Primary malignant tumors of the heart: Outcomes of the surgical treatment

Introduction

Primary cardiac neoplasms are rare tumors with an autopsy incidence of 0.0017% to 0.28%.^1–6 A minority of these tumors are malignant and their prevalence varies with age: in adults, nearly 25% of diagnosed primary cardiac tumors are malignant,^1–3 while in children it is less than 10%.^4–6 Cardiac tumors are still life-threatening diseases with a poor prognosis even when complete resection is attempted.^1–11 The aim of this study was to review surgical, hospital, and follow-up data of surgical cases at two university hospitals (S. Orsola-Malpighi Hospital and the Monzino Hospital) to identify the risk factors and prognosis of patients operated on for malignant heart tumors.

Patients and methods

The hospital records at S. Orsola-Malpighi and Monzino Hospitals were reviewed to identify patients with primary malignant cardiac tumors who underwent surgery. Between January 1979 and December 2012, 14 patients underwent surgery for primary malignant tumors at the two institutions: 11 were men (78.6%) and 3 (21.4%) were women. The mean age at surgery was 47.4 years with a range of 16.1 to 63.5 years. Secondary cardiac tumors, whether metastatic or invasive, were excluded as were primary sarcomas of the great arteries. All demographic, clinical, and diagnostic data were obtained for every patient. The methods used for diagnosis of a cardiac mass varied during the two decades and included chest radiography, transthoracic and transesophageal echocardiography, magnetic resonance imaging (MRI), and computed tomography (CT). To complete the preoperative assessment, ventricular and coronary angiography were also performed in patients with angina as well as in patients over 50 years of age and those with known coronary artery disease. Additional cardiac imaging techniques such as CT and MRI were used to clarify the extent of the mass, myocardial wall involvement, and the boundaries with mediastinal structures.

Surgical access was by a median sternotomy or left thoracotomy. Operations were performed with or without cardiopulmonary bypass and cardioplegic arrest. Standard cardiopulmonary bypass with moderate systemic hypothermia was utilized. Bicaval cannulation or superior caval cannulation associated with right femoral vein cannulation were used for venous return. All resected masses were sent for pathologic examination and characterization. Malignancy and tumor histologic type were defined at histology, whereas tumor extent and invasiveness were determined at the time of surgery by the chief surgeon assessing the anatomic characteristics of the tumors. Outpatient clinic data as well as clinical data from the cardiology and oncology departments were reviewed for follow-up data.

Analyses were performed using SPSS software (SPSS, Inc., Chicago, IL, USA). Continuous variables are reported as mean ± standard deviation and range (minimum to maximum). Estimates of overall survival were also calculated by the Kaplan-Meier method, and the curves were compared using the log-rank test.

Results

Ten (71.4%) patients were symptomatic: acute heart failure in 2 (14.3 %), arrhythmia in 7 (50%), syncope in 1 (7.1%), cardiac tamponade in 1 (7.1%), and systemic illness in 6 (42.9%). Pericardial effusion was present in 8 (57.1%) patients and pleural effusion in 3 (21.4%). A cardiac murmur was the only sign in 4 patients on physical examination. The New York Heart Association functional class at hospital admission was IV in 3 patients, II–III in 7, and only 4 patients were in class I. Rhythm disturbances were present in 7 (50%) patients: atrial flutter in 3 (21.4%), paroxysmal supraventricular tachycardia in 3 (21.4%), and total atrioventricular block requiring pacemaker implantation during the operation in one.

The main diagnostic tool was echocardiography which identified a cardiac mass in 10 (64.3%) patients. In more detailed echocardiographic evaluations, 6 (42.9%) patients presented with hemodynamic impairment caused by mitral valve obstruction or distortion in 3 (21.4%) and combined tricuspid valve, right atrium, and right ventricular outflow tract obstruction in 3 (21.4%). The diagnosis was made directly with MRI or CT in only 2 patients. Furthermore, the first patient operated on in 1979 was diagnosed by chest radiography, whereas in the patient with cardiac tamponade, the diagnosis was made in the operating room by direct observation. Four patients had metastatic disease. Metastases were located in the lung (4 patients), in the head (1 patient), and in the liver (1 patient).

The operation was performed electively in 11 patients and 3 underwent emergency surgery. A median sternotomy was carried out in all patients except for patient no. 1 who was treated by a left thoracotomy for a better approach to the posterior mediastinum and esophagus. Eight operations were undertaken with cardiopulmonary bypass and cardioplegic arrest, and in 5 patients, tumor resection was performed without the aid of cardiopulmonary bypass. The mean cardiopulmonary bypass time was 79.1 ± 35.5 min (range 51–171 min) with a mean ischemic time of 41.2 ± 16.5 min (range 15–65 min). At surgery, cardiac tumors were found as single masses in 11 (78.6%) patients and as multiple masses (2–4 masses) in 3 (21.4%). The mass was localized in the atria (10 patients; 71.4%), in the atrial and ventricular cavities (2 patients; 14.3%), and in the ventricles (2 patients; 14.3%). The tumor mass was>5 cm in diameter in most patients (8/14; 57.14%); the largest was 20 × 10 cm. Different surgical approaches for tumor resection were used: right atriotomy (3 patients), left atriotomy (4 patients), superior transseptal approach according to the Guiraudon technique (2 patients), and a left and right atriotomy (1 patient). Surgical approaches and types of operation are summarized in Table 1. Surgery consisted of partial resection in 8 patients and complete resection in 4. Complete resection was defined as the absence of gross residual disease after resection, and incomplete resection as the presence of macroscopic residual disease. We performed an exploratory sternotomy associated with multiple biopsies of the right atrium in a tumor with diffuse infiltration of the pericardium as well as the right pleura in a 16-year-old patient with right atrial, inferior vena caval, and right ventricular involvement, and a palliative cavopulmonary shunt was performed.

OPEN IN VIEWER

Table 1.

Demographic, clinical, diagnostic, operative, and postoperative data of patients with primary malignant cardiac neoplasms.

Patient no.	Sex	Age (years)	Histology	Presenting symptoms/signs	Diagnosis	Site	Metastasis	Size (cm)	Surgery	Adjuvant therapy	Follow-up mortality (months)
1	M	46.7	Myxoid liposarcoma	NYHA class II	Radiography	LA wall infiltrating posterior mediastinum	No	20 × 10	Excision	R	Yes (108.3)
2	M	16.1	Rhabdomyosarcoma	NYHA class IV pleural and pericardial effusion, systemic illness	Intraoperative	RA, RV, vena cava	No	6 × 5.3	Cavo-pulmonary shunt with Gore-Tex, incomplete mass removal		Yes (47.5)
3	M	38.2	High-grade non- Hodgkin’s lymphoma	NYHA class IIIpleural and pericardial effusion, systemic illness	TTE	RA, pericardium, great vessels	No	1.3 × 1.8	Explorative sternotomy + biopsy		Yes (23.9)
4	F	63.5	Pleomorphic malignant fibrous histiocytoma	NYHA class IIpleural and pericardial effusion	Radiography, TTE, CT, cardiac catheterization	LA	Lung	5 × 4.2	Debulking	C	Yes (18.3)
5	F	57.4	Fibrosarcoma	NYHA class II systemic illness	Radiography, TTE	Interatrial septum, LA and left appendage	No		Excision		Yes (27.5)
6	M	45.2	Angiosarcomas	NYHA class III pericardial effusion	CT	RA, RV, inferior vena cava, aortic root	No	6.5 × 5	Debulking	C	Yes (7.5)
7	M	63.1	Angiosarcomas	NYHA class IV pericardial effusion, systemic illness	MRI	RA	No	8 × 6.5 × 5.5	Excision	R	Yes (6)
8	F	45.3	Leiomyosarcomas	NYHA class I	TEE	LA	No	7 × 6.8	Excision	C	Yes (17.6)
9	M	59.1	Leiomyosarcomas	NYHA class II pleural effusion	Radiography, TTE, TEE, cardiac catheterization	LA	No	3.2 × 3.5	Debulking	C	Yes (8.1)
10	M	38.9	Angiosarcomas	NYHA class II Pericardial effusion, systemic illness	Radiography, TTE, CT, MRI, PET	RA, interatrial septum	Lung, liver, head	5 × 2.7 × 2	Debulking	C + R	Yes (13.8)
11	M	39	Osteosarcoma	NYHA II systemic illness, syncope	Radiography, CT, MRI, TTE, TEE, cardiac catheterization	LA	Lung	9 × 6.5 × 3.5	Debulking	C + R	Yes (28.9)
12	M	53	Epithelioid hemangioendothelioma	NYHA class I pericardial effusion	Radiography, CT, MRI, TTE	LV	No		Debulking		Yes (30)
13	M	63.5	Angiosarcomas	NYHA class I	Radiography, TTE, cardiac catheterization	RA	Lung		Debulking		Yes (38)
14	M	48.5	Pleomorphic malignant fibrous histiocytoma	NYHA class I pericardial effusion	Radiography, CT, MRI, TTE, PET	LV	No	2 × 4	Debulking	C	Yes (24)

On histological examination, the masses were diagnosed as: angiosarcoma in 4 patients, pleomorphic malignant fibrous histiocytoma or undifferentiated pleomorphic sarcoma in 2, leiomyosarcoma in 2, liposarcoma in 1, osteosarcoma in 1, high-grade non-Hodgkin lymphoma in 1, rhabdomyosarcoma in 1, fibrosarcoma in 1, and epithelioid hemangiosarcoma in one.

There were no hospital deaths. Mean intensive care unit stay was 58.7 ± 55.5 h (range 20–240 h), mean mechanical ventilation time was 8.9 ± 6.7 h (range 3–28 h). The main postoperative complications were bleeding requiring surgical revision in one patient, hyperaldosteronism requiring medical therapy in one patient, and pleural effusion requiring thoracentesis in another. Adjuvant therapy included chemotherapy and radiotherapy in 2 patients, chemotherapy alone in 5, and radiotherapy alone in 2. Adjuvant chemotherapy treatment data were available for only 2 patients and consisted of methotrexate, cisplatin, Adriamycin, isophosphamide in one, and isophosphamide and epirubicin (2 cycles) followed by taxane in the other. To respect the privacy of the patients followed up at other hospitals, details regarding chemotherapy in the remaining patients were not available for evaluation in the present study.

The mean survival was 28.8 ± 28 months (range 6–108.3 months). All patients died of disease recurrence or progression, and 4 were reoperated during follow-up; tumor recurrence was present in all 4, with complete resection. Patient no. 1 developed various recurrences of a primitive liposarcoma and was reoperated on 4 more times at 12, 48, 54 and 60 months after the first operation (mitral valve replacement, a second operation on the left atrium requiring redo mitral valve replacement, bypass grafting of the right coronary artery, and a new exeresis of retro-atrial tissue and on the posterior mediastinum). Thirteen months after the last operation, he underwent cobalt-beam therapy for relapse of the liposarcoma in the posterior mediastinum, and 4 months later he died of chronic congestive heart failure and respiratory insufficiency (9 years after the first operation). Patient no. 5 was reoperated on in another hospital for tumor recurrence in the left ventricle 2 years after the first surgery; unfortunately, she survived for only 6 months. Patient no. 8 had a leiomyosarcoma recurrence in the left atrium and was reoperated on 6 months after the first tumor excision. Patient no. 11 was reoperated on for tumor relapse involving the left atrium, right pulmonary veins, and aortic root 2 years later, he died 7 months later. Patient no. 14 had a relapse of the primary tumor caused by partial exeresis; he was treated with complete surgical excision 10 months after the initial operation. During follow-up, 9 patients underwent adjuvant therapy; 3 of them had complete resection at initial surgery. The mean overall survival was longer in men than women (30.5 ± 8.7 vs. 21.1 ± 3.2 months) although the difference was not statistically significant (p = 0.378). Symptomatic and asymptomatic patients had a mean survival of 27.4 ± 4.3 and 28.9 ± 9.6 months (p = 0.812), respectively. Radical resection at the first surgery was associated with prolonged survival with a mean survival of 39.9 ± 23.2 vs. 24 ± 4 months (Figure 1a) but the difference did not reach statistical significance (p = 0.318). In patients with metastatic disease, the mean survival was 24.7 ± 5.4 months compared to 30 ± 9.6 months in patients without metastatic disease (p = 0.889). The mean survival for those who underwent adjuvant therapy was 25.8 ± 10.6 months compared to 33.4 ± 4.2 months for those in whom adjuvant therapy was not performed (p = 0.241; Figure 1b). When two or more tumors occurred at multiple sites, the mean survival was 24.0 ± 6.8 months compared to 31.0 ± 10.2 months when the tumor was a single mass. Clinical characteristics, survival based on specific adjuvant treatment, recurrence, and reoperations are summarized in Table 1.

OPEN IN VIEWER

Figure 1.

Overall survival of patients who underwent (a) radical resection and partial resection and (b) patients who did or did not receive adjuvant chemotherapy.

Discussion

Primary malignant heart tumors are rare compared to metastases or direct invasion of the heart by malignant neoplasms. The first description of a heart tumor goes back to the 15th century, and the first documented resection was described in 1835. Since then, many advances in the management of heart tumors have been made, leading to a better prognosis for benign primary cardiac tumors. ¹² Primary sarcomas are the most common malignant cardiac tumors. These include: angiosarcoma, pleomorphic malignant fibrous histiocytoma or undifferentiated pleomorphic sarcoma, osteosarcoma, fibrosarcoma and myxosarcoma, rhabdomyosarcoma, leiomyosarcoma, synovial sarcoma, liposarcoma, malignant schwannoma, malignant mesenchymoma, malignant rhabdoid tumor, Kaposi’s sarcoma, and malignant hemangiopericytoma.^13,14 Regardless of the histologic type, these tumors are extremely aggressive and still have a dismal prognosis. Unlike primary benign cardiac tumors which have a higher incidence in women, ¹² malignant tumors are more common in men. In our series, 78.6% of the patients were men, with a male/female ratio of 3.7.

Unfortunately for many patients, the diagnosis is obtained when symptoms are already manifest and the tumor can seen as a systemic disease. The mean survival of symptomatic patients was no less than the asymptomatic patients (27.4 ± 4.3 vs 28.9 ± 9.6 months). Conversely, the occurrence of metastases significantly influenced survival. Diagnosis of cardiac tumors is routinely achieved with echocardiography.^15–18 However, CT and MRI are fundamental in clinical evaluation and determining the location and invasiveness. CT can adequately demonstrate the morphology, location, and extent of a cardiac mass, and its main advantage over echocardiography is in allowing assessment of extracardiac disease or metastasis. ¹⁷ In addition, cardiac-gated MRI has the advantages of a wide field of view, high contrast and spatial resolution, and multiplanar imaging capabilities, which allow precise demonstration and localization of a mass including its anatomic relationship to the cardiac chambers and any involvement of the myocardium, pericardium, or contiguous structures. Nowadays, it is routinely performed in our institutions to better define the diagnosis. ¹⁸ Molecular imaging methods such a 18 F-fluorodeoxyglucose positron-emission tomography assesses the metabolic activity of the mass, allowing characterization of the tumor and disclosing the presence of secondaries. ¹⁹ Because this was a long-term study (1979 to 2012), not all patients were evaluated by CT, MRI, or positron-emission tomography. However, the final word regarding the diagnosis of malignancy and histologic type rests with the pathologist.

As previously mentioned, malignant neoplasms of the heart still have a very poor prognosis. The surgical option alone does not usually provide effective treatment of the disease because the neoplasm often involves a large area of myocardium, leading to incomplete surgical resection, or it may have systemic diffusion. Our findings show good in-hospital results in contrast to the long-term follow-up. We had no hospital deaths but mean survival was only 28.8 ± 28 months. In a recent study by Simpson and colleagues ³ from the Mayo Clinic, the median survival for those who underwent complete surgical excision was 17 months compared to 6 months for those in whom complete surgical remission could not be achieved. In particular, patients in our series with radical resection at the time of the first surgical intervention had a mean survival was 39.9 ± 23.2 months, whereas in those with incomplete resection it was 24 ± 4 months. Recurrence of tumors is frequent even in cases where radical surgery was performed; each of our 4 patients who underwent radical surgery had a recurrence. The role of adjuvant therapy is still controversial; some studies found that conventional postoperative chemiotherapy did not appear to modify the clinical course, but others showed that combined surgical and adjuvant therapy was effective.^20–22 In our experience, it was associated with a worse prognosis: the mean survival of patients underwent adjuvant therapy was 25.8 ± 10.6 months vs. 33.4 ± 4.2 months in those in whom adjuvant therapy was not performed. We may hypothesize that this result is due to the fact that for long time, adjuvant therapy was used only in patients with high-grade malignancy and a worse prognosis. New therapeutic options might come from neoadjuvant therapies that could reduce the malignant mass thus allowing radical excision. At present, wide resection seems to be the best treatment option even if it cannot rule out tumor recurrence. To guarantee a more radical excision, cardiac autotransplants have been pursued with encouraging results. In a series of 20 patients, 17 with a primary malignant heart tumor, Blackmon and colleagues ²³ reported no hospital mortality in patients who underwent isolated heart autotransplantation, but 3 deaths in patients undergoing combined cardiac autotransplant and pneumectomy. Mass size or the degree of myocardium involvement often precludes complete excision, but the role of transplantation in such patients remains unclear. A literature review revealed 28 patients who underwent orthotopic heart transplantation, 21 of them had malignant histology. The mean survival of patients with malignant histology was 12 months even though 7 of them survived for a mean of 27 months without evidence of disease recurrence. ²⁴ In such cases, heart transplantation could be a feasible option.

Several limitations exist, especially considering the retrospective characteristics of the study, the small number of the patients, and the long follow-up period. During this time, advances in cardiac surgical techniques and better myocardial protection solutions have changed our perspective and improved patient survival. Moreover, several data were unavailable, especially regarding tumor staging and adjuvant chemotherapy. Despite these limitations, we believe this study provides insight into the natural course of primitive malignant heart tumors. We concluded that the treatment outcome for patients affected by primary malignant heart tumors remains poor. Aggressive surgery alone does not offer as good a survival rate as in other types of malignant neoplasm.