Exome array analysis of adverse reactions to fluoropyrimidine-based therapy for gastrointestinal cancer.

Fluoropyrimidines, including 5-fluororacil (5FU) and its pro-drug Capecitabine, are the common treatment for colorectal, breast, neck and head cancers-either as monotherapy or in combination therapy. Adverse reactions (ADRs) to the treatment are common and often result in treatment discontinuation or dose reduction. Factors contributing to ADRs, including genetic variation, are poorly characterized. We performed exome array analysis to identify genetic variants that contribute to adverse reactions. Our final dataset consisted of 504 European ancestry individuals undergoing fluoropyrimidine-based therapy for gastrointestinal cancer. A subset of 254 of these were treated with Capecitabine. All individuals were genotyped on the Illumina HumanExome Array. Firstly, we performed SNP and gene-level analyses of protein-altering variants on the array to identify novel associations the following ADRs, which were grouped into four phenotypes based on symptoms of diarrhea, mucositis, and neutropenia and hand-and-foot syndrome. Secondly, we performed detailed analyses of the HLA region on the same phenotypes after imputing the HLA alleles and amino acids. No protein-altering variants, or sets of protein-altering variants collapsed into genes, were associated with the main outcomes after Bonferroni correction. We found evidence that the HLA region was enriched for associations with Hand-and-Foot syndrome (p = 0.023), but no specific SNPs or HLA alleles were significant after Bonferroni correction. Larger studies will be required to characterize the genetic contribution to ADRs to 5FU. Future studies that focus on the HLA region are likely to be fruitful.

Let's get personal: predicting thiopurine and fluoropyrimidine toxicity.

The US FDA now recognizes the need to individualize treatment paradigms using biomarkers that predict response to therapy. In clinical practice the best example of this is TPMT testing, which is used to rationalize the starting dose of azathioprine and mercaptopurine. The more recent addition of drug metabolite monitoring means that thiopurine therapy can now be personalized to unprecedented levels. Of interest, parallels exist between TPMT deficiency as an explanation for thiopurine toxicity and DPD deficiency in fluoropyrimidine toxicity. For these drugs, variations in a single locus predict severe toxicity. However, while TPMT testing has translated into routine clinical practice, DPD testing has not. This article summarizes the recent research investigating interindividual differences in the metabolism of thiopurine and fluoropyrimidine drugs, and explores the attitudes which influence the uptake of pharmacogenetic testing.

Purine nucleoside phosphorylase deficiency: a mutation update.

Purine nucleoside phosphorylase (PNPase) deficiency is an autosomal recessive disorder affecting purine degradation and salvage pathways. Clinically, patients typically present with severe immunodeficiency, neurological dysfunction, and autoimmunity. Biochemically, PNPase deficiency may be suspected in the presence of hypouricemia. We report biochemical and genetic data on a cohort of seven patients from six families identified as PNPase deficient. In all patients, inosine, deoxyinosine, guanosine, and deoxyguanosine were elevated in urine, and mutation analysis revealed seven different mutations of which three were novel. The mutation c.770A>G resulted in the substitution p.His257Arg. A second novel mutation c.257A>G (p.His86Arg) was identified in two siblings and a third novel mutation, c.199C>T (p.Arg67X), was found in a 2-year-old female with delayed motor milestones and recurrent respiratory infections. A review of the literature identified 67 cases of PNPase deficiency from 49 families, including the cases from our own laboratory. PNPase deficiency was confirmed in 30 patients by genotyping and 24 disease causing mutations, including the three novel mutations described in this paper, have been reported to date. In five of the seven patients, plasma uric acid was found to be within the pediatric normal range, suggesting that PNPase deficiency should not be ruled out in the absence of hypouricemia.