Abstract
Invasive cardiac aspergillosis is a rare, potentially fatal disease which poses a significant diagnostic challenge. Combination of clinical details, imaging features, fungal markers, serology and fungal isolation with demonstration of invasion is usually necessary to establish the diagnosis. High index of suspicion is the key for early diagnosis with potential role of cardiac MRI in its early detection, delineation of extent of involvement and guidance to the appropriate site for tissue sampling, thereby allowing for improved prognosis with early institution of appropriate therapy.
Introduction
Invasive cardiac aspergillosis (ICA), an extremely rare entity is almost always seen in immunocompromised patients. The diagnosis is seldom made antemortem due to late clinical presentation and lack of established imaging features. With the advances of cardiovascular imaging techniques, especially cardiac magnetic resonance imaging (CMR), diagnosis of fungal involvement of the heart and its complications can be detected at an earlier stage by imaging appearances which can be confirmed by endomyocardial or trans-thoracic biopsy. High index of suspicion is the key for antemortem diagnosis in these cases. We present a case of ICA in an immuno-competent middle aged man with distinctive imaging features on CMR and histopathologic confirmation.
Case report
A thirty-six year old man resident of India with no known co-morbidities presented with of progressive dyspnea (New York Heart Association – NYHA II) since past 3 months. He had no past history of pulmonary tuberculosis and tested negative for HIV. Chest radiograph showed mediastinal widening and cardiomegaly. Laboratory investigations showed hemoglobin of 11.8 gm/dL, total leukocyte count of 8700 per mm3 and platelets of2.75 lakh per mm3 with normal fasting blood glucose levels. Beta D-glucan level was >500. Blood culture was sterile. ECG demonstrated low voltage complexes in limb and pre-cordial leads and right bundle branch block. Two-dimensional transthoracic echocardiography revealed biventricular asymmetrical hypertrophy (left ventricle more than right ventricle, and septum more than free wall), thickened mitral valve, aortic valve and inter-atrial septum, mild mitral regurgitation, left ventricular ejection fraction of 25–30% and mild pericardial effusion. A diagnosis of cardiac infiltrative disorder was considered.
CMR showed heterogeneous soft tissue mass in the pre-vascular space with ill-defined fat planes with the cardiac chambers. Mild pericardial effusion was seen. There was thickening with altered signal intensity of the right ventricular (RV) free wall and left ventricular (LV) wall with LV hypertrophy (Figure 1). Confluent areas of STIR hyperintensities were seen in the thickened myometrium predominantly in basal and mid-cavity level (Figure 2). Functional analysis revealed LV ejection fraction of 35% and mixed diastolic and systolic biventricular dysfunction. On late gadolinium enhancement (LGE) images confluent transmural enhancement was noted in the thickened myocardium (Figure 3).

Double inverse recovery black blood axial images (A at level of great vessels and B at mid cavity level) show heterogeneous hyperintense soft tissue mass (as compared to skeletal muscles- asterisk) in the pre-vascular space (arrow) with ill-defined fat planes with ascending aorta (a) and main pulmonary artery (p). Note circumferential thickening with altered signal intensity of the right ventricular and left ventricular walls (arrow head arrows in B) along with thickened interventricular septum (arrow in B). (C) Mild pericardial effusion is seen on axial bright blood steady state free precision (SSFP) axial image at mid cavity level (arrows).

STIR images in short axis (mid cavity) (A) and horizontal long axis (B) show hyperintense signal in the thickened biventricular walls including inter-ventricular septum (arrows) sparing of inferior wall (dashed arrows).

LGE images in short axis (mid cavity) (A) and horizontal long axis (B) show intense enhancement in thickened walls (arrows) with areas of sub-endocardial sparing in anteroseptum and apex (dashed arrows). Note inferior wall does not show any enhancement (arrow heads).
Following this the patient underwent endomyocardial biopsy that showed ulcerated endocardium with necrosis and granulomatous inflammation with few multinucleated giant cells (Figure 4A). Septate fungal hyphae confirming to morphology of Aspergillus fumigates was confirmed on Grocott's methenamine sliver stain (Figure 4B). Patient was diagnosed as myocardial invasive mediastinal aspergillosis with severe systolic dysfunction (ejection fraction – 25–30%) and grade II diastolic dysfunction.

(A) microphotograph showing ulceration of endocardium with necrosis and granulomatous inflammation with few multinucleated giant cells on periodic acid Schiff (PAS) stain x20X (B) microphotograph showing septate fungal hyphae (brown black color) on Grocott's methenamine sliver stain x20X.
Patient was treated with parenteral antifungals (injection Voriconazole for 8 weeks) and subsequently shifted to oral Voriconazole which he will continue for next 10 months. In addition he received oral steroids (Prednisolone in tapered doses for 12 weeks). He also received optimal medical therapy for heart failure. He showed progressive symptomatic improvement with follow up transthoracic echocardiography after 4 weeks revealed reduction in LV wall thickness (from 24 mm to 14 mm), however ejection fraction was maintained at 30%. Follow-up CMR is planned after 12 months of treatment or earlier in case he worsens or shows increased Beta D-glucan levels.
Discussion
ICA is an extremely rare disease with dismissal prognosis if not diagnosed and treated in time. It is almost exclusively seen in immune-compromised patients or after solid organ or bone marrow transplantation. Cardiac involvement by aspergillus is generally through hematogenous dissemination with very few cases of contiguous spread from the mediastinal disease.1–4 Four types of fungal cardiac involvement have been described (a) cardiac aspergilloma presenting as intracavitory lesion, (b) myocardial infarction secondary to embolism of aspergillus hyphae, (c) myocarditis with intra-myocardial abscesses with variable involvement of cardiac layers or pericardium and (d) fungal valvular endocarditis.1–3 Our case highlights cardiac involvement by aspergillus with contiguous spread from mediastinum, as there was no evidence of hematogenous spread on laboratory investigations, clinically patient was not that sick as often seen with hematogenous dissemination of disease and imaging was corroborative.
Contiguous involvement of heart by aspergillus is an indolent process that leads to delayed presentation and diagnosis. Therefore high index of clinical suspicion and awareness of imaging features helps in arriving at diagnosis. Nevertheless, the cases of ICA on transthoracic echocardiography demonstrate features of infiltrative cardiomyopathy. Cardiac MR is more specific and rules out other causes of infiltrative cardiomyopathies like amyloidosis, Anderson Fabry's or Sarcoidosis etc. 5 The presence of an ill defined heterogeneous soft tissue mediastinal lesion with loss of fat planes with the cardiac chambers and thickened myocardium suggests an infiltrative process, as in our case. Interstitial myocardial edematous changes appear hyperintense signal on edema sensitive short tau inverse recovery (STIR) sequences. Cardiac function is assessed by cine imaging through steady state free precision (SSFP) imaging. Late Gadolinium imaging (LGE) is important for tissue characterization which is based on the T1 shortening effect of gadolinium based contrast agents (GBCAs) and regional distribution of GBCA within the extracellular interstitium of myocardium. In cases of necrosis, infarction or inflammation, there is expansion of the extracellular matrix, where the GBCA accumulates after 5–10 min of injection. These areas therefore, appear hyperintense on LGE images. Thus, CMR helps in comprehensive anatomical and functional evaluation with improved tissue characterization delineating complete extent of cardiac involvement.
Footnotes
Declaration of conflicting interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship and/or publication of this article.
Ethical approval
Not applicable.
Informed consent
Not applicable.
