Practical Colistin Prescribing: A Clinical Guide
Practical Colistin Prescribing in the Era of Multidrug-Resistant Gram-Negative Infections: A Clinical Guide
Abstract
Colistin has re-emerged as a last-resort antibiotic for multidrug-resistant (MDR) and extensively drug-resistant (XDR) Gram-negative infections. Despite decades of use, significant knowledge gaps persist regarding optimal dosing, therapeutic drug monitoring, and strategies to minimize nephrotoxicity. This review provides evidence-based guidance on practical colistin prescribing for internists managing critically ill patients with limited treatment options.
Introduction
The global emergence of carbapenem-resistant Enterobacteriaceae (CRE), multidrug-resistant Pseudomonas aeruginosa, and Acinetobacter baumannii has necessitated the resurrection of colistin (polymyxin E), an antibiotic relegated to near-obsolescence due to toxicity concerns. Current estimates suggest that colistin remains active against 70-95% of carbapenem-resistant isolates, making it indispensable in contemporary antimicrobial stewardship.
However, colistin prescribing remains fraught with complexity. The prodrug formulation (colistimethate sodium), non-linear pharmacokinetics, narrow therapeutic index, and substantial nephrotoxicity risk demand a sophisticated understanding that extends beyond standard antimicrobial prescribing principles.
Pharmacology: Understanding the Prodrug Paradox
Pearl #1: Colistimethate sodium (CMS) is the inactive prodrug administered intravenously, which undergoes spontaneous hydrolysis to form colistin (the active antibacterial agent). This conversion occurs both in vivo and in vitro, creating significant implications for dosing and therapeutic drug monitoring.
The pharmacokinetic complexity stems from the fact that CMS and colistin have markedly different elimination patterns. CMS is predominantly renally eliminated unchanged, while colistin undergoes non-renal clearance. This explains the counterintuitive observation that patients with renal impairment may have lower colistin exposure despite higher CMS levels.
Dosing Principles
Modern colistin dosing has evolved from empirical weight-based regimens to more rational approaches based on the pharmacokinetic/pharmacodynamic (PK/PD) target of achieving adequate colistin plasma concentrations.
Loading Dose: A loading dose of 9 million international units (MIU) of CMS (approximately 300 mg of colistin base activity) should be administered to all patients regardless of renal function. This critical step achieves therapeutic colistin concentrations rapidly, as CMS-to-colistin conversion is relatively slow.
Maintenance Dosing: For patients with normal renal function (CrCl >80 mL/min), maintenance doses of 4.5 MIU every 12 hours are recommended. The European Committee on Antimicrobial Susceptibility Testing (EUCAST) epidemiological cutoff for susceptibility is ≤2 mg/L, which should guide therapeutic targets.
Oyster #1: Many clinicians incorrectly assume that higher CMS doses in renal impairment will increase colistin exposure. In reality, accumulation of CMS (the inactive prodrug) occurs, while colistin concentrations may remain subtherapeutic. Dose adjustments in renal impairment should maintain adequate colistin levels while preventing CMS accumulation.
Renal Dose Adjustment: The Evidence Base
The Garonzik nomogram, derived from population pharmacokinetic studies, provides the most evidence-based approach to dose adjustment in renal impairment. Key recommendations include:
- CrCl 50-79 mL/min: 3.8-4.5 MIU every 12 hours
- CrCl 30-49 mL/min: 3-3.4 MIU every 12 hours
- CrCl 10-29 mL/min: 2.3-2.6 MIU every 12 hours
Hack #1: For patients on continuous renal replacement therapy (CRRT), colistin is poorly cleared by dialysis membranes due to high protein binding (50-60%) and large molecular size. Loading doses remain essential, with maintenance doses of 3-4.5 MIU every 12-24 hours depending on residual renal function and CRRT intensity. Do not reduce colistin doses excessively in CRRT patients—this is a common prescribing error leading to treatment failure.
Therapeutic Drug Monitoring: When and How
Pearl #2: Therapeutic drug monitoring (TDM) for colistin is increasingly available and should be considered in specific scenarios:
- Critically ill patients with augmented renal clearance
- Obesity (>120 kg) or low body weight (<60 kg)
- Renal replacement therapy
- Treatment failure after 48-72 hours
- Pulmonary infections requiring high-dose therapy
Target steady-state average colistin plasma concentrations (Css,avg) of 2-2.5 mg/L correlate with optimal microbiological outcomes while minimizing toxicity. Trough levels >2.5-3 mg/L are associated with increased nephrotoxicity risk.
Hack #2: If TDM is unavailable (common in resource-limited settings), clinical pharmacists can calculate predicted colistin exposure using published population pharmacokinetic models incorporating patient weight, creatinine clearance, and dosing regimen. Online calculators are available at https://www.dosing.org.
Nephrotoxicity: Prevention and Management
Colistin-associated nephrotoxicity occurs in 30-60% of treated patients, typically manifesting as acute tubular necrosis with onset 7-10 days after therapy initiation. Risk factors include:
- Concurrent nephrotoxins (vancomycin, aminoglycosides, NSAIDs)
- Baseline renal impairment
- Critical illness with hemodynamic instability
- Duration >7-10 days
- Cumulative dose >300 MIU
Pearl #3: Nephrotoxicity is generally reversible upon colistin discontinuation, with renal function recovery in 60-75% of patients. However, prevention remains paramount.
Nephroprotective Strategies
Recent evidence suggests several approaches to minimize nephrotoxicity:
-
Avoid concomitant nephrotoxins: Systematic review by Dai et al. demonstrated that concurrent vancomycin increases nephrotoxicity risk (OR 2.1, 95% CI 1.5-2.9). Consider alternative Gram-positive coverage when possible.
-
Optimize volume status: Adequate hydration (targeting euvolemia) reduces acute kidney injury risk. Avoid both volume depletion and overload in critically ill patients.
-
Limit treatment duration: Restrict colistin to 7-10 days when clinically feasible. Extended courses significantly increase nephrotoxicity incidence.
-
Ascorbic acid supplementation: Preliminary data suggest that vitamin C (1-2 g daily) may reduce oxidative stress-mediated tubular injury, though definitive trials are lacking.
Oyster #2: Colistin-induced nephrotoxicity often manifests with rising creatinine but minimal changes in urine output initially. Don't wait for oliguria to adjust dosing or consider alternative therapy.
Combination Therapy: Synergy or Dogma?
The role of combination therapy with colistin remains controversial. In vitro synergy studies demonstrate additive or synergistic effects when colistin is combined with carbapenems, tigecycline, rifampin, or fosfomycin against MDR Gram-negatives.
However, clinical trial evidence is mixed. The AIDA randomized controlled trial found no mortality benefit from colistin-meropenem combination versus colistin monotherapy for CRE bloodstream infections. Conversely, the INCREMENT-CPE cohort study suggested reduced mortality with combination therapy for high-risk infections.
Hack #3: Reserve combination therapy for:
- High bacterial burden infections (pneumonia, intra-abdominal sepsis)
- Septic shock with APACHE II >15
- Infections with MIC at susceptibility breakpoint (colistin MIC ≥2 mg/L)
- Acinetobacter or Pseudomonas with known heteroresistance
For uncomplicated urinary tract infections or low-burden bacteremia, colistin monotherapy may suffice.
Site-Specific Considerations
Pneumonia
Pulmonary colistin concentrations are notoriously variable, with epithelial lining fluid levels often subtherapeutic despite adequate plasma concentrations. This pharmacokinetic limitation has prompted interest in adjunctive nebulized colistin.
Pearl #4: For ventilator-associated pneumonia, consider combining intravenous colistin (standard dosing) with nebulized colistin (2-4 MIU every 8-12 hours) to achieve optimal lung parenchymal concentrations. Meta-analyses suggest improved clinical cure rates without increased systemic toxicity.
Central Nervous System Infections
Colistin achieves poor CSF penetration (<5% of plasma concentrations) even with inflamed meninges. For MDR Gram-negative meningitis or ventriculitis:
- Use maximum systemic doses (4.5 MIU every 12 hours)
- Consider intrathecal/intraventricular colistin (10 mg daily) in consultation with infectious diseases and neurosurgery
- Combination therapy with high-dose tigecycline or fosfomycin may improve outcomes
Urinary Tract Infections
Hack #4: For MDR Gram-negative cystitis or pyelonephritis without bacteremia, consider reduced colistin doses (2-3 MIU every 12 hours) since urinary concentrations far exceed plasma levels. This approach minimizes nephrotoxicity while maintaining efficacy.
Resistance Mechanisms and Heteroresistance
Colistin resistance arises through chromosomal mutations affecting lipopolysaccharide (LPS) modification or via plasmid-mediated mcr genes. Heteroresistance—where resistant subpopulations exist within predominantly susceptible populations—occurs in 10-40% of isolates and predicts treatment failure.
Oyster #3: Standard susceptibility testing may miss heteroresistance. If clinical failure occurs despite reported susceptibility, consider repeat cultures with extended incubation or population analysis profiling if available.
Future Directions and Novel Combinations
Emerging data on colistin potentiation strategies include:
- Colistin-taniborbactam combinations showing restored activity against mcr-positive isolates
- Phage-antibiotic synergy demonstrating enhanced killing
- Lipidated colistin derivatives with reduced nephrotoxicity
Practical Prescribing Algorithm
- Confirm indication: Reserve for infections with documented or high suspicion of MDR Gram-negatives with limited alternatives
- Administer loading dose: 9 MIU regardless of renal function
- Calculate maintenance dose: Based on CrCl using Garonzik nomogram
- Assess nephrotoxicity risk: Discontinue/minimize concurrent nephrotoxins
- Consider combination therapy: For high-risk infections or MIC ≥2 mg/L
- Monitor renal function: Creatinine every 2-3 days; adjust doses accordingly
- Implement TDM: If available, target Css,avg 2-2.5 mg/L
- Limit duration: Aim for 7-10 days; de-escalate to less toxic alternatives when possible
Conclusion
Colistin remains an essential antimicrobial for MDR Gram-negative infections, but optimal use requires sophisticated understanding of its complex pharmacology. Loading doses are mandatory, renal dose adjustments should target colistin (not CMS) exposure, and nephrotoxicity prevention strategies should be implemented proactively. As resistance continues evolving, judicious colistin use within comprehensive antimicrobial stewardship frameworks becomes increasingly critical.
Key References
-
Nation RL, et al. Colistin and polymyxin B: peas in a pod, or chalk and cheese? Clin Infect Dis. 2014;59(1):88-94.
-
Garonzik SM, et al. Population pharmacokinetics of colistin methanesulfonate and formed colistin in critically ill patients. Antimicrob Agents Chemother. 2011;55(7):3284-3294.
-
Paul M, et al. Combination therapy for carbapenem-resistant Gram-negative bacteria. J Antimicrob Chemother. 2014;69(9):2305-2309.
-
Shields RK, et al. Colistin does not potentiate ceftazidime-avibactam killing of carbapenem-resistant Enterobacteriaceae. Antimicrob Agents Chemother. 2018;62(8):e00721-18.
-
Vardakas KZ, et al. Colistin versus colistin-containing combinations for treatment of Gram-negative infections: systematic review and meta-analysis. Int J Antimicrob Agents. 2018;51(4):535-547.
-
Dalfino L, et al. High-dose, extended-interval colistin administration in critically ill patients: is this the right dosing strategy? Crit Care. 2012;16(3):R139.
-
Rigatto MH, et al. Risk factors for acute kidney injury in patients treated with polymyxin B: systematic review and meta-analysis. J Antimicrob Chemother. 2014;69(10):2600-2609.
-
Tsuji BT, et al. International Consensus Guidelines for the Optimal Use of the Polymyxins. Pharmacotherapy. 2019;39(1):10-39.
Comments
Post a Comment