|dc.description.abstract||Recent publications have repeatedly highlighted the unreliability of academic
research when translated to industry. This subject has been debated in publications
such as Nature Drug Discovery, RQA monthly, Life Sci VC and Lab Times with many
reasons cited including research bias, academic self-interest, apathy, pressure to
publish and “massaging data to fit the hypothesis”.
Resistance in colon and pancreatic cancer is well documented but the literature is
conflicted. Perhaps this is because of biological variability but it could be due to
experimental design. The aim of this project is to conduct a study, mindful at each
stage of how quality management might inform the interpretation of findings
Colorectal cancer is the third most common cancer for both males and females in the
UK and makes up 13% of all cancers. 5-fluorouracil (5-FU) has been, and remains,
the first line drug of choice for treating colorectal cancer, both as a monotherapy and
in combination with other drugs. However, the response rate is around 20% when
used singly and just under 50% when used in combination.
Pancreatic cancer is the tenth most common cancer in the UK and makes up 2.55%
of all cancers. The average survival time for a pancreatic cancer patient following
surgical resection is approximately 11-20 months. This drops to 6-11 months in
unresectable localised tumour patients. For patients with unresectable, metastatic
disease the average survival time is 2-6 months.
Although 5-FU and gemcitabine remain the first line treatment for colorectal and
pancreatic cancers, there are no definitive markers of resistance which may be used
to predict response to these drugs.
Investigation into prognostic and predictive biomarkers based upon putative
resistance to these drugs is crucial to tailor therapy in patients with a potentially very
limited lifespan. Automated quantitative analysis (AQUA) is a novel image analysis
system of quantifying protein expression levels, including potential biomarkers, in
This thesis addresses the hypothesis that quantitative expression of protein
expression can be used to identify putative biomarkers of gemcitabine resistance in
pancreatic cancer, and that the methodology is transferrable to other disease types.
Using pancreatic cancer as the lead disease this project aims to measure the
expression levels of proteins critical for both pathways – thymidylate synthetase (TS),
ribonucleotide reductase subunit M1 (RRM1), ribonucleotide reductase subunit M2
(RRM2), cytidine deaminase (CDA), human equilibrative nucleoside transporter 1
(hENT1) - and pathway specific proteins – dihydropyrimidine dehydrogenase (DPYD)
and thymidine phosphorylase (TP) for 5-FU- and a gemcitabine pathway-specific
protein, deoxycytidine kinase (DCK). Thereafter, in a briefer study, the application of
the approach to colorectal cancer was investigated.
Archival FFPE blocks were used to construct tissue microarrays (TMAs) in
quadruplicate and patient demographics were collated. An antibody validation
algorithm was formulated to authenticate antibodies prior to use. A quality
programme was initiated and maintained throughout the length of the project to
ensure data integrity and reproducibility of results. The expression levels of the
proteins were quantified using Automated QUantitative Analysis (AQUA).
Low CDA and RRM1 expression were associated with longer disease free survival in
all patients. Medium CDA expression was associated with longer overall survival in
the patients who did not receive any chemotherapy. Low RRM1 expression was
linked with longer overall and disease free survival in the gemcitabine cohort.
Multivariate analysis showed that a high Lecca value was linked to increased survival
time, statistically significant in the gemcitabine group – p = 0.007 for overall survival
and p = 0.07 for disease free survival.
24% of patients who received no chemotherapy (high expression group) were still
alive at 48 months and 25% of patients who received gemcitabine (high expression
group) were still alive at 72 months.
In the low expression groups, both sets of patients had the same amount of time to
disease progression, 14 months.||en