Detection of ESBLs & AmpC David Livermore Health Protection Agency, Colindale, London

Some premises Growing resistance to 3-gen cephs Mostly ESBLs in E. coli & Klebsiella; AmpC in Enterobacter, Citrobacter, Serratia… but not always Identification of mechanism aids Epidemiological investigation / control Treatment choice Recognition of the exceptional e.g. MBLs

Resistance to 3 gen cephs; BSAC bacteraemia surveillance (%) From http://www.bsacsurv.org

EARSS resistance to 3-gen cephs in E. coli 2001 2006 http://www.earss.rivm.nl

Detecting ESBL producers 2 steps: Screen for resistance with an indicator ceph Do confirmatory test on those found resistant

Choice of indicator cephalosporin Sensitivity Specificity Cefotaxime & ceftazidime Good Cefpodoxime Moderate Cefuroxime Poor Cephalexin or cephradine Cefpirome or Cefepime

Detection of ESBLs: step 2 Seek ceph/clav synergy in ceph R isolates Double disc Combination disc Etest See http://www.hpa.org.uk

Combination discs Disc with cephalosporin + clavulanic acid Disc with cephalosporin alone

Zone differences (mm), Klebs & E. coli c’pod/clav 10+1 mg - c’pod 10 mg

Etest for ESBLs Cefotaxime Cefotaxime + clavulanate

Etest for ESBLs Cefotaxime Cefotaxime + clavulanate

ESBL confirmatory tests Pro Contra Double disc Cheap Best disc spacing varies with strain Combination disc Cheap, sensitive & specific Batch variation; Controls critical Etest Sensitive Internally controlled More expensive; False +ve's with K1 in K. oxytoca

Controls for ESBL tests +ve E. coli with TEM-3, -10 & CTX-M-15 available as NCTC 13351, 13352, 13353 No one control is perfect… and these have high levels of enzyme whilst some clinical isolate have very low levels –ve E. coli (e.g. NCTC 10418) Critical for combination discs; should give equal zones irrespective of clavulanate

Further investigating ESBLs Multiplex PCR for 5 blaCTX-M groups* TEM & SHV mutants require sequencing Beware…. Isolates may have >1 enzyme e.g. Classical TEM / SHV + TEM / SHV ESBL Many with CTX-M-15 also have OXA-1 & TEM-1 Isoelectric focusing gives fullest picture

ESBL tests for AmpC inducible species Methods optimised for E. coli & Klebsiella More difficult with Enterobacter clavulanate induces AmpC; hides ESBL Advice is to do synergy test (NOT SCREEN) with 4th gen ceph; specificity good, sensitivity moderate BSAC bacteraemia: c. 25% CephR Enterobacter have ESBL, not AmpC….. Probably an underestimate

Bacteria not to test for ESBLs Acinetobacters Often S to clavulanate alone S. maltophilia +ve result by inhibition of L-2 chromosomal b-lactamase, ubiquitous in the species

Suspect derepressed / plasmid AmpC if: Resistant 3-gen cephs, NOT cefepime & cefpirome Resistant to cefoxitin (but more ESBL producers R, too, nowadays) No ceph/clav synergy

Geometric mean MICs (mg/L): AmpC producers; 2004 London SE survey E. coli AmpC E. coli ESBL Enterobacter AmpC Enterobacter ESBL Cefotaxime 12 228 72 49 Ceftazidime 18 39 36 81 Cefepime 0.38 0.91 4.4 Cefpirome 0.57 53 2.4 10 Cefuroxime 35 >64 Cefoxitin 51 17 Potz et al., JAC 2006; 58:320-6

Some wrinkles… AmpC-derepressed M. morganii are S to pip/tazo AmpC derepressed Serratia are S to ceftazidime Cefoxitin R an unreliable marker for Providencia, Morganella & Serratia spp. Inducible & derepressed strains may appear I or S AmpC derepressed P. aeruginosa tend to be S to carbenicillin / efflux mutants are R

Confirmatory tests for AmpC Seek cefotaxime/cloxacillin synergy Cefotaxime MIC +100 mg/L cloxacillin Zones of cefotaxime 30 mg discs on agar + 100 mg/L cloxacillin No agreed interpretive standards Can also use phenylboronic acid as inhibitor

Cefotaxime combinations vs. AmpC E. coli: London SE survey Potz et al., JAC 2006; 58:320-6

Cefotaxime combinations vs. AmpC Enterobacter: London SE survey Potz et al., JAC 2006; 58:320-6

Cefotaxime / cloxacillin tests for AmpC ARMRL- reference control data

Phenyl boronic acid for detection of plasmid AmpC Expansion (mm) of FOX 30 zone with 400 mg phenyl boronic acid Fold MIC reduction for cefoxitin + 400 mg/L phenyl boronic acid Kleb MOX-1 12 128 E. coli LAT-2 64 Kleb DHA-1 14 Kleb DHA-2 13 E. coli ACC-1 4 Kleb ACT-1 ALL ESBL +ve <2 Coudron JCM 2005 43 4163

BZB: benzo(b)thiophene-2-boronic acid Disc tests for AmpC 60 E. coli & Klebsiella : cefoxitin MICs reduced >4-fold by 100 mg/L cloxacillin % with >5 mm zone expansion Cefoxitin + cloxacillin 100 mg 86% Cefoxitin + BZB 64 mg 89% Cefpodoxime + BZB 64 mg 97% Cefpodoxime + clav + BZB 64 mg 100% BZB: benzo(b)thiophene-2-boronic acid Brenwald et al., JAC 2005, 56, 600

Looks for distortion where cross intersects the cefoxitin zone 3-D test for AmpCs Plate seeded with cefoxitin S indicator strain Cut cross in agar, heavily inoculated with test strain Cefoxitin disc Looks for distortion where cross intersects the cefoxitin zone

Clover leaf (3 dimensional) test for AmpC Test strain E. cloacae, AmpC derepressed Indicator E. coli NCTC10418 Disc Cefoxitin 30 mg

Multiplex detection of plasmid AmpC genes Method of Perez-Perez & Hanson JCM 2002, 40, 2153

AmpC commercial tests ROSCO- ‘research only’ : high content (500 mg) cloxacillin & boronic acid discs for double disc synergy tests AB Biodisk- evaluating double-ended cefotetan or cefoxitin plus cloxacillin or boronic acid ETests

Hyperproduction of K1 enzyme Unique to K. oxytoca, chromosomal Indole +ve Klebsiella R cefuroxime, aztreonam, pip / tazo, ceftriaxone Borderline (S/I/R) to cefotaxime S to ceftazidime & carbapenems Weak cefotaxime or cefepime/clav synergy

MICs (mg/L) for multi-resistant UK klebsiellas Example 1 2 3 4 Cefotaxime >64 Ceftazidime CTX/clav >32 CAZ/clav 16 Cefepime Cefepime/clav 32 Ertapenem >16 Meropenem 8 Imipenem >200 isolates; 60 centres; many strains. No imipenem hydrolysis with crude extract Carbapenem resistance not transferable In the last year we have had 21 isolates from 12 centres referred to us with a characteristic antibiogram. They were referred on the basis of meropenem resistance for carbapenemase investigation. MIC’s showed a pattern whereby the ertapenem MIC was greater than the meropenem MIC which was greater than the imipenem MIC which tended to be around the breakpoint. The isolates were resistant to all antibioticss tested, including the beta-lactamase inhibitors with the exception of colistin and in some cases tigecycline. We found the isolates to be mostly unrelated by PFGE with small strain groups from single centres. No imipenem hydrolysis was evident using crude cell extract and carbapenem resistance was not transferable. We found the isolates to be negative for every carbapenemase gene tested. Woodford et al. IJAA 2007 ;29:456-9.

Mechanisms of multi-resistant UK klebsiellas OMP Profile S R R R R Mechanism is combination of porin loss & CTX-M-15 Occasionally selected during therapy Further characterisation showed that the isolates carried CTX-M and OXA-1-like genes and showed loss of at least one major outer membrane protein. Strains all carry at least one plasmid of >100Kb. IEF showed bands corresponding to CTX-M-15 and OXA-1-like with some also having a TEM enzyme, these results were confirmed by PCR. PCR mapping of the OmpK genes showed disruption of one or more of OmpK35, 36 and 37 and sequenced examples showed the disruption to be caused by insertion of an IS element into the gene. Other examples showed truncation due to a point mutation Woodford et al. IJAA 2007 ;29:456-9

Acquired carbapenemases KPC Class A, 4 variants Spreading world-wide, 2 cases in UK…. so far Often clonal, mostly Klebsiella, Enterobacter Metallo’s Class B, VIM, IMP families, also SPM, SIM, GIM Scattered, mostly non-fermenters >100 UK in since 2001, mostly VIM+ P. aeruginosa

Carbapenemase or not... KPC…. clearest R to ertapenem; no synergy in clavulanate, cloxacillin, boronic acid or EDTA tests Easy to confuse with combination of ESBL + impermeability Metallo’s…. Suspect if isolate has reduced carbapenem susceptibility, reversed by ESBL… But Frank carbapenem resistance not always seen EDTA tests not specific… many false +ves Spare aztreonam, may be affected by other mechanisms

A problem in Bolzano 209 Ceph R Enterobacteriaceae, most had ESBLs 24 lacked ceph / clav synergy- mix of E. coli, K. pneumoniae, K. oxytoca, Citrobacter Imipenem MICs 2- >32 mg/L, mostly 4-8 mg/L Meropenem, ertapenem MICs lower than imipenem Imipenem + EDTA MICs 0.12-1 mg/L All had blaVIM; mix of clonal & non-clonal!!! 19 R to aztreonam--- had CTX-M ----5 susceptible Aschbacher et al, submitted

Cica b-Test (Mast) Examine hydrolysis of chromogenic oxyimino ceph, HMRZ-86- yellow to red If +ve, test inhibition IN SEQUENCE by: Sodium mercaptoacetic acid – MBL Clavulanic acid – Class A / ESBL Benzo-thiophene-2-boronic acid – AmpC Count first positive result

Cica b-Test (Mast) No inhibitor Mercaptoacetic acid to inhibit MBL Clavulanate to inhibit ESBL Boronic acid to inhibit AmpC

Cica b-Test (Mast) blind testing of overnight cultures Mechanism inferred Reference data MBL ESBL AmpC Mixed Other Pen’ ase No activity MBL (26) 20 1 2 3 ESBL (74) 63 6 AmpC (25) 18 K. oxytoca, K1 (10) OXA carbapenemases (10) 10 P. aeruginosa OXA ESBLs (4) KPC/SME carbapenemase (2) Penicillinase (39) 5 30 Better but slower to use with antibiogram @ 48h Livermore et al., JAC in press.

Summary Labs should be able to recognise ESBL producers Even among Enterobacters Ref lab will look at difficult cases Labs should be able to recognise AmpC derepressed strains & those with plasmid AmpC Enterobacteriaceae with reduced carbapenem susceptibility need reference investigation New tests being developed