An observational study of a prospective, analytical, monocentric cohort which does not modify the patient care because the phenotyping of the dihydropyrimidine déshydrogénase DPD (lymphocyte activity) is already carried out in routine hospital. The analysis will be based on clinical, radiological and biological criteria.

Used for nearly sixty years, 5-fluorouracil (5-FU) is the oldest drug prescribed in the treatment of digestive cancers and, still today, the most prescribed drug in digestive cancer. 5-FU belongs to the class of anti-metabolites. There is an oral precursor of 5-FU which is currently available and used in digestive cancer: Xeloda® (capecitabine).

The problem is that fluoropyrimidines are 80% metabolised to 5-fluoro-5,6-dihydrouracil (5-FUH2) by a key enzyme: dihydropyrimidine dehydrogenase (DPD). There is inter-individual variability in the activity of this enzyme, partly related to genetic factors [1]. An abundant literature has already shown that a decrease in the activity of this enzyme results in an increase in the half-life of 5-FU. Thus, patients with a deficit in DPD activity have a risk of overexposure to chemotherapy in this class, and consequently an increased risk of acute, early and severe toxicity.

Since the 1990th years, many authors have shown that there is considerable inter-individual variability in the plasma concentration of 5 FU after bolus [5] or continuous infusion [6]. These variations in plasma concentrations of 5FU are probably related to a 5-FU catabolism variability related to the activity of DPD.

It has also been shown that there is a correlation between 5FU plasma level, tumor response and toxicity [7, 8]. A Phase III study showed that there was a significant impact on the response rate (33.7% VS 18.3%) p = 0.004 if doses of 5-FU were adjusted to plasma levels of 5-FU versus body surface [9]. Median survival was 16 months in the control arm versus 22 months in the experimental arm (P = 0.08), with severe grade 3-4 toxicities statistically increased in the control arm (p = 0.03). An adaptation of doses of 5-FU to the body surface, as it is done today, is in conclusion insufficient to obtain reproducible plasma concentrations due to an interindividual variability of the metabolism, linked to the activity of the DPD .

These data suggest that DPD activity, by modulating plasma concentration, may be a predictive factor of fluoropyrimidine response and also a prognostic factor. To our knowledge, there are no studies that have demonstrated a direct link between DPD activity, tumor response, and survival impact. Recently, Chamorey and al. [10] showed that the high level of DPD enzyme activity (> 0.30 nmol / min / mg protein) was significantly correlated with lower overall survival and progression-free survival. This was a retrospective analysis of 130 patients treated with fluoropyrimidines regardless of the tumor primitive (digestive, breast, ENT).

These results, presented in spite of their significance, are the result of a retrospective study that is subject, as a minimum, to selection bias. It therefore seemed important to make a prospective study, centered on digestive cancers. The prospective nature will limit selection bias and restriction to digestive cancers in which fluoropyrimidines have a main function and limit confunding biases. The selection of the first and second lines of palliative chemotherapy will increase the population studied and thus the power of the statistical analysis.

If the initial results were confirmed, this will allow DPD to be approached from a new angle and will encourage multicenter controlled prospective studies with the Antoine-Lacassagne center as the national promotor.

In our study, we decided to evaluate DPD activity using an enzymatic radio technique that directly evaluates the activity of the enzyme in blood lymphocytes, which is the original technique and the oldest.

Inclusion Criteria:

• Be > or = to 18 of age on day of signing informed consent
• ECOG 0 to 2,
• Patient with digestive cancer, all histologic type,
• Patient who will receive a metastatic or locally-advanced first or second line treatment by 5-FU or capecitabine,
• Patient with measurable lesions based on RECIST 1.1 criteria,
• Be willing and able to provide written informed consent/assent for the trial,
• Health care insurance available

Exclusion Criteria:

• Be < to 18 of age on day of signing informed consent,
• Patient without measurable lesion based on RECIST 1.1 criteria,
• ECOG > 2,
• Patients refusing to participate in the study or unable to give an oral consent,
• Contraindication of 5-FU or capecitabine treatment,
• People particularly vulnerable as defined in Articles L.1121-5 to -8 of the French Healthcare Code, including: person deprived of freedom by an administrative or judicial decision, adult being the object of a legal protection measure or outside state to express their consent, pregnant or breastfeeding women.

Metastatic or locally-advanced digestive cancer