Göran Kronvall projects

Microheterogeneity of antimicrobial susceptibility

Poster at ECCMID 2000 on Microheterogeneity of antibiotic susceptbility

Mini-review on Acridine Orange Staining in Clinical Microbiology

Acridine Orange Staining Review

Antimicrobial resistance surveillance

New rapid method for antimicrobial resistance surveillance using faecal E.coli as the indicator bacterial species.

The rapid increase in antimicrobial resistance among bacterial pathogens poses a serious threat to the community, both in developed and developing countries. There is an urgent need for the continous monitoring of the resistance situation and also for methods suitable for this task.

In the ANTRES project we have developed a rapid screening method for resistance surveillance which will meet many requirements and is suitable for surveillance in both developed and developing countries. It was first set up for screening among children (1,3) and then also in adults (2). It has already proved valuable in the ANTRES project (4).

In essence, feacal samples are streaked on a growth medium which favours Enterobacteriaceae and particularly E.coli which also gives a colour reaction on the medium (McConkey agar). After inoculation antibiotic discs are put on the agar surface and the plates incubated over night at 37ºC.

After incubation the plates are read in two ways. First, the zone diameters around the antibiotic discs are measured and interpreted for susceptibility. The zone diameter breakpoints might have to be adjusted compared to recommendations using NRI. This reading gives the susceptibility of the dominant E.coli clone found in that individual faecal specimen. Next, all growth inside zones of inhibition is recorded and this gives the value for the total resistance in all clones of E.coli. Results have indicated that the measure for total resistance might be extended to include also Enterobacteriaceae other than E.coli (5).

Evaluation of a rapid screening method for detection of antimicrobial resistance in the commensal microbiota of the gut.
Bartoloni A, Benedetti M, Pallecchi L, Larsson M, Mantella A, Strohmeyer M, et al
Trans. R. Soc. Trop. Med. Hyg. 2006 Feb;100(2):119-25

Extended antimicrobial resistance screening of the dominant faecal Escherichia coli and of rare resistant clones.
Kronvall G, Larsson M, Borén C, Kahlmeter G, Bartoloni A, Rossolini G, et al
Int. J. Antimicrob. Agents 2005 Dec;26(6):473-8

Evaluation of a rapid screening method for detection of antimicrobial resistance in the commensal microbiota of the gut.
Bartoloni A, Benedetti M, Pallecchi L, Larsson M, Mantella A, Strohmeyer M, et al
Trans. R. Soc. Trop. Med. Hyg. 2006 Feb;100(2):119-25

Multidrug-resistant commensal Escherichia coli in children, Peru and Bolivia.
Bartoloni A, Pallecchi L, Benedetti M, Fernandez C, Vallejos Y, Guzman E, et al
Emerging Infect. Dis. 2006 Jun;12(6):907-13

Multidrug-resistant commensal Escherichia coli in children, Peru and Bolivia.
Bartoloni A, Pallecchi L, Benedetti M, Fernandez C, Vallejos Y, Guzman E, et al
Emerging Infect. Dis. 2006 Jun;12(6):907-13

Grape,M.; Kristiansson,C.; Bartoloni,A.; Gotuzzo,E.; Kronvall,G.
Rapid Resistance Screening Method for Detection of Resistance Markers in Dominant Flora and Rare Clones of Faecal Samples
45rd ICAAC Abstracts (Abstract ), xxx. 2005.

Resistance to trimethoprim and to sulphonamides (Malin Grape)

Earlier studies on antimicrobial resistance in bacterial pathogens at a university hospital, Karolinska, Stockholm, in a twelve year perspective, as well as at a total of twelve hospitals in Sweden revealed significant trends. (see below)

Earlier studies

Long-term antibiotic resistance surveillance of gram-negative pathogens suggests that temporal trends can be used as a resistance

Different trends in antibiotic resistance rates at a university teaching hospital.

Antibiotic use and Escherichia coli resistance trends for quinolones and cotrimoxazole in Sweden.

These results prompted among other things an in-depth investigation regarding the molecular mechanisms for trimethoprim- and sulphonamide resistance in Gram negative pathogens.

1. New dfr2 gene as a single gene cassette in a class 1 integron from a trimethoprim resistant Escherichia coli isolate.

2. Sulphonamide resistance gene sul3 found in Escherichia coli isolates from human sources.

3. Integrons and gene cassettes in clinical isolates of co-trimoxazole-resistant Gram-negative bacteria.

4. Integrons and multidrug resistance among Escherichia coli causing community-acquired urinary tract infection in southern Indi

5. Acquired sulphonamide resistance genes in faecal Escherichia coli from healthy children in Bolivia and Peru.

6. Extended antimicrobial resistance screening of the dominant faecal Escherichia coli and of rare resistant clones

Antimicrobial susceptibility testing using disc diffusion methods (Göran Kronvall)

Species-related interpretive breakpoints were introduced on a broad scale at the clinical microbiology laboratory in Lund, Sweden, in the late 1970:ies by Göran Kronvall, at that time responsible for clinical bacteriology laboratory services in Lund.

Antimicrobial susceptibility testing using disc diffusion methods, A.

(picture above) inhibition zone diameter values from disc diffusion tests were recorded manually by the laboratory technicians in 1977 in Lund, in order to analyze zone size distributions species-wise. The different locations of the wild-type populations for the two species, E.coli and P.mirabilis for ampicillin and nalidixic acid, respectively, illustrate the need for species-related interpretive breakpoints. This is even more evident when also other bacterial species are taken into account.

Historical aspects and the concept of species-related interpretive zone diameter breakpoints

Antimicrobial susceptibility testing in the clinical laboratory is most often performed using the disc diffusion method. This method was originally standardized by Bauer et al. (the so called Kirby-Bauer method), and by Ericsson & Sherris. (see the links below)

Bauer et al

Ericsson & Sherris.

A well-known authority on antimicrobial susceptibility testing, Dr Ronald Jones, USA, wrote: " ... the disk diffusion test continues to be the most versatile, broadly accurate, and reproducible AST test we can use in the clinical microbiology laboratory." [Jones, R.N. 1992. Clinical Microbiology Newsletter 14:33-37]

The Ericsson & Sherris method was adopted early as the method of choice for susceptibility testing in Sweden, but in those days the susceptibility categories were called 1, 2, 3, and 4. In 1978 this was changed by the Swedish reference group to the more international S, I, and R categories.

In 1976 we had started to plot zone diameter histograms species-wise for the different antibiotics in Lund. This revealed homogeneous populations of isolates representing wild type susceptible isolates as well as strains with various degrees of resistance. The new zone breakpointsfrom SRGA in 1978 did not always fit individual species so exceptions with special breakpoints were introduced locally in Lund. We concluded that a proper asignment of a susceptibility category of an isolate as S, susceptible, I, intermediate susceptible, or R, resistant, to a given antibiotic required knowledge about the distribution of zone diameters for that combination of bacterial species and antibiotic. The concept of species-related interpretive zone breakpoints (see the link below) for SIR categorization was introduced. The basic rule is that an interpretive zone breakpoint should never cut a homogeneous population of isolates in a zone histogram for a given combination of bacterial species and antibiotic, be it a fully susceptible population or one with decreased susceptibility.

species-related interpretive zone breakpoints

Comparative quality control studies in the 1980:ies were performed by the Lund team involving all laboratories in Sweden. In two of the studies, Gunnar Kahlmeter, then a young doctor in the lab, was a team member. All these studies showed very clearly that there were interpretive problems using general breakpoints. The Swedish Reference Group for Antibiotics, SRGA, then appointed Gunnar Kahlmeter in 1986 to chair a methodology subgroup to further analyze the experiences gained in Lund and to produce experimental material from several other laboratories to test the concept of species-related interpretive breakpoints, a concept which was later adopted as a national standard in Sweden (see link below). For full details of the SRGA standard, interpretive breakpoints, populations of isolates with both MIC and zone diameter distributions, and other valuable information FREE OF CHARGE, both in English and in Swedish, see the SRGA home page (see link below).

National standard in Sweden

SRGA home page

Species-specific interpretation was not new, but had earlier only been suggested for occasional combinations of species and antibiotics, e.g. carbenicillin and P.aeruginosa (see link below) in 1974 and chloramphenicol and P.mirabilis (see link below) in 1980. Our extensive histogram studies of the years showed the need for a more general concept of species-related interpretive breakpoints. It should also be mentioned that Thomas F. O'Brien had pioneered studies of zone diameter histograms already in the 1960:ies.

carbenicillin and P.aeruginosa

chloramphenicol and P.mirabilis

All histogram studies so far had used simple visual inspection with comparison of interpretive breakpoints to evaluate possible problems. We felt the need for a more objective method to analyze histograms species-wise and then came up with "single-strain regression analysis" to produce species-specific regression lines

How we started using species-related breakpoints.

Antimicrobial susceptibility testing using disc diffusion methods, B.

Single strain regression analysis, SRA

One reason for using species-related interpretive breakpoints was the fact that different species often gave different regression lines, i.e. correlations between MIC values of isolates and their inhibition zone diameter values in disc diffusion tests.

Calculating species-specific regression lines is possible using Single-strain Regression Analysis, SRA. (see link below)

Single-strain Regression Analysis, SRA.

SRA was originally developed in order to provide a tool for defining interpretive zone diameter values corresponding to the MIC-limits for susceptibility issued by reference authorities.

SRA is a formula describing the relation between disc content, MIC value and inhibition zone diameter and is based on the original equations developed during the 1950, as exemplified by the work of Cooper (see link below). The difference between the SRA equation (see link below) and the regular regression line equation (see link below)in comparison with the original ones is that the SRA equation retains the amount of antibiotic in the diffusion source as a variable. This adds an interesting feature. When the MIC value of an isolate is known (for instance an official reference or control strain) then a series of disc contents producing different inhibition zone diameters will make it possible to calculate the constants of the equation. Thereby, the regression line is also defined. Moreover, this regression line was determined using only one single strain and if this isolate behaves like most isolates of that species it will also represent the regression line for that particular bacterial species.


SRA equation

regular regression line equation

SRA therefore makes it possible to determine the regression lines for different species and thereby also to set correct zone breakpoints for different species corresponding to the MIC limits for susceptibility.

For an explanation of SRA, see Kronvall & Ringertz, 1991.

Examples of breakpoint calculations using SRA:

Cephalothin, 1984

Doxycycline, 1985

Ceftazidime, 1985

Chloramphenicol, 1988

Norfloxacin , 1989

Ciprofloxacin, 1989

Gentamicin, 1981

Enterococci, 1991

Trovafloxacin, 1999

Anaerobes, 2000

Fusidic acid and C.difficile, 2000

Staphylococcus aureus, 2002

SRA was also applied to cefoperazone results (see link below) from other authors (see link below) to calculate interpretive breakpoints corresponding to MIC limits for different species, not possible using regular regression analysis.

cefoperazone results

other authors

The examples given above seem to indicate that calibration of the disc diffusion test might be possible in the individual laboratory. Such procedures are everyday tasks for the clinical chemist, but the word calibration is not yet in the vocabulary of the clinical microbiologist,

(see 'Antimicrobial susceptibility testing using disc diffusion methods, C.')

Antimicrobial susceptibility testing using disc diffusion methods, C

Calibration of the disc diffusion test and determination of optimal disc content for routine antimicrobial susceptibility testing

From the SRA experience (see link below) it was clear that interpretive zone diameter breakpoints could be determined corresponding to MIC limits set by reference authorities for individual bacterial species using SRA, single strain regression analysis. This means, in fact, that the disc diffusion test CAN BE CALIBRATED, not only for drug-bug combinations but also for individual laboratories. Calibration procedures have been routine tasks in clinical chemistry laboratories for decades.

SRA experience

Examples of calibration procedures using SRA will be given.

When a clinical microbiology laboratory in Tartu, Estonia, wanted to set up fusidic acid disc diffusion tests (see link below) for isolated Staphylococcus aureus strains, they used the NCCLS standard which lacked zone breakpoints for this antibiotic. On the other hand, SRGA issued both MIC limits and zone interpretive breakpoints, the latter though for a disc content of 50 µg. In Tartu they wanted to use a lower disc concent, 10 µg. The solution was to calculate the new zone breakpoints using SRA.

fusidic acid disc diffusion tests

Another example provided zone breakpoints for some anaerobic species and trovafloxacin (see link below) susceptibility. Although this fluoroquinolone is not available on the market, the procedure can be applied to other new antimicrobials at some stage in the clinical testing.


There are often arguments regarding disc contents for routine clinical laboratory disc testing. The power of SRA calculations can actually provide a new definition of the optimal disc content for diffusion tests:

"The lowest disc content of an antibiotic which will distinguish resistant strains of any bacterial species from strains of the intermediate or susceptible category."

This is possible to determine using SRA as was shown for fusidic acid and S.aureus , for trovafloxacin and aerobic pathogens , and for trovafloxacin and anaerobes (see links below).

fusidic acid and S.aureus

trovafloxacin and aerobic pathogens

trovafloxacin and anaerobes

Calibration was a valuable feature of SRA, the equation obtained from original formulae describing the disc diffusion test. A further extension of the SRA equation leads to the so called M-test, where you can determine the MIC value of an isolate using several disc contents,

The only definition so far of setting the optimal disc content in disc diffusion testing is presented.

Antimicrobial susceptibility testing using disc diffusion methods, D

M-test for estimation of antimicrobial MIC values of clinical isolates

The SRA equation can be modified in a way which will permit the calculation of MIC-values (see link below) for individual isolates, the so called M-test (see link below). This is due to the fact that the inhibition zone size extrapolated to zero gives a concentration value from the regression line which corresponds to the 'critical concentration' of the isolate. The principle has been described also by others, first by Shannon et al. (see link below) in 1975, who determined the penicillin susceptibility of gonococcal isolates using manual, graphic plots. Drugeon et al. (see link below) in 1987 determined the 'critical concentration' of cefotaxime and ceftriaxone for 91 bacterial isolates. In 1994 Delignette-Muller and Flandrois (see link below) described the ICD, Inhibitory Concentration in Diffusion, determined for three aminoglycosides in 70 isolates using the same principle.

calculation of MIC-values


Shannon et al.

Drugeon et al.

Delignette-Muller and Flandrois

In all these methods a series of different concentrations of the antibiotic in diffusion sources, usually paper discs, is applied. The inhibition measurements are then used to solve the equation and to calculate the 'critical concentration', or Q-zero, at zone zero. In the M-test equation the MIC value is obtained by multiplying the Q-zero value with a conversion factor. This factor is often 2, but can vary with drug and bug.

The M-test has also been used successfully for fluconazole and voriconazole susceptibility testing of Candida species (see link below).

susceptibility testing of Candida

Let us now return to inhibition zone diameter histograms. The disc method is the most common test for antimicrobial susceptibility and such results are available all over the world. How can this untapped source of susceptibility results be used for surveillance? Well, there is a method to obtain an internal calibration for comparative purposes, so called normalized resistance interpretation, NRI (see 'Antimicrobial susceptibility testing using disc diffusion methods, E.')

Antimicrobial susceptibility testing using disc diffusion methods, E.

NRI - Normalized Resistance Interpretation

Let us ask ourselves: Whick part of an inhibition zone diameter or MIC distribution (species-wise) for a given antimicrobial is unaffected by the development of resistance?

Answer: The high-zone (or the low MIC) side of the most susceptible population of strains, representing the wild-type population. When resistance occurs in an isolate of that species, the position of that isolate in the distribution changes to lower zone sizes or higher MIC values.

So, if we can use the upper zone size slope for a reconstruction of the whole wild-type distribution, then we have obtained an internal calibrator which will enable us to compare results from anywhere, from any laboratory in the world.

This can be done using the Normalized Resistance Interpretation (see link below) method, NRI.

Normalized Resistance Interpretation method, NRI

A summary of the procedure (see link below) is presented in IJAA.

A detailed analysis of parameter setting (see link below) for the NRI calculations was performed by Joneberg et al.

summary of the procedure

analysis of parameter setting

MIC distributions with regular double dilution steps provided too few points for solving the regression, but Etest results (see link below) with intermediate values included, were precise enough for NRI calculations to work. This was shown in studies of tigecycline susceptibility.

Etest results

A patent application for the NRI method has been submitted by Bioscand AB (International Patent Application WO 02/083935 A1).

Normalized Resistance Interpretation method, NRI.

The solution to standardized resistance surveys.