Acute Q Fever and Scrub Typhus, Southern Taiwan

Acute Q fever and scrub typhus are zoonoses endemic to southern Taiwan. Among the 137 patients with acute Q fever (89, 65.0%) or scrub typhus (43, 31.4%), we identified 5 patients (3.6%) who were co-infected with Coxiella burnetii and Orientia tsutsugamushi.


The Study
This study was conducted at E-Da Hospital, a teaching hospital located in Kaohsiung County in southern Taiwan, and approved by its Institute Ethics Committee (E-MRP-096-065) in 2006. Rickettsial diseases are notifiable diseases in Taiwan, and suspected cases with appropriate clinical characteristics are reported to the Center for Disease Con-trol, Taipei, Taiwan for confirmation. Because acute Q fever, scrub typhus, and murine typhus (caused by flea-borne Rickettsia typhi) are the most common rickettsial diseases in Taiwan (3)(4)(5), we requested tests for the etiologic agents of the 3 diseases simultaneously in patients who sought treatment, regardless of which disease we suspected on the basis of clinical features.
Serologic assessments for specific antibodies to C. burnetii and O. tsutsugamushi were performed by using indirect immunofluorescence antibody assay as previously described (5). Acute Q fever was diagnosed by either an antiphase II antigen immunoglobulin (Ig) G titer >320 and antiphase II antigen IgM titer >80 in a single serum sample, or a >4-fold rise of antiphase II antigen IgG titer in paired serum samples. Antigens of 3 major strains of O. tsutsugamushi (Karp, Kato, and Gilliam strains) were used to diagnose scrub typhus: either an IgM titer >80 or a >4-fold rise in IgG titer in paired serum samples for Karp, Kato, and Gilliam strains of O. tsutsugamushi. Murine typhus was diagnosed by an IgM titer >80 or a >4-fold rise in IgG titer against R. typhi in paired sera. Co-infection with C. burnetii and O. tsutsugamushi was diagnosed if the serologic results fulfilled the diagnostic criteria for both infections.
From April 15, 2004, through June 30, 2008, we identified 12 cases of murine typhus, 89 cases of acute Q fever, and 43 cases of scrub typhus; 5 persons had both acute Q fever and scrub typhus. All 5 patients with co-infections denied having fever within 3 months before admission. The demographic data and clinical manifestations of the 5 casepatients co-infected with C. burnetii and O. tsutsugamushi are listed in the online Appendix Table (available from www.cdc.gov/EID/content/15/10/1659-appT.htm). Thrombocytopenia and elevated liver enzyme levels improved after antimicrobial drug treatment.

Conclusions
In this study, we found that 5.3% (5/94) and 10.4% (5/48) patients who fulfilled the serologic diagnosis of acute Q fever and scrub typhus, respectively, were coinfected with C. burnetii and O. tsutsugamushi. Although the major arthropod vectors, animal reservoirs, and route of transmission to humans are different for C. burnetii and O. tsutsugamushi, co-infection with these 2 organisms can occur, particularly in regions where both Q fever and scrub typhus are endemic, such as southern Taiwan. Identification of co-infection improves understanding of the epidemiology of each disease and reminds clinicians to monitor patients for whom the development of chronic Q fever after acute Q fever is a high risk; such follow-up is not needed for the cases of scrub typhus alone.
Few cases of concurrent Q fever and other rickettsioses have been reported in the literature (6,7). Rolain et al.
found that 6 patients with tick-borne rickettsioses had concomitant or consecutive infection with C. burnetii; antibodies were routinely tested against antigens of C. burnetii, Rickettsia spp., Anaplasma phagocytophylum, Francisella tularensis, and Borrelia burgdorferi (6). Five and 2 of their 6 patients had an eschar and skin rash, respectively, and none had a history or clinical signs suggestive of acute Q fever. In contrast, none of our patients had an eschar, the characteristic manifestation of scrub typhus, and only 1 had a skin rash, which made a presumptive clinical diagnosis of scrub typhus difficult.
For scrub typhus, co-infection with Leptospira spp. (8-12) and R. typhi (13) has been reported. Co-infection with O. tsutsugamushi and Leptospira spp. tends to be associated with severe illness and death (8)(9)(10)(11)(12), and cases are mainly reported in Southeast Asia, particularly in Thailand (8) and Taiwan (9)(10)(11)(12). However, we did not routinely test for leptospirosis in this study. Cases of O. tsutsugamushi and R. typhi co-infection had been found in a surveillance of patients with fever of unknown causes, which were possibly cases of scrub typhus in China (13). No co-infection of O. tsutsugamushi and R. typhi was identified in our study, which might be because relatively fewer cases of murine typhus were identified.
In the 6 patients with tick-borne rickettsioses and C. burnetii coinfection reported by Rolain et al., 3 patients likely had concomitant infection caused by tick bites and the other 3 were possibly consecutive infections (6). It was difficult to identify concomitant or consecutive infection in our patients because we could not determine a definite time when patients were bitten by arthropod vectors, had contact with animal reservoirs, or were exposed to the environments abundant in C. burnetii and O. tsutsugamushi. With the assessment of serologic results, however, case-patients 3, 4, and 5 might have acquired scrub typhus first and later been infected with C. burnetii; titers of IgM and IgG against O. tsutsugamushi were high, and antibodies against C. burnetii phase II antigen were negative on first tests in casepatients 4 and 5 (online Appendix Table). In case-patient 3 who sought treatment for fever 20 days before admission, the antibody titers for C. burnetii phase II antigen and O. tsutsugamushi were rising and declining, respectively, in paired serum samples tested. Case-patient 2 might have had Q fever first and scrub typhus later because the first serum tests were positive for C. burnetii phase II IgM and IgG, but negative for O. tsutsugamushi antibodies. For case-patient 1, the rising antibody titers to both pathogens in paired serum specimens suggest a concomitant infection. Because chiggers were rarely reported as arthropod reservoirs of C. burnetii, we believe that our patients might have acquired dual infection through different routes.
The possibility of serologic cross-reaction between Q fever and scrub typhus that results in misdiagnosis of co-infection is low. Cross-reactivity of serologic test results for C. burnetii and Bartonella spp., Legionella pneumophila, L. micdadei, and Ehrlichia chaffeensis have been reported, but there were no data available on cross-reactivity for O. tsutsugamushi (6,14). In conclusion, although major arthropod vectors, animal reservoirs, and routes of transmission to humans are different for C. burnetii and O. tsutsugamushi, co-infection with these 2 organisms may occur, particularly in regions where both Q fever and scrub typhus are endemic.