Spectrum of autoantibody response to tumour associated antigens in normal population

Mathew, John (2013) Spectrum of autoantibody response to tumour associated antigens in normal population. DM thesis, University of Nottingham.

PDF (Thesis - as examined) - Requires a PDF viewer such as GSview, Xpdf or Adobe Acrobat Reader
Download (6MB) | Preview


Early detection and diagnosis of cancer has a significant impact on cancer specific mortality as shown in randomised screening trials for breast, colon and lung cancer. However, current screening tests have limitations as they reduce cancer specific mortality for breast, colon and lung cancer by only 24%, 16% and 20% respectively, and partly as a result cancer continues to be one of the common causes of death in the developed world(Aberle et al, 2011; Hardcastle et al, 1996; Larsson et al, 1996). One of the other problems associated with current screening tests is patient compliance (mammography, colonoscopy and CT) (Jonnalagadda et al, 2012; Maurer, 1995; Pooler et al, 2012).

Diagnosing cancers with a blood test by identifying tumour associated antibodies in serum is a novel method which may allow the identification of early stage cancers and hopefully it would have greater patient acceptability. These tumour associated antibodies represent an indirect amplified signal generated as a response by the immune system to tumour associated antigens secreted early on in development of cancer.

One of the common limitations of many autoantibody studies is the selection of appropriate controls - or the lack of such. One common problem is the use of limited number of normal individuals without cancer as controls, the data from which may not be representative of the normal population as a whole (Stockert et al, 1998). In addition not only the numbers of controls are often incorrect but also the age of the controls. Many studies report using ‘blood donors’ as controls(Guy et al, 1981) and clearly for most tumour types this involves both a younger population and also a relatively health population which may not always be reflective of the individuals to be screened (e.g. compare heavy smokers).

We hypothesised that autoantibody response to cancer associated antigens may alter with demographics (age, sex, and smoking) and the aim of our study was to identify the spectrum of response of tumour associated antigens in a range of demographic groups within the normal population of the East Midlands.

EarlyCDT-Lung™ is a simple commercial blood test which is reported to aid the early detection of lung cancer. The technology was initially developed in the laboratories of the Division of Breast Surgery and subsequently underwent further development by the university spinout company, Oncimmune. EarlyCDT-Lung initially measured autoantibodies (AAbs) to six cancer associated antigens (p53, NY-ESO-1, CAGE, GBU4-5, Annexinl, and SOX2) and was reported to identify up to 40% of lung cancers, at both at eariy- and late-stage disease(Boyle et al, 2011; Chapman et al, 2012; Chapman et al, 2011; Lam et al, 2011; Macdonald et al, 2012a; Macdonald et al, 2012b; Murray et al, 2010).The initial technical (Murray et al 2010) and clinical (Boyle et a;l 2011; Lam et al 2011) validation studies matched high risk individuals to every lung cancer patient. Controls were individually matched to a patient with lung cancer by age, gender and smoking history. As the 6 antigen panel had been developed and validated (Boyle et al, 2011; Murray et al, 2010) it was decided to proceed to assess the level of autoantibodies across the population and in particular to look at differences by gender and different decades of life. We used a semi-automated Enzyme linked immunosorbent assays (ELISA) to run the serum samples collected from individuals with no previous history of cancer.

Informed consent was taken prior to a detailed health questionnaire and then a blood sample by standard venipunture. The information acquired in the questionnaire included age, gender, smoking history, any autoimmune disease and family history of cancer.

Serum samples used in this thesis were collected from 2065 individuals. Male to female ratio was 1: 2.6(566:1487). Ratio of smoker versus ex-smokers versus non-smoker was 1: 2.4: 3.8 (285:672:1096). There was a fall in the number of smokers with increasing decade of life. The proportion of smokers and ex-smokers versus non-smokers remained approximately the same in both genders. Almost half the patients (964) had family history of some form of cancer. One hundred and eighty six subjects (9%) had personal history of autoimmune disease.

Analysis of autoantibody levels revealed a small but steady increase with increasing age for 4 out of the 6 antigens (p53, NY-ESO-1, CAGE and GBU4- 5). Except for CAGE, there was no significant difference in mean optical densities between males and females. For CAGE, when analysis of variance was used to adjust for run differences, there was no significant difference in mean optical densities between males and females.

Autoantibody response to all 6 cancer related antigens were consistently low in smokers. The rise in autoantibody response was more in the ex-smoker group compared to the other two groups suggesting the possibility of rebound effect when smoking is stopped. It reached statistical significance except in case of NY-ESO-1. Age matched analysis were done, and the differences were statistically significant for p53, GBU 4-5 and Annexinl. To explore further the "rebound" hypothesis further, the year of quitting for ex-smokers were extracted from the database. Any association between AAb levels and time lapse since quitting might provide support to this hypothesis. Very little difference was seen for most antigens back to 1970, but decades before that there was an observed increase in the mean AAb level for all antigens except SOX2. Further work would be required to establish such a rebound effect.

Family history and history of autoimmune disease did not have a significant impact on autoantibody levels.

Analysis of autoantibody levels in a large cohort of the normal population of the East Midlands revealed that age has a small but significant influence on the serum levels of certain autoantibodies to cancer related antigens. However, this could be confounded by the fact that incidence of cancer also increases with age, and would need further investigation and in particular longer followup of patients who have given blood in this research study to see which individuals have developed cancer.

Item Type: Thesis (University of Nottingham only) (DM)
Supervisors: Robertson, John
Chapman, Caroline
Keywords: Autoantibodies; Tumour antigens; Cancer detection
Subjects: QS-QZ Preclinical sciences (NLM Classification) > QW Microbiology. Immunology > QW501 Immunology
Faculties/Schools: UK Campuses > Faculty of Medicine and Health Sciences > School of Medicine
Item ID: 56070
Depositing User: Blore, Mrs Kathryn
Date Deposited: 04 Feb 2019 11:25
Last Modified: 07 Feb 2019 17:46
URI: https://eprints.nottingham.ac.uk/id/eprint/56070

Actions (Archive Staff Only)

Edit View Edit View