Page 78 - Read Online

Page 120 De Mattia et al. Cancer Drug Resist 2019;2:116-30 I http://dx.doi.org/10.20517/cdr.2019.04

PHARMACOGENETIC EXPLORATORY MARKERS OF FLUOROPYRIMIDINES
5-FU, is a pyrimidine analog, belonging to the antimetabolites category, and acts through its active
metabolite 5-fluoro-2-deoxyuridine-5’-monophospate that is a nearly irreversible inhibitor of thymidylate
synthase (TS, encoded by TYMS). A comprehensive overview of FP-related pathways is depicted in Figure 1.

Beside the catabolic enzyme DPD, other important pharmacogenes with an impact on the risk to develop
[30]
FP related toxicity, belong to the folate metabolism pathway . An alteration in the functionality of the
proteins involved in this pathway could modulate the activity of FP and consequently the drug toxicity
profile. Particularly, variations in genes encoding TS and 5,10-methylenetetrahydrofolate reductase
(MTHFR) represent the most investigated markers [31,32] . More recently other FP related genes as well as
specific immuno-related genetic profiles have been considered as potential markers of adverse drug reaction
occurrence. The following paper sections will cover the most recent literature regarding the emerging
pharmacogenetic data related to the development of adverse drug reactions in CRC patients treated with FP.

Dihydropyrimidine dehydrogenase
As already mentioned, DPD is the first and rate-limiting enzyme in the FP catabolic pathway converting
5-FU to dihydrofluorouracil; the activity of this protein is shown to be highly variable among individuals
spanning from partial (about 3%-5% of the entire population) to complete loss of the enzyme functionality
[33]
(about 0.2%-0.3% of the entire population) . DPD deficiency is demonstrated to be partly linked to
some genetic polymorphisms and to be responsible of life-threatening early toxic events that occur in
about 0.5% of patients receiving 5-FU [33,34] . DPYD*2A (rs3918290) DPYD*13 (rs55886062), c.2846A>T
(rs67376798, D949V) and c.1236G>A-HapB3 (rs56038477), the genetic variants included in the international
pharmacogenetic guidelines for drug adjustments do not cover all the cases of DPD deficiency and their
screening could not predict with sufficient sensitivity the risk of early severe toxic events induced by FPs.
The DPYD gene is highly polymorphic and other genetic variants are likely to impact the enzyme activity
and the risk to develop severe toxicity. Some additional DPYD variants (e.g., rs75017182, rs1801158 rs2297595,
rs17376848, rs72549309, rs1801265, rs1801160) have been suggested to be clinically relevant predictor of FPs
associated toxicity and could be considered, alone or in score combination, for improving available dosing
guidelines [Table 1] [14,18,21,22,33,35-37] . However at present robust evidence has been generated only for the
[38]
DPYD*6 (rs1801160) variant [14,37,38] . Ruzzo et al. highlighted that DPYD*6, together with DPYD rs2297595
and DPYD*2A, were not only linked to a higher risk of toxicity but also to the time-to-toxicity parameter,
emphasizing the acute occurrence of genetically determined toxicity occurrence. The introduction of the
dimension of time allows a better characterization of the gene-linked toxicity profile and is particularly
sensitive in the case of few observations (i.e., rarity of some genotypes) and genetic variants with moderate
functional effects. Moreover, the application of the recently developed high-throughput next generation
sequencing (NGS) technology to samples obtained from CRC patients exhibiting extreme toxicity
phenotypes, will allow to investigate and possibly identify additional novel and rare variants, significantly
impacting the DPD activity, that could be integrated into the pharmacogenetic algorithm to further improve
the FPs administration [18,36,39] .

The combination of the DPYD genetic variability with the measurement of the DPD enzymatic activity is
probably the most effective strategy to identify patients at risk of severe and potentially fatal FPs-associated
toxicityand has been successfully integrated in the clinical practice improving patients outcome [44,45] . Some
Dutch and French studies have recently demonstrated how Uracil levels could be used successfully to
anticipate severe toxicities [46,47] . A recent study demonstrated that an algorithm, integrating information of
high-throughput DPYD genotyping with the determination of DPD enzyme phenotype, could effectively
identify DPD deficient individuals, and is suitable for routine clinical application . A score predicting the
[33]
risk for severe toxicity based on DYPD mutation and 5-FU degradation rate among other clinic-pathological
features (i.e., age, number of drugs administered) was also developed and reported to be an easy and low-