Babesiosis in Washington State: A New Species of Babesia?

Case Report

A 41-year-old man from a rural forested area in south-central Washington State went to a local emergency room on 15 September 1991 because of a 1-week history of fever, rigors, anorexia, rhinorrhea, cough, and headache. He had previously been in good health, was not taking any medications, and had never had a blood transfusion. He had a dog, two cats, and two head of cattle and was exposed daily to tick habitats, but he did not recall any tick bites. He had not been outside the Washington-Oregon border region in many years and had never been in areas reported to be endemic for babesiosis or malaria. On evaluation, he had a few rales in his left-lung field and a platelet count of 48 10⁹/L. He was treated for a presumed bronchopneumonia, with intravenous cefazolin to be followed with oral cefixime, and was sent home.

He was hospitalized the next day (September 16) because of persistent fever, severe rigors, and dark-colored urine. He had a temperature of 39.5 C, tenderness in both upper quadrants of his abdomen without hepatosplenomegaly, a normal hematocrit (0.44) and leukocyte count, a platelet count of 46 10⁹/L, 1+ occult blood on urine-dipstick analysis with no cells detected microscopically, a normal chest radiograph, and blood and urine culture results that were negative for bacteria. He was treated with cefazolin and gentamicin without improvement. On September 18, his temperature peaked at 40.4 C; his hematocrit was 0.36, his serum lactate dehydrogenase level was 21.34 kat/L (1280 U/L), and his total bilirubin level was 20.5 mol/L (1.2 mg/dL). Intraerythrocytic ring forms attributed to Plasmodium falciparum malaria were noted on his peripheral blood smear. He was treated with mefloquine and sent home at his request.

He was rehospitalized the next day (September 19) with a temperature of 40 C, rigors, and vomiting. On re-examination of his blood smears from September 15 and 18, intraerythrocytic ring forms (in 1.2% and 3.0% of the erythrocytes) were noted as well as tetrad forms characteristic of Babesia (Figure 1). Therapy with oral quinine (650 mg, three times daily) and intravenous clindamycin (600 mg, four times daily) was begun on September 20, with symptomatic improvement by the next day. By September 23, he was afebrile, his hematocrit was 0.35, the parasitemia had decreased from 3.7% (September 20) to 0.4%, and his platelet count had increased to 143 10⁹/L. On September 24, he was sent home on oral quinine and clindamycin (600 mg, three times daily) to complete a 10-day course of therapy.

On September 30, he was evaluated because of a diffuse urticarial rash. His blood smear was normal, and he was treated with a 12-day tapering course of prednisone for a presumed hypersensitivity reaction. On November 3, when he felt feverish and had a headache, his temperature was 37.5 C, his hematocrit 0.34, and he had a detectable parasitemia of <1%. Urticaria recurred after a dose of quinine (650 mg). Although he appeared to be improving without further therapy for babesiosis, he was treated with intravenous clindamycin (1.2 grams, twice daily for 10 days). His hematocrit decreased to 0.28 on November 7; his anemia had resolved by December 12. Because of slowly resolving fatigue, he did not return to work until 6 January 1992. No parasites were detected on his blood smear in December 1991 or on smears in January, March, July, and September 1992, during which time he remained asymptomatic. Evaluation of his immunologic status with a Western-blot test for human immunodeficiency virus; a liver-spleen scan; quantitative immunoglobulins (including immunoglobulin-G subtypes); and serologic testing for antibody to tetanus toxoid, rubella virus, and streptolysin O, indicated normal results.

Methods

Serologic Studies

Serum specimens from the patient were tested at the Centers for Disease Control and Prevention (CDC) in serial fourfold dilutions by indirect immunofluorescent antibody (IFA) testing [17] for antibody to B. microti and the patient's isolate (WA1), which was propagated in hamsters inoculated with his blood (see below). A titer of at least 64 to B. microti was considered positive. Stored serum specimens from patients in the northeastern United States with B. microti-antibody titers ranging from 64 to greater than 4096 and blood smears with intraerythrocytic ring forms were assayed for IFA reactivity with WA1.

In another laboratory (University of California at Davis [UCD]) with a different IFA protocol [18], serum from the patient was assayed in serial twofold dilutions for reactivity with various Babesia isolates maintained by passage in animals (Table 1). A titer of at least 320 to B. microti was considered positive. Fluorescein-labeled, affinity-purified antibody to human IgG (Kirkegaard & Perry, Gaithersburg, Maryland) was used as the secondary antibody. Blood smears from the patient were examined for B. bovis, Babesia equi, and Babesia bigemina antigens by direct immunofluorescence testing with monoclonal antibodies specific for these species [24-26].

Animal Inoculations

Whole blood specimens from the patient were inoculated intraperitoneally (1-mL inocula) into at least two hamsters (Mesocricetus auratus) or jirds (Mongolian gerbils; Meriones unguiculatus). Giemsa-stained thin smears of blood from the inoculated animals were examined (at least 25 oil-immersion fields per slide) weekly for 6 to 8 weeks. Erythrocytes from hamsters infected with WA1 were washed in Puck's Saline G and were inoculated into a splenectomized 1-year-old female golden Labrador retriever (5.6 10⁹ parasitized erythrocytes were administered intravenously and an equal number, subcutaneously) and into a hamster (9 10⁶ parasitized erythrocytes, intraperitoneally). During the 34-day monitoring period, the dog's clinical status and hematocrit were checked daily for 20 days and then 3 times weekly, and thin smears were examined (>5000 erythrocytes/slide) daily through day 20 and then twice weekly. Pre-and postinoculation serum samples from the dog were assayed for IFA reactivity (UCD) with WA1, B. microti (GI [20] and P20 isolates), and B. gibsoni.

Southern-Blot Analysis

Babesia-infected erythrocytes (P1 pellets) were obtained as previously described [16]; erythrocytes infected with WA1, a human isolate of B. microti (2Bm) [16], and a canine isolate of B. gibsoni (6Bg) were used. Control mammalian leukocytes were separated from uninfected blood of a hamster and a dog by differential centrifugation (400 x g, 4 C, 20 min) on Ficoll-paque (Pharmacia LDB Biotechnology, Piscataway, New Jersey) gradients. After Babesia and leukocyte DNA samples were prepared [27], approximately 1 g of each DNA sample was digested with restriction endonucleases (HindIII or HaeIII; Boehringer Mannheim, Indianapolis, Indiana), as previously described [16]. DNA fragments were separated by electrophoresis in horizontal 0.8% (weight/volume) agarose gels in 45 mM Tris-borate and 1 mM ethylenediaminetetraacetic acid at 40 V for 16 to 18 hours. A Babesia-specific rDNA probe was hybridized to Southern blots of the restriction-endonuclease- digested DNAs; the probe had been produced by polymerase chain reaction amplification of sequences from B. microti DNA, with universal primers directed against highly conserved portions of the nuclear small subunit rRNA gene [16, 28]. Southern blotting, hybridization (high stringency conditions), signal generation (enhanced chemiluminescence gene detection system), and signal detection were conducted, as previously described [16].

Neighborhood Serosurvey

Serum specimens were obtained in December 1991 from the patient's three household members and from a convenience sample of persons who lived within a 20-mile radius of the patient and whose occupations or hobbies put them at risk for exposure to ticks. These persons, referred to as neighbors, completed a questionnaire about exposures and febrile illness. The serum samples were assayed for IFA reactivity (CDC) with WA1 and B. microti (Gray strain) [19]. Whole-blood specimens, obtained in February 1992 from each neighbor who had an antibody titer of at least 256 to WA1, were inoculated intraperitoneally (1-mL inocula) into at least two hamsters or jirds.

Epizootiologic Studies

From March through July 1992, 83 small mammals in the patient's neighborhood and across the border in northern Oregon were trapped. Thin smears of blood from the trapped animals were examined for intraerythrocytic parasites. The animals' spleens were preserved in a 30% glycerol solution and were later homogenized in phosphate-buffered saline, pooled in groups of five to eight, and inoculated intraperitoneally into hamsters; thin smears were examined weekly for 8 weeks.

Whole blood was obtained by cardiac puncture from 39 of the 83 animals after they were anesthetized. These animals included 27 deer mice (Peromyscus maniculatus), 5 chipmunks (Tamias sp.), 4 Townsend's moles (Scapanus townsendii), 1 shrew (Sorex vagrans), 1 California ground squirrel (Citellus beecheyi), and 1 chickaree (Tamiasciurus douglasii). Each animal's blood was inoculated intraperitoneally into a hamster or jird; thin smears were examined weekly for 8 weeks. Forty-four of the 83 animals were trapped with snap traps or otherwise killed: 11 Townsend's moles, 11 gray-tailed voles (Microtus canicaudus), 10 deer mice, 7 vagrant shrews, 1 Trowbridge's shrew (Sorex trowbridgii), 1 black-tailed jack rabbit (Lepus californicus), 1 brush rabbit (Sylvilagus bachmani), 1 western jumping mouse (Zapus princeps), and 1 coast mole (Scapanus orarius).

Attempts to collect live ticks were made from March through July 1992 by examining the trapped animals, flagging vegetation, and obtaining ticks from human and animal patients of physicians and veterinarians in the area. The ticks were identified by standard morphologic criteria. The salivary glands of partially and fully engorged ticks were excised and mounted whole on gelatin-coated slides, which were dried and fixed. The slides were stained by a modified Feulgen reaction and were examined for characteristic babesial stippling of the salivary acini [20].

Results

Serologic Studies

Serologic testing was done to determine if the species of the patient's isolate (WA1) could be identified by differential IFA reactivity with various Babesia species. His serum did not react with B. microti (Gray strain) but reacted strongly with WA1 (see Table 1); the reactivity with WA1 waned over time (Table 2). In contrast, eight stored serum samples from patients who had increased B. microti-antibody titers showed no reactivity with WA1. Strong reactivity was observed when the patient's serum was tested with B. gibsoni, a feline Babesia sp., and one of three strains of Babesia sp. from bighorn sheep, but only weak reactivity was observed with B. microti [GI and P20 isolates] and B. divergens (see Table 1). Monoclonal antibodies specific for B. bovis, B. equi, and B. bigemina did not react with WA1.

Animal Inoculations

Isolate WA1 was obtained by subinoculation of the patient's blood into laboratory animals, in which the infection was highly virulent. The 3 hamsters inoculated with blood obtained during the patient's relapse (5 November 1991) died within 8 days of inoculation; 2 were found dead on days 6 and 8, and the other had a premortem parasitemia on day 8 of greater than 95%. The isolate was serially passaged at CDC to 2 hamsters and 15 jirds, all of which died, frequently by 4 to 5 days postinoculation. The two jirds inoculated with blood obtained from the patient on 4 March 1992, when he was asymptomatic and had an unremarkable blood smear, became parasitemic and died within 18 days of inoculation. However, hamsters inoculated with blood obtained from the patient on 24 July and 22 September 1992 did not develop detectable parasitemia during the 8-week monitoring period.

The splenectomized dog inoculated with WA1-infected erythrocytes developed an antibody titer of 320 to WA1 and of 160 to B. gibsoni by day 27 post-inoculation but did not develop an antibody titer to B. microti or become febrile, anemic, or detectably parasitemic during the 34-day monitoring period. The antibody titer to B. gibsoni was still only 320 on day 34. The hamster that was inoculated with WA1 (at the same time as the dog) became parasitemic.

Southern-Blot Analysis

Hybridization of the chemiluminescent Babesia-specific rDNA probe to Southern blots of HindIII-digested DNA Figure 2 and HaeIII-digested DNA (data not shown) showed that WA1 could be clearly differentiated from B. microti and B. gibsoni. The probe hybridized weakly to mammalian cell DNA, probably because of short regions of homology with nuclear small subunit-like rRNA sequences in the leukocyte DNA.

Neighborhood Serosurvey

To determine if zoonotic transmission of WA1 is prevalent near the patient's home, serum specimens were obtained from 83 persons (3 family members and 80 neighbors), 57 (68.7%) of whom were male. Eighty-two persons (98.8%) were interviewed; their median age was 46 years (range, 8 to 77 years). Thirty-eight (46.3%) recalled finding a tick on their bodies in 1991. None of the serum samples had IFA reactivity with B. microti, but 4 (4.8%) of the 83 participants had antibody titers of 256 to WA1. The hamsters and jirds inoculated with blood from these four persons did not develop detectable parasitemia during the 6-week monitoring period. These persons, three of whom were male, were neighbors of the patient, ranged in age from 39 to 63 years, and had hematocrits greater than 0.40 at the time of the survey. All had had an undiagnosed febrile illness in 1991 with symptoms such as fatigue, anorexia, headache, as well as muscle, joint, and abdominal pain, whereas only 14 (18%) of the other 78 interviewed persons had had such an illness. All four seroreactive persons had outdoor activities at least weekly from spring through fall but did not recall finding a tick on their bodies in 1991. None had traveled to the East Coast in 1990 or 1991; one had traveled to Wisconsin.

Epizootiologic Studies

Small mammals were trapped in an attempt to determine the reservoir host of WA1. Despite concerted efforts, only 83 small mammals (see Methods) were trapped, none of which had splenomegaly or parasites detected on blood smears. The hamsters and jirds inoculated with blood or pooled homogenized splenic tissue from these animals did not become parasitemic.

Ticks were collected to determine if locally abundant ticks serve as vectors of WA1. Of the 235 ticks (226 adult Dermacentor variabilis, 4 adult Ixodes pacificus, 2 subadult Ixodes angustus, and 3 subadult Otobius megnini) that were collected, most were removed by physicians and veterinarians from human and animal patients. Despite concerted efforts, few ticks were collected from trapped rodents, and none were collected by flagging. None of the 100 Dermacentor ticks we examined showed evidence of babesial infection. The Ixodes and Otobius ticks that were collected died; therefore, their salivary glands were not examined.

Discussion

This is the first report of a human case of babesiosis acquired in Washington State and only the fourth reported case in the western United States [2, 14, 15]. It initially was misdiagnosed as a case of P. falciparum malaria, but finding intraerythrocytic tetrad forms facilitated the diagnosis of babesiosis [29]. The patient was assumed to be infected with B. microti; the infecting parasite was morphologically indistinguishable from B. microti and, like B. microti, was infective for hamsters and jirds. However, the clinical course of infection in this middle-aged immunocompetent patient and in laboratory rodents was more severe than that typically observed with B. microti. Further studies showed that the patient's isolate (WA1) was antigenically and genotypically distinct from B. microti.

Despite being relatively young (41 years old) and apparently immunocompetent, the patient had a moderately severe case of babesiosis, for which he was hospitalized. The virulence of WA1 in laboratory animals was surprising as well; hamsters infected with B. microti typically have a slower increase in parasitemia than that observed with WA1 and rarely die from the infection [30-33]. Despite receiving a 10-day course of therapy with quinine and clindamycin [34], the patient had a mild clinical and parasitologic relapse 34 days later, which was treated with a 10-day course of clindamycin. Several months after he became asymptomatic, circulating parasites were detected by animal inoculation. How commonly clinical relapse and persistence of subpatent parasitemia occur after treatment of immunocompetent persons infected with B. microti is unknown. No clinical trials have evaluated the effectiveness of clindamycin without quinine for treating human babesiosis; results of animal studies have been contradictory [35, 36].

WA1 Is Distinct from Babesia microti and Babesia gibsoni

By serologic criteria, WA1 clearly is distinct from three isolates of B. microti obtained from patients on Nantucket Island. The higher antibody titers to the GI and P20 isolates than to the Gray strain most likely are due to interlaboratory variation because of slightly different protocols for IFA testing. Intralaboratory comparisons indicated that WA1 has more antigenic similarity with morphologically similar Babesia species from domestic and wild animals than with B. microti. Although the patient's serum showed the greatest IFA reactivity with B. gibsoni, a parasite of dogs that is not known to be infective for humans, a splenectomized dog inoculated with WA1 did not develop clinical signs of babesiosis, parasitemia, or a substantial antibody titer to WA1. In contrast, five splenectomized dogs experimentally infected with B. gibsoni became febrile, anemic, parasitemic, and seroreactive to B. gibsoni (antibody titers greater than 320) by day 26 postinoculation (Conrad PA, Thomford JW. Unpublished data). These five dogs were inoculated with only 2.6 to 8.5 10⁸ parasitized erythrocytes, whereas the dog inoculated with WA1 received 13- to 43-fold more organisms.

The DNA hybridization patterns noted with a Babesia-specific rDNA probe provide further evidence that WA1 is distinct from B. microti and B. gibsoni, which, like WA1, are small Babesia (intraerythrocytic forms of 1.0 to 2.5 microns). A previous study had shown that hybridization of this probe to Southern blots of restriction-endonuclease-digested DNA from various B. microti isolates from infected patients and white-footed mice (Peromyscus leucopus) on Nantucket Island and in Connecticut all produced the same hybridization pattern, which was distinctly different from the patterns with DNA from various B. gibsoni isolates [16]. The genotypic differences between WA1 and B. microti, along with their differing antigenicity and virulence, suggest that WA1 may represent a different species of Babesiaeither a new species or a previously identified one whose zoonotic potential had not been recognized. DNA sequence analysis of a portion of the nuclear small subunit rRNA gene of WA1 and other babesial organisms is being done to provide further insight about WA1's genetic relatedness to other Babesia.

Public Health Importance of WA1

The public health importance of WA1 remains to be determined. We have not yet identified its tick vector and reservoir host, despite concerted efforts to collect ticks and trap small mammals, and we do not know how commonly persons enter the ecosystem in which transmission occurs. The results of the serosurvey suggest that four of the patient's neighbors may have previously been infected with WA1; their antibody titer of 256 to this isolate was the same as that of the patient, 1 year after his case of babesiosis was diagnosed. None of these neighbors recalled tick exposures, but persons with babesiosis commonly do not [37].

Ixodes ticks serve as vectors of zoonotic babesiosis in other geographic areas. Although in our study few Ixodes ticks were collected by trapping small mammals and none were collected by flagging, the collections were not done during the transmission season during which the patient became infected. Ixodes pacificus, which is a competent experimental vector of B. microti (Telford SR. Unpublished data) and is the main vector of Borrelia burgdorferi in the western United States [38, 39], appears to be the most likely vector of WA1. The low population density of small mammals and the exceptionally dry winter may have contributed to the low yield of I. pacificus ticks [40] during the collections. Ixodes angustus, a tick of small mammals that evidently is the vector of an unidentified rodent Babesia species in Alaska [41], is also a candidate vector of WA1, but it rarely bites humans. The apparent abundance of Dermacentor ticks in the patient's neighborhood may be because most of the ticks collected in the study were from human and animal patients of physicians and veterinarians and because Dermacentor ticks are larger and therefore more easily detected than Ixodes ticks. Although Dermacentor ticks have not been reported to be associated with zoonotic babesiosis, they have been identified as vectors of various Babesia species (for example, Babesia caballi of horses [42]) and, thus, could serve as the vector for WA1. Further efforts to identify the vector and reservoir host of WA1 should be undertaken.

Conclusion

Our report of this case of babesiosis shows that immunocompetent persons can acquire infection in the United States with Babesia species that are morphologically indistinguishable from B. microti but which have antigenic and molecular differences. Some patients elsewhere who were assumed to have been infected with B. microti may have been infected with WA1 or other Babesia species. At least 99 species of Babesia that infect mammals have been described [43, 44], but taxonomy has been based primarily on morphology and host specificity. Our report highlights the inadequacy of such criteria for definitive species identification [45] and highlights the need for increased vigilance for infections with vectorborne pathogens not previously known to infect humans. Improved serodiagnostic and molecular techniques [28] are needed for characterizing Babesia species and elucidating the epidemiology of babesiosis, an emergent zoonosis.

Abbreviation

IFA = indirect immunofluorescent antibody