Burkina Faso has 2 complementary systems of population-based meningitis surveillance (5). Surveillance for reportable diseases is conducted by the Télégramme Lettre Official Hebdomadaire (TLOH), to which district-level aggregate reports of clinically defined meningitis cases and meningitis-related deaths are transmitted weekly. Enhanced surveillance for Maladies à Potentiel Epidémique (MPE) collects case-level demographic information and results of cerebrospinal fluid (CSF) examination and laboratory testing. For this analysis, we used data from TLOH for 2007–2012 and MPE enhanced surveillance data for 2010–2012. All 63 districts reported suspected meningitis cases to both surveillance systems during 2010–2012. Aggregate laboratory testing results for the 2007–2009 meningitis seasons and enhanced surveillance data for the 2002 meningitis season also were analyzed.

We used a modified case definition to classify cases. Suspected cases of meningitis were defined as sudden onset of fever with a stiff neck or, in infants, a bulging fontanelle. Probable bacterial meningitis was defined as a suspected case for which gram-stained CSF was positive (gram-negative diplococci, gram-positive diplococci, or gram-negative bacilli); probable meningococcal meningitis cases are limited to those with gram-negative diplococci. A confirmed case of meningitis was defined as a suspected or probable case for which N. meningitidis, Haemophilus influenzae type b or Streptococcus pneumoniae was isolated in culture from blood or CSF, or antigen was detected in CSF by latex agglutination or PCR. Latex agglutination–positive cases that were later negative by PCR or culture were reclassified as suspected cases.

Beginning in 2010, real-time PCR capacity was implemented at the national reference laboratory, and detection of N. meningitidis, H. influenzae type b, or S. pneumoniae genetic material by PCR was deemed confirmatory (16,17). Serogroup was determined by antigen detection or by PCR, with PCR deemed definitive. Because serogroup Y is seen infrequently as a cause of disease in Burkina Faso, all specimens positive for serogroup Y/serogroup W by latex agglutination and not tested by PCR or culture were assumed to be serogroup W.

The meningitis season in Burkina Faso is limited to epidemiologic weeks 1–24. The pre–PsA-TT implementation period was defined as January 1, 2007–December 31, 2010, and the post–PsA-TT implementation period as January 1, 2011–December 31, 2012. Incidence rates were cumulative yearly incidence per 100,000 population and were calculated with national, regional, and district population estimates from the Institut National de la Statistique et de la Démographie. Districts crossed the World Health Organization weekly epidemic threshold when the weekly incidence rate exceeded 10 suspected cases per 100,000 population (18).

Multilocus sequence typing, porin A (PorA), and ferric enterobacin transport (FetA) genotyping was performed on 61 serogroup W isolates (6 from 2011 and 55 from 2012). Multilocus sequence typing was performed by sequencing the internal fragment of 7 housekeeping genes (19). PorA and FetA genotyping was performed by sequencing the variable regions (VRs) of porA (VR1 and VR2) and the VR of fetA. All generated sequencing data were analyzed by using the Web-based software MGIP (http://mgip.biology.gatech.edu/home.php). Molecular types (sequence type [ST], clonal complex [cc], PorA, and FetA) were determined by searching the PubMLST database (http://pubmlst.org). Strain genotype was defined as ST(cc):PorA:FetA.

Figure 1

Figure 1. Laboratory-confirmed bacterial meningitis cases and Neisseria meningitidis serogroup W cases, Burkina Faso, 2011 and 2012 meningitis seasonsData source: Maladies à Potentiel Epidémique (case-based surveillance).

During January 1–December 31, 2012, a total of 7,022 suspected meningitis cases (739 [11%] fatal) were reported through TLOH in Burkina Faso, corresponding to an incidence of 42.0 cases per 100,000 population. A total of 5,807 cases were reported through MPE case-based meningitis surveillance; 2,353 (41%) were confirmed, 415 (7%) were probable, and 3,039 (52%) were suspected meningitis cases. Of the 2,353 confirmed cases, 591 (25%) were confirmed by culture, 1,741 (74%) by PCR, and 21 (<1%) by latex agglutination. The predominant organism identified was N. meningitidis serogroup W (Table 1). More meningitis cases were reported during the 2012 than the 2011 meningitis season, and a higher proportion of cases were caused by serogroup W in 2012 than in 2011 (Figure 1).

Of persons with N. meningitidis serogroup W infection during 2012, a total of 11% were <1 year of age, 29% were 1–4 years, 29% were 5–9 years, 16% were 10–14 years, and 15% were >15 years. This distribution is similar to that for serogroup W cases during the 2002 serogroup W epidemic, according to enhanced surveillance data (serogroup W: 16% persons were <1 year of age, 28% were1–4 years, 31% were 5–9 years, 8% were 10–14 years, and 17% were >15 years). The incidence of laboratory-confirmed serogroup W was 8.7 cases per 100,000 population in 2012, compared with 0.7 per 100,000 in 2011. During 2007–2010, when <15% of cases were laboratory confirmed, the incidence of serogroup W was <0.1 cases per 100,000 population (Table 2).

Figure 2

Figure 2. Total meningitis cases reported in Télégramme Lettre Official Hebdomadaire and laboratory-confirmed Neisseria meningitidis serogroup W cases for selected districts, Burkina Faso, 2012 meningitis seasonA) District 1, population 150,000B) District 2, population...

The incidence of serogroup W varied by district, from 0.0 to 40.7 cases per 100,000 population (median 6.1/100,000) in 2012. By week 24, thirteen districts had crossed the weekly epidemic threshold of 10 suspected cases per 100,000 population (selected districts, Figure 2); N. meningitidis serogroup W was the predominant serogroup identified in 12 of these 13 districts (N. meningitidis serogroup X was the predominant organism identified in the 13th district). In no district did the cumulative incidence rate reach >100 cases per 100,000 population, the established definition of an epidemic, during 2012 (20). Three districts had reactive vaccination campaigns that used quadrivalent (A, C, W, Y) polysaccharide vaccine; in 2 of these districts, the number of cases crossed the weekly epidemic threshold. These districts were 2 of the last to cross the weekly epidemic threshold in 2012 during weeks 14 and 15. Cumulative incidence of serogroup W cases in these districts was 11.0 and 23.2 cases per 100,000 population, respectively.

We characterized 61 serogroup W isolates: 6 from 2011 and 55 from 2012, all belonging to cc11. The predominant genotype detected was ST-11(cc11):P1.5,2:F1–1, which accounted for 90% (55/61) of the total isolates. The remaining isolates differed only in the FetA and ST (ST-11[cc11]:P1.5,2:F6–3, 2/61 [3%]; ST-2961[cc11]:P1.5,2:F1–1, 2/61 [3%)]; and ST-9766[cc11]:P1.5,2:F1–1, 2/61 [3%]. Three STs were detected: ST-11 (57/61 [93%]), ST-2961 (2/61 [3%]), and ST-9766 (2/61 [3%]).

In 2012, the second year after PsA-TT introduction in Burkina Faso, the continued absence of serogroup A demonstrated the effect of this enormously successful vaccination campaign. N. meningitidis serogroup W was the predominant organism causing meningitis during the 2012 epidemic season. However, rates of serogroup W disease were substantially lower in 2012 (8.7 cases/100,000 population [Table 2]) than during the 2002 epidemic, when attack rates were ≈100–250 cases per 100,000 population (10,21).

The serogroup W clone associated with the Hajj-related outbreak during 2000 (22) has been present and a persistent cause of sporadic disease in Burkina Faso since the 2002 epidemic (R. Ouédraogo-Traoré, pers. comm.). The increase in serogroup W during 2012 might represent a natural increase in disease associated with waning population immunity in the 10 years since the last epidemic. This hypothesis is supported by the high proportion of serogroup W infections among younger children and the characterization of most serogroup W strains from 2012 as ST-11(cc11):P1.5,2:F1–1, the same genotype as the clone that caused the 2002 epidemic (11,22).

In Burkina Faso, rates of nasopharyngeal carriage of serogroup W are low (23). A previous study from Burkina Faso during the 2002 serogroup W epidemic demonstrated that serogroup W carriage infrequently induces protective immunity and that natural immunity against serogroup W may be lost (24). Serogroup replacement after nationwide vaccination with PsA-TT cannot be ruled out, however, and additional years of surveillance data, as well as nasopharyngeal carriage surveys, are required for a full understanding of whether serogroup replacement has resulted from PsA-TT vaccination. Preliminary data indicate that rates of serogroup W disease were lower during the 2013 meningitis season than during the 2012 meningitis season.

N. meningitidis has multiple genetic mechanisms to alter its antigenic profile, including capsular switching. Capsular switching has been observed among pneumococcal serotypes after the 7-valent pneumococcal conjugate vaccine was introduced in the United States, although the substantial declines in incidence of pneumococcal disease were maintained overall (25). Immediately after implementation of serogroup C meningococcal conjugate vaccine in the United Kingdom, no evidence of capsular replacement was found (26). Low rates of serogroup C disease in the United Kingdom have persisted, but serogroup Y has increased during the past several years in the United Kingdom and several other European countries (27). However, the effects of vaccination often are difficult to disentangle from natural fluctuations in meningococcal incidence and serogroup distribution.

With the recent shift in the epidemiology of bacterial meningitis in Burkina Faso after PsA-TT introduction, the current epidemic thresholds for predicting meningitis epidemics should be reevaluated. Although both serogroup W and serogroup X are known to cause epidemics, neither has been observed to cause the explosive epidemics of serogroup A or to have the periodicity observed for serogroup A epidemics. These thresholds were developed when the effects of disease were substantially higher so that epidemics could be anticipated and response activities (i.e., reactive vaccination campaigns) would prevent additional cases (18). In the 13 districts in Burkina Faso where meningitis reached the weekly epidemic threshold during 2012, the peak of reported cases was detected (Figure 2), but in no district was the cumulative annual epidemic threshold of 100 cases per 100,000 population met (20). During the 2012 meningitis season, 3 districts in Burkina Faso had vaccination campaigns that used quadrivalent polysaccharide vaccine. These are the first campaigns to use quadrivalent polysaccharide vaccine after PsA-TT vaccine; how many additional serogroup W cases were prevented and how administration of polysaccharide vaccination after PsA-TT will affect duration of protection against serogroup A disease are unknown.

The increases in serogroup W disease during 2012 were recognized because of the investment in high-quality meningitis surveillance in Burkina Faso, which underscores the importance of maintaining high-quality case-based meningitis surveillance capacity across the meningitis belt after PsA-TT introduction. Although additional years of data are needed to understand the epidemiology of meningococcal disease after PsA-TT introduction, it is evident that N. meningitidis will continue to cause sporadic disease and has potential to cause epidemics. In the meantime, countries need to effectively communicate the need for persons to remain diligent about seeking care for meningitis symptoms while not undermining public trust in the effectiveness of PsA-TT. The only available control strategy for serogroup W epidemics is reactive campaigns with trivalent or quadrivalent meningococcal polysaccharide vaccine. An affordable, multivalent vaccine, developed for use in the meningitis belt is needed in this disproportionately affected region to protect against epidemic meningitis.