Interpreting Sensitivity Reports

Paul Lewis

James W. Myers

INTRODUCTION

Since the discovery of sulfonamides and penicillin, clinicians have attempted to predict the success or failure of treatment with antimicrobial resistance testing (AST). One of the first tools utilized was the minimum inhibitory concentration (MIC).

MIC is defined as the lowest concentration of antibiotic that inhibits visible growth of bacteria. We now know that MIC alone is insufficient to guide antimicrobial selection. For example, a clinician who selects one agent because of a reported MIC of 2 over another agent with an MIC of 8 may not have selected the best agent. The agent with an MIC of 2 may only achieve plasma concentrations of 1, while the agent with MIC of 8 may maintain plasma concentration 5 times that MIC. For this reason, breakpoints are established. A breakpoint, simply put, is the drug concentration that separates organisms into resistant or susceptible categories. This takes into account the microbiologic data as well as pharmacodynamic and pharmacokinetic data to predict the likelihood of success. As our knowledge of the relationship between pharmacokinetics and pharmacodynamics continues to grow, so does the effect of antimicrobial susceptibility testing. Animal models and statistical or mathematical models such as the Monte Carlo simulation further assist in the prediction of success.

ANTIMICROBIAL SUSCEPTIBILITY TESTING

There are two categories of susceptibility testing: qualitative and quantitative.

1. Qualitative resistance testing such as disk diffusion will allow for a susceptible (S), intermediate (I), or resistant (R) interpretation. Qualitative testing is usually performed by Kirby-Bauer disk diffusion. While sufficient for less complicated conditions such as urinary tract infections, the lack of MIC data is often a limitation to qualitative testing.

2. Quantitative methods include broth macrodilution, broth microdilution, agar dilutions, E-test, and several automated systems, which yield the MIC of the antimicrobial to the organism tested. The dilutions are time consuming, cumbersome, and not practical for use in clinical practice. Oftentimes, they are used to validate other methods including automated methods.

Kirby-Bauer Disk Diffusion

A standard concentration of bacteria in a broth or saline solution is spread confluently and uniformly over a blood or Mueller-Hinton agar.
Antibiotic-impregnated disks are placed on the plate and then incubated.

The antibiotic diffuses out, establishing a concentration gradient.

Figure 52-1. Kirby-Bauer disk diffusion allows antibiotic-impregnated disks to create a zone of inhibition. The zones are measured in millimeters and are compared to standard references to determine susceptible (S), intermediate (I), or resistant (R).

At a certain point, the antibiotic concentration is insufficient to inhibit the growth of the bacteria forming a ring or zone of inhibition.
See Figure 52-1
The zone of inhibition is measured and recorded.
Based on predefined set points, the antimicrobial is classified as S, I, or R.
Advantages include simplicity, inexpensiveness, and customizability of antimicrobials tested.
Disadvantages include the lack of an MIC, being time consuming to the microbiologist, and a possibility of interpretation error by the microbiologist manually measuring the zone.

Dilution Testing

Macrobroth uses test tubes with serial dilutions of antibiotic concentrations.
Microbroth uses serial concentration in a well plate.
Agar dilution uses plates of serial concentrations.
All are inoculated with a standard concentration and incubated.
The MIC is determined by the highest concentration that contains no visible growth.
The advantage is reliability as it is the gold standard for MIC testing.
The disadvantage is that these methods are time consuming and too labor-intensive for clinical practice.

Epsilometer Test (E-test)

Uses an antibiotic-containing strip with a numeric concentration gradient.
Bacteria are plated in a uniform manner with strips placed in a radial manner and allowed to incubate.
Bacteria will grow in an elliptical pattern based on the concentration gradient.
The MIC is determined by the point at which the ellipse crosses the strip.
- See Figure 52-2
Since the strips can be placed on a wide range of media, fastidious organisms and other organisms requiring special growth media can be tested.
Strips may be supplemented with additional agents to detect certain resistance mechanisms.
Other advantages include simplicity and flexibility, useful with rarely tested antibiotics.
The disadvantage is that it is relatively expensive.

Spiral Gradient Endpoint Method

Uses an agar with increasing antibiotic concentrations spiraling out from the center of the plate.
An organism is streaked from the center of the plate outward.
The distance of growth determines susceptibility.
Although the method is expensive, inefficient, and rarely used, it may be valuable in detecting heteroresistance to vancomycin.

ESTABLISHING BREAKPOINTS

Breakpoints are established by a number of different organizations.
- Food and Drug Administration (FDA)
  - Establishes the baseline breakpoints at the time of drug approval
    
    Figure 52-2. The E-strip contains a gradient of concentrations. The MIC is determined by visualizing where the ellipse crosses the strip.
  - Pharmaceutical companies play a major role in the contribution of data.
  - Rarely reevaluates
- The Clinical and Laboratory Standards Institute (CLSI)
  - Formerly the National Committee for Clinical Laboratory Standards (NCCLS)
  - Nonprofit organization that develops consensus standards for the healthcare community.
  - Information regarding resistance mechanisms is also taken into consideration, often not available at the time of FDA approval.
- The European Committee on Antimicrobial Susceptibility Testing (EUCAST)
  - Component of the European Society of Clinical Microbiology and Infectious Diseases
  - Publishes clinical breakpoints and available for download at www.eucast.org.
- Both CLSI and EUCAST continuously update breakpoints on a yearly basis and as needed throughout the year.
  - The tables provide the MIC susceptibility breakpoint for Enterobacteriaceae (Table 52-1), Pseudomonas aeruginosa (Table 52-2), Acinetobacter spp. (Table 52-3), Staphylococcus aureus (Table 52-4), and Streptococcus pneumoniae (Table 52-5).

The establishment of clinical breakpoints is traditionally based on three key components: the distribution of the MIC’s, the pharmacokinetic/pharmacodynamic data, and the clinical outcomes

The distribution of MICs
- Wild-type population without acquired resistance
- Non-wild-type that may be harboring resistance patterns
Pharmacokinetic data of the antimicrobial agent
- In vitro drug characteristics
  - Such as stability, MIC, and zone diameters
  - Site of action—ribosome, cell wall, cell membrane, DNA
  - Model for killing—time > MIC, AUC:MIC ratio, or concentration dependent
  - Post-antibiotic effect—present if the drug continues to kill despite undetectable levels
  - Bacteriostatic versus bacteriocidal
- In vivo data such as pharmacokinetic properties and blood and tissue penetration bioavailability (if oral)
  - Serum peak concentration (Cmax)
  - Serum trough concentrations (Cmin)
  - Volume of distribution (Vd)
    - Tissue penetration
    - Cerebrospinal fluid penetration
    - Alveolar fluid penetration for pneumonia
    - Body fluids pertaining to the type of infection
  - Protein binding
  - Metabolism
  - Clearance
  - Elimination half-life
  - Special populations: renal and hepatic failure, obesity
The clinical outcomes
- Must have at least 500 isolates
- Clinical and microbiologic cure rates
- Clinical correlation is useful; however, extremely difficult to evaluate due to the multitude of confounders.

Table 52-1 Breakpoints for Enterobacteriaceae

Antibiotic	EUCAST	CLSI	Comments
Ampicillin	8	8	Used for amoxicillin
Ampicillin-sulbactam	8/4	8/4
Piperacillin-tazobactam	8/4	16/4
Ticarcillin-clavulanate	8	16/2
Cefazolin	R	2	Based on 2 g q8
Ceftriaxone	1	1	Based on 1 g q24
Ceftazidime	1	4	Based on 1 g q8
Cefepime	1	8	Based on 1 g q8 or 2 q12
Doripenem	1	1	Based on 500 mg q8
Ertapenem	0.5	2	Based on 1 q24
Imipenem-cilastatin	2	4	Low-level resistance common in Morganella, Proteus, and Providencia
Meropenem	2	4
Aztreonam	1	4	Based on 1 g q8
Levofloxacin	1	2
Gentamicin	2	4	Based on high-dose extended interval dosing
Tobramycin	2	4	Based on high-dose extended interval dosing
Amikacin	8	16	Based on high-dose extended interval dosing
Tigecycline	1	ND	Limited activity against Proteus, Providencia, and Morganella
Colistin	2	ND
Trimethoprim-sulfa-methoxazole	2/38	2/38
Fosfomycin	32	64	E. coli urinary tract isolates only
Nitrofurantoin	64	32	Urinary isolates only
R, should be reported as resistant without testing; ND, not defined.

Deterministic approach versus probabilistic approach to establishing breakpoints.

Deterministic approach
- Breakpoints established by adjusting the mean population pharmacokinetic parameters to the susceptibility of different pathogens
- Many of the pharmacokinetic/pharmacodynamic properties are not taken into consideration.
- Tends to yield potentially higher breakpoints than what is feasible in clinical practice

Probabilistic

A newer approach to breakpoint establishment that takes into account pharmacokinetic modeling
Attempts to what is likely to happen, not simply what could possibly happen

Tends to result in lower than previously established breakpoints

Table 52-2 Breakpoints for Pseudomonas aeruginosa

Antibiotic	EUCAST	CLSI	Comments
Piperacillin-tazobactam	16	16/4	Based on 4.5 g q6
Ticarcillin-clavulanate	16	16/2	Based on 3.1 g q6h
Cefepime	8	8	Based on 2 g q12 (CLSI) and q8 (EUCAST)
Ceftazidime	8	8	Based on 2 g q8
Doripenem	1	2	500 mg q8
Imipenem	4	2	Based on 1g q6 (EUCAST) and q8 (CLSI)
Meropenem	2	2	1 g q8
Aztreonam	1	8
Levofloxacin	1	2
Gentamicin	4	4	Based on high doses (5-7 mg/kg)
Tobramycin	4	4	Based on high doses (5-7 mg/kg)
Amikacin	8	16	Based on high doses (15-21 mg/kg)
Colistin	4	2

Table 52-3 Breakpoints for Acinetobacter spp.

Antibiotic	EUCAST	CLSI	Comments
Piperacillin-tazobactam	IE	16/4
Ticarcillin-clavulanate	IE	16/2
Ceftriaxone	R	8
Cefepime	R	8
Ceftazidime	R	8
Doripenem	1	NR
Imipenem	2	4	Based on 1 g q6
Meropenem	2	4
Levofloxacin	1	2
Gentamicin	4	4	Based on high doses (5-7 mg/kg)
Tobramycin	4	4	Based on high doses (5-7 mg/kg)
Amikacin	8	16	Based on high doses (15-21 mg/kg)
Colistin	2	2
Trimethoprim-sulfamethoxazole	2	2/38
Tigecycline	IE	NR
IE, insufficient evidence to define a breakpoint; R, should be reported as resistant without testing; NR, not reported.

Table 52-4 Breakpoints for Staphylococcus aureus

Antibiotic	EUCAST	CLSI	Comments
Penicillin	0.12	0.12	MIC <0.12 should be confirmed with disk diffusion.
Oxacillin	2	2
Ceftaroline	NR	NR	Resistant isolates not discovered
Vancomycin	2	2	MIC of 2 may have reduced response
Daptomycin	1	1
Tetracycline	1	4	If isolate susceptible, assume susceptible to doxycycline and minocycline
Minocycline	0.5	4	May be still be active even if resistant to tetracycline
Doxycycline	1	4	May be still be active even if resistant to tetracycline
Levofloxacin	1	1
Clindamycin	0.25	0.5	If D-test negative
Trimethoprim-sulfamethoxazole	2	2/38
Linezolid	4	4
Nitrofurantoin	NR	32
Telavancin	1	NR
Tigecycline	0.5	NR
NR, not reported.

The Monte Carlo simulation, for example, uses computer-generated model to create multiple scenarios.
The probability from low to high of achieving the desired outcome is reported.
Continuously updated as knowledge of pharmacokinetics and pharmacodynamics continues to increase and as new resistance mechanisms are introduced into the environment

Normalized MIC distributions
- MIC reported in terms of standard deviations away from the wild-type mode of distributions
- Takes into consideration that breakpoints are drug-species dependent and would allow for direct comparison between agents
- Would allow direct comparison between agents
- This may one day play an important role in objectifying antimicrobial susceptibility results.