Avoid the mistakes of the past with new drugs

The recent history of new drugs for type 2 diabetes teaches us to approach this area with caution.

Rosiglitazone has been withdrawn in the wake of the Nissen New England Journal meta analysis, before that, Troglitazone was withdrawn after several fatal episodes of hepatic failure. More recently, Pioglitazone has had a large number of cautions and contraindications attached to it, the most recent being a possible increase in the risk of bladder cancer.

Admittedly all three of these examples belong to the same class of drug but these are still salutary lessons that should serve to remind us to treat novel hypoglycaemics with caution. The UKPDS data is now relatively old and included a population of patients who were slimmer, older, smoked more and took less statins at baseline than newly diagnosed type 2 diabetes patients typically are nowadays.

Notwithstanding these limitations, this is still the most authoritative and largest RCT of glycaemic and blood pressure control in newly diagnosed T2DM and one of the largest RCTs in T2DM period. Thus, one could argue that we would still be well advised to pay heed to its results. These clearly and unequivocally showed that good glycaemic control with insulin or metformin or sulphonylureas or combinations of the above and good blood pressure control are highly effective at reducing the incidence of micro and macro vascular complications and moreover, these benefits are long lasting even after cessation of intensive intervention in the so called “legacy effect”.

In contrast, newer agents have a near absolute absence of evidence of long term efficacy, have yet to have their long term safety demonstrated and are massively expensive in comparison with metformin, gliclazide and human insulin. So, how should we approach drugs such as the GLP1 agonists? On one hand, they are effective at reducing HbA1c and have the near miraculous benefit of making weight literally fall off of many patients who take them. On the other hand they are costly, there is no data showing that their BMI and HbA1c benefits translate into a reduced incidence of complications or better still, reduced mortality, and their long term safety is far from proven.

So, what test should we apply when trying to balance these considerations? I would like to make a near heretical suggestion in this day and age of evidence-based, cost effectiveness modelling, and guideline-driven medicine; I would like to suggest that we use good old fashioned clinical judgement and a smattering of common sense! The HbA1c reductions are sizeable for many, the weight benefits are of a magnitude that are likely to yield significant cardiovascular benefits if maintained.

In ‘non-responders’ they should clearly not be continued, but their cautious use in a restricted population of responding patients combined with a high degree of clinical vigilance would therefore seem like a reasonable approach while we await more definitive long term outcome and safety data. I think the mistake that we must not make yet again is the wholesale, unthinking, blanket adoption of yet another new class of hypoglycaemic drug while there is still an absence of proof that we are really giving our patients any meaningful long-term benefit.