Epidemiological characterisation of Bordatella pertussis in Sweden, 1970-2004

Abstract: Worldwide, Bordetella pertussis causes some 50 million cases of pertussis, and an estimated 2-300,000 fatalities each year. In Sweden vaccination against pertussis was introduced in the 1950s using diphtheria-tetanus-whole-cell pertussis vaccine (DTPw) until 1979, when it was withdrawn due to concerns about effectiveness. Thereafter, there was a vaccination-free period until 1996 when diphtheria-tetanus-acellular pertussis vaccine (DTPa) was introduced. During the period without general vaccination, pertussis had become endemic, with an overall incidence of 120-150/100,000 person years. After the introduction of the acellular vaccines this trend was changed to 7-26/100,000 person years from 1998 through 2004. Nevertheless pertussis is still a health problem in Sweden, with more than 1000 reported cases each year. The primary aim of this study was to characterise the clinical isolates of B. pertussis collected during different vaccination periods in order to find possible shifts in circulating bacteria related to vaccination. A reference system for typing B. pertussis was established based on 1810 clinical isolates. Pulsed-Field Gel Electrophoresis (PFGE) was used as the main method in combination with fimbrial serotyping and gene sequencing of two other virulence factors, pertactin (PRN) and pertussis toxin (PT). We suggested 35 Swedish reference strains for PFGE traceability. This reference material is deposited at Culture Collection University of Gothenburg. Further characterisation of 1270 Swedish B. pertussis isolates collected during periods with different vaccination programs indicated shifts with respect to the PFGE profile, fimbrial serotype and prn genotype. A large number of PFGE profiles were identified but only a few were predominant during each period, and others replaced them over time. Shifts in fimbrial serotype occurred and were directly related to vaccination policy. In the DTPw or DTPa periods, Fim3 was significantly more prevalent and in the vaccination-free period Fim2. Prn1 was predominant during the DTPw period but was replaced by prn2 during the vaccine free period and DTPa period. Prn3 was sporadically observed through the study years with a peak in 1997-1998. The circulating strains differed from the vaccine strains in the studied markers and PFGE profiles. Changes in the B. pertussis population between three noted incidence peaks, in 1999-2000, 2002 and 2004, were investigated. The data showed that one profile, BpSR11, first seen in 1997, was dominating after 1999. Other BpSR11-related profiles, BpSR5 BpSR12, appeared with an increasing trend. Although vaccination with Pa has reduced disease, new variants have emerged in populations with high vaccination coverage. We compared Finnish (n=193) and Swedish (n=455) isolates circulating in 1998-2003 together with vaccine strains used in these neighbouring countries with different vaccination histories. The results suggest that the sequential epidemics were caused by clonal expansion of B. pertussis strains possibly transmitted from Sweden to Finland. BpSR11 was seen in the Swedish 1999 peak and later in the Finnish outbreak in 2003. As early as 1994, however, it was isolated in France. We analysed the association of PFGE profile and serotype to severity of disease for all children followed during the first seven years of a surveillance project. There were in all 927 children with both clinical information and strain characterisation data available. When two groups of strains characterised by PFGE profile or serotype were compared with clinical outcome, defined as duration of hospitalisation, spasmodic cough or complications, there were significantly more children with a long duration of hospital stay in the most frequent PFGE profile group (BpSR11) compared to the PFGE group of all other profiles (p=0.041). Our data indicate that the BpSR11 clone seems to have unique properties for spread and survival.