TPMT Genotype

If you're about to start azathioprine for Crohn's disease, lupus, a skin condition, or mercaptopurine for leukemia, this genotype answers one of the main safety questions before the first dose. A small fraction of people cannot inactivate these drugs through this route. In them, a standard dose can drive white blood cells down fast enough to be life-threatening.

The result does not change over your lifetime. You test once, and the answer belongs in your record for every future prescription from this drug family.

What the Gene Actually Does

TPMT is an enzyme that works inside your cells. It tags thiopurine drugs, azathioprine, 6-mercaptopurine, and 6-thioguanine, with a small chemical marker that switches them off. What is not switched off can become the active form, thioguanine nucleotides, which can get built into DNA. That is part of the point in leukemia and part of why the drugs quiet an overactive immune system. It is also why too much active drug can wipe out healthy bone marrow.

So the enzyme is a brake. In European and African-ancestry populations, roughly nine in ten people have two working copies of the gene, about one in ten has one reduced-function copy, and about three in a thousand have two reduced-function copies.

The gene is on chromosome 6p22.3. Most loss of function comes from a handful of well-known star alleles, especially 2, 3A, and *3C. Sequencing reads the coding parts of the gene rather than checking only the common positions.

Bone Marrow Suppression

This is the outcome the test exists to prevent. People with reduced enzyme function have a much higher risk of low white blood cells on thiopurines, and people with two reduced-function copies are the group at risk for the most severe toxicity.

Dosing follows from that. One reduced-function copy usually means starting well below the usual dose, often around half. Two reduced-function copies usually means avoiding thiopurines for non-cancer use or using a deeply reduced dose in leukemia protocols with close monitoring. A randomized trial in inflammatory bowel disease found that screening before treatment and reducing the dose in carriers cut blood-count complications sharply in the variant group, without making the drug work worse.

A real-world cohort of more than 1,400 people taking azathioprine for inflammatory conditions found that metabolizer status predicted who stopped the drug because of marrow toxicity. That is the stake: not an abstract lab shift, but whether you keep the treatment that is controlling your disease.

What Genotype Does Not Catch

A normal result does not prove the dose is working. Some people on thiopurines shunt the drug toward a methylated byproduct and away from the active form. That pattern can look like no benefit plus rising liver enzymes. You find it by measuring the drug's processed forms in red blood cells after treatment starts, not by assuming the inherited result told the whole story.

So this is not a good-number, bad-number test. The inherited result mainly marks low enzyme function and marrow toxicity risk. Follow-up drug levels catch underdosing, too much methylated byproduct, and missed doses once the prescription has started.

Acute Lymphoblastic Leukemia

In childhood leukemia, mercaptopurine maintenance can run for years, and the genotype helps govern how much a child can tolerate. Deficient and intermediate metabolizers accumulate more active drug and develop severe neutropenia more easily. One large UK trial cohort found that children carrying one *3A copy had better event-free survival than children with two normal copies, and that the low blood counts they experienced did not hurt outcomes. Read carefully: that does not say deficiency is harmless. It says a child who is dosed correctly for genotype can keep enough drug exposure without turning the first dosing mistake into a crisis.

A Normal Result Does Not Mean You Are Safe

This is the misunderstanding that matters most. A second gene, NUDT15, can cause severe marrow suppression through a separate route, and a TPMT-only test does not capture it. In European IBD cohorts, NUDT15 variants have carried roughly twentyfold or higher odds of thiopurine-induced myelosuppression, independently of TPMT. Together the two genes explain about half of severe blood toxicity. TPMT alone misses much of it.

Ancestry changes which gene tends to matter more. Reduced-function TPMT alleles show up in roughly one in ten people in many European-ancestry datasets, mainly because of *3A. In East Asian and South Asian groups, and in Hispanic or Latin American groups with more Amerindian ancestry, NUDT15 reduced-function carrier status is often in the teens and can exceed 20 percent in some datasets. One cost analysis found you would need to test about 23 South Asian people to prevent one case of severe marrow suppression through NUDT15, versus about 786 Europeans. If your ancestry is East or South Asian or Hispanic, a clean TPMT result is not a safety pass. Order both.

Sequencing Versus a Common-Variant Panel

Many clinical tests interrogate the common positions behind 2, 3B, and 3C, then infer common star alleles such as 2, 3A, 3B, and *3C. In people of European ancestry, that catches most enzyme deficiency. A multicenter pediatric study of 320 children found complete agreement between common-variant genotyping and sequencing on final metabolizer calls.

The gap is small but not empty, and it lands on specific people. Sequencing reads the coding parts of the gene, so it can pick up rare and novel loss-of-function alleles that a common-variant panel reports as normal. In a Swedish study of 12,663 people, extended sequencing found 15 people carrying rare or previously uncharacterized alleles who would have been called normal by a routine array. A pediatric inflammatory bowel disease cohort found a novel damaging variant in a child who could not tolerate thiopurines and carried none of the common alleles. The *8 allele is now considered reduced function, reaches about 2.3 percent allele frequency in people of African ancestry, and is absent from many conventional panels.

There is one ambiguity sequencing may not solve by itself. Two variants can sit on the same chromosome or on opposite chromosomes. That difference determines whether you have one reduced-function copy or two. Short-read sequencing and many genotyping methods often cannot phase this cleanly, so an ambiguous 3B/3C pattern may need family testing, a method that can phase variants, or enzyme activity context.

Set expectations honestly. In one 100-child inflammatory bowel disease cohort, full-exon sequencing was good at ruling out thiopurine intolerance but missed many children who later had trouble. The reason is simple: much thiopurine intolerance happens through pathways this gene does not control.

Genotype or Enzyme Activity Test?

These are two different tests and they are routinely confused. The enzyme activity test measures how much TPMT your red blood cells actually carry. The genotype reads your DNA. They agree about 95 to 96 percent of the time.

Genotype is steadier in exactly the situations that can confuse the activity test. A recent red-cell transfusion can fill your circulation with donor red cells, so an activity test may report the donor's enzyme level, not yours. Active leukemia and disturbed blood-cell production can distort the activity reading too. In a 7,195-sample series, 11 percent of people with one reduced-function copy had enzyme activity in the normal range, while 3 percent of people with two normal copies had intermediate activity. That is enough to matter when the decision is the first dose.

The activity test has one real advantage. Because it measures function directly, it can catch reduced activity from causes a limited genotype test does not explain, including rare variants nobody has characterized yet. In one pediatric cohort, it caught more cases of later drug intolerance than single-gene sequencing did. Some centers run both, which is the most informative option when a thiopurine is definitely coming.

What Can Make the Result Misleading

• Panel coverage: a negative result on a limited panel rules out only the variants the assay was designed to find. Sequencing narrows this gap, but rare or hard-to-interpret alleles can still remain.
• Ancestry: allele frequencies differ sharply between populations. A negative TPMT result carries very different weight for someone of European descent than for someone of East Asian descent, where NUDT15 is often the dominant risk gene.
• Variants of uncertain meaning: sequencing can turn up gene changes that have not been tested well enough to assign function. A report naming an uncharacterized variant is not, by itself, a dosing instruction.
• Sample source: after a donor stem-cell transplant, a blood DNA sample can reflect donor marrow rather than your inherited DNA. Buccal or saliva collection is usually the cleaner specimen when that history exists.
• Consumer-grade reports: a direct-to-consumer genetics service may list a TPMT variant, but it is not built for prescribing decisions and typically covers only selected alleles. Do not dose a drug off one.

A One-Time Test With a Long Shelf Life

Your genotype will not change, so there is nothing to retest and no trend to track. What needs tracking is downstream. If you carry a reduced-function allele and go on a thiopurine, you need serial complete blood counts, most intensively in the first weeks and months when marrow suppression typically declares itself. Measuring the drug's processed forms in red blood cells, the active thioguanine nucleotides and the methylated byproduct, tells you whether you are underdosed, correctly dosed, or generating too much of the liver-toxic byproduct. Those are the numbers that move.

Store the result somewhere it will surface again. These drugs get prescribed by gastroenterologists, rheumatologists, dermatologists, nephrologists, and oncologists, and the doctor writing the prescription in ten years is probably not the one who ordered this test.

What to Do With a Reduced-Function Result

Order NUDT15 genotyping alongside this, not after. The two genes together explain roughly half of severe blood toxicity, and neither one substitutes for the other. If a thiopurine is on the table, add a complete blood count and liver enzymes as a baseline so there is something to compare against once dosing starts.

Bring the genotype to whoever is prescribing, before the first dose. The dose reduction is substantial and specific: often around half for one reduced-function copy, and much deeper for two. Carrying reduced-function alleles in both TPMT and NUDT15 calls for a deeper cut than either alone. For a two-reduced-copy result, a phase-uncertain result, or a variant of uncertain significance, a clinical pharmacologist, medical geneticist, or genetic counselor is worth involving, because the dosing math stops being routine.

Tell your siblings, children, and parents. Each shares roughly half your genome, so a reduced-function copy in you means a meaningful chance of one in them, and they may face a thiopurine prescription someday.

One thing this result does not tell you: whether you will get any disease. Variants in this gene govern how you handle a specific drug family. They do not cause illness on their own. The result is dormant until a prescription makes it matter, and then it matters a great deal.