Clinical data to support the
interpretation of susceptibility
testing of anaerobes
Robin Howe
'Mixed anaerobes
sensitive to metronidazole'
(common practice in UK)
Pus from cerebral abscess
Primary plate after 5 days
AnO2 incubation
Courtesy Val Hall
? But would we report:'Mixed aerobes sensitive to a cephalosporin'
'Fungus present, try Athlete's Foot powder'
'Virus detected, have a hot whisky and lemon'
And would we test them like this?…..
Courtesy Val Hall
'S.O.P.' for susceptibility testing
of anaerobes
• Agar: Any (selective or non-selective)
• Inoculum: Direct sample, ?mixed, not standardised
• Antimicrobial agents: metronidazole
• Incubation: On bench till 5pm, then AnO2 18-72hrs
• Controls: None
• Interpretation:
– Any sized zone = mixed anaerobes, sens to MZ
– No zone = no anaerobes isolated
– Colonies within zone = aerobes
Courtesy Val Hall
Why is susceptibility testing of anaerobes
not generally used in clinical decisionmaking?
• Technical issues with testing
– Slow growth
– Lack of consensus regarding agar/method
• Anaerobic susceptibility patterns
– Predictable
– Unchanging over years
• Limited data to correlate in vitro results with
outcome
– Infections often polymicrobial
– Outcome affected by multiple factors (eg surgery)
Are resistance rates predictable?
Does in vitro resistance correlate with an
identifiable resistance mechanism?
B. fragilis
Prevotella
Porphyromonas
Peptostreptococcus
Fusobacterium
C. difficile
% Clindamycin Resistance
1987
2000
3
26-44
<10
67
Resistance mechanism
ermF, ermG, ermS
ermF
ermZ, ermB, ermBZ
Hecht (2004) CID39: 92
Are resistance rates predictable?
Does in vitro resistance correlate with an
identifiable resistance mechanism?
% Resistance to -lactams
-lactam/
Penicillin Cephamycins
-lactamase
inhibitors
5
8-22
>97
B. fragilis
83
Prevotella
9
Fusobacterium
0-5
21
Porphyromonas
6
Peptostreptococcus
Non-perfringens
0-5
0-88
Clostridia
0
0-10?
C. perfringens
Carbapenems
<0.2
0-5
0
Hecht (2004) CID39: 92
• Baquero (1992)
– 10% C. perfringens in
Spain resistant to pen
(MIC >0.5 mg/L)
BSAC BP 0.12 mg/L
Are resistance rates predictable?
Does in vitro resistance correlate with an
identifiable resistance mechanism?
% Resistance to -lactams
-lactam/
Carbapenems
Penicillin Cephamycins
-lactamase
inhibitors
<0.2
5
8-22
>97
B. fragilis
83
Prevotella
•B.
fragilis
9
Fusobacterium
0-5
0-5
•Class 2e cephalosporinase
21
Porphyromonas
•Altered PBPs
Peptostreptococcus
•V. common 6
•Rare
Non-perfringens
0-5
•Inhibited by0-88
clavulanate etc
•Porin
loss
Clostridia
•Cephamycinases
0
0 •Reported
0-10?
C. perfringens
•Uncommon
•cepA, cfxA
•Zinc metallo--lactamase
•cfiA, ccrA
•Present in ~4% - not usually expressed
Are resistance rates predictable?
Does in vitro resistance correlate with an
identifiable resistance mechanism?
• Reduced susceptibility to
metronidazole
– Common in
• Propionibacteria & actinomycoses
– Rare in
• B. fragilis
– nim genes
• C.difficile
REFERRALS TO ARU
•Bacteroides spp. n=78
(5% of all Bacteroides)
• Clostridium paraputrificum
n=5 (4%)
• Clostridium ramosum
n=3 (1%)
nim genes
• Nim = nitro-imidazole reductase
• Types A – G found in Bacteroides spp.
• Detected by PCR-RFLP
• Chromosomal / plasmid-borne
• Absent from some MZ resistant orgs
• Probable alternative mechanisms
• High level MZ resistance can be induced in
some nim-containing strains
Courtesy Val Hall
• nim genes identified in 2% of 1,502 B. fragilis
from 19 European countries
• nim genes identified in 2% of 1,502 B. fragilis
from 19 European countries
The role of anti-anaerobic therapy
Animal studies
• Rat model of secondary peritonitis
– Pooled faecal emulsion intraperitoneally
– initially E. coli predominates with often fatal
bacteraemia
– If survive  abscesses with B. fragilis
– Early gentamicin  no bacteraemia but late abscesses
– Early clindamycin  no effect on bacteraemic
mortality but reduced late abscesses in survivors
Onderdonk et al (1974) Infect Immun 10:1256
Animal studies
?synergistic infection
• IP injection of mixtures of three orgs
FOX
8
16
8
8
2
B. fragilis
B. thetaiotaomicron
B. ovatus
B. distasonis
E. coli
No
deaths
E. coli
B. fragilis
B. theta
Saline
Cefoxitin
Cefotetan
Amp/sul
13
4
0
7
Cefotetan
8
64
128
128
0.5
No with
abscesses in
survivors
17/17
3/26
3/30
4/23
Amp/sul
2
4
2
4
8
Bacterial counts
B.
B.
E. coli
fragilis
theta
7.4
8.7
8.3
4.8
4.7
3.9
4.5
4.0
3.6
5.4
3.2
2.8
Brook (1994) JAC 34: 791
Reports of clinical failure
associated with resistance
• Penicillin vs C. perfringens
– NIL
• Metronidazole vs Bacteroides spp
– YES
Rotimi et al (1999)CMI 5: 166
• 3 case reports
– 75 yrs female
•
•
•
•
post op Hartmann’s treated with CAZ/MTZ
Readmitted with paracolic abscess
B. frag isolated (MTZ MIC >32 mg/L)
Cured with drainage + IMI
– 40 yrs male
• Gangrenous appendix (mixed B. frag/E. coli/Pseudomonas)
• CXM/MTZ started
• Day 5 - Wound infection
– B. frag + B. ovatus (MTZ MIC >32mg/L)
– Cured with co-amox
– 37 yrs male
•
•
•
•
•
Post renal transplant
cholecystitis cholecystectomy necrotising pancreatitis
Multiple Abs – CTX/MTZ/AMIK – MEM/AMIK/CPM
Mixed isolates from lap inc B. distasonis (MIC MTZ >32 mg/L, MEM >32 mg/L)
Pt died
Reports of clinical failure
associated with resistance
• Penicillin vs C. perfringens
– NIL
• Metronidazole vs Bacteroides spp
– YES
• Penicillin vs Bacteroides spp.
– YES
Brook (1984) Arch Otolaryngol 110: 228
Gudiol (1990) Arch Intern Med 150: 2525
Reports of clinical failure
associated with resistance
• Penicillin vs C. perfringens
– NIL
• Metronidazole vs Bacteroides spp
– YES
• Penicillin vs Bacteroides spp.
– YES
• Β-lactam/β-lactamase inhibitors vs Bacteroides
spp.
– NIL
• Carbapenems vs Bacteroides spp.
– YES
• 38 year old female
–
–
–
–
–
–
–
Elective laparotomy for adhesions
Post-op IA collection treated with co-amox
Day 13 - surgical drainage & change to CTX + MTZ
BC grew B. fragilis (isolate 1)
Changed to imipenem
2 weeks later persistent empyema drained (isolate 2)
Cured with drainage/clindamycin/gentamicin
Turner et al (1995) Lancet 345: 1275
• Prospective
evaluation of 128
patients with
Bacteroides
bacteraemia
CID (2000) 30: 870
Any data relating level of
resistance to outcome?
• NO
Conclusions
• Antimicrobial resistance is variably predictable
• Resistance rates are increasing
– CLD – becoming common
– MTZ + carbapenems – emerging
• Inducibility is a concern
• Correlation between in vitro resistance and
outcome has not been established for many
anaerobic infections
– The role of surgery should not be forgotten
Finegold 1989
• Susceptibility testing of anaerobes should be done
in 4 settings:
–
–
–
–
Determine patterns of susceptibility to new agents
Monitor susceptibility patterns Nationally
Monitor susceptibility patterns locally
Assist in the management of individual patients
• Persistence of infection/ failure of usual regimes/ difficulty
making decisions based on precedent
• Brain abscess/ endocarditis/ osteomyelitis/ prosthetic device
infection/ septic arthritis