David's New Book

Wednesday, June 27, 2012

David Livermore on EUCAST and Breakpoints




David Livermore University of East Anglia

As the EU’s grand confection – a single currency for diverging economies– draws towards its denouement, it comes as some relief to point to the success of a less ambitious but more successful harmonisation: EUCAST. This has brought useful agreement, provided a little employment and may even spread a healthy contagion beyond Europe’s borders… 

Specifically, EUCAST – the European Committee on Antimicrobial Susceptibility Testing has two major achievements to its credit:

First, it has brought a harmonisation of breakpoints across Europe so that single definitions of susceptible, intermediate and resistance are now accepted throughout the continent, whereas previously there were separate national committees with different breakpoints in Norway, Sweden, the UK, France and Germany, whilst most of the southern Europe followed CLSI.   Despite slow incorporation onto certain automated testing systems, EUCAST values are increasingly adopted for day-to-day laboratory testing

Second, and in contrast to the protracted territorial squabble between CLSI and the FDA, EUCAST has reached a Memorandum of Understanding with the European Medicines Agency – the drug regulator – and serves as their advisor on breakpoints.  Decisions on indications and pathogen spectrum remain firmly with the Agency.  This agreement has functioned for 5 years, through the licensing of tigecycline, daptomycin and telavancin, with ceftaroline pending.   It precludes the unfortunate situation, not unknown across the Atlantic, where pharmaceutical company X, with a generous breakpoint from the FDA, omits to talk with the CLSI, for fear of a lower value. It also prevents the problem where device manufacturer Y is obliged to comply with FDA breakpoints to have its product licensed, whilst its customers demand a panel that tests to CLSI’s criteria, which are becoming more conservative than the FDA’s.

This article isn’t to detail the organisational structure of EUCAST, which can be found on http://www.eucast.org/organization.   Briefly, it operates with funding from the European Centres for Disease Control (ECDC) and the European Society for Clinical Microbiology and Infectious Diseases (ESCMID). Its Steering Committee, which makes the primary decisions on breakpoints, comprises representatives from European national breakpoint committees whilst its General Committee, which reviews proposals, has wider representation from across EU and non-EU Europe, also Russia, Turkey and Australia. Representatives of pharmaceutical and device manufacturers are invited to present data for review by the Steering Committee when they seek to license new agents or indications, but they do not sit on EUCAST’s decision-making committees.

Breakpoint decisions are largely predicated on pharmacodynamics, as increasingly are those of the CLSI whilst the FDA puts more weight on clinical outcomes in relation to MIC.  These pharmacodynamic breakpoints may be adjusted by a dilution up or down, to ensure that they do not slice through the MIC distribution for wild-type isolates of major species groups. They may also be adjusted on the basis of clinical evidence and experience. A category of ‘Insufficient evidence’ is noted for drugs that might be of interest against a pathogen but where clinical data are scant (e.g. daptomycin vs. enterococci), whilst low non-species-specific breakpoints are included for obscure species. Last, an epidemiological cut off (‘ECOFF’) is specified, defining the upper edge of the normal distribution of MICs for isolates without any diminution of susceptibility.  This corresponds to what others call a biological breakpoint, though EUCAST is at pains to avoid the term. ‘Intermediate’ is taken as ‘may respond at high dose’ and, for some marginal agents (e.g. macrolides against haemophili or ciprofloxacin against pneumococci) nearly all isolates count as intermediate.  This differs somewhat from CLSI, where intermediate is viewed more as a buffer to minimise errors between MIC and disc categorisations.   Once assigned – and ratified by EMA for agents  – EUCAST publishes breakpoints on its website, along with a brief rationale, a summary of which also appears in Clinical Microbiology and Infection as ESCMID’s journal of record.   Values are not set in stone (as with the FDA) and can be reviewed at the request of the manufacturer, professional societies or the regulator. There is a low bar to initiating a review, but strong justification is needed for a change.

Is the system perfect? Personally – and this comment applies as much to CLSI – I believe that the pendulum in defining resistance has swung too far from mechanisms towards pharmacodynamics, with excessive optimism in the precision of MICs.  Second, I find it incongruous that EUCAST breakpoints are allowed to split populations with common modes of resistance despite a determination not to split wild type distributions. Third, there is the unfortunate ‘loss’ (perhaps to be rectified in the future!) of the higher breakpoints used previously in the UK and elsewhere for urinary infections; as a result laboratories are suddenly finding 40% resistance to amoxicillin-clavulanate in urinary E. coli whereas they previously found 5-10%---- with no evidence that the extra ‘resistant’ isolates are associated with poor outcomes.  

But, these are lesser cavils; overall, the EUCAST system is a major improvement over the country-specific breakpoints that went before.   These never had much logic when bacteria and pharmacokinetics didn’t respect political boundaries (different currencies for countries with different economic cultures and traditions is quite another matter, though…).   Which begs the question as to whether EUCAST and CLSI breakpoints too will coalesce, especially as both organisations base their values primarily on pharmacodynamics. In the short term the answer seems to be ‘no’. Although both have recently reviewed cephalosporin and carbapenem breakpoints they’ve achieved answers a dilution or two apart.  What’s more, if both committees ever do agree completely, there will be an embarrassing little problem, for CLSI’s income derives from selling its Standards, whilst EUCAST’s guidance is available free gratis, over the internet.