This conceptual paper presents a probabilistic in vivo drug-delivery framework in which physiological variability is incorporated into release control rather than treated solely as an external source of variability. Instead of relying exclusively on fixed deterministic dosing schedules, the proposed approach explores bounded, state-dependent release behaviour to generate more adaptive exposure profiles. Parkinson’s disease is used as a motivating case study because levodopa therapy is frequently limited by gastrointestinal variability, motor fluctuations, OFF periods, and dyskinesia associated with peak–trough exposure dynamics. The framework hypothesizes that probabilistic micro-dosing strategies may reduce extreme excursions in effective exposure while increasing time spent within the therapeutic window. The manuscript is explicitly hypothesis-generating (TRL1–TRL2) and does not present clinical efficacy claims. It is intended as a conceptual basis for future modelling, experimental validation, and translational discussion in movement disorders, pharmacokinetics, drug delivery, and adaptive control systems. Potential broader applications may include epilepsy, diabetes, endocrine regulation, neuromodulation, and other therapeutic settings involving narrow therapeutic windows or fluctuation-sensitive physiology.
Bert Jan van der Werf (Fri,) studied this question.