Figure 1

Figure 1. Cowpox lesions on patient’s forearm on day 7 after onset of illness.

A 4-year-old girl from a small farm in eastern Finland was hospitalized in September 2000 because of umbilicated vesicopapules, which developed over the previous 5 days (Figure 1), and unresponsiveness to cephalexin. She had a past history of moderate atopic dermatitis. Animals at her home farm included a horse, three dogs, and a rabbit, but she had no contact with cats because of allergy. On examination, most lesions were located on her swollen red extremities, a few were found on the side of her body, and 3-mm lesions were found on the face and vulva. All lesions were at the same stage of development. She was febrile with 38°C temperature and feeling unwell. A biopsy sample from a papule was sent to a virology laboratory, where orthopoxvirus particles 230 × 300 nm in size were demonstrated by electron microscopy with negative staining.

The girl was treated in isolation at the hospital with chlorhexidine washings and wound dressings with fusidic acid. Intravenous dicloxacillin was administered to prevent secondary bacterial infections. On day 12, skin lesions progressed to deep-seated, hard, black eschars. At the same time the patient’s general condition improved. Two months later, all lesions were healed with scars.

Figure 2

Figure 2. Phylogenetic tree of selected orthopoxvirus hemagglutinin genes based on Clustal X alignment and the maximum likelihood method TreePuzzle. Virus sequences used for the analysis were raccoonpox as an outgroup (GenBank accession...

A homogenized biopsy sample was used to infect Vero cells, where a cytopathic effect typical of orthopoxviruses was seen after 2 days. The isolate, designated CPXV/FIN/T2000, was shown to contain orthopoxvirus by electron microscopy, and DNA samples from both the infected cells and the original biopsy specimen were PCR-positive for orthopoxvirus thymidine kinase gene (12). The hemagglutinin (HA) gene of the isolate was amplified (13), sequenced (948 nucleotides; accession no. AY366477), and compared to other orthopoxviruses. The CPXV/FIN/T2000 strain differed 3% to 4% at the nucleotide level from cowpox virus strains available in GenBank. The sequences were further subjected to phylogenetic analysis: they were aligned by using Clustal X with Gonnet protein matrix and analyzed by using the maximum likelihood phylogenetic software TREE-PUZZLE 5.0, applying the Hasegawa model of substitution and performing 25,000 steps. The HA sequence of CPXV/FIN/T2000 formed a separate clade with cowpox virus strain GRI-90 (Figure 2), isolated originally from another 4-year-old girl, who contracted cowpox after playing with a mole near Moscow (14). The reference cowpox virus strain Brighton grouped with camelpox and variola viruses instead of other cowpox or vaccinia viruses. However, because of the high homology of HA genes, this finding should be interpreted with caution. The remaining cowpox virus strains clustered together with vaccinia viruses; this clustering may, in some cases, be explained by origin from vaccinia virus strains that had escaped to nature (13). This finding seems not to be the case in Finland, since the HA nucleotide sequence of vaccinia virus used in Finland differed 4% both from the strain CPXV/FIN/T2000 and another Finnish cowpox virus isolate from 1989 with an identical HA sequence (data not shown). Different alignment parameters and the neighbor-joining method produced the same results with high support values. Scattered phylogenetic distribution of cowpox virus strains is supported by data presented by other researchers and may reflect an ancestral role of cowpox viruses within the Orthopoxvirus genus; some strains cluster with vaccinia viruses and others (including the reference Brighton strain) with variola virus (1).

An immunofluorescence assay (IFA) to measure specific immunoglobulin (Ig) G, IgM, and avidity of IgG antibodies (15) was established on acetone-fixed, CPXV/FIN/T2000-strain-infected Vero cells. The patient’s serum had a high IgM-antibody titer at admission and low avidity of specific IgG; after 60 days, the IgM level was low and the IgG avidity high (Table 1).

The serum samples from the patient’s pets were collected later and tested with IFA. A hunting dog had antibodies with a titer of 320; another dog had a titer of 20. Thus, a dog might have transmitted the infection from a wild rodent to the patient, although cats are thought to be the main source of human infection (5).

We further studied orthopoxvirus antibodies by IFA in Finnish fauna and humans. The seroprevalence rates for cats and horses were 3.9% and 1.6%, respectively; the rates for wild rodents (when either serum or lung or heart extracts was used) were 0% to 92%, depending on the trapping time and the location (Table 2). In Värtsilä, a rural district in eastern Finland near where the patient lives, antibodies were found in 1 (2.8%) of 36 rodents. However, variation of seroprevalence rates in rodents is also influenced by population dynamics: low prevalences are often found in the increase phase and high prevalences in the peak phase of rodent population fluctuations. The positive rodents were mainly bank voles (Clethrionomys glareolus). The two seropositive horses were from a region near where the patient lives in eastern Finland, and the three seropositive cats were free-roaming and found in southern Finland. In addition, seroprevalence rates of 50% (7/14) and 1.4% (1/73) have been found in foxes and lynxes, respectively, from a limited geographic area in Finland (16).

Sera collected at a Finnish Veterinary meeting in 2001 showed that every person >50 years had orthopoxvirus antibodies, as measured by IFA. The seroprevalence decreased gradually for younger age groups (Table 3), reflecting the gradual cessation of smallpox vaccination (the last vaccinations in Finland occurred in 1977). The average population might have lower seroprevalence rates than veterinarians because of veterinarians’ frequent contact with animals that may harbor orthopoxvirus. Consequently, younger age groups are more susceptible to smallpox and cowpox virus infections.

Although cowpox virus infection usually causes a single, painful, ulcerated vesicopustule and local lymphadenopathy, immunocompromised patients and children, especially those with atopic eczema, are susceptible to a generalized, even lethal, smallpoxlike infection (5,6). An early diagnosis and prompt recognition of the virus are essential for treating and differentiating cowpox, from other orthopoxvirus and herpesvirus infections, especially in severe cases.

Cowpox virus (orthopoxvirus) infection was diagnosed by electron microscopy, PCR, virus isolation, and serologic testing (positive IgM or low avidity of IgG antibodies). Cowpox virus strains show considerable genetic variations with different positioning in the orthopoxvirus phylogenetic tree. Circulation of cowpox virus in wild and domestic animals, together with decreased immunity in humans, may lead to increased occurrence of human cowpox, especially in atopic and immunocompromised persons who are at risk for generalized infection. The described case further suggests atopy to be a contraindication to smallpox vaccination.