Tickborne Relapsing Fever Caused by Borrelia persica, Uzbekistan and Tajikistan

To the Editor: Tickborne relapsing fever (TBRF) is caused by several Borrelia species and transmitted by argasid soft ticks of the genus Ornithodoros. The disease is endemic to many parts of the world, especially Africa (B. duttonii and B. crocidurae most prevalent), and the Mediterranean basin (B. hispanica most prevalent). In Eurasia, TBRF is mainly caused by B. persica (1–3). We report a patient who returned to France with B. persica infection after visiting Uzbekistan and Tajikistan. 
 
In September 2008, a 32-year-old man sought care at the Saint Louis Hospital in Paris, France, for a fifth relapsing episode of fever. Three weeks earlier (July 31–August 18), he had trekked through Uzbekistan and Tajikistan and spent 7 nights in a tent in Uzbekistan. The day before his return to France, he had fever (39.5°C) that lasted 5 days without other symptoms. After a 4-day interval, fever recurred for 1 day, then relapsed 3 other times (every 2 days with fever for 12 hours). Laboratory investigations performed earlier by his general physician on August 19 and 20 showed lymphopenia (0.76 × 109 cells/L and 0.44 × 109 cells/L), thrombocytopenia (94 ×109 cells/L and 80 ×109 cells/L), and C-reactive protein level 300 mg/L (reference level <6 mg/L). 
 
Physical examination when the man was afebrile found no clinical abnormalities. He did not recall tick bites, and no lesion was seen on his skin. Laboratory investigations on September 2 indicated persistant lymphopenia (1.0 × 109 cells/L), C-reactive protein 96 mg/L, and mild cholestasis (alkaline phosphatase level 125 UI/L and gamma glutamyl transferase level 92 UI/L). Thick and thin Giemsa-stained blood smears showed neither Plasmodium spp. nor Borrelia spp. However, quantitative buffy coat analysis (QBC; Becton Dickinson, Le Pont de Claix, France) showed numerous spirochetes, which prompted a rereview and careful analysis of the slides, during which spirochetes were infrequently seen. PCR and sequencing of the 16S rRNA gene, performed as previously described, from a whole blood sample identified these bacteria as B. persica (1). 
 
According to current recommendations in France, the patient was given doxycycline, 200 mg/d, for 10 days. During the first 12 hours, he was monitored for a Jarish–Herxheimer reaction, which was not observed. By the end of therapy, inflammatory syndrome and lymphopenia had resolved. 
 
Despite high-level sequence conservation, identification of Borrelia spp. by sequencing the 16S rRNA gene is reliable and useful for clinical practice (1,4–6). The sequence obtained from the patient reported here was identical (100% identity over 1,472 bp) to the B. persica {"type":"entrez-nucleotide","attrs":{"text":"HM161645","term_id":"304366745","term_text":"HM161645"}}HM161645 reference sequence available from GenBank and sampled from the Galilee region of Israel. It differed by 2 nt (99.86% identity) from B. persica {"type":"entrez-nucleotide","attrs":{"text":"U42297","term_id":"1814215","term_text":"U42297"}}U42297 and another unpublished sequence (B. persica 11/95), each from Iran. B. persica 11/95 was obtained from blood from a rodent collected in Iran and examined at the Institut Pasteur of Iran. We submitted the 2 B. persica sequences (1 from the patient reported here and 1 from the 11/95 sample) to GenBank (accession nos. {"type":"entrez-nucleotide","attrs":{"text":"HQ610930","term_id":"325149333","term_text":"HQ610930"}}HQ610930 and {"type":"entrez-nucleotide","attrs":{"text":"HQ610931","term_id":"325149334","term_text":"HQ610931"}}HQ610931, respectively). We confirmed identification of B. persica from the patient reported here by sequencing of both flaB and intergenic spacer domains (data not shown), as described (7,8). 
 
Although recently reviewed, the epidemiology of TBRF in the area of the former Union of Soviet Socialist Republics has not been extensively described (3). In most of these countries, the infection is mainly attributed to B. persica and transmitted by Ornithodoros tholozani ticks, which live in caves, soil, wall crevices, houses, and cow sheds. Mammalian reservoirs, if any, are not known (3). However, other agents have been reported, such as B. latyschewii (transmitted by O. tartakovskyi ticks, which inhabit rodent burrows) in Kazakhstan, Uzbekistan, and Turkmenistan and B. caucasica (transmitted by O. verrucosus ticks) in the western shore of the Caspian Sea (Armenia, Azerbaijan, and Georgia) (9,10). Other soft ticks (O. nereensis and O. alactagalis) have been described in central Asia. 
 
To our knowledge, TBRF cases caused by B. persica in Uzbekistan and Tajikistan have been rarely reported. The clinical illness of the patient reported here did not differ substantially from that of patients in Israel, Iran, or Jordan, where the infection is more frequently detected. The illness appears benign without chills, headache, vomiting, arthralgia, epistaxis, or hematuria. However, lymphopenia, which resolved rapidly, has not been described for other TBRF cases. 
 
Our report highlights that TBRF is endemic to countries of the former Union of Soviet Socialist Republics. Physicians should consider this diagnosis for febrile patient returning from this area. Efforts to prevent tick bites should be emphasized. Accurate microbiological diagnosis comprises molecular detection or quantitative buffy coat analyses, each of which enhances sensitivity. However, in most disease-endemic countries, diagnosis is based only on examination of direct blood smears, which can lead to false-negative results and underestimation of the actual extent of this infection (1).

In September 2008, a 32-year-old man sought care at the Saint Louis Hospital in Paris, France, for a fi fth relapsing episode of fever. Three weeks earlier (July 31-August 18), he had trekked through Uzbekistan and Tajikistan and spent 7 nights in a tent in Uzbekistan. The day before his return to France, he had fever (39.5°C) that lasted 5 days without other symptoms. After a 4-day interval, fever recurred for 1 day, then relapsed 3 other times (every 2 days with fever for 12 hours). Laboratory LETTERS investigations performed earlier by his general physician on August 19 and 20 showed lymphopenia (0.76 × 10 9 cells/L and 0.44 × 10 9 cells/L), thrombocytopenia (94 ×10 9 cells/L and 80 ×10 9 cells/L), and C-reactive protein level 300 mg/L (reference level <6 mg/L).
Physical examination when the man was afebrile found no clinical abnormalities. He did not recall tick bites, and no lesion was seen on his skin. Laboratory investigations on September 2 indicated persistant lymphopenia (1.0 × 10 9 cells/L), Creactive protein 96 mg/L, and mild cholestasis (alkaline phosphatase level 125 UI/L and gamma glutamyl transferase level 92 UI/L). Thick and thin Giemsa-stained blood smears showed neither Plasmodium spp. nor Borrelia spp. However, quantitative buffy coat analysis (QBC; Becton Dickinson, Le Pont de Claix, France) showed numerous spirochetes, which prompted a rereview and careful analysis of the slides, during which spirochetes were infrequently seen. PCR and sequencing of the 16S rRNA gene, performed as previously described, from a whole blood sample identifi ed these bacteria as B. persica (1).
According to current recommendations in France, the patient was given doxycycline, 200 mg/d, for 10 days. During the fi rst 12 hours, he was monitored for a Jarish-Herxheimer reaction, which was not observed. By the end of therapy, infl ammatory syndrome and lymphopenia had resolved.
Despite high-level sequence conservation, identifi cation of Borrelia spp. by sequencing the 16S rRNA gene is reliable and useful for clinical practice (1,(4)(5)(6). The sequence obtained from the patient reported here was identical (100% identity over 1,472 bp) to the B. persica HM161645 reference sequence available from GenBank and sampled from the Galilee region of Israel. It differed by 2 nt (99.86% identity) from B. persica U42297 and another unpublished sequence (B. persica 11/95), each from Iran. B. persica 11/95 was obtained from blood from a rodent collected in Iran and examined at the Institut Pasteur of Iran. We submitted the 2 B. persica sequences (1 from the patient reported here and 1 from the 11/95 sample) to GenBank (accession nos. HQ610930 and HQ610931, respectively). We confi rmed identifi cation of B. persica from the patient reported here by sequencing of both fl aB and intergenic spacer domains (data not shown), as described (7,8).
Although recently reviewed, the epidemiology of TBRF in the area of the former Union of Soviet Socialist Republics has not been extensively described (3) To our knowledge, TBRF cases caused by B. persica in Uzbekistan and Tajikistan have been rarely reported. The clinical illness of the patient reported here did not differ substantially from that of patients in Israel, Iran, or Jordan, where the infection is more frequently detected. The illness appears benign without chills, headache, vomiting, arthralgia, epistaxis, or hematuria. However, lymphopenia, which resolved rapidly, has not been described for other TBRF cases.
Our report highlights that TBRF is endemic to countries of the former Union of Soviet Socialist Republics. Physicians should consider this diagnosis for febrile patient returning from this area. Efforts to prevent tick bites should be emphasized. Accurate microbiological diagnosis comprises molecular detection or quantitative buffy coat analyses, each of which enhances sensitivity. However, in most disease-endemic countries, diagnosis is based only on examination of direct blood smears, which can lead to false-negative results and underestimation of the actual extent of this infection (1).

Toxoplasmosis and Horse Meat, France
To the Editor: Toxoplasma gondii parasites are obligate intracellular apicomplexans that can infect virtually all warm-blooded animals; felids are defi nitive hosts. The most common sources of human infection are ingestion of tissue cysts in undercooked meat or of food or water contaminated with oocysts shed by felids and transplacental transmission. Acquired toxoplasmosis in immunocompetent humans is frequently asymptomatic but is associated with cervical or occipital lymphadenopathy in ≈10% of patients. Severe or fatal outcomes for immunocompetent patients have been attributed to the virulence of specifi c T. gondii genotypes (1). We describe 3 cases of toxoplasmosis caused by atypical strains probably acquired by from ingestion of raw horse meat imported from Canada and Brazil. Patient 1, a 74-year-old man, was hospitalized locally (Antibes-Juan Les Pins, southern France) in March 2009 for asthenia and persistent febrile bronchitis. His medical history included severe smoking-related chronic obstructive pulmonary disease and coronary artery disease. He received broad-spectrum antimicrobial drugs and methylprednisolone. On day 23 after admission, he was transferred to our teaching hospital in Nice because of clinical deterioration and persistent fever. Disseminated toxoplasmosis was diagnosed on the basis of serologic evidence of recent primary T. gondii parasite infection and quantitative PCR detection of high Toxoplasma DNA levels in peripheral blood. Despite specifi c antitoxoplasma therapy with sulfadiazine and pyrimethamine, he remained febrile, his respiratory function worsened, and he died on day 27. Patient 2, a 24-year-old pregnant woman, was hospitalized in Draguignan, France, in December 2009 for full-term delivery. Three weeks earlier, routine serologic testing showed T. gondii parasite infection seroconversion. The newborn's and mother's ophthalmologic examinations were unremarkable. Congenital toxoplasmosis was diagnosed on the basis immunoglobulin M in the infant's serum, positive quantitative PCR of samples from the placenta, and strain isolation after inoculation of mice with a placental preparation. Sulfadiazine and pyrimethamine were started.
We performed a retrospective epidemiologic investigation of an unusual case of toxoplasmosis that occurred in March 1991. Patient 3, a 21-year-old pregnant woman living in the Nice area, was treated with spiramycine because routine serologic testing had shown T. gondii parasite infection seroconversion at 22 weeks' gestation. Amniocentesis showed T. gondii tachyzoites in amniotic fl uid by microscopic examination. At 26 weeks' gestation, the woman underwent termination of pregnancy for ultrasonography-detected fetal severe abnormalities. Fetal necropsy showed numerous cerebral, cardiac, and hepatic abscesses with T. gondii tachyzoites. A few days after pregnancy termination, the woman experienced cervical lymphadenopathy, which lasted 3 years. She reported having eaten raw horse meat regularly during her pregnancy.
Genetic analyses with microsatellite markers of the Toxoplasma spp. strains isolated from the 3 patients found 3 different atypical genotypes. Atypical strains are common in South America but unusual in France, where >95% of reported strains collected from human and animal toxoplasmosis cases belonged to the type II clonal lineage (2,3). Hence, isolation of an atypical Toxoplasma genotype from a patient in France strongly suggests contamination by a non-European strain, either during residence abroad or after ingestion of imported meat.