A pharmacometric framework for norepinephrine transporter occupancy and dose equivalence across psychotropic medications

Background

Norepinephrine transporter (NET) inhibition is a key mechanism shared across diverse psychotropic medications, yet prescribing remains largely class-based and dependent on categorical labelling rather than quantitative assessments of target engagement. Positron emission tomography (PET) imaging, while the gold standard for in vivo measurement of NET occupancy, is resource-intensive and available for only a minority of agents. Pharmacometric modelling offers a complementary, scalable approach integrating binding affinity data with pharmacokinetic parameters to estimate receptor occupancy across a broad spectrum of compounds.

Methods

NET occupancy was estimated for 26 psychotropic agents and active metabolites using a standard receptor occupancy model: Occupancy (%) = Cₙ^free / (Cₙ^free + Ki) × 100. Ki values were sourced from the NIMH Psychoactive Drug Screening Program (PDSP) database (1), and total plasma concentrations were corrected for protein binding to derive free drug concentrations. Predictions were validated against published PET-derived occupancy estimates (n = 7). Target occupancy-based dose equivalents (ED50 and ED70) were derived using logit transformation, enabling standardized comparisons of noradrenergic potency referenced to desipramine. Sensitivity analyses assessed robustness to variation in Ki (±50%), protein binding (±10%), and plasma concentration ranges.

Results

Predicted NET occupancy varied substantially across agents. Very high mean occupancy (≥95%) was achieved by desipramine and milnacipran, representing opposite extremes of protein binding, while nortriptyline, maprotiline, and doxepin reached 70-90%. Moderate occupancy (50-70%) was observed for atomoxetine, norquetiapine, and duloxetine. Venlafaxine achieved only approximately 20% mean occupancy at standard doses despite its SNRI classification, attributable to its substantially weaker NET affinity (Ki = 2840 nM versus 5.91 nM for duloxetine). Hydroxybupropion contributed meaningful NET engagement (~40%) at therapeutic steady-state concentrations, whereas parent bupropion showed negligible occupancy (<5%). ED50 estimates ranged from approximately 8 mg (desipramine) to >74 mg (duloxetine). For venlafaxine, achieving 50% NET occupancy would require doses exceeding 375 mg/day. Predicted and PET-derived occupancies were significantly correlated (r = 0.76, R² = 0.57, p = 0.049), with strong agreement for most agents (mean absolute error <11%) and milnacipran identified as an outlier (overestimation ~54%), likely due to its exceptionally low protein binding. Sensitivity analyses indicated that occupancy estimates were most sensitive to variation in Ki, particularly for agents with intermediate affinity.

Conclusions

This pharmacometric framework reveals substantial heterogeneity in NET occupancy across psychotropic medications that cannot be inferred from drug class labels alone. Protein binding and active metabolite accumulation are critical determinants of functional engagement of noradrenergic targets. The target occupancy-based dose equivalence approach, compared with desipramine, enables mechanism-informed, cross-drug comparisons of noradrenergic potency and may support treatment selection, dose optimization, adverse-event prediction, and estimation of noradrenergic burden in polypharmacy. The clinical relevance of this framework is underscored by the finding that venlafaxine’s noradrenergic effect is dose-dependent, becoming clinically meaningful primarily above 150 mg/day, with implications for its positioning relative to duloxetine. These results are hypothesis-generating; prospective PET validation in larger cohorts is required to confirm clinical applicability.

References

1. National Institute of Mental Health Psychoactive Drug Screening Program (NIMH-PDSP) Ki database (pdspdb.unc.edu).