Identifying molecular drivers and early diagnostic biomarkers in breast cancer leptomeningeal metastasis by the interrogation of cerebrospinal fluid

Treatment for breast cancer has improved overall during the last 20 years, however secondary breast cancer (when the cancer spreads to other parts of the body) remains incurable and is leading cause of breast cancer deaths. Spread to the brain occurs in up to 50% of secondary breast cancers. A form of spread to the brain is called breast cancer leptomeningeal metastasis (BCLM), where cancer grows in the lining of the brain and spinal cord. BCLM is difficult to diagnose, as is not always visible on scans, plus requires a sample of spinal fluid (CSF) to be taken by a procedure called lumbar puncture. This test often needs to be repeated multiple times to make the diagnosis, by which time often the condition has worsened.

Doctors fear the development of BCLM as the outlook is poor. Without treatment, patients live an average of 4-6 weeks. In patients who are well enough at diagnosis, treatment can be tried, which involves giving chemotherapy drugs into the CSF through repeated lumbar punctures, or through a catheter device into the brain called an Ommaya reservoir. However the treatment does not work in all patients, and it can be difficult to know when stop, as there is no test which can monitor the effect. Even with this treatment, the average survival is around 4 months.

It is not currently known how breast cancer cells spread to and grow in the lining of the brain, given that there is a blood-brain barrier which stops larger substances in the blood stream such as bacteria and cells from entering the brain. It is likely that there is a specific biological mechanism in the seeded cancer cells which allow this to happen. It is now understood that the way cancer cells behave is programmed by genetic changes in cancer cell DNA, and finding out these changes has led to the development of new cancer treatments.

The overall goals of this project are to understand how BCLM develops and to improve the testing for the condition.

To carry out this project, the first step is to collect CSF samples from patients with BCLM. This will be achieved by finding patients who are having a lumbar puncture for the diagnosis of BCLM, or who have an Ommaya reservoir device inserted for treatment.

In the laboratory, I will extract the DNA from these CSF samples, and use sequencing technologies to read the DNA code and compare to the non-cancer DNA, to find cancer-specific changes. These DNA changes will be compared to those in the patient's original breast tumour (usually removed by surgery in the past) to test whether new changes have occurred in the BCLM. I will also compare with any other areas where the cancer has spread to, for example liver or lungs, when biopsies of these have been performed in the past. These comparison can tell us whether BCLM has additional DNA changes which may explain how it is able to cross the blood-brain barrier and grow. The second step is to prove that these DNA changes can cause or influence the development of BCLM. For this, I will use methods in routine use to grow laboratory cancer cells in mice and test different DNA changes in the cells and how they affect the growth and spread of the cancer.

The third step in this project is to test whether small circulating fragments of cancer DNA, which can be found in the CSF, could be used to give an early diagnosis of BCLM from a single lumbar puncture. This can be done in parallel to the first two steps. This will involve testing the CSF in all patients being investigated for suspected BCLM, and in patients who test negative using the current method, I will test whether DNA fragments could have diagnosed BCLM earlier.

At the moment no research group in the UK or worldwide focuses specifically on this area. This is surprising given that spread to the brain from breast cancer is becoming more common, and that it has such a poor outlook. If there is to be a reduction in deaths from breast cancer, research is desperately needed into this condition.

Breast cancer leptomeningeal metastasis (BCLM) is a form of central nervous system (CNS) metastasis with a dismal prognosis of around 4 months despite currently available treatment. Despite the increasing incidence of CNS metastasis, there lack of translational research focus in BCLM. In addition, diagnosis using current methods is challenging. Overall the project aims to improve scientific understanding and develop a diagnostic biomarker by using cerebrospinal fluid (CSF).

Methods
Clinical samples will be obtained from patients with BCLM and will comprise cerebrospinal fluid (CSF), blood and primary tumour tissue; all accessed through currently active studies through the Royal Marsden and South East London Cancer Research Network (SELCRN).

1. Genomic analysis
DNA will be extracted from CSF tumour cells and/or from the CSF supernatant. DNA will undergo whole exome sequencing, and will be compared, within the same individual, to germline, primary tumour and other metastatic site DNA, and also to existing breast cancer sequencing datasets to identify recurrent or unique somatic genomic aberrations in BCLM.

2. Functional genomics
Genomic aberrations will be investigated for functional consequences through the establishment of an in vivo model. Established human tumour cell lines will be manipulated by transfection via lentiviral expression constructs, and gene knockdown by RNA and CRISPR interference.

3. Biomarker discovery
Levels of circulating tumour DNA (ctDNA) in CSF will be determined by digital droplet (ddPCR) in patients being investigated for BCLM, in parallel with standard cytological analysis. Serial CSF samples will be analysed for ctDNA dynamics to correlate this with other parameters of disease status, including the clinical decision to stop ineffective treatments.

Through the scientific methods used and findings generated, this research will impact not only breast, but other cancer types, in addition to other neurological conditions.

1. Health outcomes
Breast cancer is second only to lung cancer as the leading cancer-related cause of death in females in the UK. Metastatic disease causes the majority of deaths from breast cancer, and CNS metastases are both increasing and have the poorest prognosis. Reducing deaths from CNS metastasis is not only a major challenge in breast cancer, but also in lung cancer and melanoma, both which have a high incidence of CNS involvement. Therefore, this research has the potential to impact a substantial number of patients across these common tumour types. Additionally, this research may benefit patients with non-cancer neurological diseases, through the development of research strategies to explore meningeal pathology, which may lead to improved understanding in conditions such as Alzheimer's disease and multiple sclerosis.

2. Biotechnology and Pharmaceutical Companies

a. Biological pathways - The overarching aim of this project is to identify the biological dependencies of BCLM. This would lay foundations for the development of novel therapies, through target-based development or repurposing of existing novel compounds. Pharmaceutical companies and drug development networks will therefore be expected beneficiaries of the research outputs. In addition, I will develop patient-derived primary cell cultures which can also be used for pre-clinical testing.

b. In vivo and PDX BCLM models - PDX models can accurately recapitulate the treatment responses in patients. They provide a pre-clinical drug testing platform which better ratifies the selection of compounds for progression to phase I/II clinical trials. Patient-derived BCLM xenograft models will be developed during my project, and these would be valuable in the development of new BCLM-directed therapies.

c. Diagnostic biomarker - The clinical validation of CSF circulating tumour DNA for diagnosis and monitoring of BCLM will be of direct interest to biotechnology companies who develop the appropriate platforms. For example, Bio-Rad, who manufacture the digital droplet PCR (ddPCR) system which I will use in the research laboratory, are in the process of commercialising a number of their ddPCR assays for clinical diagnostics.

3. Policy makers

There is a growing body of evidence to support the broad clinical applicability of ctDNA in cancer. The published results of this research in relation to CSF ctDNA would be reviewed, along with feasibility and economic implications, by policy-makers such The National Institute for Health and Care Excellence (NICE) in the United Kingdom, National Comprehensive Cancer Network (NCCN) in the United States, and European Society for Medical Oncology (ESMO) across Europe, during the development of clinical practice guidelines in metastatic breast cancer.

4. Public and third sector

The Medical Research Council released a refreshed strategic plan in 2014, with the primary aim of 'picking research that delivers', in terms of improvement of health outcomes. My research project would align well with this aim, in particular under the theme 'Molecular Datasets and Disease'. This theme focuses on the utilisation of high-throughput technologies to study human cohorts, explore molecular pathways and discover new targets for therapeutics.

In terms of charity engagement, I will be based in the Breast Cancer Now Research Centre. Breast Cancer Now is the UK's largest breast cancer charity, and their vision is that by 2050, every patient with breast cancer will live. This means that advances are needed specifically in the treatment and control of metastatic breast cancer. Given the increasing incidence of BCLM, and poor survival from this, clearly this needs to be tackled to meet this vision. By the public engagement strategies described earlier, the sharing of this research activity will promote ongoing fundraising for this charity.