Intermittent drug strategies: the new era of cancer treatment

Dr Rosalie Fisher, Clinical research fellow in the Renal and Melanoma Unit at The Royal Marsden NHS Foundation Trust

In 2001, the Novartis drug imatinib revolutionised the treatment of chronic myeloid leukaemia, and so began a highly exciting era in cancer treatment: the era of targeted drug therapies. These differ from traditional chemotherapy drugs in their ability to knock down a specific biological pathway in a cancer cell that is essential for its malignant behaviour.

Targeted drugs are now used in the clinic for almost every type of adult cancer, and there is no doubt that patients have gained hugely from this progress. Not only are those with advanced cancer living longer with their disease and enjoying greater relief of their symptoms, but in some tumour types such as breast cancer, more patients are cured from the addition of a targeted agent to chemotherapy treatment. However, targeted drugs and chemotherapy share a critical shortcoming: drug resistance. Resistance to anti-cancer agents is readily manifest across the spectrum of cancer patients - in those who relapse after ‘curative’ drug therapy and in those with advanced disease - by growth of tumours and death.

There are many mechanisms by which a cancer cell evades drug therapy, but what is the over-arching, driving force in the development of drug resistance? Many in the oncology community subscribe to the ‘clonal evolution’ theory of tumour development and growth, a model based on two Darwinian principles: (1) tumours are composed of ‘clones’, groups of cells that share a common ancestor, and (2) some clones are ‘selected’ to populate the tumour over others because of their favourable characteristics. A further requirement is that the genetic material of the cancer cells is unstable, allowing clones to acquire advantageous mutations in their DNA. Recent evidence extends this model by suggesting that evolution of tumours occurs in a complicated branching, rather than step-wise, pattern. A study of kidney cancers published in the New England Journal of Medicine in 2012 reveals that the clonal make-up of different tumours within an individual patient varies markedly, and that even different regions within the same tumour carry distinct genetic profiles. Just as worrying is the evidence from lung and blood cancers that gene mutations present at a seemingly trivial level within a tumour at the start of therapy can become dominant as therapy progresses, and can have a devastating impact on what happens to the patient. Gradually therefore, we are building a picture of cancerous tumours as ecosystems or complex societies, whose populations (the cancer cell clones) are constantly shifting in space and time, in response to outside influences.

If only, as cancer researchers, our neurons would evolve so rapidly and flexibly as the cancer cells whose biology we are attempting to decipher! Perhaps then it would not have taken more than a decade to realise that we cannot give continuous drug therapy to patients with advanced cancer, whether we use targeted agents or not, with the expectation of a long period of control, let alone cure. If cancer cells adapt to therapy, it follows that therapy must also adapt.

What is the evidence that an adaptive drug therapy approach might be successful? Advanced melanoma, a highly aggressive skin cancer, provides a topical example. In 2010, decades of pessimism in this disease ended with the introduction of the BRAF inhibitor vemurafenib, which was shown to drastically improve the life expectancy of some patients with advanced melanoma. Quickly, vemurafenib treatment for BRAF-mutated melanoma was compared with chemotherapy treatment for lymphoma, as large, solid tumours in critical bodily organs melted away in a matter of weeks. With a little more experience of vemurafenib, however, the analogy was changed to that of small cell lung cancer, known for its swift response to chemotherapy, followed by rapid re-growth of tumours, and shortly after, patient death. The observation that vemurafenib controls melanoma for an average time of only six to 12 months led to a further concern – that growth of melanoma tumours may accelerate after vemurafenib is stopped, likened to a ‘releasing of the brakes’.  A recent study by our group suggests that this is not the case.  We examined the scans from 19 patients who were treated with vemurafenib, but who became resistant to it. CT scans and tumour sizes were compared during vemurafenib treatment and after it was stopped, showing that tumour growth rates actually slowed after the drug was stopped. Although surprising, these results are supported by laboratory evidence published recently in the journal Nature, indicating that melanoma tumours with the abnormal BRAF protein that are resistant to vemurafenib are also dependent on it for their continued growth. Using a mouse model, this group showed that using the vemurafenib drug in a stop-start fashion delayed the time to drug resistance. Combined work from the two groups will be presented orally at the American Association of Cancer Research meeting in April.

A similar discontinuous drug therapy strategy is being tested on a larger scale in the ‘STAR’ trial. Patients with advanced kidney cancer will be treated with the targeted drug sunitinib and are randomly assigned to one of two treatment groups. The control group will receive the standard therapy - sunitinib treatment until there is evidence that it has stopped working - while patients in the experimental group will stop the drug even though it continues to work, and will have a break off treatment until the cancer begins to grow again.

An adaptive therapy approach using targeted agents is novel; Darwin’s theory underlying it is 150 years old. We can only attempt to control the evolution of tumour cells by controlling the selection pressures to which they are subjected. An intermittent drug dosing strategy is one way to achieve this. We should acknowledge that it is not a treatment strategy that oncologists will be entirely comfortable with – our thinking has been heavily influenced by practice around antibiotic prescribing and resistance, and we have clung to the ‘if it’s working, don’t change it’ dogma. Arguably, as patients continue to die, cancer treatment isn’t working. Our doctrine must adapt – and so begins a new era of cancer treatment?