A Longitudinal Epidemiology Study of Meningococcal Carriage in Students 13 to 25 Years Old in Quebec

Disease caused by Neisseria meningitidis is associated with serious complications and a high fatality rate. Asymptomatic individuals can harbor the bacterium in the throat, a state known as “carriage,” which can lead to person-to-person spread of the pathogen. This study examined N. meningitidis carriage from 2010 to 2013 among 2 groups in the Quebec City region: ninth-grade students (aged 13 to 15 years), who were also followed in their last year of high school (eleventh grade/college entry; 16 to 18 years), and university students (18 to 25 years); both groups have been shown in some other geographic regions to have high rates of carriage. This study demonstrated that N. meningitidis carriage rates were higher among university students in dormitories than ninth-grade and eleventh-grade/college entry students. Understanding carriage rates in these age groups leads to better strategies to control N. meningitidis by targeting vaccination to those responsible for transmission within the population.

at visit 1 and an additional 112 enrolled during the visit 2 interval. Former ninth-grade students (n ϭ 433) were invited to participate in the follow-up study in eleventh grade (cohort 1 follow-up); 363 were enrolled (median age [range], 16.0 years [16 to 18 years]). Overall, 526 ninth-grade students and 339 university students completed the initial study, and 356 of 363 eleventh-grade/college entry students completed the follow-up portion of the study. Age differences notwithstanding, demographic and clinical characteristics were generally similar between cohorts 1 and 2, with 3% to 5% of subjects recorded as nonwhite and slightly more than half being female. Most participants (Ͼ90%) had received a serogroup C conjugate vaccine during the 2001 immunization campaign; between 1.5% and 3.4% at visits 1 to 3 received antibiotics within the previous 2 weeks.
NmB carriage. Regardless of methodology, NmB carriage rates were higher in university students at all visits. NmB carriage rates determined by isolate PCR analyses were 1.9% in ninth-grade students, 1.7% among cohort 1 follow-up subjects at eleventh grade/college entry, and 6.9% in university students at any visit (Fig. 2).
The NmB acquisition rate at visits 1 to 3 per 1,000 person-months was 1.9 for ninth-grade students, 0.7 for eleventh-grade/college entry students, and 3.3 for uni- N. meningitidis Carriage in Quebec versity students. Only 1.9% (10/533) of ninth-grade students and 1.9% (7/363) of eleventh-grade/college entry students compared with 6.9% (25/360) of university students were NmB carriers at any time during the study, of whom 7, 5, and 17, respectively, were carriers at Ͼ1 visit. Seven, 5, and 15 students, respectively, were carriers at Ն2 consecutive visits.  Table 1). The most frequently detected non-NmB groups in ninth-grade, eleventh-grade/college entry, and university students were NmY, with rates at any visit of approximately 1.1% to 2.2% in each age group, and NmW, with rates of 0.3% to 0.9% in ninth-grade and eleventh-grade/college entry students and 2.5% in university students. Prevalence of NmC was very low in all age groups; no NmA or NmX isolates were detected (Table 1). Among NmC carriers, 8 subjects were carriers at Ն1 visit. All 8 had previously received N. meningitidis vaccination: 6 received NmC conjugate vaccine, 1 received N. meningitidis polysaccharide vaccine, and 1 received an unknown N. meningitidis vaccine.
Acquisition rates at visits 1 to 3 for non-NmB isolates were 0.7 per 1,000 personmonths for both ninth-grade and eleventh-grade/college entry students and 1.0 per 1,000 person-weeks for university students. Acquisition rates for nongroupable isolates were 1.6 and 3.6 per 1,000 person-months for ninth-grade and eleventh-grade/college entry students, respectively, and 1.5 per 1,000 person-weeks for university students.
All meningococcal carriage. Overall carriage prevalence by isolate PCR remained relatively consistent across time points within cohorts (Table 1). Prevalence across visits ranged from 6.1% to 6.9% among ninth-grade students, from 5.0% to 6.2% among eleventh-grade/college entry students, and from 19.1% to 24.2% among university students.
Comparison of meningococcal detection and grouping methods. NmB carriage rates determined by seroagglutination, isolate PCR analyses, and direct swab PCR were 0.8%, 1.9%, and 1.9% in ninth-grade students and 4.7%, 6.9%, and 6.1% in university students at any visit, respectively ( Fig. 2A and C). NmB carriage rates among eleventhgrade/college entry students as determined by whole-genome sequencing (WGS), isolate PCR, direct swab PCR, and live cell phenotypic assay (LCPA) were 1.9%, 1.7%, 3.3%, and 0.8% at any visit, respectively (Fig. 2B). For cohorts 1 (ninth grade) and 2 (university students), NmB detection was highest by isolate PCR, followed by direct swab PCR and seroagglutination (Fig. 3). Cross-sensitivity analysis showed in each age group that direct swab PCR detected Ն50% of NmB isolates detected by the other methodologies, and isolate PCR detected Ն75% of NmB isolates detected by the other The numbers of instances in which NmB was detected for subjects by seroagglutination, isolate PCR, direct swab PCR (ninth-grade and university students); by isolate whole-genome sequencing (WGS), isolate PCR, direct swab PCR, and live cell phenotypic assay (LCPA) (eleventh-grade/college entry students); and by multiple methods are shown as Venn diagrams for ninth-grade students (A), eleventh-grade/college entry students (B), and university students (C).
methodologies. Only 2 of 23 NmB isolates from ninth-grade students and 16 of 58 NmB isolates from university students were detected by all 3 methodologies. Because isolate PCR provided a more sensitive method for determining meningococcal serogroup among carriage isolates than did seroagglutination, rates of nongroupable isolates at any visit were higher by seroagglutination than by isolate PCR (Table 2). These results are not unexpected, given that seroagglutination is a phenotypic rather than a genotypic assay. Among eleventh-grade/college entry students, WGS (n ϭ 13), isolate PCR (n ϭ 13), and direct swab PCR (n ϭ 15) each detected Ն60% of the overall number of NmB isolates detected by any methodology, whereas LCPA (n ϭ 6) detected Ͻ30%. Only 4 of 21 NmB isolates were detected by all 4 methodologies (Fig. 3).

Multilocus sequence typing (MLST) and fHBP analyses of NmB and non-NmB isolates.
Of the 48 isolates identified as NmB and subjected to whole-genome sequence analysis followed by MLST, 31.3% were typed as ST-41/44, 14.6% as ST-269, 14.6% as ST-32, and 6.3% as ST-461 complex (Fig. 4A). Nine isolates belonged to STs that have not been mapped to an existing clonal complex (CC).
At any visit, 92.3% of the 45 NmB isolates from ninth-grade students encoded fHBP subfamily A variants, as did 85.7% of isolates from the eleventh-grade/college entry students, as well as from university students (proportions of individual variants are shown in Fig. 4A). The predominant fHBP variant was A22 (33.3%); all other subfamily A variants were found in Ͻ10% of isolates with A10, A19, and A20 the next most common. Only one of the isolates (2.6%) from ninth-grade students was from fHBP subfamily B (B44), compared with 14.3% of isolates each from eleventh-grade/college entry students and the university students.

DISCUSSION
This study, focused mainly on NmB, assessed meningococcal carriage longitudinally in ninth-grade and eleventh-grade/college entry students versus university students living in residence halls and is important for understanding the epidemiology of meningococcal carriage and disease. The 2 subject groups parallel the ages preceding and corresponding to a peak in meningococcal disease incidence (6). IMD incidence among 15-to 19-year-olds in Canada was approximately 0.5 per 100,000 in 2009, with rates for NmB disease reaching nearly 0.4 per 100,000 (14). Among the same age group in Quebec in 2011, IMD incidence was 2.6/100,000 overall and 2.4/100,000 for NmB (2). In Quebec City in 2013 (last year of the study), NmB IMD incidence was 4.8 per 100,000 among those Ͻ20 years old, compared with 1.6 per 100,000 in Quebec for the same age group (15).
The study was conducted during circulation of the virulent NmB ST-269 clone (2-4),  (3,14,16). These results are the only N. meningitidis carriage data in Quebec and may serve as baseline data for investigating NmB vaccination effects on carriage in the target age group. In the only previous study of meningococcal carriage in Canada, during a 2001 outbreak of NmC IMD in British Columbia, the overall carriage rate in persons aged 11 to 55 years was 7.6%, with a significantly lower rate in adolescents aged 11 to 12 years (1.2%) than in those aged 13 to 29 years (8.0%) (17). Higher rates in our study may be due to epidemiological, methodological, geographical, or temporal differences and/or random sampling variation.
In this study, NmB carriage rates in university students (6.9%) were higher than those in ninth-grade (1.9%) and eleventh-grade/college entry students (1.7%). This pattern is consistent with results from previous studies in the United Kingdom, where carriage prevalence among subjects aged 19 to 25 years was 6.5% (6,18), suggesting an optimal vaccination window between ninth grade and university entry for future studies to assess prevention of meningococcal carriage and thus subsequent disease. NmB acquisition occurred relatively infrequently (Ͻ2 per 1,000 person-months for ninth-grade and eleventh-grade/college entry [13-to 18-year-old] students and 3.3 per 1,000 person-months for university [18-to 25-year-old] students), which is comparable to the rate of 2.8 per 1,000 persons-months estimated in 10-to 25-year-old students in the United Kingdom study (18  Results are presented for ninth-grade, eleventh-grade/college entry, and university students. Non-NmB isolates (B) from all 3 visits (n ϭ 143) were characterized by PCR (serogroup assignment, except eleventh grade by WGS) and sequence analysis (fHBP assignment). NT, nontypeable; NG, nongroupable. carriage (NmB identification at Ն2 consecutive visits) was observed in 7 ninth-grade students, 5 eleventh-grade/college entry students, and 15 university students. Loss of NmB carriage was frequent, as 25% to 40% of NmB-positive students became negative for carriage during the study. Among non-NmB meningococci, NmY and NmW were most frequently detected; NmY prevalence was comparable to that of NmB in ninthgrade and eleventh-grade/college entry students but was 3 times lower (2.2%) in university students than NmB prevalence (6.9%). These results differ from the United Kingdom study reported by Jeppesen and colleagues, wherein NmY prevalence increased with age (18). Non-NmB meningococcal acquisition also occurred relatively infrequently in both cohorts, but loss of carriage was more common. Although non-NmB carriers were more common than NmB carriers, the proportion of subjects becoming negative for carriage was lower for non-NmB than for NmB. These results are important in light of an English study in which MenB-4C vaccination did not statistically alter carriage prevalence of disease-associated NmB sequence types (ST-41/44, 32, and 269) in vaccinated university students 1 month after dose 2 (19). Administration of dose 1 took place across a 3-month enrollment period; because carriage acquisition was highest between the first 2 visits, vaccination may have occurred too late to observe the greatest effects on carriage.
Classical NmB identification methodology includes seroagglutination, which was the gold standard at the time of study design, and PCR of cultured isolates, which may be preferred based on demonstrated greater sensitivity in the current study. Direct swab PCR analysis, which does not depend on successful isolate culturing, was included to determine whether this technique may be a more efficient alternative to culture-based methods. For ninth-grade students, NmB carriage rates were similar for isolate and direct swab PCR, but for university students, isolate PCR was more sensitive. Differences between these methods may be attributed to the enrichment step associated with isolate culturing. Direct swab PCR may offer little benefit in most diagnostic settings because the isolate is not available for repeat testing but may be useful when culturing at the source is not feasible.
Isolate PCR also identified fewer isolates as nongroupable meningococci than did seroagglutination, likely due to some isolates not expressing capsule. Moreover, seroagglutination assays are somewhat subjective because of reliance on visual inspection of agglutination intensity by an operator. In addition, seroagglutination reagent usage is not standardized, and reagent availability may be inconsistent. Seroagglutination is useful for characterization of invasive isolates, which most often express capsule, but may be less effective for serogrouping carriage isolates, which do not express capsule as frequently (20).
This study is the first to report fHBP variants in NmB carriage isolates in healthy subjects in Canada. In the United States, fHBP variants from subfamily A are most frequently associated with carriage in healthy adolescents and young adults, regardless of capsule locus, whereas subfamily B variants cause invasive disease most frequently in this age group (21,22). In Canada, subfamily B variants generally cause most invasive disease in all age groups except infants (3). However, differences occur by province, with an overall predominance (including infants) of subfamily A in Ontario and subfamily B (except a predominance of subfamily A in infants) in Quebec (R. S. W. Tsang, F. B. Jamieson, B. Lefebvre, R. Gilca, S. Deeks, P. De Wals, P. Rawte, C. Tremblay, D. Law, J. Zhou, and S. Deng, 7th Vaccine and ISV Congress, poster P042, 2013). The large majority of carriage isolates in our study contained fHBP from subfamily A (A22), which is consistent with results for adolescents in the United Kingdom in which nearly 90% of NmB isolates had subfamily A variants. As carriage is considered an immunizing event, the predominance of subfamily A strains in carriage may reduce subfamily A disease in immunocompetent populations and yet result in more disease in susceptible populations such as infants.
A limitation of this study is the homogeneous study population, which should be noted when considering applicability of results to other populations. Moreover, the reported carriage rate in Quebec City may not be representative of other regions of Quebec and Canada, as has been shown for IMD isolates. Also, the population of university students in dormitories may not be representative of other populations of young adults in nonuniversity settings. In addition, the number of NmB carriers was relatively small and visits were widely spaced, preventing detection of short-term carriers. However, our study represents the largest longitudinal data set on meningococcal carriage in Canada and suggests that carriage can persist for several months, which is consistent with previous reports that 25% to 45% of carriers are persistent carriers for at least 5 to 6 months (23-25) and that 90% of persistent carriers retain the same meningococcal clone for 5 to 6 months (26).
Conclusions. This study informs the design of future studies assessing the effect of NmB vaccination on meningococcal carriage prevalence, its potential effect on herd immunity, and subsequent impact on IMD incidence. In this study, WGS of culture isolates detected more NmB than seroagglutination, isolate PCR, or direct throat swab PCR assays, suggesting the potential for a new standard for detecting N. meningitidis in future investigations of throat carriage. Additional research is also needed to better understand the significance of differences in distribution of NmB fHBP subfamilies and variants in carriage.

MATERIALS AND METHODS
Study design. This longitudinal epidemiology study was conducted at Centre Hospitalier Universitaire (CHU) de Québec in Quebec City, Canada, between November 2010 and February 2012, with a follow-up between February 2013 and December 2013. The study was approved by the Institutional Review Board of CHU de Québec. Written informed consent was obtained from each subject or a legally acceptable representative. Between November 2010 and February 2012, enrolled subjects completed 3 office visits, followed by 3 follow-up office visits for a subset of subjects between February and December 2013 (Fig. 1).
Study subjects. Participants were recruited among students attending ninth-grade classes in secondary schools (cohort 1; 13 to 15 years of age at enrollment) and among those living in dormitories at universities (cohort 2; 18 to 25 years of age at enrollment) in Quebec City. Additional follow-up was conducted for a subset of cohort 1 subjects at eleventh-grade entry (16 to 18 years of age).
Study objectives. The primary objective was to estimate NmB throat carriage prevalence in ninth-grade students, in the same students in eleventh-grade/at college entry, and in university students living in dormitories by throat swab culture and real-time PCR-based analysis of cultured isolates at 3 time points across 5 to 8 months. Samples from eleventh-grade/college entry students were analyzed by WGS.
Procedures. Ninth-grade students were enrolled at 2 separate periods. Those enrolled during the first period (November to December 2010) had 3 visits, at enrollment on day 1 and approximately 3 and 6 months later; additional subjects enrolled during the second period (January to March 2011) had 2 visits. University students visited at enrollment on day 1 and approximately 6 and 20 weeks later (Fig. 1). Subjects participated for approximately 6 months. Ninth-grade students were invited to participate in a follow-up study at eleventh-grade entry. Eleventh-grade students visited at enrollment on day 1 and approximately 3 and 8 months (at college entry) later.
Two throat swabs were collected simultaneously at each office visit. One swab (culture swab) was cultured for detection and identification of Neisseria species at the CHU de Québec laboratory. Isolates were characterized using WGS or PCR (isolate PCR) to identify common meningococcal epidemiological markers (described in detail below) and serogrouped by standard seroagglutination testing (27). The second swab was placed in Digene specimen transport medium (Qiagen, Germantown, MD; not cultured) for direct PCR-based detection of Neisseria. Direct PCR analysis of storage solution from uncultured swabs and isolate PCR was conducted at a central laboratory (Pharmaceutical Product Development, LLC, Wayne, PA). Isolate WGS, MLST, and LCPA analyses were conducted at Pfizer (Pearl River, NY). Seroagglutination assays were conducted at the CHU de Québec laboratory (Table 3).
Microbiological analysis. Specimens were collected by simultaneously swabbing the tonsils or tonsillar fossa and posterior pharynx. Immediately afterwards, the culture swab was plated directly onto Thayer-Martin improved medium. Within 5 h, the plates were transferred to an incubator (35°C, 5% CO 2 ) and monitored for up to 72 h. Colonies suspected to be N. meningitidis were subcultured on blood agar. Suspected single N. meningitidis colonies were identified as Neisseria species by oxidase testing, Gram staining, and biochemical identification (28) using the API NH kit (bioMérieux, St Laurent, QC, Canada). Serogrouping was performed by slide-agglutination as described previously (27). PCR analysis. Real-time PCR assays (29) were conducted using TaqMan primer sets (Life Technologies, Burlington, ON, Canada) for each of the 8 capsule-specific genes of interest for N. meningitidis (NmA, NmB, NmC, NmE, NmW, NmX, NmY, and NmZ). PCR assays were additionally qualified for porA and ctrA. Based on PCR results from swab culture isolates (isolate PCR) and direct swab PCR, samples were grouped into 5 categories: all meningococci, grouped meningococci, nongroupable meningococci, non-NmB meningococci, and group B meningococci (Table 4). Direct swab PCR samples were genogrouped only for NmB. Standard PCR amplification and sequencing of the fHBP gene were also performed for NmB isolates as previously described (30).

Additional genotypic and phenotypic analyses.
Characterization of the follow-up cohort meningococcal isolates (eleventh-grade/college entry) was performed by WGS. In total, 45 unique NmB isolates, representing all 3 cohorts, were further characterized by WGS and LCPA. Detailed information about WGS and LCPA analyses is provided elsewhere (20); briefly, LCPA analyses were used to determine meningococcal serogroup by bioluminescent detection of serogroup-specific monoclonal antibody binding to isolates. MLST data were obtained as previously described (31).
Statistical analysis. Based on CI estimates for various prevalence rates and sample sizes, the target enrollment was a minimum of 500 to 750 ninth-grade students and 200 to 350 university students. A convenience sample size of 360 eleventh-grade students from the former ninth-grade students was selected a priori.
The intent-to-treat population (all enrolled subjects) was used for epidemiologic endpoint analyses. Confidence intervals were calculated using the exact method based on Clopper-Pearson (2-sided). The  McNemar test using the exact method was used to compare the prevalence rates between PCR analyses and seroagglutination for each visit in ninth-grade/university students. The NmB acquisition rate was defined as the number of any new NmB carriage cases in the population in a given period of time. The rate was calculated by dividing the number of new carriage cases over time by the sum of the person-time (person-time was calculated as the sum of all initial negation subjects' duration in the study). Data from isolate PCR analyses were used to determine acquisition rates for ninth-grade and university students, and data from WGS analysis were used to determine rates for eleventh-grade/college entry students.