The serological evolution from acute Q fever to endocarditis is reported for 22 patients, with a median lag time of 3 months. From this data, we propose a follow-up strategy of serological testing at the third and sixth months after acute Q fever to obtain an early diagnosis of chronic infection.
Q fever is a zoonosis due to Coxiella burnetii infection . The disease is commonly divided into acute and chronic infections. In symptomatic acute Q fever (40% of cases), the common clinical manifestations are fever, pneumonia, and hepatitis. Chronic Q fever develops in patients who are predisposed to the disease . Patients with valvulopathies are at risk to develop Q fever endocarditis, as are immunocompromised patients. Because Q fever endocarditis is an indolent disease, it is often diagnosed only after significant valvular damage has occurred. The diagnosis of Q fever relies mainly on serological examination. Proper management of Q fever is necessary to prevent possible later endocarditis.
In previous studies, we estimated that the risk of transformation from acute Q fever to endocarditis is ∼40%, we defined therapy modalities of acute Q fever in patients with identified valvulopathy , and we modified the Duke criteria to improve the diagnosis of Q fever endocarditis . We reported the histories of patients who had developed endocarditis with previously undiagnosed valvulopathy, confirming that minimal valvulopathies, such as minimal valvular leak, mitral valve prolapse, and bicuspid aortic valve predisposed patients with acute Q fever to endocarditis . We proposed that, for all patients with acute Q fever, a transthoracic echocardiography be performed . Finally, we demonstrated that the use of PCR with serum samples may be helpful in establishing an early diagnosis of chronic Q fever .
An accurate understanding of the evolution of endocarditis from acute Q fever is needed. Literature reports based on clinical observations describe endocarditis as occurring from months to several years after acute Q fever infection, the longest interval being 20 years after infection [7, 8]. To better understand the pathological processes involved in this transition, we conducted a retrospective serological analysis to compare with clinical observations. For 22 patients, we followed the phase I IgG kinetic from acute Q fever to endocarditis and evaluated the time of occurring endocarditis.
Methods.The French National Reference Center for Rickettsial Diseases (Marseille, France) performs Q fever serological testing and confirmation for all regions throughout France, as well as for other countries. Every time the findings of this testing are consistent with Q fever, physicians are asked to provide further serum samples and clinical data. Since 1985, a total of 153,479 serum samples have been tested. Since 1991, test results have been compiled in a database. From this compilation, we retrieved 22 patients who underwent serological testing for acute Q fever and whose clinical evolution of disease was known. Fifteen of these patients were observed by 1 of the authors of this review (D.R.).
Titers of IgG, IgM, and IgA antibodies in serum samples obtained from each patient were determined using an indirect immunofluorescence assay. All serum samples were diluted at ratios of 1 : 25, 1 : 50, and 1 : 100, and were screened for total immunoglobulin. To avoid interference, rheumatoid factor adsorbent (Behring) was used to remove IgG before determination of IgM and IgA titers. If the serological screening for Q fever was positive, the final titers for IgG, IgA, and IgM were determined for both anti–phase I and anti–phase II antibodies .
Our criteria for serological diagnosis of acute Q fever were a titer of anti–phase II IgG antibodies of ⩾1 : 200 and a titer of anti–phase II IgM antibodies of ⩾1 : 50 . Patients were considered to have acute Q fever when they had titer evidence of seroconversion for specific antibodies and presented with a suggestive clinical background, including history of fever, hepatitis, or atypical pneumonia.
Patients were defined as having Q fever endocarditis on the basis of the modified Duke criteria, as previously described [4, 10]. C. burnetii anti–phase I IgG titers of ⩾800 were considered to be a major criteria for the diagnosis of endocarditis.
Results.Twenty-two patients (14 men and 8 women) were identified as having endocarditis. The mean age of these patients was 59.7 years (range, 44–76 years) (table 1). The cardinal clinical manifestations of acute Q fever were nude fever in 5 patients (22.7%), hepatitis in 14 patients (63.6%), and febrile pneumonia in 3 patients (13.6%). All findings of serological testing were compatible with acute Q fever. Among the patients who had underlying disease, we identified 1 patient with diabetes, 1 with alcoholic intoxication, and 2 with cancer (1 of whom had lymphoma). Five patients were not known to have any heart dysfunctions at the time they presented with acute Q fever. Seventeen patients (77.2%) presented with previous and known cardiovascular abnormalities. Of these, 12 patients (70.5%) presented with valvular insufficiency, 2 (11.7%) with prosthetic valves, 1 (5.8%) with mitral valve prolapse, and 2 (11.7%) with valvular stenosis. At the time of acute Q fever, cardiovascular exploration by transesophageal echocardiography was performed in only 1 patient. Except for 3 patients, all patients underwent regular follow-up serological testing after experiencing acute Q fever.
When a diagnosis of chronic Q fever was made, 17 patients (77.2%) were symptomatic. Eleven patients (50%) had cardiologic clinical abnormalities with increased murmur and valvular insufficiency or worsened cardiac failure. The other clinical manifestations were fever in 5 patients (22.7%), purpura in 1 patient (4.5%), and asthenia in 2 patients (9%). At the time of chronic Q fever diagnosis, transesophageal echocardiography was performed on the 14 patients (63.6%) who showed acute valvular insufficiency or paraprosthetic leakage. Two bicuspid aortic valves (9%), 1 mitral valve insufficiency (4.5%), and 1 mitral valve prolapse (4.5%) were identified among the patients who had unknown valvulopathic findings. Reports of 3 of these case patients have already been published . Valvular vegetations were identified in 4 patients. One patient underwent a valvular replacement. A diagnosis of endocarditis, using the modified Duke criteria, was definite in 5 patients (22.7%) and was possible in all the other patients (17 patients [77.2%]) .
The serological diagnosis of chronic Q fever was made between 1 month and 4 years, with a median time to diagnosis of 3 months, after experiencing acute Q fever. Seventeen patients (77.2%) developed chronic Q fever within the 6 months after experiencing acute Q fever (figure 1A). Of the 4 patients who experienced delayed development of chronic Q fever (up to 1 year after experiencing acute Q fever), 2 patients had a prosthetic valve, 1 patient had a mitral valve prolapse, and 1 patient had a bicuspid aortic valve. At the time of earlier diagnosis of endocarditis, the range of titers of phase I IgG were 800–12,800 (median, 800). For 5 of the patients who developed endocarditis within the first year after experiencing acute Q fever, serological follow-up results are summarized in figure 1B. A linear correlation can be observed between increasing titer values and diagnosis of endocarditis. Chronic Q fever titers were performed for patient 1 within the first month after experiencing acute Q fever, but a diagnosis of endocarditis was not made until 1 year later.
Discussion.For a long time, a diagnosis of Q fever endocarditis was made when valvular damage was severe and required surgical treatment . Because clinical manifestations leading to the diagnosis of endocarditis are protean and nonspecific, diagnosis was often delayed for several months or even years . This delay in diagnosis had a significant effect on a patient's prognosis. Today, a better knowledge of the disease, as well as better diagnostic tools (such as establishment of an accurate cutoff using microimmunofluorescence) and the serological follow-up of patients with acute Q fever and valvulopathies have allowed for earlier diagnosis [7, 10]. In our last report, we strongly encouraged physicians to look for cardiac abnormalities in patients who have acute Q fever [3, 5].
However, it is sometimes difficult to diagnose valvulopathies. It is very important to note that a diagnosis of bicuspid aortic valve was not made. Indeed, for patient 1, the diagnosis of bicuspid aortic valve was only possible using transesophageal echocardiography, in spite of the fact that the patient underwent 2 successive transthoracic echocardiographies (the second of which was performed by an expert in the field [F.T.]). In fact, transesophageal echocardiography was performed because the results of serological testing were highly suggestive of chronic Q fever (anti–phase I IgG titer, 12,800).
These cases, in conjunction with our positive serological correlations, suggest that ongoing regular serological monitoring must be performed in patients following acute Q fever. The patient is at high risk of developing endocarditis if an increase in the levels of phase I antibodies is noted. This is particularly true if valvulopathy was previously undiagnosed. In our study, we show that there is variability in the delay of passage from acute to chronic Q fever. In most cases where chronic Q fever will develop, the evolution is early, with one-half of the patients developing chronic Q fever within 3 months and up to 75% of patients developing chronic Q fever within 6 months. It is unfortunate that, in 2 of our cases, the diagnosis of endocarditis was not made until ∼4 years after the patients experienced acute Q fever. At the time of diagnosis of endocarditis, titer levels of phase I IgG were inordinately high (⩾25,800), and the patients were experiencing valvular damage and cardiac failure. As predicted by our model, lower titers (⩽1600) were not associated with cardiac failure and a need for surgery. The asymptomatic stage of Q fever endocarditis may, in part, explain the previously recognized time lag between acute Q fever and diagnosis of endocarditis. Moreover, it is now clear that some strains of C. burnetii can cause more asymptomatic infection, which may make follow-up impossible for such cases [11, 12].
In conclusion, we previously recommended that transthoracic echocardiography be used to detect valve lesions in patients with acute Q fever and to treat those who have valve lesions [3, 5]. Here, we propose that patients with acute Q fever be systematically tested—including those patients who do not have detected underlying factors—3 and 6 months after the onset of disease. Those with phase I IgG antibody titers ⩾1 : 800 should be investigated for possible infective endocarditis  using transesophageal echocardiography and PCR  to allow for early detection of the disease (appendix, figure A1).
We thank Melanie Irhig for reviewing the manuscript.
Potential conflicts of interest.D.R. is a cofounder of a start-up company (INODIAG) that is involved in serological diagnostic testing. He holds <5% of the company's capital and has never received money from it. INODIAG will not receive benefits before the year 2011. His affiliation (Université de la Méditerranée, Marseille, France) has a contract with INODIAG to assist with the production of antigens and strains. All other authors: no conflicts.