Identification of the mechanisms of 5-fluorouracil (5-FU)-induced Gastrointestinal (GI) toxicity in human colon

One of the major complications during pharmacological treatment of patients is
the occurrence of adverse drug reactions (ADRs), for which the most affected
organs are the liver, kidney, heart and the gastrointestinal-immune system. In
comparison to the other organs, less progress has been made on human*relevant
prediction of drug*induced intestinal toxicity, and it is clear that large data
gaps are currently present. Therefore, new experimentation is required to
assess GI toxicity by analyzing gene profile responses induced by exposure of
GI system to different medicines. From a series of selected compounds, this
study will specifically focus on 5-fluorouracil (5-FU)-induced intestinal
toxicity. 5-FU was chosen as the first pharmaceutical to be studied in view of
its wide use in cancer-therapy. The study will start by inclusion of metastatic
breast cancer patients that will be receiving 5-FU as part of their medical
treatment. These patients will receive an oral monotherapy consisting of cycles
of two weeks treatment with capecitabine followed by one week without
treatment. Capecitabine is a pro-drug that, after intake is converted into
5-FU, which is the actual cytostatic agent. After the first cycle, alterations
in the physiology and morphology of colon biopsies will be evaluated, and
5-FU-induced transcriptomics signatures will be established through
quantitative RNA sequencing. By doing so, we intend to obtain a better insight
in the changes in the gene profile of biopsies caused by 5-FU as well as on its
mechanism of action and toxicity. Eventually, we intend to compare this data
with in vitro organoids and animal models and determine whether there is any
relation or translatability between the models. The study can therefore be
regarded as a validation study of potential in vitro and in vivo alternative
models for studying drug induced GI-toxicity.

In this study, the primary aim is to evaluate the influence of 5-FU exposure on
the gene expression profile in human colonic biopsies from healthy tissue. In
addition, morphological changes and the molecular toxicity biomarkers
citrulline and calprotectin will be measured as secondary outcome in blood and
faeces, respectively.

This human study will include 20 metastatic breast cancer patients presenting a
healthy colon. Each participant will undergo proctoscopy without any bowel
preparation before and after a 2-weeks treatment cycle with 5-FU; rectal
biopsies and rectal swap will be taken by a specialised nurse. Rectal swaps
will be used in further studies on the impact/influence of the microbiome on
the development of GI toxicity. Additionally, patients will be asked to donate
a venous blood sample for the analysis of citrulline, 5-FU and its metabolites,
and a stool sample for the analysis of faecal calprotectin. Data analysis to
evaluate 5-FU-induced toxicity in the colon will be performed as endpoints of
the study.

During the medical treatment, subjects will follow a sequence of two different
periods: a two weeks treatment period with capecitabine followed by one week
without treatment. This will be repeated during three cycles as part of their
cancer therapy regimen. However, biopsy samples will be only collected after
one cycle. Dose will also be adjusted according to the
dihydropyrimidine-dehydrogenase (DPD) status (based on genetic testing of
polymorphisms).

The burden/risk/benefit associated with participation will be as follows:


- Rectal biopsies will be taken before and after treatment with capecitabine.
Biopsies will be collected during proctoscopy and therefore no specific bowel
preparation is necessary. This procedure is used daily in medical practice and
is a relatively safe examination method. Complications are very rare. Bleeding
may occur from biopsies, but is minimal and stops quickly. If rectal bleeding
persists, the patient must report it immediately;
- Rectal swaps, blood and faecal samples will be collected to one tube
separately each time (one sample before and one after 2 weeks of treatment).
The risks related to the collection of these samples are minimal. Rectal swaps
and faecal samples are non-invasive. Blood sampling is performed by
venepuncture, which is a standard procedure with low risk of bleeding;
- Genotyping for dihydropyrimidine-dehydrogenase (DPD) deficiencies prior to
treatment, determined by Sanger sequencing of polymorphisms: c.1905+1G>A (*2A)
in intron 14, c.1679T>G (*13) in exon 13, c.2846A>T (rs67376798) in exon 22 and
c.1129-5923C>G (rs75017182) in intron 10. This is part of the standard protocol
before treatment of these patients, and therefore does not add an additional
burden.