Diabetes: issues and challenges from a cardiologist’s perspective

Barring contraindications, manage people with diabetes with cardioprotective agents.

Issue: October 2006

ByElliot Rapaport, MD

Elliot Rapaport, MD [photo]
Elliot Rapaport

Cardiologists have been aware for decades that the prevalence of type 2 diabetes among symptomatic patients with coronary disease is significantly higher than that seen in the general population.

More recently it became apparent that the overall likelihood that a middle-aged diabetic patient without evidence of pre-existing atherosclerotic vascular disease would have a major adverse cardiovascular event, such as an acute MI, within the next seven years was comparable to the re-infarction rate among nondiabetic people who had already suffered a prior infarction.

The Adult Treatment Panel III of the National Cholesterol Education Program upgraded diabetes from a simple cardiovascular risk factor to a coronary disease equivalent, establishing it as a future risk comparable to that expected in a patient with established atherothrombotic disease.

Therefore, it becomes desirable that people with diabetes are managed with cardioprotective agents that have shown a benefit in clinical trials to the diabetic population regardless of whether clinically apparent CAD is present or not. The issue arises as to what age such agents should be started, namely low dose aspirin, statins in a dose sufficient to achieve an LDL level lower than 100 mg/dL, and ACE inhibitors.

A recent population-based retrospective study of more than 9 million people in Ontario, Canada, of which 379,003 have diabetes, demonstrated that diabetes confers an equivalent risk to aging of 15 years. It thus seems reasonable that all people with diabetes older than 40 be started on these drugs in the absence of specific contraindications. Additionally, it is important that:

BP is maintained at a level lower than 130 mm Hg/80 mm Hg with antihypertensive agents if required;
strict glycemic control is instituted to ensure that HbA1c is less than 7% with lifestyle and diet modifications and, if required, antidiabetic agents such as metformin;
achieve triglyceride levels lower than 150 mg/dL, and HDL higher than 40 mg/dL in men and 50 mg/dL in women, adding agents such as fenofibrate or niacin if needed (niacin may raise plasma glucose levels slightly but has been proven safe in clinical trials to use in people with diabetes).

Metabolic syndrome

Attention has also focused in recent years on the increasing number of people that demonstrate a clustering of several CVD risk factors that lead to an increased likelihood of future development not only of atherothrombotic CVD but also of type 2 diabetes. The term metabolic syndrome has applied to patients exhibiting this grouping of risk factors, which include abdominal obesity, hypertension, dyslipidemia and hyperglycemia.

There is an ongoing, interesting international debate involving governmental bodies, endocrine societies and national cardiovascular societies as to what constitutes the so-called metabolic syndrome and whether it is in fact a true syndrome. Certainly linking this term to the patient who presents with abdominal visceral obesity has a value in itself, as it calls attention to abdominal obesity as an endocrine organ expressing inflammatory cytokines and other mediators that lead to insulin resistance and resultant hypertension and diabetes.

Use of the term focuses the need to address not only the management of obesity but also the possibility that other cardiovascular risk factors commonly found in association with obesity may be present and need correction. These include hypertension, increased triglycerides or decreased HDL, and dysglycemia.

For that reason, I favor the definition proposed by the International Diabetes Federation that requires abdominal obesity – defined by waist measurement – as a necessary criterion, together with at least two of four other criteria – elevated triglycerides, low HDL cholesterol, elevated BP or dysglycemia. Using this definition, the prevalence of metabolic syndrome among the sample of the U.S. population examined as part of the National Health and Nutrition Examination Survey 1999-2002 was a disturbing high of 39%.

One caveat to focusing on metabolic syndrome is that it could lead some clinicians to overlook modifying or eliminating other additional cardiovascular risk factors, such as smoking, lack of adequate physical exercise or psycho-social factors, that may also be present in a patient with metabolic syndrome. It should be noted that in the Insulin Resistance Atherosclerosis Study, after multivariate adjustment, current smokers exhibited an increased incidence of diabetes compared with those who had never smoked (OR 2.66, P=.001).

Outcomes in multivessel CAD

Atheromatous disease within the epicardial coronary arteries in people with diabetes is generally more extensive and diffuse. Over time it has become apparent that diabetic patients with CAD had a worsened outlook whether managed with an interventional procedure or not. The earliest studies of saphenous vein graft closure following CABG revealed diabetes to be an independent predictor of subsequent vein graft stenosis or occlusion.

Later studies revealed that people with diabetes who are subject to an interventional procedure fared poorly compared with nondiabetic people. In the BARI trial, the prognosis either after CABG or percutaneous coronary intervention was worse among diabetic patients with multivessel disease compared with those who were not diabetic both from the standpoint of mortality and morbidity. Particularly startling to me in that trial, looking at average 5.4-year survival, was the poor outcome among people with diabetes randomized to angioplasty.

Upon completion of the trial, mortality among people with diabetes undergoing CABG was 19% whereas those receiving a PCI had a mortality of 35%, reflecting a lessened cardiac mortality (5.8% vs. 20.6%; P=.0003). The benefit in cardiac mortality with CABG appeared to be confined to those who had received at least one internal mammary artery graft.

At follow-up after a mean of 7.7 years, overall survival in the study population was 84.4% with CABG and 80.9% with percutaneous transluminal coronary angioplasty; however, among the diabetic cohort, survival had fallen to 76.4% with CABG and 55.7% with PTCA.

What was particularly striking to me in BARI was that these patients clinically had either severe angina pectoris or objective evidence of ischemia requiring revascularization on enrollment, but they were not a group of emergent patients presenting with either ST-segment elevation MI or acute coronary syndrome.

The questions that arise are not only whether CABG provides a better interventional strategy than PCI among people with diabetes but also whether PTCA is actually harmful. It should be noted that BARI was conducted prior to the routine use of stents in carrying out a PCI, ie, all patients randomized to PTCA were managed solely with balloon angioplasty.

Nevertheless, I would not have expected one out of three diabetic patients displaying the entry clinical picture in this trial to have died within five years nor close to 50% of them to have died within an average 7.7 years.

It is not entirely clear why diabetic patients with multivessel disease seemingly have done poorly with PTCA. A similarly poorer outcome was observed in the CABRI trial but not in the EAST trial. Most have attributed the poorer results with PTCA to either incomplete revascularization with the initial procedure, accelerated atherosclerosis after revascularization or to the consequences of restenosis. Even with the use of bare metal stents, which greatly reduced the incidence of restenosis, large registry databases, such as those maintained by New York, suggest diabetic patients with multivessel disease have fared better with CABG than PCI.

More recently, trials using drug-eluting stents, such as ARTS II, suggest their superiority over bare metal stents. This benefit appears to extend to the diabetic patient. The DIABETES trial of 160 diabetic patients randomized to either a sirolimus-eluting stent (Cypher, Cordis) or a bare metal stent suggests a clinical benefit with the drug-eluting stent; the incidence of either cardiac death, MI or target vessel revascularization by one year was 10% in the drug-eluting stent group but was 38.8% in the bare metal stent cohort (P<.001).

The ISAR-Diabetes trial suggested that the sirolimus-eluting stent was more effective in preventing restenosis whereas the SIRTAX trial suggested the sirolimus-eluting stent also resulted in a significant reduction in major adverse cardiac events during the first nine months compared with the paclitaxel-eluting stent (Taxus, Boston Scientific) in people with diabetes.

Interestingly, a recent German study of 1,845 consecutive patients presenting with symptomatic CAD managed with drug-eluting stents found on multivariate analysis that diabetes was not a significant predictor of restenosis. The 2005 updated American College of Cardiology/American Heart Association guidelines designate a Class IIb recommendation for PCI in patients with Class III angina who are undergoing medical therapy and have two- or three-vessel CAD with significant proximal left anterior descending disease and treated diabetes.

Hopefully, two ongoing trials will help establish the appropriate role of PCI in the era of modern stent deployment including the use of drug-eluting stents. The BARI 2D multicenter trial involves 2,800 diabetic patients with stable CAD randomized to initial elective revascularization with aggressive medical therapy or aggressive medical therapy alone. The FREEDOM trial involves randomizing 2,400 diabetic patients with multivessel disease to either a PCI with a drug-eluting stent and supporting abciximab (Reopro, Centocor) or CABG.

Studies such as the STRATEGY trial suggest clinical outcome is better with drug-eluting stents than with bare metal stents in patients undergoing a primary PCI. Two other trials directly comparing drug-eluting stents and bare metal stents in patients with an acute MI were the PREMIER and TYPHOON trials. The trials also suggest drug-eluting stents are superior, although the PASSION trial failed to confirm this finding.

In practice, drug-eluting stents are commonly used today in patients undergoing a primary PCI. The role of PCI in people with diabetes undergoing an acute MI is still evolving.

A meta-analysis of 23 clinical trials has established the overall benefit to a patient with an acute MI managed with PCI as compared with fibrinolysis when carried out without undue delay by experienced personnel. Extrapolated to the diabetic patient, it has led to many being taken directly to the cath lab.

Acute MI

Confronted with complete occlusion of the culprit vessel on angiography, even in the presence of multivessel disease, immediate deployment of a drug-eluting stent in the culprit artery is undertaken immediately rather than incur additional delay while arranging for emergent CABG to be carried out. Subsequent need for more complete revascularization can then be determined at a later date and either an additional PCI or CABG subsequently carried out. This approach is rapidly becoming standard practice, although late in-stent thrombosis remains a concern.

People with diabetes experiencing unstable angina or non-STEMI, as well as those experiencing a STEMI, have worse survival than people without diabetes. Even as early mortality from acute MI dramatically fell during the past four decades with improvements in management, the ratio of deaths among people with diabetes to people without diabetes continues to be roughly 2-to-1.

Interestingly, in a post hoc analysis of the FRISC II ACS trial, diabetes was the most important predictor of subsequent death and MI, even more so than the extent of CAD. Important in this context has been the experience with IV GPIIb/IIIa agents in various ACS randomized clinical trials. A meta-analysis by Roffi of the 30-day outcome in the six ACS trials involving 6,458 people with diabetes and 23,072 people without diabetes, where immediate PCI was not mandated, revealed that the benefit from the use of GPIIb/IIIa agents was seen only among those with diabetes.

Mortality was reduced from 6.2% to 4.6% in those with diabetes, but was unchanged in the nondiabetic patients (3% vs. 3%). Among the 1,279 diabetic patients undergoing a PCI, 30-day mortality was reduced from 4% in the control population to 1.2% among those randomized to the GPIIb/IIIa inhibitor. Routine use in the moderate or high-risk diabetic patient undergoing a PCI seems desirable.

PPAR-alpha, PPAR-gamma

Recently, attention has focused on the potential benefit of peroxisome proliferators-activated receptor alpha and gamma agonists in lessening cardiovascular events among people with diabetes. Although the recent experience with muraglitazar (Bristol-Myers Squibb), a new combined PPAR alpha–gamma agonist, was disappointing and resulted in cessation of further testing with the drug, recent separate clinical trials have been carried out in people with diabetes with fenofibrate, a PPAR alpha agonist, and with pioglitazone (Actos, Takeda), a PPAR gamma agonist.

The FIELD study randomized 9,795 people with type 2 diabetes aged 50 to 75 to receive 200 mg daily of micronized fenofibrate or matching placebo for an average follow-up of five years. The primary outcome, a composite of CHD death and nonfatal MI, was not different, although a trend favoring fenofibrate was seen with a HR 0.89 (95% CI, 0.75-1.05; P=.16). When adjusted for drop-in statin use in the course of the trial, the primary endpoint was positive with a HR 0.81 (95% CI, 0.76-0.95l; P=.01).

Furthermore, the secondary endpoint of total cardiovascular events was significantly benefited with fenofibrate. With a negative primary endpoint to a randomized clinical trial, one is constrained to recommend routine use of an agent even when numerous secondary endpoints were achieved. Thus, I cannot routinely recommend the addition of fenofibrates today to all people with type 2 diabetes. Rather, their use would seem to be primarily indicated in patients who remain either hypertriglyceridemic or persist with a low HDL despite statin therapy.

The PROactive trial tested whether pioglitazone compared with placebo in 5,238 type 2 diabetic patients reduced mortality and morbidity in high-risk patients, ie, patients with established macrovascular atherothrombotic disease. A primary composite endpoint involving six different events was not achieved after an average 34.5 months; however, a secondary composite endpoint of all-cause mortality, MI and stroke revealed a significant benefit (HR 0.84; 95% CI, 0.72-0.98; P=.027). Once again, with a negative primary endpoint, I am constrained in advocating the routine addition of the PPAR gamma agonist pioglitazone in routine secondary prevention management of the diabetic patient. The results hopefully will encourage the undertaking of additional trials of PPAR gamma agonists in people with diabetes with underlying macrovascular disease.

People with diabetes not only have an increased prevalence of CAD, but patients who meet the criterion of impaired glucose tolerance without diabetes have an increased likelihood of future major adverse coronary events. With the progressively increasing incidence of impaired glucose tolerance and diabetes reflecting, in part, the aging of our population as well as the epidemic of obesity we are currently experiencing, cardiologists must assume a front line position to ensure that cardiovascular risk factors are aggressively attacked and that our patients who have diabetes receive state of the art clinical care.

For more information:

Elliot Rapaport, MD, is a Professor Emeritus in the Cardiology Department of San Francisco General Hospital.