Following the good news surrounding the SGLT2 inhibitor, empagliflozin, and the GLP-1 receptor agonist, liraglutide, what chance a comeback for the now generic pioglitazone? Data from the Insulin Resistance Intervention after Stroke (IRIS) trial show that pioglitazone halved the progression to diabetes in people with insulin resistance and cerebrovascular disease. Progression to diabetes was a pre-specified secondary end point of IRIS and occurred in 3.8% of the 1,939 patients randomised to 45 mg/day of pioglitazone compared with 7.7% of 1,937 receiving placebo (P < .0001). This followed on from the main IRIS finding (reported earlier in 2016) that pioglitazone reduced recurrent stroke or myocardial infarction (MI) by 24%.

However, this is not the first time that pioglitazone has been shown to reduce the risk of type 2 diabetes. The Actos Now for Prevention of Diabetes (ACT NOW) (2011) demonstrated a 72% reduction in conversion from impaired glucose tolerance to type 2 diabetes. Pioglitazone has also previously showed cardiovascular benefit in the PROactive trial (2005) with a significant 16% reduction in death, MI, and stroke. These benefits have been overshadowed by concerns over side-effects, which include risk of heart failure, bladder cancer and an increased risk of bony fractures; so what did the 5-year long IRIS reveal about safety?

In IRIS, there was no increased risk for incident cancer with pioglitazone (6.9% vs 7.7%, P = 0.29) or heart failure (2.6% vs 2.2%, P = 0.35), but bone fractures did occur at a higher rate (5.1% vs 3.2%, P < 0.01). The heart failure result was probably due to exclusion of patients with heart failure at baseline and aggressive protocols for managing patients who developed oedema. Also, in IRIS there was no signal for an increase in hip-fracture which might be reassuring. Perhaps the future of piogitazone lies in post-stroke treatment protocols rather than in diabetes guidelines?

Professor Steve Bain

June 2016

The results of the Liraglutide Effect and Action in Diabetes – Evaluation of Cardiovascular Outcome Results (LEADER) trial show that liraglutide significantly reduces the primary end-point of major adverse cardiovascular events (MACE) by 13%, compared with placebo. This was driven by a significant 22% reduction in cardiovascular (CV) death although point estimates for the other MACE components (non-fatal MI and non-fatal stroke) were reduced.  All-cause mortality was also significantly lowered by 15% whilst there was no evidence of any increase in heart failure.

These results follow-on from the EMPA-REG OUTCOME study which showed a 14% significant reduction in the MACE primary endpoint for empagliflozin versus placebo. Once again, the result was driven by a reduction in CV death (38%) and there was a 32% significant reduction in all-cause mortality. There were, however, some differences between the study outcomes. The Kaplan-Meier curves parted at an earlier stage in the empagliflozin study and it demonstrated a much larger benefit on the end-point ‘hospitalisation for heart failure’. This would suggest that the mechanisms underlying these findings differ for the two anti-diabetes therapy classes (assuming that these are class effects).

Most recently come the renal outcomes from the EMPA-REG OUTCOME study. Despite one quarter of the trial participants having an eGFR less than 60 mL/min (i.e. being outside the current licence for empagliflozin initiation), there was an improvement in all renal outcomes. Empagliflozin significantly reduced the incidence of macroalbuminuria, doubling of serum creatinine with an eGFR ≤45 ml/min/1.73m² and time to first initiation of continuous renal replacement therapy.

So, how are we to take on board these dramatic data and apply them to patient care? The cynics might suggest that we need to know more before guidelines are changed since there are currently no data for type 2 diabetes patients who do not have advanced CV disease (as was the case in both studies). Furthermore, there are no data regarding the impact of combining the two classes of drugs (which could possibly negate both of their beneficial effects). Whilst these are both reasonable points, there seems to be no doubt that patients with advanced CV disease should be on one or other of these drugs and the current NICE guideline does not take this co-morbidity into account.

The last revision of the type 2 diabetes guideline from NICE was released over six years after its predecessor, although we are told more timely reviews will happen in the future. Awaiting guidance from the FDA may also be a long-drawn-out process; the recent warning regarding heart failure risk with gliptins was essentially a reanalysis of data first-published in September 2013. A change in licence for empaglflozin was submitted in January 2016 but shows no sign of being granted soon.

Perhaps clinicians, supported by medicines management groups, have to examine the current results and consider action whilst the various regulatory agencies negotiate their respective bureaucracies’?

Professor Steve Bain

June 2016