14.3 Multidrug-Resistant Pathogens & Antimicrobial Stewardship in the ICU

The intensive care unit represents an ecological epicenter for the emergence and propagation of antimicrobial resistance. The convergence of critically ill immunocompromised hosts, invasive vascular and respiratory devices, and continuous broad-spectrum selective antibiotic pressure drives nosocomial transmission of multidrug-resistant (MDR) pathogens. Optimizing critical care outcomes requires mastery of molecular resistance mechanisms, precision dosing of targeted antimicrobial agents, and rigorous execution of antimicrobial stewardship principles.

---

The ESKAPE Pathogens Overview

The Infectious Diseases Society of America (IDSA) designated the ESKAPE pathogens as the primary group of virulent, multidrug-resistant bacterial species capable of escaping the biocidal activity of conventional antimicrobial therapies:

  E ── Enterococcus faecium (Vancomycin-Resistant Enterococci / VRE)
  S ── Staphylococcus aureus (Methicillin-Resistant S. aureus / MRSA)
  K ── Klebsiella pneumoniae (ESBL- and Carbapenemase-Producing / KPC)
  A ── Acinetobacter baumannii (Carbapenem-Resistant / CRAB)
  P ── Pseudomonas aeruginosa (Multidrug- and Difficult-to-Treat Resistance / DTR)
  E ── Enterobacter cloacae complex (Inducible AmpC Beta-Lactamase Producers)

---

Gram-Positive MDR Pathogens: MRSA and VRE

1. Methicillin-Resistant Staphylococcus aureus (MRSA)

Resistance to methicillin and other $\beta$-lactams in MRSA is mediated by chromosomal acquisition of the mecA or mecC gene, which encodes an altered penicillin-binding protein termed PBP2a. PBP2a displays exceptionally low binding affinity for virtually all standard penicillins, cephalosporins, and carbapenems (with the exception of ceftaroline, an advanced-generation anti-MRSA cephalosporin).

Therapeutic Comparison for MRSA

• Vancomycin:
  • Mechanism: Glycopeptide that binds the terminal D-Ala-D-Ala peptides of nascent peptidoglycan chains, inhibiting cell wall elongation.
  • Precision Dosing & TDM: The 2020 consensus guidelines from IDSA, ASHP, and SIDP eliminated traditional trough monitoring (target 15-20 mcg/mL) for severe MRSA infections due to poor correlation with clinical cure and a 3- to 4-fold increase in acute kidney injury. Therapeutic drug monitoring now mandates AUC/MIC-guided dosing, targeting a 24-hour area under the concentration-time curve to minimum inhibitory concentration ratio of $\text{AUC}_{24}/\text{MIC} = 400\text{ to }600\text{ mg}\cdot\text{hr/L}$ (assuming a broth microdilution MIC $\le 1\text{ mg/L}$), calculated using Bayesian pharmacokinetic software.
• Daptomycin:
  • Mechanism: Lipopeptide that binds to bacterial cell membranes in a calcium-dependent manner, inserting its lipophilic tail to cause membrane depolarization, potassium ion efflux, and rapid cell death without lysis.
  • Clinical Role: Highly effective for MRSA bacteremia and right-sided infective endocarditis.

  • Warning

    Daptomycin is strictly contraindicated in pneumonia. Daptomycin is irreversibly sequestered and inactivated by pulmonary surfactant, rendering it clinically ineffective in alveolar consolidations. Monitor serum creatine phosphokinase (CPK) weekly for skeletal myopathy and rhabdomyolysis, and remain alert for drug-induced acute eosinophilic pneumonia.

• Linezolid:
  • Mechanism: Synthetic oxazolidinone that binds to the 50S ribosomal subunit (specifically the 23S rRNA), preventing assembly of the functional 70S initiation complex and halting protein synthesis.
  • Clinical Role: Excellent tissue distribution with a pulmonary epithelial lining fluid (ELF) to plasma ratio of $\approx 1.0-4.0$, In the ZEPHyR trial it achieved higher clinical cure than vancomycin for MRSA nosocomial pneumonia without a mortality difference; IDSA/ATS guidelines accept either vancomycin or linezolid.
  • Toxicities: Reversible myelosuppression (principally thrombocytopenia) occurring typically after $> 14\text{ days}$ of therapy; peripheral and optic neuropathy with prolonged courses; lactic acidosis mediated by inhibition of host mitochondrial protein synthesis; and weak, non-selective monoamine oxidase (MAO) inhibition, which predisposing patients to Serotonin Syndrome when co-administered with SSRIs, SNRIs, or meperidine.

2. Vancomycin-Resistant Enterococci (VRE)

Enterococcal resistance occurs predominantly in Enterococcus faecium. Resistance is driven by plasmid- or transposon-mediated acquisition of vanA or vanB gene clusters, which synthesize modified peptidoglycan precursors terminating in D-Ala-D-Lactate (D-Ala-D-Lac) rather than D-Ala-D-Ala. This single atomic change eliminates a crucial hydrogen bond, reducing vancomycin binding affinity by 1,000-fold.

• VanA Phenotype: Confers high-level inducible resistance to both vancomycin and teicoplanin.
• VanB Phenotype: Confers variable resistance to vancomycin with retained teicoplanin susceptibility.
• Therapy: Linezolid ($600\text{ mg}$ IV q12h) or high-dose Daptomycin ($8-12\text{ mg/kg/day}$ IV). Standard daptomycin dosing ($6\text{ mg/kg}$) is associated with rapid emergence of daptomycin-resistant enterococcal mutants.

---

Gram-Negative MDR Pathogens: ESBL, CRE, and Non-Fermenters

1. Extended-Spectrum Beta-Lactamases (ESBL)

ESBLs are plasmid-mediated enzymes (predominantly the CTX-M family, alongside SHV and TEM mutants) produced primarily by Escherichia coli, Klebsiella pneumoniae, and Proteus mirabilis. ESBLs hydrolyze third- and fourth-generation cephalosporins (ceftriaxone, ceftazidime, cefepime) and monobactams (aztreonam), but spare carbapenems and are inhibited in vitro by beta-lactamase inhibitors.

The Landmark MERINO Trial

Prior to 2018, observational studies suggested that piperacillin-tazobactam might serve as a "carbapenem-sparing" alternative for bacteremia caused by ESBL producers displaying in vitro susceptibility to piperacillin-tazobactam.

• MERINO Trial (JAMA 2018): A multicenter, international, randomized controlled non-inferiority trial randomized 379 patients with ceftriaxone-resistant E. coli or K. pneumoniae bloodstream infections to receive either piperacillin-tazobactam ($4.5\text{ g}$ IV q6h) or meropenem ($1\text{ g}$ IV q8h).
• Results: 30-day all-cause mortality was $12.3\%$ in the piperacillin-tazobactam group versus $3.7\%$ in the meropenem group (risk difference $8.6\%$). Piperacillin-tazobactam failed to demonstrate non-inferiority; a later reanalysis with central MIC re-testing suggested that some excess deaths involved isolates that were actually nonsusceptible, but carbapenems remain preferred.
• Practice Standard: Carbapenems (meropenem or imipenem-cilastatin) represent the undisputed gold standard for severe invasive infections caused by ESBL-producing Enterobacterales. Ertapenem is an effective once-daily option for non-pseudomonal infections outside the shock state.

2. Carbapenem-Resistant Enterobacterales (CRE)

CRE constitutes an urgent public health threat. Resistance to carbapenems occurs via two fundamental pathways: carbapenemase enzyme production (carbapenemase-producing CRE [CP-CRE]) or porin loss coupled with high-level ESBL/AmpC production. The molecular class of carbapenemase dictates antimicrobial selection:

                     Ambler Classification of Carbapenemases
                                       │
         ┌─────────────────────────────┼─────────────────────────────┐
         ▼                             ▼                             ▼
      Class A                       Class B                       Class D
Serine Carbapenemases        Metallo-Beta-Lactamases        OXA-48-Like Enzymes
   (e.g., KPC)               (e.g., NDM, VIM, IMP)             (e.g., OXA-48)
         │                             │                             │
• Ceftazidime-avibactam       • Cefiderocol                 • Ceftazidime-avibactam
• Meropenem-vaborbactam       • Ceftazidime-avibactam +     • Cefiderocol
• Imipenem-relebactam           Aztreonam                   (Vaborbactam & Relebactam
(Inhibited by novel BLIs)     (Zinc-dependent active site)   are INACTIVE)

• Class A (Serine Carbapenemases, e.g., KPC): Klebsiella pneumoniae carbapenemases hydrolyze all penicillins, cephalosporins, monobactams, and carbapenems. They are inhibited by novel non-$\beta$-lactam $\beta$-lactamase inhibitors (avibactam, vaborbactam, relebactam).
  • Targeted Regimens: Ceftazidime-avibactam, Meropenem-vaborbactam, or Imipenem-cilastatin-relebactam.
• Class B (Metallo-Beta-Lactamases [MBL], e.g., NDM, VIM, IMP): MBLs utilize a divalent zinc cofactor in their catalytic active site to hydrolyze all $\beta$-lactams except aztreonam. Avibactam, vaborbactam, and relebactam cannot inhibit MBLs.
  • Targeted Regimens: Ceftazidime-avibactam combined with aztreonam, the single-product aztreonam-avibactam (FDA-approved February 2025), or cefiderocol (a siderophore cephalosporin that enters bacteria via active ferric iron transporters). While aztreonam is immune to MBL hydrolysis, clinical MBL-producing strains co-produce serine ESBLs or AmpC enzymes that destroy aztreonam; avibactam neutralizes these serine enzymes, shielding aztreonam so it can eradicate the organism.
• Class D (OXA-48-like Serine Carbapenemases): Endemic to the Mediterranean and Middle East. Avibactam inhibits OXA-48; vaborbactam and relebactam do not.
  • Targeted Regimen: Ceftazidime-avibactam.

3. Difficult-to-Treat Pseudomonas aeruginosa

Pseudomonas aeruginosa possesses remarkable intrinsic and acquired resistance, including transcriptional downregulation of the OprD outer membrane porin (conferring carbapenem resistance), derepression of chromosomal AmpC cephalosporinases, and upregulation of multi-component MexAB-OprM efflux pumps.

• Ceftolozane-tazobactam: Engineered specifically to overcome pseudomonal resistance. Ceftolozane possesses high binding affinity for pseudomonal PBPs and is impervious to OprD porin mutations and active efflux pumps. It is the drug of choice for multidrug-resistant P. aeruginosa.
• Alternatives: Ceftazidime-avibactam, Imipenem-cilastatin-relebactam, Cefiderocol.

4. Carbapenem-Resistant Acinetobacter baumannii (CRAB)

CRAB expresses Class D carbapenem-hydrolyzing oxacillinases (OXA-23, OXA-24/40, OXA-58). Therapeutic options are severely constrained.

• Preferred: Sulbactam-Durlobactam plus Imipenem-Cilastatin or Meropenem: Durlobactam is a diazabicyclooctane beta-lactamase inhibitor that protects sulbactam from class A, C, and D enzymes, including OXA-type carbapenemases. Sulbactam-durlobactam was FDA-approved in May 2023 after the ATTACK trial showed lower 28-day mortality than colistin (19% vs. 32%) with less nephrotoxicity, and the IDSA AMR guidance names it the preferred agent for invasive CRAB infection, given with imipenem-cilastatin or meropenem.
• Alternative: High-Dose Ampicillin-Sulbactam: Sulbactam has intrinsic activity against Acinetobacter by binding PBP1 and PBP3. If sulbactam-durlobactam is unavailable, IDSA suggests high-dose ampicillin-sulbactam (27 g/day, providing 9 g/day of sulbactam) combined with at least one other active agent such as minocycline, tigecycline, polymyxin B, or cefiderocol.
• Polymyxin Toxicity: Colistin (polymyxin E) and Polymyxin B function as cationic detergents that disrupt outer membrane lipopolysaccharides. They carry a 30% to 50% incidence of acute kidney injury (acute tubular necrosis) and neurotoxicity (paresthesias, muscle weakness, and neuromuscular blockade).

---

Master Resistance Mechanisms and Targeted Formulary

---

Antimicrobial Stewardship Principles in the ICU

Aggressive antimicrobial stewardship programs (ASP) in the critical care setting optimize clinical cure rates, minimize selective pressure for resistant mutants, curtail Clostridioides difficile colitis, and avoid drug-induced toxicities.

1. Rapid Diagnostic Technologies (RDTs)

Multiplex real-time polymerase chain reaction (PCR) blood culture identification panels (e.g., FilmArray BCID) and matrix-assisted laser desorption ionization time-of-flight mass spectrometry (MALDI-TOF MS) identify microbial species and resistance genes (e.g., mecA, vanA, blaKPC, blaNDM) within 1 to 2 hours of blood culture positivity, enabling targeted antibiotic optimization 24 to 48 hours before conventional phenotypic susceptibility results become available.

2. Algorithmic De-escalation & The 48-72 Hour Antibiotic Timeout

Clinicians should conduct a formal "Antibiotic Timeout" at 48 to 72 hours:

• Review definitive microbiological cultures, sensitivities, and patient clinical trajectory.
• Transition from dual Gram-negative coverage (e.g., anti-pseudomonal beta-lactam plus aminoglycoside or fluoroquinolone) to monotherapy.
• Discontinue empiric MRSA coverage (vancomycin or linezolid) if nasal MRSA PCR screening is negative or cultures fail to isolate staphylococci (nasal MRSA PCR possesses a negative predictive value $> 96-98\%$ for MRSA pneumonia).
• Narrow spectrum: shift from broad-spectrum carbapenems or cefepime to targeted narrow-spectrum agents (e.g., ampicillin, cefazolin, or ceftriaxone) based on confirmed susceptibilities.

3. Evidence-Based Short Duration of Therapy

Historical 14- to 21-day antibiotic regimens increase resistance without improving cure rates. Randomized trials support defined short-course therapies:

            Evidence-Based Short Course Durations in the ICU
  ┌──────────────────────────────────────────────┬──────────────────┐
  │ Clinical Infection Syndrome                  │ Recommended Days │
  ├──────────────────────────────────────────────┼──────────────────┤
  │ Ventilator-Associated Pneumonia (VAP)        │ 7 to 8 days      │
  │ Intra-Abdominal Infection (Source Controlled)│ 4 to 7 days      │
  │ Gram-Negative Bacteremia (Source Controlled) │ 7 days           │
  │ Community-Acquired Pneumonia                 │ At least 5 days  │
  │ Catheter-Related Bloodstream (Uncomplicated) │ 5 to 7 days      │
  └──────────────────────────────────────────────┴──────────────────┘

• VAP Duration: Multiple randomized trials demonstrate that 7 to 8 days of therapy yields identical clinical cure, recurrent pneumonia rates, and hospital mortality compared to 14 days, with significantly fewer resistant superinfections.
• Intra-Abdominal Infections (The STOP-IT Trial, NEJM 2015): In 518 patients with complicated intra-abdominal infection undergoing adequate source control, a short antibiotic course of 4 days produced identical outcomes (rates of surgical site infection, recurrent sepsis, or death) compared to historical courses averaging 8 to 10 days.
• Bloodstream Infection: Several trials showed 7 days noninferior to 14 days for uncomplicated gram-negative bacteremia, and the large BALANCE trial (NEJM 2025; 3,608 hospitalized patients, including ICU patients) found 7 days noninferior to 14 days for bacteremia overall (90-day mortality 14.5% vs. 16.1%). BALANCE excluded Staphylococcus aureus, which still requires at least 14 days of IV therapy (longer if complicated).
• Community-Acquired Pneumonia: ATS/IDSA 2019 advises at least 5 days and clinical stability before stopping; severe or complicated pneumonia (empyema, S. aureus, Pseudomonas) may need longer.

4. Procalcitonin-Guided Discontinuation

Procalcitonin (PCT) is a calcitonin peptide precursor synthesized ubiquitously by parenchymal tissues in response to bacterial endotoxin and pro-inflammatory cytokines (IL-1, TNF-$\alpha$), while being suppressed by viral interferon-$\gamma$. Serum levels rise rapidly within 3 to 6 hours and possess a biological half-life of 24 hours.

• Discontinuation Algorithm: Antibiotic therapy should be discontinued when serum procalcitonin decreases by $\ge 80\%$ from peak or reaches an absolute cutoff of $< 0.25\text{ to }0.5\text{ mcg/L}$, provided the patient exhibits objective clinical stabilization.
• Trial Evidence: In PRORATA (Lancet 2010) and SAPS (Lancet Infect Dis 2016), procalcitonin algorithms reduced antibiotic exposure by about 2 to 3 days without worsening outcomes; only SAPS showed lower 28-day mortality, and ADAPT-Sepsis (JAMA 2025) confirmed shorter antibiotic courses with daily PCT-guided stopping. The 2026 SSC suggests procalcitonin plus clinical evaluation to decide when to stop, but not when to start, antimicrobials.

---

Antimicrobial Stewardship ICU Action Checklist

• Pre-Prescription Phase: Obtain dedicated sterile cultures (at least two sets of blood cultures, sputum/bronchoalveolar lavage, urine) prior to antimicrobial administration, avoiding delays $> 45$ minutes in septic shock.
• Empiric Selection: Formulate empiric coverage based on local antibiograms, prior patient colonization, and recent antimicrobial exposures within the preceding 90 days.
• 48-to-72-Hour Timeout: Formally re-evaluate therapy once microbiological identities and PCR panels return; systematically de-escalate broad empiric double-coverage to a single narrow-spectrum agent.
• Vancomycin Bayesian TDM: Transition from static trough measurements to software-guided $\text{AUC}_{24}/\text{MIC}$ calculations (target 400-600) to mitigate nephrotoxicity.
• Enforce Hard Stop Dates: Pre-specify short course durations in electronic medical records (7 days for VAP, about 4 days after source control for intra-abdominal sepsis, 7 days for uncomplicated non-S. aureus bacteremia).
• Procalcitonin Liberation Protocol: Discontinue antibiotics when procalcitonin clears by $\ge 80\%$ or drops $< 0.25-0.5\text{ mcg/L}$.