Trials intended for cardiovascular outcomes offer insight into diabetes therapies

The cardiovascular safety of diabetes treatments has long been questioned by the medical community, in particular the effect of such drugs on the risk for cardiovascular mortality, myocardial infarction and stroke. Now, two new trials addressed the cardiovascular safety of DPP-IV inhibitors, and the findings have created discussion among endocrinologists and cardiologists.

The SAVOR-TIMI 53 and EXAMINE trials were highlighted at the annual European Society of Cardiology (ESC) Congress and European Association for the Study of Diabetes (EASD) meeting and published in The New England Journal of Medicine. In SAVOR-TIMI 53, saxagliptin (Onglyza, AstraZeneca/Bristol-Myers Squibb) added to standard care neither reduced nor increased risk for ischemic events as compared with placebo in patients with type 2 diabetes at high CV risk. In EXAMINE, alogliptin (Nesina, Takeda Pharmaceuticals) added to standard care did not increase major adverse CV event rates as compared with placebo in patients with type 2 diabetes and recent acute coronary syndrome.

The trials were designed to satisfy FDA requirements that new diabetes drugs be investigated further to rule out CV risk. The FDA guidelines were developed in 2008 in the context of concerns regarding the long-term CV safety of several antidiabetic medications, such as rosiglitazone (Avandia, GlaxoSmithKline). The new data demonstrate that both SAVOR-TIMI 53 and EXAMINE fulfilled the FDA mandate of demonstrating noninferiority compared with placebo.

“Importantly and reassuringly, we now have evidence of cardiovascular and pancreatic safety with saxagliptin and alogliptin, which presumably are generalizable to the entire DPP-IV class,” Lawrence A. Leiter, MD, professor of medicine at St. Michael’s Hospital, University of Toronto, told Endocrine Today. “Clinicians have liked DPP-IV inhibitors, given the fact that they are easy to prescribe and effectively lower glucose without hypoglycemia and without weight gain. Therefore, the results are important because now they also add cardiovascular safety and pancreatic safety, and unfortunately in the world of diabetes, there aren’t a lot of our antihyperglycemic agents that have demonstrated cardiovascular safety.”

Saxagliptin data

The SAVOR-TIMI 53 trial (n=16,492) was designed to determine whether daily saxagliptin (5 mg or 2.5 mg) was noninferior to placebo when added to background therapy for the primary composite endpoint of CV death, nonfatal MI or nonfatal stroke, Deepak L. Bhatt, MD, MPH, who presented results of the SAVOR-TIMI 53 trial, said at an ESC 2013 press conference.

During a mean follow-up of 2.1 years, the primary endpoint occurred in 7.3% of patients assigned saxagliptin vs. 7.2% assigned placebo (HR=1; 95% CI, 0.89-1.12; P<.001 for noninferiority; P=.99 for superiority). An on-treatment analysis yielded similar results (HR=1.03; 95% CI, 0.91-1.17). The major secondary endpoint — a composite of CV death, MI, stroke and hospitalization for unstable angina, coronary revascularization or heart failure — occurred in 12.8% of patients assigned saxagliptin vs. 12.4% of patients assigned placebo (HR=1.02; 95% CI, 0.94-1.11). A prespecified secondary endpoint of all-cause mortality occurred in 4.9% of patients assigned saxagliptin vs. 4.2% of patients assigned placebo (HR=1.11; 95% CI, 0.96-1.27).

Reductions in blood glucose were greater with saxagliptin at 2 years (mean HbA1c reduction, 0.5% vs. 0.2%), and more patients assigned saxagliptin achieved or maintained target HbA1c <7% (40% vs. 30.3%; P<.001 for both). Also, fewer patients assigned saxagliptin required the addition or increase of any new antidiabetic medications (23.7% vs. 29.3%; HR=0.77; 95% CI, 0.73-0.82) or insulin therapy initiation for more than 3 months (5.5% vs. 7.8%; HR=0.7; 95% CI, 0.62-0.79).

“With regards to HbA1c, it’s important to note that similar to EXAMINE, SAVOR was not designed as a glucose-lowering trial,” Leiter said. “In fact, the study was designed to minimize any glycemic difference between those patients on saxagliptin and those on placebo because the investigators were encouraged to treat glycemia as per their local guidelines, and we were encouraging them to have good glycemic control in addition to good blood pressure control, lipid control, etc. Still, there was a small, but statistically significant improvement in HbA1c in those patients on saxagliptin.”

More patients in the saxagliptin group reported at least one hypoglycemic event (15.3% vs. 13.4%; P<.001). Hospitalization for hypoglycemia was infrequent and similar between groups (0.6% vs. 0.5%; P=.33).

“When the results of SAVOR were first presented at ESC, people were surprised that there was an increase of hypoglycemia observed in those patients on saxagliptin, and further analysis presented at EASD explained this. Basically, all of the excess in hypoglycemia was seen in patients on background sulfonylurea, especially in those patients who had HbA1c less than 7%,” Leiter said. “We know that the incretin agents do not cause hypoglycemia unless they’re added to another drug that causes hypoglycemia, such as sulfonylurea. In routine clinical practice, one would not add another glucose lowering agent in a patient with relatively normal HbA1c on a sulfonylurea, but if one does add a DPP-IV inhibitor, we would either stop the sulfonylurea or reduce its dose.”

Additionally, saxagliptin was shown to reduce the development and progression of microalbuminuria.

“It is encouraging that within such a short time, 2 years of follow-up, that saxagliptin has demonstrated a beneficial effect on the kidney,” researcher Itamar Raz, MD, head of the diabetes unit, department of medicine, Hadassah University Medical Center, Jerusalem, told Endocrine Today.

Rates of pancreatitis were similar with saxagliptin and placebo (0.3%; P=.77). Five cases of pancreatic cancer were reported in the saxagliptin group compared with 12 in the placebo group. The incidence of other prespecified safety endpoints were balanced, Bhatt said.

“One very important additional finding that came out of SAVOR relates to pancreatic safety,” Leiter said. “It can be argued that 2 years is not long enough to say anything with regards to pancreatic cancer; however, the concerns were raised after an even shorter duration of exposure. So, these findings with regard to pancreatic safety are important and reassuring to clinicians and patients.”

Of note, saxagliptin-treated patients had more hospitalizations for heart failure (3.5% vs. 2.8%; HR=1.27; 95% CI, 1.07-1.51). Heart failure requiring hospitalization was defined as admission or ED visit resulting in at least a 12-hour stay; clinical manifestations of heart failure; and additional or increased therapy.

“In this study, saxagliptin was shown to be a safe drug that did not cause CV harm, as was shown in the primary and secondary endpoint results. Moreover, saxagliptin did improve blood glucose control; it decreased the need for insulin and did not cause weight gain or an increase in hypoglycemia when added to metformin or insulin. This is the largest outcome study of a diabetic drug that supports both the efficacy and safety of this drug. Similar studies on older antidiabetic drugs are missing and most probably will never be done,” Raz said.

Alogliptin data

The EXAMINE trial enrolled 5,380 patients with type 2 diabetes and either acute MI or unstable angina requiring hospitalization within the previous 3 months. All were randomly assigned daily alogliptin (6.25 mg, 12.5 mg or 25 mg) or placebo in addition to existing diabetes and CV drug therapy.

During a median follow-up of 18 months, and up to 40 months, the primary endpoint — a composite of CV mortality, nonfatal MI or nonfatal stroke — occurred in 11.3% of patients assigned alogliptin vs. 11.8% assigned placebo (HR=0.96; upper boundary of the one-sided repeated CI, 1.16; P<.001 for noninferiority), William B. White, MD, chief of the hypertension and clinical pharmacology division and professor of medicine at the Pat and Jim Calhoun Cardiology Center, University of Connecticut Health Center, reported at ESC and EASD.

Outcomes were similar with alogliptin and placebo for the secondary endpoint, which was a composite of death, nonfatal MI, nonfatal stroke and urgent revascularization, White said.

Glycated hemoglobin levels were significantly lower with alogliptin treatment (mean difference, –0.36 percentage points; P<.001). Incidences of hypoglycemia, malignancies including pancreatic cancer, pancreatitis and initiation of dialysis were similar between the two groups. Rates of withdrawal due to adverse events also did not differ.

EXAMINE showed less hypoglycemic events than SAVOR-TIMI 53 and showed relative pancreatic safety, 
Simon Heller, DM, FRCP, professor of clinical diabetes and director of research and development at the University of Sheffield and Sheffield Teaching Hospitals NHS Foundation Trust, told 
Endocrine Today in an interview.

“The severe hypoglycemia was unusual but was in patients who were either on insulin or sulfonylurea and was not common. From that point of view, that’s reassuring,” Heller said. “We didn’t see any cases of carcinoma of the pancreas in either group and we saw a few cases of pancreatitis, which was reassuring. … The events are few, but would I say this definitively establishes safety? Definitely not. It wasn’t a study designed to do that in terms of pancreatitis.”

At ESC 2013, Eugene Braunwald, MD, said the EXAMINE investigators should be congratulated for “successfully completing one of the first large trials of a DPP-IV drug and showing clearly that the primary hypothesis [was] met,” although questions remained.

“I have to admit with some surprise that the reduction observed in HbA1c was not associated with more hypoglycemic events. Since the outcome of this study showed that alogliptin results were not different from placebo, it would be interesting to learn which co-therapies were added to placebo to make that arm similar to the alogliptin arm. This question is relevant and it’s important because if equality can be achieved with the addition of an inexpensive generic drug, it will be appropriate and important to conduct a cost–benefit analysis of alogliptin in this population. It’s important in this high-risk population that these additional questions will be asked,” said Braunwald, who is the Distinguished Hersey Professor of Medicine at Harvard Medical School, and founding chairman of the TIMI Study Group at Brigham and Women’s Hospital.

Heller addressed some of those questions in his interview with Endocrine Today.

“One of the questions people asked and we couldn’t answer was why if both groups — it was a double blind study — why didn’t other people increase their therapy to maintain equal glycemic control in both?” he said. “It might reflect what we call clinical inertia in that they put people on the treatment and as long as they were OK, they didn’t worry about intensifying the therapy. In our study, unlike the SAVOR-TIMI, the therapies did go up a little bit in the other groups but not up to the extent that they made up for the slightly worse glycemic control.”

At EASD in Barcelona, new data from EXAMINE showed that alogliptin did not cause a significant increase in HF hospitalization and was effective at lowering HbA1c and safe regarding pancreatic health.

“Due to the media and the questions that have been asked [since ESC] of me, my steering committee and the sponsor, we actually analyzed a post-hoc, nonspecified evaluation of the EXAMINE study, in which we at least removed the confounding of death,” White said at the meeting. “In none of the analyses, regardless of how they were constructed, did we see a significant difference in any of our endpoints.”

With adjudication, they found a composite of 7.4% in the alogliptin group and 7.5% in the placebo group (HR=0.98, P=.976). Looking at CV mortality alone, the HR was 0.84 (P=.207), and in hospitalization for heart failure, which also considered those occurring after the primary endpoint, the HR was 1.19 (P=.22). Although the overall study excluded patients with unstable CV conditions, of those randomly assigned, 15% had a history of congestive heart failure before their index acute coronary syndrome event. Of patients with that history, 18% of alogliptin patients vs. 22% of placebo patients met a primary endpoint in this post-hoc analysis, White said.

A tale of two studies

The new data can be useful in choosing among numerous available diabetes medications when treating patients with type 2 diabetes at very high CV risk.

“We need, as a field — diabetologists and cardiologists — to do more research to see what we can do to reduce the high rate of macrovascular complications in diabetes. What is clear is that neither of these DPP-IV inhibitors seems to impact favorably on CV outcomes in the intermediate term,” Bhatt said at a press conference.

David S.H. Bell, MD, told Endocrine Today that the SAVOR-TIMI 53 and EXAMINE data are interesting, especially in light of other recent research. “A meta-analysis that I conducted with colleagues and published in the American Journal of Cardiology last year showed a 55% decrease in acute coronary syndrome with DPP-IV inhibition. A meta-analysis from Europe of all DPP-IVs in general showed an approximate 30% decrease in CV events. Another study from McMaster University found a 75% decrease in CV events with these drugs.”

He said the findings in heart failure hospitalization “need further investigation, as approximately 44% of US hospital admissions for heart failure are in patients who also have diabetes.” Although the numbers reported are small, “it cannot be ignored, and this result should be looked at in higher numbers of patients,” said Bell, of Southside Endocrinology and University of Alabama at Birmingham School of Medicine. “On a positive note, the findings of a low incidence of pancreatitis are especially exciting for endocrinologists.”

Leiter said: “DPP-IV inhibitors were already a class of medication that had a lot of appeal to prescribers as well as to patients, given the fact that they are easy to take and effectively lower glucose without risk for hypoglycemia or weight gain, both of which may be seen with other glucose-lowering agents. For both physicians and patients, it is important to know that these drugs have now demonstrated cardiovascular safety.”

Carl J. Pepine, MD, professor of medicine and chief of the division of cardiovascular medicine at the University of Florida, said he is “gratified by the results from these new diabetes trials,” but concerned with the absence of data about other risk factor management.

“This is critically important since the percentage of patients achieving all four risk goals — HbA1c, systolic BP, LDL, smoking cessation — in three major federally funded trials (FREEDOM, BARI2D and the diabetes cohort of COURAGE) was only 12% to 20% at 1 year,” Pepine said. “This result was with fact that all trials had protocols to direct risk management for so-called ‘optimal medical therapy.’ So, let’s not forget that we need to do a better job by simultaneously managing multiple risk factors.”

At the EASD conference, Naveed Sattar, MBChB, MRCP, professor of metabolic medicine at the Institute of Cardiovascular and Medical Sciences at British Heart Foundation Glasgow Cardiovascular Research Centre, said, “We are in a brave new world with diabetes drugs.”

“We’re faced with the reality that in order to meaningfully reduce CV events in patients with diabetes one has to treat them with statins, get the cholesterol to target, the BP to target and to have patients stop smoking. And that reducing glucose by a small, modest amount is not going to change CV risk reduction, at least in the short period over which trials are conducted,” Sattar said. “These trials have allowed us to get closer to [the] truth quicker. They provide robust evidence on speculation, whether it was good or bad. The trials have gotten us there quicker than we would have. They’ve also picked up things that we wouldn’t have picked up unless we do the big trials — potentially a heart failure signal, which has to be taken seriously.”

However, such trials are expensive to conduct, Leiter said in an interview. “Each trial costs several hundred million dollars, and there are about a dozen such trials currently underway, so billions of dollars are being spent on these trials. It has certainly been questioned whether this is the best way to spend that money.” – by Katie Kalvaitis, with additional reporting by Katrina Altersitz

Has this research proven that medications in the DPP-IV class are worth the additional cost above and beyond a generic treatment?

Yes

The data from SAVOR-TIMI 53 and EXAMINE demonstrate the cardiovascular safety of the DPP-IV inhibitor class while data from numerous clinical trials illustrate the potential hypoglycemia and body weight advantage of this class of agents, which translates into both clinical and economic value.

I think that these agents represent specific value for many of our patients considered at risk of hypoglycemia or where there would be additional health impact related to body weight gain. Hypoglycemia and weight gain represent not only clinical challenges for individual patients, but have significant direct and indirect health economic consequences.

Though DPP-IV inhibitors are more expensive as a medication, they could benefit these patients clinically and economically long-term by reducing the hypoglycemia risk and overall weight gain.

L. Marc Evans, MRCP, MD, is a consultant diabetologist at the University Hospital Llandough in Penarth, Cardiff, UK. Disclosure: Evans reports no financial disclosure.

No

In an era where individualization of glycemic targets and therapy is stressed based on a patient’s clinical status/comorbid conditions and potential adverse effects of specific medication, the DPP-IV inhibitors represent another tool in the armamentarium for glycemic control in patients with type 2 diabetes. The side effect profile (potentially less risk of hypoglycemia and less weight gain compared with insulin-provision based therapies) is attractive when compared with agents such as sulfonylurea and insulin, but, unfortunately, the search for agents that may lower the residual CV risk associated with diabetes continues.

Clearly, further studies are required to examine the potential increase of heart failure events (which represents a serious and relatively common complication of diabetes). Until these data are available, DPP-IV inhibitors have not established themselves as superior to other classes of agents, specifically for the prevention of cardiovascular events. Metformin remains first-line therapy, and further data are needed before these potentially more expensive medications are routinely implemented in the algorithim for the treatment of diabetes.

Fortunately, the plethora of ongoing CV outcomes studies should provide further insight, and, as practicing physicians, we need to play close attention to the studies to determine the risk/benefit ratio of DPP-IV inhibitors.

David Aguilar, MD, MPH, is an assistant professor of medicine of the Section of Cardiology at the Baylor College of Medicine in Houston, Texas. Disclosure: Aguilar is a paid consultant to Bristol-Myers Squibb.