What is a Rickettsia

ESCMID PGEC Nice, 28-31 mars 2017

IS THERE A CONSENSUS ABOUT PATHOGEN

DETECTION TARGETS AND METHODS IN CULTURE-NEGATIVE ENDOCARDITIS?

Pierre-Edouard Fournier

Unité de Recherche sur les Maladies Infectieuses et Tropicales Emergentes Institut Hospitalo-Universitaire Méditerranée-Infection ESCMID Online Lectu

re Library

© by author

Of course not!

ESCMID Online Lecture Library

© by author

Infective endocarditis

Stable incidence worldwide

15 to 60 cases/million inhabitants/year in the USA and Europe.

0.16 to 5.4 per 1,000 hospital admissions

Heart surgery required: 25 to 50% of cases

Mortality: 20 to 26% during initial hospitalization, 30% overall Duval et al. J. Am. Col. Cardiol. Dis. 2012;59:1968-76

Sandre and Shafran. Clin Infect Dis. 1996:22:276-86; Martin et al. Clin Infect Dis 1997;24:669-75

Bashore et al. Curr Probl Cardiol. 2006:31:274-352

Jault et al. Ann Thorac Surg. 1997:63:1737-41; Larbalestier et al. Circulation. 1992;86:1168-74; Murdoch et al. Arch. Intern. Med. 2009;169:463-473 ESCMID Online Lectu

re Library

© by author

Evolving trend of infective endocarditis

Increase in coagulase-negative staphylococci over 50 years (intracardiac devices, prosthetic valves, hemodialysis)

Increase in Staphylococcus aureus (North America, IVDA) and Enterococcus spp. in the past decade

Increase in patient age and male/female ratio

Decrease in Streptococcus viridans and BCNE (1980s 23% 2000s 14%)

(Slipczuk et al. PLoS One. 2013;8:e82665; Dayer et al. Lancet. 2015;385:1219-28; Thanavaro & Nixon. Heart & Lung. 2014;43:334-7)


© by author

Heterogenous incidence of BCNE

North-to-South gradient

France 9%, UK 13%, Spain 14%, USA 14%, Japan 20%, Sweden 24%, Italy 25%, Germany 33%

Brazil 23%, India 31%, Pakistan 48%, Turkey 50%, Tunisia 54%, South Africa 55%, Algeria 56%, Morocco 58%, Lao PDR 61%, Thailand 69%, Egypt 69.7% Differences in the distribution of causative agents (zoonoses), differences in antibiotic use or study design (microbiological techniques used or studied populations)

Bennis A. et al. Ann Cardiol Angeiol (Paris) 1995;44:339-44. Cecchi E. et al. Ital Heart J 2004;5:249 56. Benslimani A. et al. Emerg Infect Dis 2005;11:216-24. Cetinkaya Y. et al. Int J Antimicrob Agents 2001;18:1-7. Ferrera C. et al. Rev Esp Cardiol 2012;65:891-900. Garg N. et al. Int J Cardiol 2005;98:253-60. Koegelenberg C.F. et al. QJM 2003;96:217-25. Lamas C.C. et al. Heart 2003;89:258-62. Letaief A et al. Int J Infect Dis 2007;11:430-3. Mirabel M. et al. Int J Cardiol 2015;180:270-3. Nakatani S. et al. Circ J 2003;67:901-5. Selton-Suty C. et al. Clin Infect Dis 2012;54:1230-9. Siciliano R.F. et al. Int J Infect Dis 2014;25:191-5. Tariq M. et al. Int J Infect Dis 2004;8:163-70. Watt G. et al. Am J Trop Med Hyg 2015;epub. Werner M. et al. Scand J Infect Dis 2008;40:279-85.


© by author

Main etiologies of BCNE

Empirical administration of antibiotics prior to blood cultures (50 – 70%) (Munoz et al. J. Clin. Microbiol. 2008:46:2897-901; Katsouli & Massad. Ann Thorac Surg. 2013;95:1467-74)

Fastidious microorganisms (5 – 30%, ~5% of IE)

Requiring specific media and/or prolonged incubation: Brucella spp., defective streptococci (Abiotrophia spp., Gemella spp., Granulicatella spp.), anaerobes (Finegoldia magna), HACEK bacteria, Legionella spp., Listeria spp., mycobacteria, Mycoplasma spp., Propionibacterium acnes, fungi (Aspergillus spp., Candida spp.)

Strictly (Coxiella burnetii, Tropheryma whipplei) or facultative (Bartonella spp.) intracellular bacteria


© by author

Variable role of zoonoses in BCNE

0 in South Africa (Koegelenberg et al. QJM 2003;96:217-25)

6.7% in the Lao PDR (two cases of B. henselae IE) (Mirabel et al. Int J Cardiol

2015;180:270-3)

9% in Turkey (Brucella sp. only but neither Q fever nor Bartonella sp. were investigated) (Cetinkaya et al. Int J Antimicrob Agents 2001;18:1-7)

10.3% in Brazil (2 Bartonella and 1 C. burnetii IE) (Lamas et al. Int J Infect Dis 2013

17:e65-e66)

11.9% in Egypt (Q fever, Bartonella sp. and Brucella sp.) (El-Kholy et al. Infection

2015;epub)

12.5% in Italy (3 cases of brucellosis) (Cecchi et al. Ital Heart J 2004;5:249-56)

13% in southern France (Q fever and Bartonella sp. but no Brucella sp.) (Fournier et al. Clin Infect Dis 2010;51:131-40)

17% in Thailand (Q fever, Bartonella sp., Streptococcus suis, Erysipelothrix rusiopathiae, Campylobacter fetus) (Watt et al. Am J Trop Med Hyg

2015;epub)

20% in the UK (mainly Q fever and Bartonella sp. but broad range PCR from valves was not performed) (Lamas et al. Heart 2003;89:258-62)


© by author

Non-infective etiologies of BCNE

Nonbacterial thrombotic endocarditis

Systemic lupus erythematosus (Libman-Sacks endocarditis)

Neoplasia (marantic endocarditis)

Rheumatoid arthritis

Behçet’s disease

Eosinophilic myocarditis & myocardial fibrosis (Loeffler’s endocarditis)

Allergy to pork ESCMID Online Lecture Library

© by author

Outcome of BCNE

Negative culture => etiological diagnosis delayed => increased risk of valve destruction, septic emboli and death

(Hoen et al. Clin Infect Dis. 1995; 20:501-6; Katsouli & Massad. Ann Thorac Surg. 2013;95:1467-74; Murashita et al. Eur J Cardiothorac Surg. 2005;26:1104-11; Zamorano et al. Am J Cardiol. 2001;87:1423-5)

But:

106 BCNE vs 643 BCPE (1996-2011): no statistical difference in diagnostic delay, surgery and mortality

(Ferrera et al. Rev Esp Cardiol. 2012;65:891-900) ESCMID Online Lecture Library

© by author

Is the empirical treatment of IE sufficient for BCNE?

Current guidelines = intravenous -lactam + aminoglycoside (Baddour et al. Circulation; 111:e394-434; Habib et al. Eur Heart J 2009;30:2369-413; Que and Moreillon. Nature Rev. Cardiol. 2011;8:322-36)

May not treat up to 20% of patients: fastidious bacteria (Q fever, Brucella spp., Legionella sp., Mycoplasma sp., Tropheryma whipplei), fungi A precise microbiological diagnosis is mandatory to optimize therapy


© by author

Diagnosis of infective endocarditis

A precise microbiological diagnosis mandatory to guide therapy

Culture long considered the most important diagnostic tool

Highlighted by the weight given to culture in the Duke criteria

Li et al. Clin. Infect. Dis. 2000:30:633-8


© by author

The diagnosis of BCNE: a challenge


© by author

Patient interview


© by author

Homeless, alcoholic and/or patients coming from Maghreb => B. quintana, contact with kittens => B. henselae

Patients > 50 y-o with chronic arthralgias => Tropheryma whipplei

Patients > 40 y-o with bicuspid aortic valve, contact with parturient farm animals => Coxiella burnetii

Patients coming from South America and Turkey, contact with farm animals => Brucella spp.

Young women with a history of thrombosis and/or fetal loss => systemic lupus erythematosus

Older women with arthralgias => rheumatoid arthritis

Patients > 40 y-o with embolic phenomena => marantic endocarditis

Patients with a relapsing BCNE and a porcine bioprosthesis => allergy to pork

Epidemio-clinical clues


© by author

Development of another Diagnostic score for IE

Use of a combination of aspecific clinical symptoms and biological results

Criteria independently associated to IE

Richet et al. J. Antimicrob. Chemother. 2008;62:1434-40


© by author

Blood testing


© by author

Serology

A single serum may identify the causative agent in up to 50% of culture-negative cases

Bartonella sp.: IFA: IgG > 1:800

(sensitivity 89.5%, specificity 99.6%) (Fournier et al. Clin Diagn Lab Immunol. 2002:9:795-801)

Coxiella burnetii : IFA: IgG to phase 1 > 1:800 (Se 100%, Sp 99.5%) Major Duke criterion (Rolain et al. Clin Diagn Lab Immunol. 2003:10:1147-8; Li et al. Clin Infect Dis 2000:30:633-8)

Does an IFA profile of acute QF rule out the diagnosis of endocarditis?


© by author

Endocarditis in acute Q fever ?

• 2012, 45-y-o male, no history of valvular disease • Abrupt fever and elevated transaminases • Discovery of a 10-mm aortic vegetation

(Million M. et al. Clin. Infect. Dis. 2016;62:537-44)


© by author

Endocarditis in acute Q fever

• Anticardiolipin IgG 742 GPLU (N < 20) => diagnosis

of Libman-Sacks endocarditis

• Serology => Acute Q fever (IgG2 1:200, IgM2 1:200)

• Doxycycline + hydroxychloroquine for 12 months

• Normalization of anticardiolipin Abs and TEE

• Asymptomatic on follow-up (27 months)

• Is endocarditis only a late complication of Q fever?

(Million M. et al. 2015, submitted)


© by author

Bartonella endocarditis Western blot

Cross reactions among Bartonella sp. and with Coxiella burnetii or Chlamydia => cross adsorption + WB

Sensitivity 100%, specificity 95%

(Edouard et al. J Clin Microbiol. 2015;53:824-9; Houpikian & Raoult. Clin Diagn Lab Immunol. 2003;10:95-102)

Non adsorbed Non adsorbed Adsorbed Bq Adsorbed B Adsorbed Bh Adsorbed B

1 2 1 2

1. B. quintana 1. B. quintana 2. B. henselae 2. B. henselae

1 2


© by author

Serology

Brucella sp.: IFA: Ig > 1:160

Legionella pneumophila: IFA: Ig > 1:256

Mycoplasma pneumoniae: enzyme immunoassay

Aspergillus sp.: ELISA


© by author

Antigen detection for Candida endocarditis

18 patients with proven Candida endocarditis tested for serum mannan (Platelia Candida Ag Plus [Bio-Rad, France], anti-mannan antibodies (Platelia Candida Ab Plus (Bio-Rad] and (1,3)-β-d-glucans (Fungitell assay [Associates of Cape Cod, MA])

Sensitivity 100%

(Lefort A. et al. Clin Microbiol Infect. 2012;18:E99-E109) ESCMID Online Lectu

re Library

© by author

Should we perform all serology assays in BCNE?

Microorganism Present study

(n = 676)

France 2005

(n = 348)

Medicine

2005;

84:162-73

France

(n = 88)

Clin Infect Dis

1995;

20:501-6

Great Britain

(n = 63)

Heart 2003;

89:258-62

Algeria

(n = 62)

Emerg Infect

Dis 2005;

11:216-24

C. burnetii 33.9 48 7.9 12.7 3.2

Bartonella sp. 12.7 28.4 0 9.5 22.6

Streptococcus sp. 4.6 0 1.1 6.3 3.2

Staphylococcus sp. 1.9 0 3.4 11.1 6.4

T. whipplei 1.8 0.3 0 0 0

Corynebacterium sp. 0.6 0 1.1 0 1.6

Enterobacteriaceae 0.6 0 0 0 0

HACEK bacteria 0.4 0 0 0 3.2

Brucella melitensis 0 0 0 0 1.6

Chlamydia sp. 0 0 2.2 0 0

Other bacteria 3.1 1.1 1.1 1.6 1.6

Fungi 1.2 0 0 6.3 1.6

No aetiology 36.5 22.1 82.9 50.8 54.8

Marseille study

(n = 819)

Clin Infect Dis

2010;

51:131-40

The serology panel should be adapted to local epidemiology (Bartonella endocarditis <1% in Scandinavia to > 10% in North Africa)

(Naber and Erbell. Int J Antimicrob Agents. 2007:30S:S32-6; Brouqui et al. FEMS Immunol Med Microbiol. 2006:47:1-13) ESCMID Online Lecture Library

© by author

PCR from blood

Broad range assays for bacteria (16S rRNA) or fungi

(18S rRNA) => detection and identification

Low sensitivity and specificity may be increased with pre-PCR decontamination by enzymatic digestion (Rothman et al. J Infect Dis. 2002;186:1677-81)

Multiplexed RT-PCR: LightCycler® SeptiFast (Roche): detects 19 bacterial and 6 fungal species

Less sensitive than blood culture (11/50 vs 19/50) (Casalta et al. Eur J Clin Infect Dis. 2009:28:569-573)

But may be useful in patients who have taken early antibiotics


© by author

Cardiac valve testing


© by author

Should cardiac valves be cultured? (Munoz et al. J. Clin. Microbiol. 2008:46:2897-901; Voldstedlund et al. APMIS. 2008;116:190-8)

Valve culture: sensitivity 13 - 32%, specificity 72 – 98% (Munoz et al. J. Clin. Microbiol. 2008:46:2897-901; Marin et al. Medicine. 2007;86:195-202; Boussier et al. Diagn Microbiol Infect Dis. 2013;75:240-4)

CIEDs: sonication improves culture sensitivity (Rohacek et al. Pace. 2015;38:247-53)

ESCMID Online Lectu

re Library

© by author

PCR from valvular biopsies

Highest diagnostic yield, fast

Widely used

Broad range assays (16S rRNA, 18S rRNA) +/- confirmed by

specific assays

Sensitivity 41 – 96 %, lower for paraffin-embedded biopsies

Specificity 91 - 100% (Millar et al. Scand J Infect Dis. 2001;33:673-80; Bosshard et al. Clin Infect Dis. 2003;37:167-72; Breitkopf et al. Circulation. 2003;111:1415-21; Greub et al. Am J Med. 2005;118-230-8; Fournier et al. Clin Infect Dis. 2010;51:131-40; Harris et al. Eur J Clin Microbiol Infect Dis. 2014;33:2061-6; Marin et al. Medicine. 2007;86:195-202)


© by author

Multiplexed PCR from valvular biopsies

LightCycler® SeptiFast (Roche): Se 95%, Sp 100% (Fernandez et al. Rev Esp Cardiol. 2010;63:1205-8; Leli et al. Diagn Microbiol Infect Dis. 2014;79:98-101)

SepsiTest ® (Molzym): 10 BCNE, 6 diagnoses but 3 false + (Haag et al. Diagn Microbiol Infect Dis. 2013;76:413-8)

Plex-ID ® (Abbott): PCR-electrospray ionization-MS => 66 positive in 83 paraffin-embedded valves

(Brinkman et al. J Clin Microbiol. 2013;51:2040-6)


© by author

DNA may persist in valvular tissues

After completion of antibiotic therapy

5 months to 7 years (streptococci, Bartonella sp.)

No histological lesion but past history of IE positive for the same bacterium

(Branger and Raoult. J Clin Microbiol. 2003;41:4435-7; Lang et al. Clin Microbiol Infect. 2004;10:579-81; Rovery et al. J Clin Microbiol. 2005;43:163-7)

Pitfalls of PCR


© by author

Pitfalls of PCR

Many home-made assays => lack of standardization

False positive may occur specifically using broad range

assays

Negative controls are critical

Significance of identified agents should be evaluated in

the light of epidemio-clinical data

Identication of an unusual microorganisms = > confirm

by using a second gene target


© by author

Application of NGS metagenomics to 4 patients with NVE

~300,000 reads per sample (>99% human sequences)

In two, metagenomics confirmed blood cultures (E. faecalis and S. mutans)

In two (BCNE with sterile valve culture), identification of S. sanguinis and A. defectiva (Imai et al. Int J Cardiol. 2014;172:e288-9, Fukui et al. J. Infect. Chemother. 2015;21:882-4)

Are the extra-cost (PCR x 20) and time needed worth it?

Metagenomics for the diagnosis of IE?


© by author

Histopathological examination

Major Duke criterion (Durack et al. Am J Med. 1994;96:200-22; Li et al. Clin Infect Dis. 2000;30:633-8)

Gold standard for diagnosis of IE (Castonguay et al. Cardiovasc Pathol. 2013;22:19-27; Habib et al. Eur Heart J. 2005;30:2369-413; Lepidi et

al. Infect Dis Clin North Am. 2002;16:339-61; Morris et al. Clin Infect Dis. 2003;36:697-704)

Sensitivity for native valves 73%, for prostheses 42%

Specificity 100% (Greub et al. Am J Med. 2005;118:230-8)

Crucial for BCNE

But sampling-dependent


© by author

Fluorescence in situ hybridization

Combines histopathology and molecular methods

Screening with a probe panel (pan bacteria, streptococci, enterococci, Granulicatella, B. quintana, T. whipplei)

Detection of a pathogen in 5/13 BCNE (38.5%)

Streptococci, B. quintana, T. whipplei

Valuable but requires specific probes and trained personnel

(Mallmann et al. Clin Microbiol Infect 2010;16:767-73) ESCMID Online Lectu

re Library

© by author

Immunohistochemistry

Specific mono- or polyclonal antibodies

Immunoperoxidase stain (Brouqui et al. Am J Med. 1994;97:451-8)

Capture-ELISA (Thiele et al. Eur J Epidemiol. 1992;8:568-74)

Immunofluorescence (Muhlemann et al. J Clin Microbiol. 1995;33:428-31; McCaul and Williams. Ann NY Acad Sci. 1990;590:136-47)

Q fever Bartonella sp. T. whipplei


© by author

Auto-immunohistochemistry

Patients’ own antibodies

Useful when no agent is identified by other methods

Sensitivity 80% in streptococcal IE, 100% in T. whipplei endocarditis

(Lepidi et al. J Infect Dis. 2006;193:1711-7)

T. whipplei endocarditis Streptococcal endocarditis


© by author

18FDG-PET/CT in BCNE

56-year-old man Fatigue and weight loss (–6

kgs) over 6 months Aortic bioprosthesis (7 years) TEE => thickened and partial

aortic stenosis but no vegetation

PET/CT => aortic periprosthetic FDG uptake

Serology (IFA and WB) and PCR from EDTA blood positive for B. henselae

Recovery using doxycycline – gentamicin

(Gouriet F et al. Emerg Infect Dis. 2014;20:1396-1397)


© by author

http://www.ncbi.nlm.nih.gov/pmc/articles/PMC4111175/figure/F1/

18FDG-PET/CT in BCNE

77-year-old woman Right hemiplegia and aphasia Recent spleen and right kidney

ischemic episodes Aortic bioprosthesis (4 years) TEE => diffuse aortic thickening

+ bioprosthesis stenosis and insuficiency

PET/CT => FDG uptake around the metal ring of the aortic graft

Valve replacement => detection of T. whipplei by PCR and IHC

Treatment with doxycycline and OH-chloroquine

(Jos SL et al. BMC Res Notes. 2015;8:56)


© by author

4-Dimensional-Flow MRI in BCNE

42-year-old man Headaches and word-finding

difficulties (left temporal lobe ischemic infarction)

Pandiastolic murmur TTE => aortic insufficiency TEE => bicuspid aortic valve +

suspicion of leaflet perforation MRI => regurgitation through

leaflet perforation + dilated ascending aorta

Positive Q fever serology Doxycycline + OH-chloroquine Aortic root and valve replacement (Thadani SR et al. Texas Heart Inst J. 2014;41:351-2)


© by author

Questionnaire

Bag 1: H0 1 pair of aero-anaerobic blood cultures 1 whole blood tube => serology Bartonella sp., Q fever, C. psittacii, L. pneumophila, Brucella sp., Aspergillus sp.

=> RF, antinuclear Abs, antiphospholipid

Abs, anti-pork IgE

1 whole blood tube => ACE, Ca15-3, Ca12-5, -FP

1 heparinized blood tube => cell culture 1 EDTA blood tube => PCR

Bags 2 (H2) and 3 (H4)

1 aerobic blood culture

Use of a diagnostic kit


© by author

Our diagnostic strategy 2017

Negative blood cultures

Rheumatoid factor Antiphospholipid

antibodies

Antinuclear antibodies

Dedicated RT-PCR for Bartonella spp. and

Tropheryma whipplei from EDTA blood

Q fever and Bartonella serologies

Dedicated RT-PCR for Streptococcus oralis and gallolyticus groups,

Enterococcus sp., Staphylococcus aureus, Mycoplasma hominis

Other serologies (Brucella melitensis,

Legionella pneumoniae, Mycoplasma pneumoniae,

western blot for Bartonella spp.)

Valvular biopsies (when available)

Anti-pork antibodies in patients with

porcine bioprosthesis

16S rRNA PCR for bacteria, ITS PCR for

fungi

Histological examination

Auto-immunohistochemistry

Dedicated PCR for Streptococcus oralis and gallolyticus groups,

Enterococcus spp., Staphylococcus aureus, Mycoplasma hominis, Bartonella spp., Tropheryma

whipplei

If culture is negative

If negative

If negative

If negative

If negative

ACE, CA19-9, CA15-3, CA12-5,

Α-FP


© by author

Diagnosis of endocarditis excluded In 68 patients, including:

- 1myxoma of the left atrium - 1 angiosarcoma

1,334 cases of BCNE




© by author

1,268 patients with endocarditis



1,334 cases of BCNE

795 patients

-

- Fastidious bacteria : 25 (19 T. whipplei) - Fungi: 15 - Usual bacteria : 186

-

- - -

-

- - -

With an identified aetiological agent - Zoonotic agents: 569 (420 Q fever,

149 Bartonella sp.)

- -




© by author

327 patients with possible endocarditis

126 patients with definite endocarditis

453 patients without identified aetiology



1,334 cases of BCNE


795 patients

-


-

- - -

-

- - -


149 Bartonella sp.)

- -




© by author

20 patients with

- 7 marantic - 9 - 2 - 1 Behcet disease

- - - 2 - 1

- - 9 - 2 - 1

noninfective endocarditis) - - systemic lupus erymathosus - 2 rheumatoid athritis - 1



1,334 cases of BCNE


795 patients

-


-

- - -

-

- - -


149 Bartonella sp.)

- -

453 patients without identified aetiology

327 patients with possible endocarditis

126 patients with definite endocarditis

- 1 allergy to pork - 1 - 1 - 1




© by author

Serology => 74.8% of diagnostics

PCR => 22.9% additional diagnoses

Blood: only 13.6% positive

Valves: 69.1% positive

Culture => no additional diagnosis

Auto-immunohistochemistry and differential amplification should be reserved to negative and recurrent cases

Diagnostic yield of the various methods used


© by author

Lessons from the Marseille study

No Chlamydia endocarditis (serological cross-reactions with Bartonella sp.)

No viral endocarditis (BUT: Coxsackie B2 endocarditis on an atrio-ventricular patch)

Major role of fastidious microorganisms (C. burnetii, Bartonella sp., T. whipplei)

Non-infectious aetiologies

Diagnostic strategies should be adapted to local epidemiology, notably for zoonoses (IFA)

Blumental et al. Clin Infect Dis. 2011; 52:710-6


© by author

Improving the diagnosis of BCNE Use of standardized sampling

Broad range +/- pathogen-specific PCR from valves +/- blood

Should valve culture still be used?

Importance of histopathological analysis

PET-CT and MRI may help confirm the diagnosis

Keep an open mind for new and noninfective etiologies!


© by author

What is a Rickettsia

Documents

Transcript of What is a Rickettsia