A 67-year-old Japanese man with dyspnea on effort was admitted to our hospital in May 2002. He had been treated for several years with angiotensin-converting enzyme inhibitor and warfarin sodium because of hypertension and chronic atrial fibrillation. He did not have any episodes of myocardial infarction or angina pectoris and there was no family history of ischemic heart disease or amyloidosis. We heard the third heart sound without any coarse crackles. His lower extremities had 1+ pitting edema and his neurological findings were almost normal. Resting ECG showed atrial fibrillation and asymptomatic transient ST-segment depression in leads I, _aV_L, and V₅ to V₆. Cardiac-thoracic ratio was 65% on admission. Transthoracic echocardiography showed hypertrophy in the left ventricle (LV) and LV diffuse hypokinesis with moderate tricuspid regurgitation, but no myocardial sparkling echo sign. Holter ECG, without a calcium antagonist, presented asymptomatic transient ST segment depression, suggesting myocardial ischemia (Fig. 1A). Before treatment with a calcium-antagonist, a ²⁰¹Tl-scintigram showed exercise-induced asymptomatic myocardial ischemia in the antero-septal area with ST segment depression (Fig. 1B–E). There were no abnormal findings on a routine blood test; hematochemistry including SAA protein, creatine phosphokinase MB, troponin T, and other inflammatory markers were all normal (Table 1). Following consultation a hematologist ruled out multiple myeloma.

After treatment with an angiotensin II type-1 receptor blocker (ARB) and diuretics for two weeks, we performed cardiac catheterization in July 2002. A pressure study showed almost normal data: pulmonary capillary wedge pressure 13 mmHg; cardiac index 2.7 l/min/m²; LV end-diastolic pressure 12 mmHg. The patient did not show any symptom of heart failure. An acetylcholine (ACh) provocation test (100 μg) demonstrated asymptomatic coronary artery spasm in the left anterior descending artery with ST-segment depression in V₄ to V₆ (Fig. 2A–D). Simultaneously, we recorded cardiac lactate-production in the coronary sinus sampling (cardiac lactate extraction ratio: before-Ach, 46% (control condition); after-ACh; –8%) indicating that the Ach induced myocardial ischemia by spasm. We did not perform an ACh provocation test for the right coronary artery (RCA) due to its hypoplasty. Coronary angiography showed no significant organic stenosis before ACh provocation and after injection of isosorbide dinitrate (Fig. 2B). Left ventriculography demonstrated diffuse hypokinesis with a reduced ejection fraction (EF) of 48% (Fig. 2E, F). We also performed an endomyocardial biopsy from the septal wall of the left ventricle. Histological examination showed hypertrophic cardiomyocytes and scattered microscopic focal myocardial necrosis with non-AA type amyloid deposition (Fig. 3A–D). Immunohistochemical analysis identified the amyloid deposition in the necrotic areas as a transthyretin (Fig. 3E, F), but genetic examination showed no abnormality in the transthyretin gene (data not shown). His vitreous body, stomach, and colon did not show any amyloid deposition and he presented no evidence of systemic amyloidosis.

Based on the above clinical progress, we started treatment with ARB, diuretics, warfarin, and a calcium-antagonist for heart failure and coronary spasm. Three months later with a calcium- antagonist, an exercise ²⁰¹Tl-scintigram and Holter ECG did not reveal any ischemic findings (data not shown). Although there was no significant improvement in LV-contraction on echocardiography (%FS: on admission: 24%; 3 months with a calcium-antagonist: 25%), plasma levels of B-type natriuretic peptide had decreased from 257 to 97 pg/ml and he had no symptoms of heart failure during this follow-up period.

We treated a case of heart failure with asymptomatic coronary artery spasm whose endomyocardial biopsy specimen exhibited scattered microscopic focal necrosis with amyloid deposition. It has been previously recognized that coronary artery spasm is able to cause myocardial infarction (1–3), transient attenuation in cardiac contraction or myocardial stunning (4); this suggests that coronary artery spasm may be a factor causing myocardial damage and subsequently leads to heart failure.

Previously, Sakata et al (5) performed LV-biopsy in patients with dilated cardiomyopathy-like LV dysfunction complicated by multi-vessel coronary spasm (6–9). In their report, the LV biopsy specimen showed hypertrophy in some cardiomyocytes and it also showed myocardial fibrosis without necrosis; a treatment with calcium antagonists normalized the reduced LV functioning. Interestingly, in the present case, an endomyocardial biopsy revealed not only interstitial myocardial fibrosis and/or myocyte hypertrophy, but also microscopic focal-necrosis with amyloid-deposition. We could not observe any significant improvement in LV-wall contraction as reported by Sakata et al after treatment with a calcium-antagonist (5). The reason for the discrepancy between our case and their report is uncertain at present. Their report suggests that LV-dysfunction may be reversible in patients with coronary spasm when the patients present interstitial myocardial fibrosis or myocyte hypertrophy in their biopsy specimen. Our patient's biopsy reveals not only interstitial fibrosis and/or myocyte hypertrophy, but also microscopic focal necrosis with amyloid-deposition. We hypothesize that microscopic focal necrosis and amyloid deposition may indicate extensive myocardial damage and it may therefore have been too late to start treatment with a calcium antagonist in the present patient. It has been clinically established that calcium antagonists are effective for the treatment of coronary spasm (10–12), but it is not evident whether or not calcium antagonists are clinically useful in the treatment of heart failure. Because treatment with calcium antagonists can prevent myocardial ischemia in patients with heart failure and coronary spasm, it is clinically important to determine whether the presence or absence of coronary spasm in patients with heart failure to decide an effective therapeutic strategy.

It has been known that myocarditis or microcirculatory damage can induce microscopic necrosis (13). Because we could not find any infiltration of inflammatory cells in the myocardium, myocarditis may not play a central role in the myocardial necrosis in the present case. Asymptomatic coronary spasm may cause microcirculatory damage in this case because of the improvement in the ischemic findings (normalization of ST-segment in Holter ECG and ischemic change during the exercise ²⁰¹Tl-scintigram) after treatment with a calcium-antagonist.

Generally, it is known that amyloid deposition is present mainly in the peri-microvessels and interstitials in so-called "cardiac amyloidosis", leading ultimately to myocardial damage. The present endomyocardial biopsy showed deposits of amyloid in some of the necrotic lesions. The distribution of the amyloid suggests that the amyloid deposition may be a secondary event associated with myocardial necrosis. It has been previously reported that oxidative denaturation of the amyloid precursor proteins promotes amyloid precipitations (14). We and others have reported that oxidative stress is increased in patients with heart failure (15–18) and also in coronary spasm (19, 20). This increased oxidative stress may be involved in the myocardial amyloid deposition in the present case complicated simultaneously with heart failure and coronary spasm.

In summary, we treated a heart failure patient with scattered microscopic myocardial necrosis evident in a LV biopsy specimen, which most likely resulted from repetitive myocardial ischemia which was induced by asymptomatic coronary artery spasm.