All lecturesID Practice Update

Gram-Negative Bacteria: Species, Resistance, and Clinical Associations

Resident primer (20 min) | Fellow section (30-40 min) | Edited November 10, 2026

This lecture is intended as a supplementary resource for clinicians and trainees to review evidence and discuss clinical cases.

This lecture is intended as a supplementary resource for clinicians and trainees to review evidence and discuss clinical cases.

Edited November 10, 2026. Adult clinical ID teaching. Organism identity does not replace isolate-specific susceptibility testing, infection-site assessment, and source control. Drug examples assume susceptible isolates and appropriate exposure. Verify current CLSI breakpoints and the local formulary.2

Learning objectives

  1. Predict important intrinsic resistance from species identification before the antibiogram returns.
  2. Distinguish inducible AmpC, ESBL, carbapenemases, and non-beta-lactam resistance mechanisms.
  3. Explain why seemingly active antibiotics can be poor choices at particular sites.
  4. Recognize unusual species/phenotypes that require a focused search for the source, endovascular involvement, or metastatic infection.
  5. Interpret rare nonfermenters cautiously, including when a respiratory isolate may be colonization.

Resident section

Initial interpretation of gram-negative cultures

Major organism groups

GroupCommon examplesInitial clinical implications
EnterobacteralesE. coli, Klebsiella, Enterobacter, Citrobacter, ProteusUrinary/GI/biliary sources common; ESBL, AmpC and carbapenemases vary by species
NonfermentersPseudomonas, Acinetobacter, Stenotrophomonas, AchromobacterSpecies-specific intrinsic resistance, environmental and device associations; respiratory colonization can complicate interpretation
Exposure-associated gram-negativesAeromonas, Vibrio, Pasteurella, CapnocytophagaFreshwater, salt/brackish water, and animal exposures materially change initial therapy and assessment
Enteric invasive pathogensSalmonella, Shigella, Campylobacter, YersiniaThe clinical question changes when an enteric organism is recovered from blood or another normally sterile site. The examples below focus on invasive complications.

AmpC species differences

Organism groupPractical approach
Enterobacter cloacae complex, Klebsiella aerogenes, Citrobacter freundii, Hafnia alveiAn isolate may initially test susceptible to ceftriaxone and become resistant as AmpC expression increases. Clinical reports support this risk for the first three species; inclusion of H. alvei rests mainly on laboratory evidence. For invasive disease, cefepime is preferred when the result is susceptible or susceptible dose-dependent, the required exposure is achievable, and no ESBL gene has been identified.1
Serratia marcescens, Morganella morganii, Providencia spp.Clinically significant AmpC overexpression is less common in this group. Susceptibility results usually guide selection. A high bacterial burden and limited source control, as with endocarditis, may still favor cefepime even when ceftriaxone tests susceptible.1
E. coli, Klebsiella pneumoniae, Proteus mirabilisAcquired plasmid AmpC and ESBL remain possible. Do not assume every third-generation cephalosporin resistance is inducible AmpC.

For these Enterobacterales, a cefepime MIC of 4–8 µg/mL is susceptible dose-dependent in the cited guidance. That category assumes a regimen that achieves increased exposure. The MIC alone does not identify an ESBL.1 Nitrofurantoin is not used for pyelonephritis or bacteremia. Oral fosfomycin is not initial therapy for pyelonephritis or routine bacteremia treatment; selected ESBL-producing E. coli cUTI transition use after active IV therapy, when other oral options are unavailable, is discussed in the Antibiotics Primer.

Important antimicrobial coverage gaps

DrugHigh-yield gap or caveat
CeftriaxoneCeftriaxone is unreliable against Pseudomonas. Initial susceptibility may not persist in invasive infection with a moderate-risk AmpC species. A narrower exception concerns nonsevere infection already improving on empiric ceftriaxone after adequate source control.1
ErtapenemNo reliable Pseudomonas, Acinetobacter or Enterococcus activity.
MoxifloxacinNot an antipseudomonal fluoroquinolone. Do not extrapolate from ciprofloxacin/levofloxacin. Poor urinary exposure makes it a poor UTI choice.
NitrofurantoinNot for Proteus, Morganella, Providencia, Pseudomonas or upper-tract/systemic infection.
PolymyxinsProteus, Morganella, Providencia and Serratia have intrinsic resistance; substantial toxicity limits use elsewhere.
Tigecycline / eravacyclineNot dependable for bloodstream or urinary infection because of pharmacokinetics; Proteus/Providencia/Morganella generally have poor tigecycline activity.
Daptomycin / vancomycin / linezolidNo clinical gram-negative activity due to outer-membrane or target limitations; do not count as gram-negative coverage.
AztreonamGram-negative aerobic activity including some Pseudomonas, but no gram-positive or anaerobic coverage. Many ESBL and AmpC enzymes hydrolyze aztreonam. MBL enzymes alone do not.
Meropenem-vaborbactamPrimarily useful for susceptible KPC-producing Enterobacterales. Vaborbactam does not meaningfully restore meropenem activity against DTR Pseudomonas caused by typical non-KPC mechanisms.
Ampicillin-sulbactamSulbactam has direct activity against Acinetobacter at adequate doses; conventional low-dose regimens are not automatically adequate for CRAB.

Resistance mechanisms and susceptibility results

PhenotypeMechanism or distinctionTreatment-selection implications
ESBLEnzymatic hydrolysis of many penicillins and cephalosporins, commonly CTX-MA carbapenem is generally preferred for ESBL-producing Enterobacterales bloodstream infection. In MERINO, piperacillin-tazobactam did not meet the noninferiority criterion for 30-day mortality versus meropenem in ceftriaxone-resistant E. coli or K. pneumoniae bacteremia. Cefepime was not tested in that trial; a susceptible cefepime result alone does not establish efficacy for this syndrome.1,8
Inducible AmpCSelection of derepressed mutants during therapyInvasive infection with a moderate-risk species generally favors cefepime if activity is demonstrated and the required exposure is achievable. Ceftriaxone susceptibility at baseline does not exclude resistance emerging during therapy.1
CRE without carbapenemaseOften ESBL/AmpC combined with porin lossCarbapenemase test helps distinguish options; treatment depends on AST and site.
KPCClass A serine carbapenemaseMeropenem-vaborbactam, ceftazidime-avibactam, or imipenem-relebactam are commonly active. For invasive KPC-producing Enterobacterales infection, IDSA slightly favors them in that order. Check AST.1
NDM-producing EnterobacteralesNDM is a metallo-beta-lactamase (MBL). Co-produced serine enzymes can inactivate unprotected aztreonam.For invasive infection, aztreonam-avibactam or cefiderocol are preferred if active. Ceftazidime-avibactam plus aztreonam is an alternative if the coformulation is unavailable. This comparison is specific to NDM-producing Enterobacterales; it does not establish the same choices for VIM/IMP producers or MBL-producing Pseudomonas.1
OXA-48-like EnterobacteralesClass D carbapenemaseCeftazidime-avibactam generally preferred; meropenem-vaborbactam and imipenem-relebactam do not reliably inhibit this enzyme.
Pseudomonas DTRMultiple pathways: AmpC upregulation, efflux, porin changes, target enzymesRequest testing of ceftolozane-tazobactam, ceftazidime-avibactam, imipenem-relebactam and cefiderocol. Choose based on isolate and infection site.
CRABOften OXA carbapenemases and multiple resistance mechanismsFor invasive carbapenem-resistant A. baumannii infection, sulbactam-durlobactam with imipenem or meropenem is preferred when active. Resistance to sulbactam-durlobactam or an MBL changes that reasoning. A respiratory or wound isolate may instead represent colonization.1

Distinguish MDR, XDR and DTR. DTR Pseudomonas describes resistance to a particular set of conventional antipseudomonal agents, not simply resistance to one carbapenem. New resistance can arise on treatment: repeat cultures and AST for clinical failure, relapse, or breakthrough bacteremia.

Organism-specific characteristics

Pseudomonas aeruginosa

Stenotrophomonas maltophilia

Acinetobacter baumannii complex

Achromobacter xylosoxidans

Burkholderia cepacia complex

Elizabethkingia species

Klebsiella pneumoniae: hypervirulence versus hypermucoviscosity

Proteus, Morganella, Providencia, Serratia

Gram-negative infections at uncommon sites

FindingSpecific questions to ask
Gram-negative endocarditisProsthetic valve/device? Catheter or healthcare-associated exposure? Persistent bloodstream infection? Emboli or perivalvular complications? Non-HACEK gram-negative endocarditis usually requires specialist multidisciplinary management; do not assume routine 7-day bacteremia treatment.
Gram-negative vertebral osteomyelitis or PJIUrinary or GI source? Recent procedure? Infected hardware? Isolate concordance between deep samples and blood? Need for surgery or retained-device strategy?
Recurrent Proteus bacteremia or UTIUrease-associated stone? Obstruction? Renal/perinephric abscess? Chronic catheter, nidus or prostatitis?
Klebsiella liver abscessBiliary disease versus cryptogenic infection? Hypervirulent phenotype? Ocular or neurologic symptoms suggesting metastatic focus?
Gram-negative CNS infectionNeurosurgical device, recent procedure, penetrating injury, bacteremia, contiguous source? Species and CSF penetration materially change choices.
Aeromonas after freshwater exposureNecrotizing soft-tissue infection? Rapid surgical assessment if severe. Avoid relying on first-generation cephalosporins.
Vibrio vulnificus after salt/brackish water or raw oystersAfter coastal-water exposure, rapidly progressive pain, hemorrhagic bullae, or sepsis raises concern for a V. vulnificus wound infection. Start active treatment immediately when suspected. CDC recommends a third-generation cephalosporin such as ceftazidime plus doxycycline for V. vulnificus wound infection.3 Do not wait for cultures or consultation. Suspected necrotizing infection needs urgent surgical assessment and debridement when indicated. Include broader empiric coverage if polymicrobial necrotizing infection is possible.10 Raw-oyster ingestion can present instead with bloodstream infection, particularly with liver disease.3
Pasteurella or Capnocytophaga after animal exposureBite, scratch or saliva contact? Asplenia/alcohol use for fulminant Capnocytophaga? Standard cephalexin/clindamycin regimens can be insufficient for animal-bite infections.
Campylobacter fetus bacteremiaConsider endovascular infection, especially with vascular disease or prosthetic material; distinct from uncomplicated C. jejuni enteritis.
Nontyphoidal Salmonella bacteremiaAge, immunosuppression, vascular grafts or atherosclerotic disease; evaluate for endovascular infection when clinical features warrant.

Enteric pathogens and invasive complications

Fellow section: additional resistance and phenotypes

Common errors in susceptibility interpretation

  1. In E. cloacae complex bacteremia, ceftriaxone susceptibility describes the initial isolate. AmpC expression can increase during treatment, which is why cefepime is generally favored when active, exposure is adequate, and an ESBL has not been identified. Source control remains a separate decision.1
  2. A cefepime-susceptible result in ESBL-producing E. coli bacteremia does not make cefepime a reliable definitive choice. The bloodstream syndrome and ESBL mechanism favor a susceptible carbapenem.1
  3. For invasive NDM-producing Klebsiella infection, aztreonam-avibactam and cefiderocol are preferred options if active. Ceftazidime-avibactam alone does not overcome NDM. No direct comparison establishes aztreonam-avibactam as superior to cefiderocol.1
  4. DTR Pseudomonas susceptible to a novel agent. Determine infection site, dosing, MIC and source control. Do not assume all new beta-lactams are interchangeable or that meropenem-vaborbactam is an antipseudomonal salvage option.
  5. Stenotrophomonas from BAL in a patient whose imaging and oxygen requirement are improving without active therapy. Colonization may be more plausible than invasive infection. Treat the patient and syndrome, not the culture in isolation.

CRE treatment by resistance mechanism

MechanismTypical invasive-infection options (if active)Common error
KPC-producing EnterobacteralesFor invasive infection, IDSA considers meropenem-vaborbactam, ceftazidime-avibactam, and imipenem-relebactam preferred options and slightly favors them in that order.1Assuming old polymyxin-based salvage is routinely preferable. The ranking is based on observational evidence. Check susceptibility, prior exposure, and infection site.
NDM-producing EnterobacteralesAztreonam-avibactam or cefiderocol if susceptible; ceftazidime-avibactam plus aztreonam is an alternative if the coformulation is unavailable.1Assuming ceftazidime-avibactam alone covers NDM, or applying this Enterobacterales recommendation to every MBL-producing isolate.
OXA-48-likeCeftazidime-avibactam; alternatives depend on AST and siteTreating meropenem-vaborbactam as equivalent against OXA-48.
Non-carbapenemase CREEvaluate full AST, permeability defect and beta-lactamase profileAssuming all CRE must have a carbapenemase.

This comparison assumes invasive infection, identification of the organism and resistance mechanism, and demonstrated activity. Drug exposure, renal function, site, and availability influence the choice. MBL-producing P. aeruginosa is a separate situation: for NDM, VIM, or IMP producers, cefiderocol is preferred if active.1

Susceptibility testing limitations

2026 IDSA AMR guidance: changes from 2024

The July 2026 update considers evidence through March 1, 2026. The panel used expert consensus rather than formal GRADE ratings. Preferred agents in the suggested approach are listed alphabetically. The rationale explains specific preferences, such as the slight ranking for invasive KPC-producing Enterobacterales infection. The important distinctions are which patients were studied and how well the evidence supports a difference in clinical outcomes.1

AreaWhat changed in 2026What it changes clinically / limitation
Recognizing possible ESBL productionIn E. coli, K. pneumoniae or K. oxytoca, a ceftriaxone MIC ≥4 µg/mL is a proxy for possible ESBL production.1This MIC is not a molecular diagnosis. Other mechanisms can produce the same result; the threshold does not identify ESBL production across every species.
AmpCH. alvei joins the moderate-risk group. The panel retains its 2024 advice against routine piperacillin-tazobactam for invasive infection with these species.Continuing empiric ceftriaxone may be reasonable in a nonsevere, clinically improving infection with adequate source control. That exception does not establish ceftriaxone as preferred initial definitive therapy for invasive disease.1
NDM-producing EnterobacteralesAztreonam-avibactam and cefiderocol are preferred if active. Ceftazidime-avibactam plus aztreonam is an alternative if the coformulation is unavailable. A preference for the coformulation over separate infusions does not establish superiority over cefiderocol.1The U.S. aztreonam-avibactam indication is narrower: in combination with metronidazole for complicated intra-abdominal infection due to listed susceptible gram-negative organisms in adults with limited or no alternatives. That label does not establish a bacteremia indication.7
DTR P. aeruginosaObservational comparative data help explain a preference for susceptible ceftolozane-tazobactam in pneumonia when the isolate does not produce KPC or an MBL.Carbapenemase identification can change the preferred agent. An Enterobacterales table does not transfer unchanged to Pseudomonas. Breakthrough infection warrants repeat cultures and drug-specific susceptibility testing.1
CRABThe update reinforces the preference for active sulbactam-durlobactam with imipenem or meropenem in invasive infection. If the preferred drug is delayed, high-exposure ampicillin-sulbactam with a second active agent can serve as a bridge.A bridge for delayed availability is different from an alternative needed because the isolate is resistant. Sulbactam-durlobactam resistance or an MBL requires reassessment; a colonizing isolate alone is not an indication for therapy.1
Stenotrophomonas maltophiliaMajor change: cefiderocol monotherapy designated preferred for invasive infection. Aztreonam-avibactam, preferably with a second active agent, becomes an alternative.This is a meaningful departure from the earlier combination-oriented approach, but clinical evidence for cefiderocol monotherapy remains very limited, relying heavily on susceptibility data and neutropenic animal models. Do not apply this to colonization or presume improved patient outcomes have been proven.
Nebulized antibiotics in DTR Pseudomonas or CRAB pneumoniaSignals of possible benefit have not established a routine role for adjunctive inhaled antibiotics.For DTR Pseudomonas pneumonia with an active systemic beta-lactam, routine addition is discouraged. Routine nebulized therapy is also discouraged in CRAB pneumonia.1
General frameworkUpdated 2026 CLSI breakpoints/dosing table; definitions of uncomplicated/complicated UTI aligned with the 2025 IDSA cUTI guideline.Check actual MIC and current interpretive standard. The guidance does not mean every beta-lactam on the preferred list is clinically equivalent.

Evidence limitations in the 2026 guidance

Fellow discussion questions

  1. A stable patient with a nonsevere E. cloacae intra-abdominal infection is improving on empiric ceftriaxone after adequate source control. How would the discussion change if an abscess remained undrained or the patient became unstable?
  2. A patient has NDM-producing K. pneumoniae bacteremia. Which agents are preferred, how is aztreonam protected, and why is ceftazidime-avibactam alone inadequate?
  3. An ICU BAL grows S. maltophilia. When would treatment be unnecessary? If invasive infection is established, why might cefiderocol monotherapy be favored now, and why should confidence in comparative clinical efficacy remain limited?

Source comparison: IDSA 2026 AMR Guidance, "Notable Updates from the 2024 IDSA AMR Guidance Document" (https://www.idsociety.org/practice-guideline/amr-guidance/); IDSA 2024 guidance (https://doi.org/10.1093/cid/ciae403); 2026 publication (https://doi.org/10.1093/cid/ciag481).

Clinical cases

Case 1: initially ceftriaxone-susceptible Enterobacter bacteremia

An obstructed urinary tract is the source of E. cloacae complex bacteremia. The isolate initially tests susceptible to ceftriaxone and cefepime. Why does this differ from cystitis? Obstruction is an unresolved source, and AmpC expression may undermine ceftriaxone during therapy. Drainage and an active drug that reaches the bloodstream address different parts of the problem. For cefepime, the actual MIC and achievable exposure matter.1

Case 2: Serratia bacteremia in a patient with a prosthetic valve

Two blood culture sets grow S. marcescens in a patient with a prosthetic valve. The species carries less risk of AmpC overexpression than E. cloacae complex, but a focus with high bacterial burden and limited source control may still influence drug selection.1 Bacteremia in a patient with a prosthetic valve raises concern for endocarditis without proving it. What would repeat cultures and echocardiography clarify? Which findings would prompt assessment with the valve team?

Case 3: Klebsiella liver abscess and blurred vision

A patient with K. pneumoniae bacteremia and a liver abscess reports new blurred vision. Possible endophthalmitis warrants urgent ophthalmologic assessment while the team addresses the abscess and bloodstream infection. A positive string test describes hypermucoviscosity; it cannot establish hypervirulence or decide whether the eye needs evaluation.5,9

Case 4: BAL growing Stenotrophomonas

A ventilated patient is improving without an antibiotic active against the S. maltophilia recovered from BAL, and another diagnosis explains the infiltrates. Colonization is plausible. What new clinical findings would make pneumonia more likely? Only if infection is supported do the susceptibility result and the limited human evidence for treatment options become relevant.1

Case 5: NDM Klebsiella bloodstream infection

Blood cultures grow NDM-producing K. pneumoniae. Ceftazidime-avibactam alone does not address NDM; other beta-lactamases can inactivate aztreonam given alone. Aztreonam-avibactam or cefiderocol are preferred options if active. If the coformulation is unavailable, ceftazidime-avibactam plus aztreonam is an alternative. This bloodstream example illustrates guidance for invasive NDM-producing Enterobacterales, not a separate U.S. bacteremia indication for aztreonam-avibactam.1,7

Case 6: Pseudomonas and the wrong fluoroquinolone

A urine culture grows P. aeruginosa susceptible to ciprofloxacin. A trainee recommends oral moxifloxacin. Explain why same-class labeling is not sufficient, and why moxifloxacin is also a poor urinary agent.

Case 7: painful leg after freshwater injury

Rapidly progressive pain and bullae after freshwater trauma raise concern for a necrotizing infection. Do not wait for cultures before urgent surgical assessment and empiric antibiotics. When the wound grows Aeromonas, exposure history and isolate-specific testing help narrow therapy; cefazolin alone is unreliable.10,11

Summary

A species name helps predict which susceptibility results deserve a second look. The next question is whether this isolate explains the illness and whether an active drug will reach the infected site. With an unusual isolate, knowing how the laboratory identified and tested it can be as useful as the reported MIC.

References

  1. IDSA 2026 Guidance on the Treatment of Antimicrobial-Resistant Gram-Negative Infections. https://www.idsociety.org/practice-guideline/amr-guidance/ (primary reference for AmpC, ESBL, CRE, DTR Pseudomonas, CRAB, and Stenotrophomonas; reassess if updated).
  2. Clinical and Laboratory Standards Institute. M100 Performance Standards for Antimicrobial Susceptibility Testing, current edition. https://clsi.org/standards/products/microbiology/documents/m100/ (check exact version and organism-agent breakpoints through institutional access).
  3. CDC, Clinical Overview of Vibrio vulnificus Infection. https://www.cdc.gov/vibrio/hcp/clinical-overview/index.html
  4. CDC, Clinical Overview of Shigellosis. https://www.cdc.gov/shigella/hcp/clinical-overview/index.html
  5. Russo TA, Marr CM. Hypervirulent Klebsiella pneumoniae. Clin Microbiol Rev. 2019;32:e00001-19. https://doi.org/10.1128/CMR.00001-19
  6. Huse HK, et al. Challenges in establishing epidemiological cut-off values for the Burkholderia cepacia complex. J Clin Microbiol. Published September 25, 2026. https://doi.org/10.1128/jcm.00730-26.
  7. FDA. EMBLAVEO prescribing information, indications and usage.
  8. Harris et al. Piperacillin-tazobactam versus meropenem for ceftriaxone-resistant bloodstream infection (MERINO). JAMA, 2018.
  9. Lim et al. Ocular involvement in Klebsiella pneumoniae bacteremia. Antimicrobial Agents and Chemotherapy, 2023.
  10. IDSA. Skin and soft tissue infections: suspected necrotizing infection.
  11. Aravena-Román M, Inglis TJJ, Henderson B, Riley TV, Chang BJ. Antimicrobial susceptibilities of Aeromonas strains isolated from clinical and environmental sources to 26 antimicrobial agents. Antimicrob Agents Chemother. 2012;56:1110–1112.
  12. CLSI AST News Update. Susceptibility testing in B. cepacia complex, Fall 2025.

Teaching caveat: Uncommon species, especially Achromobacter, Burkholderia, and Elizabethkingia, lack strong randomized comparative regimen evidence. This resource deliberately avoids assigning fixed treatment durations or empiric drug mandates to these organisms. Doses are not included in this primer; use organism- and syndrome-specific antibiotic modules and pharmacy dosing references for renal function, infusion optimization and PK/PD.