Löffler Endocarditis Presenting With Recurrent Polymorphic Ventricular Tachycardia Diagnosed by Cardiac Magnetic Resonance Imaging
A 56-year-old white man was admitted with palpitations, chest pressure, and dyspnea on exertion. He had an embolic stroke 2 months before admission without a defined embolic source. He had history of well-controlled hypertension and allergic asthma treated with theophylline. On admission, the resting ECG revealed sinus rhythm, right axis deviation, anterolateral ST-segment depression, and a normal QT interval (Figure 1). His cardiac biomarkers were elevated to indicate myocardial injury (troponin I, 2.66 ng/mL; creatine kinase-MB, 5.8 ng/mL). While in hospital, he developed episodes of polymorphic ventricular tachycardia (Figure 2) requiring multiple cardioversions and eventually temporary pacing. The initial peripheral blood count revealed a marked leukocytosis (22 000/μL) with a severe hypereosinophilia (47%, 10 000/μL). A transthoracic echocardiogram showed a normal-size left ventricle (LV) with normal global systolic function. The right ventricle (RV) was mildly enlarged with a mildly decreased systolic function. The LV and RV apexes appeared thickened and hypokinetic (Movie I in the online-only Data Supplement). Subsequent left and right cardiac catheterization demonstrated normal coronary arteries and elevated right and left heart filling pressures (right atrial pressure, 20 mm Hg; capillary wedge pressure, 23 mm Hg) with signs of restrictive physiology (Figure 3). An extensive workup for conditions associated with hypereosinophilia was negative. The patient had no rash or other evidence of vasculitis. Parasitosis was ruled out with serial negative stool studies and serologies. There were no elevated titers of autoantibodies, including anti-phospholipids, anti-neutrophil cytoplasmic antibody, and antinuclear antibody to suggest a connective tissue disease. A bone marrow biopsy excluded a myeloproliferative syndrome, and no cytogenetic abnormality was detected. A cardiac magnetic resonance (CMR) study with gadolinium contrast was performed to further investigate the cardiac findings. CMR confirmed normal LV global systolic function and the mildly decreased RV systolic function. The apical walls of the LV and RV were thickened and hypokinetic (Figure 4A and 4B and Movie II in the online-only Data Supplement). There was diffuse subendocardial late gadolinium enhancement (LGE) within the whole LV and RV, which was consistent with extensive endomyocardial fibrosis (Figure 3C and 3D, white arrows, and Figure 5, white arrows). Additionally, LGE images revealed foci of low signal intensity within the endocardial RV suggestive of intracavitary thrombus (Figure 4C and 4D, black arrow; Figure 5, black arrow). The T2* sequence also characterized other areas of low signal intensity in the RV, which were highly suggestive of thrombus (Figure 6, arrows). The edema-sensitive images (T2 weighted turbo-spin echo) demonstrated areas of high signal within the RV wall and within the LV myocardium (Figure 7, arrows). In view of the hypereosinophilia, the CMR findings were consistent with Löffler endomyocarditis with extensive endocardial fibrosis and intracavitary thrombi. Endomyocardial biopsy showed active myocarditis, with extensive eosinophilic infiltrates predominantly involving the endocardial surface with associated organizing mural thrombus (Figure 8). An implantable cardioverter-defibrillator was subsequently implanted for the management of recurrent ventricular tachycardia. Oral corticotherapy with prednisone was initiated despite the lack of a definite cause of the hypereosinophilia. It led to control of the ventricular arrhythmias and normalization of the eosinophils count.
Hypereosinophilic syndrome is a rare disorder characterized by persistent eosinophilia with manifestations in various organ systems. Cardiac involvement, called Löffler endocarditis, is common and is the main determinant of morbidity and mortality in hypereosinophilic syndromes.1,2 Causes of hypereosinophilia include hypersensitivity, medications, infection with parasites and fungi, lymphoma, leukemia, and other myeloproliferative syndromes. Our patient underwent an extensive workup to establish the cause of the hypereosinophilia, including a molecular assessment for fibroblast growth factor and platelet-derived growth factor receptor translocations, both of which were negative. Churg-Strauss syndrome is also associated with hypereosinophilia and can involve the heart; moreover, subendocardial scar on LGE images can be present.2 However, Churg-Strauss syndrome is a small-vessel necrotizing vasculitis associated with asthma or allergic rhinitis. Although our patient had asthma, no vasculitis was present on biopsy material, and demonstration of restrictive physiology and obliteration of the apexes on CMR more strongly supports the diagnosis of hypereosinophilic syndrome and Löffler endocarditis. Pathological findings in Löffler endocarditis include fibrous thickening of the endocardium, apical obliteration, thrombus formation, and restrictive cardiomyopathy, which manifest clinically as heart failure, thromboembolic events, and atrial fibrillation.3 Our patient is unusual because of his presentation with ventricular arrhythmias. Other symptoms are often nonspecific and may include fever and weight loss.
Multicomponent CMR allows imaging of the whole spectrum of the intracardiac manifestations of Löffler endocarditis. Classic findings include LV and RV cavity obliteration, especially the apexes, by a low-signal mass on steady-state free precession cine imaging. On LGE images, a bright and diffuse subendocardial enhancement typically is present and may be associated with low-signal structures consistent with intracavitary thrombi. CMR provides a noninvasive and reliable mean of diagnosis.4,5 Moreover, it provides exquisite visualization of the extent of the disease, which cannot be assessed by endomyocardial biopsy.5 CMR is also useful for clinical followup and treatment monitoring.6
Guest Editor for this article was Leon Axel, MD, PhD.
The online-only Data Supplement is available with this article at http://circ.ahajournals.org/cgi/content/full/122/1/96/DC1.