Abstract
What defines a successful pharmaceutical agent for use in type 2 diabetes? The ideal would be one which was taken once and resulted in a cure. This seems some way off, although with the early work on incretin hormones there did seem to be a signal in animal models that beta-cells could recover.1,2 Gastric bypass surgery also seems to indicate that there are gastrointestinal signals that could be manipulated to reverse diabetes. 3 Eight years ago I expressed a hope that we might have a cure in 15 years; there are only a few years left to run before I eat my words and many millions continue to consume pharmaceutical agents to ameliorate, although not cure, their diabetes. Nevertheless the last decade has been the time of considerable advance, and new agents are emerging. However, are they safe and effective? Does ‘safe’ mean ‘absolutely safe’ or is this comparatively safe compared with the possibility of other adverse effects such as renal failure or blindness? And what is the effect? Lowering HbA1c is a useful surrogate, but surely this is not enough in itself. Longevity might be the aspiration, and what could be a better marker of effect than fewer deaths. And here enter the regulators. The US Food and Drug Administration (FDA) espoused cardiovascular outcome trials for new diabetes treatments in 2008 in a paper entitled ‘Guidance for Industry’ 4 and headed, on each page, ‘Contains Nonbinding Recommendations’ (italics are theirs). However, these regulators have shown small inclination to vary their approach, and so the guidance is, in effect, mandated. Carefully formulated definitions apply to both endpoint measures and statistics, although the limits for the confidence intervals have no scientific basis.
Now in some respects the focus on cardiovascular outcome is a laudable development, but it may be too blinkered an approach and too risk adverse to be in the best interests of future generations of patients. Risk aversion of regulators is not compensated by any consideration of general good. Regulators sleep easily when no harm can come to anyone, and are certainly not kept awake by the worry that good medicines are consigned to history. If the current regulators and regulations were in place would we even have insulin as a therapy? After all it causes nasty hypoglycaemia. At one level it is clear that new therapies should have the capacity to extend life compared with current agents. Living longer is a recurrent medical aspiration. Nor, in the converse, would one want to be consuming agents that were dangerous or would reduce longevity. However, the clinical situation is not simple. Concentrating on cardiovascular outcome takes no cognisance of the fact that other factors may be the cause of death: what about cancer 5 or renal failure? Nor do regulators seem to consider consecutive morbidity. Those with diabetes accumulate morbidity after their first myocardial infarction: they may go onto heart failure, or blindness or amputation. And hypoglycaemia, or the lack of it, may make a huge difference to total mortality: unnoticed hypoglycaemia leads to accidents and fractures, or to aspiration pneumonia or inattention to the risks of cold weather. So total mortality may be hugely influenced by hypoglycaemia when no cardiovascular signal is to be found, and this is probably the most plausible cause of the excess mortality in the ACCORD trial. 6
There is a case to challenge the concept of longevity as an aim in itself: there are few who would prefer a long life with severe disability to one that was shorter but with an improved quality of life. So agents that reduce blindness, renal failure, peripheral vascular disease, and especially stroke may be taken by grateful patients even if cardiovascular outcomes were not improved. Secondly, the perfect should not be an enemy of the good. If regulators demand cardiovascular outcome trials this does not have a neutral effect on the pharmaceutical industries or the cost of health services. The cost of a cardiovascular outcome trial has been variously put at US$60–70 million. In 2001 the estimate of cost was US$802 million to develop a new drug and bring it to market and by 2006 the comprehensive estimate of the average cost of developing a new biotechnology product was US$1.2 billion. 7 Pharmaceutical companies wishing to launch new products need to have vast resources of financial reserves or need to merge together to ensure such reserves. Fascinatingly, the regulators are gradually forcing the competition out of the market by making drug development so expensive that merging with other companies is the only option for business survival.
Moreover, the regulators being strictly risk adverse might be in the interests of their own security or political status but the costs imposed on us all for new drugs are real. Pharmaceutical companies need to recoup the costs of outcome trials and these costs are passed onto the consumers and health services. The result is that we might have a new drug which can be consumed and which causes no hypoglycaemia, which would be welcomed by patients, which nevertheless would consume huge resources of health service budget accumulated by the pharmaceutical industry in order to pay for an outcome trial which may or may not be relevant to the diabetic communities needs or wishes.
So where do we go from here? First, one of the old filing cards I found in the Diabetes Research Laboratories years ago illustrates exactly why we need regulators (Figure 1). Stimplete Elixir 8 is thankfully not now on the market: it contained phenobarbitone to suppress electric activity in the brain, and dexamphetamine to stimulate it; it contained four vitamins in inappropriate doses and washed it down with 11% alcohol. It was used for depression, anxiety and slimming in obese patients for no clear reason. It demonstrates why regulation is essential.
A filing card issued by the Pharmaceutical Journal in March 1960. 8
Yes, regulators are here to stay, but we do need an emerging equipoise. Academic commentary and engagement with both the pharmaceutical companies and regulators is necessary if we are to avoid the nihilism of risk aversion on the one hand or unbridled capitalism on the other. There needs to be public discussion about balance of pharmaceutical-induced risks against the risks of the illness itself. We use aspirin and insulin; we use cars and lorries; we use matches and gas-fired ovens. We do not ban these on the basis that they are dangerous or have caused death and destruction. The FDA and the EMA might if they viewed them today. Society can discuss the good and the not so good, the balance of benefit against harm, and even the societal cost of a 95% confidence interval in the 1.8 relative risk of a monocomponent harmfulness.
Nor certitude, nor peace, nor help for pain; And we are here as on a darkling plain Swept with confused alarms of struggle and flight, Where ignorant armies clash by night.