Correspondence *Mayo Clinic, 5 Gonda, 200 First Street SW, Rochester, MN 55905, USA

Email cooper.leslie@mayo.edu

The case

A 47-year-old woman presented at hospital complaining of nausea, anorexia and light-headedness of 2 weeks' duration. She reported three episodes of syncope, and had developed retrosternal chest pressure and dyspnea on the day of presentation. She had no history of cardiac, pulmonary or connective-tissue disease, but did report a history of urinary tract infections and an ankle fracture 5 months before presentation. Her only medication was estrogen patches for menopause and she did not smoke or use illicit drugs.

On examination, the patient was febrile to 38.8 °C, in sinus tachycardia at 115 beats/min and had a blood pressure of 103/63 mmHg. Her blood oxygen saturation was 88% by finger-pulse oximetry. Rales were present bilaterally in the base of the lungs. Cardiovascular examination was remarkable for normal carotid upstroke, jugular venous pressure of 5 cm, distant first and second heart sounds, and no audible rub or gallops. Her liver was slightly tender and there was no peripheral edema.

An initial electrocardiogram showed sinus tachycardia with low-voltage Q waves in leads V1, V2 and V3 associated with ST-segment elevation (Figure 1). Chest radiography revealed cardiomegaly with bibasilar infiltrates, pulmonary vascular congestion and small pleural effusions. Echocardiography, performed at the time of presentation, revealed an ejection fraction of 45% with mild global hypokinesis, moderate inferior hypokinesis and mild concentric hypertrophy. The eosinophil count in the peripheral blood was 0.14 10⁹ cells/l (normal range 0.04–0.40 10⁹ cells/l). Creatinine phosphokinase levels peaked on day 2 at 67.5 ng/ml (normal level <6.2 ng/ml). Troponin T levels peaked on day 4 at 3.95 g/l (normal level <0.3 g/l).

Figure 1 Initial electrocardiogram of the patient at presentation.

The results revealed sinus tachycardia, low-voltage Q waves in leads V1, V2 and V3, and ST-segment elevation in leads V1–V4.

Full figure and legend (109K)Figures & Tables indexDownload Power Point slide (316K)

Coronary angiography revealed normal coronary artery anatomy. Despite diuretics and afterload reduction, the patient's dyspnea and chest pain worsened. Echocardiography performed on day 2 of hospitalization revealed moderate global hypokinesis with an ejection fraction of 32%, severe inferior and apical hypokinesis, and a small pericardial effusion. On day 4, the patient's ejection fraction was 25% with a moderate decrease in right-ventricular systolic function. On this day, she also developed sustained monomorphic ventricular tachycardia and progressive hypotension requiring inotropic drug treatment and INTRA-AORTIC BALLOON COUNTERPULSATION. Endomyocardial biopsy revealed extensive myocyte damage with a diffuse inflammatory infiltrate consisting of eosinophils and, to a lesser degree, histiocytes and mononuclear lymphocytes (Figure 2A,B). Because the patient had a non-perfusing rhythm, emergent cardiopulmonary bypass was carried out with cannulas placed in the right atrium and aorta. Several hours after biopsy, the patient developed ventricular fibrillation that was refractory to 40 min of standard resuscitation efforts. She underwent emergency implantation of right and left ventricular assist devices (VADs) (BVS 5000® ABIOMED Inc., Danvers, MA).

Figure 2 Microscopy of endomyocardial biopsy samples.

(A) At 16 magnification, diffuse inflammation was visible with myocyte replacement in the absence of granuloma or extensive fibrosis. (B) At 65 magnification, dead myocytes and a diffuse eosinophil-rich inflammatory infiltrate were detected, with histiocytes and mononuclear leukocytes in the absence of giant cells.

Full figure and legend (31K)Figures & Tables indexDownload Power Point slide (237K)

After VAD placement, the patient was treated with 10 mg/kg intravenous methylprednisolone daily for 3 days, followed by 1 mg/kg prednisolone daily. A combination of piperacillin and tazobactam was prescribed for broad antibacterial protection for the first 11 days following VAD placement. The presence of pulmonary infiltrates led to piperacillin and tazobactam therapy being stopped on day 12, and empirical doses of meropenem and ciprofloxacin being prescribed. The patient also received prophylactic therapy with trimethoprim–sulfamethoxazole to help prevent Pneumocystis carinii infection. Flow rates were maintained for 5 days at 6–6.5 l/min for the left VAD and 4.9–5.7 l/min for the right VAD. The flows were gradually reduced to 2.2–2.3 l/min for both VADs over the next 5 days.

After 2 weeks, the patient's left-ventricular function had improved (ejection fraction 45%) and the VADs were removed. No evidence of infection was seen, despite immunosuppression. Three months after presentation, the patient's left-ventricular and right-ventricular ejection fractions were 52% and 63%, respectively. Because of the risk of steroid-related diabetes, the dose of prednisolone was decreased to 5 mg daily after 4 weeks and mycophenolate mofetil was added at a dose of 500 mg twice daily. Four months after the patient's initial clinical improvement, her left-ventricular ejection fraction had decreased to 40% and her troponin T levels had elevated to 1.4 g/l. Because of concern about persistent myocarditis, medication with sirolimus at a dose of 1 g daily was started with subsequent clinical improvement, decreased troponin levels and normalization of the ejection fraction. The patient continues to be assessed for symptoms of recurrent myocarditis.

Discussion of diagnosis

The high percentage of eosinophils found in the biopsy sample and the patient's acute symptoms led to a diagnosis of necrotizing eosinophilic myocarditis. This condition is rare and has been reported only in individual case reports and small case series, with most cases diagnosed at autopsy.¹ The prognosis is poor, with a high proportion of patients dying within weeks of illness onset.² In some cases, there is no extracardiac pathology. A report on three patients described viral prodromes, tendencies to allergy, and rapidly deteriorating courses ending in death.¹ The disease course differs from eosinophilia-associated heart disease, in which chronic myocarditis occurs in the setting of systemic disease and extracardiac involvement. In necrotizing eosinophilic heart disease, acute onset and rapid progression of hemodynamic compromise is seen, as illustrated by syncope in the case described, as well as systolic ventricular failure.

On histologic analysis, the presence of a diffuse inflammatory infiltrate with predominant eosinophils, associated with extensive myocyte necrosis, is representative of necrotizing eosinophilic myocarditis. Analysis of pathologic tissue from two children who died of this disease revealed massive eosinophil infiltration and degranulation with deposition of major basic protein on necrotic myocardial cells.³ Degenerating eosinophils and loss of the cytoplasmic membrane leading to extracellular release of eosinophil granules can contribute to myocyte cell death.⁴ The findings under light microscopy were similar (Figure 2). Necrotizing eosinophilic myocarditis differs from typical hypersensitivity myocarditis in that the lesions are diffuse, rather than perivascular and interstitial, and myocyte necrosis is prominent.

Endomyocardial or surgical biopsy is presently the only method for confirming a diagnosis of necrotizing eosinophilic myocarditis, although noninvasive cardiac imaging is essential to assess the extent and severity of cardiac involvement. The ejection fraction can decline within days, as in this case, but recovery following steroid therapy can be similarly rapid. Dilated cardiomyopathy with decreased systolic function is the most common finding on echocardiography; the increased wall thickness due to inflammation can resemble hypertrophic cardiomyopathy, but is reversible.⁵ The presence of a transiently low voltage on electrocardiogram can also mimic an infiltrative cardiomyopathy, such as amyloidosis. Occasionally, the presentation is a focal wall-motion abnormality with pericardial effusion and tamponade.⁶ The distribution and extent of inflammation can be assessed with MRI.⁷ Other disorders should be considered in the differential diagnosis of acute necrotizing eosinophilic myocarditis (Box 1). Eosinophilic myocarditis associated with hypereosinophilic syndrome is typically more indolent than acute necrotizing eosinophilic myocarditis and evolves over several weeks or months. On presentation, the typical symptom is biventricular heart failure, although arrhythmias can lead to sudden death. Hypereosinophilia usually precedes or coincides with the onset of cardiac symptoms, but can be delayed.⁸ Eosinophilic myocarditis has also been reported in association with malignancy, parasite infection and early stages of endocardial fibrosis.

A more benign form of eosinophilic myocarditis has been reported in patients receiving chronic dobutamine treatment while awaiting heart transplantation. Up to 7.4% of these patients have eosinophils present in the explanted heart specimen. The presence of eosinophilic myocarditis at the time of transplant is not, however, associated with a decrease in transplant-free survival.⁹ In contrast to the poor natural history of necrotizing eosinophilic myocarditis, fulminant myocarditis due to active lymphocytic myocarditis has a good prognosis.¹⁰

Treatment and management

A rapid decline in cardiac function despite normal cardiac care indicates a disorder that might respond to immunosuppression. This case supports previous findings that addition of steroid therapy to optimum cardiac care in necrotizing eosinophilic myocarditis restores cardiac function and clinical recovery. Pharmacologic cardiac care should include angiotensin-converting enzyme inhibitors and -blockers, such as carvedilol or metoprolol. In the case described, we used metoprolol and lisinopril. Diuretics are sometimes needed to maintain optimum ventricular filling pressures, although they were not required in this case.

The risk of ventricular tachycardia and heart block is higher in inflammatory cardiac diseases, such as cardiac sarcoidosis^{11, 12} and giant-cell myocarditis,¹³ than in lymphocytic myocarditis or dilated cardiomyopathy.¹⁴ In the case reported here, refractory ventricular tachycardia and ventricular fibrillation required a VAD for temporary circulatory support. Alternative treatment options for ventricular fibrillation refractory to medical management include extracorporeal membrane oxygenation or a total artificial heart. As total artificial hearts are not available at our institution, and the duration of support required was anticipated to be longer than 1 week, we chose to implant VADs. These devices have been used as a successful bridge to recovery in giant-cell myocarditis and lymphocytic myocarditis; however, to the author's knowledge, their use as a bridge to recovery in necrotizing eosinophilic myocarditis is novel.

In this case, there was no evidence of infection related to the VAD after short-term, high-dose steroid therapy, but infectious disorders must be excluded histologically before immunosuppressive therapy is started. There is presently a lack of both knowledge and prospective data to guide the timing, intensity and duration of immunosuppression therapy in cases of necrotizing eosinophilic myocarditis. In the case described, immunosuppression therapy is likely to be tapered gradually after 1 year if the patient's cardiac function and levels of biomarkers (such as brain natriuretic peptide and cardiac troponin) remain normal. Another heart biopsy is not planned unless the patient's clinical status deteriorates. An alternative immunosuppression strategy could include the use of ciclosporin, but here the risk of renal failure was considered greater than the benefit of further immunosuppression therapy. As this case report shows, there is a risk of disease recurrence, which could be related to the relatively low level of immunosuppression achieved with mycophenolate mofetil and methylprednisolone therapy. The risk of disease recurrence after tapering drug doses or discontinuing immunosuppression therapy and the optimum monitoring strategy are unknown. Advances in the treatment of necrotizing eosinophilic myocarditis will depend on the development of multicenter registries, animal models, and improved case recognition by the general public and physicians through increased awareness of its characteristic clinical features.

Conclusions

Clinicians should consider uncommon disorders, such as necrotizing eosinophilic myocarditis, in patients presenting with acute cardiomyopathy symptoms and ventricular tachycardia when the clinical course is progressive despite standard cardiac care. If an endomyocardial biopsy reveals necrotizing eosinophilic or giant-cell myocarditis, the addition of immunosuppressive therapy to standard heart-failure and arrhythmia management could obviate the need for heart transplantation.

References

1. Herzog C et al. (1984) Acute necrotising eosinophilic myocarditis. Br Heart J 52: 343–348 | Article | PubMed | ChemPort |
2. Getz M et al. (1991) Acute necrotizing eosinophilic myocarditis as a manifestation of severe hypersensitivity myocarditis antemortem diagnosis and successful treatment.
   Ann Intern Med 115: 201–202 | PubMed | ChemPort |
3. deMello DE et al. (1990) Cardiac localization of eosinophil-granule major basic protein in acute necrotizing myocarditis. N Engl J Med 323: 1542–1545 | PubMed | ChemPort |
4. Corradi D et al. (2004) Eosinophilic myocarditis in a patient with idiopathic hypereosinophilic syndrome: insights into mechanisms of myocardial cell death. Hum Pathol 35: 1160–1163 | Article | PubMed |
5. Adsett M et al. (2003) Eosinophilic heart: marked left ventricular wall thickening and myocardial dysfunction improving with corticosteroid therapy. Echocardiography 20: 369–374 | Article | PubMed |
6. Kazama R et al. (2003) Therapeutic role of pericardiocentesis for acute necrotizing eosinophilic myocarditis with cardiac tamponade. Mayo Clin Proc 78: 901–907 | PubMed |
7. Chun W et al. (2004) Infiltrative eosinophilic myocarditis diagnosed and localized by cardiac magnetic resonance imaging. Circulation 110: e19 | Article | PubMed |
8. Morimoto S et al. (2003) Changes in the peripheral eosinophil count in patients with acute eosinophilic myocarditis. Heart Vessels 18: 193–196 | Article | PubMed |
9. Takkenberg J et al. (2004) Eosinophilic myocarditis in patients awaiting heart transplantation. Crit Care Med 32: 714–721 | PubMed |
10. McCarthy RE et al. (2000) Long-term outcome of fulminant myocarditis as compared with acute (nonfulminant) myocarditis. New Engl J Med 342: 690–695 | Article | PubMed |
11. Fleming HA (1994) Cardiac sarcoidosis. Lung Biol Health Dis 73: 323–334
12. Sekiguchi M et al. (1996) Cardiac sarcoidosis: diagnostic, prognostic, and therapeutic considerations. Cardiovasc Drugs Ther 10: 495–510 | Article | PubMed | ChemPort |
13. Okura Y et al. (2003) A clinical and histopathologic comparison of cardiac sarcoidosis and idiopathic giant cell myocarditis. J Am Coll Cardiol 41: 322–328 | Article | PubMed |
14. Cooper L et al. (2001) Survival in biopsy-proven cardiac sarcoidosis is similar to survival in lymphocytic myocarditis and dilated cardiomyopathy. In Proceeding of the Second International Congress on Heart Disease, 491–496 (Ed Kimchi A) New Jersey: Medimond

Contact the journal about this article

Subject areas under which this article appears: Intervention | Imaging and other investigations