Pharmacokinetics: Drug Absorption, Distribution, Metabolism, Excretion

Why pharmacokinetics matters for implantable and wearable devices

Pharmacokinetics (PK) is the study of how a drug moves through the body over time — its absorption, distribution, metabolism, and excretion, abbreviated ADME. For medical device engineers it matters because every sensor sampling rate, every calibration interval, and every alarm threshold in a drug-sensing or drug-delivering device is implicitly a claim about a timeline that pharmacology, not electronics, defines. A continuous glucose-adjacent insulin pump, a transdermal fentanyl patch controller, or an implantable chemotherapy pump all fail silently if their sampling cadence doesn't match the drug's actual concentration curve in the tissue the sensor can reach.

How drug kinetics affects sensor placement and measurement timing

Sensor placement has to match the compartment where the drug concentration you care about actually changes on a clinically meaningful timescale. A subcutaneous sensor measures interstitial fluid, which lags plasma concentration by 5–15 minutes for small hydrophilic molecules — a lag that is irrelevant for a drug with a 12-hour half-life but dangerous for one acting in minutes. If the device's control loop assumes plasma kinetics while reading interstitial fluid, every dose decision is built on stale data, and the error compounds specifically during rapid concentration changes, which is exactly when accuracy matters most.

Why steady-state concentration affects device calibration windows

Steady-state concentration — the plateau reached when drug input rate equals elimination rate — sets the earliest point at which a device can validate its own calibration against a stable reference. Calibrating a sensor against a biomarker that is still rising toward steady state produces a calibration curve that is only valid for that one transient, not for chronic dosing. As a rule of thumb, a drug reaches ~90% of steady state after roughly 3.3 elimination half-lives of constant dosing; calibrating before that point means recalibrating again once the patient's true chronic baseline is reached, which most deployed devices cannot do automatically.

Device-driven dosing feedback loops and kinetic feedback

A closed-loop dosing device is a control system wrapped around a pharmacokinetic plant, and the plant's time constants set the stability margins of the loop just as surely as an RC time constant does in analog feedback. Dose too aggressively relative to the drug's absorption delay and distribution phase, and the loop overshoots — the PK equivalent of a control loop oscillating because its sensor lag exceeds its actuator bandwidth. This is the same reasoning used in closed-loop insulin delivery, where sensor lag relative to the Blood Glucose Reference Range: Fasting, Postprandial, HbA1c must be bounded before any dosing algorithm can be certified.

How drugs cross biological barriers during absorption

Absorption is the movement of a drug from its administration site into systemic circulation, and it is gated by membrane crossing mechanisms that differ by orders of magnitude in speed.

Passive diffusion, active transport, and first-pass metabolism

Passive diffusion moves drug molecules down a concentration gradient with no energy cost and no saturation point, while active transport uses carrier proteins that saturate at high concentrations and can be competitively blocked by other substrates. First-pass metabolism is the reduction in drug concentration that happens when an orally absorbed drug passes through the liver via the portal vein before ever reaching systemic circulation; for some drugs this destroys over 90% of the administered dose before it has any systemic effect, which is why those drugs are formulated for injection, transdermal delivery, or sublingual absorption instead. For a device engineer, first-pass metabolism matters because the oral route is often unsuitable for a device-coordinated drug whose therapeutic window is narrow — the bioavailable fraction varies with liver blood flow, meal timing, and individual enzyme expression, none of which a device can measure directly.

Gastrointestinal absorption routes and bioavailability prediction

Bioavailability (F) is the fraction of an administered dose that reaches systemic circulation intact, calculated as the ratio of area under the plasma concentration curve (AUC) for the given route versus intravenous administration. GI absorption depends on gastric pH, gut motility, and whether the drug is a substrate for intestinal efflux transporters — variables that shift with food intake, co-administered drugs, and disease state, making oral bioavailability the least predictable input a device-side dosing model receives.

Transdermal and mucosal absorption kinetics for wearable delivery

Transdermal absorption is rate-limited by diffusion through the stratum corneum, giving onset times of 1–4 hours even for small lipophilic molecules, which makes patch-based wearables inherently unsuited to drugs needing rapid onset. Mucosal routes (buccal, nasal) bypass the stratum corneum and first-pass metabolism simultaneously, giving onset in minutes — a relevant trade-off when choosing where a wearable's delivery membrane should sit against skin versus mucosa.

How drugs distribute to tissues and establish effective concentrations

Distribution is the reversible movement of drug between blood and tissues once it is in systemic circulation, and it determines what concentration a tissue-adjacent sensor will actually see relative to the dose given.

Volume of distribution and tissue partitioning

Volume of distribution (Vd) is a theoretical volume, calculated as dose divided by plasma concentration, that relates the amount of drug in the body to the concentration measured in plasma — it is not an anatomical volume and can exceed total body water for drugs that sequester heavily in fat or bind tissue proteins. A drug with Vd of 0.5 L/kg stays mostly in plasma and extracellular fluid, where a blood-contacting sensor tracks it well; a drug with Vd of 20 L/kg is mostly in tissue, and plasma concentration becomes a poor proxy for the therapeutic effect the device is trying to control.

Blood-brain barrier and protein binding effects on kinetics

The blood-brain barrier restricts drugs to those that are small, lipophilic, or actively transported, which is why CNS-acting implants must deliver locally rather than rely on systemic dosing to reach a therapeutic brain concentration. Protein binding — commonly to albumin — reduces the free (pharmacologically active) fraction of a drug, and because most plasma sensors measure total drug, not free drug, a device algorithm calibrated against total concentration will misjudge effect in patients with abnormal albumin levels, such as those with hepatic or renal impairment.

Compartmental modeling for multi-tissue drug presence

Compartmental models represent the body as one or more interconnected kinetic compartments, each with its own volume and transfer rate constants, rather than tracking every individual tissue; a two-compartment model (central: blood and well-perfused organs; peripheral: muscle and fat) is the minimum needed to explain the early rapid drop and later slow decline typically seen in plasma concentration curves.

How metabolic pathways alter drug availability over time

Metabolism is the chemical transformation of a drug into other compounds, usually to make it more water-soluble for excretion, and it changes drug availability on a timescale set by enzyme kinetics, not just elimination.

Hepatic and extra-hepatic metabolism routes

The liver performs most drug metabolism through two phases — Phase I (oxidation, largely by cytochrome P450 enzymes) and Phase II (conjugation) — but extra-hepatic metabolism in the gut wall, plasma, and kidneys also contributes, and for some drugs it rivals hepatic clearance in magnitude.

Enzyme induction and inhibition in chronic therapy

Chronic co-administration of certain drugs induces or inhibits cytochrome P450 enzymes, changing a patient's clearance rate over days to weeks rather than instantly; a device algorithm with a fixed clearance assumption will drift out of calibration as enzyme expression shifts, which is why chronic therapeutic devices need periodic recalibration windows built into their design, not a one-time factory calibration.

Active metabolites and their kinetic contribution

Some metabolites retain or exceed the pharmacological activity of the parent drug, meaning a sensor or assay that measures only parent drug concentration can underestimate total therapeutic effect — a design detail that matters when selecting what a biosensor is chemically specific for.

How renal and hepatic clearance determines drug elimination

Clearance is the volume of plasma from which a drug is completely removed per unit time, and renal and hepatic clearance together account for essentially all elimination for most small-molecule drugs.

Glomerular filtration, active secretion, and reabsorption

The kidney eliminates drugs through three parallel processes: passive glomerular filtration (rate limited by glomerular filtration rate, GFR, and protein binding), active tubular secretion (carrier-mediated, can exceed filtration rate), and passive tubular reabsorption (favors lipophilic, un-ionized drug forms). Device-side estimation of renal clearance typically leans on the same Renal Biomarkers: GFR, Creatinine, BUN for Kidney Function used in standard kidney function assessment, since direct drug clearance measurement is rarely feasible in a wearable or implant.

Half-life calculation and elimination rate constants

Half-life (t½) is the time required for plasma concentration to fall by half, related to the elimination rate constant (k) by t½ = 0.693 / k, where k itself equals clearance divided by volume of distribution. Worked example: a drug with clearance of 5 L/hr and Vd of 50 L has k = 0.1 hr⁻¹, giving t½ = 6.93 hours; at that rate, four half-lives (~28 hours) bring plasma concentration to roughly 6% of its peak, the usual threshold engineers use to treat a drug as "cleared" for alarm-reset logic.

Device monitoring of clearance via biomarkers

Because real-time drug assay is rarely practical in a deployed device, clearance is inferred indirectly from surrogate biomarkers — creatinine trends for renal function, or temperature-correlated perfusion estimates from a Temperature Sensor: NTC, RTD Pt100, Infrared Thermopile — none of which measure the drug directly, so the inference always carries an uncertainty band that the dosing algorithm must accommodate rather than ignore.

How to model plasma concentration over time in devices

Plasma concentration modeling gives the device a predicted curve to sample against, rather than requiring continuous direct measurement, which is rarely available for most drugs.

One-compartment and multi-compartment models

A one-compartment model assumes instantaneous, uniform distribution and models concentration decay as a single exponential, C(t) = C₀ · e^(−kt); it is adequate for drugs with small Vd and fast distribution. A two-compartment model adds a distribution phase, producing a biexponential curve — C(t) = A·e^(−αt) + B·e^(−βt) — and is necessary whenever a device needs to distinguish early distribution-phase concentration from later elimination-phase concentration, because sampling during the wrong phase with a one-compartment assumption gives a systematically biased half-life estimate.

Area under the curve (AUC) and peak concentration prediction

AUC is the integral of plasma concentration over time and correlates with total drug exposure, while peak concentration (Cmax) and time-to-peak (Tmax) describe the absorption phase; a device predicting whether a patient is in a safe exposure range needs both numbers, not just an instantaneous reading, because two very different concentration-time curves can share the same instantaneous value at the moment of measurement.

Therapeutic drug monitoring and closed-loop sensor requirements

Therapeutic drug monitoring (TDM) is the practice of measuring drug concentration to keep it within a defined therapeutic window — below the minimum effective concentration the drug does nothing, above the maximum safe concentration it risks toxicity. A closed-loop device performing TDM needs a sensor with a sampling interval shorter than the fastest clinically relevant concentration change (often requiring sub-half-life sampling during absorption and distribution phases), a calibration validated at steady state as described above, and fallback logic for when the assumed PK model diverges from the patient's observed response.

How individual factors shift pharmacokinetics in patient populations

Individual variability in PK parameters is large enough that a device algorithm calibrated to a population average will misdose a meaningful fraction of real patients.

Age, sex, body composition, and genetic polymorphisms

Neonates and elderly patients both have reduced hepatic and renal clearance relative to typical adults, but for different reasons — immature enzyme systems in neonates, reduced organ blood flow and nephron mass in the elderly — and body composition changes (higher fat fraction) shift Vd for lipophilic drugs across both age extremes. Genetic polymorphisms in cytochrome P450 enzymes (CYP2D6 in particular) produce "poor metabolizer" and "ultrarapid metabolizer" phenotypes that can shift clearance by a factor of 5–10×, a variation no fixed-threshold device algorithm can absorb without patient-specific calibration.

Renal and hepatic impairment compensation in devices

Renal or hepatic impairment reduces clearance and prolongs half-life, requiring a device to either extend its dosing interval or reduce dose per interval — the correct compensation depends on which organ is impaired and which elimination pathway the drug actually uses, information that must come from clinical input, not be inferred solely from the device's own sensor stream.

Drug-drug interactions and device algorithm adjustment

Drug-drug interactions, most commonly through cytochrome P450 induction or inhibition, change clearance mid-therapy without any change in the primary drug's dose, which means a dosing algorithm that only reacts to its own sensor data — and has no channel for clinician-entered concomitant medication changes — will silently drift out of its intended therapeutic window.