Vyvanse

Smoothest amphetamine PK in the class — L-lysine peptide bond is cleaved by red-blood-cell aminopeptidase to release d-amphetamine over ~1-3 hours, giving 12-14hr duration with a flatter peak than Adderall IR. Skip for Dylan-archetype: same brain-development concern as Adderall (male adolescent rodent PFC dopamine axon misrouting, replicated 2023-2025) plus 12hr duration that conflicts with late-chronotype sleep migration. Strong candidate only with formal ADHD or binge-eating disorder diagnosis — Vyvanse is the only FDA-approved BED pharmacotherapy as of 2026.

Vyvanse (lisdexamfetamine, LDX) is a prodrug: the active molecule is regular d-amphetamine, but it's chemically tethered to L-lysine via a peptide bond. The intact prodrug is pharmacologically inactive — it has no affinity for DAT, NET, or TAAR1. To work, it has to be cleaved.

The activation step (where the magic lives):

• After oral absorption, LDX is taken up by red blood cells. The cytosol of RBCs (not plasma, not membrane fraction, not GI peptidases as originally thought) contains a metallo-aminopeptidase that hydrolyzes the lysine-amphetamine peptide bond (Pennick 2014). The specific enzyme has not been definitively identified; aminopeptidase B is involved but not sufficient on its own. EDTA inhibits the reaction (divalent cation dependent), and bestatin blocks it — both hallmarks of metallo-aminopeptidase activity.
• Cleavage releases free d-amphetamine + L-lysine. The conversion half-life is roughly 1 hour and the rate is only RBC-dependent — plasma esterases, GI peptidases, and liver enzymes don't contribute meaningfully. The rate is preserved across normal hematocrit and only slows when RBC count drops below ~10% of normal.
• This RBC-bound, rate-limited release is what produces the smoother PK curve. Because conversion is slow and saturable rather than instant, you get a gradual rise in d-amphetamine plasma levels rather than the sharp Tmax of IR amphetamine.

Why this matters for abuse liability:

• IV/intranasal route blocked: IV LDX 50mg in supervised abuse-liability trials produced no greater "drug liking" than placebo, while IV d-amphetamine 20mg produced significant liking. Snorted/injected LDX still has to wait for RBC hydrolysis — you can't bypass the rate-limit (Kämmerer 2024 review).
• Oral route partially protected, not fully: This is the nuance the marketing softens. Oral LDX still produces euphoria and drug-liking effects above placebo, and its oral abuse liability is comparable to oral d-amphetamine when matched for d-amphetamine exposure. The protective effect is mainly against tampering (snorting/injecting), not against swallowing extra pills. Important framing for Dylan's risk thinking — "lower abuse liability" is true for the parenteral route, modest for the oral route.

Once d-amphetamine is released, mechanism is identical to Adderall/Dexedrine:

• DAT/NET reuptake inhibition + reverse transport — d-amphetamine enters monoamine terminals via DAT/NET, displaces vesicular dopamine and norepinephrine via VMAT2 inhibition, and triggers efflux via TAAR1 activation. Net effect: large rise in synaptic DA and NE.
• MAO inhibition (modest) — d-amphetamine weakly inhibits MAO-A, prolonging monoamine action.
• Cortical and striatal effects — DA elevation in nucleus accumbens drives reinforcement; PFC NE/DA elevation drives executive function and sustained attention.

PK summary:

• Tmax of intact LDX: <1hr (it's rapidly absorbed and starts cleaving immediately).
• Tmax of d-amphetamine post-LDX: ~3.5-4.5hr (vs ~3hr for Adderall IR with sharper peak).
• Half-life of d-amphetamine: ~10-12hr.
• Effective duration: 12-14hr clinically.
• Food: high-fat meal slightly delays Tmax but doesn't change AUC — taken with or without food.
• ~1% of dose excreted as intact LDX; rest as d-amphetamine and amphetamine metabolites.