Page 259 - Personalised medicine of fluoropyrimidines using DPYD pharmacogenetics Carin Lunenburg

Phenotyping assays for predicting DPD deficiency
Phenotyping assays
General performance of assays
For the DPD enzyme activity measurements, the variance was estimated to be 8.6 ± 1.4 nmol/ (mg*h). For centre as covariate, the variance was estimated to be 6.5 ± 5.7 nmol/(mg*h). The deviations between centres in general (intra class variation) was 43.0%, therefore the general reliability of the DPD enzyme activity measurements was 57%. For the endogenous DHU/U ratio, endogenous uracil levels, 2-13C-uracil breath test and the oral uracil loading dose assay the general reliabilities were 74.1%, 92.9%, 73.5%, and 94.3%, respectively.
Endogenous DHU/U ratio
The endogenous DHU/U ratio was determined in 1,037 patients. Results of wild-type patients (non-carriers of the four DPYD variants, N=955) were compared between seventeen study centres. The endogenous DHU/U ratio differed significantly in nine study centres compared to the reference centre (lowest divergent mean DHU/U ratio 5.9, to the highest divergent mean DHU/U ratio 13.9, p<0.001). It appeared that age was significantly associated with the outcome of the DHU/U ratio (p<0.001). Age was differently distributed between the centres (p<0.001). The lowest statistically divergent mean DHU/U ratio was not as low as the suggested DHU/U ratio cut-off value (4.31)20 for DPD deficient patients, therefore no patients were excluded. The median, interquartile range (IQR) and standard deviation (SD) of each assay are shown in Table 2.
Endogenous uracil levels
10
Endogenous uracil levels were determined in 1,037 patients. Results of wild-type patients (N=955) were compared between seventeen study centres. The endogenous uracil levels differed significantly in four study centres compared to the reference centre (lowest divergent mean uracil level 8.3 ng/ml, to the highest divergent mean uracil level 18.8 ng/ml, p<0.001). It appeared that gender was significantly associated with the outcome of the uracil levels (p=0.030), with lower uracil levels in females. Males and females were significant differently distributed between the centres (p=0.046). The divergent results were substantially higher, even higher than the previously suggested cut-off value (13.9 ng/ml)20 for DPD deficient patients, therefore the data were considered unreliable and patients recruited in these centres (N=172) were excluded from further analyses. The endogenous uracil levels and endogenous DHU/U ratio were correlated to time of last meal that was eaten, to study the influence of food on the uracil levels. No correlation was found (Supplementary Figure 1), therefore time of food intake was not taken into account as covariate in further analyses.
DPD enzyme activity
The DPD enzyme activity in PBMCs was determined in 92 patients. Results of 82 wild- type patients were compared between study centres. The mean DPD enzyme activity was significantly lower in one of the four participating centres (5.23 nmol/(mg*h) versus 10.89 nmol/(mg*h) in the reference centre, p<0.001). These results were substantially lower, even lower than the cut-off value (≤5.9 nmol/[mg*h])6 for DPD deficiency, therefore the data were considered unreliable and patients recruited in this centre (N=19) were excluded from further analyses in which DPD deficiency was taken into account.
257