Help make a difference
in the lives of the nation’s childhood cancer
patients, survivors & their families.

Donate Now »

Genetics and Anti-Leukemia Therapy: The TPMT Story


by Mary Relling MD Source: Fall 2003 CCCF Newsletter

There is lots of variability among patients in medication effects. All patients differ from each other in the way that their bodies absorb, metabolize and excrete all medications. For example, if the same dosage of a medication is given to 100 children, even though the dosage is exactly the same for all children (based on body size), there can be more than a 10-fold variation in the amount of medicine that circulates in the blood. Then drugs have to get from blood to their target in tissue – for example, into tumor cells. The tumor cells can vary 10-fold in their level of responsiveness to the anticancer drugs.

For many medicines there is a wide therapeutic range. That is, the dosage that is required to achieve efficacy in a large percentage of patients is much lower than the dosage that is expected to produce adverse effects in a large percentage of patients. This means that it is possible to give all patients a dosage of a medicine that is high enough to achieve its therapeutic effects without causing many to be at risk for side effects. Medicines such as penicillin and ibuprofen are examples of medications that have a wide therapeutic range. Anticancer medicines are examples of medicines that have a very narrow therapeutic range. That is, the dosage that is required to achieve anticancer efficacy is not very much lower than the dosage that may result in life-threatening toxicity or adverse effects. Because of the large inter-patient variability in the amount of drug in the blood, and the variability in tumor responsiveness, it’s difficult to determine an average dose that will work in a high percentage of patients without causing a high percentage to have side effects. This is why adverse effects are so common for anticancer agents.

Medication dosing is further complicated, because many anticancer medications require metabolism (a chemical change caused by enzymes in the body) by proteins to be activated as anticancer agents, and most require a different type of metabolism to be detoxified and excreted.

All of this variability makes it easy to understand why the same dosages given to 100 different children may have very different effects, both desirable and undesirable, in different children.

Some differences among patients are due to genetics: Many factors influence why patients differ so widely from each other in how they handle and respond to medications. Diet, kidney function, liver function, infections, are all factors that may cause patients to differ from one another in their response to medications. For example, drinking grapefruit juice along with prednisone, vincristine, and many other medicines may increase their side effects by interfering with their metabolism and excretion from the body. But all of these factors exert their influence on the background of the genetic constitution of the individual.

Medications work by interacting with proteins, and all proteins are encoded by genes. The order of bases (A, C, G, and T) in DNA is what determines our genetic make-up. From the human genome project, we know that 1 in every 300-1000 bases can be variant from one individual to another. These genetic variations may not be substantial enough to cause disease, but they might be substantial enough to cause a slight alteration in the function of the proteins coded for by the genes. If those proteins are involved in the absorption, the metabolism, or the responsiveness of the body or of the tumor cells to medications, then those genetic variations may be part of the reason why individuals differ from one another in how they respond to a medicine. So, one lesson of the human genome project is that variation is much more common than we had previously thought. Currently, the field of pharmacogenetics is an active area of research, which is determining which, of the many thousands of genetic variations, are the most important causes of why patients differ in how they respond to medications.

TPMT metabolizes 6MP and 6TG: Thiopurine methyltransferase (TPMT) is a protein that metabolizes two of the medications that are often used to treat children with leukemia or lymphoma: 6- mercaptopurine (6MP) and 6-thioguanine (6TG). For acute lymphoblastic leukemia (ALL), the doses of 6MP and 6TG are often adjusted, along with other chemotherapy, to cause a moderate “target” degree of myelosuppression. By adjusting doses, the goal is to keep the blood counts within a certain range, from week to week. TPMT is also an enzyme that is responsible for inactivating both of these medications. The higher the TPMT activity, the less the drug is available to be activated to thioguanine nucleotide metabolites. The lower the TPMT activity, the greater the exposure of the child to active 6MP or 6TG metabolites. Therefore, the side effects from thiopurines are affected by the patient’s TPMT enzyme activity.

TPMT undergoes genetically regulated variation in activity: In the early 1980’s, it was demonstrated that this enzyme, TPMT, is genetically regulated. One in 300 individuals is completely deficient—homozygous deficient, in this enzyme activity. This means that the copy of the gene from the mother and the copy of the gene from the father both carry inactivating genetic variations. Ten percent of the population is heterozygous (one copy of the gene from one parent is “variant” and one is normal or “wild type”) and has intermediate activity. Ninety percent of the population has normal or high TPMT activity, meaning that both the maternal and paternal copies of the gene are “wild-type,” that is, do not carry any genetic variations and therefore activity is unimpeded by genetic variation. This genetic variation in TPMT activity translates into differences in the concentrations of active metabolites when patients are treated with normal dosages of 6-mercaptopurine, and to a lesser extent, 6-thioguanine. Genetic tests for TPMT can be used to determine each patient’s TPMT “genotype.”

So, the starting doses for these medications are based upon the population average, which is heavily influenced by the 90 percent of the population that have normal high wild-type TPMT activity. However, the doses tolerated by the remaining 10% may be lower. For the 1 in 300 individuals who have deficient activity, the normal doses of these medications cause severe myelosuppression due to extremely high, more than 10-fold, concentrations of these active thioguanine nucleotide metabolites. This could result in sepsis, and theoretically, death from a serious infection. It has been demonstrated that by decreasing the dosage of 6-mercaptopurine approximately 10- fold in these rare children who are homozygous deficient, one can maintain the expected degree of myelosuppression and actually still maintain high levels of thioguanine nucleotides. By knowing which patients have low TPMT activity, it is possible to avoid decreasing the dosages of other myelosuppressive therapy that may be confounding the clinical picture of myelosuppression in any one individual patient. For the 10% of the population who are heterozygote, the case is not as dramatic. About one-third of children who are heterozygous for TPMT activity experience myelosuppression severe enough to warrant a dosage decrease. Nonetheless, in this one-third of patients who are identified as TPMT heterozygotes, it is reasonable to adjust the dose of 6-mercaptopurine alone and attempt to leave the doses of the other non-myelosuppressive treatments at the normal levels. On the other hand, the 90 percent of individuals who have normal, high TPMT activity are unlikely to experience acute myelosuppression from the normal dosages of thiopurines. In general, most leukemia treatment protocols include an algorithm for adjusting the dosages of other components of therapy with no particular emphasis or de-emphasis given to the thiopurine in patients with wild-type TPMT. More detailed information can be found:

Testing for TPMT: Until recently, TMPT testing was only available as a research test. However, there are now several laboratories conducting commercially available tests for TPMT activity, for TPMT genotype, and for the active thiopurine metabolites. The genotype is a genetic test performed using each patient’s DNA. Such DNA can be extracted from a small blood sample or even a swab of the inside of the mouth – a “buccal swab.” In addition, some laboratories offer tests that measure the actual levels of the thiopurine metabolites.These thiopurine tests, which costs approximately $400.00 can be useful not only for confirming the presence or absence of TPMT deficiency, but also for assessing the degree of compliance of patients with their thiopurine therapy. Measuring compliance can be especially useful for children who are responsible for taking their own dosages; if there are no detectable blood levels of metabolites, this can be a sign that the child is not taking his or her medicines.

Who should be tested and when: It has been controversial to decide which patients should be tested for TPMT status. Should all patients be tested for TPMT status before they receive their first dose of thiopurine? Some propose that such testing is unnecessary, and suggest that only those patients who experience acute myelosuppression during treatment should be tested for their TPMT status. Others propose that testing may be helpful prior to starting thiopurine therapy, in some situations. If a child is experiencing myelosuppression after thiopurine therapy, particularly early in therapy when they have only been exposed to 6-mercaptopurine or 6-thioguanine for a short time, it is possible that TPMT deficiency or low activity is the reason. You can talk to your child’s clinician about the possibility of testing for TPMT status.

All treatment protocols for leukemia have contingencies and guidelines for adjusting the dosages of antileukemic medications. In treatment of acute lymphoblastic leukemia, decisions are often made on how to adjust the dosages of medications based on the blood counts. It is now clear that one of the factors that may have an influence on the dose of thiopurine that patients can tolerate is TPMT genetic variation. As more patients and clinicians become aware of the importance of pharmacogenetics, this and other genetic tests will work their way into the routine treatment of children with cancer.

Mary V. Relling Pharmaceutical Sciences St. Jude Children’s Research Hospital 332 N. Lauderdale Room D-1052 Memphis, TN 38105 e-mail: phone: 901 495 2348; fax: 901 525 6869

Conflict of interest statement: “Dr. Relling is a faculty member of St. Jude Children’s Research Hospital and receives funding from the National Institutes of Health to study thiopurine pharmacology. St. Jude is the owner of patent rights based on the discovery of TPMT mutations that affect thiopurine metabolism. St. Jude has licensed these rights and receives a royalty based on sales of the genetic test for these TPMT mutations. This royalty is allocated to a St. Jude account, part of which is used to fund research in Dr. Relling’s department.”