Therapeutic Drug Monitoring

The basic assumptions underlying therapeutic drug monitoring are that drug metabolism varies from patient to patient and that the plasma level of a drug is more closely related to the drug's therapeutic effect or toxicity than is the dosage.

Indications for drug monitoring

  • Drugs with a narrow therapeutic index (where therapeutic drug levels do not differ greatly from levels associated with serious toxicity) should be monitored. Example: Lithium, phenytoin, digoxin.
  • Patients who have impaired clearance of a drug with a narrow therapeutic index are candidates for drug monitoring. The clearance mechanism of the drug involved must be known. Example: Patients with renal failure have decreased clearance of digoxin and therefore are at a higher risk of toxicity.
  • Drugs whose toxicity is difficult to distinguish from a patient's underlying disease may require monitoring. Example: Theophylline in patients with chronic obstructive pulmonary disease.
  • Drugs whose efficacy is difficult to establish clinically may require monitoring of plasma levels. Example: Phenytoin.

Situations in which drug monitoring may not be useful

  • Drugs that can be given in extremely high doses before toxicity is apparent are not candidates for monitoring. Example: Penicillin.
  • If there are better means of assessing drug effects, drug level monitoring may not be appropriate. Example: Warfarin is monitored by measuring INR, not by serum levels.
  • Drug level monitoring to assess compliance is unreliable, since poor compliance cannot be distinguished from rapid metabolism without direct inpatient scrutiny of drug administration.
  • Drug toxicity is a clinical diagnosis. Drug concentrations within the usual therapeutic range do not rule out drug toxicity in a given patient. Example: Digoxin, where other physiologic variables (eg, hypokalemia) affect drug toxicity.

In summary, therapeutic drug monitoring may be useful to guide dosage adjustment of certain drugs in certain patients. Patient compliance is essential if drug monitoring data are to be correctly interpreted.

Pharmacokinetic Parameters

Five pharmacokinetic parameters that are important in therapeutic drug monitoring include:

Bioavailability. The bioavailability of a drug depends in part on its formulation. A drug that is significantly metabolized as it first passes through the liver exhibits a marked "first-pass effect," reducing the effective oral absorption of the drug. A reduction in this first-pass effect (eg, because of decreased hepatic blood flow in heart failure) could cause a clinically significant increase in effective oral drug absorption.

Volume of distribution and distribution phases. The volume of distribution of a drug determines the plasma concentration reached after a loading dose. The distribution phase is the time taken for a drug to distribute from the plasma to the periphery. Blood taken before completion of a long distribution phase may not reflect levels of pharmacologically active drug at sites of action. Examples: Digoxin, lithium.

Clearance. Clearance is either renal or non-renal (usually hepatic). Whereas changes in renal clearance can be predicted on the basis of serum creatinine or eGFR, there is no routine liver function test for assessment of hepatic drug metabolism.

Half-life. The half-life of a drug depends on its volume of distribution and its clearance and determines the time taken to reach a steady state level. After a period of 3 or 4 half-lives, the serum drug concentration will be 87.5% to 93.75% of the steady state value. Patients with decreased drug clearance and therefore increased drug half-life will take longer to reach a higher steady state level. In general, since non-steady state drug levels are difficult to interpret, therapeutic drug monitoring usually involves measurement of drug levels at steady state.

Protein binding of drugs. All routine drug level analysis involves assessment of both protein-bound and free drug. However, pharmacologic activity depends on only the free drug concentration. Changes in protein binding (eg, in renal failure or hypo-albuminaemia) may significantly affect interpretation of reported levels for drugs that are highly protein-bound. Example: Phenytoin. When the ratio of active to total drug is increased, the therapeutic range based on total drug level will not apply.

Drug Interactions

For patients receiving several medications, the possibility of drug interactions affecting drug metabolism must be considered. Example: Quinidine decreases digoxin clearance.

Time to sample blood to measure Drug Concentration

In general, the specimen should be drawn after steady state is reached (at least 4 half-lives after a dosage adjustment) and just before the next dose (trough level).
Peak and trough levels may be indicated to evaluate the dosage of drugs whose half-lives are much shorter than the dosing interval. Example: Gentamicin.

Drug monitoring may be performed in three main ways.

  1. The effect of a drug may be monitored by its clinical effect eg lowering of blood-pressure, the prophylaxis of migraines or the reduction in inflammation by steroids.
  2. The biological effect of the drug may be followed by its biochemical effects eg glucose modulation by insulin, reduction in CRP or plasma viscosity with anti-inflammatory agents, the lowering of uric acid with allopurinol or increase in prothrombin time by warfarin.
  3. A small number of therapeutic agents may be usefully monitored by their plasma concentrations. These measurements are only of value if the plasma values reflect the biological/therapeutic effect of the drug. This value is illustrated particularly well by digoxin and phenytoin both of which have side effects that are similar to the conditions which they are used to treat (phenytoin toxicity may produce fits and digoxin toxicity may produce cardiac dysrhythmias). The measurement of other drugs such as lithium and theophylline is useful as the concentrations reflect both under- or over-therapy and also correlate with toxic biological effects.
 
drug half-life in adults
target plasma concentrations
timing of sample
time taken for steady state to be established in adults
route of elimination
Carbamazepine
5-27 h
4-12 mg/L (monotherapy)
4-10 mg/L (combination therapy)
trough sample
at least 7 days
hepatic
Digoxin
40 h
0.5-1.0 microgm/L
at least 6 h after last dose
5-7 days
renal
Lithium
14-33h
0.4-1.0 mmol/L
at least 12 hr after last dose
3-7 days
renal
Phenobarbitone
 
10-40 mg/L
trough sample
10-25 days
75% hepatic
Phenytoin
20-40 h
10-20 mg/L
trough sample
8-50 days
80% hepatic 20% renal
Theophylline
smokers 4.4 h
10-20 mg/L (adults)
5-10 mg/L (neonates)
trough sample
2 days
90% hepatic 10% renal
Valproate
 
 
trough sample
30-85 h
 

Useful Points

  • If patients take their digoxin or lithium tablets in the evening or at midday, then all samples taken in a morning surgery will be taken at the correct time.
  • Digoxin toxicity is potentiated by hypothyroidism and low concentrations of potassium.
  • Steady state conditions with stable plasma levels of a drug do not occur until at least 4.5 elimination half- lives have elapsed after either initiation of therapy or a change in dose. It is not possible to interpret levels measured after a shorter time as the drug is still being distributed throughout the body.

Page updated: 20/04/10 | Updated by: Dr. Julian Barth