The Rise and Fall of Epidemic Neisseria meningitidis Serogroup W135 Meningitis in Burkina Faso,

MAJOR ARTICLE The Rise and Fall of Epidemic Neisseria meningitidis Serogroup W135 Meningitis in Burkina Faso, 2002 2005 Yves Traoré, 1 Berthe-Marie Njanpop-Lafourcade, 6 Kokou-Louis-Sewonou Adjogble, 5 Mathilde Lourd, 6 Seydou Yaro, 2 Boubacar Nacro, 3 Aly Drabo, 2 Isabelle Parent du Châtelet, 6 Judith E. Mueller, 6 Muhamed-Kheir Taha, 7 Ray Borrow, 9 Pierre Nicolas, 8 Jean-Michel Alonso, 7 and Bradford D. Gessner 6 1 Université de Ouagadougou, Ouagadougou, and 2 Centre Muraz, 3 Centre Hospitalier National Souro Sanou, and 4 Direction Régionale de la Santé, Ministère de la Santé, Bobo-Dioulasso, Burkina Faso; 5 Ministere de la Sante Publique du Togo, Lomé, Togo; 6 Agence de Médecine Préventive and 7 Institut Pasteur, Paris, and 8 Institut de Médecine Tropicale du Service de Santé des Armées, Centre Collaborateur OMS de Référence et de Recherche pour Les Méningocoques, Marseille, France; and 9 Vaccine Evaluation Department, Health Protection Agency, Manchester, United Kingdom Background. During the period 2001 2002, Burkina Faso reported its first meningitis epidemic due to Neisseria meningitidis (Nm) serogroup W135, prompting concerns that this serogroup would persist as a cause of epidemic disease. Methods. During the period 2002 2005, hospital- and population-based surveillances were conducted in 3 districts in Burkina Faso. Etiology was determined by culture, polymerase chain reaction (PCR), and latex agglutination. Reference laboratories determined phenotype and genotype. Results. Of 2004 subjects who received a lumbar puncture, 265 were identified as having Nm, including 93 who had Nm serogroup A (NmA) and 146 who had Nm serogroup W135 (NmW135). Over the study period, the proportion of cases due to NmW135 decreased by 175%, primarily because of decreased occurrence among young children and in a single district. During peak epidemic months, the annualized incidence of NmW135 decreased from 146 cases to!1 case per 100,000 population. All but 2 NmW135 isolates were phenotype W135: 2a:P1.5,2 (sequence type [ST]-11 clonal complex). All NmA isolates were phenotype A:4:P1-9 (ST-2859 of the ST- 5 clonal complex). We identified 1 isolate from serogroup Y (ST-11 clonal complex), 1 isolate from serogroup X that was similar to strains previously associated with epidemic disease, and 1 isolate from serogroup W135 of the newly described ST-4375 complex. Conclusions. For unknown reasons, serogroup W135 achieved epidemic status, primarily among young children, and then largely disappeared over a short time period. The continued circulation of multiple strains with epidemic potential emphasizes the need for ongoing surveillance and the potential benefit of vaccines that are protective across serogroups. Within the African meningitis belt that extends from Senegal to Ethiopia, most epidemics of Neisseria meningitidis (Nm) infection have been caused by serogroup A (NmA) [1]. Serogroup W135 (NmW135) emerged as a new epidemic serogroup in 2001 [2, 3]. To assist with understanding these epidemics and, thus, to pro- vide information on future epidemic potential, the current study evaluated NmW135 meningitis during parts or all of 4 epidemic seasons (December April) during 2002 2005 in 3 districts of southwestern Burkina Faso. METHODS Received 21 March 2006; accepted 9 June 2006; electronically published 22 August 2006. Correspondence: Bradford D. Gessner, 11100 Stony Brook Dr., Anchorage, AK 99516 (106617.3011@compuserve.com). Reprints: Agence de Médecine Préventive, Institut Pasteur, 28 rue du Docteur Roux, 75015, Paris, France. Clinical Infectious Diseases 2006; 43:817 22 2006 by the Infectious Diseases Society of America. All rights reserved. 1058-4838/2006/4307-0003$15.00 Case identification and surveillance sites. Acute bacterial meningitis surveillance was initiated during April 2002 in Districts 15, 22, and Houndé, encompassing Bobo-Dioulasso and its surrounding areas (total 2002 population, 880,000; population!5 years of age, 140,000; Burkina Faso Ministry of Health) [2]. Houndé passed the World Health Organization epidemic threshold of 10 suspected cases per 100,000 population per Meningococcal Meningitis in Africa CID 2006:43 (1 October) 817

week during most weeks of the first 3 months of 2003; for the remaining districts and study period, the epidemic threshold was not passed (Burkina Faso Ministry of Health, unpublished data). Using A/C polysaccharide vaccine among the 2- to 29-yearold target age group, a mass reactive vaccination campaign was conducted in the health region of Bobo-Dioulasso during the 2000 2001 epidemic season, with coverage of 20%. Before the 2001 2002 season, mass preventative campaigns were conducted with 190% administrative coverage among the target age group in all 3 districts. During the 2002 2003 season, mass reactive campaigns using A/C/W135 polysaccharide vaccine were conducted in parts of Burkina Faso, but not in the 3 study districts. During subsequent seasons, epidemic response in Burkina Faso and vaccine type were guided by the serogroup distribution of affected districts. Population-based surveillance was conducted at the referral hospital and at all 59 local health centers from April 2002 April 2003 and again from March 2004 February 2005 (activities were halted mid study because of funding limitations). From March June 2005, because of financial constraints, cases were identified only at the referral hospital. Using their clinical judgment, health care providers at study hospitals and health centers enrolled patients with suspected bacterial meningitis (according to the World Health Organization definition [http: //www.who.int/immunization_monitoring/diseases/meningitis_surveillance/en/; last accessed May 16, 2006]) even if CSF was clear. At all study sites, patients generally received a lumbar puncture and antibiotic therapy. An aliquot of CSF and a clinical case report form for each patient were transported to the reference laboratory in Bobo-Dioulasso. Laboratory methodology. The study protocol required PCR testing for all cases, sample culture according to World Health Organization methodology [4] when the time between collection of CSF and arrival at the reference laboratory was!2 h, and latex agglutination if CSF was visibly cloudy. Nm was considered the etiologic agent if it was identified by PCR, latex agglutination (Pastorex [Biorad]), or culture. All specimens positive for NmY/W135 by latex agglutination that were tested by PCR were found to be genogroup W135; consequently, we assumed that all Nm isolates found to be Y/W135 by latex agglutination were serogroup W135, except for a single specimen confirmed as serogroup Y. PCR identification was based on crga gene amplification [5]. Genogroup was determined by a multiplex PCR that used oligonucleotides for siad for genogroups B, C, Y, and W135 and mynb for genogroup A [5]. Serotype and sero-subtype were determined by reference techniques [6, 7]. Genotyping was performed by multilocus sequence typing (MLST) [8 10]. Study approval. This surveillance project was approved by the ethics review board of Centre Muraz and was supported by the Ministry of Health of Burkina Faso. RESULTS Case identification. There were 2144 subjects who were admitted for acute bacterial meningitis, of whom 2004 received a lumbar puncture. Among CSF samples, Nm was identified as the etiologic agent in 265 cases (13%), including 237 (15%) of 1583 samples tested by PCR, 164 (17%) of 948 samples tested by latex agglutination, and 130 (17%) of 784 samples tested by culture. Of the 265 cases of Nm, 93 (35%) were identified as NmA and 146 (55%) as NmW135. The remainder either could not be genogrouped or serogrouped ( n p 19) or were initially genogrouped as NmW135 but were determined later to be non-nmw135 ( n p 7). Isolate characterization. Of the 57 Nm isolates that tested positive for genogroup W135 by PCR at the Bobo-Dioulasso laboratory, 48 were phenotype W135:2a:P1.5,2 (all 12 genotyped isolates from this group were sequence type [ST] 11, belonging to the ST-11 clonal complex), 1 was W135:NT:P1.5,2 (ST-11), and 1 was W135:NT:P1.5,2 (ST-4375, which belongs to the ST-23 clonal complex). Seven additional isolates that were identified as genogroup NmW135 in the field were discovered at the reference laboratory to be phenotypes Y:14:P1.5,2 (ST-11; n p 1), X:NT:P1.5 (ST-751; n p 1), NG:NT:NST (ST- 192; n p 3), and polyagglutinable:14:p1.5,2 (ST-4375; n p 2). Thirty-five isolates from CSF that tested positive for genogroup NmA by PCR at the Bobo-Dioulasso laboratory were further evaluated, including 15 strains that were collected during the period 2002 2003 and 20 that were collected during the period 2004 2005. All 35 isolates were phenotype A:4:P1-9, and all 31 sequence-typed isolates were ST-2859 (belonging to the ST-5 clonal complex). Incidence, case-fatality percentage, and age group. To prevent overweighting the epidemic season and to limit estimates to the period during which population-based surveillance existed, incidence rates and case-fatality percentages were estimated for the periods May 2002 through April 2003 and March 2004 through February 2005. NmA incidence rates peaked among persons 1 14 years of age, whereas rates for NmW135 peaked among infants (table 1). For NmW135, a substantial minority of cases occurred among persons outside the current vaccine-eligible age group of 2 29 years. Temporal and geographic distribution. NmW135 cases peaked during April 2002 and again during the 2002 2003 season (figure 1). The proportion of Nm cases due to NmW135 was 86% during April May 2002 (the first 2 months of the surveillance system), 70% during December 2002 through April 2003, 32% during March April 2004, and 18% during December 2004 through April 2005. To evaluate changes in incidence by age group, for 4 consecutive years we compared rates during 818 CID 2006:43 (1 October) Traoré et al.

Table 1. Summary characteristics of cases of Neisseria meningitidis (Nm) infection observed during the periods May 2002 April 2003 and March 2004 February 2005 in the Bobo-Dioulasso region, Burkina Faso. Characteristic Nm infection NmW135 infection NmA infection No. of patients 221 121 (55) 83 (35) Age, mean years (range) 9.8 (0.13 81) 7.9 (0.13 66) 13 (0.43 81) Case-fatality percentage All ages, % (no. of deaths/cases) 12 (27/221) 14 (17/121) 8 (7/83)!1 year, % (no. of deaths/cases) 17 (4/24) 11 (2/19) 50 (1/2) 1 4 years, % (no. of deaths/cases) 12 (7/57) 12 (5/42) 8 (1/13) 5 14 years, % (no. of deaths/cases) 12 (12/98) 16 (7/44) 9 (4/44) 15 years, % (no. of deaths/cases) 10 (4/42) 19 (3/16) 4 (1/24) Incidence by standard age groups All ages!1 year 32 (24) 26 (19) 2.7 (2) 1 4 years 28 (57) 20 (42) 6.3 (13) 5 14 years 17 (98) 7.8 (44) 7.8 (44) 15 years 4.6 (42) 1.7 (16) 2.6 (24) Incidence by vaccine eligible groups, vaccine targeted to 2 29 year olds!2 years 32 (40) 26 (33) 2.4 (3) 2 29 years 11 (168) 5.6 (82) 5.0 (73) 30 years 8.0 (13) 3.7 (6) 4.3 (7) NOTE. Data are no. (%) of patients, unless otherwise indicated. Incidence is defined as per 100,000 persons per year. NmA, N. meningitidis serotype A; NmW135, N. meningitidis serotype W135. the peak 2 months (only data from April were available for 2002) of the 4 epidemic meningitis seasons (figure 2). NmW135 rates remained stable for persons 5 years of age but dereased substantially for younger children. Of all Nm meningitis cases, 85 (40%) occurred in the Houndé district, compared with 61 (29%) in District 15 and 65 (31%) in District 22, despite the location of the regional referral hospital in District 22. The proportions of cases due to NmA in Houndé, District 15, and District 22 were 37%, 26%, and 37%, respectively, compared with 46%, 22%, and 31%, respectively, for NmW135. When incidence rates by district were examined (figure 3), Houndé experienced the greatest NmW135 epidemic (during 2002) followed by the steepest decrease; in contrast, Districts 15 and 22 had relatively stable and low incidence rates until 2005, when rates had decreased everywhere to near zero. DISCUSSION We observed that NmW135 primarily from a single clonal complex was found in an epidemic pattern and was the predominant serogroup that was identified during the first part of the study period. The incidence was highest among young children and within a single district. Disease occurrence decreased substantially over the 4 evaluated epidemic seasons, with few cases identified by 2005. Serogroup W135 incidence increased within the context of meningitis cases due to other potentially epidemic Nm strains. One case due to NmW135 of the newly described ST-4375 was identified. NmA meningitis was caused by the new sequence type (ST-2859) within the ST-5 clonal complex that has only been identified since 2002 in Niger and Burkina Faso [11]. We identified the first documented case of invasive disease in the African meningitis belt involving a serogroup Y isolate from the same ST-11 genotype as was responsible for W135 epidemics (possibly due to capsular switching [12, 13]). Lastly, 1 serogroup X case occurred that had the same phenotype and genotype as previous epidemic serogroup X strains [14, 15]. The reasons for the emergence and decrease in NmW135 disease (while other potentially epidemic strains have remained relatively quiescent) are largely unknown. Mass reactive and preventative polysaccharide A/C vaccination campaigns could have created an ecological niche for serogroup W135. Subsequent mass campaigns with trivalent A/C/W135 polysaccharide vaccines [16] could have then led to decreases in disease. However, to argue against this explanation, Districts 15 and 22 experienced relatively low rates of NmW135 incidence despite widespread use of bivalent vaccine, and in all study districts, the proportion of disease due to serogroup W135 decreased in the absence of substantial trivalent vaccine use. Additionally, polysaccharide vaccines have little effect on carriage [17, 18] and thus would not be expected to create an ecological niche for nonvaccine serogroups. Finally, in the United Kingdom, the Meningococcal Meningitis in Africa CID 2006:43 (1 October) 819

Figure 1. Monthly occurrence of Neisseria meningitidis (Nm) cases by serogroup (including indeterminate or other [I/O]) in the Bobo-Dioulasso region, Burkina Faso. Date of change from comprehensive population-based to hospital surveillance (arrow). use of conjugate serogroup C vaccines has not led to increased disease due to other serogroups [19], despite conjugate vaccine efficacy against carriage [20]. It may be that an immunologically naive population was exposed to a virulent clone and, subsequently, experienced the development of anticapsular, antibody-based population immunity and a decrease in the rate of disease. This is consistent with the emergence and disappearance of NmW135 primarily among young children, as well as the concentration of disease followed by rapid disappearance within a single district in Burkina Faso. However, a recent study in Bobo-Dioulasso during the period February June 2003 [21] found that, despite several years of epidemic disease and substantial serogroup W135 carriage, the great majority of the population had not acquired putatively protective anti-nmw135 antibodies (as measured by IgG and serum bactericidal antibody). Additionally, those who had putatively protective antibody levels at the beginning of the study frequently lost protection over time. Because, in our study, virtually all of the strains that expressed the W135 capsule belonged to the ST-11 clonal complex, the decrease in serogroup W135 may be specific to this clone. Consistent with this hypothesis are data from France that demonstrate a reduction in the number of isolates of W135:2a: P1.5,2, ST-11 at the same time that serogroup W135 strains of other phenotypes and genotypes have persisted and diversified [8]. Isolates of W135:2a:P1.5,2, ST-11 have circulated and diversified since at least 1970 [22]. In 2000, these isolates started undergoing several clonal waves of expansion among susceptible hosts, under epidemic conditions of spread [2, 11, 23]. The current decrease in disease may have resulted from a high rate of infection (of both carriage and invasive disease) with such isolates [24], coupled with efficient stimulation of noncapsular- and noncomplement-dependent antisurface structure Figure 2. Annualized Neisseria meningitidis (Nm) incidence rates (per 100,000) by serogroup, by age group (in years), and by year of occurrence for the peak recorded months of 4 epidemic meningitis seasons in the Bobo-Dioulasso region, Burkina Faso. Total Nm incidences are greater than the sum of Nm serotype A (NmA) and Nm serotype W135 (NmW135) incidences because serogroup determination was not attempted or confirmed for all Nm isolates. 820 CID 2006:43 (1 October) Traoré et al.

B.D.G. were employees of Agence de Médecine Préventive, which receives substantial financial support from Sanofi-Pasteur, one of the funders of the study and a manufacturer of meningococcal vaccines. All other authors: no conflicts. Figure 3. Annualized Neisseria meningitidis (Nm) incidence rates (per 100,000) by serogroup, by district of residence of patients, and by year of occurrence for the peak recorded months of 4 epidemic meningitis seasons in the Bobo-Dioulasso region, Burkina Faso. Total Nm cases during the 7 evaluated months by district were 42 in Bobo-Dioulasso district 15, 51 in Bobo-Dioulasso district 22, and 63 in Houndé. Total Nm incidences are greater than the sum of Nm serotype A (NmA) and Nm serotype W135 (NmW135) incidences because serogroup determination was not attempted or confirmed for all Nm isolates. antibodies [25] and a relative genetic bottleneck. The detection of several isolates with diverse genotypes and phenotypes is consistent with a selective restriction of the dominant circulating strain (W135:2a:P1.5,2, ST-11). With the current disappearance of epidemic serogroup W135 meningitis, monovalent A conjugate vaccines once again may assume an important role. The recent history of NmW135 meningitis in sub-saharan Africa, however, reveals that new epidemic strains may occur and rapidly lead to major epidemics. Our data indicate that, although the rate of serogroup W135 disease has lessened, disease continues to be caused by multiple serogroups with epidemic potential. Continued meningitis surveillance will assist in the development and modification of vaccine strategies, including the use of multivalent conjugate vaccines and novel vaccines using antigens that are common across serogroups. Acknowledgments The clinical team consisted of Adrien Sawadogo (Centre Hospitalier National Souro Sanou), and Flore Ouedraogo, Francine Ouedraogo, and P. Korgho (Direction Régionale de la Santé, Ministère de la Santé, Bobo- Dioulasso, Burkina Faso). The laboratory team consisted of Alain Hien, Francis Hien, and Dominique Niamba (Centre Hospitalier National Souro Sanou, Bobo-Dioulasso) and Lassana Sangare (Université de Ougadougou, Ougadougou, Burkina Faso). We thank the numerous clinical and laboratory staff of Centre Muraz, Centre Hospitalier Universitaire Souro Sanou, and the peripheral health care centers who cared for patients and assisted with parts of the study. Financial support. Sanofi-Pasteur, Institut Pasteur, and The Bill and Melinda Gates Foundation. Potential conflicts of interest. B-M.N-L., M.L., I.P.C., J.E.M., and References 1. World Health Organization. Epidemics of meningococcal disease, African meningitis belt, 2001. Wkly Epidemiol Rec 2001; 76:281 8. 2. Parent du Chatelet I, Traore Y, Gessner BD, et al. Bacterial meningitis in Burkina Faso: surveillance using field-based polymerase chain reaction testing. Clin Infect Dis 2005; 40:17 25. 3. Taha MK, Parent du Chatelet I, Schlumberger M, et al. Neisseria meningitidis serogroups W135 and A were equally prevalent among meningitis cases occurring at the end of the 2001 epidemics in Burkina Faso and Niger. J Clin Microbiol 2002; 40:1083 4. 4. World Health Organization. Laboratory methods for the diagnosis of meningitis caused by Neisseria meningitidis, Streptococcus pneumoniae, and Haemophilus influenzae. WHO Communicable disease surveillance and response. WHO/CDS/CSR/EDC/99.7. Available at: http://www.who.int/csr/resources/publications/meningitis/who_cds_csr_edc_ 99_7_EN/en/. Accessed 15 August 2006. 5. Taha MK. Simultaneous approach for nonculture PCR-based identification and serogroup prediction of Neisseria meningitidis. J Clin Microbiol 2000; 38:855 7. 6. Frasch CE, Zollinger WD, Poolman JT. Serotype antigens of Neisseria meningitidis and a proposal scheme for designation of serotypes. Rev Infect Dis 1985; 7:504 10. 7. Poolman JT, Abdillahi H. Outer membrane protein serosubtyping of Neisseria meningitidis. Eur J Clin Microbiol Infect Dis 1988; 7:291 2. 8. Taha MK, Giorgini D, Ducos-Galand M, Alonso JM. Continuing diversification of Neisseria meningitidis W135 as a primary cause of meningococcal disease after emergence of the serogroup in 2000. J Clin Microbiol 2004; 42:4158 63. 9. Guiver M, Borrow R. PCR diagnosis. In: Pollard AJ, Maiden MCJ, eds. Meningococcal disease: methods in molecular medicine. Totowa, NJ: Humana Press, 2001:23 39. 10. Maiden MC, Bygraves JA, Feil E, et al. Multilocus sequence typing: a portable approach to the identification of clones within populations of pathogenic microorganisms. Proc Natl Acad Sci U S A 1998; 95: 3140 5. 11. Njanpop-Lafourcade BM, Parent du Chatelet I, Sanou O, Alonso JM, Taha MK. The establishment of Neisseria meningitidis serogroup W135 of the clonal complex ET-37/ST-11 as an epidemic clone and the persistence of serogroup A isolates in Burkina Faso. Microbes Infect 2005; 7:645 9. 12. Lancellotti M, Guiyoule A, Ruckly C, Hong E, Alonso JM, Taha MK. Conserved virulence of C to B capsule switched Neisseria meningitidis clinical isolates belonging to ET-37/ST-11 clonal complex. Microbes Infect 2006;8: 191 6. 13. Stefanelli P, Fazio C, Neri A, Sofia T, Mastrantonio P. First report of capsule replacement among electrophoretic type 37 Neisseria meningitidis strains in Italy. J Clin Microbiol 2003; 41:5783 6. 14. Djibo S, Nicolas P, Alonso JM, et al. Outbreaks of serogroup X meningococcal meningitis in Niger 1995 2000. Trop Med Int Health 2003; 8:1118 23. 15. Gagneux SP, Hodgson A, Smith TA, et al. Prospective study of a serogroup X Neisseria meningitidis outbreak in northern Ghana. J Infect Dis 2002; 185:618 26. 16. Ahmad K. Vaccination halts meningitis outbreak in Burkina Faso. Lancet 2004; 363:1290. 17. Moore PS, Harrison LH, Telzak EE, Ajello GW, Broome CV. Group A meningococcal carriage in travellers returning from Saudi Arabia. JAMA 1988; 260:2686 9. 18. Hassan-King MK, Wall RA, Greenwood BM. Meningococcal carriage, meningococcal disease and vaccination. J Infect 1988; 16:55 9. Meningococcal Meningitis in Africa CID 2006:43 (1 October) 821

19. Balmer P, Borrow R, Miller E. Impact of meningococcal C conjugate vaccine in the UK. J Med Microbiol 2002; 51:717 22. 20. Maiden MC, Stuart JM. Carriage of serogroup C meningococci 1 year after meningococcal C conjugate polysaccharide vaccination. Lancet 2002; 359:1829 31. 21. Mueller JE, Yaro S, Traoré Y, et al. Neisseria meningitidis serogroup W135 and A: carriage and immunity in Burkina Faso, 2003. J Infect Dis 2006; 193:812 20. 22. Mayer LW, Reeves MW, Al-Hamdan N, et al. Outbreak of W135 meningococcal disease in 2000: not emergence of a new W135 strain but clonal expansion within the electophoretic type-37 complex. J Infect Dis 2002; 185:1596 605. 23. Taha MK, Achtman M, Alonso JM, et al. Serogroup W135 meningococcal disease in Hajj pilgrims. Lancet 2000; 356:2159. 24. Nicolas P, Aït M Barek N, Al-Awaidy S, et al. Pharyngeal carriage of serogroup W135 Neisseria meningitidis in Hajjees and their family contacts in Morocco, Oman and Sudan. APMIS 2005; 113:182 6. 25. Welsch JA, Moe GR, Rossi R, Adu-Bobie J, Rappuoli R, Granoff DM. Antibody to genome-derived neisserial antigen 2132, a Neisseria meningitidis candidate vaccine, confers protection against bacteremia in the absence of complement-mediated bactericidal activity. J Infect Dis 2003; 188:1730 40. 822 CID 2006:43 (1 October) Traoré et al.