Blood Culture-Negative Endocarditis: Historical and Future Perspectives

1
Blood Culture-Negative Endocarditis Historical
and Future Perspectives

• Tracy Lemonovich, MD
• Clinical Instructor, Department of Medicine
• UH Case Medical Center
• Jan 18, 2011

2
Learning Objectives

• Discuss the historical perspectives of
  endocarditis
• Describe the current epidemiology of infective
  endocarditis
• Discuss advances in diagnostic strategy for blood
  culture-negative endocarditis

3
Source Levy. Br Med J. 1986293.
4
a loud systolic murmur over the precordium, a
history of prolonged low grade fever, a palpable
spleen, characteristic petechiae on the
conjunctiva and skin, and slight clubbing of the
fingers. Libman telephoned me to bring the
paraphernalia and culture media required for a
blood culture George Baehr
Levy. Br Med J. 1986293.
5
Source CDC
6
History of Infective Endocarditis (IE)

• 1542 Jean Fernel, first published report of
  endocarditis
• 1542-1800s Anatomical observations-
  abnormalities of endocardium/valves at autopsy
• Late 1800s/early 1900s William Osler and Thomas
  Horder elucidate pathophysiology clinical
  diagnostic criteria

Millar. EID. 200410.
7
History of IE Diagnostics

• 1800s Auscultation for detection of cardiac
  murmurs
• 1830-40s Elevated body temperature important
• 1870s Microscopic visualization of bacteria in
  vegetations
• 1880s Birth of bacteriology routine use of
  blood cultures

Millar. EID. 200410.
8
History of IE Diagnostics

• 1976 Use of transthoracic echocardiogram (TTE)
  in diagnosis
• 1988 Superior sensitivity of transesophageal
  echocardiogram (TEE) over TTE
• Late 1990s-present Use of molecular diagnostics

Millar. EID. 200410.
9
Current Diagnostic Criteria modified Duke
Criteria
Li. CID. 200030.
10
Modified Duke Criteria

• Major clinical criteria
• Blood culture findings positive for IE
• Typical organism from 2 separate blood cultures
  or persistently positive cultures
• Single blood culture for Coxiella burnetii or
  phase I IgG gt1800
• Evidence of endocardial involvement
• Oscillating intracardiac mass on valve or
  supporting structures, in path of regurgitant
  jets, or on implanted material
• Abscess
• New dehiscence of prosthetic valve

Li. CID. 200030.
11
Modified Duke Criteria

• Minor clinical criteria
• Predisposition predisposing heart condition or
  intravenous drug use
• Fever, temp gt38ºC
• Vascular phenomena, emboli, mycotic aneurysm,
  intracranial/conjunctival hemorrhage, Janeway
  lesions
• Immunologic phenomena glomerulonephritis, Osler
  nodes, Roth spots, RF
• Microbiologic evidence not meeting major criteria

Li. CID. 200030.
12
Sources Mandell. PPID 2009. Goldman Cecil
Medicine, 23rd ed 2007.
13
Clinical and Laboratory Findings of 2781 Patients
with Definite Endocarditis
Murdoch. Arch Intern Med. 2009169.
14
What is the Burden of Disease?

• Usual incidence 2-7 cases/100,000 person-years
• Diagnostic criteria and reporting variable
• Only 20 of clinically diagnosed cases definite
  IE
• 10-20,000 new cases/year in US
• No significant change in overall incidence last
  30 years

15
What is the Burden of Disease?

• In-hospital mortality of IE 15-20
• One year mortality approaching 40
• Despite advances in diagnostics and therapy, no
  change in mortality last 25 years

Murdoch. Arch Intern Med. 2009169.
16
Changing Epidemiology of Endocarditis
Patient risk factors
Microbiology
Medical advancements
17
Patient Risk Factors

• Aging population
• Underlying valvular disease
• Shift from rheumatic heart disease to
  degenerative heart disease
• Congenital heart disease
• Mitral valve prolapse
• Prior IE
• IV drug use (IVDU)

18
Medical Advancements and Risk of IEHealth-Care
Associated IE

• IE attributed to health-care related exposure in
  25 of patients
• Medical advancements
• Prosthetic valves
• Implantable intracardiac devices
• Indwelling vascular catheters
• Increasing use of invasive procedures
• ?Advances in immune suppressive therapies

Murdoch. Arch Intern Med. 2009169.
19
Health-Care Associated IE
Distribution of location of acquisition of 2781
patients with definite endocarditis
Murdoch. Arch Intern Med. 2009169.
20
Changing Microbiology of IE

• Staphylococcus aureus now the most common cause
  worldwide, 31 of patients
• Other gram positives important
• Viridans strep, coagulase-negative staph,
  Enterococcus
• Fastidious organisms
• HACEK 2 (0.3 in N. America)
• Haemophilus, Aggregatibacter, Cardiobacterium,
  Eikenella, Kingella
• Fungi/yeast 2
• Culture negative 10

Murdoch. Arch Intern Med. 2009169.
21
Microbiologic etiology of IE
Murdoch. Arch Intern Med. 2009169.
22
Blood Culture-Negative Endocarditis (BCNE)

• Endocarditis in which no causative organism can
  be grown in blood sample using usual lab methods
• Accounts for 2.5-31 of all IE, depending on
  case series

23
Factors Contributing to Sterility of Blood
Cultures

• Antibiotic administration preceding blood
  cultures
• Right-sided endocarditis
• Fastidious slow-growing bacteria
• Non-bacterial organisms i.e. fungi
• Non-infective endocarditis or incorrect diagnosis

24
Most Common Identified Microbiologic causes of
BCNE

• Coxiella burnetii 3-48
• Bartonella species 10-28
• Staphylococcus species 2-11
• Streptococcus species 1-6
• HACEK 0.5-3
• Fungi 1-6
• Candida, Aspergillus, Cryptococcus, endemic
  fungi, others
• Tropheryma whipplei 0.3-3
• Others Legionella, Chlamydia, Brucella

Fournier. CID. 201096.
25
Microbiologic Causes of BCNE

• Zoonotic agents important (Coxiella, Bartonella,
  Brucella)
• Limited data from US, North America
• Geographic epidemiology of zoonoses, fungi
• Antibiotics prior to blood cultures often a
  contributor (50 in most recent series)

26
Coxiella burnetii (Q fever)

• Febrile illness that occurs worldwide
• Animal reservoir cattle, sheep, goats
• Humans infected by inhalation contaminated
  aerosols
• Average 50 cases/year Q fever in US
• Endocarditis main manifestation of chronic Q fever

27
Coxiella burnetii Endocarditis

• Rarely reported in US, but likely under diagnosed
• Usually diagnosed by serologic assay, phase I IgG
  titer gt1800
• Major Duke diagnostic criteria
• PCR testing and immunohistochemistry of valve
  also have been used

28
Coxiella burnetii Endocarditis
Sources Fournier. CID. 201096. Mandell. PPID
2009.
29
Bartonella species

• Endocarditis linked to B. henselae and B.
  quintana
• Both species globally endemic
• B. henselae transmission to humans via cats
• Etiology of cat scratch disease
• B. quintana cause of trench fever
• Vector human body louse

30
Bartonella Endocarditis

• B. quintana associated with alcoholism,
  homelessness
• Significant proportion afebrile, advanced
  valvular disease, embolic phenomenon
• Diagnosed with culture, serologic assay IgG
  gt1800, PCR testing, or histology/immunohistochemi
  stry of valve

Raoult. Arch Int Med. 2003163.
31
Bartonella Endocarditis
Source Fournier. CID. 201051
32
Fungal Endocarditis and BCNE

• Candida most common cause of fungal IE 70
• Most have positive blood cultures
• Often related to central venous catheters,
  cardiac surgery, chemotherapy, IVDU
• Non-candidal fungal IE unlikely to be blood
  culture positive
• Fungal IE common cause of prosthetic valve BCNE-
  16 in recent series

Thuny. Heart. 201096.
33
Fungal Endocarditis and BCNE

• Aspergillus 2nd most common cause of fungal IE
  after cardiac surgery
• Immune suppression important risk factor
• Mortality high (80 in one series)
• Diagnosis by valve tissue staining/culture, PCR,
  serology, ?galactomannan antigen

Gumbo. Medicine. 200079.
34
Fungal Endocarditis and BCNE

• Other important fungi
• Cryptococcus
• Rare cause of IE, blood cultures often positive,
  serum antigen, valve tissue staining/culture
• Endemic fungi Histoplasma, Coccidioides
• Rare, but likely underestimated
• Diagnosis by valve tissue staining/culture,
  urine/serum antigen, PCR testing, serology
  supportive
• Others Saccharomyces, Cladosporium, others

35
Histoplasma capsulatum Endocarditis
Source Jinno S. J Clin Microbiology. 201048.
36
Tropheryma whipplei

• May be more frequent cause of BCNE than
  previously thought- 2.6 in recent series
• May be only manifestation of Whipples disease
• Improved diagnostics
• PAS staining of valve tissue, PCR testing,
  immunohistochemistry

Fournier. CID. 201096.
37
T. whipplei Endocarditis
Sources Fournier. CID. 201051 Jeserich. Ann
Intern Med. 1997126
38
Other Microbiologic Causes of BCNE

• Legionella species
• Rare cause of IE, described as cause of
  prosthetic valve IE
• Diagnosis by culture (difficult), serology,
  urinary antigen, PCR testing
• Brucella melitensis
• Rare (1 of BCNE) endemic to Mediterranean,
  Middle East, Asia, Africa
• Acquired via animal exposure, unpasteurized milk
• Diagnosis by blood culture (variable), serology,
  PCR testing

39
Other Microbiologic Causes of BCNE

• Chlamydophila (formerly Chlamydia) species
• Rare but reported in literature
• Serology may cross-react with Bartonella
  antibodies
• Mycoplasma species
• Also rare but well-described
• Diagnosis by serology, PCR testing
• Viruses
• Cause of myocarditis (enteroviruses) but not BCNE

Fournier. CID. 201096.
40
Non-Infectious Causes of BCNE

• Likely an important cause of BCNE, prevalence not
  well known
• 2.5 of BCNE in recent series
• Marantic, Libman-Sacks/autoimmume (SLE,
  rheumatoid arthritis, Behcets, anti-phospholipid
  antibody-related)
• Diagnosis clinical signs/symptoms, detection of
  autoantibodies

Fournier. CID. 201096. Library.med.utah.edu
41
Diagnostic Studies for BCNE

• Blood cultures
• Routine extended incubation does not improve
  yield
• HACEK organisms easily isolated with 5 day
  incubation in current blood culture systems
• Terminal subculture, lysis centrifugation culture
  may improve yield of certain organisms
• Brucella, fungi

Petti. J Clin Microbiol. 200644.
42
Diagnostic Studies for BCNE

• Serologic testing
• Molecular testing
• Blood, valve tissue, embolic vegetations
• Valve tissue PCR sensitivity 40-60, specificity
  near 100
• False negative pre-operative antibiotics
• False positive non-viable bacteria after
  treatment, contaminated tissue

Rovery. J Clin Microbiol. 200543.
43
Diagnostic Strategy for BCNE

• Largest diagnostic case series of BCNE
• 819 cases evaluated 2001-2009
• Most cases from France
• Definite and possible IE by modified Duke
  criteria
• Largest series in which PCR detection from
  valvular biopsies performed
• Use of several new diagnostic techniques

Fournier. CID. 201096
44
Serological Testing

• Coxiella burnetii
• Bartonella quintana and henselae
• Legionella pneumophila
• Brucella melitensis
• Mycoplasma pneumoniae
• 47.8 of pts with microbiological diagnosis made
  by serology

Fournier. CID. 201096
45
Molecular Testing

• Blood
• Broad-range PCR for bacteria (16s rRNA) and fungi
  (18s rRNA), some viruses
• Included specific primers for Coxiella,
  Bartonella sp., T. whipplei, Chlamydia sp., CMV,
  Enterovirus
• If valve tissue available
• Broad-range PCR for bacteria and fungi
• If other testing negative, primer extension
  enrichment reaction (PEER) PCR

Fournier. CID. 201096
46
Other Valvular Testing

• All suitable specimens cell cultured
• Bacteria detected identified by PCR testing
• Histopathological analysis
• Autoimmunohistochemistry
• Done for specimens in which all other techniques
  negative

Fournier. CID. 201096
47
Additional Testing

• All patients tested for rheumatoid factor,
  antinuclear antibodies, anti-DNA antibodies
• If all testing negative, physicians contacted
  regarding diagnosis of neoplastic or autoimmune
  disease made elsewhere

Fournier. CID. 201096
48
Distribution of 819 Suspected BCNE Cases
Fournier. CID. 201096
49
Yield of Additional Diagnostic Testing

• Most diagnoses made by serological testing (47.8
  of cases with microbiologic diagnosis)
• Chronic Q fever 77, Bartonella 22.5
• PCR second best diagnostic technique
• Blood poor sensitivity 36 of 257 (13.5)
  specimens tested positive
• Valve PCR high yield 157 of 227 (69) specimens
  positive
• Strep, fastidious bacteria no viruses identified

Fournier. CID. 201096
50
Proposed Diagnostic Strategy for BCNE
51
Limitations of New Diagnostics

• In most recent study, 36.5 of cases still
  undiagnosed after extensive testing
• Range 22-83 in various studies
• Overall sensitivity of diagnostic strategy varies
  by population
• Lower sensitivity if definite AND possible
  endocarditis included
• Poor sensitivity of commercially available PCR
  detection kits of blood

52
Is this strategy applicable to our US population?

• Lower incidence of zoonoses in US vs. Europe
  (Coxiella, Bartonella)
• Likely under reported/diagnosed Coxiella highly
  prevalent in US cattle, 22 veterinarians
  seropositive
• ?Similar rates of fungal etiologies
• Higher in N. America in studies of culture
  positive IE
• Availability of PCR testing

Kim. EID. 200511 Whitney. CID. 200948 Murdoch.
Arch Intern Med 2009169
53
Future Challenges in BCNE

• Additional studies of epidemiology of BCNE in
  US/North America
• Use of serologic testing for zoonoses based on
  exposure history vs. universal testing
• Role of non-infectious causes of BCNE
• Specificity of auto-antibody testing

54
Future Challenges in BCNE

• Improvements in PCR testing techniques,
  availability
• Development of highly sensitive PCR assays of
  blood for staph, strep
• Role of fungal serological testing, antigen
  assays
• Urine/serum antigens, galactomannan,
  (1,3)ß-D-glucan levels
• Proposed modification of Duke criteria to include
  molecular testing

55
Questions?