<?xml version="1.0" encoding="UTF-8"?>
<opml version="2.0">
  <head>
    <title>097-02 Drug discovery, preclinical development, and clinical trials</title>
    <ownerName>Integrated Medical Foundations</ownerName>
  </head>
  <body>
    <outline text="Drug discovery, preclinical development, trials">
      <outline text="Development as iterative translation">
        <outline text="Connects a modifiable mechanism with benefit"/>
        <outline text="Rarely a simple target-to-cure sequence">
          <outline text="Disease may not depend on the target"/>
          <outline text="Models capture only one feature"/>
          <outline text="Active compound may lack exposure or safety"/>
        </outline>
        <outline text="Failure at any interface invalidates results"/>
      </outline>
      <outline text="Target identification and validation">
        <outline text="Genetics, pathology, omics, phenotypic screens"/>
        <outline text="Association is not validation"/>
        <outline text="Stronger target evidence">
          <outline text="Manipulation changes disease-relevant biology"/>
          <outline text="Direction and timing understood"/>
          <outline text="Human variation predicts compatible effects"/>
        </outline>
        <outline text="Lifelong genetic loss differs from late inhibition"/>
        <outline text="Valid targets may be inaccessible or unsafe"/>
      </outline>
      <outline text="Screening and hits">
        <outline text="Phenotypic: integrated biology, harder mechanism"/>
        <outline text="Target-based: tractable, may oversimplify"/>
        <outline text="Orthogonal assays confirm activity"/>
        <outline text="Counter-screens catch artefacts and off-targets"/>
        <outline text="A hit is a starting compound, not a drug">
          <outline text="Confirm identity, purity, concentration-response"/>
        </outline>
        <outline text="Structure-activity relationships guide chemistry">
          <outline text="Lipophilicity aids passage, adds binding"/>
        </outline>
      </outline>
      <outline text="Lead optimisation and biomarkers">
        <outline text="Pharmacokinetics and pharmacodynamics together"/>
        <outline text="Absorption through active metabolites set exposure"/>
        <outline text="Target engagement biomarkers"/>
        <outline text="Downstream biomarkers show biological consequence"/>
        <outline text="Clinical endpoints: feel, function, survive"/>
        <outline text="A biomarker is not automatically a surrogate">
          <outline text="Non-causal marker or off-target harm"/>
        </outline>
      </outline>
      <outline text="Preclinical models">
        <outline text="In vitro: purified protein to co-culture">
          <outline text="Complexity costs control and throughput"/>
          <outline text="Cell lines mutate and lose phenotypes"/>
          <outline text="Organoids lack circulation and immunity"/>
        </outline>
        <outline text="Compare with achievable unbound human exposure"/>
        <outline text="Animals integrate kinetics, physiology, toxicity">
          <outline text="Species differ in targets, metabolism, immunity"/>
          <outline text="Homologous, phenotypic, or empirically predictive"/>
        </outline>
        <outline text="Rigorous design reduces exaggerated effects"/>
      </outline>
      <outline text="Safety, toxicology, and manufacturing">
        <outline text="Safety pharmacology: heart, lungs, CNS"/>
        <outline text="Toxicology: repeat dosing, genotoxicity, reproduction"/>
        <outline text="NOAEL is a boundary, not zero risk"/>
        <outline text="Starting dose integrates exposure and uncertainty"/>
        <outline text="Manufacturing determines delivered exposure">
          <outline text="Impurities may have distinct toxicity"/>
          <outline text="Biologics: folding, glycosylation, immunogenicity"/>
          <outline text="Process changes need comparability evidence"/>
        </outline>
      </outline>
      <outline text="Early clinical development">
        <outline text="First-in-human: safety, tolerability, kinetics"/>
        <outline text="Single and multiple ascending dose cohorts">
          <outline text="Predefined stopping rules"/>
        </outline>
        <outline text="Healthy volunteers or patients by risk"/>
        <outline text="Escalate on cumulative exposure, not dose alone"/>
        <outline text="Phase 2: dose, regimen, preliminary efficacy">
          <outline text="Dose choice is an estimation problem"/>
          <outline text="Several doses and exposure-response modelling"/>
        </outline>
        <outline text="Enrichment raises signal, narrows applicability"/>
      </outline>
      <outline text="Confirmatory trials and endpoints">
        <outline text="Phase 3 tests benefit and harm"/>
        <outline text="Randomisation balances prognostic factors"/>
        <outline text="Allocation concealment protects enrolment"/>
        <outline text="Blinding reduces differential care and assessment"/>
        <outline text="Intention-to-treat preserves randomisation">
          <outline text="Per-protocol is vulnerable to selection"/>
        </outline>
        <outline text="Composite endpoints driven by minor components"/>
        <outline text="Surrogates need an established link to benefit"/>
        <outline text="Superiority, non-inferiority, equivalence">
          <outline text="Non-inferiority needs a working active control"/>
          <outline text="Poor adherence, crossover mimic similarity"/>
        </outline>
        <outline text="Significance is not clinical importance"/>
      </outline>
      <outline text="Adaptive designs and pharmacovigilance">
        <outline text="Prespecified adaptation preserves error control"/>
        <outline text="Platform trials add or remove arms"/>
        <outline text="Interim looks need statistical adjustment"/>
        <outline text="Independent data-monitoring committees"/>
        <outline text="Post hoc adaptation is ordinary bias"/>
        <outline text="Rare, latent, pregnancy effects after approval"/>
        <outline text="Spontaneous reports lack denominators"/>
        <outline text="Investigate signals proportionately"/>
      </outline>
      <outline text="Development decisions">
        <outline text="Probability, consequence, opportunity cost"/>
        <outline text="Coherence outweighs a single small p value"/>
        <outline text="Update target, model, molecule, dose separately"/>
        <outline text="Answers must stay coherent over time"/>
      </outline>
    </outline>
  </body>
</opml>
