Permeabilisation of brain metastases for early detection and treatment

The spread of cancer from a primary tumour to the brain remains one of the greatest challenges in cancer treatment and life expectancy from diagnosis is typically of the order of 3-9 months. Blood vessels in the brain are much more impermeable than those elsewhere in the body and possess what is known as the blood-brain barrier, which prevents the movement of both drugs and diagnostic "dyes" from the blood into the brain. This blood-brain barrier is one of the main reasons that earlier diagnosis, when treatment might be more effective, is not possible. We are interested, therefore, in finding ways to make the blood vessels that are close to these tumours leaky, when the tumours are still very small, to enable both earlier diagnosis and more effective treatment. We have recently shown in experimental models that by injecting a protein called TNF into the blood stream we can make the vessels close to secondary tumours in the brain leaky, whilst leaving the rest of the vessels in the brain intact. This approach works because the blood vessels associated with these secondary tumours are different to normal brain blood vessels and express a protein receptor (TNFR1), which TNF can interact with and cause the vessels to become leaky. We have shown that using this approach we can selectively deliver both imaging dyes for early diagnosis and common cancer drugs to very small secondary brain tumours when they are still very small. TNF can be toxic at high doses, but the amounts we are using are much lower than this and, therefore, would be tolerable. Nevertheless, we believe that if we can develop a more selective protein that interacts only with TNFR1, and no other protein receptors, we can further reduce the amount needed and any possibility of toxic side-effects. We will test this new protein in a number of different ways to confirm its specificity for TNFR1 and its ability to cause vessel leakiness. The huge advantage of this approach over alternative solutions to the problem of the blood-brain barrier, is that once the vessels are leaky delivery of any diagnostic dye or drug to these tiny brain tumours is possible. The outcome of these studies will enable us to apply for funding for a clinical trial to take this approach forward to the clinic.

Secondary spread, or metastasis, to the brain is a major challenge in cancer therapy and prognosis is extremely poor. A significant barrier to both early detection and effective treatment is the blood-brain barrier (BBB), which is intact during the early stages of tumour development and only heterogeneously permeable at later stages. The presence of an intact BBB excludes both therapeutic and diagnostic imaging agents from metastases. However, we have shown in mouse models that intravenous injection of the cytokine tumour necrosis factor (TNF) selectively induces BBB permeabilisation at sites of micrometastasis in the brain, leaving the rest of the BBB intact. Using this approach we could both detect metastatic colonies using contrast-enhanced MRI and deliver a common cancer therapeutic, trastuzumab, to micrometastatic colonies. Our work indicates that the permeabilisation is mediated primarily though activation of the TNF type 1 receptor (TNFR1); this receptor was found selectively on vessels closely associated with brain metastases both in mouse models and in human brain metastasis tissue. The toxicity of high-dose TNF limits its use as a tumoricidal therapy. However, we have demonstrated that its permeabilising activity occurs at concentrations below the maximum tolerated dose. Moreover, given the TNFR1 specificity of the permeabilising effect, we now propose to develop a highly selective TNFR1 agonist, which will further reduce the dose required and eliminate toxicity associated with TNFR2 activation. We will assess binding specificity and permeabilising action in vitro and in vivo and undertake preliminary toxicology studies. At the end of this project, we will be in a position to undertake cGMP manufacture, full toxicology studies and a Phase I/IIa clinical trial. Our solution, to the problem of an intact BBB in early brain metastasis, will open a window of opportunity for both diagnosis and treatment that currently does not exist.

An estimated 20-40% of cancer patients develop brain metastases, with most originating from lung, breast, melanoma, colorectal and kidney. In the USA and UK, respectively, this accounts for ~170,000 and ~15,000 new cases each year, with an expected increase to >20,000 in the UK by 2020.

We anticipate screening at risk patients using the permeabilising strategy both at time of primary cancer diagnosis and periodically following treatment of the primary. This approach will enable brain metastases to be detected and treated early, when efficacy is greatest. Our approach would yield substantial gains in extending the clinical potential of systemic cancer drugs. It is well established that the greatest benefits are gained by early treatment of metastatic spread; our approach would make this possible for brain metastasis. In some cases, costs and morbidity of more radical radiotherapy or surgery may be avoided through the use of systemic therapies.

Additionally, early detection will confer the following advantages:
(i) confirm curative potential of radical therapy for primary when brain metastases absent;
(ii) enable decision towards palliative care, when brain metastases are untreatable, improving quality of life;
(iii) eliminate unnecessary, debilitating prophylactic treatments when brain metastases are absent.

As an example, a health economics assessment of our strategy in non-small cell lung cancer brain metastasis, based on increased efficacy of existing therapeutic approaches, yielded an estimated ICER of £6,684 per quality-adjusted life year (QALY) gained when used in both diagnosis and treatment. Indicating cost-effectiveness based on the £20,000 per QALY threshold used by NICE.

More broadly, the BBB is an obstacle to many high molecular weight compounds. The development of adjunct therapy to selective permeabilise the BBB would be enormously attractive for pharma across numerous brain pathologies associated with a regionally activated brain endothelium where upregulation of the TNFR1 receptor is likely. For example, the remyelination anti-LINGO therapy for MS (Biogen) is currently being administered at 100mg/kg once weekly, which equates to 7g of antibody. Not only is this prohibitively expensive, but it is also likely to generate unwanted side effects. Targeted delivery would enable this dosing regime to be substantially reduced.

As a further example, we have recently demonstrated that the vascular endothelium is activated at the edges of primary brain tumours; the most actively invasive area of the tumour, but where the BBB is intact. Consequently, access of therapeutics and diagnostic agents to these areas is limited. As a result surgical removal and/or radiotherapy is frequently sub-maximal and, critically, does not remove or eliminate the most actively invasive cells at the tumour margins leading to recurrence. Similarly, systemically administered therapy has little likelihood of success owing to incomplete tumour penetration. By facilitating access of diagnostic agents (both MRI and new fluorescent imaging agents that target tumour cells, surgical removal and/or radiotherapeutic targeting of all tumour tissue could be greatly improved. At the same time, delivery of systemic therapeutics throughout the tumour could be facilitated, opening new avenues of therapy.

Finally, temporary opening of the BBB has recently been shown to facilitate clearance of amyloid-beta from the brain and improve cognitive function in a mouse model of Alzheimer's disease (which is associated with an inflammatory response), suggesting further potential applications of our approach.