Monitoring Water Sources for Environmental Reservoirs of Toxigenic Vibrio cholerae O1, Haiti

To the Editor: In the March 2014 issue of Emerging Infectious Diseases, Alam et al. reported a survey of water sources in Haiti conducted to isolate Vibrio cholerae (1). Each month from April 2012 through March 2013, they sampled 15 sites at 3 rivers and 1 estuary in West Department. From 179 water samples and 144 aquatic animals and plants, they obtained 7 V. cholerae O1 isolates, including 3 ctx-positive toxigenic strains. 
 
Unfortunately, the results for all 7 V. cholerae O1 isolates were aggregated, and no details were provided about the exact time and location of collection of samples corresponding to the 3 ctx-positive strains. The authors posed the question of whether V. cholerae O1 has become established in environmental reservoirs in Haiti, subsequently warning that “as long as the causative microorganism is present in the environment, eradication of the disease will not be possible.” 
 
However, after challenging their results with more accurate epidemiologic data, we found that these 3 ctx-positive toxigenic strains could more likely have been present in the sampled rivers as a result of recent fecal contamination (Figure). Indeed, many cholera cases were reported in the corresponding communal sections (i.e., the smallest Haitian administrative unit, average 25 km2) when the samples containing the 7 V. cholerae O1 isolates were collected. In this context of an ongoing cholera epidemic associated with persisting rainfall (Figure), generalized open-air defecation inevitably leads to contamination of water sources. It is therefore impossible to determine whether V. cholerae–positive rivers constitute perennial reservoirs of the bacteria or whether they act only as transient vectors of the pathogens. 
 
 
 
Figure 
 
Weekly cholera incidence in communal sections of Haiti in which water samples were positive for Vibrio cholerae O1; accumulated precipitation in the studied area by week during April 2012–March 2013; and number of environmental sites from which ... 
 
 
 
The recent dramatic decrease in cholera transmission may provide a good opportunity to address this issue (2). We thus encourage Alam et al. to continue the search for ctx-positive toxigenic V. cholerae O1 strains in surface waters, especially during cholera-free periods.


Stanislas Rebaudet and Renaud Piarroux
. Each month from April 2012 through March 2013, they sampled 15 sites at 3 rivers and 1 estuary in West Department. From 179 water samples and 144 aquatic animals and plants, they obtained 7 V. cholerae O1 isolates, including 3 ctx-positive toxigenic strains.
Unfortunately, the results for all 7 V. cholerae O1 isolates were aggregated, and no details were provided about the exact time and location of collection of samples corresponding to the 3 ctx-positive strains. The authors posed the question of whether V. cholerae O1 has become established in environmental reservoirs in Haiti, subsequently warning that "as long as the causative microorganism is present in the environment, eradication of the disease will not be possible." However, after challenging their results with more accurate epidemiologic data, we found that these 3 ctx-positive toxigenic strains could more likely have been present in the sampled rivers as a result of recent fecal contamination ( Figure, http://wwwnc.cdc.gov/EID/article/21/1/14-0627-F1.htm). Indeed, many cholera cases were reported in the corresponding communal sections (i.e., the smallest Haitian administrative unit, average 25 km 2 ) when the samples containing the 7 V. cholerae O1 isolates were collected. In this context of an ongoing cholera epidemic associated with persisting rainfall (Figure), generalized openair defecation inevitably leads to contamination of water sources. It is therefore impossible to determine whether V. cholerae-positive rivers constitute perennial reservoirs of the bacteria or whether they act only as transient vectors of the pathogens.
The recent dramatic decrease in cholera transmission may provide a good opportunity to address this issue (2). We thus encourage Alam et al. to continue the search for ctx-positive toxigenic V. cholerae O1 strains in surface waters, especially during cholera-free periods. To the Editor: Hemorrhagic fever with renal syndrome (HFRS) is caused by hantavirus. During 1995-2005, China reported 20,000-50,000 cases of HFRS annually, which represents 90% of HFRS cases worldwide (1-3). In China, HFRS is caused mainly by 2 serotypes of hantavirus: Hantaan virus (HTNV) and Seoul virus (SEOV) (4). Pathogenic hantavirus serotypes are considered to be strictly associated with their serotype-specific reservoir hosts. HTNV is associated with the striped field mouse (Apodemus agrarius), and SEOV is associated with the brown rat (Rattus norvegicus) and the black rat (Rattus rattus) (4,5). HTNV causes a severe form of HFRS, characterized by renal failure that in some cases is followed by pulmonary edema and disseminated intravascular coagulation; the estimated death rate is 5%-15%. SEOV causes a moderate form of HFRS (6).
We captured 1,276 animals comprising 5 rodent species and 1 shrew species (Table) and analyzed serum antibody against hantavirus of each animal using an antigen sandwich ELISA Kit (Shanghai Jiahe Biotechnology, Shanghai, China). The serum was considered to contain antibodies against hantavirus when the optical density (OD) 450nm of the sample was greater than the threshold. The threshold was calculated by using the equation: threshold = the average OD of the negative control + 0.15. ELISA results showed that 23.3% of animals were seropositive to hantavirus antigen (Table). The seropositive rate to hantavirus was 44.0% in Asian house shrews (Suncus murinus), 25.3% in house mice (Mus musculus), 15.4% in Chinese hamsters (Cricetulus griseus), 10.3% in brown rats, 10.1% in striped field mice (Apodemus agraius), and 3.0% in greater long-tailed hamsters (C. triton). The seropositivity rate for rodents was higher during summer (May-August) and lower during spring (March and April) and winter (October and November) but not significantly different among the months.
To determine what types of hantavirus infected the animals, we amplified viral RNA of HTNV and SEOV from animal lung samples using reverse transcription PCR with serotype-specific primers (7); 2.1% of animals had viral RNA of HTNV, and 2.1% had viral RNA of SEOV (Table). HTNV RNA was detected in striped field mice (6.3%), house mice (1.4%), and brown rats (0.6%). The hantavirus-positive animals were captured in February, April, and November for stripped field mice; November for brown rats; and April and November for house mice. SEOV was detected in brown rats (8.2%) and Asian house shrews (1.7%). These SEOV-positive animals were captured in January, March, May, June, and July for brown rats and March and November for Asian house shrews. The phylogenetic analysis of sequences amplified by reverse transcription PCR is presented in the online Technical Appendix Figure (http://wwwnc.cdc. gov/EID/article/21/1/14-0960-Techapp1.pdf). The nucleotide sequences of the PCR products have been deposited in GenBank (accession nos. KM357423-KM357452).
Hantavirus had been considered to be strictly associated with specific reservoir hosts and to have the same geographic distribution pattern as these reservoir hosts. All hantaviruses that caused human diseases had been associated with rodents, including members of Murinae, Arvicolinae, and Sigmodontinae spp. Insectivore hantaviruses were not known to cause human disease. The rodent hantavirus and the insectivorous hantaviruses were thought to have co-evolved with their specific rodent and insectivorous hosts over millions of years (8). One observed geographic clustering of hantavirus strains, and the association of hantaviruses with their reservoirs, might have been caused by an isolation-by-distance mechanism (9,10) and mixture of both host switching and co-divergence (10). Our study demonstrated that HTNV not only infects its traditional host, the striped mouse, but also infects house mice and rats; SEOV infects not only rats but also shrews, suggesting host expansion for both HTNV