Although cardiac myxoma is one of the most common types of benign cardiac tumors, infected myxomas are rare. Approximately 90% of these patients present with non-specific symptoms and signs, including myalgia, muscle weakness, arthralgia, fever, weight loss, fatigue, and skin manifestations, making the diagnosis of infected cardiac myxoma challenging [1].

This report describes a patient who presented with culture-negative infected cardiac myxoma. Polymerase chain reaction (PCR) following successful surgical resection showed that the myxoma was infected with Haemophilus parainfluenzae. These findings emphasize the importance of alternative methods for diagnosing infected cardiac myxoma, especially when the culture results are negative for bacteria.

A 39-year-old woman without any other medical or family history presented to the emergency department of a hospital with uncontrolled fever, chills, and myalgia that had affected her intermittently for the previous 5 months. Upon an original examination at a local hospital, she was prescribed medication such as non-steroidal anti-inflammatory drugs, but her symptoms were not controlled. An evaluation at the local hospital 1 month earlier by transthoracic echocardiography (TTE), transesophageal echocardiography, and positron emission tomography resulted in a diagnosis of left atrial myxoma as the cause of her continued fever. Although surgery was planned 2 months later, she was referred to Asan Medical Center due to the aggravation of high temperature (over 38.3°C), headache, general pain, and myalgia.

A physical examination in the emergency room showed that her blood pressure was 111/69 mm Hg, her heart rate was 110 beats/min, her body temperature was 38.2°C, and her respiratory rate was 18/min. Laboratory data showed a white blood cell count of 10,200/L, a hemoglobin concentration of 12.7 dg/L, a platelet count of 243,000/mL, and a serum C-reactive protein concentration of 4.55 mg/dL. Daily blood and urine cultures following admission were negative for growing bacteria. Chest X-rays and electrocardiography were normal. Computed tomography showed a mass in her left atrium, abutting the interatrial septum and suggesting a myxoma. TTE showed a huge echogenic mass with a stalk, with the mass having irregular surfaces and being hypermobile (Fig. 1). Based on TTE, there was a suspicion that the mass posed a high risk of embolic infarction because of its shape. Thus, we decided to proceed with an urgent operation. Three days passed from hospitalization to surgery. Although vancomycin was administered for 2 days before surgery, an infectious disease physician recommended the discontinuation of antibiotic treatment, which we agreed to.

Figure 1. Echocardiography results. (A) Preoperative transthoracic echocardiogram. Apical 4-chamber view showing the large left atrial myxoma, visible as a huge 3.7×3.1 cm echogenic mass with a stalk. (B) Preoperative transthoracic echocardiogram. Parasternal long-axis view, showing the mass in the left atrium. (C) Postoperative transthoracic echocardiogram. Parasternal long-axis view showing that the mass in the left atrium had been removed.

The mass was approached using a 5 cm lateral mini-thoracotomy through an incision in the right fourth intercostal space. Cardiopulmonary bypass was performed through the right femoral artery using Perclose-ProGlide (Abbot Laboratories, Chicago, IL, USA) and the right femoral vein. The operation included antegrade infusion through root cannulation under conditions of moderate hypothermia. The total cardiopulmonary bypass time was 62 minutes and the aortic cross-clamp time was 30 minutes. The excised mass measured 3×4 cm, was oval in shape, had a broad 1-cm stalk, abutted the interatrial septum, and was very fragile (Fig. 2). The tumor, including the atrial septum, was resected. The defect site was closed with 4-0 Prolene. To find the cause of the fever, we requested biopsy tissue from the Department of Laboratory Medicine, extracted DNA, and performed pan-bacterial PCR.

Figure 2. Gross finding. The excised mass had an oval shape with a 1-cm-long stalk and was very fragile.

The patient was transferred to an intensive care unit after surgery and extubated 6 hours later. She was moved to the general ward on postoperative day 2. TTE on postoperative day 4 showed no remnant mass with a normal-sized left ventricle and normal systolic function, including an ejection fraction of 60%. A pathologic examination of the surgically excised mass confirmed the diagnosis of left atrial myxoma.

The patient had no postoperative complications, her fever and myalgia subsided, and she was discharged on day 7 without additional antibiotics. At a follow-up visit at the outpatient clinic 1 week after discharge, pan-bacterial PCR of the surgical specimen revealed H. parainfluenzae (at 99.87%), although a follow-up blood culture did not reveal any bacteria. The infectious disease specialist recommended that antibiotics be administered to treat endocarditis, and she was prescribed levofloxacin (750 mg/day, oral). However, after 7 days, she experienced dizziness, and was switched to ciprofloxacin (500 mg, twice daily, oral) for 3 weeks.

The study protocol was approved by the institutional review board of the Asan Medical Center (2020-1821).

Although cardiac myxoma is one of the most common types of benign cardiac tumors, infected cardiac myxoma is very infrequent [1]. No cases have been described in Korea, and only 1 case of H. parainfluenzae-infected cardiac myxoma was reported (in 1988) between 1966 and 2014 according to a search of the MEDLINE database [2,3]. Thus, the exact prevalence of Haemophilus-associated cardiac myxoma remains undetermined.

The diagnosis of infected cardiac myxoma may be a challenge. Up to 90% of patients with cardiac myxomas present with nonspecific symptoms and signs, including myalgia, muscle weakness, arthralgia, fever, weight loss, fatigue, and skin manifestations [4,5]. These signs and symptoms also occur in patients with uninfected cardiac myxomas, as well as those with other types of endocarditis, including those with rheumatic fever or immunologic diseases [6]. However, infected myxoma of the heart is associated with worsening clinical signs (i.e., high fever, systemic embolism, and the presence of microorganisms refractory to antibiotic treatment); therefore, it requires careful diagnosis and treatment [2,3,7].

Comparisons of the pathologic characteristics of infected and uninfected myxoma have shown that necrosis, hemorrhage, and inflammation may be characteristics of both [7]. Although little is known about the pathological characteristics of infected cardiac myxoma, the findings in this patient suggested several differences from uninfected myxoma [1,6]. Although cardiac myxoma usually presents with myxoid stroma, the H. parainfluenzae-infected tumor in the present patient presented with a pathology that included clumped red blood cells due to severe hemorrhage. Pathologic slides showed uncommon borderlines around the tumor, including plasma cell infiltration and other features of chronic inflammation (Fig. 3).

Figure 3. Pathologic examination of the tumor. (A) Slide showing that the mass was an atrial myxoma, with hematoxylin and eosin staining of the infected portion showing no bacteria (×40). (B) Slide showing an area of hemorrhage with plasma cells, suggesting a diagnosis of infectious cardiac myxoma. In typical cardiac myxoma, this part consists of myxoid stroma (×200).

Infected cardiac myxomas have been classified into 3 types based on their clinical and pathological characteristics [3]. According to these criteria, a definitive diagnosis of infected myxoma requires culture-positive results or microorganisms on a pathologic examination. However, we must consider that cases of infected myxoma are uncommon. Additionally, research into bacteria that cannot be detected using current culture tests is limited. The molecular approach has a sensitivity of 72% and a specificity of 100%, whereas the culture-based approach has only a sensitivity of 26% and a specificity of 62%. Even for infective endocarditis, which has a relatively large number of reported cases, fastidious organisms are reported to cause 5%–15% of all cases and 50% of culture-negative infective endocarditis [8]. For this reason, diagnostic methods other than bacterial culture should be considered for infected myxoma patients. In particular, 16S rRNA gene sequencing is useful when antibiotic treatment is initiated before blood culture, when pathogens are difficult to multiply using conventional culture methods, and when some pathogens that pose serious biological risks, such as Coxiella burnetii and Brucella spp, need to be identified [8,9]. Our finding of a PCR-detected H. parainfluenzae-infected myxoma in a patient with negative blood and urine culture results suggests the need for additional diagnostic methods for surgical specimens. Broad-range PCR of surgical specimens is a powerful method for determining the infective agent when traditional microbiological and serologic diagnoses do not provide a definitive result.

In conclusion, the findings from this patient indicate the difficulty of determining whether infected cardiac myxoma can cause fever of unknown origin, especially as infected cardiac myxoma has no specific pathologic characteristics. Areas of severe inflammation with infiltrating plasma cells may, however, be indicative of infected cardiac myxoma.

Second, the atypical presentation of infected cardiac myxoma in this patient highlights the challenges in diagnosing patients with negative blood culture results. PCR was found to be a reliable tool for diagnosing this patient, suggesting the need for alternative laboratory tests in the diagnosis of infected cardiac myxoma. PCR may be especially useful in determining the etiologic cause of infected myxoma in culture-negative patients.

Author contributions

Conceptualization: SAK, SHJ. Data curation: SAK, WKP, SHJ. Formal analysis: SAK, SHJ. Methodology: SAK, WKP, SHJ. Project administration: SAK, SHJ. Visualization: SAK. Writing–original draft: SAK. Writing–review & editing: SAK, SHJ. Final approval of the manuscript: SAK, SHJ.

Conflict of interest

No potential conflict of interest relevant to this article was reported.

Funding

This research did not receive any specific grant from funding agencies in the public, commercial, or not-forprofit sectors.