A prodrug is an inactive or low-activity molecule (often an ester, carbamate or phosphate mask on the real drug) that is designed for better absorption, solubility or targeting. Once inside the body it is bioactivated — usually by esterases, phosphatases or CYP450 enzymes cleaving off the masking group — into the pharmacologically active parent drug, which is then eliminated by the usual clearance pathways.
This is a two-step sequential first-order process, the same math used to model real prodrug pharmacokinetics (and, formally, a radioactive decay chain):
dP/dt = -k1·P (prodrug hydrolysis)
dA/dt = k1·P - k2·A (active drug appears, then clears)
P(t) = P0 e^(-k1 t)
A(t) = P0 k1/(k2-k1) · (e^(-k1 t) - e^(-k2 t))
Rather than solving the ODE directly, each of the N molecules in the 3D view is simulated individually as a stochastic event: on every timestep dt, a prodrug molecule converts with probability 1 − e^(−k1·dt), and an active-drug molecule is cleared with probability 1 − e^(−k2·dt) — the exact discrete-time form of a Poisson process with rate k. Summed over hundreds of molecules this individually-random behaviour reproduces the smooth exponential curves above almost exactly, which is also why real drug concentration data is never perfectly smooth at the molecular scale.
- k1 (hydrolysis) — how fast the masking group is cleaved. Higher k1 means the prodrug converts almost immediately (little benefit over dosing the active drug directly); very low k1 means most of the dose is excreted unconverted.
- k2 (elimination) — how fast the liberated active drug is cleared. The ratio k1/k2 sets whether the process is "activation-limited" or "elimination-limited", which is exactly the flip-flop kinetics problem real pharmacokineticists have to watch for.
- Real examples of this exact mechanism: enalapril → enalaprilat (ester hydrolysis, ACE inhibitor), valacyclovir → acyclovir (esterase), and codeine → morphine (CYP2D6 O-demethylation).