Using tumour, peripheral blood and sentinel nodal transcriptomics to understand the interaction between melanomas and the host

UK melanoma incidence has increased more rapidly since the 1970's than any other cancer and in 2011 there were 13,348 affected patients. Although the increased incidence in recent years has been greater in people over the age of 60, melanoma occurs in a significant proportion in very young adults. Although the majority of melanoma patients, thankfully are cured by their surgery, until very recently indeed, the treatment for melanoma which has spread to other organs has been terrible. The progress in the last 6 years has been that even advanced melanoma has been successfully treated (apparently for the long term) with treatments that work by activating immune responses in the patient: effectively encouraging the patient's own immune system to reject their cancer. Although these long term responses have been ground breaking much needs to be done to capitalise on the progress. There is a need , to predict who will respond, to find drugs for the non-responders and especially to identify patients early in their disease who require, and will benefit from immunotherapy to prevent progression (so-called adjuvant treatment). In order to meet these needs, a better understanding of what determines how patients' immune systems interact with the tumour cells is needed.
Existing research has shown that in some patients, their immune cells invade the tumour and may actually destroy it. However, it is clear that the tumour cells often then sense that attack and are able to turn off those immune responses. We have preliminary evidence that a number of factors can modulate this process: that a harmful sort of inflammation in the whole body and in the tumour may promote the progression of the tumour and appears to be associated with a suppression of immune responses. Many inflamed tumours have ulceration of the skin surface seen by the pathologist down the microscope, which is associated with poorer survival. This ulceration is more common in smokers and less common in people with higher levels of vitamin D in their blood. In this application we seek funding to better understand ulceration, the role of inflammation, smoking and vitamin D in survival and the interaction between tumour cells and the immune system. The project is designed to do this using a form of genetics, which is the expertise of the group. We are recruiting to a study in Leeds, in which patients give blood samples and allow us to take samples from their tumours and their lymph glands stored in NHS laboratories. We will use laboratory techniques to look at how the genes in those tumours, blood samples and lymph glands are expressed. We will then use computer based tests (bioinformatics) to estimate the activity of different parts of the immune system in those data. In so doing we will be able to study how inflammation, ulceration, smoking, vitamin D, and different tumour types associate with different patterns of immune activity.
If we can thereby identify what allows some patients to mount an immune response, how the cancers turn off that immune response and what factors affect these processes, then we argue that we will be able to both develop low toxicity treatments which can be used in early disease and to identify patients who will benefit from those treatments. We will understand if there is evidence that harmful inflammation such as that associated with smoking and obesity suppresses immune responses to cancer and reduces survival and the effects of vitamin D. We will therefore be able to better advise patients at diagnosis about factors which may be important to their prognosis and if inflammation does play a role in suppressing immune responses to melanoma then we will explore common drug therapies such as aspirin and other non-steroidal anti-inflammatory drugs as adjuvant therapies.
In the final period of the project we will use what we have learned to devise tests (predictive biomarkers) to target drug treatments to those who will benefit most.

We will address the objectives described using data from 2 studies
1. We will recruit 300 newly diagnosed melanoma patients to the Ulceration and Immunity Study (164 to date), in which patients are bled at presentation and 4 intervals in the first year.
a)The participants allow sampling of the stored paraffin embedded primary tumours and sentinel nodes, and the generation of transcriptomic data from those tissues using the Affymetrix HTA.
b) Vitamin D levels are measured at each time point and patients supplemented if indicated by Leeds clinical protocols.
c) Peripheral blood white cell RNA is fixed at the bedside using Leukolock filters for gene expression analysis.
d) The transcriptomes of melanoma primaries, blood and nodes will be analysed agnostically using pathway analysis (Metacore). Using bioinformatics we will derive data sub-sets which are indicative of the presence of different components of the immune system (the immunome), as pioneered by INSERM U872, Laboratory of Integrative Cancer Immunology, in order to investigate the interaction between tumour and immune cells.
e) Previously we have generated primary tumour transcriptomes from 222 melanoma primaries from the Leeds Melanoma Cohort in which application of the bioinformatic approach developed by the INSERM group resulted in the identification of clusters which were strong independent predictors of survival. We will test these clusters in the new samples from the Ulceration and Immunity Study and determine the associations between these tumour sub-groups and the transcriptomic derived "immunomes', in order to understand whether the 'clusters" have survival significance because of the immunological processes evidenced in blood and nodes.
2. We will recruit a second set of patients commencing systemic therapy for stage 4 disease and seek consent to use blood, tumour and nodal samples in which the signatures derived from the Ulceration and Immunity Study can be tested as predictive biomarkers

1. Melanoma patients
(a) Given the incidence of melanoma under the age of 35, many potential years of life are lost to melanoma. This project will inform the development of adjuvant therapies with the hope of reducing recurrence and melanoma deaths. Preliminary data suggest that 'smouldering' inflammation may play a role in melanoma mortality. This chronic inflammation is associated with smoking and reduced by non-steroidal anti-inflammatories. This suggests potential adjuvant therapies that are low-cost and relatively non-toxic (such as Cox 2 inhibitors) for at least a proportion of patients.
(b) The identification of prognostic biomarkers and predictive biomarkers for immunotherapies will bring real life benefits to patients (better outcome and less toxicity).

2. Economic/commercial entities
We plan to develop a robust blood test to assess suppression of immune responses in cancer patients. We are exploring the commercial potential of such a test with the Leeds University Stratified Medicine Hub. Engaging suitable economic partners for the development of such a test would clearly be of economic benefit to the nation.

3. The NHS
The identification of effective predictive biomarkers will have economic benefits for the NHS in terms of targeting expensive therapies at those most likely to benefit and reducing unnecessary toxicity.

4. Cancer researchers/clinical trials
By fixing RNA at venesection we will enable routine collection of samples for immune assays in busy clinics for the first time. This approach may be valuable to researchers trying to develop similar predictive biomarkers in other cancers and indeed in other medical conditions where immunosuppression is important, such as autoimmune disease or organ transplants.

5. Bioinformatics research
Whilst the potential for transcriptomics in many areas of health is recognised, the science of bioinformatics remains poorly developed. This project will provide insight into this development in conjunction with the Leeds MRC Clinical Bioinformatics Centre.

6. People at risk of melanoma/general public
(a) It may be that even if smouldering inflammation does play a role in suppressing immune responses to melanoma, that the 'dye is cast' at the time of diagnosis; reversing inflammation after diagnosis may be too late to promote the development of immune responses without potent immunomodulatory drugs such as Ipilimumab. Prophylactic reversal of inflammation may therefore be vital in those at increased risk of melanoma such as those with the atypical mole syndrome (2% of the population) or a strong family history of melanoma and therefore important to the nation's health.
(b) We have good evidence that smoking at diagnosis increases the likelihood of death and this information could help some people to stop smoking to their benefit whether cancer patients or not.

7. Patients at risk of, or suffering from, other forms of cancer
We contend that host immune responses play a protective role in melanoma and that studying this cancer will inform the management of other cancers.