Wednesday, October 4, 2023
Presenter: Anthony Keech, MD (University of Sydney, Australia)
The largest intervention study to date in patients with diabetes mellitus, the Fenofibrate Intervention and Event Lowering in Diabetes (FIELD) trial, has shown that patients with type 2 diabetes who took fenofibrate for 5 years had significant reductions in total cardiovascular disease events, particularly nonfatal myocardial infarction (MI) and coronary revascularization.[1] However, the primary endpoint of the FIELD trial, major coronary events, was not significantly reduced by fenofibrate compared with placebo. The FIELD investigators believe this result could be partly due to the high rate of statin use in the placebo group (other lipid-lowering therapies were permitted in this trial), which may have masked a larger treatment benefit from fenofibrate, but other possible reasons for the failure of fenofibrate to produce the expected treatment effect remain unclear. In addition to being presented in Dallas, Texas, at the American Heart Association (AHA) meeting, the main results of the FIELD trial were published simultaneously in The Lancet.[2]
Patients with type 2 diabetes are at 3- to 4-fold risk for developing cardiovascular disease compared with nondiabetic individuals. Fenofibrate is known to reduce blood levels of total cholesterol, low-density lipoprotein (LDL)-cholesterol, triglycerides, and apolipoprotein B and to increase high-density lipoprotein (HDL)-cholesterol. No large, prospective, randomized trials of fibrates in patients with type 2 diabetes had previously been reported. Therefore, the rationale for FIELD was to determine whether early intervention with fenofibrate could prevent cardiovascular events in middle-aged to elderly patients with type 2 diabetes, with or without preexisting valvular disease or lipid abnormalities.[3]
To be eligible for FIELD, subjects had to be between the ages of 50 and 75 years, with total cholesterol levels 3.0-6.5 mmol/L (116-251 mg/dL) plus either a total-to-HDL cholesterol ratio ≥ 4.0 or a triglyceride level > 1.0 mmol/L (> 89 mg/dL), and had to have type 2 diabetes mellitus defined according to World Health Organization criteria (onset of diabetes at age > 35 years). Patients were excluded if they had triglyceride levels > 5.0 mmol/L (> 443 mg/dL). These lipid entry criteria were consistent with recruiting for people who would not qualify for fully subsidized lipid-modifying treatment under the government guidelines in the countries in which they were recruited.
FIELD investigators recruited a total of 9795 patients (mean age, 62 years; 63% men) at 63 clinical centers in Australia, Finland, and New Zealand.[4] The median duration of diabetes in these patients was 5 years, and the median HbA1c was 6.9%, indicating good glycemic control. A high proportion, 56%, had a history of hypertension, and only 22% had prior cardiovascular disease.
All patients received dietary advice over a 4-week period. They then began a 6-week single-blind placebo run-in, followed by a 6-week active run-in period on fenofibrate 200 mg (comicronized formulation) daily for all patients. At this point, they were randomized in a double-blind fashion to treatment with fenofibrate 200 mg or matching placebo for 5-7 years or until ≥ 500 primary outcome events (fatal coronary heart disease [CHD] or nonfatal MIs) were recorded.
Although patients were to have no indication for lipid-lowering therapy on randomization, at any time after randomization they could be started on any appropriate therapy at the discretion of the patient's usual physician. Thus, FIELD evaluated the role of fenofibrate on a background of "usual care," which in practice meant statin therapy. In the end it was found that during the trial, a higher proportion of the patients on placebo than those on fenofibrate went on to statin therapy (17% vs 8%, respectively), and as will be seen, this must be factored into interpretation of the results.
After 4 months of treatment, compared with placebo, fenofibrate had reduced total cholesterol, LDL-cholesterol, and triglycerides by 11%, 12%, and 29%, respectively, and raised HDL-cholesterol by 5%. These differences decreased throughout the trial in the total population, but especially in patients who were on additional lipid-lowering drugs. In patients not on any lipid-lowering therapy other than fenofibrate, the changes in lipids at 4 months were similar to those in the total population and were maintained throughout the study, except for the rise in HDL-cholesterol, which fell from 5% to 2%.
After a median of 5 years, there were 544 primary outcome events (nonfatal MI- or CHD-related death). The primary outcome was reduced by 11% in the fenofibrate group vs the placebo group, but this difference was not significant (Table 1). This difference was accounted for by a significant decrease in nonfatal MI, but a nonsignificant increase in CHD mortality. The excess of CHD deaths in the fenofibrate group (110 vs 93 events in the placebo group) was not due to a significant increase in any particular cause of death, although it was mostly due to an increase in sudden cardiac death (70 vs 64 events, respectively).
| Placebo (%) | Fenofibrate (%) | HR | 95% CI | P | Value |
|---|---|---|---|---|---|
| Coronary events | 6 | 5 | 0.89 | 0.75-1.05 | .16 |
| CHD mortality | 2 | 2 | 1.19 | 0.90-1.57 | .22 |
| Nonfatal MI | 4 | 3 | 0.76 | 0.62-0.94 | .010 |
CHD = coronary heart disease; CI = confidence interval; HR = hazard ratio; MI = myocardial infarction
When statistical adjustment was made for some patients receiving new statin therapy, however, the reduction in coronary events in the fenofibrate group increased to 19% vs placebo, and this was significant (P = .01).
The secondary outcome, total cardiovascular disease events (a composite of cardiovascular disease mortality, MI, stroke, and coronary and carotid revascularization), was significantly reduced by 11% (Table 2). This was mainly driven by the 24% reduction in nonfatal MI and a significant 21% reduction in the need for coronary revascularization. Nonsignificant reductions in other secondary outcomes, including stroke, were also seen in the fenofibrate group.
| Placebo (%) | Fenofibrate (%) | HR | 95% CI | P Value | |
|---|---|---|---|---|---|
| Total CVD events | 13.9 | 12.5 | 0.89 | 0.80-0.99 | .035 |
| CVD mortality | 2.6 | 2.9 | 1.11 | 0.87-1.41 | .41 |
| Total mortality | 6.6 | 7.3 | 1.11 | 0.95-1.29 | .18 |
| Total stroke | 3.6 | 3.2 | 0.90 | 0.73-1.12 | .36 |
| Nonhemorrhagic stroke | 3.2 | 2.9 | 0.91 | 0.73-1.14 | .43 |
| Coronary revascularization | 7.4 | 5.9 | 0.79 | 0.79-0.93 | .003 |
| All revascularization | 9.6 | 7.8 | 0.80 | 0.70-0.92 | .001 |
CI = confidence interval; CVD = cardiovascular disease; HR = hazard ratio
Again, after statistical adjustment for new statin therapy, the reduction in total cardiovascular disease events with fenofibrate vs placebo was significant, at 15% (P = .004).
A significant reduction in cardiovascular disease events with fenofibrate was seen in the 7664 patients with no prior cardiovascular disease (19%, P = .004), but not in the 2131 with prior cardiovascular disease. Clear benefit of fenofibrate was also seen in patients aged < 65 years (P = .003), but not in those aged ≥ 65 years. Gender, high vs low waist measurement, or presence/absence of hypertension or metabolic syndrome did not influence the effect of fenofibrate.
Fenofibrate also reduced the rate of patients progressing to microalbuminuria (3.5 to < 3.5 mg/mmol) or macroalbuminuria (≥ 3.5 mg/mmol), and fewer patients regressed in terms of urinary albumin (Table 3). This change was highly significant over 5 years (P = .002).
| Microalbuminuria/Macroalbuminuria | Placebo (%) | Fenofibrate (%) |
|---|---|---|
| Regression | 8.2 | 9.4 |
| No change | 74.6 | 73.2 |
| Progression | 11.0 | 9.5 |
Compared with the placebo group, patients in the fenofibrate group also had a significant 30% reduction in need for first retinal laser therapy (P = .0003), an 18% reduction in hospitalizations for angina pectoris (P = .04), and a 31% reduction in need for amputations (P = .04). Taken together, these results suggest a beneficial effect of fenofibrate on the microvasculature, although the FIELD investigators admitted that the mechanism for this is unknown. These results do mean that FIELD is the first study to suggest that fenofibrate reduces both diabetes-related kidney disease and eye disease.
Fenofibrate was demonstrated to be well tolerated, alone and in combination with statin therapy, with a similar number of treatment discontinuations to that in the placebo group. Serious adverse events were reported in 0.8% of patients on fenofibrate and 0.5% of those on placebo. A small, nonsignificant increase in noncardiovascular deaths was seen in the fenofibrate group, due exclusively to a higher rate of death from cancer (3.4% vs 3.0%, respectively). Patients on fenofibrate had small increases in risk of pulmonary embolism (1.1% vs 0.7%, P = .022) and pancreatitis, a known but rare side effect of fibrates (0.8% vs 0.5%, P = .031). Rhabdomyolysis was rare (< 0.1% incidence) in both groups, and no case was reported in patients taking statin therapy. Fenofibrate was not more commonly associated with elevated levels of alanine aminotransferase or creatine kinase, but creatinine levels (> 200 micromol/L) were more often reported (1.5% vs 1.0%).
"FIELD provides important information on the role of early intervention with fenofibrate in people with type 2 diabetes," said Prof Keech. The results are likely to be of particular importance in patients without previous cardiovascular disease. "Use of fenofibrate should now be considered in the context of well-established statin therapy, where its main use is likely to be in combination," he and the other FIELD investigators believe.
AHA-designated discussant Henry N. Ginsburg, MD (College of Physicians and Surgeons, Columbia University, New York, NY), said that it was difficult to compare the results of FIELD with those of previous fibrate trials, since those trials provide no clues to explain the FIELD results. He stressed that it was not clear whether lipid changes obtained with fibrates were key factors in determining outcomes, especially since the high rate of statin therapy added during the trial could have had an important effect on the outcome. Future studies to determine whether an additional therapeutic agent adds benefit should always be conducted on the background of (near) optimal therapy (preferably controlled by the investigators) rather then "usual care" therapy, Dr. Ginsburg recommended.
Commenting in an editorial accompanying the FIELD paper in The Lancet,[5] Helen Colhoun, MD (University College Dublin, Ireland), said that neither the results in the primary prevention group nor the results in the secondary prevention group were conclusive. Her conclusion is that the evidence of benefit from FIELD is not sufficient to warrant a guideline recommendation for increased fenofibrate use in patients with diabetes, nor do the results "provide convincing evidence of the benefit of fenofibrate therapy in patients already at target serum LDL cholesterol." Prof. Colhoun was co-principal investigator of the Collaborative Atorvastatin Diabetes Study (CARDS),[6] a trial of primary prevention with atorvastatin 10 mg daily in almost 3000 patients with type 2 diabetes without high LDL-cholesterol. Last year, these investigators reported a 37% reduction in the primary outcome, a composite of acute coronary heart disease events, coronary revascularization, or stroke, with atorvastatin after almost 4 years.
Both Dr. Ginsburg and Prof. Colhoun noted that more answers to questions about the safety and efficacy of combined use of fenofibrate and a statin in diabetes patients should come from the Action to Control Cardiovascular Risk in Diabetes (ACCORD) trial. This ongoing National Heart, Lung and Blood Institute-sponsored multicenter trial has a double 2 x 2 factorial design and is recruiting 10,000 patients, 5800 already on standard or intensive glycemic control, for randomization to simvastatin with or without addition of fenofibrate.[7] The results of ACCORD are expected to be announced in 2009-2010.
