Page 76 - Read Online

Page 118 De Mattia et al. Cancer Drug Resist 2019;2:116-30 I http://dx.doi.org/10.20517/cdr.2019.04

Up to date, even though more than 160 single-nucleotide polymorphisms (SNPs) have been reported
in DPYD, only four of them presented a sufficient evidence of a clinical impact to be included in the
current guidelines. The most dysfunctional variants known for their role in impairing DPD activity are
DPYD*2A (rs3918290) and DPYD*13 (rs55886062) associated with an almost complete protein deficiency in
homozygous individuals. C.2846A>T (rs67376798) and c.1236G>A (DPYD-Hap-B3; rs56038477) are instead
related to a moderate loss of protein function. A number of studies described the impact of these SNPs on
the outcome of an FP-based treatment in term of risk of developing severe to life-threatening toxicity both in
large randomized clinical trials [14,15] , and in retrospective collection of unselected patients from the current
[16]
clinical practice . More recently, the pre-emptive genotyping of DPYD*2A, coupled with a front-line dose
[17]
reduction was demonstrated to be feasible and effective in preventing toxicity occurrence . In an effort to
provide a global DPD metabolizing status for each patients and to provide personalized dosing guidelines
based on the patients genotype for the four variants, a gene activity score was developed, taking into account
[18]
the diplotype allelic combination of the variants . Based on the single allele score, a patient’s specific
estimation of the protein functionality is obtained, conferring to the patients a gene activity score ranging
from 0 (completely dysfunctional) to 2 (completely functional), allowing to tailor the FP starting dose based
on the patients genotype. The most recent version of shared pharmacogenetic guidelines are based on the
[19]
definition of the patient gene activity score .
Our group demonstrated in a large population of 763 CRC patients treated with a chemotherapeutic
regimen including FP in whichever setting of treatment, that carriers of at least one variant in the four
DPYD SNP panel were significantly more exposed to the risk of both acute (within the first three cycles)
and chronic clinically relevant toxicity, defined as grade 3 or higher non hematological or grade 4 or
[10]
higher hematological . When stratifying the patients according to their DPYD activity score the trend
for the toxicity risk was significantly increasing with the decrease of the patients activity score. The
significant relationship between DPYD genotype and toxicity risk has been confirmed also when FP-based
chemotherapy is given in combination with radiotherapy in more than 800 patients, confirming that even in
this specific clinical setting an upfront dose adjustment based on DPYD genetic profile would be beneficial
[20]
for patients safety .

Available pharmacogenetic guidelines for FP are currently based on the screening of four genetic variants
in the DPYD gene (www.pharmgkb.org). DPD is the first and rate-limiting enzyme of FP catabolic pathway
[Figure 1]. A recent publication reported the results of a prospective clinical trial testing the utility and
feasibility to apply current pharmacogenetic guidelines in the clinical practice to patients treated with an
[19]
FP-based therapy . The trial enrolled more than 1,100 cancer patients and demonstrated that an up-front
adjustment of FP dosage based on the genotype of the patients contributed to expose the patients to a more
homogeneous drug plasma level. This resulted in a similar risk to develop severe toxicity between risky
variants carriers and non-carriers. The risk to develop severe toxic side effects in poor metabolizer patients
[21]
was decreased in comparison to historical cohorts of patients treated according to the standard practice .
A clinical implementation experience of the DPYD four relevant SNPs typing prior to FP treatment has
been reported with a positive effect on patients outcome and with a successful up-take by the treating
oncologists . However this is not the common practice in the majority of the Health Care Systems, and the
[22]
[23]
clinical utility of the test is far from being widely acknowledged . One of the major concerns is the lack of
formal health technology assessment studies, including cost-effectiveness and cost-consequences evaluation.
We previously reported that the costs required to manage toxicity related to chemotherapy are associated to
[24]
the patient’s genotype for specific pharmacogenetic toxicity risk variants . Last year we performed a cost-
analysis in a large group of Italian CRC patients treated with FP-based chemotherapy demonstrating that
carriers of the four DPYD variants have higher toxicity management costs than non-carriers (carriers €2,972;