What Every Pharmacist Should Know about Antimicrobial Susceptibility Testing
Steph’s Note: Today, This week, we welcome back our favorite ID expert, Dev Chatterji, PharmD, BCPS, BCIDP. You may remember him from previous tl;dr posts like this one about pneumonia, this one about urinary tract infections, this one about cefidericol, or this one about how to be a successful resident. He’s back this time to share his brain when it comes to interpreting patients’ microbiology results. Tell us what we need to know, Dev!
Also - Be sure to check out our Antibiotic Pocket Guide to help put your newfound susceptibility knowledge into practice!
The clinical microbiology lab is an invaluable partner in providing key information that can be used to guide the selection of optimal antibiotic regimens for individual patients. In terms of direct relevance to the management of patients with infections, antimicrobial susceptibility testing (AST) is probably the single most important activity performed in the clinical microbiology laboratory.
Despite reviewing and anticipating AST results routinely in clinical practice, I don’t really recall receiving much didactic education on laboratory diagnosis and antibiotic sensitivity testing during pharmacy school. I was only required to take one microbiology class as a prerequisite before applying to pharmacy school, so you can imagine how well that information stuck with me. If there’s an S next to the antibiotic, it’s fine to use, but if it says R, then it’s not ok…easy enough, right?
Mic drop (no pun intended…)
Seriously though, that’s probably not too far off from how the majority of clinicians probably interpret microbiology susceptibility reports. However, as we get deeper into this era of multi-drug resistant organisms, it’s important to have a working understanding of how to interpret antimicrobial susceptibility reports and be familiar with some of the nuances that can drive drug selection.
Susceptibility Testing
Let’s start with a brief review of susceptibility testing. Most of us are familiar with the conventional identification of bacteria and susceptibility testing we learned about in school. This consists of gram staining followed by phenotypic susceptibility testing using ‘standard’ methods such as broth dilutions or disk diffusion.
The goal of these susceptibility tests is to provide a quantitative metric that can then be translated into a standardized interpretation modality. In the U.S., these standards come from the Food & Drug Administration (FDA). To create these interpretive standards, the FDA collaborates with standards development organizations such as the Clinical Laboratory Standards Institute (CLSI) or the U.S. or European Committees on Antimicrobial Susceptibility Testing (USCAST, EUCAST, respectively).
This quantitative metric is referred to as the MIC (minimum inhibitory concentration), which is defined as the lowest concentration of the antibiotic that inhibits growth of the organism. Examples of conventional methods used to determine the MIC are as follows:
Dilution
Involves preparing specific concentrations or dilutions of antibiotics in a growth medium and dispensing them in test tubes or a well plate. A standardized bacterial suspension is added to the test tubes or well plates and incubated for a defined time. The tubes or plates are subsequently examined for visible bacterial growth and the concentration of the first tube or well in which bacterial growth is absent is noted—this is the MIC.
A visualization of the dilution method. See how the lowest concentration of antibiotic that inhibits growth is 32 mcg/mL? Therefore the MIC for this sample is 32 mcg/mL. (Image)
Disc Diffusion (aka, the Kirby-Bauer Method)
Involves first placing an inoculum of bacteria onto the surface of an agar plate. Then disks containing defined amounts of antimicrobials (commercially prepared) are placed on the agar surface and incubated for a defined time (usually 24 hours) during which antibiotic can diffuse through the plate and form a circular concentration gradient around the disk. The ‘zones’ of growth inhibition around each disk are measured and the diameter is used as an indirect measurement of the MIC. The zone diameters are interpreted using CLSI guidelines to correlate with a MIC and determine susceptibility.
A visualization of the disk diffusion method. See how there are zones of growth inhibition around each letter-labeled disk of antibiotic? The size of the zone helps to determine the MIC. (Image)
Practically speaking, when microbiology labs are reporting susceptibility results on a large scale, automated systems are utilized. There are many commercial FDA-approved automated platforms available which enable more timely and efficient reporting of results and are much less labor intensive than manually performing the susceptibility testing methods I mentioned above.
These automated systems provide a MIC through the use of various algorithms and rules that can be programmed within the systems. With these automated systems, an MIC can be determined for a panel of antimicrobial agents simultaneously.
Interpretation of Susceptibility Results
Now that you have an understanding of how we obtain the quantitative metric from susceptibility testing, it’s important to understand how to use that metric in a clinical setting. Clinical breakpoints or MIC breakpoints are what microbiology labs use to qualitatively categorize bacteria and other microorganisms in order to help guide clinicians on potential agents to utilize. The breakpoint is the highest plasma concentration of the drug that can safely be achieved in a patient to define susceptibility to an organism. (They’re essentially MIC values that predict probability of treatment success.)
Clinical breakpoints are standardized for a specific organism and each antibiotic tested for that organism, not for an individual patient. On the other hand, the MIC is at the patient level. Susceptibility testing combines these 2 factors by performing tests on an individual patient’s cultures to determine an MIC for a particular antibiotic against a particular pathogen in that specific patient.
The patient’s MIC is compared to a predetermined standardized breakpoint and subsequently categorized. Standards development organizations such as CLSI or EUCAST determine what the clinical breakpoint for a specific antibiotic would be to a specific organism. To determine breakpoints, these standards development organizations use data such as PK/PD properties of the drug in question, resistance mechanisms, and other clinical data (dose, indication, etc.). Clinical breakpoints are universal with respect to the standards organization being used. The typical qualitative categories are:
Susceptible (S)
MIC is less than the determined breakpoint for the organism. At typical doses, the antibiotic would have a high probability of being effective.
Intermediate (I)
MIC is approaching or at the determined breakpoint. The antibiotic in question may be effective at higher doses (if safely attainable) at specific infection sites based on PK/PD of the antibiotic.
Resistant (R)
MIC is greater than the determined breakpoint. The antibiotic would be unlikely to achieve therapeutic success at safely attainable doses.
To better explain, check out the diagram below:
Imagine this is a snapshot of a susceptibility report for a patient with positive cultures for this organism above. For this patient, after performing susceptibility testing, you can see that Antibiotic A resulted in a MIC of 0.5 mcg/mL. And let’s say it has been determined (per CLSI standards) that the MIC breakpoint for Antibiotic A against this organism has been set at 2 mcg/mL. In this case, on the susceptibility report, the interpretation for Antibiotic A would therefore be reported as susceptible (S) since the MIC is well below the determined breakpoint and has a high probability of being effective in this patient.
Now how about a more practice-based example.
Here is what a sample microbiology susceptibility report may look like:
Note that on typical susceptibility reports from the microbiology lab (much like this one), all you are shown is the MIC of each of the test antibiotics for the pathogen identified and the associated interpretation. The clinical breakpoint is not listed on the report.
When interpreting a report and deciding which definitive therapy to go with, there are a number of things to consider with regards to antibiotic selection that go well beyond the susceptibility report. These include: site of infection, PK/PD of the drug in question, patient’s allergy history, patient’s organ function (e.g., renal function), and a number of other factors.
However, here are a couple basic rules of thumb for the susceptibility report itself:
Think beyond the “S”. Yes, generally, if the interpretation is “S”, you can rest assured the antibiotic will have a high probability of being effective against the pathogen. However, there are many things to think about beyond effectiveness against the pathogen…you must be mindful of the fact that:
Susceptibility testing is in vitro
Susceptibility testing does not always take into consideration items mentioned above such as individual patient factors, PK/PD of the drug at the site of infection, institution protocols, etc.
Do not compare MIC between drugs. The MIC cannot be interpreted based on absolute values—remember, each antibiotic has a range of concentrations tested and a determined breakpoint that’s individualized for the specific antibiotic against a specific pathogen.
In the example above, without taking into consideration all other drug or patient-specific factors, cefepime would not have a higher probability of success in this patient than piperacillin/tazobactam solely based on the fact that the MIC value for cefepime is 2 and the MIC for piperacillin/tazobactam is 4.
In the same vein, the agent with the lowest numerical MIC is not necessarily the agent you should choose.
Beyond the Susceptibility Report
For all the overachievers out there, in the spirit of “thinking beyond the S,“ I figured it was prudent to expand a bit further on how susceptibility data along with PK/PD can optimize antimicrobial selection and dosing. (By the way, I was kidding…all pharmacists are overachievers.)
We discussed above the attainment of the MIC and how the MIC is used to guide clinicians regarding the activity and potential efficacy of a given antimicrobial agent against a given pathogen. However, the MIC isn’t the whole story! It gives somewhat superficial information regarding potential antimicrobial effect, and relying solely on the MIC and the subsequent qualitative categorization can have several limitations—especially when you are dealing with difficult to treat pathogens.
One of the most important considerations in determining the efficacy of an antimicrobial agent is the relationship between the MIC of an antimicrobial for the organism in question and the exposure of the organism to that given antimicrobial. Exposure of the organism to the agent in the patient is dependent on the dose given and the pharmacokinetic/pharmacodynamic (PK/PD) properties of the drug.
Remember, the PK of an antimicrobial characterizes drug exposure and concentration per unit time; the PD characterizes the drug response as a function of the concentration-time profile. The three most common PK/PD indices used to predict an antimicrobial’s efficacy are:
Time > MIC
This is the percentage of time that free drug concentrations exceed the MIC during a dosing interval.
Cmax or Peak:MIC
This is the ratio of maximum free drug concentration to the MIC.
AUC/MIC
This is the ratio of free drug AUC over a 24-hour period to the MIC.
For more enhanced background reading on these PK/PD concepts (if tldr just isn’t quite enough for your voracious knowledge appetite), I encourage you to check out these articles.
Anyhow, the rationale for being familiar with these PK/PD indices is because knowing which PK/PD index correlates with the best response for a particular antimicrobial allows for optimization of dosing and ensures a higher probability of successful treatment. For example, the efficacy of aminoglycosides is best correlated with optimizing the Cmax: MIC ratio. Therefore, utilizing the extended interval dosing strategy is the dosing modality to get the ‘best bang for your buck’ with that class of antimicrobials.
The table below lists examples of commonly used antimicrobials and their corresponding pharmacodynamic exposure goal:
Another example of a tool to help guide optimal dosing is known as Monte Carlo simulation. Monte Carlo simulation is a mathematical modeling tool that uses some real patient data to simulate a larger patient population to test different dosing schemes and their likelihood of achieving a particular pharmacodynamic target. For example, one prototypical example of a impactful Monte Carlo simulation is the strategy to dose ß-lactams as continuous or prolonged infusions.
Translating Antimicrobial Susceptibility Testing Results to Patient Care
Using this information to provide patient-centered pharmacotherapy is easier said than done given all the complexities involved. However, understanding the basics of susceptibility testing combined with being familiar with PK/PD nuances and available data on optimizing antimicrobial dosing puts us pharmacists in a key spot for driving medication management in these situations. A recent example where I was not only able to help guide therapy, but also debunk some misconceptions on how to interpret a susceptibility report, went as follows:
Patient with a complicated medical history, including malignancy, was admitted to the ICU with respiratory failure. He had a subsequently complicated hospital course requiring tracheostomy. After he was transferred from the ICU, he developed hospital acquired pneumonia (HAP). His sputum culture resulted as such:
As you can see, there was heavy growth of Pseudomonas aeruginosa in the sputum, one of those ‘difficult to treat’ nosocomial gram-negative pathogens. This particular isolate happened to be relatively sensitive to all tested antimicrobials—so how do you hone in on the optimal agent to use AND optimally dose that agent?
Here’s how my thought process went…
Without even looking at the MICs at first, I considered the site of infection. It was a respiratory infection, so I knew we needed an agent with good tissue penetration. Drugs with a low volume of distribution, such as aminoglycosides, would not be optimal as monotherapy. Next, I knew I needed to review the patient’s medication allergies and assess his renal function. Using those factors to weight risk and benefit of available options, I then weighed that against the potential adverse effects and collateral damage of those possibilities.
And I still hadn’t even brought MICs into the mix yet!
Only after these initial thoughts did I begin to assess the microbiology results. Using the AST results, my discussion with the medical team was between a ß-lactam or quinolone since the MICs for ciprofloxacin and levofloxacin were very low. Given the potential adverse effects/collateral damage of a full quinolone course, the team agreed not to use this class up front. (We did consider the quinolones could have utility as oral options to complete a course of therapy if the patient got to the point of being ready to discharge.)
Given the patient didn’t have any allergy history, we agreed to pursue a ß-lactam. The anti-pseudomonal ‘work horses’ at my institution are cefepime, pip/tazo, and meropenem. So thinking through how to differentiate those options…
Meropenem
The CLSI susceptibility breakpoint is </= 2.
This patient’s MIC was < 0.5; therefore, there was a high likelihood of treatment success.
The team wanted to go with meropenem given it had the lowest absolute MIC. But REMEMBER, do not compare MICs between agents. (Of course, this case is a little trickier as you will see below.)
From a stewardship standpoint, nonetheless, it’s a good idea to try to utilize carbapenem-sparing options, if possible.
Cefepime
The CLSI susceptibility breakpoint is </= 8.
This patient’s MIC was 4; therefore, there was also a high likelihood of treatment success with this agent.
Piperacillin/Tazobactam
The CLSI susceptibility breakpoint is </= 16/4.
This patient’s MIC was 16/4.
Since the piperacillin/tazobactam MIC was right at the susceptibility cusp, does this mean we should absolutely have avoided this agent?
Not necessarily, there should still be a relatively strong probability of success. But one way we can tip the scales in our favor in this battle is to exploit piperacillin/tazobactam’s PK/PD characteristics by considering a prolonged infusion dosing strategy. This should further solidify likelihood of target attainment by optimizing time > MIC since B-lactam antibiotics are exhibit time-dependent effects on bacteria.
In this case, there is no “wrong” answer per se, but you do want to be able to optimize your options. We had a discussion about stewardship and using a carbapenem-sparing option, despite the less optimal piperacillin/tazobactam MIC. Cefepime would have been a fine option as well, but the team wanted to maintain anaerobic coverage given the patient had a tracheostomy.
In the end, we went with piperacillin/tazobactam, and I recommended the extended infusion dosing strategy to optimize our likelihood of hitting the pharmacodynamic target.
Final Thoughts
As you can see, there are a number of things to take into consideration when assessing microbiology data, and you should always combine the microbiology information with patient specific factors. And of course, always remember your PK/PD!
Remember the MIC and subsequent clinical breakpoint only give you ‘on the surface’ information regarding antimicrobial effect—using the MIC alone to make your therapeutic decisions is doing a disservice as you are taking all those complex drug-bug interactions and putting them into a summative qualitative value (S, I, or R). Use the MIC and the antimicrobial susceptibility test results as one piece of the clinical puzzle, but always consider pathogen and patient-specific factors.