Page 246 - Personalised medicine of fluoropyrimidines using DPYD pharmacogenetics Carin Lunenburg

Chapter 9
After further investigations and consult with the attending nephrologist, it appeared that this patient had a history with allogeneic hematopoietic stem cell transplantations. Saliva samples from both patient and stem cell donor were obtained and genotyped, which showed a CYP3A5*1/*3 genotype for the patient and CYP3A5*3/*3 genotype for the donor. At Erasmus MC, in 2011 a discrepancy for CYP3A5*3 was found. Thereafter, 300 patients were genotyped at Erasmus MC in a single run, using TaqMan. Comparing the outcome with PCR-RFLP revealed three discrepancies. Upon sequencing, the TaqMan assay appeared to have been wrongly addressing the wild type status, possibly due to allele dropout caused by a variant located at one of the TaqMan primers.
At Erasmus MC, in 2016 a DPYD result was found discrepant between PCR-RFLP and TaqMan-based validated assays, PCR-RFLP giving heterozygosity *1/*13, indicating a 50% dose reduction, whereas the other assay indicated wild type for the four SNPs tested (regular starting dose). Both methods were repeated, confirming the earlier results. Direct sequencing revealed that wild type was the correct outcome. This was the only discrepant finding for DPYD.
At Erasmus MC, the genotype of three patients showed discrepancies for CYP2D6*6. The AmpliChip assay showed wildtype for CYP2D6*6, whereas TaqMan determined CYP2D6*6. The tests were repeated, and sequencing of the samples revealed the correct genotype (CYP2D6*6). Like the CYP3A5*3 discrepancies, a possible explanation was a variant located at one of the AmpliChip primers, resulting in allele dropout and thereby misclassification, missing the CYP2D6*6 allele. Previously, another discrepancy for CYP2D6*6, possibly caused by allele dropout, was described by Rasmussen et al.13 We are not aware of publications on other discrepancies of CYP3A5*3 or DPYD*13.
In total, eight discrepancies were found at Erasmus MC in 72,910 SNPs tested in duplo (0.01%). For LUMC, one discrepancy was found in 16,932 SNPs tested in duplo (0.006%), which could be attributed to chimerism in a patient with previous stem cell transplantation. In total, the probability of finding a discrepant result when using two independent techniques thus calculates to be 0.01%.
244