Cotrimoxazole - optimal dosing in the critically ill

Pharmacokinetics

The volume of distribution (Vd) for each drug has been determined in healthy subjects with TMP having a much larger volume of distribution than SMX based on differences in lipid solubility [1]. There are only very limited data available on the impact of critical illness on the Vd of the two drugs, despite the widely recognized changes that occur with other antibiotics. In critically ill patients requiring mechanical ventilation for Pneumocystis carinii pneumonia, the Vd of TMP was 1.6 L/kg versus 1.4 L/kg for non-ventilated patients, and the Vd for SMX was 0.5 L/kg versus 0.4 L/kg [2]. In trauma patients, although reporting low Acute Physiology and Chronic Health Evaluation II (APACHE II) scores of 1 to 24, the Vd of TMP was found to be 2.1 L/kg and for SMX was 0.5 L/kg [3]. No studies have reported the magnitude of change in Vd of the drugs in patients with septic shock requiring large volume resuscitation or vasopressors. Data are also lacking on the impact of obesity on the Vd, and resulting dosing of the drug. Therefore, dosing regimens should be based on actual body weight.

Both TMP and SMX are eliminated from the body predominantly by renal excretion [1, 4]. Approximately 20% of SMX is metabolized in the liver to N₄-acetylsulfamethoxazole which is subsequently excreted in the urine [5]. N₄-acetylsulfamethoxazole lacks relevant antibacterial activity [1]. The remainder of SMX is cleared by the kidney as unchanged drug [4]. The renal excretion of SMX is increased when the urine is alkaline [5]. Similarly, only a small portion of TMP (10 to 20%) is metabolized by the liver to inactive metabolites, which are subsequently conjugated and excreted in the urine [1]. The remaining portion of TMP elimination is via renal secretion of unchanged drug [4]. Unlike SMX, TMP renal clearance is increased with acid urine [1].

Renal dysfunction

The total amount of SMX (unchanged and metabolites) recovered in the urine of patients with renal impairment remains similar to patients with normal renal function (approximately 80% of an oral dosage), but the portion of the drug remaining unchanged decreased as renal function decreased, suggesting increased metabolism as renal function deteriorates [8]. The magnitude of the increase in clearance can not be estimated clinically or from published literature, but may be of relevance in producing drug concentrations equivalent to patients with normal renal function. The lack of accumulation of active SMX with renal impairment has important therapeutic ramifications since the metabolite does not have antimicrobial activity. If the SMX dosage is reduced when treating patients with renal impairment, the concentration of the active unchanged SMX may decrease, since its clearance is not impaired and may even increase, and potentially compromise the therapeutic effect.

Both unchanged drug and the conjugated metabolites of TMP will accumulate as renal function decreases unless a dosage reduction is implemented.

Liver dysfunction

The pharmacokinetics of both SMX and TMP are not altered significantly by liver impairment [5]. The half-life of both SMX and TMP was not prolonged, versus normal subjects, in patients with mild or severe liver dysfunction, although the characterization of the liver function was poorly defined [21].

The metabolism of SMX appears to be influenced by the activity of N-acetyltransferase 2 (NAT2), with the existence of genetic polymorphism altering the activity of the enzyme [22]. Patients without the native NAT2*4 allele do not clear SMX as rapidly resulting in accumulation of drug and increased exposure [22]. However, the published research methodology does not describe the ability to differentiate active unaltered SMX from the various metabolites. Since they found a correlation between drug accumulation and creatinine clearance, the potential that the finding of increased exposure in patients without native NAT2*4 could be due to accumulation of inactive SMX metabolites. No data exist to describe the effect of critical illness on activity of NAT2 or the ramifications of dosing of patients with the various polymorphisms. Since the individual patients’ genetic makeup is not usually known at the time of initiating TMP/SMX therapy, currently the clinician is unable to assess or consider the need for dosage adjustments.

Dialysis

Similar to findings in non-dialysis dependent renally impaired patients, unchanged SMX concentrations do not increase in patients with renal impairment sufficient to necessitate dialysis [8]. However, the concentrations of TMP and of SMX metabolites increased in dialysis-dependent patients compared to patients with severe non-dialysis dependent renal impairment [8]. Surprisingly, given the drug’s prolonged use, only limited description of the effect of dialysis on removal of each drug during a dialysis session is available; none for critically ill patients. Pharmacokinetic evaluation in a single patient during a hemodialysis session suggested efficient removal of both antibiotics, with total (unchanged and metabolites) SMX clearance increasing approximately 3.5-fold over non-dialysis periods (83.9 versus 24.4 ml/min) [23]. Approximately 50% of total (unchanged and metabolites) SMX and TMP body stores was removed by a four hour dialysis with a 1 m² cuprophane dialysis membrane at a blood flow rate of 200 ml/min and dialysate flow rate of 500 ml/min [24]. These limited data suggest that hemodialysis is effective in removing SMX metabolites and TMP. However, since active SMX does not accumulate with standard dosing in dialysis patients, the optimum dosing of the combination becomes challenging. Dosage reduction may compromise the activity of unchanged SMX while standard dosages will result in accumulation of potentially toxic SMX metabolites and TMP. A potential approach would be to provide full dosages for the first 24 to 48 hours with subsequent dosage reduction based on clinical response [25].

SMX and its metabolites are partially removed by continuous ambulatory peritoneal dialysis [5], although the amount cleared would be anticipated to be clinically insignificant [23, 26]. Instillation of 2 L of 2.5% glucose peritoneal dialysis solution four times daily contributed only 7% of total TMP body clearance and 8% of unchanged SMX clearance in chronic CAPD patients [27]. SMX metabolites are similarly poorly removed by CAPD as demonstrated by a low clearance rate found with four times daily dwells [16]. Rapid cycling peritoneal dialysis at a rate of 3 L/hour efficiently (60%) removed SMX, but had minimal effects on TMP clearance [28]. Hence, peritoneal dialysis should not be considered to reduce the potential for accumulation of either drug to toxic concentrations, and the remaining renal function of the patient should guide dosing.

Removal rates during continuous veno-venous hemodiafiltration (1.8 m² high-flux polysulfone filter at blood flow rates of 180 ml/min, dialysate flow rate of 1,500 ml/hour and fluid replacement of 1,500 ml/hour) produced clearance rates for both drugs similar or exceeding normal renal function [29]. Extended daily dialysis (1.3 m² high flux polysulfone filter at blood flow rates of 140 ml/min, dialysate flow of 170 ml/min for 440 minutes) resulted in substantial clearance of TMP, with less clearance of SMX, and no report of clearance of SMX metabolites [30]. These authors caution clinicians regarding dosing reduction during continuous veno-venous hemodiafiltration and the potential for such reduction to result in inadequate drug concentrations [30].

Toxic concentrations

Pharmacodynamics

Both TMP and SMX act by blocking steps in the production of metabolically active folate [53]. Sulfonamides act by blocking the production of dihydrofolate from precursor components (p-aminobenzoate, pteridine, and glutamate) [54]. This inhibition is irrelevant for human metabolism since humans can utilize dihydrofolate absorbed from the diet while bacteria must manufacture the dihydrofolate from precursors. TMP works on a subsequent step of folate metabolism by competitively inhibiting dihydrofolate reductase, the enzyme responsible for production of the active tetrahydrofolate [54]. Humans are not adversely affected by the TMP-induced blockage of dihydrofolate reductase since this enzyme in bacteria is greater than 1,000-fold more sensitive than human dihydrofolate reductase. Therefore, at concentrations achieved with therapeutic dosages, two steps in folate metabolism within bacteria are blocked, while human metabolism is unaffected.

Surprisingly, the determination of the pharmacodynamic parameter (concentration or time) most related to bacterial kill has not been clearly determined [55]. TMP/SMX has variable bacteriostatic and bacteriocidal action. Early work suggested that TMP/SMX has bacteriocidal activity against gram positive organisms, such as Staphylococcus aureus[56]. What little available evidence evaluating concentration versus time killing suggests that TMP/SMX has concentration-dependent activity against gram-positive organisms, such as Staphylococcus aureus[54, 57]. For gram-negative organisms, it has been hypothesized that TMP/SMX has concentration-dependent killing of E. coli, although supporting data are lacking [57]. Early investigations suggested only bacteriostatic activity with concentrations achievable in urine due to interaction with nutrients which inhibit the activity of the drugs [58]. Subsequent limited evaluation suggests that TMP/SMX is bacteriostatic against sensitive E. coli at concentrations of 2 mcg/ml TMP and 40 mcg/ml SMX, supporting time-dependent killing [59]. However, others have shown that at TMP concentrations greater than 2 mcg/ml (>12 mcg/ml SMX) bacteriocidal activity can be detected [20].

For other organisms, the pharmacodynamics are even less clear. Limited evidence from Burkholderia pseudomallei treatment suggests concentration effects through achievable in vivo concentrations [60]. This concentration-dependent killing was not long standing, and converted to bacteriostatic effects when studied using time-kill methodology [60]. Only bacteriostatic activity could be demonstrated against Nocardia spp. at TMP/SMX concentrations of 3.4/17 mg/L; a concentration achievable with a 7.5 mg/kg/day TMP dose) [61].

In summary, TMP/SMX possesses bacteriostatic activity, which is time dependent, with the potential for concentration dependent bacteriocidal activity for specific organisms, possibly in selected infected sites within the body. Since the pharmacodynamics are unclear, clinicians should select dosages that would be anticipated to produce adequate concentrations, as described by successful therapeutic studies, throughout the dosage interval.

Target concentrations

Dosing recommendations in the critically ill

The clinician is frequently placed in a position of uncertainty regarding the MIC of the infecting organism, and the appropriate dose to achieve the target concentrations for optimal pharmacodynamics, and resulting efficacy. Clinical trials to demonstrate the optimal dosing regimens, as discussed above, are generally lacking. The benefit of prospective therapeutic drug monitoring of TMP/SMX concentrations in critically ill patients has not been evaluated in the published literature. Intuitively, therapeutic drug monitoring would be anticipated to be beneficial. However, the capability of most clinical laboratories to provide such a service is limited. Therefore, the clinician is forced to extrapolate the available pharmacodynamic and pharmacokinetic information available to select a dosage regimen. Due to the lack of published literature on the effects of critical illness and treatment on the pharmacokinetics of TMP/SMX, the clinician must attempt to estimate dosing requirements. Similarly, the lack of data on the impact of obesity on dosing requirements forces the clinician to base dosing on actual body weight to ensure achievement of desired drug concentrations. The urgency of obtaining effective concentrations in a patient who is critically ill from an infection warrants the use of doses larger than in patients with limited symptoms from the infection. By utilizing the limited available data on pharmacokinetics, pharmacodynamics, and potential for dose-related toxicities, the following doses are recommended for achieving the desired target concentrations for potential pathogens in critically ill patients.

For patients with Pneumocystis jovenii, Stenotrophomonas maltophilia, Burkholderia pseudomallei or Nocardia spp. infection - 15 mg/kg/day TMP IV or PO (with corresponding 75 mg/kg/day SMX).

(The drug could be given in two divided dosages based on the pharmacokinetics, but the relationship between Cmax concentrations and toxicity has been only evaluated in a limited non-critically ill population [7]. Whether the toxicity is more closely correlated with the Cmax or total exposure to the drugs is not known. To potentially minimize the impact of the Cmax on development of toxicities, the drug can be administered in smaller doses more frequently, three to four per day.)

For methicillin-resistant Staphylococcus aureus (MRSA) or extended-spectrum β-lactamase-producing (ESBL) Enterobacteriaceae - 160 mg TMP IV or PO twice daily (with corresponding 800 mg SMX).

Greater than 30 ml/min - full dose;

Fifteen to 30 ml/min - half dose (the concentration of active SMX may be reduced, relying on TMP activity only);