Many of us woke last Friday morning to news coverage of the Cancer Drugs Fund, with reporting describing the Fund as “a waste of money” based on a paper published in Annals of Oncology by Aggarwal et al.
The paper lays out a series of evaluations of the drugs provided by the Cancer Drugs Fund (CDF) in the years before its 2016 reform. The analyses represent a range of different perspectives, including the benefit experienced by patients, clinicians, and funding decision-makers. The first of these is an analysis of the results from the index trial for each drug – the key trial that provided evidence for the drug’s efficacy in each indication. It is this analysis that provides the basis for statements in the media that the average benefit was an extra 3.2 months of survival. Specifically, the analysis showed that of the studies reporting a statistically significant overall survival benefit (of which there were 18 out of the 47 drugs assessed), the median overall survival benefit was 3.2 months, ranging from 1.4 to 15.7 months.
What may not have come across in the media coverage, is these are not new data. This is not a report of actual, real-world outcomes from patients who accessed treatments through the Cancer Drugs Fund. As the authors point out, we didn’t do a good job of collecting such data, at least in the early stages of the CDF’s life. CASMI’s own observations are similar; we have been tracking the re-appraisal of drugs from the Cancer Drugs Fund now going through NICE, and have seen little evidence of real-world data from usage in the CDF being used to update the evidence base (CASMI unpublished observations).
This paper therefore doesn’t claim to have established the benefits of these drugs in’ normal’ use. However, the authors reference published studies that suggest less benefit is likely to be realised in the ‘real world’ than in the trials, notably when trial patients are younger and fitter than the clinical population. A glance at the paper’s table of CDF trials shows, for example, that the vast majority of patients in these trials were relatively well (‘performance status 0 or 1’, on a 0-5 scale where 0 is fully active, and 5 is dead) – if that’s not the case for the clinic population, we might expect they would do less well.
That isn’t to say the CDF drugs provided no benefits to anyone. The result reported from a clinical trial is an average (typically the median) so some patients will always do better than that; as several of the cancer charities reminded us in their responses to this paper, there are examples of patients who have done well on these treatments. And of course some will do worse. But until we can tell which patients are going to do well, the median or mean represent the best we know about the expected outcome.
The conclusions of the study perhaps shouldn’t surprise us – over half of the drugs in the CDF had after all been rejected by NICE, who found the clinical benefit insufficient to justify the cost. That’s why these drugs were in the CDF in the first place. Interestingly, though, 11 of these NICE-rejected drugs are now being reappraised by NICE, and in 8 cases so far, NICE has reversed its previous decision and recommended the drugs for funding. These decisions are based on largely the same data as the original decision (CASMI unpublished observations). However, the committee materials in the public domain are extensively redacted as commercial in confidence, suggesting a degree of pricing flexibility by the manufacturers in order to keep these drugs in use. We await the outcome of one of the more challenging of these decisions – trastuzumab emtansine (Kadcyla) in breast cancer – with a Technology Appraisal Committee meeting in early May.
-Dr Liz Morrell