At the bench, the same elevated carbapenem MIC can sit on top of very different biology: a plasmid-borne carbapenemase, a metallo-β-lactamase, or a quieter combination of porin loss and β-lactamase overproduction. The isolate may look similar in routine susceptibility testing, but the consequences for infection control, molecular screening, and treatment selection are not interchangeable.
That distinction is where the real diagnostic work begins. A laboratory does not simply need to report that an organism is resistant. It needs to clarify how resistance is being produced, whether a transferable determinant is present, and which apparent treatment options are likely to fail because the resistance mechanism has been misidentified.
Carbapenemases are a family, not a single resistance mechanism
The Ambler classification remains one of the most useful ways to organize carbapenemase biology. It is not merely a naming system. Each class reflects a different catalytic strategy, a different set of genetic determinants, and a different pattern of activity against β-lactamase inhibitors.
The clinically important groups include:
- Class A carbapenemases, particularly KPC enzymes such as KPC-2 and KPC-3. These are serine β-lactamases and are commonly associated with Klebsiella pneumoniae.
- Class B metallo-β-lactamases, including NDM, VIM, and IMP. These enzymes depend on metal ions for activity and have a mechanistic profile that separates them sharply from KPC and OXA-48-like enzymes.
- Class D oxacillinases, especially OXA-48-like enzymes such as OXA-48 and OXA-232. Their expression and phenotypic footprint can be less obvious than that of some KPC-producing isolates.
This classification matters because a molecular result is not just a label to place in the electronic record. It may explain why one inhibitor combination retains in-vitro activity against a particular isolate while another mechanism remains completely outside its coverage.
Ceftazidime-avibactam illustrates the point clearly. The combination demonstrates in-vitro activity against Class A carbapenemases such as KPC-2 and against Class D OXA-48-like producers. It does not have activity against Class B metallo-β-lactamases such as NDM, VIM, and IMP. A laboratory report that identifies only carbapenem resistance, without clarifying the enzyme class where possible, leaves a substantial part of the clinical interpretation unfinished.
The enzyme is only one piece of the phenotype
Carbapenemase production can be strong enough to produce a recognizable resistance pattern, but expression level, host organism, permeability, and additional β-lactamases all influence what appears on the susceptibility panel. Two isolates carrying related resistance genes may not produce identical MIC profiles. Conversely, an isolate without a detectable carbapenemase gene may still show clinically significant carbapenem resistance through non-carbapenemase pathways.
That is why molecular detection and phenotypic testing should not be treated as competing philosophies. PCR or another targeted assay can identify a known determinant quickly, while culture-based susceptibility testing shows the organism’s functional response to antimicrobial exposure. Each answers a different question.
In practical terms:
1. The phenotype asks what the isolate can withstand.
2. The molecular assay asks which known genetic mechanism may be driving that behavior.
3. The interpretive report connects the two without pretending that either is complete on its own.
This is especially important in laboratories that process specimens from patients with recent healthcare exposure, prolonged antimicrobial treatment, or transfer from facilities where multidrug-resistant organisms are already circulating. In those settings, a result can have consequences beyond the individual culture. It may trigger additional screening, infection prevention measures, and genomic or epidemiologic investigation.
A carbapenem-resistant result is a warning; a mechanism-resolved result is a map.
KPC, plasmids, and the ST258 Klebsiella pneumoniae lineage
KPC, or Klebsiella pneumoniae carbapenemase, occupies a central place in the history of carbapenem-resistant Enterobacteriaceae. KPC was first described in 1996, and the associated blaKPC gene is plasmid-borne. That mobility is one reason KPC is so consequential: the resistance determinant is not confined to a single stable chromosomal lineage.
The gene has also been frequently associated with the globally dominant ST258 clone of K. pneumoniae. It is useful to keep the two ideas separate. The plasmid provides a vehicle for genetic transfer, while the bacterial lineage provides a successful background in which the determinant can persist and spread. In a real surveillance program, both dimensions matter.
A molecular diagnostics workflow that detects blaKPC can answer an urgent question quickly, but it does not automatically reveal the full epidemiology of the isolate. Targeted testing may identify the resistance gene; broader genomic surveillance may be needed to understand whether isolates are closely related, whether the same plasmid architecture is circulating, or whether the gene has appeared in a different organism.
Why gene detection is not the same as gene expression
The phrase “blaKPC expression” is often used casually, but the laboratory needs to be precise about what has actually been measured. A gene detection assay generally demonstrates the presence of a target sequence. It does not necessarily quantify transcription, enzyme production, or the contribution of that determinant to the final MIC.
Phenotypic expression can be shaped by several surrounding variables:
- the promoter and regulatory context of the gene;
- the copy number and structure of the carrying plasmid;
- permeability of the outer membrane;
- the presence of additional β-lactamases;
- changes in porin production;
- and the broader genetic background of the isolate.
This is why a molecular result should be interpreted alongside colony identification and antimicrobial susceptibility testing rather than treated as a complete substitute for them. In a working laboratory, the challenge is not simply to find a gene. It is to understand how the gene is operating inside this organism, in this specimen, under these testing conditions.
For technologists, this is also where the manual work remains visible. The isolate must be recovered cleanly, identified accurately, and tested in a way that produces interpretable results. A beautiful molecular assay cannot rescue a contaminated culture, a mixed population, or an organism identification that was never resolved properly.
Non-carbapenemase resistance: when permeability and β-lactamase activity combine
One of the most important misconceptions in CRE work is the assumption that every carbapenem-resistant Enterobacteriaceae isolate must be producing a carbapenemase. Non-carbapenemase-producing CRE exist, and their resistance can emerge through the interaction of two processes:
1. reduced entry of the drug through the outer membrane; and
2. increased hydrolysis or inactivation by other β-lactamases.
In K. pneumoniae, loss or alteration of porins such as OmpK35 and OmpK36 can reduce carbapenem penetration. On its own, a permeability change may not be sufficient to create a high-level resistant phenotype. But when porin deficiency occurs alongside AmpC cephalosporinase overproduction or ESBL production, the combined effect can push the organism into clinically significant carbapenem resistance.
This is a classic example of why resistance mechanisms should be understood as systems rather than isolated genes. The organism does not need a carbapenemase if enough drug is prevented from reaching its target and the β-lactamase burden is high enough to remove what does enter.
The diagnostic problem with porin-mediated resistance
A targeted carbapenemase PCR panel may be negative in a non-carbapenemase-producing CRE isolate. That negative result is useful, but it is not the same as a susceptible result. It tells the laboratory that the common targets included in the assay were not detected; it does not erase the phenotype.
A sensible interpretation therefore requires the results to remain in conversation with one another:
| Laboratory finding | What it may suggest | What it does not prove |
|---|---|---|
| Carbapenem resistance with a detected blaKPC target | A Class A carbapenemase mechanism is likely present | The gene alone does not describe expression level or all resistance contributors |
| Carbapenem resistance with an blaNDM, blaVIM, or blaIMP target | A Class B metallo-β-lactamase mechanism is likely | The result does not predict activity of every possible combination regimen |
| Carbapenem resistance with no tested carbapenemase target | A non-carbapenemase mechanism is possible, including porin loss with AmpC or ESBL production | It does not prove that no carbapenemase exists outside the assay’s target range |
| Reduced carbapenem susceptibility with complex β-lactamase and porin changes | Combined permeability and hydrolytic mechanisms may be operating | A single phenotype cannot always identify the exact genetic pathway |
When I discuss these results with microbiologists, the recurring lesson is simple: a negative molecular screen should narrow the reasoning, not end it. In particular, laboratories should be cautious about presenting molecular detection panels as if they capture every route to CRE. Most are designed around important known targets. They are not universal inventories of resistance biology.
AmpC, ESBLs, and the outer membrane
AmpC overproduction and ESBL production are not interchangeable categories, but both can contribute to carbapenem resistance when paired with impaired outer-membrane permeability. The interaction is often more informative than the presence of either enzyme alone.
For example, an isolate may carry an ESBL that substantially compromises extended-spectrum cephalosporin activity, while porin loss reduces the amount of carbapenem reaching the periplasm. The result can be a phenotype that looks like carbapenemase-mediated resistance on a first pass, even though the underlying mechanism is different.
This is one reason routine susceptibility testing remains an essential part of molecular diagnostics for CRE screening. The work may feel less glamorous than sequencing or rapid amplification, but the culture plate and the dilution panel still provide the functional ground truth against which the molecular findings must be interpreted.
Class B metallo-β-lactamases create a particularly difficult diagnostic gap
NDM, VIM, and IMP belong to the Ambler Class B metallo-β-lactamases. Their catalytic mechanism depends on metal ions, and they are distinct from the serine enzymes in Class A and the oxacillinases in Class D.
The practical consequence is substantial. A laboratory cannot assume that an antimicrobial strategy active against KPC-producing organisms will also be useful against an NDM-producing isolate. The enzyme class changes the therapeutic landscape, and the distinction must be made early enough to influence clinical decision-making.
Why phenotypic and molecular tools need each other
Phenotypic carbapenemase detection may include methods such as the Carba NP test, the Modified Hodge Test, and microbroth dilution antimicrobial susceptibility testing. Each approach comes with a different balance of speed, labor, interpretive complexity, and sensitivity to the biology of the isolate.
A molecular assay can be faster and more specific for the targets it includes. Phenotypic testing, meanwhile, can reveal enzyme activity or an overall resistance pattern that does not fit neatly into the molecular panel. Neither approach should be forced to answer a question it was not designed to answer.
The bench-level workflow often involves several layers:
- recovery of the organism from the clinical specimen;
- confirmation that the culture is pure and representative;
- species-level identification;
- initial susceptibility testing;
- phenotypic assessment for carbapenemase activity when indicated;
- targeted molecular detection of important carbapenemase genes;
- and escalation to sequencing or broader genomic analysis when the phenotype and targeted results do not align.
The awkward cases are often the most educational. A resistant phenotype with no detected target may point toward porin loss, ESBL or AmpC overproduction, an unrepresented carbapenemase, or a technical issue that needs to be revisited. A positive gene result with an unexpected phenotype may prompt questions about gene expression, assay specificity, mixed culture, or the contribution of other mechanisms.
This is not a workflow that rewards rushing. It rewards careful incubation, clean isolation, repeat testing when the pattern is discordant, and communication between the technologist, the clinical microbiologist, and the infection prevention team.
Molecular diagnostics for CRE screening: speed without false certainty
Screening for carbapenem-resistant organisms is often discussed as though the central decision were simply culture versus PCR. In practice, the choice depends on what the laboratory needs to know.
Culture can recover viable organisms and support susceptibility testing, further characterization, and epidemiologic work. Molecular assays can identify selected resistance determinants rapidly, which may be valuable when isolation decisions or urgent infection-control actions cannot wait for a full culture workflow.
But molecular diagnostics are only as broad as their target design. A panel that detects blaKPC and blaNDM may be highly useful for the organisms and mechanisms it covers, while still missing OXA-48-like enzymes, VIM, IMP, or a non-carbapenemase mechanism if those targets are not included. The result must therefore be read with the assay’s scope in mind.
A practical reasoning sequence
When a suspected CRE isolate reaches the laboratory, I find it helpful to keep the reasoning in a defined order rather than jumping immediately to a single gene hypothesis:
1. Confirm the organism and the purity of the culture.
A mixed culture can produce contradictory phenotypic and molecular findings. Species identification is not a clerical preface; it shapes the likely resistance mechanisms and the next testing step.
2. Review the carbapenem susceptibility pattern.
Look at the overall β-lactam profile rather than one reported value. The pattern may suggest a broader β-lactamase background, a permeability problem, or a phenotype that warrants urgent carbapenemase testing.
3. Assess for carbapenemase activity with an appropriate phenotypic method.
Methods such as Carba NP or the Modified Hodge Test may contribute useful evidence, while microbroth dilution provides quantitative susceptibility information.
4. Use targeted molecular detection to identify important known genes.
Targets may include genes associated with KPC, NDM, VIM, IMP, or OXA-48-like enzymes, depending on the assay and local laboratory strategy.
5. Resolve discordance rather than averaging it away.
A resistant phenotype with a negative target panel deserves investigation. So does an unexpected susceptibility pattern in an isolate carrying a carbapenemase gene.
6. Escalate surveillance when the pattern suggests transmission.
If several isolates show related resistance determinants or unusual phenotypes, genomic surveillance can help distinguish repeated independent events from a circulating clone or mobile element.
This approach preserves the strength of both culture and molecular testing. It also protects the laboratory from one of the most damaging shortcuts in antimicrobial resistance work: treating a rapid answer as a complete answer.
Ceftazidime-avibactam and the danger of mechanism-free reporting
The therapeutic implications of carbapenem resistance are where the molecular details become immediately consequential. Ceftazidime-avibactam is active in vitro against Class A carbapenemases such as KPC-2 and against Class D OXA-48-like producers. It lacks activity against Class B metallo-β-lactamases, including NDM, VIM, and IMP.
That single distinction should be enough to discourage reports that use “CRE” as the final level of explanation whenever a mechanism can be identified. CRE is a clinically important category, but it is not a treatment mechanism.
For a KPC-producing isolate, ceftazidime-avibactam may remain relevant based on the isolate’s susceptibility result and the clinical context. For an OXA-48-like producer, the same combination may also show in-vitro activity. For an NDM-producing organism, however, ceftazidime-avibactam alone should not be represented as active. The laboratory must not imply efficacy against Class B producers without an appropriate co-formulation strategy, such as one involving aztreonam, and even then the interpretation requires clinical expertise rather than a simple gene-to-drug shortcut.
This is also why antimicrobial susceptibility results should not be detached from the molecular report. A clinician seeing a resistant Enterobacteriaceae isolate needs a coherent explanation of what was detected, what was not detected, and how confidently the mechanism accounts for the phenotype.
The most useful CRE report does not stop at resistance. It explains the biology that makes resistance actionable.
From isolated results to genomic surveillance
Individual molecular results become more powerful when they are connected over time. The detection of blaKPC in one K. pneumoniae isolate is important for that patient. Repeated detection of the same determinant in related isolates, especially within a recognized lineage such as ST258, raises a different question: is the laboratory seeing a transmission network, a persistent reservoir, or repeated introductions?
Genomic surveillance can help examine those possibilities. It may also clarify whether a resistance gene is moving through a clone, across plasmids, or among different Enterobacteriaceae species. Targeted PCR is usually not designed to answer those questions. It is a focused tool, and its strength is speed and specificity for defined targets.
The surveillance value of a result depends on disciplined laboratory practice upstream. The isolate must be preserved appropriately, linked to accurate metadata, and characterized consistently. Small process details matter: the quality of the subculture, the confidence in species identification, the way an unusual phenotype is documented, and whether apparently duplicate isolates are recognized as such.
In the daily life of a microbiology service, this is where the human labor can disappear from the final dashboard. The graph shows a gene. The spreadsheet shows a date. The infection-control team sees a cluster. Behind each data point are plates read under time pressure, susceptibility panels checked for plausibility, and technologists deciding whether a result is ordinary enough to release or unusual enough to stop and investigate.
The diagnostic lesson
Carbapenem resistance in Enterobacteriaceae is best understood as an interaction between enzyme activity, genetic mobility, membrane permeability, and the organism’s broader β-lactamase background.
KPC, NDM, VIM, IMP, and OXA-48-like enzymes belong to different Ambler classes and carry different implications. The plasmid-borne blaKPC gene and its association with ST258 illustrate how a mobile determinant and a successful bacterial lineage can reinforce one another. Non-carbapenemase CRE remind us that porin loss, especially involving OmpK35 and OmpK36, can combine with AmpC or ESBL overproduction to produce resistance without a detectable carbapenemase gene.
For the laboratory, the practical position is steady rather than dramatic: culture remains important, phenotypic testing remains informative, and molecular assays are most valuable when their scope and limitations are explicit. A negative gene panel does not cancel a resistant phenotype. A positive gene result does not replace susceptibility testing. And the name of the resistance mechanism matters because it can change both the infection-control response and the range of plausible treatment options.
The best molecular microbiology is not the fastest isolated signal. It is the careful joining of signal, culture, phenotype, and context until the organism’s behavior makes sense.