Professor Adrian Newland answers questions about the evolution of the cancer drug prioritisation process, its current status, and the challenges that lie ahead
When did the London Cancer New Drugs Group come into operation, what were its aims, and how did you become involved?
In 2002, the tyrosine kinase inhibitor (TKI) imatinib gained its licence for the treatment of chronic myeloid leukaemia (CML). As a trial drug, it was supplied for free, but Novartis planned to start charging for imatinib 3 months after approval. At the time, several hospitals in London, particularly Hammersmith Hospital, had many patients with CML taking imatinib, so a group was put together from five London Cancer Networks to decide what should happen when the NHS had to start paying for the drug.
The London Strategic Health Authority built on this collaboration to develop the London Cancer New Drugs Group (LCNDG). The initial aim of the LCNDG was to review all new cancer drugs in order to foresee and forestall any similar problems in the future, and it was the development of this group that was the start of drug prioritisation. The LCNDG had the support from a variety of agencies, particularly the London Medicines Information Services based at Guy’s Hospital. I was asked to chair the LCNDG, because, at the time, I was involved with the then National Institute for Clinical Excellence (NICE) as a member of the panel that reviewed new cancer drugs and so already had some experience with drug prioritisation, albeit in a low-key and advisory capacity. From inception, the LCNDG reviewed upcoming drugs that were identified by pharmacists as of interest and drugs that clinicians were starting to use in practice to make recommendations to primary care trusts on the clinical effectiveness of the drugs under consideration, whether they should be approved for use in London, and whether the commissioners would provide local funding.
Bearing in mind that NICE existed at the time, why was London doing its own reviews?
An opinion from NICE is very important, as local commissioning groups have to fund drugs approved through the NICE technology appraisal process within 3 months. However, it took NICE up to 3 years after the licensing of a drug to provide a recommendation, creating a significant delay for oncologists who wanted access to new drugs, which were often widely available internationally. This was, of course, not just a London problem, but our remit only covered the London area.
The members of the LCNDG aimed to make a recommendation within 6 months. We held meetings every 2 months, and once a review was commissioned, a draft recommendation was developed and sent out for comments from the committee members and other interest groups. The comments were considered, and a final recommendation was made and circulated to commissioners in London. We did not undertake the detailed health economic analysis that NICE would; instead, our recommendations were based on clinical effectiveness, including overall survival and progression-free survival (PFS), toxicity, and, from 2010 onwards, quality of life, as data on this were not widely available in the early days.
The LCNDG report had no weight, except that we had a lot of discussions with commissioners, and the vast majority of clinicians were prepared to accept our recommendation that a drug was useful and should be funded or that it was not and should not be funded. Our recommendations were published a minimum of 14 months before NICE had opined, so our patients and clinicians were getting access to new drugs significantly earlier than they would through the NICE process. When we retrospectively compared our recommendations with those from NICE, we had 96% congruence: they rejected one drug that we accepted, and, bizarrely, we rejected one that they accepted.
All of our papers, with detailed analyses and recommendations, were available through our website. Some networks, including the North West Cancer Network based in Manchester, adopted our methods, while many followed our work but adopted our recommendations rather than developing their own. Ultimately, our work was used by 50% of England cancer networks.
As a refinement of the process, we started to provide a prioritisation list of drugs for consideration over the year. Any clinician could put forward a drug that they felt worth introducing, and the LCNDG would then review the evidence supporting the drug at a yearly meeting involving oncologists from all specialties. The oncologist would have to make the case in front of their peers, and the discussions were open and transparent. The process was widely accepted by the clinicians, and this approach resulted in a list of drugs that should be funded, a list of orphan drugs that should also be supported, a list of drugs that should not be funded, and a grey list of drugs that could be funded if the money was available. By building on this existing analytical approach, it was possible to develop the CDF very quickly when it was introduced.
Why was the Cancer Drugs Fund created and what were its original objectives?
Around 2010–2011, the government was put under pressure to reduce delays in patients in the UK getting access to drugs that were available in the US and Europe. The Cancer Drugs Fund (CDF) was set up as a mechanism to give patients access to these drugs while waiting for NICE (by then renamed as the National Institute for Health and Care Excellence) to publish its recommendations (see Box 1). The CDF was announced in summer 2010 and after a 3-month period of development kicked off in the autumn for a 6-month pilot period. The actual CDF started in spring 2011 using methods developed in the interim period.
The methods used for the national CDF were based on the analytical methods that had been used by the LCNDG over the previous 3 years, with important input from the North West Cancer Network, which had developed similar processes. This group was chaired by Professor Peter Clark, who subsequently went on to chair the national CDF group.
Box 1: Original aims and objectives of the Cancer Drugs Fund1
- The Cancer Drugs Fund was created in 2011 in order to ‘enable patients to access the cancer drugs their doctors think will help them’
- It was intended to provide a means of improving patient access to cancer drugs prior to the anticipated reform of arrangements for branded drug pricing on expiry of the Pharmaceutical Price Regulation Scheme (PPRS) at the end of 2013
- Its objectives were to:
- provide maximum support to NHS patients
- put clinicians and cancer specialists at the heart of decision making, consistent with the Government’s wider policy of empowering health professionals and enabling them to use their professional judgment about what is right for patients
- act as an effective bridge to the Government’s aim of introducing a value-based pricing system for branded drugs in 2014
- It intended to cover treatments not routinely available on the NHS in four categories:2
- drugs that were yet to be appraised by NICE
- drugs that would not be considered by NICE due to small patient population for which they were licensed but that were not covered by specialised commissioning arrangements
- drugs that were not recommended by NICE, mainly on the grounds of cost effectiveness
- drugs that could not be appraised by NICE, as clinicians wish to use them outside their licensed indication to treat forms of cancer with a similar biology to that for which they are licensed
The government established the CDF in 2011 as a temporary solution to act as an effective bridge to a value-based pricing system for branded drugs that was planned to be introduced by 2014. Why has value-based pricing still not been introduced?
Value-based pricing, in principle, recognised that it was reasonable to pay more for a drug for a rare orphan condition than for a fourth ‘me-too’ drug. Value-based pricing was a good concept in theory, but it did not work in practice because of game playing among the pharmaceutical companies (I suspect some of which put prices up with the intention of reducing them later so that they then appeared magnanimous). Consequently, despite much discussion and some confrontation, the concept of value-based pricing died; the CDF continued extending from 2011 onwards, but when value-based pricing did not emerge as planned in 2014, the CDF carried on in its original format until it closed on 31 March 2016.
The CDF received a lot of criticism for exceeding the allocated budget year on year and for funding drugs with limited clinical benefits. Was this criticism justified? Why did the CDF keep exceeding its annual budget?
The CDF was more or less successful in bringing through new drugs, but the cost did escalate from £50 million for the interim CDF in 2010–2011 to £360 million a year in 2015–2016. Although some interest groups had raised concerns about potentially higher costs at the fund’s inception, this was not the primary problem.3 The main reason for exceeding the budget was that a mechanism to remove drugs from the list had not been built in, as the scheme had been expected to have a limited lifespan and drugs were therefore anticipated to transfer to general commissioning following the normal NICE process when the CDF closed.
The CDF also reviewed drugs early in their lifecycle when data on clinical effectiveness were immature and with little real-world experience. The expectation was that NICE would review the drug again later and that it would not continue to be funded if it did not live up to its early promise. However, towards the end of the CDF, both the number of cancer drugs assessed by NICE and the proportion they approved reduced dramatically, because if there was any uncertainty, NICE would not fund a drug safe in the knowledge that the CDF, which worked on looser criteria than NICE, would likely step in. This essentially meant that fewer cancer drugs transferred into general commissioning.
When the CDF committee first realised that there was likely to be a significant overspend, the basic criterion of minimum mean survival of 2 months for a drug to be considered was increased to 3 months, which meant that two-thirds of drugs would no longer be eligible; this change was not introduced, but it did act as a salutary warning of (in general) the small gains from the drugs we were looking at. In the final year of the CDF, a cost element was also introduced to the assessment. This was not a full health economic assessment, but reviewers considered the potential value of a drug given its cost and outcomes in terms of overall survival, PFS, toxicity, quality of life, and novelty. A drug would have a positive recommendation if it scored highly on all of these elements, but the committee was less willing to support a drug that was expensive and only produced a marginal increase in overall survival. The CDF had a cut-off cost, above which drugs would not be funded unless companies were willing to negotiate a patient access scheme with NHS England. We did try to minimise costs toward the end, but the very nature of the CDF meant that it was always going to steadily accumulate costs.
From the introduction of the CDF in 2011 up until July 2016, limited usable data (such as information on life expectancy, quality of life, and side-effects) were collected. Was criticism of the CDF due to this lack of usable data fair?
When the CDF was set up, it was generally felt that we would be able to look at overall survival data based on the Oxford Cancer Intelligence Unit, but it did not have sufficient funding, and so data on overall survival were never collected. In London, we attempted to get over this, as the LCNDG would only approve a drug for 3 months at a time, with subsequent approval for the next 3 months only if patients were confirmed to be receiving it and were still alive.
The data revealed some important information—for example, we found that 10% of patients never received a drug that we approved and 40–60% of courses of parenterally administered drugs were not completed because of significant side-effects. These findings can be explained, as most trials of drugs recruit young people with no other morbidities so that they are robust enough to remain in a trial for the duration and there is enthusiasm on the part of both patients and clinicians to complete the study. However, we know that most patients with cancer have different demographics, particularly those with solid tumours, who are older than 60 years, often have other co-morbidities, and are less able to tolerate toxicities. This does raise the question of whether a treatment that is toxic enough to stop patients taking it and only gives a mean improvement in survival of 3–4 months is offering any genuine survival advantage or just causing significant problems in their remaining weeks or months. Indeed, one concern about the CDF was that an oncologist with a patient who had failed three or four lines of therapy could just offer another new drug approved by the CDF rather than having an honest and realistic discussion with the patient about their long-term outcomes, even when the standard of care with supportive therapy might achieve similar survival but better quality of life. The criticism that we had limited usable data is valid up to a point, but we did feel we were being unduly criticised for something that was out of our control.
What is the Systemic Anti-Cancer Therapy (SACT) database and will its data resolve this problem? Are there any important questions that SACT data will not answer?
The SACT dataset is the national mandatory collection of systemic anti-cancer therapy activity from all NHS England chemotherapy providers. The current SACT database will collect the previously missing data for patients of all ages and co-morbidities, so NICE will be able to analyse drugs in the future knowing the proportion of patients who received a drug and its impact on mortality. The SACT database has already collected an enormous amount of data on thousands of regimens that can be analysed on a regional and hospital basis and could potentially be analysed on a consultant basis. Currently, regimens cannot be correlated with mortality, although this should be possible within the next year. Current developments in the SACT database, which are overseen by Professor David Dodwell from Leeds, are promising and should support NICE’s current work.
The CDF was reformed in 2016 to become a ‘managed access’ fund paying for new drugs for a set period before they are definitively approved or rejected by NICE. Do you think this reform has been successful?
The CDF reform (see Box 2, Figures 1 and 2)4,5 has involved much discussion and hard work by numerous stakeholders who all want it to be a success. One key development is that NICE should be reviewing all drugs as they are authorised and quickly decide whether they are clinically effective and should be funded through the reformed CDF to allow collection of further data; this is a positive move that essentially mirrors the purpose of the LCNDG in 2002. The new CDF model gives NICE more flexibility. If initial data are good enough to show clinical effectiveness, NICE can agree to funding. If the data indicate that a drug has promise but probably is not cost effective based on current evidence and price, NICE can discuss CDF-funded patient access schemes with the manufacturer, while additional data are collected for the SACT database from real-life patients rather than the highly selected populations in company-funded studies. After 2–3 years, the drug can be reassessed to see whether survival data confirm its early promise and, if not, it can be defunded, which was not an option with the original CDF. At the moment, I would cautiously say that signs for the reformed CDF are encouraging.
Box 2: Aims and objectives of the reformed Cancer Drugs Fund from 20164,5
- New CDF aims to:4
- provide patients with faster access to the most promising new cancer treatments
- drive stronger value for money for taxpayers in drug expenditure
- offer those pharmaceutical companies that are willing to price their products responsibly, a new fast-track route to NHS funding for the best and most promising drugs via an accelerated NICE appraisal process and a new CDF managed access scheme
- The current CDF holds a national list of drugs maintained by NHS England:5
- the fund for 2017–18 is worth £340 million
- NICE assesses all new systemic cancer treatments within 90 days of authorisation:5
- if NICE recommends ‘yes’, CDF is used to make the drug available immediately rather than waiting 90 days for implementation; after 90 days, the drug is funded from NHS England commissioning budget rather than CDF
- if NICE recommends ‘no’, treatment can still be accessed through individual funding requests on a case-by-case basis
- NICE may recommend a drug ‘for use within the CDF’, where there is plausible potential for routine commissioning but still clinical uncertainty; a period of managed access (usually no longer than 2 years) is implemented, during which the drug is funded by the CDF to allow collection of more data on its clinical effectiveness, after which NICE reappraises the treatment
- CDF may also be used to fund off-label cancer treatment, which is determined by NHS England as part of its annual specialised commissioning process, as NICE cannot appraise drugs being used outside of their marketing authorisation.5
Figure 2: How drugs are added to the Cancer Drugs Fund7
What do you see as being the challenges in cancer drug prioritisation over the next 3–5 years? What challenges do you foresee related specifically to specialised commissioning?
Specialised commissioning faces numerous challenges. The changing demographic towards an ageing population will require a shift in focus from trying multiple lines of chemotherapy to recognising the value of end-of-life care. Problems with the Liverpool Care Pathway, which was poorly handled in some areas, set end-of-life care back, which is a shame, as this is an aspect of care that really needs to be developed and supported.
The changing demographic towards an ageing population will require a shift in focus from trying multiple lines of chemotherapy to recognising the value of end-of-life care
Companion diagnostics will be increasingly important with the development of drugs that are suitable for a small proportion with a specific genetic marker. For example, the drug crizotinib is only suitable for the 3% of patients with lung cancer who have an overactive version of the enzyme anaplastic lymphoma kinase, who respond very well to crizotinib but not at all to other chemotherapies. Drug combinations are another topic of increasing interest, as the 100,000 Genomes Project has shown that patients with malignancies often have several genetic abnormalities at presentation. If a patient is given a drug that targets only the predominant abnormality, the others, even if they are present in a very small percentage of cells, are likely to continue to grow and may cause relapse down the line. As the best chance of curing cancer is first time around, combinations of drugs that hit several of those molecular abnormalities may be needed from presentation.
Examples with a single molecular abnormality are few, but the potential in these rare conditions is exemplified in CML. The BCR-ABL chromosome rearrangement is generally the only abnormality in patients with CML; 15 years ago, mean survival was 3 years beyond diagnosis at most, yet suppression of this abnormality with imatinib or nilotinib (another TKI) has produced survival rates approaching 90% at 10–12 years, and some patients are even able to stop treatment and may be cured.
This reiterates that targeted treatments should be given early, although this would, of course, have enormous financial implications. It also raises questions from other clinicians in other specialties about why money is being spent on cancer but not on non-malignant conditions such as multiple sclerosis and dementia.
Do you have any concluding comments?
Survival in patients with cancer has improved progressively since the introduction of the Cancer Strategy by Professor Sir Michael Richards in 2000.6 This has been in small part due to better access to the newer cancer drugs following the schemes introduced as part of this strategy, but it is necessary to accept that surgery and radiotherapy are responsible for a larger number of cures. However, it is important to realise that the main gains reflect improvements in the whole cancer pathway. Early diagnosis and support therapy are crucial, as are the development of good-quality end-of-life care and support for the increasing number of people living with and beyond cancer. At a very basic level, early diagnosis remains poor in many parts of the country. The development of screening programmes and patient education remain key to early diagnosis, and the emphasis on prevention needs support.
Increasing understanding of molecular mechanisms involved in the malignant process is leading to an increasing array of newer agents, many of which are active across a range of tumours. This has also increased the importance of companion diagnostics to direct treatment to the most appropriate patients and led to the realisation of personalised treatment. To properly assess these newer, and very expensive, treatments, the new processes introduced by NICE will need support, data collection, and careful evaluation, so that we can access drugs that are both clinically and cost effective. The lessons learnt from the CDF have been absorbed and are being used to great effect, taking drug assessment forward. To justify the continued focus on spending on cancer drugs by specialist commissioning, the new processes will be important to defend expenditure on cancer drugs against other competing and equally compelling demands.
Professor Newland was previously a member of the National Chemotherapy Clinical Reference Group (2014–2016), chair of the NICE Diagnostics Advisory Committee (2010–2017), member of the National Cancer Drugs Fund (2010–2016), former chair of the London Cancer Drugs Fund (2010–2016), former president of the Royal College of Pathologists (2005–2008), former chair of the London Cancer New Drugs Group (2002–2016), and clinical director of the North East London Clinical Network (2001–2013).
- Department of Health. The Cancer Drugs Fund. Guidance to support operation of the Cancer Drugs Fund in 2011–12. Leeds: DH, 2011.
- Department of Health. Improving outcomes: a strategy for cancer. London: DH, 2011.
- Cancer Drugs Fund will not be enough, say cancer pharmacists. Pharm J 2010; 285: 395. Available at: www.pharmaceutical-journal.com/news-and-analysis/cancer-drugs-fund-will-not-be-enough-say-cancer-pharmacists/11028440.article (accessed 11 January 2018).
- NHS England. Appraisal and funding of cancer drugs from July 2016 (including the new Cancer Drugs Fund): a new deal for patients, taxpayers and industry. London: NHS England Cancer Drugs Fund Team, 2016.
- Bate A. NHS commissioning of specialised services. London: House of Commons Library, 2017.
- Department of Health. The NHS cancer plan: a plan for investment, a plan for reform. London: Department of Health, 2000.]
- NICE website. Cancer Drugs Fund. Committee decision-making: recommendations for use within the CDF. Available at: www.nice.org.uk/about/what-we-do/our-programmes/nice-guidance/nice-technology-appraisal-guidance/cancer-drugs-fund (accessed 24 January 2018).