• unplanned admission to hospital for congestive heart failure at 2 years.
• cardiovascular death for congestive heart failure at 2 years.

• all cause mortality [ Time Frame: 2 years ] [ Designated as safety issue: No ]
• worsening of the symptoms (that is defined by either a decrease by <1 Mets in the SAS questionnaire score or an increase by >I class in the NYHA functional class for at least 3 months as compared with the baseline) [ Time Frame: 2 years ] [ Designated as safety issue: No ]
• an increase in brain natriuretic peptide (BNP) by > 30% of the value at the randomization in patients with BNP < 200 pg/ml at the randomization [ Time Frame: 2 years ] [ Designated as safety issue: No ]
• a need for modification of the treatment for heart failure (changes in oral medicine for at least one month or addition of intravenous drug(s) for at least 4 hours) [ Time Frame: 2 years ] [ Designated as safety issue: No ]

• all cause mortality
• worsening of the symptoms (that is defined by either a decrease by <1 Mets in the SAS questionnaire score or an increase by >I class in the NYHA functional class for at least 3 months as compared with the baseline)
• an increase in brain natriuretic peptide (BNP) by > 30% of the value at the randomization in patients with BNP < 200 pg/ml at the randomization
• a need for modification of the treatment for heart failure (changes in oral medicine for at least one month or addition of intravenous drug(s) for at least 4 hours)

The purpose of this study is to compare therapeutic effects of furosemide, a short-acting loop diuretic, and azosemide, a long-acting one, in patients with heart failure, and to test our hypothesis that long-acting diuretics are superior to short-acting types in heart failure.

The mortality and morbidity of heart failure are still high despite emerging evidences that have shown beneficial effects of ACE inhibitor, beta-blocker, ARB, and aldosterone receptor antagonist. Diuretics are the most prescribed in heart failure patients in attenuating symptoms due to fluid retention, and diuretics are recommended as essential medicines in patients with heart failure symptoms and/or fluid retention. However, the effects of a long-term administration of diuretics on morbidity and mortality have not been adequately assessed in the prospective clinical study, and the retrospective analysis did not necessarily indicate the diuretic-induced improvement of mortality. McCurley et al demonstrated the adverse effects of furosemide in a tachycardia-induced heart failure model (J Am Coll Cardiol 2004; 44: 1301-1307). Yoshida et al. demonstrated that the administration of furosemide did not improve mortality rate, while the administration of azosemide, a long-acting loop diuretic, improved mortality rate in a hypertensive heart failure model (Cardiovasc Res 2005; 68: 118-127). If the effects on mortality and/or morbidity of heart failure patients are different among classes of diuretics, we should choose a class to provide better prognosis. Thus, we designed a multicenter prospective study, J-Melodic (Japanese Multicenter Evaluation of LOng- versus short-acting Diuretics In Congestive heart failure) to obtain a clinical evidence about the effects of diuretics in heart failure.

Comparison: Congestive heart failure patients matched with the following conditions will be recruited: (1) clinical diagnosis of heart failure based on a slight modification of the Framingham criteria within 6 months before the entry, (2) twenty years or older, (3) NYHA II or III, (4) loop diuretic(s) is (are) administered currently, (5) no change in baseline therapy and symptoms of heart failure within a month. After screening for eligibility and obtaining written informed consent, patients will be randomized to either azosemide or furosemide treatment in a 1:1 ratio. In any arms, patients are treated with standard therapy including digitalis, mineralocorticoid receptor blockers, ACE inhibitors, ARB, beta-blockers, and calcium channel blockers. Patients discontinued taking previous loop diuretic(s) and were directly rolled over to one of the two arms with either azosemide 30-60 mg/day or furosemide 20-40 mg/day, without a placebo run-in period. The dose of each diuretic will be appropriately adjusted according to symptoms of each patient, and patients will be maintained for the rest of the study. Thereafter, patients are reviewed every 2 to 8 weeks. The planned minimum follow-up period for each patient is 2 years, and electrocardiography, chest X-ray and blood sample will be conducted at the study entry and every 12 months after the randomization.

The primary outcome is a composite of cardiovascular death and unplanned admission to hospital for congestive heart failure. The secondary outcomes are listed as follows: all cause mortality; worsening of the symptoms [that is defined by either a decrease by (1) 1 Mets in the SAS questionnaire score or an increase by (2) I class in the NYHA functional class for at least 3 months as compared with the baseline]; an increase in brain natriuretic peptide (BNP) by more than 30% of the value at the randomization in patients with BNP less than 200 pg/ml at the randomization; unplanned admission to hospital for congestive heart failure, or a need for modification of the treatment for heart failure (changes in oral medicine for at least one month or addition of intravenous drug(s) for at least 4 hours).

• Drug: furosemide
• Drug: azosemide

• Active Comparator: Patients with chronic heart failure are given furosemide.
• Active Comparator: Patients with chronic heart failure are given azosemide.

Inclusion Criteria:

• Clinical diagnosis of heart failure based on a slight modification of the Framingham criteria as previously described within 6 months before the entry
• Current status of heart failure is NYHA II or III.
• Currently, loop diuretic(s) is (are) administered.
• No change in baseline therapy and symptoms of heart failure within a month

Exclusion Criteria:

• Current symptomatic hypotension
• Hypertension that has not been controlled to the satisfaction of the investigator
• Hemodynamically significant (in the investigators opinion) LV outflow tract obstruction (due to either aortic stenosis or ventricular hypertrophy)
• Acute coronary syndrome
• Primary pulmonary hypertension or pulmonary hypertension not due to LV dysfunction
• Serious cerebrovascular disease
• Acute myocardial infarction within the last 3 months
• Patients who require intravenous inotropes
• Cerebrovascular accident within the last 3 months
• Percutaneous coronary intervention or open heart surgery within the last 3 months
• On the waiting list for percutaneous coronary intervention or open heart surgery
• Serum creatinine > 2.5 mg/dl
• Serious liver disease
• Any change in cardiovascular drug therapy within a month prior to randomization
• History of chronic obstructive pulmonary disease or restrictive lung disease
• Diabetes mellitus that has not been well controlled (fasting blood glucose>200 mg/dl、HbA1c > 8%)
• Any life-threatening acute disease
• Patients with implantable cardiac defibrillator
• Other diseases likely to cause death or serious disability during the period of the study
• Patients unable to walk without personal aid