MDMA pharmacokinetics: A population and physiologically based pharmacokinetics model-informed analysis

Huestis and colleagues situate this work in the context of MDMA (3,4-methylenedioxymethamphetamine) development for post‑traumatic stress disorder (PTSD). Earlier clinical trials indicate MDMA-assisted therapy can be effective and generally well tolerated, and pharmacokinetic (PK) properties of MDMA include nonlinear behaviour, primary hepatic elimination via CYP2D6, and rapid mechanism‑based inhibition (MBI) of CYP2D6 that can limit its own metabolism. The proposed clinical regimen for therapeutic sessions uses split dosing (two portions given 1.5–2 h apart) at maximum split doses of 120 mg (80 + 40 mg) and 180 mg (120 + 60 mg), with sessions spaced weeks apart rather than daily dosing.

Simulations used a Simcyp Caucasian healthy volunteer population with default CYP2D6 metaboliser frequencies (56.8% extensive, 32.5% intermediate, 8.2% poor, 2.5% ultra‑rapid). Virtual trials matched clinical demographics: for dose regimen comparisons, 10 virtual trials of 10 volunteers (50% female) aged 20–50 years were run to derive Cmax, AUC0‑44 (approximate five half‑lives) and AUC0‑4 (acute exposure). Sensitivity analyses varied fm CYP2D6 from 0.05 to 0.95 using atomoxetine as a representative sensitive CYP2D6 substrate to probe DDI magnitude. Metformin was chosen as a victim probe to evaluate potential transporter‑mediated DDIs; simulations included uncertainty scenarios reducing in vitro Ki values by up to 3‑fold for MATE1 and up to 15‑fold for OCT1/OCT2.

DDI simulations using atomoxetine as a sensitive CYP2D6 substrate found an fm CYP2D6 value of 0.85 best recovered observed clinical interactions and was used in modelling. Sensitivity analysis predicted a mild interaction (AUC ratio ≥1.25) beginning at fm CYP2D6 = 0.05, a moderate interaction (≥2‑fold) at fm CYP2D6 = 0.20, and a strong interaction (≥5‑fold) at fm CYP2D6 ≥ 0.75, illustrating substantial potential for MDMA to increase exposure of CYP2D6 substrates depending on their dependence on that pathway. For metformin, using in vitro Ki values for OCT1, OCT2 and MATE1 yielded no predicted interaction; however, reducing Ki values in sensitivity scenarios (3‑fold for MATE1 and 15‑fold for OCTs) produced a weak interaction with predicted AUC and Cmax elevations >1.5‑fold but <2‑fold. The data indicate less than 20% of MDMA and MDA are excreted unchanged renally.

The PBPK simulations captured observed data within accepted predictive limits and provided insight into split‑dosing dynamics and drug‑drug interaction (DDI) risk. Simulated split dosing produced similar overall exposure to single doses while lowering early exposure (AUC0‑4) and modestly delaying Tmax, an outcome the authors link to clinical rationale for split dosing: more gradual onset and extended peak effects without increasing total duration, and potential mitigation of early dose‑dependent physiological adverse effects such as elevated blood pressure and heart rate. PBPK DDI analyses reinforced MDMA's role as a potent CYP2D6 mechanism‑based inhibitor and indicated that co‑administration with CYP2D6 substrates, even those with low fm CYP2D6, can lead to clinically meaningful increases in exposure depending on substrate dependence and therapeutic index. Regarding renal transporters, the authors report that in vitro inhibition constants exceed expected clinical MDMA concentrations and that only conservative reductions in Ki produced weak interactions with metformin; therefore, transporter‑mediated DDIs are unlikely to be broadly clinically meaningful. The investigators acknowledge the observed study effect may reflect differences in washout periods, prior drug use history or other unmeasured factors between MPKF and NIDA datasets, and they note some parameter estimates had large uncertainty, warranting cautious interpretation.

The authors conclude that MDMA can be administered without regard to food because a high‑fat/high‑calorie meal did not alter Cmax or AUC, only delayed Tmax. PopPK modelling identified no covariates that necessitate clinically meaningful dose adjustments in the target population. PBPK simulations indicate split dosing reduces exposure during the first 4 h while maintaining overall exposure, which may facilitate a gentler onset and reduce severity of early adverse events. Finally, MDMA's principal DDI liability is as a precipitant via potent CYP2D6 inhibition; transporter‑mediated interactions are unlikely to be broadly meaningful but could be weak under conservative in vitro‑to‑in vivo assumptions.