Use of a single-tube nested real-time PCR assay to facilitate the early diagnosis of acute Q fever.

Q fever is a ubiquitous zoonosis caused by Coxiella burnetii, a small pleomorphic Gram-negative intracellular bacillus that occurs worldwide (1). The clinical manifesta- tions of Q fever are varied, with influenza-like illness, hepa- titis, or pneumonia being the most common presentations in acute infections and infective endocarditis in chronic infec- tions. Since the clinical presentations of Q fever are usually nonspecific, a definite diagnosis mainly requires laboratory confirmation. Serology is currently the diagnostic method of choice for Q fever, with indirect immunofluorescence assay (IFA) be- ing the most commonly used method (1,2). However, as antibodies may not appear until late into the course of the disease, it is often difficult to make an early diagnosis based solely on serology for patients who present early. Since early diagnosis would be helpful both for treatment of patients and to initiate a timely response in the event of a Q fever outbreak, the development of a diagnostic assay based on direct antigen detection or polymerase chain reaction (PCR) would be an invaluable supplement to serology (3,4). Q fever is a notifiable disease in Taiwan, and physicians are required to submit blood samples from patients to the rickettsial reference laboratory of the Taiwan Centers for Disease Control (TCDC) for confirmation. Our current diagnostic method relies on serology using a commercial IFA kit (Focus Diagnostics, Cypress, Calif., USA). The crite- rion for confirmation of acute Q fever is a titer of anti-phase II IgM ≥ 80 in any acute-phase serum sample, or a fourfold increase of anti-phase II IgG titers between paired acute- and convalescent-phase serum samples in patients with symptoms suggestive of Q fever. To facilitate the early diagnosis of acute Q fever, a modified PCR-based assay, namely a single-tube nested real-time PCR (STN-RT PCR) assay, has been developed in our laboratory. The results of our PCR testing and a recommended diagnostic algorithm for Q fever are presented herein. A total of 93 blood samples from patients with confirmed acute Q fever by IFA were tested with our PCR assay. Each patient provided only one blood sample for testing. If a patient had provided two or more blood samples, only the first was retrieved. A further 86 blood samples from sub- jects suspected to have Q fever but who turned out to be Q fever seronegative were selected as negative controls. The sensitivity of using either the STN-RT PCR assay alone or together with serology for the diagnosis of Q fever was compared with that of serology alone by McNemar's test. The influence of IgM antibody in blood samples on the de-