ideasIdea
Ancestry-aware pharmacology: drug-level sub-studies across populations before approval
Drugs are processed differently by people with different genetic backgrounds, but doses are set mostly in white and East Asian populations. Every new drug should be studied for how it behaves across ancestries, with dosing advice by genotype rather than by race.
Registrational programmes include pharmacokinetic and pharmacogenomic sub-studies powered across major genetic ancestry groups, genotyping known variants (CYP2D6, CYP3A5, UGT1A1, DPYD, NUDT15, HLA alleles) and reporting exposure and toxicity by genotype. Labels carry genotype-based dosing where exposure differs, replacing crude race-based statements. African, South Asian and Indigenous populations are prioritised because they are least represented in existing PK data.
Hypothesis
Ancestry-aware programmes will identify clinically relevant exposure differences for a meaningful fraction of new oral oncology agents, and genotype-guided dosing will reduce grade 3+ toxicity in affected groups without loss of efficacy.
Rationale
Known examples include NUDT15 variants and thiopurine toxicity in East Asian and Hispanic patients, UGT1A1*28 and irinotecan, and DPYD variants and fluoropyrimidines; several were discovered only after approval and harm.
What would test it
Mandate the sub-study for new oral agents for three years and count label changes attributable to it; run a genotype-guided dosing RCT for one agent with a discovered difference.
Maturity
early clinical
Who has to act
regulator
Cost to try
Medium ($1M to $50M)
Years to first evidence
4
Bottlenecks it attacks
- Trials do not represent the people who get cancer · Older, Black, Hispanic, Asian, rural, poor and multimorbid patients are under-represented, so results may not apply to them.
- Wrong doses · Most drug doses were chosen as the highest a person can tolerate, which is often more than they need.