automation/semi-automated methods and antibiotic ... · pdf fileautomation/semi-automated...

Automation/Semi-Automated methods and Antibiotic

Susceptibility Testing – Efficiency and Pitfalls

Tom Olma Acting Pathology West Network Microbiology Discipline Scientist, In-charge of

General Microbiology Unit, CIDMLS, ICPMR, Westmead, Pathology West

Unable to reliably predict susceptibility

Keep people in jobs

Specific therapy to avoid resistance

Give profits to drug companies

To treat infections

Reduce pain and suffering

Save lives

Epidemiology

Why do we do susceptibility testing?

Time taken to perform Manual

Instrument based

Automated!

Interpretation system Discriminate populations S and R “break points”

Cost: Low consumable high labour

High consumable low labour

Buy in

Auto validation/reporting

Is it efficient?

Treat patients When need it!!

Prevent infections “don’t worry about the shot Gun…give the red pill and nuke

it!!!!”

Interpretation systems CLSI. EUCAST, CDS, etc Pharmokinetics Pharmodynamics Sounds good to me!

Which v’s witch doctor

Standardisation Reproducible Is it a science? ~~~~~~~Measurement of uncertainty! Antibiotic Stewardship

Is it effective?

And everywhere you look there is resistance!

Is there something intrinsically wrong with what we are doing?

Are we so innocent?

General failure to have an early warning system that alerts that resistance is occurring rather than has occurred! The lag between awareness and action

Biological Resistance v’s Clinical Resistance

Antibiotic Susceptibility Testing – Efficiency and Pitfalls

EUCAST

Eucast susceptibility testing based on wild type isolates and applying pharmokinetics and pharmodynamics principles.

Data driven

NCCLS now CLSI has had a dramatic rethink

Game Changer!!

Phenotypic test methods: based on antimicrobial activity and breakpoints MIC determination (broth, agar, gradient diffusion)

Disk diffusion (BSAC, CA-SFM, CLSI, SRGA, EUCAST...)

Automated systems (Vitek, Phoenix, Microscan, Sensititre, ...)

Genotypic test methods: based on the detection of a resistance gene or its product mecA, vanA, vanB…

PBP2a, β-lactamase detection

By deduction–“expert rules” If mecA-positive, then report beta-lactam antibiotics as R

If erythromycin-R, then report azithro-and clarithromycin as R

If know ID predict intrinsic resistance

Methods for antimicrobial susceptibility testing

By products:

Enzymes – strip methods, Antigen detection

Proteins – Maldi-Tof

Reflects mechanism

Many mechanisms may be involved and need multiple detectors unless broad approach taken –phenotypic methods.

Methods for antimicrobial susceptibility testing

Main objectives

• To produce antimicrobial susceptibility testing (AST) results in a automated (or semi-automated) mode

• To standardize AST avoiding uncontrolled differences

• Offer AST in a shorter period of time than manual methods

• To interpret AST results (clinical categorization / interpretation)

• To make it more efficient and easier to perform and interpret

• Reduce errors

• INTERFACE!!

• Do we reduce cost?

Antimicrobial susceptibility automatic systems

MIC based automated systems

Instrument based Disk Systems

Adagio

BIOMIC

Antimicrobial susceptibility automated systems

None of the current automated susceptibility testing

devices can be considered genuinely automated …

• Automated systems consist of devices with computer-assisted

incubation, reading, interpretation and reporting functions (NO HANDS)

• Semi-automated systems require off-line incubation*. The panels are

automatically read with computer-assisted interpretation and

reporting*manual loading of each panel into the system is required.

• Manual systems use commercial (eventually in-house) panels that are

read by laboratory personnel. Results are either recorded by hand or

manually entered into a computer for interpretation and reporting

All instruments have implemented computer programs

Only for validated isolates

Only for Rapid Growers

Lot of exceptions/issues across board and need to be careful of their testing capabilities inducible resistance

Cannot report Tazo for Serratia marcescens needs to be retested using another method

If Pseud Resistant to Timentin and Tazo – need to test for Tazo separately

Cannot report meropenem for Aeromonas spp.

Proteus mirabilis susceptible to ampicillin and augmentin yet resistant to cefazolin – on retesting cephazolin susceptible

Beta lactamase testing for penicillin – MIC to penicillin not reliable when susceptible.

To Panacea or to sea!

Significant Coagulase negative staphylococci – when resistant to oxacillin reliable indication for fluclox resistant when susceptible should be tested for Mec A low level expression.

Fail to detect Heterogeneous resistance in many organisms.

Heterogeneous resistant Van A Enterocococci – machine says were susceptible – Etest showed colonies within the zone of inhibition PCR .

Emphasises the limitations and need to work with companies to resolve them

Limitations continued…..

MALDI-TOF

Game Changer!!

Rapid Low cost Entry level

? Susceptibility methods

ID used to predict intrinsic resistance of certain organisms


Most automatic susceptibility testing devices have incorporated …

– Interface connections with laboratory information systems (LIS)

– Quality control computer programs

– Computer programs or expert systems: • “antibiogram interpretive reading”: to interpret phenotypes and

infer resistance phenotypes

• “expert rules”: to perform actions based in clinical evidences and resistance mechanisms knowledge in response to specific antimicrobial susceptibility test results

– Programs to manage results for epidemiological purposes


Classification • MIC based systems

– agar dilution (no longer exists!)

– Broth dilution (specialised)

– microdilution: Microscan, Sensititre, Phoenix, …

– growth curves: VITEK legacy, VITEK2

• Disc diffusion based systems – -BIOMIC System

– --ADAJIO System

• ……

Data obtained from Evangelista & Truant. In: Commercial methods in Clinical Microbiology. 2002

Adapted from presentación by Dr. Rafael Cantón Antimicrobial susceptibility testing –practical

implementation of the EUCAST breakpoints and methods 20thECCMID, Vienna, Austria, 10 –13 April, 2010

• Lower ranges of concentrations are needed (EUCAST break points are mostly lower than CLSI)

– instability of certain antibiotics might affect accuracy (essential and categorical agreements)

• carbapenems, β-lactam-β-lactamase inhibitor combinations, …

– major discrepancies (false resistance) could be observed, particularly with isolates expressing low level resistance mechanisms

Issues with EUCAST breakpoint implementation


• S and R breakpoints can be ...-

– too close (essential and categorical agreements can be affected)

e.g. ciprofloxacin and enterobacteriaceae (S≤0.5 / R >1)

vancomycin and staphylococci (S≤2 / R >2)

– too separate (a wider concentration range is needed in the panel)

• e.g. aztreonam and P. aeruginosa (S≤1 / R >16)*

*The R breakpoint was increased from 8 to 16 mg/L to avoid dividing the wild type MIC distribution. The R breakpoint relates to high dose therapy. The S breakpoint is set to ensure that wild type isolates are reported I.


• A desirable attribute...

... to include drug concentrations equal to ECOFFs*

allowing detection of wild type organisms (no-R mechanism)

*epidemiological cut off values

A philosophical and technical change...

... breakpoints are interpreted and expressed differently

Where to now?

No Genuine automation after 60 years !

Total Laboratory Automation

Game Changers!

SensidiscA Maldi-TofA

Back to the future

Automated Disk diffusion testing

“Look Ma….. No hands ….. Maybe”

• Genuine automation a step closer!

• Cheaper to use but high entry level

• Low consumable and low labour cost

Questions

automation/semi-automated methods and antibiotic ... · pdf fileautomation/semi-automated...

Documents