Therapeutic Drug Monitoring: Case-based PK/PD, case interpretations, and clinical decision-making

Administrative setup and classroom dynamics
The session opens with the lecturer checking that everyone is attentive and comfortable. He mentions administrative logistics: a QR attendance code will be used later in the workshop rather than at the start to prevent students from scanning and leaving immediately. A secretary will ensure the code is not up at the beginning, and there will be a sign-off list for data or QR-code issues. He emphasizes patient engagement and intends to keep the session interactive, aiming to make therapeutic drug monitoring (TDM) engaging despite the inherent dryness of “drugs.” He notes this may be his last workshop in this particular series, foreshadowing a transition to other topics, possibly for later years. He invites questions but also cautions that pharmacology questions require understanding mechanisms of action to be worthy of marks. He also stresses practical exam strategy: focus on differential reasoning rather than memorizing every fact. He asks for attendance, ensures audio is working, and sets expectations for participation. He encourages students to stay until the end for a coherent case-based exploration of core concepts, including mechanisms, monitoring, and clinical decision-making. He mentions previously completed questions and uses it as a bridge to today’s cases.

Key concepts introduced for the session
The lecturer frames pharmacology around several core ideas: (1) not all questions can be answered by rote; understanding mechanism of action and pharmacokinetic principles is essential to derive sensible differentials and therapeutic plans; (2) some drugs display enzyme induction or inhibition, have narrow therapeutic indexes, or require therapeutic drug monitoring (TDM); (3) in pharmacology, unlike other medical disciplines, small changes in dose can produce disproportionately large changes in drug exposure for drugs with saturable metabolism (zero-order kinetics); (4) timing of sampling is critical in TDM, especially for agents with non-linear kinetics; trough concentrations (lowest concentration just before the next dose) often inform toxicity risk and dosing adequacy, whereas peak concentrations relate to efficacy for concentration-dependent antibiotics; (5) drug binding to plasma proteins (e.g., albumin) affects the pharmacologically active unbound fraction; in hypoalbuminemia, the unbound fraction can increase, altering interpretation of total drug levels.

Foundational pharmacokinetic concepts for TDM

First-order vs zero-order kinetics: In first-order kinetics, the elimination rate is proportional to concentration, described by
\frac{dC}{dt} = -k C, \quad k = \frac{Cl}{Vd} where $k$ is the elimination rate constant, $Cl$ is clearance, and $Vd$ is the volume of distribution. In zero-order kinetics, elimination proceeds at a constant rate independent of concentration:
\frac{dC}{dt} = -k_0
These differences matter: zero-order kinetics (saturable metabolism) can make small dose changes produce large concentration shifts; the lecturer uses the alcohol analogy to illustrate saturation: once the liver’s capacity is overwhelmed, metabolism cannot keep up, leading to disproportionate increases in plasma concentration.
Half-life and steady state: For most drugs, steady state is achieved after about
t{ss} \approx 4-5\, t{1/2} \,,
where $t{1/2}$ is the half-life. For phenytoin in the case discussed, the half-life is given as approximately t{1/2} \approx 22\ \text{hours}. Changes in dose under zero-order kinetics can alter concentrations even at steady state, because the system can saturate at higher concentrations.
Therapeutic drug monitoring (TDM) framework: TDM is particularly relevant for drugs with narrow therapeutic windows, significant inter-individual variability, or important drug–drug interactions. The aim is to optimize exposure by considering pharmacokinetics (absorption, distribution, metabolism, excretion) and pharmacodynamics (drug effect) in the clinical context.
Drug binding and albumin correction: Many drugs are highly protein-bound. Only the unbound fraction is pharmacologically active. Total drug level reflects both bound and unbound drug; in hypoalbuminemia, the unbound fraction rises, potentially increasing pharmacologic effect or toxicity. Clinicians may adjust interpretation via correction factors or by measuring unbound drug in selected cases; labs often provide the correction as part of the report.
Conceptual distinction: clinical toxicity vs laboratory toxicity is context-dependent. Therapeutic ranges are probabilistic, derived from populations, and must be interpreted alongside patient factors (age, organ function, comorbidities, concurrent medications).

Case 1: 43-year-old male with epilepsy on phenytoin and fluoxetine
Clinical vignette and key points
A 43-year-old man with long-standing epilepsy presents with ataxia and slurred speech. He is known to be on phenytoin 200 mg daily and recently started on fluoxetine 20 mg daily for depression (a week prior). He is brought to the emergency center with nausea, vomiting, and confusion. Age is highlighted as an important red flag for risk stratification in pharmacology because it shapes baseline physiology and drug handling. The patient has a history of epilepsy from childhood (diagnosed at age seven), which influences the differential: in a young patient, common acute etiologies such as metabolic disturbances or infections may predominate, while in a long-standing epileptic with chronic therapy, drug interactions or toxicity become central considerations.

Key pharmacology and mechanism points

Phenytoin is a classic antiepileptic whose kinetics can be saturable (zero-order) at therapeutic or supra-therapeutic levels. The patient’s dose is 200 mg daily, which the lecturer notes as not particularly high for some clinical contexts, but the case deliberately uses potentially flawed data to illustrate reasoning.
Phenytoin exhibits enzyme interactions and can be subject to both induction and inhibition, affecting levels of concomitant medications via hepatic metabolism. Fluoxetine, a selective serotonin reuptake inhibitor (SSRI) metabolized in the liver, can interact with phenytoin through shared hepatic enzyme pathways (cytochrome P450 system). The lecturer notes the metabolic involvement of CYP enzymes (specifically CYP2D6 for fluoxetine) and highlights the potential for clinically significant interactions between phenytoin and fluoxetine via hepatic metabolism, requiring concentration monitoring.
The session emphasizes that when managing a patient on phenytoin who develops confusion or ataxia, one should consider differential diagnoses and potential drug interactions, plus non-drug etiologies (electrolyte disturbances, hypoglycemia, head injury, cancer with seizure). The first-order vs zero-order kinetics concept matters because dose changes can dramatically affect concentration under saturable metabolism.

Clinical decision-making steps and interpretation

Steps for the clinician include prioritizing stabilization, checking for potential dosing errors, assessing organ function (liver, in particular, for phenytoin and fluoxetine metabolism), and evaluating other drugs and interactions.
The lecturer highlights the importance of measuring a phenytoin concentration to guide management and to decide on dose adjustments. The presented case yields an elevated phenytoin level: $C_{phenytoin} = 51$ (in whatever units the course uses, typically mg/L for phenytoin). The therapeutic range given in the slide is [10,20], so a measured value of 51 is interpreted as toxic.
Timing of sampling is crucial for interpretation in zero-order kinetics. The last dose was at 19:00, and the sample was taken at 17:32 the next day. In zero-order kinetics, timing relative to dosing significantly affects interpretation: trough samples (just before the next dose) are most informative for toxicity risk, whereas random timing can mislead.
The half-life of phenytoin is around t_{1/2} \approx 22\ \text{hours} in this discussion, affecting how soon troughs equilibrate and how to interpret trough concentrations. If a trough is taken during or soon after a dose, levels can appear spuriously high.
The concept of “loading dose” did not arise for phenytoin in this particular case, but the general TDM framework recognizes that if a patient has a high concentration, one should evaluate dosing, organ function, and interactions before adjusting therapy.

Practical implications and clinical pearls from the case

The patient also has a depression diagnosis and is taking fluoxetine, which is relevant because fluoxetine can affect hepatic metabolism of other drugs. The presence of concurrent antidepressant therapy should prompt consideration of hepatic enzyme interactions and possible increases in phenytoin exposure.
When interpreting a phenytoin trough near or in the supra-therapeutic range, consider whether the sample time was truly trough, whether hepatic function might be altered, possible interactions, and whether measurement was a total vs. free phenytoin concentration. Hypoalbuminemia can significantly affect free phenytoin levels; while total level can be misleading, unbound phenytoin concentration is more directly related to toxicity risk.
The management logic includes stabilizing the patient, reviewing the medication list for interactions and dose issues, aligning sampling timing with pharmacokinetic principles, and adjusting therapy accordingly while monitoring for improvement or persistence of toxicity.

Case 1: key concepts reinforced in the discussion

Zero-order kinetics and saturable metabolism can magnify the effect of small dose changes on drug exposure.
Timing of concentration measurement is critical for accuracy in interpreting phenytoin levels.
Drug–drug interactions (phenytoin with fluoxetine) can meaningfully alter drug exposure and should be anticipated in planning treatment.
The role of albumin in drug binding can influence interpretation of total drug levels; unbound levels are most clinically relevant but are less commonly measured due to cost.

Case 2: Digoxin in a 61-year-old with heart failure and atrial fibrillation
Clinical vignette and setup
A 61-year-old female with heart failure and AF is on a regimen that includes digoxin and furosemide, with other background optimizers such as ACE inhibitors and beta-blockers. She presents with confusion and color vision disturbances. Digoxin is a drug with a very narrow therapeutic window and a long history of use for selected patients with AF and heart failure. The lecturer uses the case to emphasize how TDM applies to drugs with narrow therapeutic indices and how electrolyte disturbances and renal function influence digoxin toxicity and efficacy.

Mechanism and pharmacodynamics: why electrolytes matter

Digoxin acts as a sodium–potassium ATPase (Na+/K+-ATPase) inhibitor with high affinity for the potassium binding site. Inhibition increases intracellular calcium via reduced activity of the Na+/Ca2+ exchanger, enhancing calcium entry and increasing cardiac contractility (positive inotropy). This mechanism benefits patients with heart failure and AF who require improved cardiac output.
Because digoxin’s action is tied to intracellular calcium, the electrolyte milieu, especially potassium and magnesium, modulates its effect and toxicity. Hypokalemia increases digoxin binding and toxicity risk because less competing potassium reduces site availability, effectively increasing digoxin’s binding to Na+/K+-ATPase. Conversely, hyperkalemia reduces digoxin binding and can mitigate some of its effects, but hyperkalemia itself is dangerous and can complicate management.
The clinical takeaway is that digoxin toxicity risk rises with hypokalemia and renal impairment, and management must address electrolyte abnormalities in addition to adjusting digoxin dosing.

Pharmacokinetics and monitoring strategy for digoxin

Digoxin has a narrow therapeutic window, with the lecture illustrating a reduced therapeutic range of [10,20] (units not explicitly specified in the transcript, but historically digoxin is often discussed around ng/mL). The high value (51) in the case clearly indicates toxicity in this teaching scenario.
The drug’s pharmacokinetics and monitoring must consider steady-state dynamics. It typically takes about 5 \text{ half-lives} to reach steady state for many drugs, but with digoxin, steady-state estimation and monitoring are still essential due to its narrow window and renal clearance dependence.
The case emphasizes that digoxin concentration alone does not determine management; the clinical picture and electrolyte status are equally important. If a digoxin level is high but the patient is clinically stable, treatment decisions may differ from cases with high levels plus toxicity symptoms.
The mechanism slide highlights the sodium–potassium ATPase binding paradigm to explain why electrolyte disturbances modify digoxin effect and toxicity risk. A low potassium level increases digoxin binding and toxicity risk; thus, correcting potassium and other electrolytes is a first-line corrective step in digoxin toxicity.

Interpretation of the digoxin level and practical management

When analyzing digoxin levels, consider timing relative to the last dose and whether the patient is at steady state. In the lecture, steady-state timing is reinforced by the idea that a trough at steady state is most informative for assessing ongoing exposure and safety.
The discussion also covers the importance of signs and symptoms (e.g., color vision changes, confusion) that accompany digoxin toxicity, reinforcing that laboratory values must be interpreted in the clinical context.
If toxicity is suspected, management begins with correcting electrolytes (noting that hypokalemia increases digoxin binding and toxicity), followed by dose adjustments if needed, and repeated monitoring of digoxin levels and renal function.
The relationship between digoxin pharmacodynamics and electrolytes is summarized with the conceptual statement that reduced potassium increases digoxin’s effect because digoxin binds to Na+/K+-ATPase more readily when extracellular potassium is low.

Case 2: take-home messages

Digoxin requires careful monitoring due to its narrow therapeutic window and reliance on renal clearance. Electrolyte status, particularly potassium, must be managed to avoid exacerbating toxicity.
The therapeutic approach combines measurement of digoxin concentration with evaluation of clinical status and electrolyte levels, with particular attention to potential interactions with diuretics and other medications that influence electrolytes.
Clinicians should not interpret digoxin levels in isolation; the clinical context and laboratory data (electrolytes, renal function) drive management decisions.

Case 3 and Case 4: Amikacin vs. Vancomycin – PK/PD concepts, MIC, and practical monitoring
Clinical vignette and setup
A hospital case involves a patient with a severe infection, treated with amikacin (an aminoglycoside) and piperacillin–tazobactam. The case then pivots to a discussion of vancomycin in similar contexts. The emphasis is on therapeutic drug monitoring and the pharmacokinetic/pharmacodynamic (PK/PD) distinctions that drive dosing decisions for these antibiotics.

Amikacin: concentration-dependent killing and PK/PD targets

Amikacin is an aminoglycoside that exhibits concentration-dependent killing and a post-antibiotic effect. Its efficacy hinges on achieving a high peak concentration relative to the minimum inhibitory concentration (MIC) of the target organism. The MIC is the minimum concentration required to inhibit visible growth of the organism after a defined incubation period.
The key PK/PD concept is: the antibiotic kills best when the peak concentration exceeds the MIC (i.e., high $C_{max}$ relative to MIC). The post-antibiotic effect means that even after drug levels fall below MIC, bacterial suppression may continue for a time, which influences dosing intervals.
In practice, trough concentrations are used primarily to assess toxicity risk and renal function, whereas peak concentrations are used to assess efficacy in concentration-dependent antibiotics like amikacin. For patients with renal impairment, dosing intervals are often extended (e.g., dosing every several days rather than daily) to accommodate reduced clearance while striving to maintain an effective peak above MIC.
The case explains that the trough level (e.g., $C{ trough }$) is used to monitor toxicity and renal status, while peaks ($C{max}$) are used to assess whether the drug exposure has reached the level necessary to achieve bacterial kill. If the trough is below the target or if renal function is impaired, dosing intervals or amounts must be adjusted accordingly. The concept of the post-antibiotic effect supports maintaining adequate exposure without unnecessarily high troughs.

Vancomycin: time-dependent killing and AUC/MIC considerations

Vancomycin is time-dependent rather than concentration-dependent, meaning its efficacy correlates with the duration that drug concentrations stay above the MIC rather than the peak level alone. Target exposure is commonly expressed as AUC/MIC (area under the concentration–time curve over 24 hours divided by MIC). The AUC/MIC target is used to optimize both efficacy and toxicity risk, with higher AUC associated with better killing up to a point before nephrotoxicity risk increases.
A loading dose is often given with vancomycin to rapidly achieve therapeutic levels, especially in severe infections like sepsis or pneumonia where time to efficacy is critical. After loading, maintenance doses are adjusted to maintain concentrations within the therapeutic window, taking into account renal function and the MIC reported by microbiology.
In patients with renal impairment, maintenance dosing intervals are lengthened and dose amounts may be reduced to achieve an appropriate exposure while avoiding accumulation. The loading dose is generally maintained because it helps reach effective concentrations quickly, reducing the time needed to achieve therapeutic exposure.
The AUC/MIC concept requires accounting for renal function variability and dynamic physiological changes in critically ill patients, which may necessitate frequent monitoring and dose adjustments during the course of treatment.

Key practical ideas from Case 3/4

Sampling strategy depends on the PK/PD properties of the drug: for aminoglycosides like amikacin, trough monitoring is essential for safety, while peak monitoring is used to confirm efficacy when clinically indicated. For vancomycin, troughs were historically used, but modern practice often emphasizes AUC-guided dosing to balance efficacy and nephrotoxicity risk; loading doses help reach target exposure quickly.
MIC is central to interpretation of susceptibility testing. If the MIC is low and the organism is susceptible, a lower exposure may be effective; if MIC is high (intermediate or resistant), achieving effective exposure may require higher doses or may be contraindicated due to toxicity risk.
In renal impairment, dosing strategies shift to longer intervals and reduced maintenance doses, while loading doses may still be used to reach therapeutic targets rapidly. The practical takeaway is that TDM for antibiotics is dynamic and context-dependent, requiring frequent reassessment and integration of lab results (MIC, trough/peak values, AUC estimates) with clinical status.

Clinical and philosophical points about TDM and interpretation

The lecturer stresses the difference between statistical significance and clinical significance, noting that statistically significant changes in a parameter (e.g., blood pressure or lab value) do not automatically justify changing management if the clinical impact is negligible. This is an important caution in interpreting PK/PD data and trial results.
The importance of context is emphasized repeatedly: patient age, organ function (liver, kidney), nutritional status (albumin), comorbidities, and polypharmacy all modulate pharmacokinetics and pharmacodynamics, affecting interpretation of levels and decisions about dosing and monitoring.
A recurrent theme is that TDM is a tool to guide clinical judgment, not a substitute for it. Interpreting levels requires considering timing, steady-state status, organ function, potential drug interactions, and the patient’s clinical trajectory. When uncertain, labs are valuable but must be interpreted in the clinical context rather than in isolation.

Closing notes and QR code logistics

The instructor returns to logistics: a QR code for attendance will be displayed at the end of the session after a brief pause so that the secretary can ensure it is not displayed during the lecture. He asks students to wait approximately 30 seconds before scanning to avoid issues, and he mentions a sign-in sheet for data concerns.
He ends by pointing out that this is his last session in this series, and he invites students to see him later in oncology or other settings if their paths cross, signaling a transition in roles rather than an end to learning.

Summary of key takeaways for the exam and clinical practice

For phenytoin: be mindful of zero-order kinetics, potential interactions with psychotropics (e.g., fluoxetine) via hepatic metabolism, and the importance of corrected interpretation in the context of albumin binding. Concentration monitoring should be timed to obtain trough measurements to assess toxicity risk in saturable kinetics.
For digoxin: recognize the narrow therapeutic window and the central role of electrolytes, particularly potassium, in toxicity risk. Digoxin toxicity presents with neuro-visual symptoms (e.g., color vision changes) and cognitive changes; management includes correction of electrolytes, careful dose adjustment, and consideration of renal function. Use PK/PD concepts to interpret levels in conjunction with the patient’s clinical status.
For amikacin and vancomycin: understand concentration-dependent vs time-dependent killing. Use trough monitoring for aminoglycosides to assess toxicity risk and renal function, and consider peak concentrations for efficacy. For vancomycin, use AUC/MIC targets to guide dosing, with loading doses to rapidly achieve therapeutic exposure, and adjust maintenance dosing in renal impairment. MIC interpretation guides susceptibility and dosing decisions.
Across all cases, emphasize the integration of pharmacokinetic principles, pharmacodynamics, organ function, and clinical status to inform safe, effective therapy. Remember that dosing decisions are a blend of science and individualized patient context, not a one-size-fits-all rule.

Formulas and key equations referenced

First-order elimination kinetics:
\frac{dC}{dt} = -k C, \quad k = \frac{Cl}{V_d}
Half-life: t_{1/2} = \frac{\ln 2}{k}
Zero-order elimination kinetics:
\frac{dC}{dt} = -k_0
Time to steady state (approximate): t{ss} \approx 4-5 \ t{1/2}
Concentration-dependent antibiotics (example: amikacin): reach and maintain concentrations above MIC, with $C{max}$ relative to MIC and a post-antibiotic effect (PAE). The concept is often summarized as ensuring $C{max} > MIC$ for efficacy, while troughs are monitored for toxicity and renal function.
Time-dependent antibiotics (example: vancomycin): efficacy correlates with time above MIC and AUC/MIC-guided dosing; loading dose reduces time to achieve therapeutic exposure. AUC/MIC target is used to balance efficacy with nephrotoxicity risk.
Pharmacodynamic concepts (MIC and targets): MIC is the minimum inhibitory concentration; susceptibility categories are typically reported as susceptible, intermediate, or resistant based on MIC and organism-determined breakpoints.
Protein binding concept: The pharmacologically active fraction is the unbound fraction, which increases when albumin is low; total drug levels may not reflect the active drug exposure unless corrected or unbound levels are measured.

End of notes