ideasIdea
Look for the resistant sub-population before the first dose
Resistance mutations often exist in a tiny fraction of cells before treatment starts. A very sensitive test at diagnosis could find them and prompt a combination from day one.
Duplex or error-corrected sequencing detects variants below 0.01 percent allele fraction. Pre-existing resistance clones (EGFR T790M before first-generation TKIs, KRAS in colorectal cancer before anti-EGFR, ESR1 in breast cancer) predict early failure. The proposal is a baseline ultra-deep panel of known resistance alleles for every patient starting a targeted drug, with a pre-specified rule to add the mechanism-matched second agent if a clone is found.
Hypothesis
Ultra-deep baseline detection of a resistance allele predicts progression within six months on monotherapy, and patients with detectable pre-existing clones who receive an upfront combination have progression-free survival matching those without such clones.
Rationale
Pre-existing resistance was shown for T790M and for KRAS in colorectal cancer; assays have improved by two orders of magnitude since those studies.
What would test it
Prospective observational study in 300 patients starting osimertinib or an anti-EGFR antibody: ultra-deep baseline plasma and tissue, blinded, correlated with time to progression; if predictive, a randomised combination trial in the clone-positive group.
Maturity
preclinical evidence
Who has to act
research
Cost to try
Medium ($1M to $50M)
Years to first evidence
3
Bottlenecks it attacks
- Tumour heterogeneity and clonal evolution · A tumour is many tumours. Treatments that kill most cells leave the rest to grow back, changed.
- Acquired resistance to every therapy · Nearly every targeted therapy stops working within months to a few years as the tumour adapts.