Screening Tests for ESBLs in H'influenzae

1
Screening Tests for ESBLs in H.influenzae
Stephen Glen Tristram School of Human Life
Sciences, University of Tasmania, Launceston,
Tas, Australia
Introduction TEM derived extended spectrum
ß-lactamases (ESBLs) and associated cephalosporin
resistance are well established in the
Enterobacteriaceae1. In comparison, despite the
fact that TEM-1 ß-lactamase positive ampicillin
resistant (BLPAR) strains of H. influenzae are
now common2, ESBLs are yet to be detected in this
organism2,3. The possibility of ESBLs emerging in
H. influenzae has been suggested3, and there has
been one report of a TEM-15 in H.
parainfluenzae4. Recently a range of ESBLs were
cloned into H. influenzae, and while the MICs of
extended spectrum cephalosporins were raised they
were not above the NCCLS breakpoints3. Based on
this observation, and the problems initially
encountered in the detection of ESBLs in the
Enterobacteriacea1, it is possible that ESBLs
have already emerged in H. influenzae and gone
undetected. The detection of ESBLs in H.
influenzae may be further complicated by the
emergence of ß-lactamase negative ampicillin
resistant (BLNAR) strains that have reduced
susceptibility to all ß-lactams, including
extended spectrum cephalosporins, due to
alterations in penicillin binding proteins
(PBPs). This may be especially so in those more
recently emerging strains that combine altered
PBPs with the presence of a ß-lactamase (BLPACR).
The aim of this study was to evaluate a range
of screening tests to establish their suitability
for the detection of ESBLs in H. influenzae.
Results and Discussion ALL the ESBL producing
clones tested as SUSCEPTIBLE to cefotaxime by the
standard NCCLS, BSAC and CDS disc diffusion
methods (data not shown), and keyhole effects or
distorted zones of inhibition were NOT seen
between the respective co-amoxiclav and
cefotaxime discs. Keyhole effects were seen with
the TEM-3,4 and 5 producing strains but not the
TEM-1 producing strains with the modified double
disc diffusion test. The keyhole effects were
more obvious with the lower strength cefotaxime
discs, and also more obvious with the plain
clavulanic acid discs compared to the
co-amoxiclav discs, but in all cases were highly
dependant on disc spacing and much less distinct
than those seen with ESBL producing
Enterobacteriaceae and they could NOT be
recommended for routine use. ALL of the ESBL
producing strains were CORRECTLY detected and
NONE of the other strains were WRONGLY detected
as possible ESBLs using the combination discs
(data not shown). ALL of the ESBL producing
strains were CORRECTLY detected using the
co-amoxiclav pre-diffusion disc replacement
tests, but of the other strains tested, 3 out of
5 BLNAR strains (and those with the highest
cefotaxime MIC) and the single BLPACR strain were
FALSELY detected as possible ESBLs when a
cefotaxime 0.5 µg disc was used. These strains
were CORRECTLY designated when higher strength 5
or 30 µg cefotaxime discs were used. (see Table 1
and Fig 1). This apparent synergy between the
cefotaxime and clavulanic acid is actually simply
due to the amoxycillin in the co-amoxiclav disc
and occurs as a result of relatively greater
decrease in susceptibility to most cephalosporins
compared to ampicillin or amoxycillin in BLNAR
strains5. When these strains were re-tested with
a cefotaxime 0.5 µg disc pre-diffused with a 2 µg
amoxicillin disc rather than a co-amoxyclav 2/1
µg disc, similar results were seen (data not
shown), but when combination discs with plain
clavulanic acid and a cephalosporin disc are
used, NO increase in zone size is seen.
Table 1. Selected Disc Diffusion Screening Test
Results
Methods Previously described strains of H.
influenzae producing a range of cloned TEM type
ß-lactamases3 (see Table 1) were used to
determine the ability of various screening tests
to detect ESBLs. Standard Disc Diffusion.
NCCLS, BSAC and CDS disc diffusion susceptibility
tests for cefotaxime were performed according to
the relevant methods to see if the different
breakpoints used in the various methods would
affect the results. A co-amoxiclav disc was also
tested and placed 22 mm from the cefotaxime disc
(as would occur in a 6 disc dispenser) to check
for keyhole effects. Other Disc Diffusion
Methods. All other tests described were performed
on chocolate agar (Columbia agar base with 8
chocolatised horse blood) and incubated at 35oC
in an atmosphere of 5 CO2 for 16-18 hrs. Inocula
were prepared to a 0.5 McF standard in 0.9
saline from overnight cultures, and inoculated by
flooding the plate. Modified Double Disc
Diffusion. A centrally placed cefotaxime disc
(30, 5 or 0.5 µg) was tested with 3 separate
co-amoxiclav 2/1 µg discs placed at increasing
distances in a cloverleaf arrangement. This test
was also repeated using freshly prepared 10 µg
clavulanic acid discs in place of the
co-amoxiclav discs. Co-amoxiclav Pre-diffusion
Disc Replacement. A 2/1 µg co-amoxiclav disc was
placed on an inoculated plate and allowed to
diffuse for 60 minutes before being removed and
replaced with a cefotaxime disc (30, 5 or 0.5
µg) or a cefpodoxime disc (10 µg). Another
cefotaxime or cefpodoxime disc of equal strength
was included on the plate for comparison. An
increase in zone diameter 5mm for the
pre-diffused disc relative to the plain disc was
considered suggestive of an ESBL. Combination
Discs. Commercially available discs (Oxoid,
Australia) containing a combination of either
cefotaxime (30 µg) or cefpodoxime (10 µg) and
clavulanic acid (10 or 1 µg) were tested in
parallel with a corresponding plain cephalosporin
disc. An increase in zone diameter 5 mm for the
clavulanic acid supplemented discs compared to
the plain discs was considered suggestive of an
ESBL. Further Evaluation. The co-amoxiclav disc
replacement and the combination disc tests were
selected as suitable for further evaluation and
used to test a selection of 20 BLPAR and 5 BLNAR
strains, and a single BLPACR strain (see Table 1).
Fig 1d
Fig 1f
Fig 1e
Conclusion Strains of H. influenzae showing
reduced susceptibility to extended spectrum
cephalosporins should first be tested for
ß-lactamase production, and if positive, screened
for the presence of an ESBL using a suitable
test. Co-amoxiclav pre-diffusion disc replacement
tests using co-amoxiclav 2/1 µg discs replaced
with a cefotaxime disc (either 5 or 30 µg, but
NOT 0.5 µg) and a 10 µg cefpodoxime disc, or
combination disc test using commercially
available cefpodoxime/clavulanic acid 10/1 µg and
cefotaxime/clavulanic acid 30/10 µg discs are
recommended. ESBL producing strains should should
give an increase in the diameter of inhibition
zone size 5mm compared to the plain
cephalosporin disc. Double disc diffusion tests
for demonstration of keyhole effects are NOT
recommended.

• References
• 1. Bradford, P. (2001) Clin Microbiol Rev, 14,
  933 - 951.
• 2. Hoban, D Felmingham, D. (2002) J Antimicrob
  Chemother, 50S1, 49 - 51.
• 3. Tristram, S. (2003) J Antimicrob Chemother,
  51, 39 - 43.
• 4. Pitout, M. et al. (2002) In Program and
  Abstracts, 42nd ICAAC.
• 5. Dabernet, H. et al. (2002) Antimicrob Agents
  Chemother, 46, 2208 - 2218.