West Nile Virus Infection in Humans and Horses, Cuba

A surveillance system to detect West Nile virus (WNV) was established in Cuba in 2002. WNV infection was confirmed by serologic assays in 4 asymptomatic horses and 3 humans with encephalitis in 2003 and 2004. These results are the first reported evidence of WNV activity in Cuba.

Serum specimens from horses in Havana and Havana Province were tested for antibodies to WNV by using a competitive enzyme-linked immunosorbent assay (ELISA) with monoclonal antibodies 3.1112G and 6B6C-1 as described by Blitvich et al. (13). We tested 210 serum specimens from horses collected as part of an infectious anemia study. The immunoglobulin M (IgM) test was not performed because horses were never suspected of having WNV and did not have any history of suspected viral encephalitis or other illness or symptoms. An inhibition value >30% was used as the diagnostic criterion to identify flavivirus antibody ( Table 1).
The Cuban Health Ministry and Medical Services division conducted surveillance for encephalitis of unknown origin in patients >30 years of age. Serum and cerebrospinal fluid specimens were shipped at 4°C to the Tropical Medicine Institute "Pedro Kouri." Human sera were screened for WNV IgM and IgG by using commercial IgM and IgG ELISA kits (Focus Technologies, Cypress, CA, USA) according to manufacturer's instructions. Hemagglutination-inhibition (HI) tests were also undertaken with WNV and Saint Louis encephalitis virus (SLEV) antigen (14).
Reactive serum samples were further tested by a plaque reduction neutralization test (PRNT) with WNV (NY99, Ontario, Canada, 2001 isolate), SLEV (Parton strain, American Type Culture Collection catalog no. VR-1265), and dengue virus (dengue 2, NG-C strain). PRNT was performed to confirm WNV-specific antibody and was carried out as described previously (15) by using a neutral red double-overlay procedure. Horses or human patients were considered seropositive for a particular flavivirus if the 90% PRNT titer for that virus was >4-fold greater than the neutralization titers determined for other viruses used in the assay. Endpoint titrations were defined as the highest dilution of serum that reduced plaque formation by >90%.
Most (58%) of the 1,217 birds tested were resident species of Cuba, primarily chestnut manakins (Lonchura malaccas), blue jays (Cyanocitta cristata), herring gulls (Larus argentatus), yellow-faced grassquits/olive finches (Tiaris olivacea), and northern parulas (Parula americana). None of the birds were positive for WNV RNA by RT-PCR. We attempted to isolate WNV from most samples, but virus was not cultured. Most dead birds were received in good condition for testing. Nineteen (9.0%) of the 210 horses had serum specimens with antibodies to flaviviruses (Table 1). Four and 8 animals had WNV-and SLEV-specific antibodies, respectively, in the PRNT. Two horses seropositive for WNV came from Havana City, and 2 others came from Havana Province. Seven serum samples had antibodies to undetermined flaviviruses on the basis of results of neutralization assays, and further virus characterization was not per-formed. None of the horses, including those that were positive for flavivirus antibodies, showed any signs of illness at the time of serum collection.
Serum specimens from 13 patients with encephalitis were tested for possible WNV infection. Two of these patients (M.M. and K.R.) were identified as part of WNV surveillance in 2003. Both acute-phase and convalescentphase serum specimens from these patients were positive for flavivirus antibody by IgM and IgG ELISAs ( Table 2). Cerebrospinal fluid from patient K.R. was also positive for IgM by WNV ELISA. Acute-phase and convalescentphase serum samples from M.M. and K.R. were also positive by HI assay and showed seroconversion to WNV. Convalescent-phase serum samples were tested for WNVspecific antibody by PRNT, and the neutralization titers were 320 and 160, respectively. Neutralizing antibodies against SLEV or dengue virus were not detected in the serum specimens of either patient. These persons are the first to have confirmed cases of WNV-associated illness in Cuba. Both patients had histories of febrile illness, muscle weakness, and encephalitis, and both were hospitalized. These persons had jobs that required them to spend large amounts of time outdoors, and they lived in communities in Santi Spiritus and Villa Clara in central Cuba. Another person (O.G., Table 2), who also resided in Santi Spiritus, had a low WNV titer by HI assay but had neutralizing antibodies to WNV, which suggests a past WNV infection. This patient was identified during surveillance in 2004 but may have been exposed to WNV in 2003.
Serum specimens from the 10 remaining patients were negative for WNV IgM but were positive for flavivirus IgG by ELISA; most of these were also positive by HI assay ( Table 2). One person appeared to have been exposed to dengue virus, and 4 had antibodies to unidentified flaviviruses. Five patients had serum with SLEV-specific antibodies. Seroconversions were not demonstrated in any of these persons, so we cannot say whether their illnesses were associated with SLEV or dengue virus infections.

Conclusions
We report the first evidence of antibodies to WNV in horses and humans in Cuba. The fact that human and horse infections have been detected strongly suggests that a local amplification cycle has been established in Cuba. The mode of entry of the virus into Cuba is unknown.
In North America, avian death from WNV infection has been well documented (3)(4)(5). However, none of the dead birds collected during this study showed evidence of viral infection. Although 1,217 animals were tested, a more intensive dead bird surveillance program may be needed to identify animals that die from WNV infection. Resident bird species in Cuba may be less susceptible to WNV infection, and death rates among birds in the Caribbean may be lower than those observed in Canada and the United States.
This study also provides evidence that suggests WNV and SLEV may co-circulate in Cuba. Further studies are required to confirm these observations and to characterize the transmission cycles involved. Finally, expansion of existing mosquito control programs in Cuba, which currently focus on Aedes aegypti and dengue prevention, may be required to respond to this new public health threat.