Therapeutic Drug Monitoring And Toxicology

Principles of therapeutic drug monitoring (TDM)

Therapeutic drug monitoring (TDM) is used for drugs with a narrow therapeutic index -- the gap between effective and toxic concentrations is small. The single most tested concept is sample timing. A trough is drawn immediately before the next dose (lowest concentration, used to confirm the drug stays above the minimum effective level and to avoid toxicity for renally cleared drugs). A peak is drawn at a drug-specific interval after the dose to confirm efficacy. Always document time of last dose and time of draw; mistimed samples are uninterpretable. Steady state is generally reached after about five half-lives.

Free drug is the active fraction. For highly protein-bound drugs such as phenytoin, hypoalbuminemia raises the free (active) fraction even when total drug looks normal -- a classic trap. Reference methods include immunoassay (EMIT, FPIA) and chromatography (HPLC, LC-MS/MS) for confirmation.

Major toxicology analytes

Acetaminophen (paracetamol) is the most important overdose. Plot the serum level against time on the Rumack-Matthew nomogram; a level above the treatment line means hepatotoxicity risk, and the antidote is N-acetylcysteine (NAC), which replenishes glutathione. Draw the level no sooner than 4 hours post-ingestion.

Salicylates (aspirin) cause a characteristic mixed acid-base picture: early respiratory alkalosis from stimulated breathing followed by a high-anion-gap metabolic acidosis. Treatment includes urine alkalinization with bicarbonate.

Ethanol is the most commonly measured drug of abuse; the legal driving limit is 0.08% (80 mg/dL). Methanol and ethylene glycol cause severe high-gap acidosis and an osmolal gap; antidote is fomepizole (or ethanol) plus dialysis.

Co-oximetry and metals

• Carboxyhemoglobin (COHb): carbon monoxide binds hemoglobin ~200x more avidly than oxygen; measured by co-oximetry. A standard pulse oximeter reads falsely normal because it cannot distinguish COHb. Treat with 100% oxygen.
• Methemoglobin: iron oxidized to Fe3+ cannot carry oxygen; causes chocolate-brown blood and cyanosis unresponsive to oxygen. Antidote is methylene blue.
• Lead: inhibits heme synthesis (raises FEP/ZPP, causes basophilic stippling); chelate with EDTA or succimer.

Drugs-of-abuse screening uses immunoassay with GC-MS or LC-MS/MS confirmation, since screening immunoassays cross-react and produce false positives (e.g., poppy seeds and opiates).

Pharmacokinetic calculations and correlations

The exam may give a half-life and ask for steady state or accumulation. After one half-life 50% of the drug remains; the practical rule is that steady state arrives at about five half-lives and that the same number of half-lives are needed for clearance after stopping a drug. Digoxin toxicity is a favorite correlation: it worsens with hypokalemia and hypomagnesemia because low potassium increases digoxin binding to the cardiac Na/K-ATPase, so a digoxin level must always be read alongside the potassium.

Phenytoin demonstrates nonlinear (saturable) kinetics -- a small dose increase near saturation can cause a large, disproportionate rise in serum level and sudden toxicity.

For phenytoin, valproate, and other highly albumin-bound drugs, a low albumin or uremia displaces drug from protein, so the total level looks therapeutic while the free (active) level is toxic; report or calculate free drug in these patients. A worked osmolal-gap toxicology case: a patient with severe high-anion-gap acidosis, an osmolal gap of 25, and crystalluria is presumptive ethylene glycol poisoning, treated urgently with fomepizole and dialysis. Choosing acetaminophen or simple lactic acidosis without accounting for the osmolal gap is the trap.

Carbon monoxide and methemoglobin both lower oxygen delivery while the arterial pO2 stays normal, because pO2 measures dissolved oxygen, not hemoglobin saturation -- a distinction frequently tested.

Specimen integrity and confirmatory testing

Toxicology depends heavily on preanalytical control. Ethanol must be drawn with a non-alcohol skin prep (use povidone-iodine, never an isopropyl swab) to avoid contaminating the specimen, and the tube should be filled to minimize headspace if a volatile is involved. Forensic and workplace specimens require a documented chain of custody; a break invalidates the result regardless of the analytical accuracy.

Screening immunoassays are designed for sensitivity and therefore cross-react -- amphetamine assays flag certain decongestants, and opiate assays flag poppy seeds -- so a positive screen always requires GC-MS or LC-MS/MS confirmation, the definitive identification method, before a result is reported as positive. The recurring exam principle is that a presumptive screen is never a final answer: timing, collection technique, chain of custody, and confirmatory testing must all be satisfied, and recognizing which step was violated is the judgment the toxicology items are built to assess.

Antidotes and key toxidromes to memorize

The exam pairs poisons with their antidotes and signature lab findings, so build a quick reference. Acetaminophen -> N-acetylcysteine, plotted on the Rumack-Matthew nomogram. Methanol/ethylene glycol -> fomepizole (or ethanol) plus dialysis, with a high anion gap and an osmolal gap; ethylene glycol additionally shows calcium oxalate crystals in urine. Methemoglobinemia -> methylene blue, with chocolate-brown blood. Carbon monoxide -> 100% oxygen (hyperbaric in severe cases), with elevated carboxyhemoglobin on co-oximetry. Organophosphates -> atropine plus pralidoxime, with depressed cholinesterase. Opioids -> naloxone.

Benzodiazepines -> flumazenil. Iron -> deferoxamine. Lead -> EDTA or succimer, with basophilic stippling and raised protoporphyrin. Digoxin toxicity -> digoxin-specific Fab fragments, made worse by hypokalemia. Linking the clinical toxidrome, the confirming laboratory pattern, and the correct antidote in a single step is the integrated correlation skill these therapeutic-drug-monitoring and toxicology questions are designed to reward.