Thiopurines: A delicate balancing act


Although only a minority of people have a genetic variant for the TPMT gene, these patients face a high risk of death if prescribed thiopurine drugs.

Illustration by Sarah Nagorcka

Illustration by Sarah Nagorcka


Gene Dosage is a monthly column by Janan Arslan that finds out what genome science is uncovering about each individual's unique response to drugs and pharmaceuticals. Janan is a graduate student and pharmacogenomics researcher with a keen interest in personalised medicine.

If you are ever diagnosed with an autoimmune disorder, an inflammatory bowel disease or a blood cancer, there's a good chance you'll be prescribed a thiopurine. First synthesised in the 1950s, this class of drug includes the immunosuppressive drug azathioprine and its metabolite 6-MP.

Thiopurines are prodrugs ‒ to be of any benefit to the patient, they need to be converted upon absorption. Although generally considered effective, thiopurine use can sometimes lead to serious adverse events, including liver damage, inflammation of the pancreas, and reduced red and white blood cells and platelets. These side effects are primarily attributable to polymorphisms of the gene thiopurine methyltransferase (TPMT).

Throughout my pharmacogenomics career, I’ve mostly seen patients who were normal metabolisers for TPMT. However, on the rare occasion that we found a person with a reduced or non-functioning TPMT gene... let’s just say we were very quick to pick up a phone and call their prescribing physician.

Metabolism for thiopurines is quite complex. When consumed, about 90% of azathioprine is immediately converted to 6-MP, which can then follow one of three pathways. In one pathway, the 6-MP is converted to thioguanine nucleotides (TGNs), the compounds that make thiopurines immunosuppressive and toxic to cells. This conversion is achieved via a multi-step enzymatic pathway.

Azathioprine tablet: an immunosuppressant used in organ transplants. NLM Pillbox/Wikimedia Commons (public domain)

Azathioprine tablet: an immunosuppressant used in organ transplants. NLM Pillbox/Wikimedia Commons (public domain)

Alternatively, 6-MP metabolites can be inactivated by one of two enzymes: TPMT or xanthine oxidase. Any 6-MP inactivated by these enzymes is unavailable to form TGNs, and so can't produce the intended effects of the drug.

TGNs are purine antagonists that can incorporate themselves into the DNA of white blood cells. Purines, such as adenine and guanine, are the building blocks of DNA and are essential for completely functioning cells. The presence and build-up of TGNs alter the synthesis and function of the DNA, damaging the DNA and blocking cell proliferation. It is this function that induces immunosuppression and provides the drug’s efficacy.

When everything is running and functioning smoothly, moderate levels of TGNs are formed, providing the patient with the required therapeutic effect. However, deficiencies in the pathway can lead to the increased availability of 6-MP. The more 6-MPs there are with no other competing pathways, the more TGNs will be formed. High levels of TGN are responsible for the drugs’ life-threatening adverse events, such as myelosuppression, where a reduction of bone marrow results in lower levels of white and red blood cells and platelets, and therefore lowered immunity.

In pharmacogenomics, the focus is primarily on the gene TPMT, which is available for clinical testing, is extensively studied, and is considered to be largely responsible for the metabolism and adverse event outcomes of thiopurine administration. If a patient has a non-functional TPMT gene (where they inherit a null copy of the gene from both their mother and father), the level of TGNs will be greatly elevated, even if a patient is prescribed standard thiopurine doses. 

Some patients may have moderate TPMT capacity, where one copy of the inherited gene is normal and the other is damaged. For these patients, standard doses of thiopurines may still result in adverse events. Although their ability to metabolise thiopurine is less hindered than that of complete non-responders, monitoring and titrating doses to suit these patients is still recommended.

Structure of the TPMT protein. Emw/Wikimedia Commons (public domain)

Structure of the TPMT protein. Emw/Wikimedia Commons (public domain)

As far as pharmacogenomics tests go, TPMT genotype testing is probably one of the most established tests available. Most pharmacogenomics companies offer TPMT testing as part of their “menu”, and in Australia there are now several laboratories that offer the test. As I mentioned in an earlier column, TPMT genetic testing is one of only two pharmacogenomics tests that are covered by Medicare. This has allowed hospitals, such as the Royal Children’s Hospital in Melbourne, to incorporate and offer the test to their patients more easily.

Aside from TPMT genotyping, an alternative method that can be used clinically involves measuring the TPMT activity of red blood cells. Whereas the genetic test identifies variants that alter the enzymatic abilities of the TPMT, reviewing red blood cell activity is a direct measure of what is happening to the patient during the course of their treatment. These are typically referred to as genotype and phenotype tests, respectively.

The benefit of TPMT genotyping is that it is not affected by blood transfusion. If a patient has recently undergone blood transfusion, testing the red blood cells directly could give misleading results. The results may also be affected by anaemia (deficiency in red blood cell production) and the age of the red blood cells. That being said, prescribers generally use both methods of testing for completeness (although naturally there are some arguments for both test types).

Inherited differences in the TPMT gene were first described by Weinshilboum and Sladek in 1980. Since their publication, many genetic variants have been identified, with the variants known as *2, 3A, 3B and 3C accounting for the majority (approximately 95%) of genetic differences. Despite the plethora of genetic variants that could affect the function of the TPMT gene, the literature suggests only 0.3% of patients have low or non-functional TPMT activity, while 10% have moderate activity. This leaves about 89% of individuals who have normal TPMT activity.

So, if only a small fraction of patients are affected, why bother? What’s the point of testing everyone who will be prescribed a thiopurine drug if there’s only a small chance that a patient has a non-functioning TPMT? I can’t begin to tell you how many times I’ve had someone ask me this. And my answer has always been very simple: those who do have genetic variants are at risk of experiencing fatal side effects. That in itself is motivation enough to get a pharmacogenomics test.