Diagnosing Thyroid Cancer Using a Blood Test
This project will utilize two powerful technologies for diagnosing endocrine diseases:
proteomics and genetic (RNA and DNA) markers. Proteomics is a relatively new, rapidly
expanding and exciting area of biomedical research (Robin et al, 2009, Frolich et al, 2009).
Posttranslational modifications of proteins are critical for function. Modified proteins may
be markers of cancer phenotypes and therefore be useful tumor markers (Narimatsu et al,
2010). Proteomic research in thyroid cancer is in its infancy (Krause et al, 2009). The
available studies on thyroid cancer have utilised tissue rather than serum samples,
nonetheless the results are encouraging (Brown et al, 2006, Wang et al, 2006, Netea-Maier
et al, 2008, Krause et al, 2007, Moretz et al, 2008).
Genomic markers of thyroid cancer have been described and are increasingly being used on
biopsy material for accurate diagnosis. Among the described markers point mutations (BRAF
V600E, NRAS codon 61, HRAS codon 61), gene rearrangements (RET / PTC1, RET / PTC3, PAX8 /
PPARgamma) and other polymorphisms have been found to be useful (Nikiforova and Nikiforov,
2009, Ohori et al, 2010). There is good evidence that in recurrent thyroid cancer small
numbers of thyroid cancer cells can be detected in peripheral blood, in sufficient
quantities to detect thyroid-specific mRNA by RT PCR (Karavitaki et al, 2005, Barbosa et al,
2008, Milas et al, 2009). Most of these studies have focused on the detection of
thyroglobulin mRNA with moderate success. A significant difficulty with this approach is
that detection of thyroglobulin mRNA in peripheral blood cannot distinguish between the
presence of normal thyroid tissue or thyroid cancer.
The project is a collaborative venture between Newcastle Biomedicine, the NHS, and
Biosignatures Ltd (a North-East based proteomics diagnostics company). Biosignatures has
invested a great deal of research in optimizing sample handling and sample analysis, thus
giving rise to plasma proteomic protocols that are stable and suitable for large comparative
studies (Elliott et al, Jackson et al, 2010, Bramwell et al, 2007). The data generated from
plasma 2D gel electrophoresis and mass spectroscopy is analysed by proprietary "supervised
learning" technology. The system is given multiple examples of group classes (disease cases)
and from this derives a signature pattern ('proteomic fingerprint') that allows the classes
to be discriminated. This signature will then be validated against a novel patient dataset
to ensure robust disease status discrimination. The combination of this research and
technology can produce blood derived signatures of disease in an applied clinical setting
(Cash and Argo, 2009, Borthwick et al, 2009).
Thyroid cancer affects 2000 new patients in the UK per annum (Cancer Research UK). Once the
initial treatment of thyroid cancer is completed (thyroidectomy followed by radioiodine
ablation), monitoring is essential to detect residual disease or recurrence. Recurrence
rates in thyroid cancer are as high as 30% (Mazzaferri and Kloos, 2001) and can declare
themselves decades after initial treatment, so that patients have to be monitored regularly
for life. Monitoring for disease recurrence consists of iodine scans, an ultrasound scan of
the neck 6-8 months after initial treatment, and 6-12 monthly blood tests thereafter for the
serum marker thyroglobulin. Thyroglobulin is a valuable marker in many people with thyroid
cancer (Spencer and Fatemi, 2006). Unfortunately in approximately 30% of patients antibody
interference with the assay renders this test unreliable (Spencer and Fatemi, 2006). In such
cases patients are subjected to repeated scans, though a negative scan has a far less
predictive value than a negative thyroglobulin blood test when the analyte can be measured
reliably. We have selected thyroid cancer as the primary topic of study for proof of concept
for the following reasons:
- Current diagnostics technology (measurement of serum thyroglobulin) suffers from
interference of measurement of the analyte in 30% of cases, rendering this tumour
marker entirely unreliable when such antibodies are present. Attempts using
conventional biochemical analytical technology to overcome this problem over the past 3
decades have failed. Thus a proteomics/genomics approach has only to perform with a
better than 70% specificity to provide a superior diagnostic test.
- The potential cost savings to the NHS by the development of such a diagnostic test (by
avoidance of expensive scans) will be considerable.
- Exposure of patients to radiation from repeated scans will be reduced with obvious
- The study is non-interventional, will induce no additional discomfort, and is expected
to have no impact on the care received by participants at this stage.
- Extrapolation of such technology to the evaluation of thyroid nodules (present
clinically in 5% of the adult population) and even screening of the population for
thyroid malignancy, would have profoundly beneficial preventative and public health
Observational Model: Case Control, Time Perspective: Prospective
Proteomic markers of differentiated thyroid cancer
The primary objective of the study is to derive molecular (proteomic) diagnostic signatures that that can distinguish patients with recurrent / residual thyroid cancer from those with no residual disease.
National Health Service, UK: