Integrated Quantitative Systems Pharmacology model of opioid use disorder, linking competitive mu-opioid-receptor (MOR) occupancy from illicit fentanyl/heroin, methadone, buprenorphine, naloxone, and naltrexone to the mesolimbic reward pathway, chronic tolerance/opponent-process withdrawal (locus-coeruleus noradrenergic rebound), craving/relapse risk, dose- dependent respiratory depression/overdose, methadone-associated QTc prolongation, and treatment-retention outcomes, together with symptomatic alpha2-agonist (lofexidine) withdrawal management and harm-reduction comorbidity pathways.

Opioid use disorder is a chronic relapsing brain disease driven by repeated mu-opioid-receptor (MOR) agonism in the mesolimbic reward pathway: opioid binding disinhibits VTA dopaminergic neurons (via MOR-mediated suppression of tonic GABAergic interneurons), triggering nucleus-accumbens dopamine release, euphoria, and incentive-salience learning that consolidates drug-cue associations (Volkow 2016 NEJM). With repeated exposure, homeostatic counter-adaptation (cAMP superactivation, receptor downregulation) produces tolerance, while the locus coeruleus — acutely suppressed by opioid MOR-Gi signaling — rebounds into noradrenergic hyperactivity when drug levels fall, driving the autonomic withdrawal syndrome (mydriasis, piloerection, GI cramping, myalgia, anxiety — captured by the Clinical Opiate Withdrawal Scale, COWS) and, together with dynorphin/kappa-receptor- mediated anti-reward dysphoria, intense craving and relapse risk (Koob & Volkow 2016 Lancet Psychiatry). The most acute danger is dose-dependent respiratory depression: at brainstem (pre-Bötzinger complex) MOR saturation, full agonists (illicit fentanyl — now often adulterated with xylazine — heroin, methadone) can produce fatal hypoventilation, especially with benzodiazepine co-use; naloxone competitively displaces agonist to acutely reverse this, though its short half-life creates renarcotization risk against longer-acting opioids. Evidence-based pharmacotherapy — methadone (full agonist, QTc-prolonging), buprenorphine (high- affinity partial agonist with a ceiling effect that caps respiratory depression, risking precipitated withdrawal if induced too early), and naltrexone (full antagonist, oral or monthly extended-release depot, which also creates overdose risk if tolerance is lost during treatment lapses) — stabilizes MOR occupancy, suppresses withdrawal/craving, and roughly halves mortality versus no opioid-agonist treatment (Sordo 2017 BMJ), while lofexidine offers non-opioid symptomatic withdrawal relief and counseling/contingency management improves treatment retention.

1. Etiology & risk factors (genetic OPRM1 variant, chronic pain/iatrogenic exposure, social/adolescent vulnerability)
2. Opioid receptor pharmacology & signaling (MOR/KOR/DOR, Gi/Go, cAMP, GIRK, β-arrestin, analgesia/GI/miosis)
3. Mesolimbic reward pathway (VTA GABA/DA neurons, NAc dopamine release, euphoria, incentive salience, cue memory)
4. Neuroadaptation — tolerance & withdrawal substrate (cAMP superactivation, CREB, locus coeruleus, dynorphin/KOR anti-reward, HPA/CRF)
5. Withdrawal clinical syndrome (autonomic hyperactivity, GI cramping, myalgia, anxiety/insomnia, COWS composite)
6. Respiratory depression & overdose cascade (brainstem MOR, pre-Bötzinger complex, hypoventilation → apnea → death)
7. Drug PK/PD — illicit fentanyl/heroin (full agonist, high potency, xylazine/carfentanil adulteration)
8. Drug PK/PD — methadone (full agonist, long half-life, hERG/QTc, NMDA antagonism)
9. Drug PK/PD — buprenorphine ±naloxone (partial agonist, ceiling effect, precipitated-withdrawal risk)
10. Drug PK/PD — naloxone (acute reversal) & naltrexone (sustained blockade, XR depot)
11. Drug PK/PD — lofexidine/clonidine (α2-agonist, symptomatic withdrawal management)
12. Craving, relapse & behavioral/psychosocial therapy (cue reactivity, stress, contingency management, retention)
13. Comorbidity, harm reduction & biomarkers (IV-use infection risk, benzodiazepine synergy, naloxone distribution, mortality)

• Drug PK (6 drug classes, 15 compartments) — illicit fentanyl (gut/IN + central + peripheral), methadone (gut + central), buprenorphine (SL depot
  • XR-SC depot + central), naloxone (IN/IM depot + central), naltrexone (oral + XR-IM depot + central), lofexidine (gut + central).
• MOR occupancy — computed algebraically each step from a competitive multi-ligand Emax equation (relative Ki affinities, ligand-specific intrinsic efficacy), with an enforced partial-agonist ceiling for buprenorphine-dominated regimens.
• Disease/PD network (8 compartments) — MOR tolerance/adaptation index, locus-coeruleus noradrenergic tone, withdrawal (COWS 0-48), craving (VAS 0-100), respiratory-drive index (with an overdose/apnea threshold), methadone-associated QTc, phasic dopamine/euphoria, and treatment- retention index.

1. Patient profile — adjustable age/weight/use-history & baseline severity.
2. Drug PK — log-scale plasma concentrations for the six tracked drug moieties.
3. MOR occupancy/tolerance — total receptor occupancy, tolerance index, phasic dopamine/euphoria.
4. Withdrawal & craving — locus-coeruleus tone, COWS score, craving VAS.
5. Respiratory/overdose risk — respiratory-drive index vs. the apnea threshold, overdose-flag events.
6. Clinical endpoints — methadone QTc and treatment-retention index.
7. Scenario comparison — runs all 10 scenarios with the chosen profile.
8. References — key citations and link to the full bibliography.

# 1) Render the mechanistic map
dot -Tsvg oud_qsp_model.dot -o oud_qsp_model.svg
dot -Tpng -Gdpi=150 oud_qsp_model.dot -o oud_qsp_model.png

# 2) Simulate scenarios in R
install.packages(c("mrgsolve","dplyr","tidyr","ggplot2","shiny",
                   "shinydashboard","DT"))
library(mrgsolve)
mod <- mread_cache("oud_mrgsolve_model.R")
out <- mrgsim(mod, ev(amt=16, cmt="GUT_BUP", ii=24, addl=14), end=336, delta=0.1)
plot(out, c("MOR_occupancy_total","COWS_score","Respiratory_idx","Craving_VAS"))

# 3) Launch the dashboard
shiny::runApp("oud_shiny_app.R")

• Designed for research, education, and hypothesis generation — not clinical decision support, addiction-medicine guidance, or a substitute for validated overdose-risk prediction tools.
• MOR occupancy is modeled via a simplified competitive multi-ligand Emax equation with literature-informed relative affinity/efficacy surrogates, not a mechanistic receptor-binding kinetic (association/dissociation rate) model; real pharmacology includes slower buprenorphine receptor dissociation kinetics not explicitly represented here.
• The respiratory-depression Hill function and apnea threshold are illustrative surrogates for hypoxic-ventilatory-response suppression, not validated overdose-risk predictors — inter-individual variability (opioid-naive vs. tolerant, renal/hepatic impairment, genetic CYP variability) needs omega() blocks for population simulations.
• Xylazine and other non-opioid adulterant effects are represented as a simple additive respiratory-suppression term; they are not naloxone-reversible in reality, which the model reflects only as an unmodified additive offset rather than a full non-opioid sedation pathway.

Inherits the repository LICENSE.