Learning outcomes
By the end of this lesson you should be able to: define minimum inhibitory concentration and minimum bactericidal concentration and explain which one a clinician needs in endovascular infection; describe how disc diffusion, MIC determination and gradient strips are performed and standardised; interpret a susceptibility report using susceptible, intermediate and resistant categories; recognise the phenotypic signatures of MRSA, ESBL producers, carbapenemase producers and inducible clindamycin resistance; explain how breakpoints are set from microbiology, pharmacokinetics and clinical outcome data; and apply pharmacokinetic–pharmacodynamic principles to explain why a larger dose of a time-dependent drug is not the answer.
Prerequisites
Lessons 1 to 3. You need Gram stain behaviour, the envelope mechanisms of drug action and the genetic basis of resistance — otherwise a susceptibility report is a list of letters with no mechanism behind them.
The big idea
AST is the translation layer between the test tube and the patient. A susceptibility report answers one question: if I give this drug to this patient at a standard dose, is the drug likely to reach a concentration at the site of infection that exceeds what the organism needs. Everything about how the test is set up — the medium, the inoculum, the incubation, the interpretation — exists to make that prediction reliable.
Two facts bound the science, and both matter. First, 'susceptible' is not a property of the organism alone: it is a property of the organism, the drug, the dose, the route and the site. A MIC of 2 mg/L may be treatable in a urine with a highly concentrated drug, but not in cerebrospinal fluid with a poorly penetrating one. Second, the laboratory cannot individualise the dose — it reports against standard exposure, so the clinician must still adjust for organ function, renal replacement and source control.
The two quantitative parameters
The MIC is reported for the drug, not for the patient, and the adjacent figure that determines whether it is achievable is the drug's pharmacokinetics. A cephalosporin with a lung-tissue concentration fifty times its serum concentration can treat a pneumonia caused by an organism whose MIC would be unacceptable in endocarditis. This is why a report showing 'intermediate' is not a rejection of the drug — it is a statement that the drug may work if exposure can be increased, by dose, by frequency, or by treating a site where the drug concentrates.
How the tests are performed
| Method | Principle | Standardisation that must be honoured | What it reports |
|---|---|---|---|
| Disc diffusion (Kirby–Bauer) | A filter disc impregnated with one antibiotic concentration is placed on an inoculated agar plate; antibiotic diffuses radially and a zone of inhibition forms | Mueller–Hinton agar, a 0.5 McFarland inoculum, 35±2°C, 16–18 hours, and a CLSI- or EUCAST-defined disc content | A zone diameter in millimetres, converted to a category |
| Broth microdilution (MIC) | Two-fold dilutions of drug in broth are inoculated with a fixed density; the last clear well is the MIC | Mueller–Hinton broth, 0.5 McFarland, 35±2°C, 16–20 hours; controls each run | An MIC in mg/L, paired with a breakpoint |
| Agar dilution | The same dilutions, but poured as plates, so many isolates can be tested against one concentration | Same conditions; used for reference and for fastidious organisms | An MIC in mg/L |
| E-test and gradient strips | A strip carries a continuous exponential gradient of antibiotic; the ellipse of inhibition is read where its edge crosses the strip | Standard agar and inoculum; used for fastidious organisms and for research | An MIC read directly against the strip, in mg/L |
| Automated systems (Vitek, Phoenix, BD Phoenix) | Instrumented broth microdilution cards with rapid colourimetric or fluorometric readout | Same incubation and density logic, with instrument-specific algorithms | MICs and categories in 4–12 hours |
| Molecular detection (mecA, vanA, blaCTX-M, blaNDM) | PCR or hybridisation probes for the resistance determinant itself | Validated targets and controls; independent of growth | The gene, and therefore a phenotype, before culture is available |
NoteStandardisation is the whole reason AST can be compared between laboratories in London and Lagos. Mueller–Hinton agar is specified because it is low in inhibitors and free of the para-aminobenzoic acid that would neutralise sulphonamides; its batch is controlled for calcium and magnesium because cation content alters aminoglycoside activity against Pseudomonas; and a 0.5 McFarland inoculum — about 1.5 x 10^8 colony-forming units per millilitre — is specified because a heavier inoculum can make a susceptible organism read falsely resistant.
| Control strain | Why it is used | What an out-of-range result means |
|---|---|---|
| Escherichia coli ATCC 25922 | A defined wild-type susceptibility profile for a broad drug range | Disc content, medium or technique has drifted; the batch is void |
| Staphylococcus aureus ATCC 25923 | A wild-type Gram-positive profile, essential for beta-lactam discs | The medium's thymidine content or inoculum density is wrong |
| Pseudomonas aeruginosa ATCC 27853 | Tests susceptibility of a difficult Gram-negative aerobe | Cation content or incubation atmosphere is out of specification |
| Enterococcus faecalis ATCC 29212 | Controls glycopeptide and aminoglycoside MICs | The dilution series or the reading time is wrong |
Reading the report: categories, not absolutes
| Category | Definition | Clinical action | Common example |
|---|---|---|---|
| Susceptible (S) | The organism is inhibited at a concentration achievable with a normal dose at the site of infection | Standard therapy is expected to succeed | E. coli UTI with amoxicillin–clavulanate MIC 2 mg/L |
| Intermediate (I) | Inhibited only at a concentration achievable with a maximum tolerated dose, or with poor margin of error; or close to the breakpoint in a technically borderline range | Clinically usable if exposure can be increased (higher dose, longer infusion, site where the drug concentrates) | Ceftazidime MIC 8 mg/L in Pseudomonas, treating a urinary infection where the drug concentrates 50-fold |
| Resistant (R) | Inhibited only at a concentration not achievable by a safe dose at the site | Do not use; therapy failure is expected | MRSA reported resistant to all beta-lactams |
Watch out'Intermediate' is not a synonym for 'nearly works' — it is an explicit instruction to think about exposure. The old 'moderately susceptible' label was retired precisely because it invited inaction. An intermediate result demands a decision: increase exposure, treat a site where concentration is favourable, or change drug.
Breakpoints are derived from three converging data sets: the MIC distribution of the target species (the wild-type population must be above the cut-off, resistance mechanisms below it); pharmacokinetic–pharmacodynamic targets — what exposure the drug achieves in blood and tissue and what multiple of MIC is needed; and clinical outcome data from well-characterised patients. Standards bodies (CLSI in the United States, EUCAST in Europe) publish the resulting breakpoints, and each revision is a new clinical judgement, not merely a fresh arithmetic mean.
Phenotypic resistance signatures you must recognise
| Phenotype | How the laboratory shows it | What is happening genetically | Therapeutic consequence |
|---|---|---|---|
| MRSA | Cefoxitin disc resistant (or a mecA/PBP2a-positive result), whatever the oxacillin disc reads | mecA on the SCCmec cassette encoding PBP2a, with low affinity for all beta-lactams | No beta-lactam is usable; treat with a glycopeptide, linezolid, daptomycin or ceftaroline |
| ESBL producer | Clavulanate synergy — the zone expands towards a clavulanate-containing disc | Plasmid blaCTX-M hydrolyses third-generation cephalosporins and aztreonam | Avoid cephalosporins; a carbapenem is standard for serious infection |
| AmpC over-producer | Chromosomally encoded beta-lactamase inducible by clavulanate; susceptible to cefepime only marginally | Promoter or attenuator mutation producing constitutive AmpC | Cefepime with caution; carbapenem for invasive infection |
| Carbapenemase producer | Meropenem resistant; modified carbapenem inactivation test or a carbapenem disc enhancement positive | KPC, NDM, VIM, IMP or OXA-48 enzymes, frequently with co-resistance | Ceftazidime–avibactam or colistin; contact isolation and epidemiological investigation |
| MDR / XDR / PDR (Magiorakos definitions) | Non-susceptible to at least one, three or all agents in three or more classes | Accumulation of multiple acquired mechanisms | Combination therapy and expert consultation mandatory |
| Inducible clindamycin resistance | Erythromycin resistant, clindamycin appears susceptible in the standard disc test, but the D-test shows flattening of the clindamycin zone | erm gene methylating 50S ribosomal target, switched on by erythromycin | Never report clindamycin as usable on the disc test alone — use a D-test or report both resistant |
NoteThe modified Hodge test and the carbapenem inactivation test for carbapenemases have largely replaced the modified Hodge test in modern laboratories, and both exist because a carbapenem-resistant organism cannot be assumed to be the only resistance problem — many carbapenemase producers are simultaneously resistant to aminoglycosides and fluoroquinolones, so the full panel, not the single drug, determines therapy.
Pharmacokinetics and pharmacodynamics explain the MIC
Why does one drug need a high peak and another a long time above the MIC? Because the kill pattern differs. Time-dependent drugs (beta-lactams) kill by how long the concentration stays above the MIC, so the target is the percentage of the dosing interval above it; concentration-dependent drugs (aminoglycosides) kill by the peak achieved, so the target is a peak-to-MIC ratio; and exposure-dependent drugs (fluoroquinolones, vancomycin) are best predicted by the area under the curve divided by the MIC.
| Class | Kill pattern | PD target | Why simply raising a single dose does not help |
|---|---|---|---|
| Beta-lactams (penicillins, cephalosporins, carbapenems, monobactams) | Time-dependent | Free drug above the MIC for 40–70 per cent of the interval | A larger dose raises the concentration, but the kinetics already clear the drug below the MIC quickly; prolonged or continuous infusion, or shorter intervals, is the effective strategy |
| Aminoglycosides | Concentration-dependent, with a post-antibiotic effect | Peak/MIC ratio of 8–10 | Dividing the dose destroys the peak; the target is a single high daily dose with a trough near zero to protect the kidney |
| Fluoroquinolones | Exposure-dependent, concentration-augmented | AUC/MIC ratio above 125 for Gram-negatives | Raising the dose raises resistance selection pressure at the same time as AUC; toxicity at the tendon and QT interval limits it |
| Vancomycin | Exposure-dependent against Gram-positives | AUC/MIC of 400–600 (mg·h/L per MIC) | Peak-guided dosing has been replaced by AUC-guided dosing because a target peak correlates poorly with nephrotoxicity and outcome |
A patient with bacteraemia has a Pseudomonas aeruginosa isolate with a ceftazidime MIC of 4 mg/L, just above the susceptible breakpoint and 'intermediate' by the report. The intensive care team changes from bolus dosing to an extended infusion over three hours, so the concentration stays above the MIC for far more of the dosing interval. The organism was never changed; only the exposure was, and the conjugate pharmacology made an 'intermediate' result clinically workable.
Special situations the test cannot answer
| Situation | Why routine AST gives a wrong answer | What to do |
|---|---|---|
| Meticillin resistance reported as susceptible | Heteroresistant MRSA expresses PBP2a in a small subpopulation, so the standard incubation reads susceptible | Use cefoxitin or mecA/PBP2a detection; treat as MRSA |
| Biofilm-associated infection on a prosthetic joint | Organisms in a biofilm are 10–1000 times less susceptible than planktonic cells | AST is not interpretable; treatment is device removal plus prolonged combination therapy |
| Bactericidal therapy in endocarditis | A static MIC result gives no MBC, and some agents (for example a macrolide against pneumococcus) are bacteriostatic | Determine the MBC, or deliberately choose a combination that is reliably bactericidal |
| Polymicrobial infection | Each isolate is reported separately, and the drug chosen must work against all of them | Test every isolate and target the least susceptible member of the group |
| Slow-growing organism | The organism may not have reached log phase within the standard incubation | Extend incubation, or use a concentration-independent molecular test |
| Obligate anaerobe | Standardised methods for anaerobes are technically demanding and less widely validated | Agar dilution in an anaerobic atmosphere, with reference laboratory confirmation |
Susceptibility report -- worked reading
Organism: Escherichia coli (blood culture)
Amoxicillin-clavulanate MIC 4 mg/L S
Ceftriaxone MIC <= 1 mg/L S
Piperacillin-tazobactam MIC 4 mg/L S
Gentamicin MIC 2 mg/L S
Ciprofloxacin MIC 4 mg/L R
Trimethoprim MIC > 4 mg/L R
Interpretation: empiric ciprofloxacin must be stopped; the beta-lactam classes
above remain active and any of them can treat a urinary source at adequate
concentration. Gentamicin synergy with a beta-lactam is available if a
bactericidal effect is required.Common mistakes
Watch outTreating a bigger zone as 'better'. The test compares one isolate against one drug under standardised conditions; a larger zone only means a lower MIC relative to the breakpoint, and it cannot be compared between different discs or different drugs, because disk contents, diffusion properties and breakpoints all differ.
Watch outBelieving the MIC is the minimum concentration that kills. It is the minimum concentration that inhibits visible growth. The MBC, which is measured by subculturing to drug-free medium, is the number that speaks of killing — and only that number is interpretable in endocarditis or in a neutropenic patient.
Watch outChoosing a drug purely because the report says 'S'. The report ignores allergy, renal function, pregnancy, route of administration, penetration into the site and cost. A susceptible aminoglycoside for meningitis is useless because it does not cross the blood–brain barrier; a susceptible macrolide for typhoid is useless because Salmonella is an intracellular organism that the drug cannot reach effectively.
Check your understanding
A meningitis isolate reports an 'intermediate' result for ceftriaxone. Why is this result more worrying than the same wording on a urinary isolate?
Why is a 0.5 McFarland inoculum specified for disc diffusion rather than simply 'a visible amount of growth'?
A Staphylococcus aureus isolate is erythromycin resistant but clindamycin susceptible on the routine disc test. What must the laboratory do before reporting clindamycin as usable?
Multiple choice
Which definition is correct?
An enterococcus isolated from a heart valve shows an MIC of 8 mg/L to ampicillin. Why does the laboratory determine the MBC rather than simply reporting the MIC?
Why is a cephalosporin given as a prolonged or continuous infusion rather than as a larger bolus in a seriously ill patient?
Summary and next
Susceptibility testing converts bacterial genotype into a clinically usable prediction: the MIC says at what concentration growth is inhibited, the MBC whether it is killed, and the categories — susceptible, intermediate, resistant — what that means for a standard dose. The methods are standardised to the millimetre and the McFarland unit because a zone diameter is only meaningful against published breakpoints, and breakpoints are derived from MIC distributions, pharmacokinetics and clinical outcomes together. Recognising the phenotypes — MRSA, ESBL, carbapenemase, inducible clindamycin — is reading the same resistance genetics you met in the previous lesson.
Next: the first lesson of module 2, innate immunity. You have been studying the bacteria; now you meet the host — the barriers, the phagocytes, complement and pattern-recognition receptors that respond within minutes and need no previous encounter with the organism.