Antimicrobial susceptibility testing in Europe the role of

Antimicrobial susceptibility testing in Europe - the role of national breakpoint committees and EUCAST Gunnar Kahlmeter, EUCAST gunnar. kahlmeter@ltkronoberg. se NCCLS (USA) BSAC wp DIN CA-SFM CRG NWGA SRGA

E. coli vs. cefotaxim S< / R> R (%) 2/2 1. 8 % 4 / 32 0. 6 % CRG 4/8 0. 9 % DIN 2/8 1. 1 % NCCLS 8 / 32 0. 6 % NWGA 1/2 1. 8 % SRGA 0. 5 / 1 2. 6 % ND - (WT< 0. 5) 5. 0 % BSAC CA-SFM EUCAST tentative clinical brpt EUCAST epidemiological cut off

Simulated local outbreak!

The breakpoint compromise! Clinical aspects – breakpoints must have clinical value! Epidemiological aspects – breakpoints must allow (early) detection of acquired resistance mechanisms! Methodological aspects – breakpoints must allow reproducible S-, I- and R-categorization in the lab! <2. 0

EUCAST European Committee on Antimicrobial Susceptibility Testing formed in 1997 and restructured in 2002 convened by European Society for Clinical Microbiology and Infectious Diseases (ESCMID) National Breakpoint Committees in Europe and financed by ESCMID National Breakpoint Committees in Europe DG-SANCO of the European Union (3 year grant from May 2004)

EUCAST • EUCAST General Committee: - one representative, appointed by the appropriate medical associations, from each European country - one representative each from ISC and FESCI - Chairperson and Scientific secretary (appointed by ESCMID) - meets once a year at ECCMID - all Steering Committee proposals are referred to the General Committee for comments before decision • EUCAST Steering Committee: - Chairperson and a Scientific Secretary (appointed by ESCMID) - one representative each from the European national breakpoint committees (presently 6) - two representatives from the EUCAST General Committee - Czech Republic and Greece 2002 -2004 - Russia and Spain 2004 -2006 • EUCAST industry network - The network consists of all interested manufacturers of pharmaceuticals and susceptibility testing devices. All are invited to take an active part in EUCAST activities - Steering Committee proposals are referred to the industry network for comments before decision - relevant industry members can apply for inclusion on the email list by contacting the EUCAST secretariat

EUCAST General Committee 2004 Austria Prof Helmut Mittermayer Belgium Prof Jan Verhaegen Bosnia Dr Selma Uzunovic-Kamberovic Bulgaria Prof Krassimir Metodiev Croatia Dr Arjana Tambic-Andrasevic Czech Republic Dr Pavla Urbaskova Denmark Dr Niels Frimodt-Møller Estonia Dr Paul Naaber Finland Dr Antti Nissinen France Prof Claude-James Soussy Germany Prof Bernd Wiedemann Greece Prof Alkiviadis Vatopoulos Hungary Dr Éva Bán Iceland Dr Karl Gustaf Kristinsson Ireland Dr Martin Cormican Italy Prof Pietro Emanuele Varaldo Latvia Dr Arta Balode Lithuania Prof Arvydsa Ambrozaitis Netherlands Prof John Degener Norway Dr Martin Steinbakk Poland Prof Waleria Hryniewicz Portugal Prof Jose Melo Cristino Romania no official representative Russia Dr Olga Stetsiouk Serbia Dr Lazar Ranin Slovak Republic Prof. Milan Niks Slovenia Dr Jana Kolman Spain Dr Francisco Soriano Sweden Dr Barbro Olsson-Liljequist Switzerland Prof Jaques Bille Turkey Dr Deniz Gür UK Prof Alasdair Mac. Gowan Yugoslavia no official representative ISC – Prof Paul Tulkens FESCI – Prof David Livermore Email network of industry with interest in antimicrobials Chairperson Gunnar Kahlmeter, Sweden Scientific Secretary Derek Brown, UK

EUCAST Steering Committee Membership • Chairperson • Scientific Secretary • BSAC (The UK) • CA-SFM (France) • CRG (The Netherlands) • DIN (Germany) • NWGA (Norway) • SRGA (Sweden) Gunnar Kahlmeter Derek Brown Alasdair Mac. Gowan Fred Goldstein Johan W. Mouton Arne Rodloff Martin Steinbakk Anders Österlund • General Committee rep Olga Stetsiouk (Russia) • General Committee rep Francisco Soriano (Spain) 2002 - 05 2002 - 05 2004 - 06

EUCAST definitions of clinical breakpoints available at www. eucast. org Clinically Susceptible (S) • a microorganism is defined as susceptible by a level of antimicrobial activity associated with a high likelihood of therapeutic success • a microorganism is categorized as susceptible (S) by applying the appropriate breakpoint in a defined phenotypic test system Clinically Intermediate (I) • a microorganism is defined as intermediate by a level of antimicrobial activity associated with indeterminate therapeutic effect • a microorganism is categorized as intermediate (I) by applying the appropriate breakpoints in a defined phenotypic test system Clinically Resistant (R) • a microorganism is defined as resistant by a level of antimicrobial activity associated with a high likelihood of therapeutic failure. • a microorganism is categorized as resistant (R) by applying the appropriate breakpoint in a defined phenotypic test system EUCAST has re-defined susceptible, intermediate and resistant and defined the terms wild type and non-wild type microorganism…. . and agreed… Clinical breakpoints may be altered with legitimate changes in circumstances Clinical breakpoints are presented as S<x mg/L; I>x, <y mg/L; R>y mg/L

EUCAST definitions of epidemiological cut off values available at www. eucast. org Wild type (WT) • a microorganism is defined as wild type (WT) for a species by the absence of acquired and mutational resistance mechanisms to the drug in question. • a microorganism is categorized as wild type (WT) for a species by applying the appropriate cut-off value in a defined phenotypic test system. • wild type microorganisms may or may not respond clinically to antimicrobial treatment. Microbiological resistance - non-wild type (NWT) • a microorganism is defined as non-wild type (NWT) for a species by the presence of an acquired or mutational resistance mechanism to the drug in question. • a microorganism is categorized as non-wild type (NWT) for a species by applying the appropriate cut-off value in a defined phenotypic test system. • non-wild type microorganisms may or may not respond clinically to antimicrobial treatment. Epidemiological cut-off values will not be altered by changing circumstances. The wild type is presented as WT<z mg/L and non-wild type as NWT >z mg/L

EUCAST publications 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. European Committee on Antimicrobial Susceptibility Testing. (2000). Terminology relating to methods for the determination of susceptibility of bacteria to antimicrobial agents. EUCAST Definitive Document E. Def 1. 2. Clinical Microbiology and Infection 6, 503 -8. European Committee on Antimicrobial Susceptibility Testing. (2000). Determination of antimicrobial susceptibility test breakpoints. EUCAST Definitive Document E. Def 2. 1. Clinical Microbiology and Infection 6, 570 -2. European Committee on Antimicrobial Susceptibility Testing. (2000). Determination of minimum inhibitory concentrations (MICs) of antibacterial agents by agar dilution. EUCAST Definitive Document E. Def 3. 1. Clinical Microbiology and Infection 6, 509 -15. European Committee on Antimicrobial Susceptibility Testing. (2001). Linezolid breakpoints. EUCAST Definitive Document E. Def 4. 1. Clinical Microbiology and Infection 7, 283 -4. European Committee on Antimicrobial Susceptibility Testing. (2003). Determination of minimum inhibitory concentrations (MICs) of antibacterial agents by broth microdilution. EUCAST Discussion Document E. Def 5. 1. Clinical Microbiology and Infection 9 (issue 7 insert) 1 -10. Ridgway, G. L. , Bébéar, C. M, et al. (2001). Antimicrobial susceptibility testing of intracellular and cell-associated pathogens. EUCAST Discussion Document E. Dis 6. 1. Clinical Microbiology and Infection 7 (issue 12 insert), 1 -10. Rodriguez-Tudela, J. L. , Barchiesi, F. , Bille, J. et al. (2003). Determination of minimum inhibitory concentrations by broth microdilution of fermentative yeasts. EUCAST Discussion Document E. Dis 7. 1. Clinical Microbiology and Infection 9 (issue 8 insert), 1 -8. Drobniewski, F. (2002). Antimicrobial susceptibility testing of Mycobacterium tuberculosis. EUCAST Discussion Document E. Dis 8. 1. Clinical Microbiology and Infection 8 (issue 10 insert), 1 -10. Kahlmeter G, Brown DFJ, Goldstein FW et al. (2003) European harmonization of MIC breakpoints for antimicrobial susceptibility testing of bacteria. Journal of Antimicrobial Chemotherapy 52, 145 -148. Kahlmeter G & Brown D. Harmonisation of European breakpoints – can it be achieved? Clinical Microbiology Newsletter, in press. Published documents, discussion documents and tentative decisions are posted on the EUCAST website. Tentative decisions and discussion documents - after a period of consultation they will be submitted for publication and/or be made available on the EUCAST website as final documents or decisions (www. eucast. org).

EUCAST Subcommittee on Antifungal Susceptibility Testing (EUCAST AFST) • develop reference methods for antifungal susceptibility testing • set breakpoints for antifungal drugs Financed through EUCAST processes for breakpoint setting, decisions and consultation

EUCAST collaborations • EMEA – SOP being developed which renders EUCAST the official European breakpoint committee (EUCAST breakpoints in the SPCs). • Expert groups and reference laboratories (Neisseria, Salmonella, European Veterinary working group, NEQAS) • EARSS – EUCAST permanently on the EARSS advisory board • NCCLS – NCCLS/EUCAST broth dilution method for the determination of MIC-values harmonised through CEN and ISO; finished document end of 2004. – NCCLS/EUCAST common QC type strain MIC-target values and ranges – NCCLS/EUCAST harmonised FQ breakpoints for staphylococci. – Collaborative process for revision of cephalosporin and carbapenem MIC breakpoints (first joint meeting in Tampa, January 2005). • Pharmaceutical industry information and consultation network

Liaison with NCCLS

Breakpoints for new antimicrobials • Daptomycin • Tigecycline

EUCAST procedure for setting breakpoints The next 9 slides describe the EUCAST procedure for harmonising European breakpoints.

1. Data on dosing, formulations, clinical indications and target organisms are reviewed and differences which might influence breakpoints are highlighted Example: ciprofloxacin National breakpoint committees BSAC UK CA-SFM France CRG Netherlands DIN Germany NWGA Norway SRGA Sweden Most common dose 500 x 2 oral 400 x 2 iv 500 x 2 oral 200 x 2 iv 250 x 2 oral 200 x iv 500 x 2 oral 200 x 2 iv 200 -400 x 2 oral 400 x 2 iv 500 x 2 oral 400 x 2 iv Maximum dose schedule 750 x 2 oral 400 x 3 iv 750 x 2 oral 400 x 2 iv data pending 750 x 2 oral 400 x 3 iv oral, iv oral, iv Dosage Available formulations Clinical data There is clinical evidence for ciprofloxacin to indicate a poor response in systemic infections caused by Salmonellae with low-level fluoroquinolone resistance (MIC>0. 064 mg/L) EUCAST has suggested that the epidemiological cut off value (S<0. 064/R>0. 064 mg/L) be used in Salmonellae systemic infections. These strains are best found using a nalidixic acid 30 µg screen disc in routine susceptibility testing. There is agreement in EUCAST that ciprofloxacin activity against Enterococci and Streptococci, including S. pneumoniae, is insufficient to categorize wild type bacteria “susceptible”.

2. Multiple MIC-distributions are collected, the wild type MIC distribution is defined and tentative epidemiological cut-off values determined (WT < X mg/L) Epidemiological cut off: WT<2. 0

3. Existing national clinical breakpoints are compared Ciprofloxacin was used in this example: Breakpoints prior to harmonisation (mg/L) S< / R> General breakpoints BSAC CA-SFM CRG DIN NWGA SRGA ND 1/2 1/2 0. 125/2 1/2 Species related breakpoints not yet NCCLS no Enterobacteriaceae 1/1 0. 12/2 0. 12/1 1/2 Pseudomonas spp. 1/4 ND 1/1 1/2 0. 12/2 0. 06/2 1/2 Acinetobacter spp. Staphylococci 1/1 Streptococci 1/1 excluded 0. 12/2 excl 2/2 (I)* excluded 0. 12/2 (I)* excluded 0. 12/2 1/2 0. 12/0. 5 0. 12/0. 25 1/- S. pneumoniae Enterococci Haemophilus/Moraxella spp. 1/1 Corynebacteria N. Meningitidis N. Gonorrhoeae P. Multocida Anaerobes excl 1/1 0. 06/0. 12 0. 03/0. 25 0. 06/0. 12 0. 06/0. 25 ND ND 0. 12/0. 25 excluded ND excluded no no 0. 06/- Campylobacter spp. 1/1 Helicobacter pylori 2/2 0. 06/1 no no 0. 06/0. 5

4. Using available PK/PD data, Monte Carlo simulations are performed and a tentative breakpoint calculated ”Minimum requirement for Scategory” is that the high MIC value of the wild type MICdistribution is consistent with the MIC derived from the PK/PD index needed for optimal efficacy based on free drug”.

5. Clinical data relating outcome to MIC-values and resistance mechanisms are assessed in relation to the tentative breakpoint …. and you´ve just heard Professor Mac. Gowan adress that issue

6. Tentative breakpoints are checked against target species wild type MIC distributions to avoid splitting the wild type …it was decided to set the break-point at S≤ 0. 125 and R>2 mg/L, rendering wild type S. pneumoniae intermediately susceptible to ciprofloxacin. <2 mg/L Epidemiological cut off: WT<2. 0 To permit reproducible susceptibility testing, splitting of the wild type must be avoided: 0. 5 and 1 mg/L were not acceptable…. and 2 mg/L was considered too high…

7. Consultation process on tentative breakpoints - national committees - EUCAST general committee - pharmaceutical industry, AST device manufacturers - others via EUCAST website 8. Re-appraisal of tentative breakpoints following comments 9. Further consultation if required

EUCAST has now finalised breakpoints for fluoroquinolones, glycopeptides, aminoglycosides and linezolid and is now EUCAST breakpoint tables are published working on cephalosporins, carbapenems and aztreonam. at www. eucast. org

EUCAST websites are found at www. eucast. org The EUCAST website is a section of the official ESCMID website – it gives details of all EUCAST activities including - constitution and organisation - committee member lists - meetings - EUCAST documents - EUCAST power-point presentation - clinical MIC breakpoint tables - MIC distributions for wild type bacteria and fungi - epidemiological MIC cut-off values.

www. eucast. org This is the first screen of the EUCAST general website.

Choose to display in English, French or German First screen of the EUCAST program for the display of wild type MIC distributions in microorganisms. www. eucast. org

EUCAST aggregated wild type MIC distributions - the fact that wild type microrganisms of the same species exhibit identical MIC (and inhibition zone) distributions irrespective of origin provides an opportunity to aggregate multiple distributions and define a reference. • • Data from scientific papers, breakpoint committees, reference laboratories, pharmaceutical industry, surveillance networks (EARSS, Sentry, Alexander Project, National surveillance networks) All submitted full-range MIC distributions accepted • To date no systematic exclusions. • October, 2004 – 3500 MIC distributions

S. pneumoniae and ciprofloxacin MIC distributions This slide shows a portion of the data set for S. pneumoniae and ciprofloxacin. Each MIC distribution is from a different investigator, surveillance program, breakpoint committee or pharmaceutical company. The median of the uni- or of the first part of the bi- or multimodal distribution has been marked in blue.

Aggregated S. pneumoniae ciprofloxacin MIC-data. Values of >1% are shown in graph as bars.

Choose to display in English, French or German www. eucast. org

Specify the drug or the bug (never both) - after a moment a table of MIC-distributions is shown. Click on any species in the left hand column to display the data as a bar chart, with EUCAST epidemiological cut-off values and harmonised European clinical breakpoints.

(1) To define epidemiological cut-off values

(2) As a template for calibration of methodology (accuracy and imprecision). ”We have defined the result of antimicrobial susceptibility testing!”

(3) Reference MIC database for breakpoint setting - to avoid clinical breakpoints that divide wild type bacteria

(4) As MIC reference database

EUCAST wild type MIC distributions – templates for calibration of MIC determinations Laboratories which cannot fit their own MIC data to the EUCAST reference distribution should look into the following possibilities: • • • The method used for MIC determination in the local set of data is not adequately standardised or calibrated, The species identification is incomplete There are too few determinations to allow identification of the part of the distribution that constitutes the wild type microorganisms (which usually corresponds to the four lowest dilution steps).

Examples from the EUCAST wild type MIC distribution program. 1

The objectives of EUCAST are • to form in EUCAST, under the auspices of the European Society of Clinical Microbiology and Infectious Diseases", a professional network of - the national breakpoint committees and experts on antimicrobial susceptibility testing and - industry involved in the production and marketing of antimicrobial agents or of in-vitro diagnostic medical devices used in antimicrobial susceptibility testing; • to set common European breakpoints for surveillance of antimicrobial resistance; • to identify national differences in clinical breakpoints and to harmonise breakpoints for existing and new antimicrobial drugs; • to produce, disseminate and update a series of documents on the technology of in-vitro antimicrobial susceptibility testing, promoting standardisation of methods used in different parts of Europe and comparability of results obtained by different technologies; • to encourage internal and external national and international quality assessment schemes; • to collaborate with European and international groups concerned with antimicrobial susceptibility testing and/or the epidemiology of antimicrobial resistance; • to advise European Community Institutions on the technology and interpretation of antimicrobial susceptibility testing; • to work with groups outside Europe (eg NCCLS) to achieve international consensus on susceptibility testing; • to devise and participate in educational and training programmes for antimicrobial susceptibility testing (workshop with EARSS in 2005, two workshops for national breakpoint committees in 2005 & 2006).

The End

E. coli and ciprofloxacin in ECO • SENS Wildtype distributions do not vary with geographical origin! G Kahlmeter, JAC, 2003.

Wildtype distributions do not vary with geographical origin!

Origin

Summary of EUCAST procedure for setting clinical breakpoints 1. Data on dosing, formulations, clinical indications and target organisms are reviewed and differences which might influence breakpoints are highlighted. 2. Multiple MIC-distributions are collected, the wild type MIC distribution is defined and tentative epidemiological cut-off values determined (WT < X mg/L). 3. Existing national clinical breakpoints are compared. 4. Using available PK/PD data, Monte Carlo simulations are performed and a tentative breakpoint calculated. ”Minimum requirement for S-category” is that the high MIC value of the wild type MIC-distribution is consistent with the MIC derived from the PK/PD index needed for optimal efficacy based on free drug”. 5. Clinical data relating outcome to MIC-values and resistance mechanisms are assessed in relation to the tentative breakpoint. 6. Tentative breakpoints are checked against target species wild type MIC distributions to avoid splitting the wild type. 7. Consultation process (national committees, EUCAST general committee and pharmaceutical industry and AST device manufacturers). 8. EUCAST clinical breakpoint tables are published on the internet (www. eucast. org) with links to tables and graphs of wild type distributions of MIC values.

How to implement EUCAST breakpoints • The national breakpoint committees have committed themselves to implementing EUCAST breakpoints – which means that anyone using the national European systems will gradually adhere to the EUCAST breakpoint system • Breakpoints as presented in EUCAST tables can be directly applied to MIC distributions (local and national surveillance, EARSS, etc) • Systems for automated susceptibility testing can be set up with EUCAST MIC breakpoints.

EUCAST wild type MIC distributions and epidemiological cut-off values – the concept JAC 2003; 52: 145 -148 EUCAST developed the concept of antimicrobial wild type MIC distributions and epidemiological cut-off values (JAC 52: 145 -148, 2003). Software was created to receive and display large volumes of MIC data for bacteria and fungi over the Internet. It is freely available at http: //www. eucast. org. Distributions are displayed in an aggregated format. Tables and graphs show the part of the MIC distribution which, when EUCAST defines the ”epdemiological cut-off value”, is defined as the ”wild type distribution”. The epidemiological cut-off value separating microorganisms without (wild type) and with acquired or mutational resistance (non-wild type) and clinical breakpoints are, if defined, shown on the bottom line of the graph. The epidemiological cut-off value (left hand lower corner) is shown as WT≤ X mg/L. The clinical breakpoints (right hand lower corner) are shown as S≤ Y mg/L and R> Z mg/L.

EUCAST wild type MIC distributions and epidemiological cut-off values – methods and data Origin of MIC data Each distribution is comprised of aggregated MIC data including individual MIC distributions from - publications in international journals - breakpoint committees - antimicrobial surveillance systems such as EARSS, SENTRY, the Alexander Project - pharmaceutical companies and susceptibility testing device manufacturers. Thus, unless otherwise specifically stated, distributions include results obtained with different methods. These methods do not give exactly the same results but the results rarely vary by more than one doubling dilution step. In this way the aggregated EUCAST MIC distributions contain the random variation between different investigators and the systematic variation seen between different methods. Origin of the organisms included in the MIC distributions The data are from tests on bacteria and fungi collected from man and animals, of any geographic origin and over a wide timeframe. MIC methods represented Species-specific distributions of MIC values collected from all over the world are included in the database. The distributions shown represent full range MIC values determined with methods described by EUCAST, BSAC (UK), CA-SFM (France), CRG (The Netherlands), DIN (Germany), NCCLS (USA), NWGA (Norway), and SRGA (Sweden) or methods calibrated to these methods (eg. commercial methods which give full range MIC values).

EUCAST wild type MIC distributions – templates for calibration of MIC determinations Exclusion of data All submitted full-range MIC distributions have been accepted. There has been no systematic exclusion of data from one contributor or from one method. The contributions are screened by the EUCAST Steering Committee and less than 10% have been excluded from the aggregated distributions. However, all data are held in the database and are accessible to the Steering Committee. The most common reason for exclusion has been that the data were not full-range MICs so that a significant proportion of MICs were outside the tested range. Laboratories which cannot fit their own MIC data to the EUCAST reference distribution should look into the following possibilities: • The method used for MIC determination in the local set of data is not adequately calibrated, • The species identification is incomplete, • There are too few determinations to allow identification of the part of the distribution that constitutes the wild type microorganisms. This usually corresponds to the four lowest dilution steps.

EUCAST wild type MIC distributions – why are only the MICs of wild type microorganisms displayed? The distributions consist of MIC-values determined over 30 years or more. Resistance frequencies obtained through the aggregated MIC distributions would not be representative of current antimicrobial resistance frequencies and would be confusing and misleading. When the epidemiological cut-off value has been determined by the EUCAST Steering Committee it blocks display of the non-wild type microorganisms (red bars, upper figure) and shows only the part representing the wild type (lower fig).

EUCAST wild type MIC distributions – how to contribute data Everyone is invited to contribute data All who have full-range MIC data for bacteria or fungi are invited to contribute data as long as MICs are determined with an accepted standardised method, which should be named. Once entered on the database the data will not be identifiable as separate distributions but will help build the aggregate reference distributions. The biologically resistant (non-wild type) part of the distribution will be seen only by the EUCAST Steering Committee. Submitting data to the EUCAST database does not interfere with publication of data. Where can I get more information? Contact EUCAST – email addresses and information can be obtained through the EUCAST website at http: //www. eucast. org