Antimicrobial dosing in continuous renal replacement therapy

May 2002

Continuous renal replacement therapies (CRRT), such as continuous venovenous hemofiltration (CVVH), hemodialysis (CVVHD) and hemodiafiltration (CVVHDF), are routinely used to manage acute renal failure in the critically ill patient. Acute renal failure is associated with a high mortality rate, even if uremia is well controlled. Moreover, sepsis complicates disease progression and is a major predictor of poor outcomes. Optimizing drug therapy is essential in this patient population.

In vivo data on the removal of medication during chronic renal replacement techniques is limited. However, clinical research of antibiotic therapy during CRRT by renal, infectious disease and critical care specialists has been steadily increasing. This article focuses on important concepts and information regarding antibiotic therapy in patients receiving CRRT.

Dose adjustments

Before calculating an adjusted antimicrobial dose for a patient on CRRT, the antibiotic must be at least 25% renally eliminated since drugs cleared from the blood through nonrenal mechanisms will not be significantly affected by CRRT. There are two major factors and several minor factors that affect clearance of drug during CRRT. The first is protein binding. Only unbound drug is available for filtration by CRRT. Highly protein bound drugs have less removal by CRRT. The second is the volume of distribution of the drug. A large volume of distribution reflects a drug that is highly concentrated in body tissues. Since these agents are not particularly accessible to the systemic circulation, they are unlikely to be delivered to the renal replacement circuit in quantities sufficient to result in significant removal from the body. Other minor drug properties (molecular weight, drug charge, water or lipid solubility, dialysis membrane binding potential) only minimally alter drug elimination during CVVH.

Renal excretion is a frequent route of drug clearance for aminoglycosides, most cephalosporins, and fluoroquinolones. Volumes of distribution and protein binding are variable pharmacokinetic parameters with these agents, particularly in critically ill patients. Therefore, clinical investigations are our best predictor of antibiotic elimination during CRRT.

Source: Dorie W. Hoody, PharmD

Aminoglycosides and vancomycin warrant specific attention. Aminoglycosides are often used for gram-positive synergy and exhibit concentration-dependent killing. Effective dosing of these agents can be achieved by monitoring serum-drug concentrations during CVVH. Aminoglycoside therapy should be initiated with a loading dose. Redosing should occur once serum-drug concentrations are <1 mg/L (see Table 1).

Vancomycin is a large molecule, (1,400 daltons) but it is effectively removed during CRRT. It exhibits time-dependent killing but does not exhibit postantibiotic effects. Therefore, vancomycin trough concentrations should be closely monitored to assure prompt redosing once the desired serum trough concentration (typically 8-12 mg/L) has been achieved.

All antibiotics have not been studied during continuous renal replacement therapy. Therefore, mathematical equations using known pharmacokinetic and dynamic variables can be used to predict appropriate doses. B. Joos et. al. constructed a theoretical model for predicting the clearance of antibiotics during CVVH. The fraction of unbound drug, the fraction of nonrenal drug clearance, and total body clearance (expressed as ml/min) from manufacturer or historical literature is needed to calculate appropriate doses (see Table 2).

The fraction of unbound drug (F_u) is calculated using Equation 1. The adjustment for CVVH is calculated by multiplying F_u with the ratio of the rate of ultrafiltration blood flow (UFR) to total body clearance (Cl_n), which is then added to the fraction of nonrenal drug clearance (Q_o). Then this adjustment value is multiplied by the normal dose, resulting in the adjusted dose for CVVH, according to Equation 2.

This method is simple for continuous hemofiltrative techniques, but cannot be applied to continuous venovenous hemodialysis or hemodiafiltration. As shown in Table 1, experimental data of drug removal using hemodiafiltration suggest increasing the frequency of antimicrobial administration is needed to maintain adequate tissue concentrations.

Drug elimination during CVVHD or CVVHDF depends on dialysate flow rate, ultrafiltration flow rate and the sieving coefficient. Therefore, drug clearance during CVVHD or CVVHDF primarily depends on diffusive rather than convective forces. Dose adjustments for any CRRT can be calculated with application of sieving coefficient values, estimated clearance for the CRRT and a simple pharmacokinetic equation.

The first step in calculating an appropriate antibiotic dose for any form of CRRT is to identify through manufacturer data that the ratio of renal clearance to total body clearance is >0.25 (25%). Then the residual total body clearance of the patient must be estimated. Residual renal and hepatic function should be considered if the drug is eliminated through either of these pathways. Next, the drug sieving coefficient (Sc) should be obtained. Drugs that freely pass through dialysis membranes have a Sc of 1; drugs unable to permeate the membrane will have a Sc of 0. Most sieving coefficient values can be obtained from in vitro or in vivo (preferred) data from the manufacturer. If the Sc data is unavailable, the free fraction of drug (fu) can be used in place of Sc in Equation 4. Additionally, Sc can be calculated after obtaining the variables in Equation 3, but this is rarely done for one patient. Figure 1 describes the calculations needed for dose adjustments depending on the type of CRRT.

A number of other mathematical approaches have been used to estimate the maintenance dose for antiinfective agents during CRRT. However, the most appropriate method is to use published literature as summarized in Table 1. When using these recommendations or mathematical equations, patients should be closely monitored for drug effectiveness and toxicity with regimens adjusted accordingly. Until the results of further research work become available, physicians should be familiar with proposed equations and stay abreast of literature addressing drug dosing during CRRT.

For more information:

• Joos B, et al. Pharmacokinetics of antimicrobial agents in anuric patients during CVVH. Nephrol Dial Transplant. 11:1582-1585.
• Joy MS, Matzke GR, et al. A primer in continuous renal replacement therapy for critically ill patients. Annals of Pharmacotherapy 32:362-375.
• Davies JG, et al. Drug removal in continuous haemofiltration and haemodialysis. BJHM 54(10):524-528.