Codeine and tramadol: painkillers that fail in both directions
Most drug-gene stories have one failure mode. Codeine has two, at opposite ends of the same scale, and the dangerous one is not the one people expect.
Also sold as: Co-codamol, Panadeine, Ultram, Tramal.
What codeine and tramadol are
Codeine and tramadol are opioid painkillers, prescribed for moderate pain and often sold in combination tablets with paracetamol. They sit below morphine in potency and above the ordinary anti-inflammatories, which is why they turn up so widely: after dental work, after minor surgery, for injuries and for some chronic pain.
Why your genes come into it
Neither drug does much on its own. Codeine has to be converted into morphine before it relieves anything, and tramadol depends on the same enzyme for its more potent metabolite. CYP2D6 does that conversion, and the amount of enzyme people carry varies more than for almost any other pharmacogene. At the slow end, very little active opioid is produced, so the tablet is close to a placebo and people are sometimes told their pain cannot be that bad. At the fast end the conversion runs hard, and a standard dose can produce an opioid level nobody intended. That second failure mode is why regulators restricted codeine in children and in breastfeeding mothers after deaths were linked to ultrarapid metabolism.
What published guidance says
The rows below restate CPIC's published guidance for each possible CYP2D6 result. They describe what a prescriber weighs when choosing an analgesic, not a dose for you to adopt. If codeine or tramadol is not working, the useful move is to say so to the prescriber rather than to take more of it.
CYP2D6
| If your CYP2D6 result is | What that means | Published CPIC guidance |
|---|---|---|
| Poor Metabolizer | Minimal conversion to the active opioid → little analgesia. | CPIC: the guideline suggests an analgesic not metabolised by CYP2D6 (e.g. morphine or a non-opioid), avoiding codeine/tramadol. |
| Intermediate Metabolizer | Reduced conversion to the active opioid → possibly reduced analgesia. | CPIC: the guideline notes monitoring response; an alternative analgesic may be considered if response is inadequate. |
| Normal Metabolizer | Normal conversion to the active opioid. | CPIC: standard use per label. |
| Ultrarapid Metabolizer | Rapid conversion can produce unsafe active-opioid levels. | CPIC: the guideline suggests avoiding codeine/tramadol and using an analgesic not metabolised by CYP2D6. |
Can your DNA file answer this?
This is the honest limit worth reading before you buy anything. CYP2D6 is the one pharmacogene a genotyping chip cannot resolve: the gene sits beside two near-identical pseudogenes and the variation that decides the result is structural, meaning whole-gene deletions, duplications and hybrid copies. An array reads pre-selected single positions and can see none of that. Aimosti therefore does not report CYP2D6 from a 23andMe or AncestryDNA file at all. It is part of the Deep Read panel, which runs from aligned reads in a BAM or CRAM file, and even there a diplotype the reads do not settle is reported as Indeterminate. If a service offers you a codeine result from chip data, the data behind it cannot support the answer.
A genotyping chip reads only a few hundred thousand pre-selected positions, about 0.02% of your genome, chosen for common variation. It does not sequence the rest, so it cannot find the rare or novel pathogenic variants the clinical and carrier modules look for: a “no finding” from chip data means “this chip never looked”, far more so than with whole-genome sequencing. Chip data suits common-variant traits, pharmacogenomic tag SNPs, and haplogroup ancestry, not clinical or carrier screening.
Common questions
Codeine has never worked for me. Does that mean I am a poor metabolizer?
It is one possible explanation and a fairly common one, but a report of no effect is not a genotype. Pain has many moving parts, and plenty of people who convert codeine normally still find it inadequate for the pain they have. What a CYP2D6 result adds is a mechanism a prescriber can act on rather than a subjective account they have to take on trust.
Which painkillers are not affected by CYP2D6?
Published guidance for reduced and ultrarapid metabolizers points toward analgesics that do not depend on this enzyme, and morphine is the example CPIC names. Which alternative is appropriate depends on what you are treating and your other medications, so it is a prescribing decision rather than a substitution you can make from a table.
Why were children stopped from getting codeine?
Because the ultrarapid end of this gene turned out to be dangerous. Children who convert codeine unusually fast can reach opioid levels high enough to suppress breathing after ordinary doses, and deaths were reported, including in breastfed infants whose mothers were taking it. Regulators restricted paediatric use as a result. The restriction is population-level, applied without knowing anyone's genotype, precisely because the genotype was not routinely known.
Can I get this from my 23andMe file?
No, and Aimosti will not pretend otherwise. CYP2D6 needs aligned reads to call properly. This is the clearest case on the site of a question consumer chip data cannot answer, which is why the coverage section above says so plainly rather than burying it.