ORIGINAL ARTICLES 721 IMAJ • VOL 12 • DECEMbEr 2010 Background: Traditionally, medication dosage was based on clinical and demographic parameters, but drug meta- bolism was recently recognized as an important factor for proper dosing and prediction of side effects. Metabolic considerations are crucial when administering drugs with a narrow therapeutic index, such as those of the thioguanides family (azathioprine and 6-MP). These can cause life- threatening myelosuppression due to low activity of a critical metabolic enzyme, thiopurine S-methyl transferase. A number of single nucleotide substitutions encoding variant enzymes account for most enzyme deficiencies. Objectives: To determine the frequency of individuals from different Israeli ethnic groups who may be at risk for drug toxicity from drugs of the thioguanide family due to enzymatic variants. Methods: DNA analysis was performed using polymerase chain reaction methods. We tested TPMT allelic variants TPMT*3A (G460A, A719G), TPMT*3B (G460A) and TPMT*3C (A719G) in five subpopulations in Israel: mixed-origin Israeli Jews, Arabs, Druze, Jews of Kurdish extraction, and Ethiopian Jews. Results: The Druze (P = 0. 0002) and Ethiopian Jewish (P = 0.015) subpopulations had a significantly unique dis- tribution of allelic variants compared to the rest of the Israeli population. The Druze subpopulation showed a high number of TPMT variants with decreased activity, and a homozygote for TPMT*3A/ *3A was detected. Ethiopian Jews were found to carry mainly the TPMT*3C variant, also observed in other studies of African populations. Conclusions: It is advisable that Druze patients be tested for the TPMT enzyme before starting treatment with 6-MP or azathioprine. Such testing may also be considered for other Israeli ethnic subgroups. IMAJ 2010; 12: 721–725 pharmacogenetics, drug metabolism, adverse effects, 6-MP, azathioprine Evaluating Frequencies of Thiopurine S-Methyl Transferase (TPMT) Variant Alleles in Israeli Ethnic Subpopulations Using DNA Analysis Ofri Ronen 1 * , Sara Bar Cohen PhD 2 ** and Deborah Rund MD 2 ** 1 Hebrew University Medical School and 2 Department of Hematology, Hadassah-Hebrew University Medical Center (Ein Kerem Campus) Jerusalem, Israel ABSTRACT: KEY WORDS: *This study was performed in partial fulfillment of the MD thesis requirements of the Hebrew University Medical School, Jerusalem, Israel ** These authors contributed equally to the manuscript TMPT = thiopurine S-methyl transferase T he accumulation of knowledge on genetic variations among populations has enabled the development of what has come to be known as "personalized medicine." Much pharmacogenomic research has been conducted to enable identification of individual patients at risk for adverse effects of drugs. is is important when considering that iatrogenic severe side effects of medications are responsible for 6–7% of all hospital admissions, prolong the average hospitalization by 2 days and cause about 100,000 deaths each year in the United States [1]. e search for pharmacogenomic markers to identify patients most likely to have increased risk of toxic effects of drugs has focused on changes in genes that code for drug- metabolizing enzymes. A change in enzymatic activity can lead to the accumulation of a drug or its metabolites which can be toxic, especially when there is a narrow therapeutic index [1]. Drugs of the thioguanides family, such as azathio- prine and 6-MP, are good examples of this type of medication. ey can cause life-threatening myelosuppression due to low activity of a critical metabolic enzyme, thiopurine S-methyl transferase. ose drugs are used in many medical subspe- cialties such as hematology (for treating acute lymphocytic leukemia), dermatology (for pemphigus vulgaris), gastroen- terology (for inflammatory bowel disease) and surgery (organ transplantation). It has long been known that treatment with these drugs might be toxic and even life-threatening because of myelosup- pression. As early as 1979, it was found that leukemia patients respond differently to 6-MP. Weinshilboum and Sladek [2] studied the activity of TPMT in the erythrocytes of 298 ran- domly taken samples. ey observed a trivalent distribution in which 88.6% had high activity, 11.1% had low activity and 0.3% had undetectable activity. From that distribution, using the Hardy-Weinberg equation, they noted the existence of an autosomal co-dominant inheritance of two alleles encoding high and low TPMT activity [2]. Subsequently, pharmacogenomic studies have revealed that some individuals are deficient in an enzyme crucial for metabolizing thioguanides, known as thiopurine S-methyl transferase. In 1996, Szumlanski et al. [3] mapped the TPMT