doi: 10.1161/CIRCULATIONAHA.107.783126
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
(Circulation. 2008;118:S7-S15.)
© 2008 American Heart Association, Inc.
Original Articles |
Survival After Resection of Primary Cardiac Tumors
A 48-Year Experience
From the Brigham and Women’s Hospital (A.W.E.), and Harvard Medical School, Boston, Mass; and the Mayo Clinic (J.A.D., R.C.D., C.J.M., T.A.O., F.J.P., H.V.S.), Rochester, Minn.
Correspondence to Andrew W. ElBardissi, MD, MPH, Department of Surgery, Brigham and Women’s Hospital and Harvard Medical School, 75 Francis Street, Boston, MA 02115. E-mail aelbardissi{at}partners.org
 |
Abstract |
|---|
 Top Abstract IntroductionPatients and MethodsResultsDiscussionReferences |
|---|
Background— Primary cardiac tumors are rare but have the potential to cause significant morbidity if not treated in an appropriate and timely manner. To date, however, there have been no studies examining survival characteristics of patients who undergo surgical resection.
Methods and Results— From 1957 to 2006, 323 consecutive patients underwent surgical resection of primary cardiac tumors; 163 (50%) with myxomas, 83 (26%) with papillary fibroelastomas, 18 (6%) with fibromas, 12 (4%) with lipomas, 28 (9%) with other benign primary cardiac tumors, and 19 (6%) with primary malignant tumors. Operative (30 day) mortality was 2% (n=6). Univariate analysis indicated that patients who underwent resection of fibromas and myxomas had superior survival characteristics in comparison to the remainder of tumor variants; these results were consistent after adjusting for age at surgery, year of surgery, and cardiovascular risk factors. Based on actuarial characteristics of the 2002 U.S. population, patients who underwent myxoma resection had survival characteristics that were not significantly different from that of an age and gender matched population (SMR 1.11, P=0.57) whereas those who underwent resection of fibromas (SMR 11.17, P=0.002), papillary fibroelastomas (SMR 3.17, P=0.0003), lipomas (SMR 5.0, P=0.0003), other benign tumors (SMR 4.63, P=0.003), and malignant tumors (SMR 101, P<0.0001) had significantly poorer survival characteristics. Furthermore, malignant tumors in younger patients were highly fatal (HR 0.899, P<0.0001). Although the most significant predictor of mortality was tumor histology, survival was also influenced the by the duration of CPB and NYHA III/IV; the impact of these risk factors varied with time. The cumulative incidence of myxoma recurrence was 13% and occurred in a younger population (42 versus 57 years, P=0.003) with the risk of recurrence decreased after 4 years.
Conclusions— Surgical resection of primary cardiac tumors is associated with excellent long-term survival; patients with cardiac myxomas have survival characteristics that are not significantly different from that of a general population. Predictors of mortality are primarily related to tumor histology but also include clinical characteristics such as symptomatology and duration of CPB.
Key Words: cardiac tumor • myxoma • lipoma • fibroma • malignant sarcoma • survival • surgical resection
|
Introduction |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
Primary cardiac tumors are rare entities, occurring with a lifetime incidence of 0.0017% to 0.02%.1 In fact, malignant tumors that metastasize to the heart are nearly 50 times more common than tumors originating from the heart.2 Patients in the former group face a poor prognosis, and treatment is often not directed at cardiac involvement but rather on treating the origin of the cancer or providing palliative care. Comparatively, patients with primary cardiac tumors face an uncertain treatment strategy that ranges from conservative management to aggressive surgical intervention. Of those tumors that originate from the heart, approximately 75% are benign with nearly 50% of these being myxomas.3 Other primary benign tumors include lipomas, fibromas, hemangiomas, teratomas, and rhabdomyomas, whereas primary malignant tumors of the heart are predominantly sarcomas.4 Clinically, tumors of the heart cause symptoms by three mechanisms: intracardiac obstruction, systemic embolization of tumor fragments, or constitutional symptoms by unclear mechanisms.5 Because of the rarity of primary tumor occurrence, diversity of histology, and anatomic origin, there are only small case series in the literature describing the epidemiology, presentation, and outcome after medical and surgical treatment. Furthermore, while resection for benign tumors appears to be curative, long-term survival after resection is poorly defined. The purpose of this analysis is to more clearly define the survival characteristics and prognostic indicators of long-term survival after resection of primary cardiac tumors.
|
Patients and Methods |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
From December 11, 1957 to April 19, 2005, 521 consecutive patients presented to our institution and were diagnosed with a cardiac mass suggestive of a cardiac tumor. Of these, 329 patients were diagnosed with a cardiac tumor that originated from the heart, 323 (98%) patients were treated surgically, and 6 (2%) were treated conservatively. Of the 323 patients included in this study, 94% (n=304) had benign tumors and 6% (n=19) had malignant tumors. Benign tumors included myxomas (n=163, 50%), papillary fibroelastomas (PFEs) (n=83, 26%), fibromas (n=18, 6%), lipomas (n=12, 4%), and other benign tumors which were grouped together to gain statistical power (n=28, 9%; calcified amorphous tumors [n=8], hemangiomas [n=7], teratomas [n=6], unilocular developmental cysts [n=5], and rhabdomyomas [n=2]). Malignant tumors (n=19) were also grouped together and included, leiomyosarcomas (n=8), angiosarcomas (n=6), malignant fibrohistiosarcomas (n=3), and rhabdomyosarcomas (n=2). The Mayo Foundation Institutional Review Board approved this study. The authors had full access to and take full responsibility for the integrity of the data. All authors have read and agree to the manuscript as written.
Statistical Analysis
Demographic and other patient-related data were obtained from Mayo Clinic medical records. Follow-up information was obtained from subsequent clinic visits, written correspondence from local physicians, and questionnaires mailed to patients or families. The results were analyzed in multiple phases. Pearson’s chi-squared tests were used to compare characteristics between categorical variables relating to presenting characteristics and operative data of specific tumor types/locations. Analysis of variance was used to compare characteristics between continuous variables relating to presenting characteristics and operative data of specific tumor types/locations. An automated backward logistic regression model was used to determine associations between cardiopulmonary bypass time (CPB) and operative factors. Primary end points were predefined and included death from any cause, tumor recurrence, and New York Heart Association Classification (NYHA) status at follow-up.
To analyze survival after surgical resection, a Cox proportional hazards model was initially used to compare the unadjusted hazard ratios between histological tumor variants. Using a baseline survival function estimate derived from the Kalbfleisch-Prentice estimator (pages 114 to 1186), median survival was calculated for each histological type and adjusted for the year of surgery, age at surgery, and presence of coronary artery disease. The observed mortality rates between histological types were subsequently compared with expected mortality rates using the U.S. Census Bureau 2002 national life-tables by means of a 1-sample log-rank test7; differences between an age- and gender-matched population was assessed by means of a standardized mortality ratio (SMR).7 After adjusting for year of surgery, predefined demographic and clinical risk factors were entered into a Cox proportional hazards model to identify the most significant mortality risk factors. Model selection was performed using the approach described by Collett (pages 80 to 898) and is briefly described here. After univariate analysis for each predictor, those with a parameter probability value <0.15 were selected for initial entry into the multivariate model. After initial fitting, nonsignificant variables were eliminated using backward selection (P<0.1). Nonsignificant univariate predictors were subsequently tested using forward selection (P=0.1), and all possible 2-way interactions were tested using forward selection (P=0.1 for entering). Finally, all nonsignificant main effects (unless a component of an interaction term) and nonsignificant interactions were removed at a probability value <0.05. The proportional hazards assumption was evaluated for all significant predictors; for those assumptions that did not hold, interaction terms with follow-up time were evaluated to allow for hazard ratio variation. To assess the discriminatory ability of the final model, a Kaplan-Meier survival curve of patients in the 3 tertiles of the linear predictor was created. SAS version 9.1 (SAS Institute Inc) was used for statistical analysis.
|
Results |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
Overall, the mean age at tumor diagnosis was 54±19.5 years and females were disproportionately affected (61%, n=196, P<0.0001). The most common presenting complaints were dyspnea (n=136, 42%), acute embolic event (n=80, 25%), and chest pain (n=73, 22%); approximately 64% (n=207) of patients presented in NYHA III/IV. Cardiovascular risk factors included hypertension in 32% (n=103), history of coronary artery disease in 18% (n=58), diabetes mellitus in 9% (n=30), and hypertrophic obstructive cardiomyopathy (HOCM) in 4% (n=13).
Presenting characteristics of patients categorized by tumor histology are shown in Table 1. There was significant variation in age at diagnosis between different tumor types; fibromas were diagnosed in a younger cohort of patients compared to other histological variants. Cardiovascular risk factors were similar between the different tumor types; however, patients with fibromas had a lower incidence of hypertension. There was a high incidence of HOCM in patients with PFEs. Functional status at the time of presentation also differed significantly among the different histological tumor types; patients with fibromas had greater cardiovascular functional capacity (NYHA I), however these patients also presented with a higher incidence of palpitations and documented ventricular arrhythmias. Constitutional symptomatology was the most common presenting complaint in patients with malignant tumors; a presenting characteristic that was uncommon with other tumor variants.
|
Diagnosis
The majority of patients were diagnosed with echocardiography (n=278, 86%), however, angiography (n=33, 10%), cardiac MRI (n=9, 3%), and computed tomography (n=9, 3%) were also used. Tumor and hemodynamic characteristics (compiled from all diagnostic modalities) are shown in Table 2. There was significant variation in tumor histology by tumor location. Specifically, left and right atrial tumors were primarily myxomas, whereas valvular tumors were predominantly papillary fibroelastomas. There was a relatively equal distribution of the various histological subtypes of ventricular tumors; however, cardiac fibromas occurred exclusively in the ventricle. Additionally, primary tumors of the great arteries were predominantly malignant. There was a higher incidence of mitral regurgitation in patients with left atrial and aortic valve tumors; however, it should be noted that 10 of 12 (83%) patients with aortic valve tumors and mitral regurgitation also had HOCM and experienced regurgitation secondary to systolic anterior motion of the mitral valve. Mitral and tricuspid stenosis secondary to tumor prolapse was relatively common in patients with left and right atrial tumors, respectively, whereas pulmonary regurgitation was a common consequence of right ventricular tumors.
|
Surgical Approach and Resection
Table 3 details operative characteristics by tumor location. There was no significant difference in cardiopulmonary bypass (CPB) time, cross-clamp time, or CPB temperature by tumor location. After adjusting for year of surgery, a multivariate linear regression model indicated that CPB time was significantly longer in patients with fibromas (β=58.25, P=0.008), those who underwent aortic valve repair/replacements (β=14.32, P=0.04), and those who underwent tricuspid valve repair/replacements (β=18.84, P=0.04). Approach to tumor resection varied according to tumor location. The majority of atrial, mitral valve, and tricuspid valve tumors were resected by a single atrial approach, whereas aortic valve tumors were commonly resected via aortotomy. Ventricular tumors demonstrated the greatest variety of resection techniques; approximately one third were resected by ventriculotomy; however, other techniques such as extracardiac resection, atrial, and aortic approaches were also used. The majority of tumors were resected with negative margins, however tumors originating from the great arteries often underwent debulking as a consequence of their malignant histology. Cardiac reconstruction with prosthetic material was commonly performed after atrial tumor resection. Conversely, ventricular reconstruction with prosthetic material was uncommon; only 1 patient required septal closure with a prosthetic patch. Although the rate of concomitant coronary artery bypass grafting procedures was relatively constant between tumor locations, septal myectomy was performed at a high rate in patients with aortic tumors (all PFEs with HOCM). Valve procedures also varied according to tumor location; mitral valve repair and replacements were the most commonly performed procedures, and tumors that originated from the mitral valve frequently required mitral repair. Of note, there was a low incidence of valve replacement in patients with valve tumors.
|
The incidence of postoperative complication was 7.4% (n=24), and included bleeding requiring reoperation, renal insufficiency requiring dialysis, pulmonary embolism, pneumonia, pulmonary effusion, respiratory insufficiency, and sternal wound infections. Operative (30 day) mortality was 2% (n=6). After discharge, all patients with a diagnosis of a malignant tumor underwent at least 1 round of adjuvant chemotherapy, the most common regimen being ifosamide, adrianmycin, and granulocyte-colony stimulating factor; radiation was not used in any cases.
Survival Analysis
Follow-up was obtained in 89% of patients at a mean of 6.16±6.88 years and extended up to 38 years. Univariate analysis of the different histological subtypes of tumors derived from the Kalbfleisch-Prentice estimator (Table 4, Figure 1) revealed that patients with PFEs, other benign tumors, and malignant tumors had poorer long-term survival when compared to patients who underwent myxoma resection. After adjusting for age at surgery, year of surgery, and other cardiovascular risk factors, patients who underwent resection of PFEs, lipomas, and malignant tumors had significantly poorer survival characteristics (Table 1) when compared to patients with myxomas. Of note, while it was determined a priori to force year of surgery in the model, this variable did not appear to be an effect on survival (HR 2.55, P=0.11).
|
|
U.S. Population survival data from the 2002 Census Bureau was subsequently used to compare tumor histology survival characteristics with an age and gender matched population. Comparisons were made by means of standardized mortality ratio, which provides an estimation of survivorship between a specified group of patients and the standard population; statistical significance was determined by a 1-sample log rank test as previously described.7 Patients with cardiac myxomas did not have survival characteristics that were significantly different from an age and gender matched population (SMR 1.11, 95% CI 0.73 to 1.71, P=0.57) whereas patients with PFEs had poorer survival characteristics with a 10-year survival rate of 61% compared to an expected survival of 81% (SMR 3.17, 95% CI 1.53 to 6.59, P=0.0003). Furthermore, while patients who underwent fibroma resection did not have survival characteristics different from those with myxoma resection, survival was worse when compared to an age- and gender-matched population with a corresponding 10-year survival rate of 75% compared to an expected survival of 98% (SMR 11.17, 95% CI: 1.62 to 76.77, P=0.002). Patients with lipomas (actual 10-year survival of 51% versus 85% expected, SMR 5.0, 95% CI 1.72 to 14.6, P=0.0003), other benign tumors (actual 10-year survival of 70% versus 86% expected, SMR 4.63, 95% CI 1.37 to 15.63, P=0.003), and malignant tumors (actual 5-year survival of 21% versus 90% expected, SMR 101, 95% CI 54.65 to 235.67) also had poorer survival characteristics in comparison to an age and gender matched population.
Multivariate Analysis
A multivariate model was created, however evaluation of the proportional hazards assumption for all significant predictors led to the inclusion of a CPB time and follow-up time interaction term and a NYHA and follow-up time interaction term. As such, Table 5 illustrates the final multivariate model which had excellent discriminating ability (P<0.0001) as demonstrated in Figure 2. The strongest predictors of survival were related to tumor histology: malignant and other benign tumors. The effect of malignant tumors interacted with age, suggesting that malignant tumors occurring in younger patients had a poorer prognosis than those which occurred in an older demographic. NYHA functional classification III/IV at presentation was also found to be a strong predictor of death (HR 5.56, P<0.0001). Additionally, the interaction term of NYHA III/IV*follow-up time was also found to be significant in our study (HR 0.883, P<0.0001) indicating that the effect of being highly symptomatic at presentation on the hazard of death declines as time increases. Similarly, the effect of prolonged CPB on the hazard of death decreased as follow-up time increased (HR 0.996, P<0.0001).
|
|
The incidence of tumor recurrence by tumor histology (myxoma versus other) is shown in Figure 3; the cumulative incidence of myxoma tumor recurrence was 13% and increased steadily up to 4 years after which there was a low hazard of tumor recurrence. There was no association between the method of surgical excision and the incidence of tumor recurrence, however patients who experienced a myxoma recurrence were younger than those who did not (42 years versus 57 years, P=0.003). The site of recurrence was at the location of the original tumor in 81% of cases. Analysis of NYHA functional classification (dichotomized into NYHA I/II versus III/IV) at follow-up revealed a significant logistic regression model; patients who presented with aortic stenosis (OR 21, P=0.001) and those that experienced a cardiac tumor recurrence (OR 7.41, P=0.007) were most likely to be severely symptomatic at follow-up in comparison to the remainder of the cohort.
|
|
Discussion |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
Primary cardiac tumors are rare, making it difficult to design studies that effectively characterize the impact on survival after surgical resection. Thus, while the first report of a primary cardiac tumor was made in 18459 and the first successful surgical resection performed in 1955,10 this is the first study to comprehensively examine the long-term survival characteristics of patients who undergo surgical resection for primary cardiac tumors. While some of the earliest surgical series have described patients who underwent cardiac myxoma resection as having exceptional long-term survival characteristics,11 this is the first study to document that these patients do not have survival characteristics that differ from an age- and gender-matched population. Comparatively, our analysis of survival after malignant tumor resection reaffirmed the dismal survival characteristics that have been previously reported. Furthermore, it is clear from this analysis that malignant tumors occurring in younger patients are more aggressive and associated with a particularly poor prognosis. Interestingly, other benign tumors, which represent a grouping of four exceptionally rare tumors, had an unexpectedly poor survival; a finding that may be attributable to the complexity of these tumors and the population in which they occur. After adjusting for these tumor types, predictors of mortality were found to be the duration of CPB time and NYHA III/IV at presentation—both of these hazards on mortality diminished with time.
We identified preoperative (NYHA) and operative (CPB time) risk factors that varied with time. Preoperative NYHA functional classification has been implicated as a predictor of survival in many types of cardiac surgery.12 Although not surprising that this finding remains true in tumor resection, as time after surgery increases, the disparity between the survival characteristics of those who were severely symptomatic and those who were minimally symptomatic diminishes, suggesting that surgical resection of cardiac tumors can have a dramatic effect on survival despite seemingly high preoperative risk classification. CPB time in this analysis was associated with concurrent cardiac procedures (ie, valve replacements) and complex tumors (ie, fibromas), however, also demonstrated a similar effect; the hazard of prolonged CPB time on mortality decreased over time when adjusting for the other significant predictors of morality. In sum, although histology is the greatest predictor of survival, we believe these findings to indicate that many cardiac tumors represent a reversible cause of heart failure, and despite the extent of surgery required for resection or severe symptomatology before surgery, many patients who are otherwise characterized as high risk are likely to enjoy long-term survival if treated surgically.
The findings reported here also indicate that certain measures may be used to improve diagnosis and expedite surgical management. PFEs are the most underdiagnosed cardiac tumor and may be the most common.13 Despite the increasing use of echocardiography for the diagnosis of intracardiac masses, PFEs are often not suspected because of their minimal effect on both symptomatology and hemodynamics. In a recent study from our institution,14 47% of all patients with papillary fibroelastomas were diagnosed incidentally. The strong association with HOCM as well as surgical, radiation, or hemodynamic trauma15–17 provide compelling evidence that patients with primary cardiovascular disease be monitored regularly with echocardiography in an effort to identify tumors early. It is important to emphasize that these confounding concurrent cardiac risk factors are likely to be more causally associated with the long-term survival described here than the PFE and subsequent surgical resection. However, the potential for significant morbidity if resection is not pursued should not be underestimated; a high rate of thromboembolism has been described,14 and as such, the presence of such tumors should encourage immediate surgical treatment.
Gerbode18 described the first cardiac tumor recurrence. Since then, there have been a number of reports describing patients with multiple recurrences19,20 occurring predominately in patients with myxomas. Whereas the risk of myxoma recurrence has been approximated to be 3% for sporadic tumors and 22% for familial myxomas,21 we present an overall rate of myxoma recurrence to be 13%. There have been many hypotheses on the cause of recurrence, including multifocality, intraoperative seeding, and malignancy transformation21,22; however, there continues to be uncertainty regarding the etiology. We provide evidence that myxoma recurrence is highly associated with younger patients and not associated with the method of surgical excision. As such, tumor recurrence may be attribtuable to a patient’s biological propensity rather than the method of surgical excision, despite tumor recurrence frequently occurring at the site of previous tumor resection. Consequently, it is important for the necessary postoperative screening to occur. The hazard of myxoma recurrence increases linearly for the first 4 years after tumor resection, after which the risk of recurrence is quite low; a pattern of myxoma recurrence that has been previously described with few patients experiencing a recurrence at greater than 4 years after resection.23 Based on this finding, all patients (especially younger patients) who undergo tumor resection should be followed closely with echocardiography semiannually for the 4 years after resection with the site of previous tumor resection closely monitored. Other tumors that recurred were predominantly malignant, however, also included patients with PFEs.
Our approach to malignant tumor resection is conservative and consists of debulking without extensive cardiac resection. Despite aggressive adjuvant chemotherapy after resection, the median survival of patients with malignant tumors was less than 1 year. This finding is consistent with previous studies that documented survivals of 16.5 months24 and 9.6 months.25 Indeed, our multivariate model indicates malignant tumor histology is the most significant clinical predictor of mortality; however, we have also identified that younger patients tend to have particularly aggressive tumors. While some have advocated for highly aggressive surgical resection including cardiomyoplasty26 and autotransplantation,27 it is unclear whether either of these approaches result in superior survival outcomes. It is currently our clinical practice to pursue surgical resection when there is a likelihood of complete resection or if patients are severely symptomatic. The treatment algorithm for patients with malignant cardiac tumors is poorly defined. While there are promising clinical trials of chemotherapeutic regimens for sarcoma treatment,28–30 there continues to be no standard of care. As such, malignant primary cardiac tumors present the greatest clinical challenge to surgeons and must be evaluated on a case by case basis.
Limitations
We present the findings of a retrospective study of patients that underwent surgical resection of primary cardiac tumors spanning a 48-year time period. The limitations of retrospective cohort studies apply, especially when considering the time span of this study. Although we attempted to account for surgical era by forcing year of surgery into our prediction model, a significant limitation of this study is a failure to adequately compensate for improvements appreciated after surgical resection in later time periods as the result of improved diagnosis and surgical techniques. Additionally, as a result of the large time span of this study, we chose to use the 2002 U.S. census survival data which may have overestimated the risk of mortality after tumor resection since the median date of surgery was 1996. A major limitation in the interpretation of this data are the survival characteristics of patients who underwent resection of "other benign tumors". These tumors represent a conglomeration of tumor types that were grouped to gain statistical power. Unfortunately, there are most likely to be other confounders that impact survival which our study is not designed to detect. Thus it is important to note that these results only describe the association with long-term survival rather than causation. Despite these limitations, we believe this study provides insight into the natural history of cardiac tumor disease while highlighting the predictors of long-term outcome after surgical resection.
Conclusions
Cardiac tumors can be resected with minimal morbidity and mortality. The strongest predictors of mortality are closely linked to tumor histology, but also include clinical characteristics such as symptomatology and duration CPB time. Patients with cardiac myxomas, the most common cardiac tumor, enjoy survival that is not significantly different than that of a standard population; however, after resection, patients with myxomas should be closely followed with echocardiograms for a minimum of 4 years to evaluate for tumor recurrence; particularly in younger patients. Whereas the majority of benign tumors enjoy excellent survival, patients with malignant tumors have a dismal prognosis and younger patients with malignant tumors have particularly poor survival characteristics.
|
Acknowledgments |
|---|
Disclosures
None.
|
Footnotes |
|---|
Presented at the American Heart Association Scientific Sessions, November 4–7, 2007, Orlando, Fla.
|
References |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
1. Silverman N. Primary cardiac tumors. Annals of Surgery. 1980; 191: 127–138.[Medline] [Order article via Infotrieve]
2. Abraham KP, Reddy V, Gattuso P. Neoplasms metastatic to the heart: review of 3314 consecutive autopsies. Am J Cardiovasc Pathol. 1990; 3: 195–198.[Medline] [Order article via Infotrieve]
3. McAllister HA Jr, Hall RJ, Cooley DA. Tumors of the heart and pericardium. Curr Probl Cardiol. 1999; 24: 57–116.[Medline] [Order article via Infotrieve]
4. Shapiro LM. Cardiac tumours: diagnosis and management. Heart. 2001; 85: 218–222.
5. Larrieu AJ, Jamieson WR, Tyers GF, Burr LH, Munro AI, Miyagishima RT, Gerein AN, Allen P. Primary cardiac tumors: experience with 25 cases. J Thorac Cardiovasc Surg. 1982; 83: 339–348.[Abstract]
6. Kalbfleisch P. The Statistical Analysis of Failure Time Data (Wiley Series in Probability and Statistics) Hoboken: John Wiley & Sons; 1980.
7. Finkelstein DM, Muzikansky A, Schoenfeld DA. Comparing survival of a sample to that of a standard population. J Natl Cancer Inst. 2003; 95: 1434–1439.
8. Collett D. Modelling Survival Data in Medical Research. Boca Raton, LA: Chapman & Hall/CRC.
9. King T. On simple vascular growths in the left auricle of the heart. Lancet. 1845: 428–429.
10. Crafoord C. Discussion of: Glover RP. Late results of mitral commissurotomy. In: Henry Ford Hospital International Symposium on Cardiovascular Surgery: Studies in Physiology, Diagnosis and Techniques: Proceedings of the Symposium, Detroit, MI: W.B. Saunders; 1955:202–211.
11. Castells E, Ferran V, Octavio de Toledo MC, Calbet JM, Benito M, Fontanillas C, Granados J, Obi CL, Saura E. Cardiac myxomas: surgical treatment, long-term results and recurrence. J Cardiovasc Surg (Torino). 1993; 34: 49–53.[Medline] [Order article via Infotrieve]
12. Shroyer AL, Coombs LP, Peterson ED, Eiken MC, DeLong ER, Chen A, Ferguson TB Jr, Grover FL, Edwards FH. The Society of Thoracic Surgeons: 30-day operative mortality and morbidity risk models. Ann Thorac Surg. 2003; 75: 1856–1864;discussion 1864–5.
13. Edwards FH, Hale D, Cohen A, Thompson L, Pezzella AT, Virmani R. Primary cardiac valve tumors. Ann Thorac Surg. 1991; 52: 1127–1131.[Abstract]
14. Ngaage DL, Mullany CJ, Daly RC, Dearani JA, Edwards WD, Tazelaar HD, McGregor CG, Orszulak TA, Puga FJ, Schaff HV, Sundt TM, III, Zehr KJ. Surgical treatment of cardiac papillary fibroelastoma: a single center experience with eighty-eight patients. Ann Thorac Surg. 2005; 80: 1712–1718.
15. Klarich KW, Enriquez-Sarano M, Gura GM, Edwards WD, Tajik AJ, Seward JB. Papillary fibroelastoma: echocardiographic characteristics for diagnosis and pathologic correlation. J Am Coll Cardiol. 1997; 30: 784–790.[Abstract]
16. Cesena FH, Pereira AN, Dallan LA, Aiello VD, Mansur AJ. Papillary fibroelastoma of the mitral valve 12 years after mitral valve commissurotomy. South Med J. 1999; 92: 1023–1028.[Medline] [Order article via Infotrieve]
17. Kearney DL, Titus JL, Hawkins EP, Ott DA, Garson A Jr. Pathologic features of myocardial hamartomas causing childhood tachyarrhythmias. Circulation. 1987; 75: 705–710.
18. Gerbode F, Kerth WJ, Hill JD. Surgical management of tumors of the heart. Surgery. 1967; 61: 94–101.[Medline] [Order article via Infotrieve]
19. St John Sutton MG, Mercier LA, Giuliani ER, Lie JT. Atrial myxomas: a review of clinical experience in 40 patients. Mayo Clin Proc. 1980; 55: 371–376.[Medline] [Order article via Infotrieve]
20. Blondeau P. Primary cardiac tumors-French studies of 533 cases. Thorac Cardiovasc Surg. 1990; 38 Suppl 2: 192–195.[Medline] [Order article via Infotrieve]
21. McCarthy PM, Piehler JM, Schaff HV, Pluth JR, Orszulak TA, Vidaillet HJ Jr, Carney JA. The significance of multiple, recurrent, and "complex" cardiac myxomas. J Thorac Cardiovasc Surg. 1986; 91: 389–396.[Abstract]
22. Waller DA, Ettles DF, Saunders NR, Williams G. Recurrent cardiac myxoma: the surgical implications of two distinct groups of patients. Thorac Cardiovasc Surg. 1989; 37: 226–230.[Medline] [Order article via Infotrieve]
23. Gray IR, Williams WG. Recurring cardiac myxoma. Br Heart J. 1985; 53: 645–649.
24. Basso C, Valente M, Poletti A, Casarotto D, Thiene G. Surgical pathology of primary cardiac and pericardial tumors. Eur J Cardiothorac Surg. 1997; 12: 730–737;discussion 737–8.
25. Donsbeck AV, Ranchere D, Coindre JM, Le Gall F, Cordier JF, Loire R. Primary cardiac sarcomas: an immunohistochemical and grading study with long-term follow-up of 24 cases. Histopathology. 1999; 34: 295–304.[CrossRef][Medline] [Order article via Infotrieve]
26. Chachques JC, Argyriadis PG, Latremouille C, D'Attellis N, Fornes P, Bruneval P, Couetil JP, Carpentier AF. Cardiomyoplasty: ventricular reconstruction after tumor resection. J Thorac Cardiovasc Surg. 2002; 123: 889–894.
27. Reardon MJ, Walkes JC, Benjamin R. Therapy insight: malignant primary cardiac tumors. Nat Clin Pract Cardiovasc Med. 2006; 3: 548–553.[CrossRef][Medline] [Order article via Infotrieve]
28. Hensley ML, Maki R, Venkatraman E, Geller G, Lovegren M, Aghajanian C, Sabbatini P, Tong W, Barakat R, Spriggs DR. Gemcitabine and docetaxel in patients with unresectable leiomyosarcoma: results of a phase II trial. J Clin Oncol. 2002; 20: 2824–2831.
29. Leu KM, Ostruszka LJ, Shewach D, Zalupski M, Sondak V, Biermann JS, Lee JS, Couwlier C, Palazzolo K, Baker LH. Laboratory and clinical evidence of synergistic cytotoxicity of sequential treatment with gemcitabine followed by docetaxel in the treatment of sarcoma. J Clin Oncol. 2004; 22: 1706–1712.
30. Fata F, O'Reilly E, Ilson D, Pfister D, Leffel D, Kelsen DP, Schwartz GK, Casper ES. Paclitaxel in the treatment of patients with angiosarcoma of the scalp or face. Cancer. 1999; 86: 2034–2037.[CrossRef][Medline] [Order article via Infotrieve]
This article has been cited by other articles:
 |
|
 |
|
  C. Kiblbock, W. Schutzenberger, J. Kammler, J. Demmer, and F. Leisch Transient loss of consciousness and a heart murmur BMJ, September 2, 2009; 339(sep02_2): b3323 - b3323. [Full Text]
|
|
|
|
 |
|
 J. F. M. Marchini, A. J. Rodrigues, and A. Schmidt A ventricular thrombus mimicking a tumour BMJ Case Reports, September 1, 2009; 2009(aug31_1): bcr0620091944 - bcr0620091944. [Abstract] [Full Text]
|
|
|
|
 |
|
 G. Thiene, M. Valente, M. Lombardi, and C. Basso CHAPTER 20 Tumours of the Heart ESC Textbook of Cardiovascular Medicine, January 1, 2009; 2(1): med-9780199566990-chapter - med-9780199566990-chapter. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||