LETTERS

Table. Risk factors and outcome for 27 neutropenic cancer patients with bacteremia Serogroup X in due to multidrug-resistant (MDR) or drug-susceptible Acinetobacter baumannii infection Patients with Patients with Meningococcal drug-susceptible MDR All patients, A. baumannii, A. baumannii, Disease, no. (%) no. (%)* no. (%)* Western Characteristic (N = 27) (n = 12, 44%) (n = 15, 56%) Risk for bacteremia To the Editor: Although >12 Central venous catheter 19 (70.4) 9 (75.0) 10 (66.7) different serogroups of Neisseria Acute leukemia 11 (40.7) 6 (50.0) 5 (33.3) meningitidis exist, most disease out- Previous prophylaxis with quinolones 14 (51.9) 8 (66.7) 6 (40.0) breaks across the African Previous therapeutic treatment epidemic belt are caused by serogroup with cephalosporins 15 (55.6) 8 (66.7) 7 (46.7) A and, less frequently, by serogroups Previous therapeutic treatment C and W135 (1). N. meningitidis se- with carbapenems 8 (29.6) 4 (33.3) 4 (26.7) rogroup X was fi rst described in the Outcome 1960s and has been found to cause a Septic shock 4 (14.8) 2 (16.7) 2 (13.3) Death 2 (7.4) 1 (8.3) 1 (6.7) few cases of invasive disease across *Insignificant difference between patients with drug-susceptible infection and those with MDR North America, Europe, and Africa infection (p<0.05 by univariate analysis). (2). In Africa, small serogroup X out- Septic shock developed in 4 Vladimir Krcmery* breaks have been described in (14.8%) of the 27 neutropenic patients and Erich Kalavsky† (9 cases over a 2-year period) and in with A. baumannii–associated bac- *St Elizabeth School of Health and Social (134 cases between 1995 and teremia, and 2 (7.4%) of the 27 died Sciences, Bratislava, Slovakia; and †Trna- 2000) (3,4). In 2006, however, 51% of (Table). However, we did not fi nd a va University, Trnava, Slovakia 1,139 confi rmed cases of meningococ- statistically signifi cant association cal meningitis in Niger were found to between death among patients with be caused by serogroup X (5). Before bacteremia caused by MDR A. bau- Reference the 2005-06 meningococcal epidemic season, no published reports had de- manni (1 death) compared with death 1. Sunenshine RH, Wright MO, Maragakis among those with bacteremia caused LL, Harris AD, Song X, Hebden J, et al. scribed serogroup X isolates in East by A. baumannii strains susceptible to Multidrug-resistant Acinetobacter infec- Africa. We report the involvement tion mortality rate and length of hospi- the carbapenems, ciprofl oxacin, and of N. meningitidis serogroup X in an talization. Emerg Infect Dis. 2007;13: outbreak of in amikacin (1 death) (Table). This fi nd- 97–103. ing is similar to that described by Su- Western Kenya. nenshine et al. (1) in the general ICU Address for correspondence: Vladimir Krcmery, In January 2006, the Ministry of population and in neutropenic cancer St Elizabeth School of Health and Social Health of Kenya and Médecins sans patients with bacteremia; however, Sciences, Department of Oncology, Heydukova Frontières were notifi ed of a suspect- multivariate analysis was not conduct- 10, 812 50 Bratislava, Slovakia; email: vladimir. ed meningococcal disease outbreak in ed to control for severity of illness and [email protected] West Pokot District, bordering Ugan- coexisting illness. In conclusion, neu- da, in Western Kenya. On the basis of the initial outbreak investigation, the tropenic cancer patients with bactere- Letters mia due to MDR A. baumannii infec- Letters commenting on recent articles outbreak was assessed to have begun tion do not appear to be at increased as well as letters reporting cases, out- in late December 2005. Subsequent breaks, or original research are wel- risk for death compared with patients active surveillance, using the same come. Letters commenting on articles clinical case defi nition of sudden fever with bacteremia due to antimicrobial should contain no more than 300 drug–susceptible A. baumannii. words and 5 references; they are more onset with stiff neck, altered mental likely to be published if submitted status, or both, showed 74 suspected within 4 weeks of the original article’s cases through mid-March 2006, with This work was supported by grant publication. Letters reporting cases, a case-fatality rate of 20%. No cases 06/07/VEGA of the Ministry of Education outbreaks, or original research should were reported after March 2006. of Slovak Republic and approved by the contain no more than 800 words and 10 references. They may have one Over the course of the outbreak, Ethics Committee of St Elizabeth Cancer Figure or Table and should not be di- cerebrospinal fl uid samples were ob- Institute in Bratislava. vided into sections. All letters should tained from 18 patients. Due to low contain material not previously pub- population density, poor access to lished and include a word count. seminomadic populations, and the

944 Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 13, No. 6, June 2007 LETTERS limited nature of the outbreak (rela- can assume that there was 1 menin- 2003 shows dominance of sequence type tively small numbers dispersed over gitis outbreak that started in eastern 5 (ST-5) and ST-11 complexes. J Clin Mi- crobiol. 2005;43:5129–35. a wide geographic region), obtaining and spread to Western Kenya. 2. Gagneux S, Wirth T, Hodgson A, Ehrhard specimens from untreated patients in Although initial laboratory testing in I, Morelli G, Kriz P, et al. Clonal group- West Pokot proved diffi cult. Three of Uganda suggested the presence of se- ings in serogroup X Neisseria meningiti- the 5 fi rst samples were found to show rogroup A, among 23 specimens sub- dis. Emerg Infect Dis. 2002;8:462–6. 3. Djibo S, Nicolas P, Alonso JM, Djibo A, gram-negative diplococci on staining, sequently sent to the Oslo laboratory, Couret D, Riou J, et al. Outbreaks of se- the next 2 were negative on Pastorex 11 were identifi ed as serogroup X by rogroup X meningococcal meningitis in rapid latex agglutination test (Bio-Rad PCR and 3 were serogroup W135 (6). Niger 1995–2000. Trop Med Int Health. Laboratories, Hercules, CA, USA) Therefore, the outbreaks in both Ke- 2003;8:1118–23. 4. Gagneux SP, Hodgson A, Smith T, Wirth (during the outbreak investigation), nya and Uganda involved multiple N. T, Ehrhard I, Morelli G, et al. Prospective and a subsequent 13 were sent to the meningitidis serogroups. In West Po- study of a serogroup X Neisseria men- African Medical and Research Foun- kot, Kenya, the Ministry of Health and ingitidis outbreak in northern Ghana. J dation (AMREF) laboratory in Nairo- Médecins sans Frontières conducted Infect Dis. 2002;185:618–26. 5. Boisier P, Nicolas P, Djibo S, Taha M-K, bi, Kenya, for culture and susceptibil- a vaccination campaign using the tri- Jeanne I, Maïnassara HB, et al. Menin- ity testing. From these 13 specimens, valent polysaccharide vaccine against gococcal meningitis: unprecedented in- 2 yielded a pure growth of N. menin- serogroups A, C, and W135. cidence of serogroup X–related cases in gitidis serogroup X, while no growth Before 2006, previous disease 2006 in Niger. Clin Infect Dis. 2007;44: 657–63. was observed for the remaining 11 outbreaks caused by serogroup X had 6. Lewis R. Meningococcal meningitis sero- specimens. These 2 cultures were sub- not reached the magnitude of those group X: a new player in Ugandan epide- sequently confi rmed as serogroup X caused by serogroups A, C, or W135; miology. WHO Health Action in Crisis. by the World Health Organization Col- they tended to evolve independently 2006;1:6. 7. Centers for Disease Control. Epidemic laborating Centre for Meningococci in of the occurrence of both serogroups meningococcal disease—Kenya and Tan- Oslo, Norway. Multilocus sequence A and C and to be self-limited (3,4). zania: recommendations for travelers, typing and sequencing of the porA Although most of Kenya is not includ- 1990. MMWR Morb Mortal Wkly Rep. and fetA genes as described (http:// ed in the African meningitis belt, large 1990;39:13–4. pubmlst.org/neisseria/), showed that epidemics of meningococcal disease Address for correspondence: Helen S. Cox, the infecting strain belonged to a new have been reported previously (7). In Burnet Institute for Medical Research and sequence type, ST-5403, and that it conclusion, we would like to highlight Public Health, PO Box 2284, Melbourne, was subtype P1.19,26 and FetA type the presence of N. meningitidis sero- Victoria 3001, Australia; email: hcox@burnet. F3-27. This sequence type is unrelated group X in East Africa, its potential edu.au to other serogroup X isolates from involvement in disease outbreaks, and Africa, including those from the lat- the diffi culties it may cause for labora- est serogroup X outbreak in Niger, but tory confi rmation and, consequently, it resembles a sequence type isolated for making an appropriate epidemic in the in the 1970s. In response. addition to the testing at AMREF and in the Oslo laboratory, the 13 samples Sadiki Materu,* Helen S. Cox,† were also analyzed by PCR at the US Petros Isaakidis,† Naval Medical Research Unit No. 3 in Bienvenu Baruani,† Cairo, Egypt. Overall, 5 of these 13 Thomas Ogaro,‡ specimens were positive for serogroup and Dominique A. Caugant§ X (including the 2 samples found to be *African Medical and Research Founda- serogroup X at AMREF and confi rmed tion, Nairobi, Kenya; †Médecins sans Fron- by PCR in Oslo) and 1 each was posi- tières–Spain, Nairobi, Kenya; Ministry of tive for serogroups C, W135, and Y. Health, Nairobi, Kenya; and §Norwegian At the same time as this outbreak Institute of Public Health, Oslo, Norway in Western Kenya, a meningococcal meningitis outbreak was being moni- References tored across the border in the Kar- amoja region of northeastern Ugan- 1. Nicolas P, Norheim G, Garnotel E, Djibo da. Seminomadic populations move S, Caugant D. Molecular epidemiology of isolated in the freely across the 2 countries, and we African meningitis belt between 1988 and

Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 13, No. 6, June 2007 945