Something has to change. We cannot continue tolarate the high cost of new cancer drugs up to, and in excess of, $100,000 (£80,000). This is unsustainable and we need to be thinking seriously about how we can lower the prices.

What can be done to lower the cost while enabling patients receive the best treatment?

One in two people in the UK are expected to get cancer in their lives and the statistics are similarly shocking in many other Western nations – better treatments are clearly urgently needed. And better treatments are coming. For example, the innovative new targeted drugs and immunotherapies deliver significant benefit to cancer patients.

But we need to be able to afford them and the spiralling costs of these new drugs are threatening to make cancer treatment unsustainable. As one patient advocate recently put it, innovation is meaningless if nobody can afford it.

Even in 2012, 12 of 13 newly-approved cancer drugs were marketed at over $100,000 a year. These high prices are particularly concerning when clinicians often have to combine several of these drugs for the best results. For example, the hugely promising combined immunotherapy treatment of nivolumab and ipilimumab is priced at around $250,000. This exceeds the median cost of a home in the US, which is $240,000. In the UK, this game-changing treatment has recently been rejected for use on the National Health Service (NHS) for people with advanced head and neck cancer because of cost. This cannot go on.

But, often perhaps with good reason, we have come to accept high prices for cancer drugs. Many drugs fail to make it through. Research and development costs are high. And large clinical trials are very expensive. These factors are commonly cited as major contributors to cost. However, as we move further into an era of personalised medicine, do the arguments behind such high prices still hold up?

We can often select patients genetically when trialling targeted cancer drugs with great success, which means we can avoid the costly phase III trials with thousands of patients. In 2016 a study of only 50 patients was needed by the FDA to approve crizotinib for lung cancer patients with ROS1 mutations.

As we see FDA approvals expand the patient population across multiple cancer types, we would expect the price of the treatment to decrease but this rarely happens.

Recently, with colleagues from the MD Anderson Cancer Center and the Netherlands Cancer Institute, we considered the problem of unsustainable cancer drug prices in the journal Cell (http://www.cell.com/cell/abstract/S0092-8674(17)30124-1) and suggested a fundamentally different approach to drug discovery and development, which would keep lifesaving cancer treatments affordable.

We considered new models of discovery drug development, some low-hanging fruit to harvest as soon as possible, and ways in which the pharmaceutical industry could help itself to bring prices down.

Pharma companies, as commercial enterprises, could be more efficient – this is one element that contributes significantly to the high cost of cancer drugs. For example, there is massive unnecessary duplication with multiple companies working on the same small group of molecular targets. Many pharma companies perform very similar trials with comparable drugs but do not share their data. There are over 800 clinical trials testing T cell checkpoint immunotherapies, which together aim to enrol over 166,000 patients.

Imagine if the companies behind these studies collaborated more – far fewer studies would be needed. Key results would be generated quicker. Unfortunately it is people with cancer who will pay the price for this sort of inefficiency.

Another area for further improvement is the use of biomarkers for patient selection. Although there has been a welcome increase in the use of tests to dictate which patients should be included in a clinical trial – leading to stratified or personalised medicine – there is still room for improvement. For example, the new immune checkpoint inhibitors were initially approved without biomarkers to determine who would respond. Use of biomarkers should bring down costs by reducing the size and length of clinical trials.

Both scientific and organisational inefficiencies like this appear to be contributing to the escalating costs of drug development – and certainly could be reduced to help bring prices down.

Of course academic organisations are not perfect either. But many of the discoveries in basic science that have led to new categories of cancer drugs have originated from academia. And there is certainly room for academia to work more closely and efficiently with industry.

In fact, I believe an increasing proportion of innovative drug discovery can be driven forward by academic institutions, as is done at the ICR. For example, we have been successful in discovering the CYP17 inhibitor abiraterone for advanced prostate cancer and in identifying a biomarker strategy for the PARP inhibitor olaparib in people with ovarian cancer with BRCA mutations.

In academic institutions, the traditional model of self-supporting researchers driving independent programmes is increasingly evolving into a ‘team science’ approach with expert multidisciplinary teams which have the budgets needed and the required expertise in their staff to power drug discovery. The Cancer Research UK Cancer Therapeutics Unit at the ICR and the Institute for Applied Cancer Science at MD Anderson are two good examples of this model. It is important that in the new models we create we have sufficient expertise, experience and resources to ensure that drugs are progressed both intelligently and rapidly so that cancer patients can receive drugs as quickly as possible, as well as at affordable prices.

Academic drug discovery has the advantage of being free and incentivised to take on risky challenges that pharma could not. For example, we know that a large number of cancer genes are yet to be drugged – we need to complete the job of drugging the cancer genome and it is no small task. And there are types of genes that are not the classical targets – such as RAS, MYC and p53. It is very high risk to go after these targets where there is very little knowledge of the biology and we may not have the technology to drug such challenging targets yet.

Academia can make a huge impact here. We have shown we can identify and validate targets (although there is room for improvement in reproducibility and robustness); develop linked biomarkers to improve patient selection; show the target can be drugged; demonstrate proof of concept in animal models, for example with chemical probes; take a candidate drug through preclinical development; and conduct early stage clinical studies to show that a treatment works and can be tolerated in cancer patients.

The reach of non-profit drug discovery and development is expanding. Hospitals with deep oncology expertise that are part of, or linked to, academic cancer research institutes are often the location of all three phases of clinical trials. Some academic pharmacies are certified to develop and produce drug formulations that are suitable for clinical use. And academic clinicians have proven themselves to be highly capable of running large investigator-initiated, as well as pharma-sponsored, clinical trials.

Academia therefore has the skills and the tools to discover drugs and take them through clinical trials. But academic drug development is commonly stalled at the clinical testing stage and this is for several reasons. Most academic groups would not have the stringent quality control required to manufacture clinical grade drugs at large-scale. They also do not usually have the funds needed for the expensive toxicology testing required by regulators. And, even if academia could surmount these two hurdles, they clearly would not be equipped to handle the marketing and sales of the drugs, nor would that be appropriate.

Many academic researchers want their discoveries to benefit patients and so these obstacles drive them towards working with industry, which can work well. But companies of course want to ensure the maximum return for their investment at this stage and academics lose any influence over pricing if they want to see their discoveries reach the clinic.

One solution to the current challenges is to form new companies that are partnerships between academia and industry. And it may even be possible for some generic drug producers to work with academia in this way. These firms bring generic versions of drugs to market at greatly reduced prices and are therefore used to working with lower profit margins, and they have the infrastructure in place to manufacture at large scale, as well as the marketing and sales capabilities. But many other models are also possible.

By industry and academia working together as small specialised companies on targeted drugs, with associated biomarkers of response, which would require only require small numbers of patients on trials to satisfy regulators then if savings are passed on, drug prices could be far lower than what we currently see.

But this model would only succeed if the academic organisations can resist selling their discoveries to the highest bidder and stay true to their societal responsibilities. To that end these organisations would need to agree a price ceiling, and a pricing strategy that ideally also ensures accessibility in lower- and middle-income nations, this would need to be done during any negotiations with partners and investors. The academic partners would receive royalties on sales of the drugs, which would help sustain them and fund the discovery and development of new drugs. Patients would gain access to much needed drugs without bankrupting themselves or health care providers.

Currently there are not many academic drug discovery centres in the world that are ready to take on such a partnership so this clearly is not an overnight solution to the global crisis in cancer drug pricing. But over time academic centres could begin to move towards this model and create the kind of competition needed to gradually drive down the prices that result from the conventional model.

In the short-term, academia, working in these new partnerships, can help to bring prices down by attempting to harvest the low-hanging fruit that are not profitable enough for industry to consider. Researchers can take patent-expired drugs and, using biomarkers that suggest efficacy, and explore new indications. Academia can also seek out effective drug combinations using drug candidates that were abandoned because they were not a success on their own.

If we are to end this era of $100,000 cancer drugs, we are going to have to make some big changes to the whole way drugs are discovered and developed. The proposals I and my colleagues have outlined are disruptive but there is a clear and urgent necessity to lower cancer drug prices to keep lifesaving drugs available and affordable for patients. It just will not be possible for healthcare systems like the NHS to afford the newest and best cancer drugs if prices continue to rise as rapidly as they are now.