivation of Bordetella pertussis by the immune system.

Materials and Methods

Descriptive analysis

The age distribution of the 87 patients of group A is given in Figure 2. Pertussis-vaccination in The Netherlands at the time of the study consisted of 4 immunisations in the first year of

life, with a nationally produced whole cell pertussis vaccine. Fifteen of the 19 patients < 1 year were not vaccinated at onset of pertussis or were incompletely vaccinated (fewer than three immunizations). In only one of those 15 was it documented that the patient had received immunisations against B.pertussis after the pertussis episode (note: at the time of the study it was normal practice to abandon [further] vaccination against B.pertussis in infants who had pertussis). In the other 68 patients five were not vaccinated. For two subjects the vaccination status was unknown (aged 30.2 and 35.6 years).

In three of the 87 patients in group A during serological follow-up after pertussis a second infection with B.pertussis was serologically documented, respectively 3.4, 5.8 and 6.6 years

after the first (see below for details). In those three patients the follow-up of the first infection with B.pertussis was considered to have ended with the last serum obtained before onset

of the second infection with B.pertussis. The serological follow-up of the second episode in two of these three patients was > 6 months and inclusion of the follow-up of these two

second episodes therefore resulted in follow-up in group A of 89 episodes of infection with B.pertussis in 87 patients. From the 89 follow-up periods in group A, 403 sera were available.

The distribution of those sera over time since onset of pertussis is shown in Table 1. The mean number of sera for each follow-up period was 4.5 (range of 2-12). As shown in Figure

3 the time between onset of infection with B.pertussis and obtaining the last follow up serum sample ranged from 6 months to 10.7 years with a median of 3.1 years.

The age distribution of the 34 patients of group B is also given in Figure 2. Since the national vaccination program in The Netherlands against pertussis started in 1952, the 33 patients of group B who were born before 1952 (i.e. all the nuns) had not been vaccinated.

The one patient born after 1952 (the employee) had been completely vaccinated. From the 34 pertussis-episodes of the 34 patients of group B 99 sera were available, with 64 sera in the first 6 months after onset, six sera from six patients in the second half year and 29 sera from 29 patients in the second year (Table 1). The mean number of sera per patient/episode was 2.97 (range of 2 to 3). The time between onset of pertussis and obtaining the last follow up serum sample ranged from a minimum of 11.1 months to a maximum 17.6 months with a median

The second youngest age group (1-4 years) appears to show a response which is slightly different from the other age groups (slower increase, lower magnitude). As already stated,

data in this group show markedly high dispersion, as may also be appreciated by comparing error bars (illustrating variation) in Figure 5 (see also Fig.4 and Tables 2 and 3).

Discussion

This study in 121 pertussis patients aged 0-94 years shows that elevated levels of IgG-PT induced by infection with B.pertussis consistently decline again to below detection levels

within several years, unless this course is interrupted by re-infection with B.pertussis, which was observed in three cases. In each of the different age groups there was wide variation in

rapidity, intensity and rate of decay of IgG-PT and between groups there was a wide overlap of the ranges of these variables. Although the pattern of IgG-PT response is not significant

age dependant, there was a tendency for the IgG-PT response after infection to be faster and stronger and the decay to be more rapid with increasing age. The pattern in the 1-4 y age

group was different from all other age groups in that the IgG-PT response tended to be slower mounting and smaller in amplitude and in most cases decayed rapidly.

Our study is unique with respect to the high number of participating patients, the wide age range of the patients, the long follow up times in part of the patients and also the method of

analysis using a hierarchical model for induction and decay of antibodies after infection. The few studies of the course of IgG-PT after infection with B.pertussis that have been published all show decline over time but assessment of possible differences in children and adults is lacking and follow up times are relatively short. In the studies of Hodder et al (9) and Heininger et al (32) the course of IgG-PT over a period of 20 months after pertussis in 48 adults aged >65 years (9) and over a period of 28 months in 11 adults (32) was remarkably similar to the pattern over the same time span in adults in our study. In those studies the internationally standardized IgG-PT ELISA, recommended by FDA, was used of which the relationship with our ELISA is known (see Materials and Methods section; (29)).

For instance, the geometric mean IgG-PT titre of 11 adult pertussis patients was 242 IU/ml at 2 months after onset of infection and 45 IU/ml at 28 months after onset (32). However, in none of those patients IgGPT had declined below detection levels, which may be related to the limited follow up time.

Two other studies, one comprising both adults and children (2) and one comprising young children (33), also showed a significant decline of geometric mean IgG-PT titre 1 year after

pertussis. The study with the longest follow up time was from Tomoda et al. who measured IgG-PT during an outbreak of pertussis in a semi-closed adult community and again 5 years

later (34). He showed that in the large majority of the 21 pertussis-patients of that outbreak (37.5 + 12.1 years) IgG-PT was > 50 U/ml several weeks after onset of pertussis; 5 years later

the IgG-PT in all patients had declined to < 10 U/ml and in most to undetectable levels.

These findings support our conclusion that in the majority of cases or perhaps in all cases of pertussis, the subsequent decay of infection-induced IgG-PT does not level off well above

detection thresholds but progresses to undetectability.

The relatively short persistence, in all age categories, of infection-induced peak levels of IgG-PT supports our previous finding that a diagnostic cut-off of 100 U/ml (equivalent to

125 IU/ml) for serodiagnosis of actual or very recent infection with B.pertussis is valid for all ages (19). Although vaccination can induce IgG-PT levels > 100 U/ml, interference with

serodiagnosis of pertussis is minimal because IgG-PT induced by vaccination with whole cell pertussis vaccines or acellular pertussis vaccines also declines quite rapidly. Multiple studies show a decrease after primary and booster vaccination to very low or undetectable levels within 1 to 8 years (13, 35-38).

Taranger et al: followed 813 children after pertussis vaccination with a monovalent pertussis toxoid vaccine. There was a strong IgGPT response to a geometric mean of 143 IU/ml at 1-2 months after vaccination and a rapid decrease to a geometric mean of 8 IU/ml at 21-32 months after vaccination.

One explanation for our finding of a relatively slow and low IgG-PT response in the 1-4 years age group, the large majority of whom had been (recently) vaccinated, might be that children who suffer from B.pertussis infection shortly after vaccination are a separate category, i.e. are children with a relatively poor immunoresponse to pertussis antigens. Indeed, Taranger et al: showed that previously vaccinated children who within 33 months after completing vaccination developed pertussis upon exposure in their household had had a significantly lower IgG-PT response 1-2 months after the third vaccination with a monocomponent (PT) acellular vaccine (mean peak-response 79 U/ml) than children who did not develop pertussis upon household exposure within that time-frame (mean peak response after vaccination 212 U/ml).

The tendency of the IgG-PT response to be more rapid and strong with increasing age may indicate the involvement of specific memory immunity through encounters with B.pertussis

antigens earlier in life. Also Granström et al (40) have shown that adults have a faster peak response than children after a natural infection. The intuition that such rapid and strong

responses would persist longer evidently is not true. Perhaps the rapid decay of IgG-PT after a rapid and strong response may be explained by a faster eradication of the pathogen

and shorter duration of antigenic stimulation of the immune system. The phenomenon as such has been noted before: Blennow & Granström (41) showed that children receiving a

booster vaccination with an acellular vaccine containing pertussis toxin and filamentous haemagglutinin showed a more rapid decay of neutralizing antibodies to pertussis toxin in

the Chinese hamster ovary cells assay than after the primary vaccination. This despite the fact that after the booster vaccination the median of the neutralizing antibodies titres was higher than after the primary vaccination.

The three re-infections documented in this study were in children whose IgG-PT at the time of the second infection had declined to <10 U/ml. Although the number of re-infections was

small, their timing and incidence was compatible with the statement that natural infection initially induces protection against re-infection but that susceptibility to (re-)infection reemerges when IgG-PT concentrations have fallen to < 10 U/ml and that subsequently susceptibility for disease in association with infection increases over time. In a previous paper,

we have described two other patients who had infection with B.pertussis 12 years after the first episode. In contrast to the three re-infections 3.4, 5.8 and 6.6 years after the first in this

study, the re-infections after 12 years were associated with typical symptoms i.e. long-lasting paroxysmal cough.

In conclusion this study shows that the high IgG-PT concentrations induced by infection with B.pertussis consistently decline to low or undetectable levels within 5-6 years, this decline is associated with emergence of susceptibility for re-infection and disease. Although the pattern of decline is largely independent of age, there is a tendency for older people to have a more rapid increase, higher peak and faster decline. This study also shows that a diagnostic cut-off of 100 U/ml is not only valid in children (19) but is true for all ages.

Abstract

Objectives: Bordetella pertussis circulates even in highly vaccinated populations and there is a considerable amount of asymptomatic infection in adults. For designing more effective

vaccination schedules it is important to quantify the age-dependent relation between the number of notified cases and the number of infections.

Methods: We used a statistical relationship between the time-since-infection and the IgG pertussis titers, derived from a longitudinal data set, to estimate time since infection for all

individuals in a cross sectional population-based study (1995-1996) based on their titers.

Age-specific incidence of infection with B. pertussis was calculated and compared with the age distribution of notified cases of pertussis in 1994-1996.

Results: Estimated incidence of infection was 6.6% per year for 3 to 79-years olds, annual incidence of notified cases 0.01%. Estimated age-specific incidence of infection was lowest

for 3-4 year-olds and increased up to the age of 20-24 years. The number of notified cases was highest for 3 to 9 year olds.

Conclusions:

B. pertussis infections occur frequently in the Dutch population, particularly in adults whose reported incidence is low. The age-distribution of infections differs notably from those of notified cases. Vaccination strategies should take into account age-specific circulation and contact patterns between age groups.

Introduction

Despite widespread vaccination, infection with Bordetella (B.) pertussis remains a cause of considerable morbidity even in countries with high vaccination coverage. The continuing

circulation of the pathogen is attributed to waning of vaccine-induced immunity, which leads to the occurrence of pertussis infection among previously vaccinated children, adolescents

and adults [6-9]. In those groups, an infection with B. pertussis often goes undiagnosed, since the disease in many cases is atypical and mild and laboratory diagnosis is very insensitive, if only culture and/or PCR is available. Like others, we have shown that availability of pertussis serology and particularly the definition of reliable criteria for positivity associated with high levels of IgG to pertussis toxin (IgG-PT) in a single serum sample (one-point serology) can greatly enhance the detection of pertussis infections.

Inclusion of positive one-point serology as laboratory confirmation for notification in the Netherlands has increased the notification rate considerably, especially among older children

and adults. However, the use of surveillance data of notified infections for understanding the epidemiology of pertussis remains limited by the lack of registration- or notification

discipline. Furthermore, the fraction of infections diagnosed or reported is age-dependent because the severity of the disease as well as the sensitivity of diagnostic methods is agedependent. Circulation of B. pertussis in vaccinated children, adolescents and adults plays an important role in the continuing transmission of the pathogen to infants too young to be

vaccinated, in whom disease is most severe and possibly fatal.

To design better preventive measures, for example by determining the optimal ages for booster vaccinations, insight is needed into the age-specific incidence of all infections with B. pertussis as opposed

to only those symptomatic infections that are diagnosed and reported. Information about the seroprevalence of IgG-PT antibodies in the general population in combination with knowledge about the rate

of decline of IgG-PT antibody levels after infection with B. pertussis offers the opportunity to study the incidence of infection in various age-groups irrespective of clinical course, diagnosis and reporting frequency. Pertussis toxin is expressed only by B. pertussis and cross-reacting antigens have not been described.

Furthermore IgG-PT responses occur in most patients with B. pertussis infection and high levels persist only temporarily [10]. We estimated the incidence of B. pertussis infections

in the population using a novel two-stage approach. A statistical description of the decline in antibody levels after infection as derived from a small scale longitudinal study was

combined with data about the age-specific distribution of IgG antibodies against pertussis toxin (IgG-PT) in sera derived from a large scale cross sectional study of the general population to estimate the age-specific incidences of infection for the age range 3-79 years. These were compared to notification data of reported clinical cases of pertussis. Implications for vaccination-strategy are discussed.

Analysis

In Teunis et al: a skewed hyperbolic function was used to describe the relationship between the log time since infection and the log antibody titer. This four-parameter function

was fitted to the data for each individual patient from the longitudinal study of diagnosed clinical pertussis patients yielding a set of response curves with variation among individuals

(Figure 1). On a linear time scale, the antibody response rises very fast (within a few days) and then declines very slowly over a period of several years. Therefore, the rising part of the

response curve may safely be neglected and an inverse function can be determined on the basis of the long monotone declining part of the response. The inverse can be determined

for each individual response curve separately. For every value of the log titer an average time since infection can be calculated.

Discussion

Incidence and age profile of infections

There has been a long-standing discussion in the literature, whether vaccination against infection with B. pertussis reduces transmission of infection as opposed to only reducing

the incidence of clinical infections [27-29]. While there is agreement about the fact that vaccination does indeed reduce transmission, to estimate the amount of that reduction requires

knowledge about the fraction by age of clinical cases among all infections. This fraction is determined by a complex interplay of the age-dependent force of infection, immunity and

reporting behavior. Techniques of the analysis of serologic data have made it possible to identify individuals with recent B. pertussis infections and compare seroprevalence data with

notifications of symptomatic disease IgG-PT levels in the Netherlands can be interpreted as markers of recent infection, because firstly the amount of pertussis toxin in the Dutch whole cell vaccine is very low and second, because vaccine-induced IgG-PT levels are minimal and short-lived. However, also in populations in which pertussis vaccines are used that contain moderate to high amounts of pertussis toxin, vaccine induced IgG-PT declines to barely detectable levels within 2 to 4 years.

Two major issues clearly emerge from our analysis.

1. the estimated incidence of infections with B. pertussis is considerable in all age groups and much higher than the reported incidence. We estimated that around 6.6% of the Dutch population had experienced infection with B. pertussis in the year before serum sampling, while, in contrast, the incidence of notifications in 1994 to 1996 amounted to 0.01% per year.

2. the age-specific profile of the reported cases diverges remarkably from the estimated age-specific profile of incidence of infection with B. pertussis. While the highest incidence of reported symptomatic cases is observed among children aged 3 to 9 years, the incidence of infection is lowest among 3-4-year-olds, increases with age and peaks for 20 to 24-year-olds. Therefore, most cases are notified in those age-categories with the lowest incidence of infection.

Our cross-sectional study was performed in a limited time period (1995-1996). Therefore, the estimated incidence of infection and its age-distribution are a snapshot in time and it is

not clear whether those estimates apply to other time periods. We believe that the age-specific profile of incidence of infection with B. pertussis is rather stable over time. Repeating the

cross-sectional study can only assess whether or not this is correct. However, some support for our hypothesis is found in the similarity of the sero-profile of IgG-antibodies against pertussis toxin in 548 vaccinated children of 1-12 years in 1980, the sero-profile of about 800 individuals of all ages in 1992-1993 (pilot study for the large cross-sectional study) and the sero-profile in the present study in

1995-1996. Also, the fact that our findings in adolescents and adults are in agreement with the high incidences found in prospectively followed cohorts of small numbers of adolescents and adults in

the USA in other time period may be seen as supporting our hypothesis. In those studies significant increases in sequential sera of IgA/IgG antibodies against pertussis toxin were taken to indicate infection with B. pertussis. In the studied cohorts incidences of 3.3% to 8% per year were estimated. As discussed by Teunis et al.: the immune response of IgG-antibodies against pertussis toxin after infection

with B. pertussis shows large variation among individuals. Here, we worked with the point wise (for each titer value) averages of the individual response curves (Figure 1). This means that we neglected the dependence within individual responses but, as we applied this procedure to a large representative population sample, we assumed that variation on the individual level averages out on the population level. We assumed that the longitudinal study population and the cross-sectional study population are identical in their responses to pertussis infection.

With the size of the longitudinal sample as it is available to us at present, we can only say that the responses in different patient categories did appear to be similar.

Several factors are responsible for the large discrepancies between reported pertussis cases and the estimated cases of infection in their incidences and age-profiles. The amount of

underreporting varies by age, because severity of disease, medical care seeking and diagnostic power are varying with age. Indeed, a high rate of underreporting has been observed mainly in older children, adolescents and adults.

Recently, Strebel et al.: active case finding among older children and adults (10-49 years) from a well-defined population, who consulted the physician with cough lasting at least 14 days. An incidence

of symptomatic infection with B. pertussis of 0.5% per year was found, which was about 100 fold higher than the incidence of notified cases in that age-category.

A similar high incidence of symptomatic infections with B. pertussis among adults was found in a highly vaccinated region of France.

Another set of factors that influence the age-specific incidence profile of pertussis infections is related to the dynamics of transmission and immunity. The transmission of air-borne infections is strongly determined by the age-dependent patterns of mixing in a population. The contact rates between age groups, in conjunction with age-specific fractions of susceptible, infectious and immune individuals, channel the transmission of the infection through the population. The peaks observed in Figure 4 of age-specific incidence in the population might be related to high contact rates, the lower infection rate for those aged 60 years and older to a lower contact rate in that age group. The gradual increase with age of the incidence of infection to a peak of 10.8% in 20-24 year-olds suggests that there is a high variability in the duration of vaccine-induced immunity, which in some may be less than 2 years, in others more than 5 years. However, the incidence of notified cases of pertussis is highest among

3-9 year-olds, suggesting a strong reporting bias in that age category but also showing that susceptibility for symptomatic infection with B. pertussis may re-emerges shortly after vaccination.

In the USA, sero-prevalence data of IgG-PT in the population have been seen as an indication that immunity after vaccination with whole-cell vaccines wanes after about 10 years. However, in that study the investigators did not take the decline of IgG-PT after infection into account.

The mean incidence of 6.6%, if constant in time, indicates that on average, within a period of 20 years the entire population experiences infection, i.e. that vaccination against pertussis will be followed

on average by 3 episodes of natural infection during life. Indeed, there are strong indications that immunity wanes also after natural infection and that re-infection is possible.

German investigators estimated the duration of the protective period following natural infection at 20 years. Recently, we observed laboratory-confirmed mildly symptomatic

infection with B. pertussis in 4 children 3,5-12 years after the first laboratory-confirmed clinically typical episode of pertussis.

Implications for vaccination

In the Netherlands, the age-specific profile of notifications with the highest incidence in those aged 3 to 9 years led to the decision of the Dutch Health Council to introduce a booster

vaccination with acellular pertussis vaccine at 4 years of age. Since siblings play a role in transmission to vulnerable infants, one expects the incidence of severe pertussis in

infants to decline. However, on the basis of the incidence estimates presented here, we expect that the introduction of the booster vaccination at 4 years on the long run will lead to a shift of the peak

of infections to older age groups. Those older age groups, being the peak child bearing ages, may have more contacts with vulnerable infants, implying that on the long run booster vaccination might lead

to an increase of the incidence of severe pertussis in infants.

This is consistent with the fact that in the USA, where a booster vaccination at preschool age has been included in the vaccination-scheme for several decades, the proportion of infants

among the notifications of pertussis is relatively high.

Conclusion

More insight is needed into the role of adults as compared to siblings in the transmission of B. pertussis to young unvaccinated infants. The results of the present study support the findings of others in that adults are an important source of infection. More effective than booster vaccinations for adults given in 10 yearly intervals might be a strategy that directly targets (future) parents and caregivers. Mathematical modeling studies are needed for a precise quantitative analysis of the effects of different vaccination strategies on the age-specific incidence of (symptomatic and asymptomatic) pertussis infections. Previous modeling studies had to cope with the lack of data concerning the force of infection and the age dependent fractions of symptomatic and notified cases of infection. While we are

still far from having a solid quantitative basis on which to build reliable mathematical models, we think that our study investigated an essential link between the transmission dynamics as described by mathematical models and notification data of pertussis infections.

To our knowledge, the methodology we used to estimate the frequency of infection with B. pertussis has not previously been described and can be used more generally to estimate infection frequency from sero-prevalence data.

Kinkhoest en Zwangerschap

F.G.A. Versteegh 1, B. Muller 1, J.C.M. van Huisseling 2, J.F.P. Schellekens 3, J.J. Roord 4 published in part as “ Kinkhoest en Zwangerschap” in Ned Tijdschr Obstet Gynec

1: Groene Hart Ziekenhuis, afdeling Kindergeneeskunde, Gouda

2: Groene Hart Ziekenhuis, afdeling Gynaecologie/Verloskunde, Gouda

3: Rijksinstituut Voor Volksgezondheid en Milieu, Bilthoven

4: Vrije Universiteit Medisch Centrum, afdeling Kindergeneeskunde, Amsterdam

Abstract

Whooping cough is a serious, contagious disease. With the recent increase in Bordetella pertussis infections in the vaccinated population the risk for perinatal whooping cough (in

mother and child) has increased. The infants too young to be vaccinated are at greatest risk for severe complications and fatal outcome of pertussis. Because Bordetella pertussis infections in adults ànd

in young infants may present as atypical disease the diagnosis may be missed in first instance. Symptoms, diagnosis, treatment and prophylaxis are discussed. The purpose of this paper is to improve knowledge about the risks and prevention of perinatal pertussis in mother and child.

Inleiding

Kinkhoest, een zeer besmettelijke luchtweginfectie met de obligaat humane bacterie Bordetella pertussis (Bp), is, ondanks vaccinatie, blijven voorkomen en de incidentie lijkt toe te nemen, vooral

bij gevaccineerde kinderen en volwassenen. (1,2) Ook het aantal gevallen onder de leeftijd van 3 maanden neemt toe, met een mortaliteit van 1-2 %. (2,3) Kinkhoest bij pasgeborenen vindt bovendien

vaker plaats dan wordt herkend. Het probleem van perinatale kinkhoest wordt onderschat, alhoewel er de laatste jaren toenemend aandacht voor is. (4-16) Tegen deze achtergrond is het voor de gynecoloog/obstetricus belangrijk om te beseffen dat zwangeren die kinkhoest in de laatste maanden van hun zwangerschap doormaken hun pasgeboren kind kunnen besmetten en dat bij een bewezen infectie behandeling van de zwangere en preventieve behandeling van de pasgeborene geïndiceerd is. Het doel van dit artikel is obstetrici en verloskundigen van een belangrijk en mogelijk toenemend probleem bewust te maken door op de grote risico’s van perinatale Bordetella pertussis infectie te wijzen en gegevens wat betreft symptomen, diagnose, behandeling en profylaxe op een rij te zetten.

Etiologie, Vaccinatie, Incidentie:

Kinkhoest bij de mens wordt meestal veroorzaakt door Bp, soms door de Bordetella parapertussis. De bacterie hecht zich aan het trilhaarepitheel van de nasopharynx, waarna hij zich lokaal vermenigvuldigt en toxinen produceert die schade in de luchtwegen aanrichten waardoor de typische uitputtende, aanvalsgewijze hoest ontstaat. De bacterie zelf is niet invasief. Wel kunnen er systemische effecten van geabsorbeerde toxines optreden, zoals een sterke leucolymfocytose.

Sinds 1954 worden in Nederland kinderen op de leeftijd van 3, 4, 5 en 11 maanden gevaccineerd tegen kinkhoest met een whole cell vaccin dat geproduceerd wordt door het RIVM. Sinds 1999 zijn de eerste drie immunisaties vervroegd van 3, 4 en 5 naar 2, 3 en 4 maanden, om de kwetsbaarheid van de jongste kinderen zo klein mogelijk te maken. Sinds 2000 is bovendien een boostervaccinatie met een acellulair vaccin op 4 jarige leeftijd ingevoerd, om de circulatie van Bp en ziektegevallenbij lagere schoolkinderen te verminderen. Wel heeft de Inspectie voor de Gezondheidszorg zeer recent zijn grote zorg uitgesproken over de toenemende kritische houding van ouders tegen het vaccineren o.a. onder invloed van de Nederlandse Vereniging

Kritisch Prikken.

Noch vaccinatie, noch een infectie levert een blijvende immuniteit op. Infecties met Bp, hetzij symptoomloos, hetzij met milde atypische symptomen, hetzij met typische symptomen (kinkhoest) kunnen

dus ook op latere leeftijd optreden. Ook mensen die een infectie met Bp hebben doorgemaakt kunnen later opnieuw die infectie krijgen, vaak gemitigeerd en deels symptoomloos. (17-23) Waarschijnlijk maakt ruim 6 % van de bevolking jaarlijks een infectie met Bp door.

Nederland heeft al decennia lang een hoge vaccinatiegraad (ongeveer 96%). Desondanks worden jaarlijks 300 tot 500 kinderen in het ziekenhuis opgenomen wegens kinkhoest (60% jonger dan 3 maanden, 18% tussen 3 en 12 maanden). Vrijwel alle sterfgevallen ten gevolge van kinkhoest (in Nederland 2-5 per jaar volgens registratie van het CBS) vallen in de groep jonger dan 3 maanden.

Symptomatologie:

Het typische ziektebeeld kenmerkt zich door drie stadia. Het eerste stadium is het catarrale stadium dat een à twee weken duurt en wordt gekarakteriseerd door niet specifieke symptomen van milde hoest, verkoudheid en soms lichte koorts. Dit is de meest besmettelijke fase. Hierna volgt het paroxysmale stadium dat zich kenmerkt door plotseling heftige en herhaaldelijke hoestaanvallen met de typische gierende inademing aan het einde van de paroxysme. Deze periode kan enkele weken tot maanden duren. Vervolgens treedt het convalescentie stadium in met een geleidelijke afname van de klachten.

Kinkhoest en Zwangerschap:

De belangrijkste bron voor een infectie met Bp zijn vermoedelijk volwassenen met geringe symptomen die kinderen in hun omgeving besmetten. Waarschijnlijk zijn de moeders de

belangrijkste infectiebron voor kinderen, en met name pasgeborenen. Maar ook andere gezinsleden kunnen als infectiebron fungeren.

Het doormaken van een infectie met Bp geeft geen levenslange bescherming. De antistoffen verdwijnen na een infectie met Bp relatief snel. Na een jaar is de gemiddelde waarde voor IgGPT 32 U/ml. Onbekend is op welk niveau er geen bescherming meer is, echter algemeen wordt aangenomen dat bij een titer lager dan 20 U/ml er onvoldoende bescherming is.

Indien de moeder antistoffen opbouwt tegen Bp, tijdens de zwangerschap of kort daarvoor, gaan deze via de placenta over naar het kind. (26-31) Uit de beschikbare data blijkt, dat indien de infectie

met Bp vroeg genoeg in de zwangerschap plaatsvindt en de moeder voldoende tijd heeft om antistoffen te vormen, er weinig risico is voor het kind. Indien de infectie met Bp bij moeder optreedt kort

voor de partus (laatste 1 à 2 maanden), is er onvoldoende bescherming voor de baby en als de besmetting bij moeder optreedt na de bevalling, is er evenmin bescherming en ook niet als de infectiebron

een van de anderen gezinsleden is (of een van de bezoekers).

Dat maternale antistoffen bescherming aan het kind kunnen bieden is in verschillende studies aangetoond. Al in 1943 toonden Cohen en Scadron aan dat antistoffen van moeders die tijdens de zwangerschap werden gevaccineerd tegen kinkhoest overgingen naar de pasgeborenen. Oda laat in muizen zien dat babymuizen geboren uit gevaccineerde moeders beschermd waren tegen Bp infectie

en dat ook colostrum een dergelijk beschermend effect heeft. (27,30) Ook Arciniega laat zien dat antistoffen tegen Bp in muizen passieve bescherming kunnen geven en dat deze antistoffen ook gevonden worden in de sera van pasgeborenen.

Er zijn slechts enkele studies beschreven betreffende kinkhoest gedurende de zwangerschap. Meestal verloopt de ziekte voor de moeder zonder al te veel problemen. Gedurende een langetermijn studie

van zwangere vrouwen met een besmettelijke ziekte tijdens de bevalling vonden Granström et al in de periode 1975-1985 35 vrouwen met kinkhoest, waarvan 19 in 1984-1985. Kinkhoest bleek daarmee

in de laatste twee jaar van die studie de meest voorkomende “kinderziekte” bij zwangere vrouwen te zijn.

MacLean en Calder: beschrijven 4 gevallen van kinkhoest in de zwangerschap van de 43 zwangerschappen (9,3%) die zich in die periode voordoen in een praktijk in Schotland.

Als besmetting van de pasgeborene tijdens of kort na de bevalling heeft plaatsgevonden ontwikkelt het ziektebeeld zich meestal rond de tweede, derde week. Besmetting in utero is wel eens vermoed, maar nooit bewezen. De klachten bestaan uit voedingsproblemen, snelle ademhaling en hoesten. Bij de jongste kinderen kan echter het klassieke ziektebeeld van kinkhoest achterwege blijven en staan algeheel ziek zijn en apneus op de voorgrond. Ook bradycardiën kunnen het eerste symptoom zijn. Deze aspecifieke presentatie is waarschijnlijk ook de reden dat het ziektebeeld vaak niet of laat herkend wordt bij hele jonge kinderen. Het ziektebeeld wordt mogelijk verergerd als er sprake is van dubbelinfecties.

Beiter: beschrijft een casus, waarbij de kinkhoest bij moeder niet werd herkend, de baby een week na de geboorte ziek wordt en overlijdt. Er bleek sprake van een dubbelinfectie met adenovirus.

MacLean: vindt bij 1 van zijn 4 beschreven patiënten één co-infect met Influenza B, en Smith et al: beschrijven 9 kinderen met ernstige vroege kinkhoest, waarvan 4 met een virale

co-infectie. Van de 9 baby’s overlijden er 6, waarvan 3 met een co-infectie.

De incidentie van perinatale pertussis in Nederland is onbekend echter uitgaande van 200.000 partussen per jaar in Nederland en een infectiefrequentie bij volwassenen van ongeveer 6% per jaar (zie hoofdstuk 6), zou dit betekenen dat er per jaar ongeveer negenduizend vrouwen een infectie met Bp doormaken in de zwangerschap. Dit will zeggen dat ongeveer 3000 vrouwen een infectie met Bp doormaken in de laatste twee maanden van de zwangerschap en de eerste maand erna. Zoals beschreven hoeft het klinisch beeld niet altijd even duidelijk te zijn en kan variëren van een milde luchtweginfectie tot typische kinkhoestklachten. Daartegenover staat dat ook andere verwekkers van luchtweginfecties een kinkhoestachtig beeld kunnen veroorzaken.

Gezien de hoge morbiditeit en mortaliteit van perinatale pertussis bij de neonaat is het belangrijk, zowel in de laatste weken van de zwangerschap als rondom de partus, bedacht te

zijn op het ziektebeeld kinkhoest bij de moeder en na afname van diagnostiek moeder en kind direct te gaan behandelen.

Diagnostiek:

De diagnostiek bij volwassenen en oudere kinderen met een verdenking op een infectie met Bp bestaat uit nasopharyngeale kweek en/of PCR op een nasopharyngeale wat en/of

serologie. De kweek is erg gevoelig maar slechts korte tijd betrouwbaar af te nemen, 1 à 2 weken na de eerste ziektedag. Voor de PCR geldt hetzelfde, al is de afnameperiode 3 tot 4

weken na de eerste ziektedag. Serologie wordt pas enige weken na de eerste ziektedag positief en is daarom in de acute fase niet behulpzaam. Bij baby’s dient ongeacht de ziekteduur altijd naast de serologie een kweek en/of PCR te worden ingezet, aangezien bij hen de afweer tegen Bp langzamer op gang komt en er langer een kans bestaat op een positief resultaat. Daarnaast zijn er ook andere ziekteverwekkers (o.a. adenovirus, influenzavirus, para influenzavirus, respiratoir syncytieel virus, Mycoplasma pneumoniae), die vergelijkbare hoestbeelden kunnen geven. Bovendien geven co-infecties waarschijnlijk meer morbiditeit. Voordat de uitslagen van de diagnostiek echter bekend zijn gaat er enige tijd voorbij. Het is dan ook belangrijk na afname van diagnostiek snel met behandeling te beginnen.

Behandeling:

Medicamenteus:

Granström et al adviseren de moeder te behandelen met Erythromycine 3 keer daags 250 tot 500 mg gedurende 10 dagen. Dit is niet nodig als de klachten langer dan 7 weken geduurd hebben op moment van diagnose, omdat het dan zelden meer voorkomt dat de patiënt nog Bp bij zich draagt en besmettelijk is. Wanneer de bevalling plaatsvindt binnen 7 weken na het ontstaan van de kinkhoest wordt geadviseerd bij opname een Erythromycine kuur te geven, zelfs als de moeder al een Erythromycine kuur gehad heeft. Dit omdat er geen 100 % zekerheid is dat de eerste kuur de Bp volledig heeft verwijderd. Ook de nieuwe macroliden (clarithromycine, azithromycine) zijn effectief gebleken in de behandeling van kinkhoest.

De bevalling moet gebeuren in een eenpersoons verloskamer en na de bevalling dienen moeder en kind op een eenpersoonskamer behandeld te worden. Het kind van de moeder met kinkhoest krijgt, gedurende 5 dagen na de geboorte, 40 tot 50 mg. Erythromycine per kilo lichaamsgewicht. Er is geen bezwaar tegen borstvoeding. Voordat moeder en kind naar huis ontslagen worden dienen alle gezinsleden, die mogelijk verdacht kunnen worden van een kinkhoestinfect, ook behandeld te worden met Erythromycine, gedurende minstens 5 dagen, omdat zij als mogelijke infectiebron een risico kunnen vormen voor de pasgeborene. Een probleem hierbij is natuurlijk dat bij volwassenen de kinkhoestinfectie praktisch symptoomloos of met atypische symptomen kan verlopen, zodat het moeilijk

kan zijn mensen, die een risico vormen in de omgeving van het kind, te identificeren.

Behandeling met pertussis immunoglobuline lijkt succesvol, maar wordt niet op grote schaal gebruikt en is in Nederland niet beschikbaar.

Preventie:

Kinkhoest is, ondanks de bestaande vaccinatie programma’s, nog steeds een probleem, met name voor de heel jonge kinderen die niet of onvolledig gevaccineerd zijn. Om te komen

tot een wereldwijde aanpak is de Global Pertussis Initiative opgericht, een internationale samenwerking van deskundigen vanuit verschillende disciplines.

Er zijn verschillende vaccinatie strategieën na het eerste levensjaar mogelijk voor de preventie van kinkhoest:

-het vaccineren van alle volwassenen: hiervoor is het in veel landen (ook Nederland) nog nodig een infrastructuur op te zetten om alle volwassenen hiervoor te motiveren en vervolgens te kunnen vaccineren.Voor een effectieve reductie van het aantal kinkhoest gevallen bij jonge kinderen is waarschijnlijk een vaccinatiegraad van >85% nodig en moet de vaccinatie iedere 10 jaar herhaald worden.

Dit wordt al aanbevolen in Canada en Oostenrijk.

-het vaccineren van alle adolescenten: omdat de immuniteit door vaccinatie tegen kinkhoest sterk gedaald is na 10-12 jaar vormen ook adolescenten een belangrijke bron voor pertussis infectie. De geschatte incidentie van kinkhoest in de leeftijdsgroep van 20-24 jaar is het hoogst van alle leeftijden (bijna 11%)(zie hoofdstuk 6). Revaccinatie alleen bij adolescenten (en niet bij volwassenen) zal onvoldoende bescherming bieden. Ook hier geldt dat in veel landen eerst een programma opgezet zal moeten worden om ook adolescenten te motiveren en vervolgens te vaccineren. Dergelijke programma’s worden aanbevolen in b.v. Australië, België, Canada, Duitsland, Finland, Frankrijk, Oostenrijk en Zwitserland.

-revaccineren op de leeftijd van 4-6 jaar: het revaccineren van kinderen in deze leeftijdsgroep gebeurt al in veel landen, in Nederland sinds 2000. Met het opnieuw vaccineren van kinderen rond het 4e jaar wordt verwacht dat de immuniteit tegen kinkhoest zal aanhouden tot verder in de adolescentie.

-selectief vaccineren van aanstaande moeders, familie en nauwe contacten: vaccinatie van zwangeren in het 3e trimester met de nieuwe acellulaire vaccins staat op dit moment ter discussie, mede gezien de onbekende risico’s voor de foetus. Middels een dergelijke vaccinatie zouden echter wel beschermende antistoffen aan het kind kunnen worden overgedragen, zoals bij een natuurlijke infectie, die mogelijk voldoende bescherming zouden kunnen bieden gedurende de eerste twee maanden tot de eerste vaccinatie. Een andere mogelijkheid is moeder direct post partum te vaccineren, en familieleden en andere nauwe contacten ook, deze zo mogelijk al voor de bevalling. Ook kan nog overwogen worden vrouwen met kinderwens te vaccineren voor de zwangerschap.

-het vaccineren van personeel werkzaam in de gezondheidszorg of kinderopvang: Ziekenhuispersoneel loopt ook het risico kinkhoest op te lopen, zowel thuis als in het ziekenhuis, door patiënten met kinkhoest. Granström et al geven aan dat voor medisch personeel het risico kinkhoest te krijgen van opgenomen patiënten met kinkhoest minimaal is. Ze adviseren wel dat medewerkers met hoestklachten niet zouden moeten mogen zorgen voor patiënten (moeders en pasgeborenen) totdat een diagnose is gesteld. Ook in de recente

studie van Riffelmann et al blijkt dat er geen hogere antistoftiters voorkomen bij ziekenhuis personeel in vergelijking met bloeddonors en matrozen, echter zij geven niet aan of er ten tijde van het onderzoek patiënten met Bp infecties waren opgenomen.

Kurt et al en Linnemann et al: laten echter zien dat personeel wel degelijk kinkhoest kan overbrengen op patiënten.

Weber et al: geven een overzicht van verschillende studies waarin kinkhoest epidemieën in ziekenhuizen worden beschreven, veroorzaakt door patiënten, ziekenhuis

medewerkers en bezoekers. Er is dus oplettendheid geboden ten aanzien van hoestklachten in het algemeen en kinkhoest in het bijzonder bij ziekenhuispersoneel.

-verbeteren/aanpassen huidige vaccinatieschema bij jonge kinderen: Vaccinatie van de neonaat kort na de geboorte heeft weinig zin. Englund et al laten zien dat een preëxistent hoger gehalte aan maternale antistoffen tegen kinkhoest bij het jonge kind een verminderde antistofreactie geeft na vaccinatie met een whole cell vaccin, maar niet bij een acellulair vaccin. Bovendien zijn pasgeborenen nog niet goed

in staat antistoffen te vormen.

Conclusies en Aanbevelingen:

1. Het niet overwegen van een infectie met Bp bij hoogzwangere vrouwen met luchtweginfecties kan fatale gevolgen hebben voor de pasgeborene.

2. Een infectie met Bp kan voorkomen bij volwassenen ook al zijn zij gevaccineerd of hebben zij eerder een infectie met Bp doorgemaakt. Bovendien kan een infectie met Bp bij volwassenen gemitigeerd en zelfs onherkenbaar verlopen.

3. Als de moeder geen infectie met Bp gedurende de zwangerschap of het jaar ervoor heeft doorgemaakt zijn er waarschijnlijk geen of onvoldoende beschermende antistoffen die transplacentair kunnen worden overgedragen.

4. Het klinisch beeld van een infectie met Bp bij pasgeborenen verschilt van het klassieke kinkhoestbeeld en wordt daardoor vaak laat of niet herkend.

5. Bij (verdenking op) perinatale pertussis bij de moeder dienen moeder èn kind adequaat behandeld te worden evenals naaste familie. Overwogen dient te worden ook (para)medisch personeel dat in nauw contact is geweest met de patiënt te behandelen.

6. Andere infectieuze verwekkers kunnen een kinkhoestachtig beeld veroorzaken.

7. Dubbelinfecties van een infectie met Bp met een andere verwekker kunnen een ernstiger beeld laten zien.

8. Om de bescherming van pasgeborenen tegen kinkhoest te verbeteren zijn nieuwe vaccinatiestrategieën nodig.

Serologic evidence of possible mixed infections with Bordetella pertussis and other respiratory pathogens

F.G.A. Versteegh 1, E.A.N.M. Mooi-Kokenberg 2, J.F.P. Schellekens 3, J.J. Roord 4

Submitted

1: Groene Hart Ziekenhuis, Department of Pediatrics, Gouda, the Netherlands

2: Groene Hart Ziekenhuis, Department of Medical Microbiology and Infection Control, Gouda, the Netherlands

3: National Institute of Public Health and the Environment, Diagnostic Laboratory of Infectious Diseases and Perinatal Screening, Bilthoven, the Netherlands

4: Vrije Universiteit Medical Center, Department of Pediatrics, Amsterdam, the Netherlands

Abstract

Background: In pertussis-like respiratory infections, once pertussis has been laboratory confirmed, other potential causative pathogens will seldom be looked for. Probably most

mixed infections are found accidentally and since these mixed infections might cause a more severe disease we performed a retrospective study of their incidence.

Patients and methods: We selected from 2 groups of patients with serologically confirmed Bordetella (B.) pertussis infection those in whom serology for other respiratory pathogens

had also been performed. Group 1 consisted of 50 pertussis patients with 51 episodes of B.pertussis infection selected from 100 patients with serologically confirmed pertussis.

They participated in a long term follow up after a B.pertussis infection. In group 2, 31 pertussis patients were selected from 98 consecutive patients with positive pertussis serology from one routine practice.

Results: In 23 of 82 pertussis infections (28%) serological evidence of one (n=21) or two (n=2) additional infections were demonstrated. These involved para-influenza virus (n=6),

RSV (n=6), Mycoplasma pneumoniae (n=5), adenovirus (n=4), influenza A virus (n=3) and influenza B virus (n=1).

Conclusions: We conclude that in patients with B.pertussis infection, co-infection with another respiratory pathogen is often present.

Introduction

Because both viral and bacterial infections are common during early childhood, dual infections are not unexpected. However, clinical manifestation of such combined infections

may be difficult to interpret despite an increase in knowledge concerning microbial agents associated with acute respiratory infections of childhood. For many years there has

been an interest in the dynamic interaction between pathogens and the influence on disease interference. Studies on Bordetella (B.) pertussis infection often associated adenoviruses

and other respiratory viruses with the pertussis-like syndrome in the absence of clear evidence of B.pertussis infection. The presence of more than one respiratory pathogen in an ill

infant may have important diagnostic, therapeutic, prognostic, epidemiological and infection control consequences. Despite the relatively high frequency with which both viral and

B.pertussis infection may occur in the community, the occurrence of a combination of two or more microbiological agents (or mixed infection) is underestimated, both in infants

and in adults. We retrospectively investigated the incidence of mixed infections with other respiratory pathogens in pertussis patients.

Pathogens in community acquired coughing in children: a prospective cohort study

F.G.A. Versteegh 1, G.J. Weverling 2, M.F. Peeters 3, B. Wilbrink 4, M.T.M. Veenstra-van Schie 1, J.M. van Leeuwen-Gerritsen 1, E.A.N.M. Mooi-Kokenberg 5, J.F.P. Schellekens 4, J.J. Roord 6

Abstract

To investigate in children with prolonged coughing the frequency of different respiratory pathogens, the rate of mixed infections and possible differences in severity of disease between

single and mixed infections, we performed a 2 years prospective study.

Children from 0-18 years with coughing lasting 1 to 6 weeks were included. Sera of 135 children (with 136 episode of prolonged coughing) were tested for antibodies against different

viruses and bacteria. Swabs were taken for culture and PCR on different viral and bacterial pathogens.

One or more pathogens were found in 91 patients (67%). In 49 (36%) one infectious agent was found, in 35 (26%) 2 and in 7 (5%) more than 2. The most frequent pathogens encountered

were rhinovirus (43 patients-32%), Bordetella pertussis (23 patients-17%) and respiratory syncytial virus (15 patients-11%).

The most frequent encountered mixed infection consisted of a combination of Bordetella pertussis and rhinovirus (n=14-10%).

No significant differences were observed regarding clinical symptoms between patients without pathogens and those in whom one or more pathogens were found, except that patients with

a mixed infection were significantly older. There is a strong seasonal influence in the number of infections but not in the number of mixed infections.

In children with prolonged coughing there is a high frequency of mixed infections occurring during the whole year regardless of the season. Mixed infection however does not cause

increased disease severity. Although patients with mixed infection were significantly older, no clinical symptoms were found that discriminate for specific pathogens.

Introduction

Prolonged coughing is a frequent symptom in children and often is associated with respiratory infection. Although prolonged coughing is a prominent feature in Bordetella (B.) pertussis

infection, infections with other respiratory pathogens may cause prolonged coughing as well. Mixed infections with a combination of 2 or more pathogens do occur. However data on

clinical manifestations of such combined infections may be difficult to interpret and it often concerns retrospective studies with different results. It has been suggested that mixed

infections may cause more severe illness, especially in younger children. In an earlier retrospective observational study (data not published) among 81 children with serologically

proven B.pertussis infection we found in 28% of the patients evidence for concomitant infections with other respiratory pathogens. It is known that many respiratory pathogens may

cause prolonged coughing in children, however prolonged coughing is not always caused by a respiratory infection.

We conducted this study to investigate the role of different respiratory pathogens in prolonged coughing in children and the frequency of possible mixed infections. We also studied whether

patients in whom one or more pathogens were detected suffered from more severe disease than patients without pathogens.

Summary and Conclusions

In this thesis we report the results of the studies we performed to answer the questions, which were put forward in chapter 1. The first part of this thesis focuses on the interpretation

of Bordetella (B.) pertussis serology and the epidemiological consequences of our results evaluating surveillance data. The second part of this thesis shows data on other respiratory

pathogens besides B.pertussis in mixed infections and prolonged coughing. In this last chapter the most relevant findings will be summarized following the questions from

chapter 1. Also practical implications and suggestions for future research will be proposed.

1: which titer of immunoglobulin G antibodies against pertussis toxin (IgG-PT) in one point serology is proof for recent infection?

Our results show that an IgG-PT level of at least 100 U/ml is a specific tool in laboratory confirmation by one-point serology of patients with a suspected pertussis infection.

Independently of age, these levels are diagnostic of very recent or actual infection with B.pertussis. Such levels are present in less than 1% of the population and are reached in

most pertussis patients within 4 weeks after disease onset. High IgG-PT levels persist only temporarily. The IgG-PT levels after natural infection with B.pertussis decrease within less

than 1 year to a level below 100 U/ml for almost all patients who had high IgG-PT levels.

The regression model we used in this study predicts that peak levels > 100 U/ml occur 4-8 weeks after infection. In the declining phase a level of 100 U/ml is reached in 4.5 months after

infection onset. A level of <40 U/ml is reached within 1 year after disease onset. At diagnosis the number of patients with IgG-PT levels ≥ 100 U/ml is considerably larger (4.5-fold) than the number

of patients with at least a 4-fold increase in IgG-PT. IgG-PT levels > 100 U/ml provide a useful laboratory tool for the diagnosis of pertussis in both the individual patient as in epidemiological studies.

2: may a patient suffer from B.pertussis infection more than once in a lifetime?

In chapter 3 we demonstrate that children, vaccinated or not, may have serologic evidence of re-infection with B.pertussis. Although it is known that people may suffer from re-infection

with B.pertussis, these patients are the first in whom symptomatic re-infection with B.pertussis has definitely been proven. The infection took place 3.5-12 years after the first infection. But it is obvious that clinical complaints not necessarily comply with typical pertussis infection. It was shown that the severity and duration of respiratory symptoms in patients with proven and relatively early re-infections with B.pertussis increased with time elapsed since the first infection.

B.pertussis infection should be considered in patients with symptoms of typical or atypical whooping cough, irrespective of age, their vaccination status or previous whooping cough.

3: what is the natural course of antibodies against B.pertussis after infection?

5: what is the best way to protect newborns and not yet (fully) vaccinated babies against B. pertussis infection?

In chapter 7 we review the risks of perinatal infection with B.pertussis, the diagnosis, treatment and prophylaxis. Also we consider the possible influence of different vaccination schemes on the rate

of B.pertussis infection in young children and the disease burden in older age groups.

Although the best way seems to be to vaccinate the adolescent and adult population every 10 years, the feasibility of this regime requires large investments in making people aware of the necessity of

these booster vaccinations and to develop a new infrastructure to deliver these vaccinations. Another option is to vaccinate pregnant women or women before pregnancy and their families.

6: how often mixed infections occur in B.pertussis infection?

During the prospective study on the longitudinal course of IgG-PT we regularly observed mixed infections. Therefore in chapter 8 we retrospectively looked for mixed infections in this group and

in a group of patients in whom the diagnosis pertussis was made in 1998. In 28 % of the patients with proven B.pertussis infection there was evidence of mixed infection with para-influenza virus,

RSV, Mycoplasma pneumoniae, adenovirus or influenza virus. Since infection with more than one pathogen might cause more severe or prolonged disease we suggest the possibility of dual infection

should be investigated more often.

7: what is the role of other respiratory pathogens in the pertussis like syndrome?

To investigate the role of different respiratory pathogens in prolonged coughing in children and to analyze the clinical impact of mixed infections of B.pertussis with other respiratory pathogens we performed a prospective study in children with coughing complaints lasting 1 to 6 weeks (chapter 9). 1/3 of the patients had one pathogen (n=49, 36%), 1/3 more respiratory pathogens (n=42, 31%).

The most frequent pathogens encountered were rhinovirus, B.pertussis and respiratory syncytial virus. In the patients with a mixed infection

the most frequent combination was B.pertussis and rhinovirus. Infections with more than one pathogen occurred during the whole year regardless of the season. No signs of more disease

severity could be demonstrated in children with more than 1 pathogen, although children with more than 1 pathogen were significantly older than those with none or 1 pathogen.

There were no clinical data found that discriminated between pathogens, whether pathogens were found or not, or differences in treatment.

The threshold for IgG-PT of 100 U/ml as proof of recent infection made it possible to improve the diagnosis of B.pertussis infection, by one-point serology. It becomes clear from the long term data on notification of B.pertussis infection, two-point and one-point serology that the discrepancy between notifications and positive serology became smaller since the acceptation of one-point serology and the notification rate has increased. Before 1996 all serological tests for pertussis were performed at the LIS-RIVM. However, since 1998 at least three of the 16 regional Public Health Laboratories and also some other (hospital) laboratories have started to perform serology with commercial available assays. Consequently, the population coverage of serological surveillance based on serological data of LIS-RIVM is now estimated to have decreased from 100% to less than 50%. This explains why the notification rate since 1999 is higher than the number of patients with positive serology at LIS-RIVM. Because there is no longer one central laboratory for the diagnosis of B.pertussis infections it is more difficult to gain insight in the epidemiology of B.pertussis infections in the Netherlands; notification data differ significantly from estimated infections and are very much lower.

It became clear that people may suffer from B.pertussis infection more than once in a lifetime.

When the time between infections is longer the more severe disease occurs in the second infection. There is some evidence that the moment of re-infections might be influenced by

booster vaccination against B.pertussis. This, together with the increasing awareness about the importance of adolescents and adults as a source of B.pertussis infection, makes it important to gain more insight in the incidence of B.pertussis infection and the influence of booster vaccination. Therefore studies should be designed to investigate this interaction in order to develop new vaccination strategies.

The development and validation of models of the kinetics of IgG-PT after B.pertussis infection makes it possible to compare patients of different ages, from different places or different time periods.

This will help to increase the knowledge about the epidemiology of B.pertussis infection. Since there seems a tendency of an age dependent course of IgG-PT after B.pertussis infection further studies

are needed to confirm this finding. If this age dependency holds true it influences the results of demographic studies about the incidence of B.pertussis infection. It might also influence decisions about vaccination schedules for adult booster vaccination.

Comparison of the different methods used in this thesis to calculate the course of IgG-PT after B.pertussis infection is therefore the first step. When this swifter and higher response in older people

means they will suffer from B.pertussis infection with less pronounced disease and therefore will not as easily be recognized as pertussis patients, more attention should be paid to identify B.pertussis infection in this age group.

Although there is a significant number of mixed infections both in B.pertussis infection and in prolonged coughing, we could not demonstrate more severe disease in mixed infection.

Also many people are carrier of respiratory infectious agents without disease. This leaves open the question of the relevance of mixed infections. When mixed infections are not true mixed infections

but consecutive infections they still may prolong disease. Also an increased knowledge about mixed infection might influence the management of patients.

Most important is how to control B.pertussis infection, especially for the vulnerable, not or partially vaccinated, young children. Isolating patients is effective but impossible to execute

regarding the number of patients and the difficulty to identify the majority of the older patients because of their aspecific or subclinical symptoms. Another possibility is antibiotic

prophylaxis for the ones at risk and all people around this patient, or only for those at risk.

Although passive immunization with human hyperimmune anti-pertussis gammaglobulin is still used occasionally elsewhere, it is no common practice in the Netherlands, where the hyperimmune anti-pertussis gammaglobulin is not available.

The best way to control B.pertussis infection is vaccination either with whole-cell or acellular vaccine. Since asymptomatic re-infection may occur rather a short period after infection

or vaccination protective immunity against disease lasts longer than the immunity against infection. Thus it seems that natural infection or vaccination provides better protection

against the toxic effects of B.pertussis infection than against adhesion but the protection is not life-long. The duration of the protective immunity acquired by vaccination lasts about as long as by natural infection (4-12 years vs. 4-20 years), although data are not unequivocal.

The interaction between vaccination and boosting of pertussis immunity by B. pertussis infection on one hand and waning immunity against B. pertussis on the other influences the epidemiology and

the dynamics of pertussis in the population. Therefore it is necessary to study the influences of changes in vaccination schedules.

The studies presented in this thesis made us understand better some things about B.pertussis infection but also raised new questions. New studies have to be performed to answer some of these questions but for sure will raise other ones.

Kinkhoest: nieuwe inzichten in diagnose, voorkomen en klinische aspecten

Samenvatting

Kinkhoest is een besmettelijke ziekte van de luchtwegen, veroorzaakt door de kinkhoest bacterie: Bordetella pertussis. Deze infectie kan voorkomen op alle leeftijden, maar de

klachten zijn het ergst bij kleine kinderen. Jaarlijks hebben wereldwijd 20-40 miljoen mensen kinkhoest, vooral in de ontwikkelingslanden. Er wordt geschat dat per jaar ongeveer 200 000-400 000 mensen aan kinkhoest overlijden.

De eerste epidemie van kinkhoest is beschreven door Guillaume de Baillou (1538-1616) in 1578, maar al veel eerder, in 1190, was in Engeland een ziektebeeld bekend dat kinkehost

werd genoemd. En Dodonaeus (1517-1585) beschrijft in zijn Cruijde boeck uit 1554 al een therapie voor kieckhoest!

Kinkhoest manifesteert zich als een heftig hoesten met stikaanvallen, maar afhankelijk van voorafgaande vaccinaties en/of eerdere infecties kan de ziekte milder verlopen. Het klinische

beeld kent drie stadia, 1: de catarrale fase, gekenmerkt door een verkoudheidsbeeld en beginnend hoesten, 2: de paroxysmale fase, met de typische hoestaanvallen, en 3: de herstelfase,

waarin de klachten langzaam afnemen. De totale ziekteduur bedraagt ongeveer drie maanden.

In het Chinees wordt de ziekte dan ook de “honderd dagen ziekte” genoemd.

De bacteriesoort Bordetella kent op dit moment acht verschillende typen: Bordetella pertussis,

Bordetella parapertussis, Bordetella bronchiseptica, Bordetella avium, Bordetella hinzii, Bordetella holmesii, Bordetella trematum en Bordetella petrii. Bordetella pertussis komt alleen

bij de mens voor, de andere komen vooral bij dieren voor maar kunnen ook klachten geven bij mensen, vooral de Bordetella parapertussis.

Bij een infectie met Bordetella pertussis komt een aantal ziekmakende stoffen vrij, zoals onder andere pertussis toxine, filamenteus haemagglutinine, fimbriae en pertactin. Tegen deze

stoffen, antigenen, vormt het lichaam antistoffen die een rol spelen in de immuniteit tegen infectie met Bordetella pertussis. Het aantonen van deze antistoffen wordt gebruikt in de

diagnostiek van kinkhoest.

De afgelopen tien jaar is er een duidelijke toename van het aantal kinkhoestgevallen geconstateerd, vooral in ontwikkelde landen met een hoge vaccinatiegraad. In Nederland was tussen 1989 en 1994

de gemiddelde incidentie (aantal nieuwe gevallen) van kinkhoest 2,4 per 100.000 mensen op basis van meldingen en 2,3 per 100.000 mensen op basis van positieve serologie (aantonen van antistoffen

in bloed). In 1996 was er sprake van een sterke stijging van de incidentie tot 27.3 per 100.000 en sindsdien is deze hoog gebleven. Een belangrijke oorzaak voor de toename van Bordetella pertussis infecties is dat er veranderingen zijn opgetreden in

Samenvatting

Kinkhoest is voortdurend aanwezig in de bevolking, en daarbovenop komen epidemieën voor met een cyclus van ongeveer 4 jaar in landen waar gevaccineerd wordt en een cyclus van 2 tot 3 jaar elders.

In Nederland zijn deze cycli, ondanks de vaccinatie, sinds 1996 ook om de 2 à 3 jaar. Bordetella pertussis is erg besmettelijk en wordt vooral door hoestdruppeltjes van mens op mens overgedragen.

Natuurlijke infectie of vaccineren geeft een langdurige maar geen levenslange bescherming tegen kinkhoestinfectie. Door gedeeltelijke bescherming kan een milder ziektebeeld ontstaan, dat vaak niet

als kinkhoest herkend wordt. Dit is vooral het geval bij adolescenten en volwassenen. Deze mensen kunnen echter wel anderen besmetten, vooral ongevaccineerde of gedeeltelijk gevaccineerde jonge kinderen.

De diagnose kinkhoest wordt gesteld op klinische symptomen: langer dan twee weken hoesten, met daarbij of hoestbuien, en/of kinken (gierende inademing) en/of overgeven

direct na het hoesten. De diagnose kinkhoest kan bevestigd worden door het kweken van de Bordetella pertussis of door het aantonen van deze bacterie middels de polymerase chain

reaction (PCR). Ook kunnen er antistoffen tegen antigenen in het bloed aangetoond worden.

Daarbij moet in twee verschillende bloedmonsters minstens een viervoudige stijging van de antistoffen optreden. Een positieve kweek met Bordetella pertussis is de gouden standaard

voor het vaststellen van kinkhoest, maar het kweken van de Bordetella pertussis is erg moeilijk.

PCR maakt het mogelijk ook nog Bordetella pertussis aan te tonen als de bacterie niet meer leeft; maar 3 à 4 weken na de infectie is de kans op een positief resultaat van de kweek of

PCR niet erg groot meer. Als de ziekte langer aanhoudt heeft het meer zin antistoffen in het bloed aan te tonen. Bepaling van de antistof immunoglobuline G tegen pertussis toxine (IgGPT) is de meest gebruikte en best gevalideerde test. IgG-PT wordt uitsluitend aangemaakt na een infectie met Bordetella pertussis, niet met andere Bordetella’s. Antistoffen tegen andere antigenen worden ook wel gebruikt, maar zijn minder gestandaardiseerd. Bovendien zijn deze antigenen ook bij infecties met andere Bordetella’s te vinden. Het vaststellen van een viervoudige stijging van de antistoffen in het bloed wordt bemoeilijkt doordat veel patiënten pas later tijdens hun ziekte medische hulp zoeken. Dan hebben de antistoffen vaak al (bijna) hun piek bereikt (meestal na 4 à 5 weken) en kan geen viervoudige stijging meer worden aangetoond. Heel jonge kinderen hebben nog een onrijpe afweer. Daardoor wordt de bacterie langzamer uit het lichaam verwijderd en heeft het zin ook na langere tijd nog een

kweek en/of PCR in te zetten.

Een klinisch beeld lijkend op kinkhoest kan ook veroorzaakt worden door andere verwekkers als adenovirus, influenzavirus, parainfluenza virus, respiratoir syncytiaal virus, Chlamydia pneumoniae of Mycoplasma pneumoniae. Ook zijn gemengde infecties beschreven met verschillende verwekkers.

Kinkhoest kan behandeld worden met antibiotica, macroliden hebben daarbij de voorkeur.

Vaccinatie beschermt tegen infectie met Bordetella pertussis. Meestal gebeurt dit met een combinatievaccin, waarin ook difterie, tetanus, en polio zitten. Jarenlang is gebruik gemaakt

van vaccins op basis van geïnactiveerde bacteriën, het whole-cell vaccin. Vanwege de bijwerkingen zijn nieuwe vaccins ontwikkeld met daarin een of meer geselecteerde antigenen,

zoals pertussis toxine. Dit zogenaamde acellulaire vaccin is in 2001 opgenomen in het Rijks Vaccinatie Programma.

Doel van dit proefschrift

Door de plotselinge stijging in 1996 van het aantal kinkhoest gevallen in Nederland stelden wij ons een aantal vragen:

- Kunnen we met één enkel bloedmonster de diagnose kinkhoest stellen?

- Kun je meer dan één keer in je leven kinkhoest krijgen?

- Hoe snel verdwijnen de antistoffen tegen kinkhoest na een infectie uit je bloed?

- Hoe vaak komen kinkhoest infecties elk jaar voor en bij wie?

- Hoe kun je de meest kwetsbaren (pasgeborenen en jonge kinderen die nog niet volledig zijn gevaccineerd) het best beschermen tegen kinkhoest?

- Hoe vaak komen menginfecties voor bij kinkhoest en zo ja welke verwekkers spelen een rol?

- Is ernstig hoesten altijd kinkhoest en wat is de rol van andere verwekkers van hoestklachten in het beeld lijkend op kinkhoest?

Opzet

Vanaf 1993 werd bij ongeveer 100 patiënten die kinkhoest hadden doorgemaakt, als er om een andere reden toch bloed afgenomen moest worden, na verkregen toestemming, wat extra

bloed afgenomen voor het opnieuw bepalen van het IgG-PT. Daarnaast werd bij een deel van deze patiënten gekeken of er nog andere verwekkers van luchtweginfecties aangetoond konden worden.

Deze gegevens werden op zich en in samenhang met andere gegevens bewerkt. Daarnaast werd in een tweejarige prospectieve studie vanaf 2001 onderzocht welke verwekkers aangetoond konden

worden bij kinderen met langdurig hoesten.

Resultaten

In hoofdstuk 1 wordt een overzicht gegeven van de huidige stand van zaken met betrekking tot Bordetella pertussis infecties. Daarnaast wordt uiteen gezet wat de vraagstellingen zijn van dit proefschrift.

Hoofdstuk 2 beschrijft de studie waarin wij aantonen dat een titer (gehalte) van IgG-PT van 100 U/ml of hoger bewijzend is voor een recente Bordetella pertussis infectie, en dat daarmee voor de diagnose kinkhoest bepaling in één bloedmonster (eenpuntsserologie) mogelijk is. Minder dan 1% van de bevolking heeft titers boven de 100 U/ml. In het wiskundig model dat we gebruiken in deze studie bereikt de IgG-PT titer een piek na 4-8 weken, waarna een geleidelijke daling optreedt. Na ongeveer 4½ maand wordt in de dalende fase gemiddeld de 100 U/ml grens weer bereikt, en na een jaar bedraagt de gemiddelde titer < 40 U/ml. Het aantal patiënten met een IgG-PT titer > 100 U/ml in één bloedmonster is 4,5 maal zo groot als het aantal patiënten met een viervoudige titerstijging in twee bloedmonsters. Een IgG-PT titer > 100 U/ml is daarmee een zeer bruikbaar hulpmiddel in de diagnostiek van kinkhoest in de individuele patiënt èn in epidemiologische studies.

In hoofdstuk 3 beschrijven we vier patiënten die serologisch bewezen voor de tweede keer een kinkhoest infectie doormaken, 3,5 tot 12 jaar na de eerste infectie. Drie van de 4 kinderen zijn in hun eerste levensjaar gevaccineerd. Ofschoon wel vermoed werd dat mensen opnieuw kinkhoest kunnen doormaken, zijn deze patiënten de eersten waar dit ook in beide gevallen is bewezen. Bij deze patiënten was de tweede infectie niet altijd herkenbaar als een typische kinkhoestinfectie. Wel waren de klachten ernstiger en leken meer op een typische kinkhoestinfectie naarmate de eerste infectie langer geleden had plaatsgevonden. Het is dus belangrijk bij mensen met hoestklachten ook aan een kinkhoestinfectie te denken, ondanks eerdere vaccinatie en/of kinkhoestinfectie.

In de hoofdstukken 2,4 en 5 zijn verschillende methoden/wiskundige modellen gebruikt om de daling van het IgG-PT te beschrijven. Deze drie methoden zijn het gevolg van opeenvolgende verbeteringen en voortschrijdend inzicht. In de in hoofdstuk 2 gebruikte methode wordt uitsluitend rekening gehouden met de gegevens uit de dalende fase van het IgG-PT.

Daarom werd in hoofdstuk 4 een model ontwikkeld dat ook gebruik maakt van de eerste stijgende fase van de antistofrespons. Het hierbij gebruikte model van een scheve hyperbool past beter op de gevonden data dan het eerste model, maar geeft nog steeds weinig aanknopingspunten bij het interpreteren van de gevonden immuun respons.

Vervolgens werd in hoofdstuk 5 het beloop van de antistofrespons in de tijd beschreven middels een dynamisch model van de interactie tussen de bacterie en het immuunsysteem.

Dit model gaat er van uit dat de bacterie exponentieel groeit in de gastheer. Tegelijkertijd echter wordt de bacterie onschadelijk gemaakt door de antistofreactie. Het onschadelijk

maken van de bacterie hangt af van de kans dat antistoffen bacteriën tegenkomen. Daarnaast hangt de productie van antistoffen af van de ontmoetingskans tussen antistoffen en bacteriën.

Dit zogenaamde roofdier-prooimodel is het meest eenvoudige model om de interactie tussen gastheer en ziekteverwekker te beschrijven.

Samenvatting

met langdurig hoesten en om te zien wat de klinische invloed is van gemengde infecties van Bordetella pertussis met andere verwekkers van luchtweginfecties werd een tweejarige

prospectieve studie opgezet.

Deze studie wordt in hoofdstuk 9 beschreven. Bij ongeveer een derde van de patiënten werd geen verwekker (n=45, 33%) aangetoond, bij een derde één verwekker (n=49, 36%), en bij een derde meerdere (n=42, 31%). De meest voorkomende verwekkers zijn rhinovirus, Bordetella pertussis en respiratoir syncytiaal virus. Bij patiënten met een gemengde infectie is de meest voorkomende combinatie Bordetella pertussis en rhinovirus. Gemengde infecties komen het hele jaar voor ongeacht het seizoen. Kinderen met meer dan een verwekker zijn niet zieker, al zijn ze wel significant ouder dan kinderen zonder aantoonbare verwekker of slechts één verwekker. Ook zijn er geen klinische gegevens gevonden die een onderscheid mogelijk maken tussen de verschillende verwekkers, de aanwezigheid van verwekkers, of verschillen in behandeling.

Het feit dat een IgG-PT titer van 100 U/ml in een enkel bloedmonster een bewijs is van een recente kinkhoestinfectie heeft de diagnostiek van kinkhoest verbeterd en vereenvoudigd.

Ook konden we aantonen dat mensen vaker dan één keer kinkhoest kunnen krijgen, gevaccineerd of niet. Tevens bleek dat hoe langer de tijd was tussen twee kinkhoestinfecties hoe zieker de patiënt

de tweede keer was. Het modelleren van het beloop van IgG-PT titers na een kinkhoestinfectie maakt het mogelijk patiënten van verschillende leeftijden, verschillende plaatsen en verschillende tijdvakken te vergelijken. Hierdoor zal onze kennis over de epidemiologie van kinkhoest verder toenemen. Wel zal er verder onderzoek gedaan moeten worden naar het mogelijke leeftijdsbepaalde beloop van IgG-PT. Het door ons gevonden leeftijdspatroon heeft invloed op de uitkomsten van studies naar het voorkomen

van kinkhoest en op de vaccinatieschema’s voor volwassenen. De snellere en hogere reactie bij ouderen op een kinkhoestinfectie betekent dat zij minder ernstig ziek worden en daardoor ook moeilijker

te herkennen zijn als een kinkhoestpatiënt. Er moet daarom extra aandacht besteed worden om kinkhoest in deze leeftijdsgroep te kunnen herkennen. Zowel bij kinkhoestinfecties als bij langdurig hoesten bestaan een groot aantal gemengde infecties. Er zijn echter geen aanwijzingen voor ernstigere ziekte bij gemengde infecties. Ook zijn veel mensen drager van een verwekker van luchtweginfecties zonder dat zij ziek zijn. Is er dan wel sprake van gemengde infecties? Meer kennis omtrent gemengde luchtweginfecties kan van invloed zijn op de behandeling van de patiënt.

Het belangrijkste is kinkhoest onder controle te krijgen ter bescherming van het jonge, niet of gedeeltelijk gevaccineerde, kind. De beste manier hiervoor is vaccinatie. Aangezien een klachtenvrije herinfectie met Bordetella pertussis al kort na een infectie of vaccinatie kann ontstaan lijkt de bescherming tegen ziek worden langer te bestaan dan de bescherming tegen infectie. Maar deze bescherming

is niet levenslang. De interactie tussen vaccinatie en het versterken van de immuniteit door natuurlijke infectie met Bordetella pertussis aan de ene kant en de verdwijnende immuniteit na verloop van tijd aan de andere kant is van invloed op de epidemiologie en de dynamiek van kinkhoest in de bevolking. Het is daarom belangrijk de door ons verzamelde gegevens te gebruiken voor het bestuderen van de invloed van veranderingen in vaccinatieschema’s op de epidemiologie van kinkhoest.

Vorwort/Suchen. Zeichen/Abkürzungen. Impressum.