LETTERS

References Yersinia pestis in rural area in the district of Ambatofi- nandrana, Madagascar. Domestic 1. Hayes EB. Zika virus outside Africa. Pulex irritans Emerg Infect Dis. 2009;15:1347–50. were collected with candle traps inside http://dx.doi.org/10.3201/eid1509.090442 Fleas during 5 houses during 3 nights (Table). Fleas 2. Lanciotti RS, Kosoy OL, Laven JJ, Plague Outbreak, were also caught on small mammals Velez JO, Lambert AJ, Johnson AJ, trapped inside houses and outside in the et al. Genetic and serologic properties of Madagascar Zika virus associated with an epidemic, sisal fences and rice fields (Table). A Yap State, Micronesia, 2007. Emerg In- total of 319 fleas belonging to 5 fect Dis. 2008;14:1232–9. http://dx.doi. To the Editor: Yersinia pestis in 5 genera were collected inside and org/10.3201/eid1408.080287 (family Enterobacteriaceae) is a bac- outside the houses, an average of 44 per 3. European Centre for Disease Preven- terium that can cause high rates of tion and Control. Zika virus—Pacific house (maximum 71): Pulex irritans, (08). ProMed. 2014 Mar 16. http:// death in susceptible mammals and can Echidnophaga gallinacea, and Cteno- www.promedmail.org, archive no. 2014 provoke septicemic, pneumonic, and cephalides canis fleas were collected 0316.2335754. bubonic plague in humans (1). This inside the houses (244, 76.5%), and S. 4. Schwan K. Zika virus—Pacific (07). zoonotic pathogen can be transmitted ProMed. 2014 Mar 9. http://www.pro- fonquerniei and X. cheopis fleas were medmail.org, archive no. 20140309. directly by infectious droplets or by collected outside (75, 23.5%). The hu- 2322907. contact with contaminated fluid or tis- man , P. irritans, was the most col- 5. Tappe D, Schmidt-Chanasit J, Ries A, sue or indirectly through flea bites (1). lected flea species (233, 73.3%), fol- Ziegler U, Müller A, Stich A. Ross River Plague was introduced into Mad- virus infection in a traveller returning lowed by S. fonquerniei (62, 19.4%), X. from northern Australia. Med Microbiol agascar in 1898 from rat-infested cheopis (13, 4.1%), E. gallinacea (10, Immunol (Berl). 2009;198:271–3. http:// steamships that had sailed from affect- 3.1%), and C. canis (1, 0.3%). dx.doi.org/10.1007/s00430-009-0122-9 ed areas (2). Now, Madagascar is 1 of Bacterial DNA was extracted 6. Tappe D, Rissland J, Gabriel M, 2 countries in Africa that have report- Emmerich P, Gunther S, Held G, et al. from 277 fleas of 5 species: 233 P. First case of laboratory-confirmed Zika vi- ed cases of human plague every year irritans, 24 S. fonquerniei, 9 X. cheo- rus infection imported to Europe, Novem- since 1991 (3). During January 2008– pis, 10 E. gallinacea, and 1 C. canis. ber 2013. Euro Surveill. 2014;19:20685. January 2013, the number of human PCR to detect Y. pestis was performed 7. Fonseca K. Zika virus—Canada ex plague cases reported in Madagascar ′ Thailand. ProMed. 2013 May 29. http:// by using primers YP1 (5 -ATC TTA www.promedmail.org, archive no. 2013 ranged from 312 to 648 per year. Of CTT TCC GTG AGA AG-3′) and YP2 0529.1744108. these, 61.8%–75.5% were laboratory (5′-CTT GGA TGT TGA GCT TCC 8. Kwong JC, Druce JD, Leder K. Zika confirmed (National Plague Labora- TA-3′) to amplify a 478-bp fragment virus infection acquired during brief tory of the Ministry of Health, pers. travel to Indonesia. Am J Trop Med Hyg. (4). Y. pestis DNA was then amplified 2013;89:516–7. http://dx.doi.org/10.4269/ comm.). Most (>83%) confirmed cas- and genotyped by Beckman Coul- ajtmh.13-0029 es were bubonic plague, which most ter Genomics Inc. (Takeley, United 9. Foy BD, Kobylinski KC, Chilson Foy commonly results from flea bites, Kingdom). The positive control was Y. JL, Blitvich BJ, Travassos da Rosa A, suggesting that these bites were the Haddow AD, et al. Probable non-vec- pestis reference strain (strain 6/69, 3 8 tor-borne transmission of Zika virus, most common mode of Y. pestis trans- × 10 bacteria/mL; Institut Pasteur de Colorado, USA. Emerg Infect Dis. mission. In Madagascar, Xenopsylla Madagascar). 2011;17:880–2. http://dx.doi.org/10.3201/ cheopis fleas have been known as the Detection of Y. pestis was carried eid1705.101939 primary plague vector in urban areas, out on 274 fleas belonging to 5 flea 10. Kutsuna S, Kato Y, Takasaki T, Moi ML, Kotaki A, Uemura H. Two cases of Zika whereas Synopsyllus fonquerniei fleas species: 230 P. irritans (181 unfed and virus imported from French Polynesia to have been usually involved in plague 49 engorged), 24 S. fonquerniei (15 Japan, December 2013 to January 2014. transmission in rural areas (2). unfed and 9 engorged), 9 X. cheopis Euro Surveill. 2014;19:pi=20683. In January 2013, a total of 9 sus- (8 unfed and 1 engorged), 10 E. galli- pected bubonic plague cases, 3 con- nacea (blood-feeding status not iden- Address for correspondence: Jonas Schmidt- firmed, were reported in Soavina, a tified), and 1 unfed C. canis. Y. pestis Chanasit, Bernhard Nocht Institute for Tropical

Medicine, WHO Collaborating Centre for Table. Fleas collected inside and outside houses in Soavina, Madagascar, Arbovirus and Haemorrhagic Fever Reference January 2013 and Research, Bernhard-Nocht-Strasse 74, 20359 Species Total no. (%) No. (%) inside No. (%) outside Hamburg, Germany; email: [email protected] Pulex irritans 233 (73.0) 233 (95.5) 0 Ctenocephalides canis 1 (0.3) 1 (0.4) 0 Echidnophaga gallinacea 10 (3.1) 10 (4.1) 0 Search past issues of EID at Synopsyllus fonquerniei 62 (19.4) 0 62 (82.7) wwwnc.cdc.gov/eid Xenopsylla cheopis 13 (4.1) 0 13 (17.3) Total 319 (100) 244 (100) 75 (100)

1414 Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 20, No. 8, August 2014 LETTERS was detected in 9 P. irritans fleas (7 Author affiliations: Institut Pasteur de Mad- Serologic male [6 unfed and 1 engorged] and 2 agascar, Antananarivo, Madagascar (J. [engorged] female) from 3 houses, in- Ratovonjato, S. Boyer); and World Health Surveillance for cluding the house where a confirmed Organization Collaborating Centre, Institut West Nile Virus human case of plague had occurred Pasteur de Madagascar, Antananarivo (M. in Dogs, Africa (online Technical Appendix, http:// Rajerison, S. Rahelinirina) wwwnc.cdc.gov/EID/article/20/8/13- DOI: http://dx.doi.org/10.3201/eid2008.130629 To the Editor: West Nile fe- 0629-Techapp1.pdf). Eight sequences ver is caused by the West Nile virus (GenBank accession nos. KJ361938– References (WNV), a mosquito-borne member KJ361945) were obtained and share of the Flavivirus. Birds are 99% nucleotide homology with plas- 1. Bulter T. Plague and other Yersinia infec- the natural reservoir of the virus, minogen activator genes of Y. pestis tions. In: Greenough WB III, Merigan TC, which is maintained in nature in a editors. Current topics in infectious dis- published in GenBank (accession nos. ease. New York: Plenum: 1983. p. 71–92. mosquito–bird–mosquito transmis- AF528537, AY305870). No Y. pestis 2. Brygoo ER. Epidemiology of the plague at sion cycle. WNV has been detected was detected in the 24 S. fonquerniei, Madagascar [in French]. Arch Inst Pasteur in many regions worldwide, includ- 9 X. cheopis, 10 E. gallinacea, or 1 C. Madagascar. 1966;35:7–147. ing North America, Europe, Africa, 3. World Health Organization. Report on canis fleas collected. global surveillance of epidemic-prone the Near East, and Asia (1). WNV Although only X. cheopis and S. infectious diseases—plague [cited 2013 has been shown to cause meningoen- fonqueniei fleas had previously been Apr 15]. http://www.who.int/csr/resources/ cephalitis in humans and horses. In described as plague vectors in Mada- publications/plague/CSR_ISR_2000_1/en/ the United States, seroconversion in index5.html gascar, P. irritans fleas were most 4. Hinnebusch J, Schwan TG. New method dogs was detected 6 weeks before a commonly collected during this field for plague surveillance using polymerase human case was reported (2). Thus, study; engorged and unfed male and chain reaction to detect Yersinia pestis in dogs could be considered as sentinels female P. irritans fleas carried Y. pes- fleas. J Clin Microbiol. 1993;31:1511–4. for WNV infection, but their role as 5. Laudisoit A, Leirs H, Makundi RH, tis. Other studies have found P. irri- Van Dongen S, Davis S, Neerinckx S, reservoir is unlikely because of short- tans fleas in the plague risk area in oth- et al. Plague and the , Tanzania. term and low levels of viremia (3). er countries in Africa (5,6); one study Emerg Infect Dis. 2007;13:687–93. http:// In this study, we determined the se- found that P. irritans fleas may play a dx.doi.org/10.3201/eid1305.061084 roprevalence of WNV in dogs living 6. Eisen RJ, Gage KL. Transmission of flea- role in plague epidemiology in Tanza- borne zoonotic agents. Annu Rev Entomol. close to humans in different environ- nia (5). Data on P. irritans fleas in rats 2012;57:61–82. http://dx.doi.org/10.1146/ ments to assess their role as sentinels make it unlikely that these fleas are annurev-ento-120710-100717 of this potentially severe zoonosis. involved in rat-to-human transmis- During 2003–2012, blood sam- sion of Y. pestis in Madagascar. Dur- Address for correspondence: Jocelyn ples were collected from 753 adult ing 1922–1995, a total of 118,608 rats Ratovonjato, Medical Entomology Unit, Institut dogs from France and 6 countries in were caught and examined in Mada- Pasteur de Madagascar, BP 1274, Antananarivo Africa (Table). Samples were cen- gascar, but only 148 P. irritans fleas 101, Madagascar; email: [email protected] trifuged within 24 h after collection, were identified, and none have been separated, frozen at –20°C, and sent found on rats since 1996 (http://www. to the virology laboratory of the In- pasteur.mg/spip.php?rubrique124). stitut de Recherche Biomédicale des The high density of P. irritans fleas Armées (Marseille, France). Each we observed in villages where plague sample was systematically tested for outbreaks occurred in late 2012 and IgG against WNV by using an in- early 2013 (http://www.pasteur.mg/ house ELISA with inactivated WNV spip.php?rubrique124) supports the as antigen. Serum samples were con- possibility that P. irritans fleas played sidered positive if the optical density a role in domestic human-to-human at 450 nm was >3-fold the mean of transmission of Y. pestis during that for negative antigen. Because of these outbreaks. the antigenic cross-reactivity among flaviviruses, all positive samples were Jocelyn Ratovonjato, further tested by Western blot for Minoarisoa Rajerison, WNV-specific antibodies (4); serop- Soanandrasana Rahelinirina, revalence was calculated on the basis and Sébastien Boyer of Western blot–confirmed cases only.

Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 20, No. 8, August 2014 1415