When I use a word . . . Commercial clinical trials

A review

Earlier this year, the UK government appointed Lord James O’Shaughnessy, senior partner at a consultancy firm, Newmarket Strategy, a board member of Health Data Research UK, and a former Conservative health minister, to conduct an independent review into what was described as “the UK commercial clinical trials landscape.”1

The background to this review was that there had reportedly been a 44% fall in recruitment of patients to phase 3 commercial clinical trials between 2017 and 2021. This had had the effect, according to an analysis by the Association of the British Pharmaceutical Industry (ABPI), of dropping the UK down the global rankings for numbers of phase 3 industry sponsored clinical trials from fourth place in 2017, after the USA, Germany, and Spain, to 10th in 2021, now behind the USA, China, Spain, Japan, Germany, Canada, France, Poland, and Italy.2 This was despite the fact that during the same time the UK had risen from fourth to third place in regard to phase 1 trials and from 10th to second in regard to phase 2 trials. The implication was that there were difficulties in recruiting patients to the phase 3, so-called pivotal, trials that companies use in applying to regulatory authorities for marketing authorisations.

However, rather than analysing the raw numbers of trials being carried out in each country at each phase, expressing the numbers of trials being carried out per million population would have given a better reflection of what had happened in different countries. China, for example, was second only to the USA in the ABPI’s 2021 table in terms of the total numbers of all three categories of trials, phases 1, 2, and 3; however, it was 10th in the table when the data were corrected for population, and would perhaps have been even lower, had other countries’ performances also been reported.

By a per million population analysis, the UK’s performance stayed steady at fourth place globally in regard to phase 1 trials, fell slightly—from fourth to fifth place—in regard to phase 2 trials, and fell from sixth to seventh place in regard to phase 3 trials. So the drop in performance on this measure compared with other countries was not as great as the ABPI claimed on the basis of the absolute numbers of trials being performed. Nor did the ABPI include information on the sizes of the trials performed, nor the innovativeness of the products being tested, nor the outcomes.

Nevertheless, as a crude measure, the numbers of trials per million population in the UK did fall markedly, by almost 50%. Despite that, the UK has more or less held its position globally, because fewer phase 3 trials have been carried out everywhere, falling from 2784 by the top 10 countries in 2017 to 2126 in 2021. In contrast, the numbers of phase 1 and phase 2 trials have not changed. This is consistent with the fact that most failures in drug development occur at phase 2,3 when the early promise of a new medication is not fulfilled, although failures can also occur at phase 3.

Other reasons may therefore at least partly explain the reduction in the numbers of phase 3 trials besides difficulty in recruitment. For example, in 2021 fewer drugs were newly licensed in the UK (35, of which few were regarded as innovative) than in the EU (40) or the USA (52).4 A more granular analysis is needed.

Lord O’Shaughnessy’s review has just been published,5 having been carried out with commendable speed, so quickly in fact that one might be forgiven for unworthily wondering whether at least some of the recommendations to which the government would be likely to agree were decided in advance. The report runs to 74 pages and includes 27 recommendations relating to eight stated problems. The government’s response has been published simultaneously, supporting the recommendations and detailing agreement to implement five of them, at a cost of about £120m over the next three years.6

A detailed analysis will require more time for reflection. Instead, I should like to highlight aspects of clinical trials that are not mentioned in the report.

Trial quality and participant adherence

Many factors contribute to the quality of a clinical trial. Absence or at least minimisation of biases, for example; a high rate of adherence to the protocol; participant adherence to the intervention; and a low rate of drop-outs. Of these, participant adherence is perhaps the most neglected.

As the former US surgeon general C Everett Koop once said, “Drugs don’t work in patients who don’t take them.” Poor adherence in clinical trials has several potential consequences, leading to biased results, reduced statistical power, and impaired causal inferences.7 These include:

● failure to confirm efficacy or underestimation of the extent of efficacy;

● by the same token, underestimation of the risks of harms;

● impaired development of innovative drugs and drugs for rare diseases.

Lord O’Shaughnessy’s report says nothing specific about how good participant adherence can be achieved and measured. Adherence to trial protocols gets a passing mention.

Harms

The word “harm” does not appear in the report in relation to the adverse effects of trial interventions. The term “side effects” is mentioned in relation to phase 1 studies. Harm from clinical negligence is mentioned. Terms such as “adverse events,” “adverse effects,” and “adverse reactions” do not appear. The word “risk” appears several times, but not in relation to adverse drug reactions or a comparison with benefits, the so-called benefit-risk ratio, better termed the benefit to harm balance. A single reference to “the risk assessment process” might be related to adverse drug reactions, but could equally apply to “the risk, or perceived risk, of carrying out clinical trial activities to the trial sites themselves,” mentioned elsewhere.

Most early phase clinical studies are too small to detect anything but the most frequent harms. And there is a risk that by rushing to approve a new medication that appears to be highly beneficial, important harms will be missed.

Take, for instance, lorcaserin. The history of medicines that have been developed in the hope of managing obesity has featured more withdrawals than survivals. When we surveyed medicines that had been licensed for use in the management of obesity since 1950, we found that 25 had been withdrawn from the market because of adverse drug reactions.8

Lorcaserin, a selective 5HT_2C receptor agonist, developed for use as a weight reducing agent, gained marketing approval for obesity in the USA in 2012, albeit after initial hesitancy by the Food and Drug Administration (FDA) because of concerns about harms. Concerns about the risks of tumours, psychiatric disorders, and valvulopathies led the European Medicines Agency (EMA) to refuse the manufacturers of lorcaserin marketing approval. When we reviewed the evidence,9 having analysed the results of a large clinical trial in overweight or obese subjects with high cardiovascular risk profiles, whose authors concluded that lorcaserin generated sustained weight loss without increased risks of major cardiovascular events, we came to a different conclusion. We concluded that lorcaserin caused minimal weight loss, that its harms profile was as yet incompletely understood, and that it was unlikely to be a cost effective intervention. We also noted the long list of exclusion criteria for participation in the trial and thought that the published results were not likely to have strong external validity. A few months later, the manufacturer withdrew lorcaserin from the market, at the request of the FDA, because of adverse effects.10

A final thought

Innovation in pharmacological interventions is desirable, as is rapid development of new compounds to improve public access to effective medications. However, any rush to develop new pharmacological interventions should be tempered by assuring that phase 3 clinical trials are of high quality, by paying attention to such matters as participant adherence and the potential harms of newly developed medications as well as their benefits, in an attempt to maximise the benefit to harm balance.