Journal of Infection (2007) 54, 313e318

www.elsevierhealth.com/journals/jinf

REVIEW

Q fever
Sally J. Cutler a,*, Maha Bouzid b, Ronald R. Cutler a
a b

University of East London, School of Health and Bioscience, Romford Road, Stratford, London E15 4LZ, UK University of East Anglia, Schools of Health and Biological Sciences, Norwich NR4 7TJ, UK

Accepted 24 October 2006 Available online 4 December 2006

KEYWORDS
Coxiella burnetii; Q fever; Zoonoses; Zoonosis; Tick-borne disease

Summary An outbreak of Q fever occurred in Scotland during this summer and was reported in news headlines. Despite these newsworthy headlines, Q fever remains poorly understood. The causative organism, Coxiella burnetii, has a worldwide distribution, with the notable exception of New Zealand. Even with its ubiquitous nature, Q fever is rarely reported. We explore some of the underlying reasons for this apparent under diagnosis together with some of the diagnostic challenges posed by this obligate intracellular pathogen. The host range for this microbe spans arthropods, through to birds and a diverse range of mammals including livestock, companion animals and man. In most, infection remains sub-clinical, however, in some, infection can cause severe and life-threatening complications. Furthermore, possible long-term persistence within those infected, may result in long-term sequelae disassociated from initial risk factors or acute clinical presentation. We review current thinking on C. burnetii, and identify some of our current knowledge gaps. 2006 The British Infection Society. Published by Elsevier Ltd. All rights reserved.

Discovery to today
The name Q fever was rst coined in 1937 by Derrick.1 The causative organism was rst isolated from ticks collected at Nine Mile Creek in Montana, USA, and subsequently cultivated by Cox. Meanwhile, Burnet successfully recovered this microbe from samples taken from Derricks patients. He noted the rickettsial nature or this organism, which was subsequently named Coxiella burnetii. C. burnetii can infect a broad spectrum of susceptible hosts including domestic animals (livestock and pets),
* Corresponding author. Tel.: 44 0208 2234162. E-mail address: s.cutler@uel.ac.uk (S.J. Cutler).

wildlife and even non-mammalian species including reptiles, sh, birds and ticks. Human infection is usually via shedding of C. burnetii in milk, faeces, urine and birth products from infected ruminants. Infection is mainly acquired through inhalation of infectious aerosols, at parturition (including normal births) or at slaughter. In humans, infection is often asymptomatic, and in its mild form can be mistaken for other u-like illnesses. It can, however, be associated with chronic or even fatal consequences, usually through endocarditis. Though this organism is endemic in the environment, infection is rarely reported. Approximately 100 human cases are reported in Britain each year, but diagnosis can be problematic.2 Similar levels have been detected in other countries such as the USA where again, this is likely to be unrepresentative of the

0163-4453/$30 2006 The British Infection Society. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.jinf.2006.10.048

314 true infection levels. In the USA, Q fever became a nationally reportable disease for most States in 1999, since which a steady rise in human cases has become evident.3 Sero-epidemiological surveys are patchy, but those conducted give incidence of 18e37% of blood donors from Morocco, Tunisia, Zimbabwe, Nigeria and North Africa.4 Targeted studies of an at risk populations of farmers in the UK revealed 29% were seropositive.5 Though rarely in the public eye, several notable outbreaks have occurred recently in the UK. During 1981, 29 cases of Q fever were diagnosed among residents living adjacent to a route used by farm vehicles in the outskirts of Newport, Wales.6 Perhaps our most notorious outbreak was in April 1989 near Solihull, West Midlands, England. This outbreak resulted in 147 conrmed cases, possibility assisted by windborne spread.7 Another outbreak was centred upon cardboard box factory in Newport during the summer of 2002.8 Unlike the Solihull outbreak that was associated with livestock during the lambing/calving season, the cause of this outbreak remained elusive. More recently, in July 2006 and outbreak was reported among abattoir workers near Stirling, Scotland (ProMed archive number 20060721.2001). Eleven cases were conrmed, but this could rise. The clinical spectrum of Q fever is expanding, however, diagnosis frequently correlates with the geographical location of enthusiasts, suggesting infection elsewhere remains undetected. The highly infectious nature of this debilitating containment level 3 pathogen has led to its inclusion as a list B potential biological warfare agent by the CDC. This, however, has revitalised current research interest into this microbe.9 We look forward to the results of these efforts that may help to decipher the biological mysteries of Q fever.

S.J. Cutler et al. The obligate intracellular requirements for growth, C. burnetii resulted in its initial grouping among the Rickettsiaceae. Following DNAeDNA hybridisation studies, this was challenged with the suggestion that it had greater resemblance with the gamma division of Proteobacteria.11 This was later conrmed with whole genome sequencing.12,13 Following phylogenetic revision, C. burnetii currently sits as a monotypic species closely related to the Legionellales. What is becoming increasingly apparent is that C. burnetii isolates fall into different groups. This has been shown using various microbiological methods including plasmid proling, intragenic spacer typing,14 pulsed-eld electrophoresis restriction endonuclease digests,15 variable number tandem repeat typing16,17 and DNA hybridisations to microarrays.18 Furthermore, new Coxiella-like organisms have been detected, such as Rickettsiella grylli, an intracellular pathogen found in grasshoppers,13 and an as yet unnamed type in ticks collected from pelicans.19 In the light of these ndings the phylogenetic relationships of this group will probably have to be re-addressed in future years.

The human sentinel

Infection of livestock with C. burnetii often goes unnoticed. Indeed, sporadic abortion of up to 5% is considered normal and does not prompt investigation. Furthermore, the lack of overt clinical signs exhibited among infected livestock make diagnosis a challenge. The microscopic staining methods used to detect infection in livestock, reinforced with serology, could be signicantly improved. As a result, infection is identied when it spills into the human population through zoonotic exposure. If a cluster of human cases occur, investigation of the likely source will likely follow, thus mankind serves as a sentinel for disease activity. Most cases in man follow exposure to livestock, rather than tick exposure, which may also carry the pathogen (see Fig. 1).

The microbe
C. burnetii can exist in different morphological forms. The small cell variant (SCV) has been likened to a spore-like structure with the ability to persist amidst extreme environmental extremes.10 These enable the organism to persist on formites for over a year, withstand disinfectants and survive extremes of temperature. In contrast, the large cell variant (LCV) has evolved to exploit and persist within the acidied phagolysosomes of monocytes and macrophages. TNF is stimulated together with other host cytokines. It is postulated that a disregulation of the cytokine network with a resulting impairment of bactericidal responses may enable persistence in chronic cases. Furthermore, different phases have been reported with phase I being associated with mammalian infection, while phase II, typically of reduced virulence, associated with deletion or truncation of the lipopolysaccharide of C. burnetii. The switch from virulent to the avirulent phase II can follow propagation of these microbes. Differences have been noted with both rates of phagocytosis and intracellular proliferation between phase I and phase II variants. Confusion can arise from the initial serological response during acute infection being against phase II antigen, with antibodies against phase I correlating with more chronic disease.

Figure 1 Ticks can be infected with C. burnetii. Their significance for causing human disease has yet to be established. Figure shows a nymph and larval stage of Ixodes ricinus ticks.

Q fever

315
Table 2 Clinical spectrum of Q fever seen in a UK study Frequency in UK survey43 1117 47% 5% 2% 2% 0.3% 0.1% 50 casesb

When to suspect Q fever

A large proportion of infected cases remain asymptomatic (up to 60%), however, whether these represent those exposed to the infection, but successfully cleared the agent, or whether they remain sub-clinically infected remains a hotly debated topic. What is becoming more apparent is the ability of this organism to result in long-term sequelae, probably associated with chronic persistence. This has been supported by positive PCR assays for the organism after antimicrobial treatment and clinical recovery.20 Recently it has been suggested that C. burnetii infection could be associated with chronic fatigue syndrome.20e23 In those with clinical signs, acute disease is characterised by u-like febrile episodes accompanied by respiratory signs of atypical pneumonia or hepatitis, sometimes with clinical hepatomegaly, but rarely jaundice (see Tables 1 and 2).10,24 Acute disease manifestations appear to uctuate with geographical locations, suggesting that strain differences may inuence clinical outcome,24 or that route of inoculation may inuence disease presentation.10 Less frequently, but especially in those with underlying cardiac valvular damage, prosthetic valves, or immunocompromising conditions, endocarditis may follow in about 5% of cases. Here, Q fever endocarditis cases fall into the group of sterile blood culture endocarditis patients owing to the requirements for intracellular cultivation. Chronic disease can manifest as vascular infection, particularly in those with aortic abnormalities,25 or osteoarticular infection (including osteoarthritis or osteomyelitis), especially in those with joint prosthesis or who are immunocompromised.10,26 Chronic granulomatous hepatitis,24 chronic pulmonary disease,24 lymphadenopathy,26 meningitis or encephalitis,27e30 either in conjunction with endocarditis, or without, have all been associated with chronic infection, see Table 2. Diagnostic delays correlate with the frequency of complications such as hepatosplenonegaly.24 In pregnancy, infection has been associated with adverse outcomes, indeed in one review of 15 patients who acquired infection during pregnancy, only ve successfully gave birth and only two infants were of normal birth

Clinical manifestation Number of cases in study over 10 years Pneumoniaa Hepatitisa Pericarditis Myocarditis Meningitis Erythema nodosum Endocarditis
a b

Some cases had both hepatitis and pneumonia. Cases presented with endocarditis during 10 year study period, thus unknown denominator.

Table 1 Q fever

Frequency and Incidence of clinical categories of Infection Category Prevalence/Associated conditions 60% 38% 2% Rare 0.2e0.5% Endocarditis Vascular infection Osteoarticular infection Osteoarticular infection

Primary Asymptomatic Acute, self-limiting Acute, hospitalised Acute, during pregnancy Chronic Percentage of acute progressing Valve lesions and/or cancer Vascular aneurysm or prosthesis Cancer Paediatric patients

weight.31 These adverse outcomes are believed to occur either through placentitis or direct foetal infection.24 A striking absence of clinical signs in children of under 15 years has been noted.32 Similarly, clinical disease is more frequently correlated with males than females. This has led to the suggestion that female hormones may offer a protective effect reducing the incidence of disease among this group.33 Treatment of chronic disease is usually with a combination of doxycycline and hydroxychloroquine.34 Problems have been encountered with many therapeutic regimes with recovery of viable organisms even following 12 months of treatment,35 probably resulting from the low pH values in the phagolysosomes in which these organisms reside. Addition of hydroxychloroquine restores the bactericidal activity of doxycycline through increasing the intracellular pH, however, can be associated with photosensitivity or retinal toxicity. C. burnetii are inherently resistant to both b-lactams and aminoglycoside antimicrobials, while different isolates vary in their susceptibility to erythromycin.10 Monotherapy is associated with clinical relapse upon cessation of treatment.34 The therapeutic regime in chronic disease is long and protracted, often for at least 18 months,3 however, a therapeutic regime of three years is recommended for endocarditis monitored serologically every six months during therapy and every three months thereafter.36 Our understandings of the host-microbial factors that modulate clinical expression are still in their infancy. Evidence is growing that cell mediated immunity plays a crucial role in controlling disease expression. Whether latently infected individuals will undergo disease reactivation during later times of immunosuppression remains unresolved.

A stealthy disease in livestock and wildlife

This situation is mirrored in livestock where infection is rarely reported, however, 21% of UK dairy herds may produce evidence of exposure (Defra Zoonoses Report 2004). Infection in livestock has largely focussed on abortion or delivery of weak offspring. Goats appear to be highly susceptible to abortion as demonstrated following

316 experimental infection with a range of inocula during pregnancy, all of which resulted in abortion.37 As in the experimental study above, natural infection in sheep and goats has resulted in abortions.38,39 Although abortion is less common in cattle, infection still poses a zoonotic risk through prolonged excretion of the organism in milk. Wildlife too also frequently infected, however, few studies have addressed natural infection in wildlife species. Remarkably, these infections in animal species appear largely asymptomatic.

S.J. Cutler et al. Various PCR methods have been designed for amplication of C. burnetii, many using the multicopy insertion sequence IS1111. These assays have been demonstrated to detect between 6.5 and 10 genome equivalents.41 Again, the outer surface protein found only in acute isolates, acute disease antigen A (adaA), can be used in a molecular assay to conrm acute infection.40 Although PCR is useful for detection of early acute cases or following antigen shedding in livestock, or applied to tissue samples from focal manifestations, these assays should be recommended in addition to serology.

Diagnostic dilemmas
The major diagnostic dilemma is when to suspect a case of Q fever? The protean clinical manifestations, often initially u-like can easily be overlooked or misdiagnosed. Many early acute cases spontaneously resolve without antimicrobial intervention. Furthermore, the disease is endemic to much of the world, thus it is not inconceivable that a patient may have a serological titre totally unrelated to their current medical problems. For these reasons, detection of IgM antibodies and use of paired samples to demonstrate seroconversion is particularly important. Controversy exists over diagnostic cutoffs in various serological assays. Some have found unacceptably low sensitivity using recommended diagnostic thresholds while others have elevated cutoff thresholds to overcome specicity issues. There are also reports of poor correlation between assays using different strains of C. burnetii as antigen. No comprehensive comparative study of commercially available diagnostics for either man or animals could be found at the time of writing this review. Some advocate the lowering of diagnostic thresholds in event of an outbreak, thus enabling more potential cases to be identied. The concern with this practice is the possibility of compromising diagnostic specicity. For indirect immunouorescence assays (IFA), titres against phase II antigen in excess of 1:50 are considered diagnostic for IgM, while those above 1:200 are used for IgG for acute disease. Those with chronic disease usually have high titres against both phase I and phase II antigens, often in excess of 1:1600. IgM may still be evident in chronic cases. Several ELISA assays have been developed, for human and veterinary application. Cutoff values for these, like some commercial IFA kits, may, however, vary according to the manufacturer of the assay. A complement xation test (CFT) is still a prescribed diagnostic assay for livestock testing, however, gives inferior performance compared with IFA or ELISA formats. Further serological advances are offered through the identication of an antigen only associated with acute isolates, acute disease antigen A (adaA). This has been cloned, expressed and developed into an ELISA format, where it has retained its immunological activity, thus offering a potential assay for differentiation of acute from chronic disease.40 As an obligate intracellular pathogen, cultivation can only be achieved through use of embryonated eggs or cell culture methods. A miniturised tissue culture method known as shell vials has proved particularly useful for growth of these microbes. Resulting growth is identied by direct immunouorescence staining using polyclonal serum conjugated with FITC, Gimenez or Giemsa staining, or by PCR.

Strategies for control and intervention

In common with all zoonotic diseases, control of the disease in animals will inuence the level of disease seen in man. Interrogation of C. burnetii genomic data is being actively pursued to identify likely vaccine candidates. We are still, however, far from producing safe, effective well characterised vaccine. Occupationally at risk groups in Australia are vaccinated with a heat-inactivated phase I vaccine, however, this requires prior testing of potential recipients to avoid adverse reactions in those found to be seropositive.42

Prospects for the future

The genomes of several more isolates of C. burnetii and phylogenetically close relatives are currently undergoing annotation. It is hoped that these may provide insights into those factors essential for virulence. Greater research funding has been allocated towards this microbe encouraging detailed studies of the pathological mechanisms employed by this intriguing microbe in order to subvert and multiply within the phagolysosome and indeed persist within the host for prolonged periods. Understanding the role of both humoral and cell mediated immunity in both control of infection and modulation of clinical presentation is an essential prerequisite for devising vaccinational strategies for control of this infection in target populations.

Serodiagnostics
B

Acute infection characterised by response against phase II antigen. Chronic infection against both phase I and phase II antigens. Acute infection IgM titres against phase II of >1:50 or IgG titres of >1:200. Chronic infection with high titres against phase I or phase II (often >1:1600) with or without the presence of IgM. ELISA available for human and veterinary diagnosis. Cloned immunogenic antigens being explored. CFT remains a prescribed test for livestock, but generally considered inferior to IFA or ELISA.

Q fever

317
7. Hawker JI, Ayres JG, Blair I, Evans MR, Smith DL, Smith EG, et al. A large outbreak of Q fever in the West Midlands: windborne spread into a metropolitan area? Commun Dis Public Health 1998;1:180e7. 8. van Woerden H, Mason B, Nehaul L, Smith R, Salmon R, Healy B, et al. Q fever outbreak in industrial setting. Emerging Infect Dis 2004;10:1282e9. 9. Madariaga MG, Rezai K, Trenholme GM, Weinstein RA. Q fever: a biological weapon in your backyard. Lancet Infect Dis 2003;3: 709e21. 10. Marrie TJ. Coxiella burnetii pneumonia. Eur Respir J 2003;21: 713e9. 11. Vogel JP. Turning a tiger into a house cat: using Legionella pneumophila to study Coxiella burnetii. Trends Microbiol 2004;12:103e5. 12. Seshadri R, Paulsen IT, Eisen JA, Read TD, Nelson KE, Nelson WC, et al. Complete genome sequence of the Q-fever pathogen Coxiella burnetii. PNAS 2003;100:5455e60. 13. Seshadri R, Samuel J. Genome analysis of Coxiella burnetii species e insights into pathogenesis and evolution and implications for biodefense, in Rickettsioses: from Genome to Proteome, Pathobiology, and Rickettsiae as an International Threat; 2005. p. 442e450. 14. Glazunova O, Roux V, Freylikman O, Sekeyova Z, Fournous G, Tyczka J, et al. Coxiella burnetii genotyping. Emerging Infect Dis 2005;11:1211e7. 15. Thiele D, Willems H, Kopf G, Krauss H. Polymorphism in DNA restriction patterns of Coxiella burnetii isolates investigated by pulsed eld gel electrophoresis and image analysis. Eur J Epidemiol 1993;9:419e25. 16. Arricau-Bouvery N, Hauck Y, Bejaoui A, Frangoulidis D, Bodier C, Souriau A, et al. Molecular characterization of Coxiella burnetii isolates by infrequent restriction site-PCR and MLVA typing. BMC Microbiol 2006;6:38. 17. Svraka S, Toman R, Skultety L, Slaba K, Homan WL. Establishment of a genotyping scheme for Coxiella burnetii. FEMS Microbiol Lett 2006;254:268e74. 18. Beare PA, Samuel JE, Howe D, Virtaneva K, Porcella SF, Heinzen RA. Genetic diversity of the Q fever agent, Coxiella burnetii, assessed by microarray-based whole-genome comparisons. J Bacteriol 2006;188:2309e24. 19. Reeves WK, Loftis AD, Priestley RA, Wills W, Sanders F, Dasch GA. Molecular and biological characterization of a novel Coxiella-like agent from Carios capensis, in Rickettsioses: from Genome to Proteome, Pathobiology, and Rickettsiae as an International Threat. 2005. p.343e5. 20. Marmion BP, Storm PA, Ayres JG, Semendric L, Mathews L, Winslow W, et al. Long-term persistence of Coxiella burnetii after acute primary Q fever. Q J Med 2005;98:7e20. 21. Helbig K, Harris R, Ayres J, Dunckley H, Lloyd A, Robson J, et al. Immune response genes in the post-Q-fever fatigue syndrome, Q fever endocarditis and uncomplicated acute primary Q fever. Q J Med 2005;98:5665e74. 22. Iwakami E, Arashima Y, Kato K, Komiya T, Matsukawa Y, Ikeda T, et al. Treatment of chronic fatigue syndrome with antibiotics: pilot study assessing the involvement of Coxiella burnetii infection. Intern Med 2005;44:1258e63. 23. Wildman MJ, Smith EG, Groves J, Beattie JM, Caul EO, Ayres JG. Chronic fatigue following infection by Coxiella burnetii (Q fever): ten-year follow-up of the 1989 UK outbreak cohort. Q J Med 2002;95:527e38. 24. Maurin M, Raoult D. Q fever. Clin Microbiol Rev 1999;12:518e53. 25. 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; 26:116e21. 26. Foucault C, Lepidi H, Poujet-Abadie JF, Granel B, Roblot F, Ariga T, et al. Q fever and lymphadenopathy: report of four

Summary points
 An endemic zoonotic disease often underdiagnosed.  Infection can be acute, chronic or latent with potential reactivation. New pathological associations are expanding the clinical spectrum of Q fever.  Therapeutic management with doxycycline and hydroxychloroquine.  Vaccines not generally available.  A robust and highly infectious containment level 3 pathogen listed by the CDC as a potential biothreat agent.

Future prospects
 Improved diagnostics through use of cloned diagnostic antigens improving standardization and negating the need to cultivate large amounts of this containment level 3 intracellular pathogen. This will facilitate improved epidemiological surveillance.  Production of efcacious vaccine that would not interfere with diagnostic assays.  Comprehension of microbial virulence factors and mechanisms of pathogenesis: B Why are most cases in adult men? B What are the host susceptibility factors? B What host-microbial interactions inuence successful clearance, latency or reactivation? B Are all types of C. burnetii of equal virulence? B Do certain types correlate with clinical presentation?

Acknowledgements
We thank Med-Vet-Net for their support (www.medvetnet. org).

References
1. Derrick E. The Query Fever e the Elkington Oration. Queensland Health 1964;1:1e20. 2. Cutler S, Paiba G, Howells J, Morgan K. Q feverea forgotten disease? Lancet Infect Dis 2002;2:717e8. 3. Karakousis PC, Trucksis M, Dumler JS. Chronic Q fever in the United States. J Clin Microbiol 2006;44:2283e7. 4. Fournier P-E, Marrie TJ, Raoult D. Diagnosis of Q fever. J Clin Microbiol 1998;36:1823e34. 5. Thomas DR, Salmon RL, Coleman TJ, Morgan-Capner P, Sillis M, Caul EO, et al. Occupational exposure to animals and risk of zoonotic illness in a cohort of farmers, farmworkers, and their families in England. J Agric Safety Hyg 1999;5:373e82. 6. Salmon MM, Howells B, Glencross EJ, Evans AD, Palmer SR. Q fever in an urban area. Lancet 1982;1:1002e4.

318
new cases and review. Eur J Clin Microbiol Infect Dis 2004;23: 759e64. Bernit E, Pouget J, Janbon F, Dutronc H, Martinez P, Brouqui P, et al. Neurological involvement in acute Q fever: a report of 29 cases and review of the literature. Arch Intern Med 2002;162: 693e700. Drancourt M, Raoult D, Xeridat B, Milandre L, Nesri M, Dano P. Q fever meningoencephalitis in ve patients. Eur J Epidemiol 1991;7:134e8. Kofteridis DP, Mazokopakis EE, Tselentis Y, Gikas A. Neurological complications of acute Q fever infection. Eur J Epidemiol 2004;19:1051e4. Shaked Y, Samra Y. Q fever meningoencephalitis associated with bilateral abducens nerve paralysis, bilateral optic neuritis and abnormal cerebrospinal uid ndings. Infection 1989;17: 394e5. Raoult D, Tissot-Dupont H, Foucault C, Gouvernet J, Fournier PE, Bernit E, et al. Q fever 1985e1998. Clinical and epidemiologic features of 1,383 infections. Medicine (Baltimore) 2000;79: 109e23. Raoult D, Marrie T, Mege J. Natural history and pathophysiology of Q fever. Lancet Infect Dis 2005;5:219e26. Leone M, Honstettre A, Lepidi H, Capo C, Bayard F, Raoult D, et al. Effect of sex on Coxiella burnetii infection: protective role of 17 beta-estradiol. J Infect Dis 2004;189: 339e45. Raoult D, Houpikian P, Dupont HT, Riss JM, Arditi-Djiane J, Brouqui P. Treatment of Q fever endocarditis: comparison of 2 regimens containing doxycycline and ooxacin or hydroxychloroquine. Arch Intern Med 1999;159:167e73.

S.J. Cutler et al.

35. 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;33:428e31. 36. Elliott TSJ, Foweraker J, Gould FK, Perry JD, Sandoe JAT. Guidelines for the antibiotic treatment of endocarditis in adults: report of the Working Party of the British Society for Antimicrobial Chemotherapy. J Antimicrob Chemother 2004;54:971e81. 37. Arricau BN, Souriau A, Lechopier P, Rodolakis A. Experimental Coxiella burnetii infection in pregnant goats: excretion routes. Vet Res 2003;34:423e33. 38. Berri M, Crochet D, Santiago S, Rodolakis A. Spread of Coxiella burnetii infection in a ock of sheep after an episode of Q fever. Vet Rec 2005;157:737e40. 39. Masala G, Porcu R, Sanna G, Chessa G, Cillara G, Chisu V, et al. Occurrence, distribution, and role in abortion of Coxiella burnetii in sheep and goats in Sardinia, Italy. Vet Microbiol 2004; 99:301e5. 40. Zhang G, To H, Russell KE, Hendrix LR, Yamaguchi T, Fukushi H, et al. Identication and characterization of an immunodominant 28-kDa Coxiella burnetii outer membrane protein specic to isolates associated with acute disease. Infect Immun 2005; 73:1561e7. 41. Klee S, Tyczka J, Ellerbrok H, Franz T, Linke S, Baljer G, et al. Highly sensitive real-time PCR for specic detection and quantication of Coxiella burnetii. BMC Microbiol 2006;6:2. 42. Zhang G, Samuel JE. Vaccines against Coxiella infection. Expert Rev Vaccines 2004;3:577e84. 43. Pebody RG, Wall PG, Ryan MJ, Fairley C. Epidemiological features of Coxiella burnetii infection in England and Wales: 1984e1994. Commun Dis Rep CDR Rev 1996;6:R128e32.

27.

28.

29.

30.

31.

32. 33.

34.