A Randomized Clinical Trial of Trimetazidine, a Partial Free Fatty Acid Oxidation Inhibitor, in Patients With Heart Failure

Patients

This was an individually randomized open-label study. We recruited from the HF clinic of our institution 65 consecutive patients (5 women, age 65 ± 7 years, range 52 to 81 years) with HF (in 35, secondary to ischemic heart disease; all cases evaluated by coronary angiography), New York Heart Association (NYHA) functional class II to IV. As a rule, to ascertain/exclude the presence of coronary disease, all patients attending our clinic with a diagnosis of HF undergo stress/rest cardiac imaging testing and, in case of positive results, coronary angiography. All patients were receiving standard treatment with angiotensin-converting enzyme inhibitors, beta-blockers, long-acting nitrates, digoxin, diuretics, and antiplatelet drugs, as required. After obtaining informed consent, patients were randomly allocated to either conventional therapy plus trimetazidine (20 mg three times daily) (34 patients, 18 post-ischemic) or conventional therapy alone (31 patients, 17 post-ischemic). The randomization was provided by individual sealed envelopes prepared in advance at the investigation site according to a computer-generated random list. The enrollment period lasted 7 months, and the trial was scheduled to last 12 months. Diabetes was present in 4 patients randomized to trimetazidine and in 5 randomized to conventional therapy alone. Investigators were allowed to modify the dose of conventional treatment and to lower or discontinue the dose of trimetazidine, as necessary.

Patients were considered for study in the presence of: 1) persistent symptoms (dyspnea on exertion, fatigue, orthopnea, paroxysmal nocturnal dyspnea or edema, alone or in combination) despite optimized treatment of HF for at least 12 weeks and optimized up-titration of angiotensin-converting enzyme inhibitors and beta-blockers, with stable doses for the last 4 weeks; and 2) an ejection fraction (EF) <45% by transthoracic echocardiography. Patients were excluded in case of an acute myocardial infarction or unstable angina pectoris within 3 months, primary valvular disease, history of any alcohol abuse within 6 months, high-grade arrhythmias, significant renal insufficiency (serum creatinine ≥2.2 mg/dl), coronary lesions suitable for revascularization, left ventricular aneurysm or known active neoplastic process, and any orthopedic or neurological illness that could limit their ability to exercise. At the end of the follow-up period, all patients underwent functional evaluation; the physicians performing the tests were blinded with regard to the arm to which patients had been assigned. The following examinations were performed.

Collection of medical history and physical examination

Symptoms relative to HF were classified according to NYHA functional class. Patients also completed the left ventricular dysfunction questionnaire (LVD-36) to measure the impact of left ventricular dysfunction on daily life and well-being (23). Overall quality of life was evaluated on a visual analog scale (range 0 to 100). Additionally, from when the assay was available in our institution, in 14 and 17 patients of the trimetazidine and conventional therapy group, respectively, a blood sample for serum brain natriuretic peptide levels (NT-pro BNP, Roche Diagnostics, Basel, Switzerland) was drawn in the morning before testing.

Exercise testing

Treadmill exercise tests were performed in the morning, in the fasting state, according to the Bruce protocol. A computer-assisted system for exercise (X-SCRIBE Stress Testing System, Mortara Electronics, Milwaukee, Wisconsin) was used. Blood pressure (cuff sphygmomanometer) and the 12-lead electrocardiogram were recorded at baseline, during the third minute of each exercise step, and throughout recovery. Exercise was terminated for the appearance of >2 mm rectilinear or down-sloping ST-segment depression, for severe dyspnea, angina, fatigue, ventricular tachycardia, or for a blood pressure decrease >10 mm Hg. The cardiologist supervising the exercise test was blinded to the therapy assigned to the patient. Heart rate, systolic/diastolic blood pressure, and rate-pressure product were measured at rest, at the onset of dyspnea and angina, at the appearance of 1-mm ST-segment depression, and at peak exercise. Time to the onset of angina, 1-mm ST-segment depression, and peak exercise were also recorded. To quantify the energy spent during exercise testing, the metabolic equivalent system (METS) was also used. A MET is a unit of energy that approximates the amount of oxygen required under basal conditions at rest and is equivalent to 3.5 ml O₂× kg⁻¹× min⁻¹.

Echocardiography

All studies were performed with a Sonos 5500 (Hewlett-Packard, Andover, Massachusetts) with broad-band transducers capable of second harmonic imaging (Hewlett-Packard S4 with 1.8/3.6 MHz transducer) and recorded on half-inch VHS tape for later review. Fractional shortening was calculated by M-mode measurements in the short-axis view. Measurements of left ventricular end-diastolic volume (EDV) and left ventricular end-systolic volume (ESV) were obtained from the apical four-chamber view by using the single-plane Simpson rule, from which left ventricular EF was calculated as: (EDV−ESV)/EDV×100. The endocardium of the end-diastolic image and the end-systolic image was visually identified and manually digitized. The left ventricular apex was defined as the point on the endocardial contour furthest from the midpoint of the mitral valve plane and the left ventricular long-axis as a line connecting the mitral plane midpoint and the apex.

Statistical analysis

Values are given as mean ± 1 SD or as percentages when appropriate. Analysis was performed in the whole population and in patient subgroups, according to the etiology of HF (ischemic/nonischemic). Differences in mean values between groups were assessed using the 2-tailed Student ttest. Variables not showing a gaussian distribution (assessed by Shapiro-Wilk test) were compared by Mann-Whitney Utest. The chi-square test was performed to investigate NYHA functional class differences between treatments. All calculated p values are 2-tailed and considered as significant when <0.05. Analysis was performed with SAS 6.12 (SAS Institute Inc., Cary, North Carolina).

Functional status

In the trimetazidine group, NYHA functional class decreased 1 grade in 21 patients, decreased 2 grades in 1 patient, increased 1 grade in 2 patients, and remained stable in the remaining 3 patients, whereas in the conventional therapy group it decreased 1 grade in 4 patients, increased 1 grade in 9 patients, increased 2 grades in 2 patients, and remained stable in the remaining 12 patients (p < 0.0001) (Table 3). Accordingly, in the trimetazidine group, LVD-36 score decreased from 18.8 ± 7.0 to 15.1 ± 6.3 (p = 0.038) and quality of life score increased from 63 ± 19% to 71 ± 15% (p = 0.07), whereas in the conventional therapy group, LVD-36 score increased from 18.4 ± 8.8 to 19.2 ± 9.5 (p = 0.07) and quality of life score decreased from 66 ± 12% to 61 ± 20% (p = 0.2; p = 0.04 vs. trimetazidine) (Fig. 1,Table 4). Apart from patients who died during the follow-up period, patients randomized to conventional therapy alone had a higher incidence of cumulative adverse cardiovascular events compared with patients randomized to trimetazidine (Table 5).

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

Effect of trimetazidine and conventional therapy alone on quality of life (0% to 100%), at baseline (solid bars)and at follow-up (open bars). Compared with patients treated with conventional therapy alone, in whom a clear trend toward deterioration was observed, those treated with trimetazidine showed a significant improvement of quality of life.

Ergometric results

Table 4shows detailed results. No significant resting hemodynamic differences were observed between the 2 groups at baseline and at follow-up. Trimetazidine significantly increased peak METS (from 7.37 ± 2.49 to 8.74 ± 2.39, p = 0.04) and total exercise time (from 314 ± 149 s to 402 ± 153 s, p = 0.04). Conversely, these parameters remained stable in the conventional therapy group (METS from 7.35 ± 2.15 to 7.24 ± 1.57, p = 0.8; exercise time from 324 ± 147 s to 316 ± 125 s, p = 0.8). Both exercise time and METS were higher in the trimetazidine group compared with the group on conventional therapy (p = 0.026 and p < 0.006, respectively).

Echocardiographic results

Table 4shows detailed results. Treatment with trimetazidine maintained stable left ventricular EDV (from 150 ± 41 ml to 153 ± 46 ml, p = 0.7) and significantly decreased left ventricular end-systolic volume (from 98 ± 36 ml to 89 ± 27 ml, p = 0.04); accordingly, EF significantly increased from 34 ± 7% to 41 ± 6% (p < 0.001). On the contrary, in the conventional therapy group, both left ventricular EDV and ESV remained stable (from 142 ± 43 ml to 149 ± 63 ml, p = 0.2, and from 92 ± 34 ml to 97 ± 34 ml, p = 0.1, respectively); accordingly, EF did not significantly change (from 36 ± 5% to 34 ± 8%, p = 0.09) (Fig. 2).In the trimetazidine group, ESV was significantly lower and EF higher than in the conventional therapy group (p = 0.048 and p < 0.0001, respectively).

• Download figure
• Open in new tab
• Download powerpoint

Figure 2

Individual effects of trimetazidine (A)and conventional therapy alone (B)on left ventricular ejection fraction (LVEF), at baseline, and at follow-up. Trimetazidine significantly improved LVEF, whereas a trend toward a decline in left ventricular function was observed in patients treated with conventional therapy alone.

Analysis according to etiology of HF

Table 6shows detailed results. Quality of life and LVD-36 improved only in ischemic patients treated with trimetazidine. Both in ischemic and in nonischemic patients, a trend toward improvement in exercise parameters was observed in those treated with trimetazidine. Maladaptive left ventricular remodeling was observed only in ischemic patients on conventional therapy alone, as evidenced by a significant increment of left ventricular ESV, and consequently in a significant reduction of EF. Conversely, ejection EF significantly increased in patients treated with trimetazidine, regardless of the etiology of HF.

Manipulation of cardiac metabolism

A number of different approaches have been used to manipulate energy metabolism in the heart. These involve both indirect measures, as well as the use of agents that directly act on the heart to shift energy substrates use away from fatty acid metabolism and toward glucose metabolism, which is more efficient in terms of ATP production per mole of oxygen used. One way to increase glucose oxidation and to decrease fatty acid metabolism in the heart is to decrease circulating fatty acid levels. This can be achieved by the administration of glucose-insulin solutions (24), nicotinic acid (25), and beta-adrenergic blocking drugs (26,27). Another approach consists of directly modifying substrate use by the heart. Pharmacological agents that inhibit fatty acid oxidation include direct beta-oxidation inhibitors, the so-called 3-ketoacyl-coenzyme A thiolase inhibitors, such as trimetazidine (14) and ranolazine (28). Inhibition of oxidative phosphorylation and fatty acid substrates has been shown to shift substrate use from fatty acid to glucose (29,30). This could also explain why trimetazidine seems particularly effective in diabetic patients (18,31). However, in the present study we could not confirm this issue because the diabetic population was small and equally distributed in the two groups.

Metabolic alterations in HF

Wasting of subcutaneous fat and skeletal muscle is relatively common in HF and suggests an increased use of noncarbohydrate substrates for energy production (32,33). In fact, fasting blood ketone bodies (34) as well as fat oxidation during exercise (35) have been shown to be increased in patients with HF. Insulin resistance has been found to be associated with HF (36), and the consequent impaired suppression of lipolysis could determine the development of ketosis. Experimental studies have shown that sodium dichloroacetate stimulates pyruvate dehydrogenase activity by inhibiting pyruvate dehydrogenase kinase (37,38). Stimulation of pyruvate dehydrogenase activity leads to enhanced glycolysis of glucose and use of lactate by the myocardium for aerobic respiration. Myocardial consumption of FFAs is simultaneously inhibited, with the overall effect of a change of substrate use from predominantly nonesterified FFA to glucose and lactate (39), finally resulting in improved left ventricular mechanical efficiency (40). Similar mechanisms are likely to be at the base of the beneficial effects of trimetazidine in HF patients observed in the present study. These beneficial effects can be explained by the fact that glucose and lactate are more efficient fuels for aerobic respiration, yielding a 16% to 26% improvement in the oxygen consumption efficiency of the myocardium (41). Indeed, experimental and clinical data support the concept that shifting the energy substrate preference away from fatty acid metabolism and toward glucose metabolism is an effective approach to treating heart disease (42–44).

Effects of trimetazidine on cardiac metabolism

From the results of this study, it seems that trimetazidine was equally effective in treating patients with HF of ischemic and nonischemic origin. Indeed, in HF, similar to what happens during acute myocardial ischemia, glucose and lactate oxidation are decreased and fatty acid oxidation is increased, increasing the oxygen requirement per ATP molecule produced. In fact, a major factor in the development and progression of HF is already a reduced availability of ATP, determining a metabolic state that has been defined as energy starvation (45). Fantini et al. (13) observed that the use of palmitoyl-carnitine by isolated cardiac mitochondria is inhibited by high doses of trimetazidine, in the absence of significant changes in pyruvate and citrate oxidation. Additional studies also suggest that trimetazidine acts by affecting myocardial substrate use because the drug inhibits use of fatty acid substrates and shifts metabolism from fatty acid to glucose oxidation (14). Specifically, the agent inhibits beta-oxidation by selectively blocking the activity of 3-ketoacyl coenzyme A thiolase, the last enzyme of the oxidative chain (14). By inhibiting fatty acid oxidation, trimetazidine stimulates total glucose use, including both glycolysis and glucose oxidation. Additionally, trimetazidine increases the incorporation of long-chain fatty acids into the cardiomyocyte membrane (46), thus significantly reducing the availability of cytosolic FFAs and acylcarnitine, which can have deleterious effects on calcium handling. These effects are probably operative on both cardiac and skeletal muscle determining improved cardiac efficiency and peripheral glucose extraction and use.

Finally, a recent study has shown that energy deficiency in HF might result from increased uncoupling proteins (i.e., less efficient ATP synthesis) and depleted glucose transporter protein (i.e., reduced glucose uptake) (47). On this ground, the adoption of drug therapies (such as trimetazidine and ranolazine) aimed at interrupting the metabolic vicious circle in HF has been advocated (48).

Conclusions

The results of this study support the concept that shifting the energy substrate preference away from fatty acid metabolism and toward glucose metabolism by trimetazidine, a ketoacyl-coenzyme A thiolase inhibitor, is an effective adjunctive treatment in patients with HF, in terms of symptoms, exercise tolerance, quality of life, and left ventricular function improvement. These beneficial effects seem to be operative in patients with both ischemic and nonischemic HF. This would be particularly effective in the former, which has a prognosis that is definitively worse. Although highly suggestive, the question of whether these effects could translate into decreased morbidity and mortality needs further investigation. To this end, we believe that time has come to evaluate the effects of partial fatty acid oxidation inhibition in patients with HF in a multicenter, randomized, placebo-controlled trial.