Page 223 - Personalised medicine of fluoropyrimidines using DPYD pharmacogenetics Carin Lunenburg

Evaluating the implementation of prospective DPYD genotyping
days was required and capecitabine therapy was permanently terminated (Table 4, patient 13).
Discussion
In this study, the successfulness of routine application of a prospective DPYD screening program followed by pharmacogenetically guided dose recommendations was studied. The percentage of patients in which screening was performed was relatively high: 86.9% of all eligible (newly prescribed 5-FU or capecitabine) patients. In the study period, 13.1% of the patients were not screened prior to receiving 5-FU or capecitabine therapy, which on average comes down to one patient per month. Follow-up of dose recommendations given by the pharmacist were applied in all cases except one, resulting in a high acceptance.
Our study has several limitations. Due to the retrospective design of our study, available data may not always have been fully complete. For example for some patients, it was not possible to retrieve why DPYD screening was not requested or whether a patient actually started fluoropyrimidine therapy. In addition, the study was performed with data obtained in a real world clinical setting instead of a regulated and controlled case report form. We had to manually check patient files to obtain specific information and not all physicians may have systematically annotated CTC-AE grading continuously to describe toxicity. Due to the low number of DPYD variant carriers our study was not powered to formally test the effect of DPYD screening on fluoropyrimidine-induced toxicity and only explorative analyses could be performed.
In this study, we determined the level of routine application of DPYD screening in daily practice, which increased at the end of the study period to 90–100%. This might indicate that prescribers were undergoing a learning or acceptance curve following the initial start, and were getting used to apply DPYD genotyping increasingly in their daily routine.
We believe patients do not need to be genotyped if previous 5-FU or capecitabine usage without toxicity is known or if patients were genotyped (DPYD) or phenotyped (DPD) previously. However, within the 41 (13.1%) remaining patients legitimate reasons can still exist (e.g., well-tolerated treatment before 2013 with 5-FU or capecitabine), but might not have been filed in the medical record. Therefore we can conclude the 90–100% (≤ 1 patient not tested per month) rate was an effective prospective DPYD screening implementation. Disputable is, if this clinical acceptance can become 100% continuously. In order to support the clinical implementation, the use of a clinical decision support system might be suitable. In LUMC a clinical decision support system entitled adverse drug event alerting system (ADEAS) is used in daily practice in the hospital pharmacy of LUMC.22 This system is used by hospital pharmacists to systematically select patients at risk of possible adverse drug events. It retrieves data from several information systems, and uses clinical rules to select the patient at risk of adverse drug events.
As mentioned before, sensitivity of genotyping is relatively low (<14.5% for DPYD*2A and c.2846A>T combined).11 Even if all patients with a DPYD variant are identified and treated with an appropriately reduced dose, not all fluoropyrimidine-related toxicity can be prevented. Adding a DPD phenotyping test may increase sensitivity, but is expensive and logistically challenging to implement in clinical practice.13 SNPs located in other genes than DPYD
8
221