Antimicrobial Susceptibility Testing

Antimicrobial susceptibility testing (AST) determines whether a bacterial or fungal isolate is likely to respond to an antimicrobial agent.

Methods are disc diffusion (Kirby-Bauer, zone diameter in millimetres), gradient strip tests (Etest, minimum inhibitory concentration read from the strip), broth microdilution (the reference method, giving the minimum inhibitory concentration), automated systems (VITEK, MicroScan, Phoenix) and rapid phenotypic methods (direct-from-blood-culture disc tests, short incubation systems),

together with molecular detection of resistance genes (mecA, vanA, vanB, blaKPC, blaNDM, blaOXA-48, blaCTX-M) and screening tests for specific mechanisms (cefoxitin for methicillin resistance, inducible clindamycin resistance D-test, modified carbapenem inactivation, carbapenemase inhibitor tests, extended spectrum beta lactamase confirmation).

Results are reported as susceptible (S), susceptible with increased exposure (I, where a higher dose or concentration at the site overcomes the resistance) or resistant (R), using clinical breakpoints that combine microbiological, pharmacokinetic and clinical outcome data.

Antimicrobial resistance is associated with about 5 million deaths each year worldwide, of which 1.2 million are directly attributable, with the highest burden in sub-Saharan Africa and South Asia.

Priority pathogens are carbapenem resistant Enterobacterales, Acinetobacter baumannii and Pseudomonas aeruginosa, methicillin resistant Staphylococcus aureus, vancomycin resistant enterococci, third generation cephalosporin resistant Enterobacterales, multidrug resistant tuberculosis, and drug resistant Neisseria gonorrhoeae and Candida auris.

Indications. Any clinically significant bacterial isolate from a sterile site (blood, CSF, joint, pleural, peritoneal fluid), urine, deep wound, respiratory tract and other sites where treatment is required. Not for commensals or contaminants. Fungal susceptibility for invasive Candida and Aspergillus isolates.

Surveillance of local resistance rates (antibiogram), infection control (identification of multidrug resistant organisms), epidemiological studies, and direction of de-escalation in stewardship programmes.

Specimen. A pure isolate from culture (not a direct swab), grown on non-selective medium, from 18 to 24 hours of growth. A standardised inoculum (0.5 McFarland turbidity, about 1.5 times 10 to the power 8 colony forming units/mL) on Mueller-Hinton agar (with blood or special supplements for fastidious organisms), incubated at 35 °C for 16 to 20 hours (24 hours for some agents). Direct testing from positive blood culture broth is used for earlier results.

Reference values. Categories depend on organism, drug and breakpoint document in use, so each laboratory reports S, I or R (or the minimum inhibitory concentration) using its current breakpoints. Quality control strains (for example, Escherichia coli ATCC 25922, Staphylococcus aureus ATCC 25923 and 29213, Pseudomonas aeruginosa ATCC 27853) must give zone diameters and minimum inhibitory concentrations within accepted ranges.

Interpretation.

• Susceptible: treatment success is likely at the usual dose. Susceptible, increased exposure: success is likely at higher dose, prolonged or continuous infusion, or at sites where the drug concentrates (urine).
• Resistant: treatment failure is likely, and the agent is avoided. Resistance in vitro with clinical success can occur in urinary infection with drugs concentrated in urine.
• Methicillin resistant Staphylococcus aureus (cefoxitin resistant or mecA positive): resistant to all beta lactams except ceftaroline and ceftobiprole.
• Extended spectrum beta lactamase producers (Escherichia coli, Klebsiella): resistant to penicillins and cephalosporins, aztreonam, with carbapenems the usual treatment for serious infection (piperacillin-tazobactam is not reliable in bloodstream infection).
• Carbapenemase producers (KPC, NDM, OXA-48, VIM, IMP): resistant to carbapenems, and therapy uses newer combinations (ceftazidime avibactam, meropenem vaborbactam, cefiderocol, aztreonam avibactam for metallo-beta-lactamases), colistin or tigecycline, as the mechanism directs.
• Inducible clindamycin resistance (D-test positive): clindamycin is avoided in serious staphylococcal infection despite apparent susceptibility.
• AmpC-producing organisms (Enterobacter, Citrobacter, Serratia): risk of resistance emerging on third generation cephalosporins despite initial susceptibility, so cefepime or a carbapenem is chosen for serious infection.
• Vancomycin resistant enterococci (vanA, vanB): linezolid or daptomycin.
• Minimum inhibitory concentration near the breakpoint ("susceptible close to breakpoint") needs clinical judgement, dose optimisation and drug monitoring (vancomycin, aminoglycosides).
• Intrinsic resistance (for example, Klebsiella to ampicillin, Pseudomonas to many agents, enterococci to cephalosporins) means testing is not needed and the drug is not reported.
• Cascade reporting: narrow spectrum agents are reported first, with broader agents released only when resistance is found, to support stewardship.

Critical values and findings needing immediate action.

• Carbapenem resistant organism, vancomycin resistant enterococcus, MRSA, extended spectrum beta lactamase producer or Candida auris: call the clinician, start contact precautions, isolate the patient, notify infection control and review treatment.
• Resistance in a blood culture isolate to the empirical antibiotic: change therapy within hours.
• Pan resistant isolate: infectious disease and microbiology review for combination therapy, and research study or compassionate access drugs.
• Rifampicin resistance in Mycobacterium tuberculosis: management of multidrug resistant tuberculosis, notification and expert review.
• Treatment failure with a susceptible report: repeat testing for heteroresistance, inoculum effect, emergence of resistance on therapy, and drug exposure.

Interferences and pitfalls.

• Testing of mixed cultures, contaminants or colonisers produces misleading results, and treatment follows clinical infection and not the antibiogram.
• Inoculum density, medium depth, pH, cation content, incubation time and temperature, disc potency and storage all alter zone sizes. Quality control must be run on each batch.
• In vitro susceptibility does not guarantee clinical success: poor penetration (cephalosporins and aminoglycosides into CSF or bone, daptomycin in the lung), host factors, foreign material, biofilms, abscesses and inadequate dosing cause failure.
• Different breakpoint systems give different categories, and changes in breakpoints alter reported resistance rates over time.
• Heteroresistance (colistin, carbapenems in Acinetobacter and Klebsiella) is missed by routine testing, and broth microdilution is needed for colistin.
• Automated systems can miss some resistance mechanisms (vancomycin intermediate S. aureus, certain carbapenemases, mecC), and confirmatory methods are required.
• Antibiograms from small numbers (under 30 isolates) are unreliable, and cumulative reports should use the first isolate for each patient in each period.
• Fungal susceptibility is less standardised, and echinocandin resistance in Candida glabrata (FKS mutations) and azole resistance in Aspergillus fumigatus (cyp51A) need specialised testing.
• Laboratory capacity is limited in low resource settings, where use of discs with quality control, and surveillance networks, are the minimum.

Special groups.

• Children and neonates: dose and drug choice are limited, and breakpoints may differ for paediatric dosing.
• Pregnancy: tetracyclines, fluoroquinolones and aminoglycosides are generally avoided regardless of susceptibility.
• Renal and hepatic impairment: dose adjustment and drug monitoring.
• Immunosuppressed patients and device infections: bactericidal agents, prolonged courses and combination therapy.
• Regions: resistance rates differ greatly, with high extended spectrum beta lactamase and carbapenemase rates in South Asia, the Middle East and parts of Africa and Southern Europe, and low rates in Northern Europe, and local antibiograms should direct empirical therapy.

Follow up tests. Minimum inhibitory concentration by gradient strip or broth microdilution for critical drugs, phenotypic confirmation tests (extended spectrum beta lactamase, carbapenemase, D-test), molecular resistance gene detection, whole genome sequencing for outbreaks, drug level monitoring (vancomycin, aminoglycosides, voriconazole), repeat cultures to document clearance, infection control screening swabs, and local antibiogram review.