From the Southern Association for Vascular Surgery

Abdominal aortic aneurysm and Coxiella burnetiiinfection: Report of three cases and reviewof the literatureCarmine Sessa, MD, PhD,a Lulzim Vokrri, MD,a Paolo Porcu, MD,a Max Maurin, MD, PhD,b

Jean Paul Stahl, MD,c and Jean-Luc Magne, MD,a Grenoble, France

Coxiella burnetii, the etiologic agent of Q fever, is mainly responsible for endocarditis with negative blood culture results,but only a few cases of C. burnetii infections of aortic aneurysms have been published. We report three cases of abdominalaortic aneurysms treated in patients with Q fever infection with simultaneous endocarditis (n � 1) and previous historyof cardiac valve replacement for endocarditis (n � 1). A coeliac aortic aneurysm was diagnosed in one patient treated foracute Q fever with persistent serologic results showing chronic infection despite adequate antibiotic therapy and withoutendocarditis. Resection of the aneurysm cured the chronic infection, and C. burnetii was identified by culture of theaneurysmal wall. In the two other cases, chronic infection of C. burnetii was diagnosed by serologic examination aftersurgery for an abdominal aortic aneurysm. One patient with negative blood culture results had amaurosis fugax due toendocarditis and required aortic valve replacement; recurrent fever without evidence of valve dysfunction or infectiondeveloped in one patient who had had prosthetic cardiac valve replacement 6 months earlier for endocarditis. Aorticaneurysms were treated with in situ prosthetic grafts and long-term antibiotic therapy. At a mean follow-up of 12 years,no septic aortic complications occurred, and serologic test results have remained negative. The presence of an aorticaneurysm and cardiac valve disease seems to be a predisposing factor for chronic C. burnetii infection. Diagnosisparticularly relies on the physician’s awareness of this condition and is confirmed by serologic examination. Aorticaneurysm resection is mandatory to cure the chronic infection and must be associated with long-term antibiotic therapy.

(J Vasc Surg 2005;42:153-8.)

Coxiella burnetii is the etiologic agent of Q fever, azoonosis with a worldwide distribution, especially in ruralareas.1,2 Symptoms of Q fever are polymorphic and non-specific; infection may be acute or chronic, although almost60% of Q fever cases are asymptomatic.3-5 Endocarditiswith negative blood culture results is the most commonclinical presentation of chronic Q fever and is found inabout 60% to 70% of the cases.6-9 Other manifestations areosteoarticular infection, chronic hepatitis, chronic pulmo-nary infections, and isolated fever.10-12 Underlying cardio-vascular disease, such as cardiac valve defect and aorticaneurysm, is a risk factor for chronic Q fever infection,although the incidence of the disease is probably underes-timated.12-14

Because the clinical manifestations of C. burnetii vas-cular infection are nonspecific, the diagnosis may be recog-nized only when serologic tests are performed systemati-cally in patients with arterial aneurysms, vascular grafts, andendocarditis who live in an endemic area and who presentwith unexplained fever and negative blood culture re-sults.15-24 C.burnetii can also be isolated by culture25 and

From the Departments of Vascular and Thoracic Surgery,a Bacteriology,b

and Infectious Diseases,c Centre Hospitalo-Universitaire de Grenoble.Competition of interest: none.Reprint requests: Carmine Sessa, Department of Vascular and Thoracic

Surgery, BP 217, Centre Hospitalo-Universitaire de Grenoble, 38043Grenoble, France (e-mail: csessa@chu-grenoble.fr).

0741-5214/$30.00Copyright © 2005 by The Society for Vascular Surgery.

doi:10.1016/j.jvs.2005.03.022

specific polymerase chain reaction (PCR) amplificationfrom blood and aortic aneurysm specimens.15-21, 26-30

We report three cases of aortic aneurysms operated onin patients with Q fever and with simultaneous endocarditisin one patient and a prior history of cardiac valve replace-ment for endocarditis in another.

CASE REPORTS

Case 1. This case has previously been reported.18 A 65-year-old woman, who was a heavy smoker, presented in February 1988with general physical deterioration, weight loss, fever, and abdom-inal pain due to granulomatous hepatitis diagnosed by a biopsyspecimen. An abdominal computed tomography (CT) scanshowed a 40-mm saccular coeliac aortic aneurysm (Fig 1). Cardiacechography showed no evidence of endocarditis.

Serologic tests revealed acute C. burnetii infection (Fig 2).The patient was started on doxycycline for 3 weeks, and presented1 month after the discontinuation of antibiotics with recurrentfever. Given the recurrent infection, rifampin was given withdoxycycline until November 1988. Serologic tests showed thepersistence of phase II antibodies (immunoglobulin G [IgG],1,600) and the appearance of phase I antibodies (IgG, 400).Diagnosis of chronic Q fever was confirmed, and the patientcontinued to receive antibiotic therapy until September 1989because of the persistence of fever and the increment of phase I IgGantibodies �1,600.

In May 1990, a new episode of fever in the absence ofendocarditis led to resection of her aortic aneurysm (Fig 3), withplacement of a prosthetic graft. C. burnetii was isolated by culture

from blood, the aneurysmal aortic wall, and the thrombus at the

153

JOURNAL OF VASCULAR SURGERYJuly 2005154 Sessa et al

National Reference Center and World Health Organization Col-laborative Center for Rickettsial Diseases in Marseille.

Antibiotic therapy with doxycycline and rifampin was main-tained for 3 years and then stopped when her phase I IgG antibodytiter normalized. After a 15-year follow-up, the patient has had norecurrence of fever. Phase I IgG antibodies have been stable(�200). A CT scan showed moderate and stable dilatation of theaortic graft (Fig 4) and of the abdominal aorta (Fig 5) of �40 mmin diameter.

Case 2. A 47-year-old man presented in February 1992 withan episode of subacute ischemia of the right lower limb as aconsequence of popliteal artery aneurysm thrombosis. A vascularevaluation showed a 35-mm saccular aneurysm of the abdominalaorta extending to the iliac arteries. The patient had a cardiacmurmur. He presented with an episode of sweating after havingspent time in the mountains with pets.

His aortic aneurysm was treated by resection and placement ofan aortobiiliac graft. On postoperative day 2, the patient had twoepisodes of amaurosis fugax. The results of a carotid duplex scanwere normal. Cardiac echography showed endocarditis of theaortic valve with presence of vegetations. Transesophageal echog-raphy confirmed the presence of vegetations and showed a periaor-tic abscess. Blood culture test results remained negative, and aserologic examination showed phase I IgG antibodies �1,600,allowing a diagnosis of chronic Q fever. Surgery of the aortic valvewas undertaken with a prosthetic St. Jude valve (St Jude Medical,St. Paul, Minn). No organisms grew on the aortic valve, althougha specific culture for C. burnetii was not done.

Antibiotic therapy with doxycycline-rifampin was started for 6months and then switched to doxycycline-ofloxacin for 4 years.With a follow-up of 13 years, serologic test results have remainednegative, with phase I IgG antibodies �200. No complicationsfrom the cardiac valve or the aortic graft have occurred.

Case 3. A 55-year-old man underwent resection of an ab-dominal aortic aneurysm extending to the iliac arteries with place-ment of an aortobiiliac bifurcated prosthetic graft in October1992. An inflammatory reaction was noted on the right iliac

Fig 1. Computed tomography scan of the patient in case 1showsa saccular coeliac aortic aneurysm.

aneurysm involving the ureter. The patient’s history was relevant

for aortic valve replacement with aortocoronary bypass 6 monthsearlier.

In the postoperative period from cardiac surgery, an episode offever of unknown origin developed. Results of blood culturesremained negative, and fever resolved after 3 weeks of antibiotics.During the postoperative period from the aortic surgery, a newepisode of fever occurred. Results of blood cultures were negative.

Given the patient’s previous history of aortic valve replace-ment and the presence of an aortic aneurysm, a C. burnetii infec-tion was suspected and confirmed by serologic tests that showedchronic infection with a high titer of phase I antibodies (1,280).Ciprofloxacin was started for 1 year and then switched to doxycy-cline for 6 months because of tendinitis. A later follow-up exami-nation revealed moderate dilatation of the aortic graft thatprompted renewal of ciprofloxacin for 18 months. The graft hasremained stable after 12 years of follow-up, without evidence ofinfection. Serologic test results for C. burnetii have remainednegative, with phase I IgG antibodies �200.

DISCUSSION

Infection of an aortic aneurysm or vascular graft is anuncommon complication. The more commonly involvedorganisms are staphylococcus and salmonella, which can beeasily identified by culture.31,32 However, about 25% ofinfected aneurysms and 5% to 8% of graft infections remainwithout bacterial identification.33,34 This can be because ofwidespread use of antibiotics before specimens are tested orthe presence of intracellular pathogens, such as C. burnetii,that are difficult to grow in culture. Diagnosis can beachieved by serologic examination or PCR amplification.3,4

C. burnetii is the etiologic agent of Q fever, a zoonosiswith a worldwide distribution in rural and, recently, inurban areas.3,4 Q fever is endemic and has been described inalmost every country except New Zealand. The reservoir islarge and includes many wild and domestic animals.1-4 Theaerosol route by inhalation is the primary mode of humancontamination by C. burnetii,2 which may occur directlyfrom infected animals. The organism may also be spread bythe wind,2 and Q fever may occur in patients who live inurban areas without any evident contact with animals.

Infection in humans is usually asymptomatic or mani-fests as a mild disease such as fever, weight loss, andabdominal pain, with spontaneous recovery. Because theclinical presentation of Q fever is polymorphic and nonspe-cific, the incidence of Q fever among humans is probablyunderestimated. Diagnosis particularly relies upon physi-cian’s interest and awareness of the symptoms of Q fever aswell as access to reliable diagnostic laboratories. However,when evoked, a definitive diagnosis of the disease is easyand is confirmed from serologic examination, with phase IIand phase I antibodies distinguishing acute from chronicdisease.

It is now well established that development of the chronicinfection is more often due to predisposing host factors thanthe virulence of the bacterium. Patients at risk from chronic Qfever include those with previous cardiac valve defects and, toa lesser extent, those with aneurysms or vascular grafts as well

as immunocompromised patients.12-14,15-21

Imm

JOURNAL OF VASCULAR SURGERYVolume 42, Number 1 Sessa et al 155

Endocarditis with negative blood culture results is themost frequent clinical presentation of chronic Q fever,accounting for 60% to 70% of the cases.6,8 Embolic mani-festations that involve the brain and upper and lower limbsare mainly due to endocarditis and are observed in 20% ofpatients.10-11,15,19 The incidence of endocarditis due to C.burnetii infection is 5% in France and 3% in the UnitedKingdom; the incidence is poorly defined in the United

0

500

1000

1500

2000

2500

3000

Acute phase�diagnosis

Antibodies titer

Serologies from the

National Reference Center and World Health Organization

Collaborative Center for RickettsiDiseases in Marseille

* *

1988 1989

Fig 2. Serologic profile of patient in case 1. Phase IIrespectively. *Recurrent fever after antibiotics discontinuaneurysm resection with 3 years antibiotic therapy. IgG,

Fig 3. An intraoperative view of coeliac aortic aneurysm in thecase 1 patient.

States.4

Cardiovascular disease may favor passage from acute tochronic Q fever despite adequate antibiotic therapy. This isdue to the affinity of C. burnetii for cardiovascular tissuesand its ability, as an obligate intracellular microorganism, tosurvive after gaining entry into that monocytes and macro-phages that are present in the aortic thrombus and indamaged cardiac valves.35 This situation has been observedin our experience and gives pathophysiologic support to thefact that host conditions—and above all, endocarditis andaortic aneurysms—are the predisposing conditions for thedevelopment of chronic Q fever despite adequate antibiotictherapy.4,18,20 In fact, the patient of case 1 presented withan infection relapse after antibiotic discontinuation, and hertiter of phase I IgG antibodies decreased only after theaortic aneurysm was resected (Fig 2).

Recent studies have suggested that surgery is manda-tory for abdominal aortic aneurysms �50 mm in diameter,but indications are not well defined in the setting of C.burneti.36,37 Because an aortic aneurysm represents a riskfactor for the development of chronic C. burnetii disease,we believe that with the currently available antibiotic regi-men, surgery should be undertaken for aortic aneurysms�50 mm in patients with or without endocarditis if feverscontinue and if the results of serologic examinations remainelevated despite adequate antibiotic treatment.

In case 1, fever discontinued and the phase I IgGantibody titer decreased only after resection of the coeliac

Serological follow-up in Grenoble

years

*

1991 1992 1994 1995 1996 1998 2003

Phase II-IgG

Phase II-IgA

Phase I-IgG

Phase I-IgA

Phase I are indicative of acute and chronic infection,; **recurrent fever despite antibiotics; ***coeliac aorticunoglobulin G; IgA, immunoglobulin A.

al

*

**

1990

andation

aortic aneurysm, with the patient being considered cured

JOURNAL OF VASCULAR SURGERYJuly 2005156 Sessa et al

after 15 years (Fig 2). Despite dilatation of the aortic graft(Fig 4) and a 40-mm abdominal aortic aneurysm (Fig 5),serologic results have remained negative, with no signs ofrecurrent or persistent infection.

The association of an abdominal aortic aneurysm withproven (case 2) or suspected endocarditis (case 3) has notbeen previously published in the literature. In our series, it islikely that case 2 was a noninfected aneurysm with simulta-neous, undiagnosed C. burnetii endocarditis, whereas case 3

Fig 4. A computed tomography scan taken at the 15-ytation.

Fig 5. A computed tomography scan taken at the 15-yeaneurysm.

was a noninfected aneurysm with prior endocarditis that was

undertreated and not diagnosed as C. burnetii infection. Infact, there was no intraoperative evidence of aortic aneurysminfection, although case 3 had slight periaortic inflammation.It is probable that aortic aneurysm and endocarditis are aunified expression of C. burnetii infection in that they repre-sent a cardiovascular predisposing condition for developmentof chronic infection.

The 13 cases of C. burnetii infection reported literatureinvolve nine aneurysms (infrarenal aorta, 6; ascending

llow-up of the patient in case 1 shows aortic graft dila-

ow-up of the patient in case 1 shows an abdominal aortic

ear fo

ar foll

aorta, 1; thoracic aorta, 1; and suprarenal aorta, 1) and four

JOURNAL OF VASCULAR SURGERYVolume 42, Number 1 Sessa et al 157

vascular grafts (aortobifemoral bypass, 2; abdominal aorticgraft, 1; and subclavian-to-subclavian bypass, 1).15-21 Theclinical manifestations at diagnosis were a fever of �38°C in7 patients, weight loss in 6, lumbar or abdominal pain (orboth) in 4, embolic phenomena in 2, symptoms of hepatitisin 1, and myalgia, nausea, and vomiting in 1.

The diagnosis of Q fever is generally established byserologic tests22-24 and, more recently, by DNA amplifica-tion using PCR-based techniques.26-30 Isolation of C. bur-netii by culture is done only in very few laboratories becauseof the risk of transmission to laboratory personnel and thelack of sensitivity of the technique, particularly when anti-biotics have been previously administered. In all the casesreported in the literature, including our three cases, thediagnosis was achieved by serologic tests, and C. burnetiiwas isolated by culture in 10 cases. PCR amplification wasmade in 10 cases and was positive in 7 cases.

Serologic testing allows differentiation between acuteand chronic Q infection from the antibodies profile (Fig2).22,23 Acute Q fever is characterized by antiphase IIantibodies, whereas antiphase I antibodies are seen in thechronic phase. A single serologic examination enables thediagnosis of chronic Q fever. Nevertheless, the differentserologic techniques are not well standardized, and theresults are difficult to compare between the different labo-ratories.

The National Reference Center and World HealthOrganization Collaborative Center for Rickettsial Diseasesin Marseille has established the threshold of sensitivity forthe antibodies titer of C. burnetii infection.23,24 An IgGanti-phase II titer of �200 and an IgM antiphase II titer of�50 are diagnostic of recent infection. In chronic Q fever,IgA antiphase I titers are �100 and IgG antiphase I titersare �800.

More recently, C. burnetii DNA amplification from clin-ical specimens by PCR has been used successfully.29 With thismethod, C. burnetii DNA can be detected retrospectively infrozen samples and even in paraffin-embedded tissues. 29,30

An undiagnosed or untreated C. burnetii infection ofan aortic aneurysm or a vascular graft can lead to severecomplications. Of the 13 patients reported in the literature,four (31%) died, and postoperative septic complicationsdeveloped in seven (54%).15-21 Five (38%) septic pseudo-aneurysms of the vascular prosthesis occurred and wereadjacent to a vertebral osteomyelitis in three patients and toan aortoduodenal fistula in two patients. Infectious com-plications occurred in three patients before the diagnosis ofQ fever was made. In two patients, these complicationsdeveloped despite an adequate antibiotics treatment for 2months. Two patients with complications recovered after areoperation, received antibiotics for a mean duration of 3years, and are now considered cured. Of the five patientswithout postoperative complications, two are consideredcured after 3 years of treatment, as shown by a normaliza-tion of their serologic titers. Three patients are still receiv-ing antibiotics.

In our experience, the patient in case 3 was found to have

a slight dilatation of the aortic graft at the 2-year follow-up

that was still stable 12 years later. At 15 years after surgery, thepatient in case 1 also had moderate dilatation of the aortic graft(Fig 4), with a �40-mm-diameter enlargement of the abdom-inal aorta (Fig 5). CT scans of both patients did not showevidence of recurrent infection, and normal titers of phase IIgA antibodies suggested the chronic infection was cureddespite the aortic aneurysm.

Surgical treatment of the aneurysm and slight dilatationof the aortic graft must be associated with antibiotic ther-apy, although its optimum duration cannot be determined,because no definite criteria for a C. burnetii cure are cur-rently available. Suggestions have ranged from 1 year ofantibiotic therapy4,38,39 to indefinite administration.40 Thecurrent recommendation for the treatment of Q fever en-docarditis is at least 3 years of double coverage with doxy-cycline-ofloxacin or doxycycline-rifampin. This same anti-biotic protocol has been extrapolated and recommendedfor the treatment of vascular infection for a minimum of 18to 36 months.41 More recently, the combination of doxy-cycline with chloroquine has been proposed for 18 monthsin the treatment of Q fever endocarditis, although it has notyet been evaluated in the treatment of vascular infection.

In the literature, C. burnetii vascular infections in pa-tients without endocarditis have been treated with mono-therapy in five cases, with doxycycline-ofloxacin in threecases, and with doxycycline-rifampin in two cases. Fivepatients received antibiotic therapy for a mean duration of 3years and were considered cured as assessed by a normalclinical examination and a decline in serologic titers of IgGphase I antibodies; four patients were still receiving antibi-otic therapy.20 In our experience, the duration of theantibiotics regime was 3 years in the patient in case 1 and 4years in the other two patients who underwent surgery forinfrarenal aortic aneurysms and who presented with simul-taneous proven (case 2) or suspected (case 3) C. burnetiiendocarditis.

Duration of antibiotic therapy should be determinedfrom clinical, biologic, and serologic monitoring of phase Iantibody titers. Raoult et al6,41,42 proposed a decrease ofphase I IgG and IgA antibody titers to �200 as the mainpredictive criterion of clinical cure. High antibody phase 1titers to C. burnetii suggest a high risk of infection relapseor the development of vascular complications. This situa-tion warrants close follow-up and long-term or lifelongantibiotic treatment. However, serologic evaluation is rec-ommended monthly for the first 6 months after antibioticsare discontinued and every 3 months for 2 years beforeconsidering the patient cured.3,4,6

In conclusion, underlying cardiovascular disease, suchas a valve defect and an aortic aneurysm, is a predisposingfactor to chronic infection with C. burnetii. Diagnosisparticularly relies on the physician’s awareness of suchinfection and specialized diagnostic laboratories. Q fevershould be suspected and serologic tests systematically per-formed in patients with endocarditis and negative bloodcultures, in febrile patients with an aortic aneurysm, arte-rial, or cardiac prosthesis, and in those with prolonged or

unexplained fever, abdominal pain, or weight loss who live

JOURNAL OF VASCULAR SURGERYJuly 2005158 Sessa et al

in an area endemic for Q fever. C. burnetii can also beidentified by culture or by PCR of cardiac valve or aortictissue samples. Surgical treatment must be associated withlong-term antibiotic therapy, the duration of which shouldbe based on titers of phase I IgG antibodies.

REFERENCES

1. Davies TR, Edwards Y, Morgan A, Caul EO. Prevalence of Q fever in arural practice. J Public Health 1997;19:324-7.

2. Tissot-Dupont H, Torres S, Nezri M, Raoult D. A hyperendemic focusof Q fever related to sheep and wind. Amer J Epidemiol 1999;150:67-74.

3. Fournier PE, Marrie TJ, Raoult D. Diagnosis of Q fever. Minireview.J Clin Microbiol 1998;36:1823-34.

4. Maurin M, Raoult D. Q fever. Clin Microbiol Rev 1999;12:518-53.5. Raoult D, Tissot-Dupont H, Foucault C, Gouvernet J, Fournier PE,

Bernit E, et al. Q fever 1985-1998. Clinical and epidemiologic featuresof 1,383 infections. Medicine (Baltimore) 2000;79:109-23.

6. Raoult D, Levy PY, Harlé JR, Etienne J, Massip P, Goldstein F, et al.Chronic Q fever: diagnosis and follow-up. N Y Acad Sci 1990;590:51-60.

7. Brouqui P, Dupont HT, Drancourt M, Berland Y, Etienne J, Leport C,et al. Chronic Q fever. Ninety-two cases from France including 27without endocarditis. Arch Intern Med 1993;153:642-48.

8. Raoult D, Marri T. Q fever. Clin Infect Dis 1995;20:489-96.9. Muhlemann K, Matter L, Meyer B, Schopfer K. Isolation of Coxiella

burnetii from heart valves of patients treated for Q fever endocarditis.J Clin Microbiol 1995;33:428-31.

10. Weir WR, Bannister HG, Chambers S, DeCock K, Mistry H. Chronic Qfever associated with granulomatous hepatitis. J Infect 1984;8:56-60.

11. Raoult D, Bollini G, Gallais H. Osteoarticular infection due to Coxiellaburnetii. J Infect Dis 1989;159:1159-60.

12. Rigaud JL, Sirol J, Barabe P, Delprat J, Maistre B, Segonne J, et al.Rickettsial arteritis due to Coxiella burnetii. Arch Mal Coeur Vaiss1977;70:185-90

13. Lovey PY, Morabia A, Bleed D, Peter O, Dupuis G, Petite J. Long termvascular complications of Coxiella burnetii infection in Switzerland:cohort study. BMJ 1999;319:284-6.

14. Wildman M, Ayres JG. Cardiovascular risk factors cannot be ignored(letters). BMJ 2000;320:58.

15. Ellis ME, Smith CC, Moffat MAJ. Chronic or fatal Q fever infection: areview of 16 patients seen in North-East Scotland (1967-1980). QJ Med 1983;52:54-66.

16. Raoult D, Piquet P, Gallais H, De Micco C, Drancourt M, Casanova P.Coxiella burnetii infection of a vascular prosthesis (letter). N EnglJ Med 1986;351:1358-9.

17. Fergusson RJ, Shaw TR, Kitchin AH, Matthews MB, Inglis JM,Peutherer JF. Subclinical chronic Q fever. Q J Med 1985;57:669-76.

18. Micoud M, Brion JP, Boulard JC, Magne JL, Gratacap B, Stahl JP, et al.Infection of aortic aneurysm with Coxiella burnetii. Lancet 1991;338:584.

19. Piquet P, Raoult D, Tranier P, Mercier C. Coxiella burnetii infection ofpseudoaneurysm of an aortic bypass graft with contiguous vertebralosteomyelitis. J Vasc Surg 1994;19:165-8.

20. Fournier PE, Casalta JP, Piquet P, Tournigand P, Branchereau A,Raoult D. Coxiella burnetii infection of aneurysms or vascular grafts:report of seven cases and review. Clin Infect Dis 1998;26:116-21.

21. Mejia A, Toursarkissian B, Hagino RT, Myers JG, Sykes MT. Primaryaortoduodenal fistula and Q fever: an underrecognized association?

Ann Vasc Surg 2000;14:271-3.

22. Peacock MG, Philip RN, Williams JC, Faulkner RS. Serological evalu-ation of Q fever in humans: enhanced phase I titers of immunoglobulinsG and A are diagnostic for Q fever endocarditis. Infect Immun 1983;41:1089-98.

23. Tissot-Dupont H, Thirion X, Raoult D. Q fever serology: cutoff deter-mination for microimmunofluorescence. Clin Diagn Lab Immunol1994;1:189-96.

24. Fournier PE, Casalta JP, Habib G, Messana T, Raoult D. Modificationof the diagnostic criteria proposed by the Duke endocarditis service topermit improved diagnosis of Q fever endocarditis. Am J Med 1996;100:629-33.

25. Musso D, Raoult D. Coxiella burnetii blood cultures from acute andchronic Q fever patients. J Clin Microbiol 1995;33:3129-32.

26. Mallavia LP, Whiting LL, Minnick MF, Heinzen RA, Reschke D,Foreman M. Strategy for detection and differentiation of Coxiellaburnetii strains using the polymerase chain reaction. Ann N Y Acad Sci1990;590:572-81.

27. Chen SY, Hoover HA, Thompson A, Williams JC. Characterization ofthe origin of DNA replication of the Coxiella burnetii chromosome.Ann N Y Acad Sci 1990;590:491-503.

28. Frazier ME, Heinzen RA, Mallavia LP, Baca OG. DNA probes detect-ing Coxiella burnetii strains. Acta Virol 1992;36:83-9.

29. Stein A, Raoult D. Detection of Coxiella burnetii by DNA amplificationusing polymerase chain reaction. J Clin Microbiol 1992;30:2462-6.

30. Stein A, Raoult D. A simple method for amplification of DNA fromparaffin-embedded tissues. Nucleic Acids Res 1992;20:5237-8.

31. Sessa C, Farah I, Voirin L, Magne Jl, Brion JP, Guidicelli H. Infectedaneurysms of the infrarenal abdominal aorta: diagnostic criteria andtherapeutic strategy. Ann Vasc Surg 1997;11:453-63.

32. Moneta GL, Taylor LM, Yeager RA, Edwards LM, Nicoloff AD,McConnell DB, et al. Surgical treatment of infected aortic aneurysm.Am J Surg 1998;175:396-99.

33. Yeager RA, Porter JM. Arterial and prosthetic infection. Ann Vasc Surg1992;6:485-91.

34. Muller BT, Wegener OR, Grabitz K, Pillny M, Thomas L, SandmannW. Mycotic aneurysms of the thoracic and abdominal aorta and iliacarteries: experience with anatomic and extra-anatomic repair in 33 cases.J Vasc Surg 2001;33:106-13.

35. Harris RJ, Storm PA, Lloyd A, Arens M, Marmion BP. Long-termpersistence of Coxiella burnetii in the host after primary Q fever.Epidemiol Infect 2000;124:543-9.

36. Mortality results for randomised controlled trial of early elective surgeryor ultrasonographic surveillance for small abdominal aortic aneurysms.The UK Small Aneurysm Trial Participants. Lancet 1998;352:1649-55.

37. Boyle AJ, Jelinek MV. Rethinking the approach to abdominal aorticaneurysms. Lancet 2001;357:2140.

38. Cacoub P, Wechsler B, Chapelon C, Thibon M, Godeau P. Q feverendocarditis and treatment with the fluoroquinolones. Arch InternMed 1991;151:816-8.

39. Wilson HG, Neilson GH, Galea EG, Stafford G, O’Brien MF. Q feverendocarditis in Queensland. Circulation 1976;53:680-4.

40. Pedoe HD. Apparent recurrence of Q fever endocarditis followinghomograft replacement of aortic valve. Br Heart J. 1970;32:568-70.

41. Raoult D. Treatment of Q fever. Antimicrob Agents Chemother 1993;37:1733-6.

42. Raoult D, Houpikian P, Tissot Dupont H, Riss JM, Arditi-Dijane J,Brouqui P. Comparison of 2 regimens containing doxycycline and ofloxa-cin or hydroxychloroquine. Arch intern Med 1999;159:167-73.

Submitted Dec 30, 2004; accepted Mar 15, 2005.