Darapladib failed both of its phase 3 trials and was never approved — but re-analyzing the trial genetics found real gene-by-treatment interactions, including one locus where the same allele that raises risk on placebo is protective on the drug.
Darapladib, a lipoprotein-associated phospholipase A2 (Lp-PLA2) inhibitor, failed to improve outcomes in its two large phase 3 trials — STABILITY (stable coronary heart disease) and SOLID-TIMI 52 (acute coronary syndrome) — and was never approved for use. This page is not about a drug anyone takes today. It is about what a pharmacogenetic re-analysis of those two trials, combining 23,981 genotyped participants, found anyway: real, statistically strong gene-by-treatment interactions on major coronary events (MCE).
The clearest finding is rs12290663, near ANO3. On placebo, the risk allele raised MCE risk — hazard ratio 1.22 (p=4.28×10⁻⁵). In darapladib-treated patients, the same allele was protective — hazard ratio 0.79 (p=2.07×10⁻⁵). The genotype-by-treatment interaction itself reached p=6.03×10⁻⁹, the only interaction in the whole study to cross genome-wide significance.
rs192427471, near ANKRD50, shows a related but different pattern: a strong risk effect on placebo (HR 2.02, p=2.87×10⁻⁸) that simply disappeared under darapladib (HR 0.75, not significant) — the drug neutralizing a risk rather than reversing it.
rs147204125, near ACOT6, runs the other way: no effect on placebo (HR 1.05, not significant), but a clear risk increase specifically in darapladib-treated patients (HR 2.11, p=3.15×10⁻⁸) — the one locus here where being on the drug, not off it, carried the added risk.
Two more variants, rs149232047 and rs151269874, are tagged in GWAS Catalog against diarrhea in darapladib-treated patients — a documented side effect that affected up to 13% of trial participants. Both reach genome-wide significance (p=5×10⁻⁹ and p=2×10⁻⁹). That sits awkwardly next to the paper's own abstract, which states plainly that "no major loci for tolerability were found." Both things are true from what is verifiable here — the statistics cross the line, and the authors did not treat them as a major result — and this page states the tension rather than resolving it in either direction.
Four more variants push that tension further. GWAS Catalog tags rs11915606 (BTD, OR 7.23, p=4×10⁻⁸), rs145044782 (AC006465.3, OR 10.88, p=3×10⁻⁸), rs144610116 (FAM114A1, OR 89.29, p=7×10⁻⁹) and rs62568141 (near FOXB2, OR 3.36, p=7×10⁻¹¹) against a more specific "moderate or severe diarrhoea" outcome. The very large odds ratios here — 89.29 in one case — are the signature of a rare variant with only a handful of observed events, not a confidently estimated effect size: genome-wide significant by p-value, but with far more uncertainty around the point estimate than the number alone suggests.
Darapladib was never approved and nobody is prescribed it. Nothing here is a treatment recommendation. This page documents a real pharmacogenetic re-analysis of a failed drug trial — genuinely interesting genetics, not clinically actionable today.
Gene-by-treatment interactions like rs12290663's are exactly the kind of finding that motivates precision-medicine trial design — a drug that fails on average can still work, or backfire, in a genetically defined subgroup. That this happened in a drug that ultimately failed is itself part of the lesson: overall trial failure does not mean no subgroup was actually affected by the drug, in either direction.
What a 23andMe/AncestryDNA export or raw VCF can and can't tell you about Darapladib Response in Coronary Artery Disease comes down to these specific, well-studied positions — not a diagnosis.
near ANO3 · rs12290663
See detailed info →ACOT6 · rs147204125
See detailed info →near CENPW · rs149232047
See detailed info →RP11-77K12.8 · rs151269874
See detailed info →near ANKRD50 · rs192427471
See detailed info →BTD · rs11915606
See detailed info →FAM114A1 · rs144610116
See detailed info →AC006465.3 · rs145044782
See detailed info →near FOXB2 · rs62568141
See detailed info →The studies behind these variants recruited participants from different ancestries — a result found in one population doesn't always transfer to another. Based on 9 of 9 linked studies with a resolved discovery ancestry.
Databases, guidelines and references
Darapladib is a lipoprotein-associated phospholipase A2 (Lp-PLA2) inhibitor that failed to improve outcomes in two large phase 3 cardiovascular trials and was never approved for use.
Because re-analyzing the trial genetics found real, statistically strong gene-by-treatment interactions — including one locus where the same allele is harmful on placebo and protective on the drug — genuinely interesting pharmacogenetics, independent of whether the drug itself is in use.
rs12290663, near ANO3: the risk allele raised coronary event risk on placebo (HR 1.22) but was protective in darapladib-treated patients (HR 0.79) — a genotype-by-treatment interaction reaching p=6.03×10⁻⁹.
No. Darapladib is not an approved or prescribed drug, so nothing here is actionable clinically. This documents a research finding from a failed trial's genetics, not a guideline.
Free to reuse. This page's text is original writing from freely-available research, licensed CC BY 4.0 — reuse it, including commercially, with attribution to MyGeneLog™. It's general research-derived information, not medical advice or a diagnosis — see Terms of Use.