Genetic variation associated with thiopurine methyltransferase (TPMT) deficiency and toxicity to azathioprine and mercaptopurine therapies is the classic example of pharmacogenetics in clinical practice. Red cell thioguanine and methylmercaptopurine levels enable further dose optimisation. Non-responders with a high methylmercaptopurine : thioguanine ratio may benefit from split dose therapy or switching to allopurinol and low dose thiopurine therapy. Allopurinol and low dose thiopurine therapy may rescue patients experiencing hepatotoxicity associated with high methylmercaptopurine levels. Variation in the nudix hydrolase NUDT15 is prevalent in South and East Asian ethnicities and is a more frequent cause of severe thiopurine toxicity than TPMT deficiency, but uptake of testing has been slow.

The association between severe fluoropyrimidine toxicity and DPD deficiency was reported in 1985. Five variants in the DPYD are associated with early, severe toxicity to capecitabine and 5-fluorouracil chemotherapies. Testing for these variants prior to the start of therapy is cost effective, reduces hospital admissions due to toxicity and saves lives. Following reports in the popular press, patients are increasingly requesting testing prior to therapy. However, uptake of testing has been slow.

In conclusion, genetic variation in three enzymes are associated with severe, life-threatening toxicity to purine and pyrimidine drug analogue therapy. Testing for only one of these, TPMT, is in routine clinical practice.