HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via CrossRef Citing Articles via Google Scholar Google Scholar Articles by TROTTER, C. L. Articles by KACZMARSKI, E. B Search for Related Content PubMed PubMed Citation Articles by TROTTER, C. L. Articles by KACZMARSKI, E. B Agricola Articles by TROTTER, C. L. Articles by KACZMARSKI, E. B

Fatal outcome from meningococcal disease – an association with meningococcal phenotype but not with reduced susceptibility to benzylpenicillin

Immunisation Division, PHLS Communicable Disease Surveillance Centre, London and *PHLS Meningococcal Reference Unit, Manchester, UK

Corresponding author: Ms C. Trotter (e-mail: Ctrotter{at}phls.org.uk).

Received 24 Jan. 2002; revised version received 30 March 2002; accepted 7 April 2002.

	Abstract

TOP Abstract Introduction Materials and methods Results Discussion References

 
Penicillin has been the mainstay of treatment for meningococcal disease. Isolates of Neisseria meningitidis that are less susceptible to penicillin have been reported in several countries and in recent years have become more common. The clinical significance of this reduced susceptibility has not been investigated on a large scale. Hence, N. meningitidis isolates from culture-confirmed cases of meningococcal disease in England and Wales, between 1993 and 2000, were routinely serogrouped, serotyped and tested for susceptibility to penicillin. These data were linked to death registrations and analysed retrospectively. The changing trends in susceptibility were described and multivariate logistic regression was used to examine associations between strain characteristics and fatal outcome. The frequency of N. meningitidis isolates less susceptible to penicillin increased from <6% in 1993 to >18% in 2000. In particular, isolates expressing serogroup C with serotype 2b and serogroup W135 had a higher frequency of reduced penicillin susceptibility (49% and 55%, respectively). There was no evidence of an association between fatal outcome and infection with a less penicillin-susceptible isolate. Fatal outcome was associated with serogroup and serotype, with the odds of death for cases infected with C:2a and B:2a strains three-fold higher when compared with the baseline. For this large dataset the serogroup and serotype of the infecting strain influenced mortality from meningococcal disease and may be markers for hypervirulence. No association was found between reduced penicillin susceptibility and fatal outcome, but the increasing frequency of isolates less susceptible to penicillin highlights the need for continued surveillance.

	Introduction

TOP Abstract Introduction Materials and methods Results Discussion References

 
Penicillin has been the mainstay of treatment for meningococcal disease in England and Wales and remains the treatment of choice in many countries. Isolates of Neisseria meningitidis described as ‘moderately resistant’ to penicillin (usually defined as penicillin MIC of 0.1–1.05muµg/ml) have been described since the mid-1980s and are now widely reported [1–4]. In several countries (e.g., the UK, Spain, Australia) less susceptible isolates have become increasingly common [5–7], but most isolates would be expected to remain susceptible to appropriate therapeutic doses of penicillin [8]. Treatment failure has been reported rarely [9, 10], and the only account of a treatment failure in the UK [9] concerns a patient treated with a low dose of penicillin. However, detailed case studies of children infected with isolates with reduced susceptibility to penicillin suggest that these infections result in higher mortality [11] and, therefore, reduced penicillin sensitivity may be very important clinically.

In the UK in the 1990s, >200 deaths per year were caused by meningococcal disease and the national case fatality rate based on routine reports remained high (c. 8%). Factors that have been shown to affect the outcome of meningococcal disease include age [12, 13], clinical presentation (with a higher case fatality rate associated with septicaemia than meningitis [12–14]), treatment factors [15, 16] and the characteristics of the infecting organism, including serogroup and serotype [12, 14, 17, 18]. Interventions to improve the outcome of meningococcal disease include the encouragement of early administration of penicillin [19]. ‘Hypervirulent’ meningococcal strains (i.e., those associated with a more severe outcome, including death) are characterised by their periodic emergence in association with recognised phenotypic markers, i.e., serogroup, serotype and sero-subtype [14].

This study was prompted by the absence of reported population-based analyses assessing the impact of reduced penicillin susceptibility on the outcome of meningococcal disease. It also aimed to determine whether there was an association between hypervirulent strains and reduced penicillin susceptibility. The PHLS Meningococcal Reference Unit (MRU) routinely tests isolates from cases of meningococcal disease in England and Wales for susceptibility to antibiotics. Data on patient characteristics (age, sex, region) and isolate characteristics (serogroup, serotype and sero-subtype) are also recorded. Deaths from meningococcal disease are registered with the Office for National Statistics (ONS), and these were linked to MRU data. All cases of meningococcal disease that were confirmed by culture at MRU between 1993 and 2000 were analysed retrospectively to describe the epidemiology of less penicillin-susceptible isolates and to explore any associations between reduced penicillin susceptibility, phenotype and fatal outcome.

	Materials and methods

TOP Abstract Introduction Materials and methods Results Discussion References

 
Data sources

All cases of culture-confirmed meningococcal disease between Jan. 1993 and Dec. 2000 for which penicillin MICs had been measured were identified from the MRU database. Data included age, sex and region of patient, date of receipt of isolate, isolate characteristics and penicillin MIC. The threshold for an isolate to be considered as having reduced susceptibility to penicillin was an MIC of 0.15muµg/ml, and the result was coded accordingly as full or reduced susceptibility. To examine isolate characteristics, the most common strains (defined by serogroup and serotype of the isolate) were identified, to give 15 strain types, with the less common isolates grouped together. The MRU data were linked to ONS death registrations by matching records on name, date of birth, sex and region.

Determination of benzylpenicillin MICs of N. meningitidis isolates

The MIC of benzylpenicillin of N. meningitidis isolates submitted to MRU was determined by the agar incorporation method. Briefly, sterile serial doubling dilutions (0.005–5.0 mg/L) of benzylpenicillin (Sigma) were made. The antibiotic was added to pre-prepared agar volumes to achieve the desired appropriate dilution in a final volume of 20 ml. The culture medium used was Columbia blood agar (CBA; Oxoid) supplemented with defibrinated horse blood 5%. Light suspensions of pure viable cultures were further diluted 100-fold before inoculation on to the pre-dried agar plates. Batches of 20 test and 5 control organisms were tested on each set of plates. The CBA plates were inoculated and allowed to dry before incubation at 37°C for 48 h. The MIC was recorded as the dilution at which <10 colonies were observed. Results were accepted if the MICs of the controls were within one dilution of their mean value. Organisms were categorised according to the MIC as follows: fully susceptible <0.1 µg/ml, reduced susceptibility 0.1–1 µg/ml, resistant >1 µg/ml.

Phenotyping

All isolates were characterised by serogroup determination by co-agglutination with polyclonal antisera and serotyped and sero-subtyped by whole-cell dot-blot with serotype and sero-subtype monoclonal antibodies as described previously [20].

Statistical analysis

Differences in the geometric mean penicillin MICs and the frequency of less susceptible isolates over time were investigated by the t test. Associations between penicillin susceptibility, fatal outcome, time, patient and strain characteristics were investigated with unconditional logistic regression in STATA. In the first instance, susceptibility to penicillin (fully susceptible versus reduced susceptibility) was explored by including age group, year and strain characteristics in the logistic regression model. Secondly, factors affecting fatal outcome were investigated by including age group, year, susceptibility to penicillin and strain characteristics in the main model. To test for an association between increasing MIC and fatal outcome, susceptibility to penicillin was considered as a categorical variable in an alternative model.

	Results

TOP Abstract Introduction Materials and methods Results Discussion References

 
There were 11 459 records from cases of meningococcal disease between 1993 and 2000 for which penicillin MICs were available. The age distribution of cases followed the pattern typical of meningococcal disease, with the highest incidence in infants and the under-5-year-olds and a smaller secondary peak in those aged 15–19 years. There was an excess of male cases, particularly in the younger age groups. Most of the isolates were serogroup B (60%) or serogroup C (34%), with other serogroups isolated rarely (W135 2.5%, Y 1.6%, NG 1.1%, Z, A and 29E <1%). The most common phenotypes isolated between 1993 and 2000 were B:4:NT/P1.4/NT (n=1155), C:2a:P1.5/NT/NT (n=1150) and B:NT:NT/P1.15/NT (n=948). Overall 974 (8.5%) cases had a fatal outcome.

Reduced sensitivity to penicillin (MIC=0.15muµg/ml) was found in 1447 (12.6%) isolates over the study period. The geometric mean penicillin MIC increased significantly from 0.057 µg/ml in 1993 to 0.074 µg/ml in 2000 (p <0.0001) and the frequency of less susceptible isolates increased from <6% in 1993 to >18% in 2000 (p <0.001) (Fig. 1).

View larger version (22K): [in this window] [in a new window]   Fig. 1. Proportion of isolates less susceptible to penicillin () and geometric mean inhibitory concentrations (•) by year.

Isolates with reduced susceptibility to penicillin were more likely to be associated with specific serogroups and serotypes (Table 1). In particular, isolates identified as serogroup C in combination with serotype 2b, and serogroup W135 had a much higher frequency of reduced penicillin susceptibility (49% and 55%, respectively, compared with 13% in the whole sample, p <0.0001 in each case). However, the frequency of these phenotypes among cases of meningococcal disease was relatively low (5.5% of all isolates were C:2b, 2.5% were W135).

There was no evidence of any association between fatal outcome and reduced susceptibility to penicillin (Table 2). When the association between fatal outcome and penicillin MICs was examined, there was a suggestion that the odds of death were increased when the MIC was >0.64 (OR=2.2), but the confidence interval was very wide and did not approach statistical significance. A test for trend was also non-significant at the 5% level (p=0.083).

Fatal outcome varied significantly by serogroup and serotype (p <0.001, Table 1). Overall, relative to group B, the odds of a fatal outcome were significantly higher for serogroup C cases (OR=1.71, 95% CI 1.48–1.98) and cases infected with other serogroups (OR=1.07, 95% CI 0.79–1.44). Isolates identified as C:2a and B:2a gave the highest odds of death, almost three-fold higher than the baseline (group B:1), and B:15 isolates were also significantly associated with a fatal outcome.

Older patients were more likely to die than younger patients, with the lowest risk of death in the 5–9 age group (Fig. 2). The odds of death did not vary substantially between 1993 and 1998 but were significantly lower in 1999 and 2000.

View larger version (9K): [in this window] [in a new window]   Fig. 2. Odds of a fatal outcome in cases of meningococcal disease by age group (adjusted for strain characteristics and year) (OR, ; 95% CI, uu. uu. uu. uu.).

	Discussion

TOP Abstract Introduction Materials and methods Results Discussion References

 
The study found no association between reduced penicillin susceptibility and fatal outcome, despite the increasing frequency of isolates with reduced penicillin susceptibility over time. A recent study in Australia [21] also confirms this lack of association. The implication is that penicillin remains a suitable treatment for meningococcal disease. This may be especially relevant in countries where third-generation cephalosporins are not widely available. However, given the increasing frequency of less susceptible isolates and the increasing geometric mean MIC, the antibiotic susceptibility profiles of N. meningitidis isolates should continue to be monitored.

Phenotypic characteristics, i.e., serogroup and serotype, appear to be good markers for both reduced susceptibility to penicillin and hypervirulence. The proportion of isolates with reduced penicillin susceptibility was much higher for C:2b and W135 compared with other strains. These associations have also been shown in Canada (W135) [22] and Spain (C:2b) [23]. Phenotype C:2a appears to be a marker for hypervirulence, with higher fatality rates also associated with W135 isolates (which express the serotype 2a). Other studies in Norway [12] and France [17] in the 1990s have also associated C:2a strains with an increased risk of death, but this was not shown in a retrospective study of meningococcal disease between 1959 and 1981 in the Netherlands [14]. In the early 1990s a virulent serogroup C strain emerged in Canada associated with serotype 2a [24], which was identified by multilocus enzyme electrophoresis (MLEE) as ET-15 of the ET-37 complex, and later by multilocus sequence typing as ST-11. Genotyping of the representative isolates referred to MRU shows that the hypervirulent C:2a phenotype identified here is also primarily ST-11.

It is assumed that the strains examined here are representative of the entire population of meningococcal strains that cause invasive disease in the UK. However, not all cases of meningococcal disease are confirmed in the laboratory. Although there is no evidence to suggest this is the case, bias could be introduced if there are major differences in disease-causing strains that are diagnosed clinically compared with those that are confirmed in the laboratory. Over the study period 16 026 isolates were submitted to the MRU, of which 72% were included in this analysis. Since 1996, MRU has offered a service to identify isolates by the PCR, but isolates confirmed by PCR alone were excluded from this study as they were not tested for penicillin susceptibility. Again, bias could be introduced if those isolates that are only PCR-positive are different to those that are confirmed by culture; however, this was judged to be unlikely.

Meningococcal infection is increasingly confirmed by non-culture methods [25], for which no antibiotic sensitivity data are available. One of the mechanisms of decreased susceptibility to penicillin in meningococci has been identified as sequence changes in the gene coding for the penicillin-binding protein (PBP-2 gene, or penA) resulting in a reduced affinity for penicillin [26]. Molecular techniques to identify the genetic basis of reduced susceptibility to penicillin include PCR [27] and improved strain characterisation by means of nucleic acid sequence typing schemes such as multilocus sequence typing (MLST) [28].

The use of genotypic classifications is also critical in defining new strains and mapping their clonal origin, especially compared with phenotypic markers, which only describe surface antigens, many of which are under selective pressure. However, it may be some time before molecular methods can replace traditional phenotyping methods – especially in the routine surveillance of meningococcal disease – so identifying phenotypic markers of hypervirulence can provide valuable information.

The identification of hypervirulent strains may help to inform treatment and control strategies. The introduction and spread of the C:2a, ET-15 clone in the UK in the mid-1990s was associated with large increases in disease incidence. This led to the accelerated testing, licensing and introduction of the UK meningococcal group C conjugate vaccination programme [29]. Recent studies have shown that the case fatality rate of meningococcal disease can be reduced by improved clinical management in specialised centres [15, 16]. However, attempts to further reduce mortality by better case management rather than by vaccination could be limited by the virulence of the infecting strain. In countries where this hypervirulent clone has not yet become established, monitoring of disease-causing phenotypes could act as an early warning system to trigger a vaccination campaign should these clones appear.

	Acknowledgments

 
The authors would like to thank Nick Andrews for statistical advice and helpful discussions and Professor Norman Noah for his comments on the manuscript.

	References

TOP Abstract Introduction Materials and methods Results Discussion References