Page 127 - Personalised medicine of fluoropyrimidines using DPYD pharmacogenetics Carin Lunenburg

Supplement
Rationale of
the therapeutic recommendation
NOTE 1: The dose reduction described here is well substantiated for *1/*2A and c.1236G>A/ c.1236G>A. The dose reduction for patients with c.2846A>T (c.2846A>T/c.2846A>T or c.1236G>A/c.2846A>T) is based on, among other factors, the dose reductions identified for *1/c.2846A>T.
NOTE 2: If a patient has two different genetic variations that result in a partially functional DPD enzyme (e.g. c.2846A>T and c.1236G>A), this recommendation applies if the variations are on a different allele. If both variations are on the same allele, the gene activity score is between 1 and 1.5, depending on whether and how the two gene variations influence each other and on other factors that influence the DPD activity. Whether a gene activity score
of 1 or 1.5 needs to be assigned in the case of two different genetic variations can only be 4 determined by measuring the enzyme activity (phenotyping).
For 25 patients with genotype *1/*2A, one with genotype *1/*13, one with genotype c.2846A>T/c.2846A>T and one with genotype c.1236G>A/c.2846A>T, the weighted average of the dose adjustments calculated based on 5-FU clearance or AUC was a reduction to
45% (18-49%, median 33%). Because the relatively low median was caused by the low values found in the two smallest studies (n = 1 and n = 2 respectively), it was decided to base the dose recommendation on the weighted mean. The weighted mean of 45% was translated to 50% to be more achievable in clinical practice. This is similar to the dose reduction to 56% and 60% of the standard dose found by Deenen 2011 and Meulendijks 2016 when investigating patients with respectively *1/*2A and c.1236G>A/c.1236G>A in whom toxicity-guided dose adjustments were made. It is also similar to the mean tolerated dose of 55% found by Henricks 2017 JCO Precis Oncol for 2x c.1236G>A/c.1236G>A, 1x c.1236G>A/c.2846A>T and 1x c.2846A>T/c.2846A>T, although in this study a strongvariation between patients (and genotypes) was found. In addition, Deenen 2016 found no difference in toxicities between 18 patients with *1/*2A on an initial dose of maximally 50% of the standard dose and *1/*1-patients on the standard dose. Lunenburg 2016 found no grade
≥ 3 toxicity when treating three patients with *1/*2A with an initial dose of 50% of the standard dose.
There are no data on clearance or AUC for c.1236G>A/c.1236G>A and only scarce data
on clearance or AUC or on maximum tolerated dose in clinical practice for c.2846A>T/ c.2846A>T and c.1236G>A/c.2846A>T. Deenen 2011 found a dose reduction to 74% of
the standard dose for patients with *1/c.2846A>T when toxicity-guided dose adjustments were made. Extrapolation of the required dose reduction for *1/c.2846A>T would lead to
a required dose reduction to 50% for c.2846A>T/c.2846A>T. This is equivalent to the dose reduction for *1/*2A and c.1236G>A/c.1236G>A, which are also in the gene activity score 1 group.
Instead of dose adjustment, physicians may also choose an alternative.
Increase in the AUC of 5-FUby 103% (16x *1/*2A), 127% (1x c.1236G>A/c.2846A>T) or 766% (1x c.2846A>T/c.2846A>T).
52-80% decrease in clearance.
69-109% increase in half-life.
7 of the 10 studies and two meta-analyses found an increased risk of grade ≥ 3 toxicity. Increased grade ≥ 3 toxicity: OR = 4.67-24.9; RR = 4.40-9.76. The highest ORs were found for haematological toxicity. There was a 74-793% increase in the percentage of patients with grade ≥ 3 toxicity. Out of 48 patients with genotype *1/*2A in published cohort studies, 73% developed grade ≥ 3 toxicity. The allele frequency of *2A in a group with grade III/IV toxicity was 1548-2879% higher. Toxicity generally occurred in the first cycle. Six patients died due to toxicity, including two that had used capecitabine.
No association with grade ≥ 3 toxicity was found for breast cancer patients receiving adjuvant/neoadjuvant therapy with 5-FU, epirubicin and cyclophosphamide in a phase II study that showed 94% grade ≥ 3 toxicity and in a small study of 21 patients with grade ≥ 3 toxicity. 5-FU toxicity is not common in breast cancer patients treated with this combination therapy.
table continues
Kinetic consequence
Clinical consequence
125