Abstract
Background
Previous studies on primary cardiac tumors were mainly based on small case series collected from a limited number of institutions. Contemporary data of patients with primary cardiac tumors treated with or without surgery in a nationwide clinical setting are limited.
Methods
Using the Diagnosis Procedure Combination database, we retrospectively identified 1317 patients hospitalized with a primary cardiac tumor (1023 myxomas, 63 non-myxomas, 72 sarcomas, 41 malignant lymphoma, 118 unspecified tumors) at 486 hospitals in Japan from July 2010 to March 2013. The outcome was overall in-hospital mortality, defined as in-hospital death occurring during the initial hospitalization or during rehospitalization. We examined the associations of baseline factors with overall in-hospital mortality and undergoing surgical resection using multivariable logistic regression analyses.
Results
Overall, 914 (69.4%) patients underwent surgery and 403 (30.6%) did not. The surgery group was younger (median age, 67 years vs. 71 years, p < 0.001) and was more likely to be treated at an academic hospital (38.9% vs. 27.8%, p < 0.001) than the no-surgery group. The surgery group also had a higher Barthel index and a higher conscious level and showed a lower frequency of extracardiac malignancies than the no-surgery group. The likelihood of undergoing surgery was associated with coexisting cerebral infarction [adjusted odds ratio (95% confidence interval), 1.96 (1.23–3.12)] and academic hospital [1.58 (1.20–2.09)]. Patients with lower Barthel index and coexisting extracardiac malignancies were less likely to undergo surgery. Overall in-hospital mortality was 2.1% and 13.4% in the surgery and non-surgery groups, respectively. Older age, lower Barthel index, lower consciousness level, coexisting metastatic extracardiac malignancy [2.95 (1.24–7.01)], and sarcoma [21.04 (8.28–53.42)] were associated with higher overall in-hospital mortality, while academic hospital [0.41 (0.20–0.84)] and surgical resection [0.39 (0.20–0.74)] were associated with lower mortality.
Conclusions
Several background factors were associated with prognosis and surgery in patients hospitalized with primary cardiac tumors.
Keywords
Introduction
Primary cardiac tumors are rare, with a prevalence ranging from 0.001% to 0.030% in various autopsy series [
1
, 2
, 3
, 4
, 5
, 6
]. Cardiac tumors can be identified according to their constitutional symptoms and cardiac complications, including systemic or pulmonary embolism, heart failure caused by circulatory obstruction, and cardiac dysfunction associated with direct tumor invasion [1
, 2
, 7
, 8
, 9
, 10
, 11
]. Cardiac tumors can also be found incidentally in imaging studies such as echocardiography, computed tomography, and magnetic resonance imaging [8
, 12
, 13
, 14
].- Beroukhim R.S.
- Prakash A.
- Buechel E.R.V.
- Cava J.R.
- Dorfman A.L.
- Festa P.
- Hiavacek A.M.
- Johnson T.R.
- Keller M.S.
- Krishnamurthy R.
- Misra N.
- Moniotte S.
- Parks W.J.
- Powell A.J.
- Soriano B.D.
- et al.
Characterization of cardiac tumors in children by cardiovascular magnetic resonance imaging: a multicenter experience.
J Am Coll Cardiol. 2011; 58: 1044-1054
The Japanese annual national statistics of the Japanese Association for Thoracic Surgery (JATS) have shown that the number of surgically resected cardiac tumors has been increasing [
15
, 16
, - Masuda M.
- Masuda M.
- Kuwano H.
- Okumura M.
- Arai H.
- Endo S.
- Doki Y.
- Kobayashi J.
- Motomura N.
- Nishida H.
- Saiki Y.
- Tanaka F.
- Tanemoto K.
- Toh Y.
- Yokomise H.
Thoracic and cardiovascular surgery in Japan during 2013: annual report by The Japanese Association for Thoracic Surgery.
Gen Thorac Cardiovasc Surg. 2015; 63: 670-701
17
, 18
, 19
]. Possible reasons for this increase are greater awareness among physicians and wider availability of imaging modalities for cardiac tumors than in the past. However, previous studies investigating patient characteristics and outcomes of cardiac tumors have been largely restricted to single-center case series [6
, 11
, 20
, 21
, 22
, 23
] or multicenter studies with limited study populations or small sample sizes [14
, - Beroukhim R.S.
- Prakash A.
- Buechel E.R.V.
- Cava J.R.
- Dorfman A.L.
- Festa P.
- Hiavacek A.M.
- Johnson T.R.
- Keller M.S.
- Krishnamurthy R.
- Misra N.
- Moniotte S.
- Parks W.J.
- Powell A.J.
- Soriano B.D.
- et al.
Characterization of cardiac tumors in children by cardiovascular magnetic resonance imaging: a multicenter experience.
J Am Coll Cardiol. 2011; 58: 1044-1054
24
, 25
, - Neuville A.
- Collin F.
- Bruneval P.
- Parrens M.
- Thivolet F.
- Gomez-Brouchet A.
- Terrier P.
- de Montpreville V.T.
- Le Gall F.
- Hostein I.
- Lagarde P.
- Chibon F.
- Coindre J.M.
Intimal sarcoma is the most frequent primary cardiac sarcoma: clinicopathologic and molecular retrospective analysis of 100 primary cardiac sarcomas.
Am J Surg Pathol. 2014; 38: 461-469
26
]. In addition, patient data in most of the previous studies were collected for study periods of >10 years. Large contemporary multicenter studies of patients with primary cardiac tumors who were treated with or without surgery are lacking.- Padalino M.A.
- Vida V.L.
- Boccuzzo G.
- Tonello M.
- Sarris G.E.
- Berggren H.
- Comas J.V.
- Di Carlo D.
- Di Donato R.M.
- Ebels T.
- Hraska V.
- Jacobs J.
- Gaynor J.W.
- Metras D.
- Pretre R.
- et al.
Surgery for primary cardiac tumors in children: early and late results in a multicenter European Congenital Heart Surgeons Association (ECHSA) study.
Circulation. 2012; 126: 22-30
When a cardiac tumor is highly suspected, surgical resection is the mainstay of treatment for both benign and malignant cardiac tumors [
1
, 2
, 7
, 8
, 9
, 10
]. Most benign cardiac tumors are curable with complete resection [2
, 8
, 10
, 13
], while malignant cardiac tumors are less likely to be completely resected and recur more frequently than benign tumors even after complete resection [20
, 21
, 22
, 23
, 24
, 27
]. Although small benign tumors may be managed nonsurgically with echocardiographic follow-up, a recent observational study reported an increased risk of cerebrovascular complications and mortality in patients with suspected papillary fibroelastomas who did not undergo surgical resection [[28]
]. Meanwhile, patients in poor general condition or with significant comorbidities may be nonsurgically treated irrespective of the type of cardiac tumor because of perioperative risk. No previous study has examined factors associated with the likelihood of undergoing surgical resection in patients with primary cardiac tumors. Although the prognosis of benign tumors is generally good, malignant tumors, especially cardiac sarcomas, have a rapidly fatal course even after surgery [20
, 21
, 22
, 23
, 24
, 27
]. It remains unclear which patient factors are associated with poor in-hospital prognosis in patients with primary cardiac tumors. The objective of the present study was therefore to investigate factors associated with in-hospital mortality and the likelihood of undergoing surgical resection in patients hospitalized with primary cardiac tumors, using data from a large-scale inpatient database.Methods
Study design and data source
This was a nationwide retrospective cohort study using the Diagnosis Procedure Combination (DPC) database, which has been described in detail previously [
29
, 30
]. The Institutional Review Board of The University of Tokyo approved the study and waived the requirement for patient informed consent because all data were anonymized. The DPC database included approximately 18.3 million inpatient data from approximately 1000 Japanese hospitals for 33 consecutive months between 1 July 2010 and 31 March 2013, representing approximately 50% of all inpatient admissions in Japan [29
, 30
].The DPC database includes the following data: patient age and sex; diagnoses, comorbidities, and complications recorded with Japanese text and the International Classification of Diseases, 10th Revision (ICD-10) codes; status of activities of daily living, which can be used for calculation of the Barthel index [
[31]
]; consciousness level at admission measured with the Japan Coma Scale [- Sanchis J.
- Bonanad C.
- Ruiz V.
- Fernandez J.
- Garcia-Blas S.
- Mainar L.
- Ventura S.
- Rodriguez-Borja E.
- Chorro F.J.
- Hermenegildo C.
- Bertomeu-Gonzalez V.
- Nunez E.
- Nunez J.
Frailty and other geriatric conditions for risk stratification of older patients with acute coronary syndrome.
Am Heart J. 2014; 168: 784-791
[30]
]; drugs and devices; surgical and nonsurgical procedures; length of hospital stay; discharge status (dead or alive); and unique hospital identifiers.Patient selection
From the DPC database, we selected patients with a confirmed diagnosis of a primary cardiac tumor using ICD-10 codes (Supplementary Table 1). Subtypes of benign cardiac tumors (myxoma, rhabdomyoma, fibroma, papillary fibroelastoma, schwannoma, and lipoma) and malignant cardiac tumors (angiosarcoma, rhabdomyosarcoma, liposarcoma, fibrosarcoma, myxosarcoma, synovial sarcoma, and malignant lymphoma) were identified from the Japanese-text diagnoses in the DPC database. For cardiac myxomas, the tumor location (left atrium, right atrium, left ventricle, or right ventricle) was also identified according to the Japanese-text diagnosis. Non-myxomas included rhabdomyomas, fibromas, papillary fibroelastomas, schwannomas, and lipomas. Sarcoma included angiosarcoma, rhabdomyosarcoma, liposarcoma, fibrosarcoma, myxosarcoma, and synovial sarcoma. A primary cardiac tumor recorded without a specific Japanese-text diagnosis was defined as ‘unspecified type of tumor’ in the present study. Eligible patients with primary cardiac tumors were classified into patients who underwent surgical treatment for cardiac tumors (surgery group) and those who did not (no-surgery group). In the present study, surgical treatment for cardiac tumors included surgical resection of cardiac tumors with or without coronary artery bypass grafting or valve surgery (valvuloplasty or valve replacement).
Baseline variables
Baseline variables included the following: patient age and sex; Barthel index at admission; Japan Coma Scale at admission; comorbidities already present at admission (primary extracardiac malignancy, metastatic extracardiac malignancy, cerebral infarction, congestive heart failure, pulmonary embolism, chronic pulmonary disease, chronic renal failure, and chronic liver disease); and type of hospital (academic or non-academic) (ICD-10 codes for comorbidities are shown in Supplementary Table 1).
Outcome measures
The measured outcome was overall in-hospital mortality, which was defined as in-hospital death occurring during the initial hospitalization or during rehospitalization. For patients who underwent surgery, we also identified postoperative mortality, which was defined as in-hospital death occurring during the hospitalization in which surgical treatment was performed.
Statistical analysis
Categorical variables are presented as numbers and proportions and were compared using the Chi-square test. Continuous variables are presented as median and interquartile range (IQR) and were compared using the Mann–Whitney U test or Kruskal–Wallis test. We examined factors associated with likelihood of undergoing surgical resection of cardiac tumors in a multivariable logistic regression model with adjustment for the following variables: age, sex, Barthel index at admission, Japan Coma Scale at admission, comorbidities at admission, and type of hospital. We also examined factors associated with in-hospital death in a multivariable logistic regression model with adjustment for the following variables: age, sex, Barthel index at admission, Japan Coma Scale at admission, comorbidities at admission, type of hospital, type of cardiac tumor (myxoma, non-myxoma, sarcoma, malignant lymphoma, or unspecified tumor), and performance of surgical resection.
All hypothesis tests had a two-sided significance level of 0.05. SPSS version 22 (IBM Corp., Armonk, NY, USA) was used for all statistical analyses.
Results
Study population
We identified 1317 patients (1753 hospitalizations) with a primary cardiac tumor who were hospitalized at 486 acute-care hospitals in Japan. The median age was 68 years (IQR, 58–76 years). There were 1023 (77.7%) patients with a myxoma, 63 (4.8%) with a non-myxoma, 72 (5.5%) with a sarcoma, 41 (3.1%) with malignant lymphoma, and 118 (9.0%) with an unspecified type of tumor (Table 1). Of the eligible patients with a primary cardiac tumor, 914 (69.4%) underwent surgical resection of the tumor and 403 (30.6%) did not.
Table 1Frequencies of each type of primary cardiac tumor.
| n (%) | No. of surgically treated patients in each tumor group (%) | |
|---|---|---|
| Total no. of patients | 1317 (100.0) | 914 (69.4) |
| Benign cardiac tumor | 1086 (82.5) | 901 (83.0) |
| Myxoma | 1023 (77.7) | 773 (75.6) |
| Site of myxoma | ||
| LA | 759 (57.6) | 595 (78.4) |
| RA | 74 (5.6) | 48 (64.9) |
| LV | 14 (1.1) | 8 (57.1) |
| RV | 11 (0.8) | 5 (45.5) |
| Unspecified site | 165 (12.5) | 117 (70.9) |
| Non-myxoma | 63 (4.8) | 28 (44.4) |
| Rhabdomyoma | 27 (2.1) | 1 (3.7) |
| Fibroma | 22 (1.7) | 15 (68.2) |
| Papillary fibroelastoma | 8 (0.6) | 8 (100.0) |
| Lipoma | 5 (0.4) | 3 (60.0) |
| Schwannoma | 1 (0.1) | 1 (100.0) |
| Malignant cardiac tumor | 113 (8.6) | 29 (25.7) |
| Sarcoma | 72 (5.5) | 23 (31.9) |
| Angiosarcoma | 53 (4.0) | 14 (26.4) |
| Rhabdomyosarcoma | 7 (0.5) | 2 (28.6) |
| Fibrosarcoma | 6 (0.5) | 3 (50.0) |
| Myxosarcoma | 3 (0.2) | 3 (100.0) |
| Liposarcoma | 2 (0.2) | 1 (50.0) |
| Synovial sarcoma | 1 (0.1) | 0 (0.0) |
| Malignant lymphoma | 41 (3.1) | 6 (14.6) |
| Unspecified type of tumor | 118 (9.0) | 84 (71.2) |
Numbers are shown as n (%). LA, left atrium; RA, right atrium; LV, left ventricle; RV, right ventricle.
Baseline characteristics, treatment, and outcomes according to tumor type
Table 2 shows the clinical characteristics according to type of primary cardiac tumor. Age and sex distributions differed according to type of tumor. Patients with non-myxomas were the youngest among the five tumor groups and accounted for the majority of patients aged <10 years. All identified malignant tumors in patients aged <50 years were sarcomas, while the proportion with malignant lymphoma among older patients with a malignant tumor increased with age. Patients with myxomas were more often female than patients with other tumors. The Barthel index at admission was lower in patients with sarcomas and malignant lymphoma than in those with myxomas. The Japan Coma Scale at admission was similar across the groups. Patients with malignant lymphoma showed the highest frequency of primary extracardiac malignancy and congestive heart failure at admission, while patients with sarcomas showed the highest frequency of metastatic extracardiac malignancy at admission. Patients with myxomas showed the highest frequency of cerebral infarction at admission, followed by those with non-myxomas. Among the five tumor groups, the proportion having surgical resection was highest in patients with myxomas. Overall in-hospital mortality was highest in patients with sarcomas, followed by those with malignant lymphoma. Postoperative mortality was also highest in patients with sarcomas. No patient with a non-myxoma died during hospitalization.
Table 2Baseline characteristics, surgical treatment, and in-hospital outcomes according to type of primary cardiac tumor.
| Myxoma (n = 1023) | Non-myxoma (n = 63) | Sarcoma (n = 72) | Malignant lymphoma (n = 41) | Unspecified type of tumor (n = 118) | p-value | |
|---|---|---|---|---|---|---|
| Age, years, median (IQR) | 69 (60–77) | 45 (0–67) | 54.5 (38.5–63.5) | 77 (70–83) | 64 (49–73) | <0.001 |
| 0–9 | 2 (0.2) | 26 (41.3) | 3 (4.2) | 0 (0.0) | 6 (5.1) | |
| 10–19 | 6 (0.6) | 3 (4.8) | 1 (1.4) | 0 (0.0) | 5 (4.2) | |
| 20–29 | 13 (1.3) | 1 (1.6) | 6 (8.3) | 0 (0.0) | 0 (0.0) | |
| 30–39 | 17 (1.7) | 1 (1.6) | 11 (15.3) | 0 (0.0) | 5 (4.2) | |
| 40–49 | 69 (6.7) | 2 (3.2) | 12 (16.7) | 0 (0.0) | 14 (11.9) | |
| 50–59 | 124 (12.1) | 5 (7.9) | 11 (15.3) | 1 (2.4) | 17 (14.4) | |
| 60–69 | 282 (27.6) | 11 (17.5) | 18 (25.0) | 8 (19.5) | 32 (27.1) | |
| 70–79 | 341 (33.3) | 8 (12.7) | 7 (9.7) | 17 (41.5) | 28 (23.7) | |
| 80–89 | 149 (14.6) | 6 (9.5) | 3 (4.2) | 11 (26.8) | 11 (9.3) | |
| ≥90 | 20 (2.0) | 0 (0.0) | 0 (0.0) | 4 (9.8) | 0 (0.0) | |
| Sex: male | 410 (40.1) | 36 (57.1) | 41 (56.9) | 20 (48.8) | 58 (49.2) | 0.002 |
| Barthel Index at admission | ||||||
| 100 | 701 (68.5) | 26 (41.3) | 32 (44.4) | 14 (34.1) | 93 (78.8) | <0.001 |
| 50–95 | 98 (9.6) | 5 (7.9) | 15 (20.8) | 8 (19.5) | 7 (5.9) | |
| 0–45 | 121 (11.8) | 3 (4.8) | 13 (18.1) | 15 (36.6) | 9 (7.6) | |
| Missing data | 103 (10.1) | 29 (46.0) | 12 (16.7) | 4 (9.8) | 9 (7.6) | |
| Japan Coma Scale at admission | ||||||
| 0 (alert) | 968 (94.6) | 59 (93.7) | 66 (91.7) | 39 (95.1) | 117 (99.2) | 0.319 |
| 1–3 (drowsy) | 36 (3.5) | 2 (3.2) | 2 (2.8) | 1 (2.4) | 0 (0.0) | |
| 10–30 (somnolent) | 13 (1.3) | 1 (1.6) | 3 (4.2) | 0 (0.0) | 1 (0.8) | |
| 100–300 (comatose) | 6 (0.6) | 1 (1.6) | 1 (1.4) | 1 (2.4) | 0 (0.0) | |
| Comorbidity at admission | ||||||
| Primary extracardiac malignancy | 87 (8.5) | 1 (1.6) | 5 (6.9) | 6 (14.6) | 4 (3.4) | 0.041 |
| Metastatic extracardiac malignancy | 10 (1.0) | 0 (0.0) | 29 (40.3) | 4 (9.8) | 3 (2.5) | <0.001 |
| Cerebral infarction | 120 (11.7) | 6 (9.5) | 2 (2.8) | 1 (2.4) | 10 (8.5) | 0.048 |
| Congestive heart failure | 278 (27.2) | 9 (14.3) | 13 (18.1) | 16 (39.0) | 26 (22.0) | 0.016 |
| Pulmonary embolism | 8 (0.8) | 0 (0.0) | 1 (1.4) | 0 (0.0) | 1 (0.8) | 0.879 |
| Chronic pulmonary disease | 52 (5.1) | 4 (6.3) | 1 (1.4) | 1 (2.4) | 3 (2.5) | 0.377 |
| Chronic renal failure | 20 (2.0) | 4 (6.3) | 1 (1.4) | 1 (2.4) | 8 (6.8) | 0.008 |
| Chronic liver disease | 29 (2.8) | 1 (1.6) | 0 (0.0) | 0 (0.0) | 6 (5.1) | 0.201 |
| Type of hospital: academic | 348 (34.0) | 26 (41.3) | 28 (38.9) | 12 (29.3) | 54 (45.8) | 0.079 |
| Surgical resection of cardiac tumors | 773 (75.6) | 28 (44.4) | 23 (31.9) | 6 (14.6) | 84 (71.2) | <0.001 |
| Plus CABG | 87 (8.5) | 5 (7.9) | 3 (4.2) | 1 (2.4) | 19 (16.1) | 0.017 |
| Plus valve surgery | 57 (5.6) | 3 (4.8) | 3 (4.2) | 0 (0.0) | 9 (7.6) | 0.440 |
| Overall in-hospital mortality | 32 (3.1) | 0 (0.0) | 28 (38.9) | 8 (19.5) | 5 (4.2) | <0.001 |
| Postoperative mortality | 10/773 (1.3) | 0/28 (0.0) | 2/23 (8.7) | 0/6 (0.0) | 2/84 (2.4) | 0.060 |
Numbers are n (%) unless otherwise indicated. IQR, interquartile range; CABG, coronary artery bypass grafting.
a Overall in-hospital mortality was defined as in-hospital death occurring during initial hospitalization or during rehospitalization.
b Postoperative mortality was defined as in-hospital death occurring during the hospitalization when surgical treatment was performed.
Surgery group vs. no-surgery group
Table 3 shows a comparison between the surgery group and no-surgery group. Patients in the surgery group were significantly younger, had a significantly higher Barthel index and a significantly lower Japan Coma Scale at admission, and showed a significantly lower frequency of primary and metastatic extracardiac malignancies than those in the no-surgery group. Patients in the surgery group were more likely to be treated at an academic hospital than at a non-academic hospital. Overall in-hospital mortality was significantly lower in the surgery group than in the non-surgery group.
Table 3Comparison between patients treated with and without surgical resection.
| Surgery group (n = 914) | No-surgery group (n = 403) | p-value | |
|---|---|---|---|
| Age, years, median (IQR) | 67 (58–74) | 71 (58–80) | <0.001 |
| Sex: male | 380 (41.6) | 185 (45.9) | 0.143 |
| Barthel Index at admission | |||
| 100 | 680 (74.4) | 186 (46.2) | <0.001 |
| 50–95 | 81 (8.9) | 52 (12.9) | |
| 0–45 | 76 (8.3) | 85 (21.1) | |
| Missing data | 77 (8.4) | 80 (19.9) | |
| Japan Coma Scale at admission | |||
| 0 (alert) | 884 (96.7) | 365 (90.6) | <0.001 |
| 1–3 (drowsy) | 20 (2.2) | 21 (5.2) | |
| 10–30 (somnolent) | 7 (0.8) | 11 (2.7) | |
| 100–300 (comatose) | 3 (0.3) | 6 (1.5) | |
| Comorbidity at admission | |||
| Primary extracardiac malignancy | 49 (5.4) | 54 (13.4) | <0.001 |
| Metastatic extracardiac malignancy | 9 (1.0) | 37 (9.2) | <0.001 |
| Cerebral infarction | 103 (11.3) | 36 (8.9) | 0.204 |
| Congestive heart failure | 251 (27.5) | 91 (22.6) | 0.063 |
| Pulmonary embolism | 6 (0.7) | 4 (1.0) | 0.517 |
| Chronic pulmonary disease | 42 (4.6) | 19 (4.7) | 0.924 |
| Chronic renal failure | 23 (2.5) | 11 (2.7) | 0.822 |
| Chronic liver disease | 27 (3.0) | 9 (2.2) | 0.460 |
| Type of hospital: academic | 356 (38.9) | 112 (27.8) | <0.001 |
| Overall in-hospital mortality | 19 (2.1) | 54 (13.4) | <0.001 |
Numbers are n (%) unless otherwise indicated.
Factors associated with the likelihood of undergoing surgery
In the multivariable logistic regression analysis, lower Barthel index and coexisting primary or metastatic extracardiac malignancy were significantly associated with a lower likelihood of undergoing surgical resection of a tumor (Table 4). Coexisting cerebral infarction and academic hospital were significantly associated with a higher likelihood of undergoing surgical resection of a tumor.
Table 4Multivariable logistic regression analysis of likelihood of undergoing surgery.
| Odds ratio | 95% CI | p-value | |
|---|---|---|---|
| Age, by 10-year increase | 1.01 | 0.94–1.08 | 0.796 |
| Female sex (reference: male sex) | 1.21 | 0.93–1.58 | 0.157 |
| Barthel Index at admission | |||
| 100 | Reference | ||
| 50–95 | 0.41 | 0.27–0.61 | <0.001 |
| 0–45 | 0.22 | 0.15–0.33 | <0.001 |
| Missing data | 0.22 | 0.15–0.33 | <0.001 |
| Japan Coma Scale at admission | |||
| 0 (alert) | Reference | ||
| 1–3 (drowsy) | 0.70 | 0.35–1.41 | 0.315 |
| 10–30 (somnolent) | 0.48 | 0.16–1.42 | 0.187 |
| 100–300 (comatose) | 0.34 | 0.08–1.44 | 0.143 |
| Comorbidity at admission | |||
| Primary extracardiac malignancy | 0.29 | 0.19–0.46 | <0.001 |
| Metastatic extracardiac malignancy | 0.13 | 0.06–0.29 | <0.001 |
| Cerebral infarction | 1.96 | 1.23–3.12 | 0.005 |
| Congestive heart failure | 1.35 | 1.00–1.84 | 0.052 |
| Pulmonary embolism | 1.17 | 0.30–4.58 | 0.825 |
| Chronic pulmonary disease | 0.79 | 0.43–1.44 | 0.436 |
| Chronic renal failure | 0.95 | 0.42–2.11 | 0.891 |
| Chronic liver disease | 1.23 | 0.52–2.95 | 0.635 |
| Type of hospital | |||
| Non-academic hospital | Reference | ||
| Academic hospital | 1.58 | 1.20–2.09 | 0.001 |
CI, confidence interval.
Factors associated with overall in-hospital mortality
In the multivariable logistic regression analysis, older age, lower Barthel index, higher Japan Coma Scale, coexisting metastatic extracardiac malignancy, and sarcoma were significantly associated with higher overall in-hospital mortality (Table 5). Academic hospital and surgical resection of a tumor were significantly associated with lower overall in-hospital mortality.
Table 5Multivariable logistic regression analysis of overall in-hospital mortality.
| Odds ratio | 95% CI | p-value | |
|---|---|---|---|
| Age, by 10-year increase | 1.42 | 1.14–1.76 | 0.001 |
| Female sex (reference: male sex) | 0.72 | 0.40–1.30 | 0.273 |
| Barthel Index at admission | |||
| 100 | Reference | ||
| 50–95 | 1.94 | 0.81–4.64 | 0.138 |
| 0–45 | 3.94 | 1.80–8.59 | 0.001 |
| Missing data | 3.02 | 1.29–7.07 | 0.011 |
| Japan Coma Scale at admission | |||
| 0 (alert) | Reference | ||
| 1–3 (drowsy) | 0.79 | 0.23–2.80 | 0.719 |
| 10–30 (somnolent) | 1.44 | 0.27–7.58 | 0.669 |
| 100–300 (comatose) | 8.70 | 1.39–54.3 | 0.021 |
| Comorbidity at admission | |||
| Primary extracardiac malignancy | 1.45 | 0.59–3.60 | 0.421 |
| Metastatic extracardiac malignancy | 2.95 | 1.24–7.01 | 0.014 |
| Cerebral infarction | 0.79 | 0.26–2.37 | 0.677 |
| Congestive heart failure | 1.18 | 0.63–2.22 | 0.603 |
| Pulmonary embolism | 1.19 | 0.09–16.04 | 0.897 |
| Chronic pulmonary disease | 0.89 | 0.19–4.22 | 0.883 |
| Chronic renal failure | 1.16 | 0.24–5.68 | 0.855 |
| Chronic liver disease | 0.90 | 0.10–8.36 | 0.925 |
| Type of hospital | |||
| Non-academic hospital | Reference | ||
| Academic hospital | 0.41 | 0.20–0.84 | 0.016 |
| Type of tumor | |||
| Myxoma | Reference | ||
| Non-myxoma | – | – | – |
| Sarcoma | 21.04 | 8.28–53.42 | <0.001 |
| Malignant lymphoma | 2.17 | 0.81–5.80 | 0.124 |
| Unspecified tumor | 2.29 | 0.79–6.62 | 0.127 |
| Surgical resection of tumors | |||
| No | Reference | ||
| Yes | 0.39 | 0.20–0.74 | 0.004 |
Discussion
This nationwide cohort study found that several factors were associated with in-hospital prognosis and undergoing surgical tumor resection in patients hospitalized with a primary cardiac tumor. A lower Barthel index and coexisting malignancy at admission were significantly associated with a lower likelihood of undergoing surgical treatment, while coexisting cerebral infarction and an academic hospital were significantly associated with a higher likelihood of undergoing surgical treatment. A lower Barthel index, coexisting metastatic extracardiac malignancy, and sarcoma were significantly associated with higher overall in-hospital mortality, while an academic hospital and surgery were significantly associated with lower overall in-hospital mortality.
Comparison with previous studies
Previous studies on primary cardiac tumors have been limited because of the rarity of this type of tumor. Since 1997, the JATS has conducted annual surveys in Japan and has published the national statistics of thoracic surgery every year [
16
, - Masuda M.
- Masuda M.
- Kuwano H.
- Okumura M.
- Arai H.
- Endo S.
- Doki Y.
- Kobayashi J.
- Motomura N.
- Nishida H.
- Saiki Y.
- Tanaka F.
- Tanemoto K.
- Toh Y.
- Yokomise H.
Thoracic and cardiovascular surgery in Japan during 2013: annual report by The Japanese Association for Thoracic Surgery.
Gen Thorac Cardiovasc Surg. 2015; 63: 670-701
17
, 18
, 19
]. The JATS survey covers almost all cases of cardiac surgery in Japan and includes data on 30-day mortality and hospital mortality according to the diseases and operative procedures. These JATS survey data are useful for physicians and surgeons to identify the number and risk of each cardiac surgery in Japan. However, previous reports on cardiac tumors, including the JATS surveys, focused on patients treated with surgery and did not include data about patients treated without surgery. Thus, we performed this retrospective cohort study to investigate factors associated with mortality and the likelihood of undergoing surgical resection in patients hospitalized with primary cardiac tumors by using a dataset including both surgically and non-surgically treated patients.Previous studies with relatively large sample sizes (n = 124–323) reported that 82–94% of cardiac tumors were benign and 6–18% were malignant [
6
, 7
, 11
, 15
], which were consistent with the present study. The proportion of myxomas among primary cardiac tumors was 78–89% in the present study and previous Japanese studies [7
, 15
] compared with 42–50% in previous American studies [6
, 11
]. This difference may be attributable to racial differences between American and Japanese patients. In this study, patients with non-myxoma were younger and had a lower comorbidity burden than patients with myxoma, probably because non-myxoma is more frequently found in healthy pediatric or adolescent patients than in patients with other types of cardiac tumor. Our results also showed that the proportion with malignant lymphoma among patients with malignant cardiac tumors increased as patient age increased.Baseline factors and surgical treatment
When a cardiac tumor is suspected in a clinical setting, consideration of the indications for surgical resection is important [
1
, 2
]. However, surgical resection of a cardiac tumor is often performed after cardiovascular complications occur. Our multivariable analysis showed that coexisting cerebral infarction at admission was significantly associated with a higher likelihood of undergoing surgical treatment. This may support the notion that patients with cerebral embolism of tumors are more likely to undergo surgery promptly to prevent additional embolism. Our analysis also showed that several comorbidities and the status of activities of daily living affected the choice of treatment in patients with primary cardiac tumors. Patients with metastatic malignancy or a poor activity level were less likely to undergo surgery, possibly because the risk of surgery was deemed to outweigh the potential benefit. We believe that our results support the following strategy in patients with cardiac tumors: surgical resection should be performed (1) before cardiovascular complications occur, and (2) before metastasis occurs in patients with suspected malignant tumors. Although it is often difficult to identify cardiac tumors before the onset of symptoms associated with the tumors, our results may suggest that prompt resection of tumors should be considered once a cardiac tumor is clinically identified. However, in patients of older age or who have a poor activity level, the benefits of prophylactic surgical resection should be weighed against its operative risks.Surgical treatment and mortality of malignant cardiac tumors
In the present study, only 31.9% of patients with a cardiac sarcoma underwent surgical resection. One possible reason for this low proportion is that patients with a sarcoma had advanced-stage tumors in the present study. Approximately 40% of patients with a sarcoma had metastatic extracardiac malignancy. Patients with sarcoma had a higher proportion of congestive heart failure in the present study than in a previous study (18.1% [13/72] vs. 2.9% [1/34]) [
[20]
]. These high comorbidity burdens may have precluded attending physicians from performing surgical resection in patients with a cardiac sarcoma. The reason for the low proportion of surgical resection (14.6%) in patients with malignant lymphoma could be that chemotherapy is generally the first-line therapy for cardiac malignant lymphoma, unlike other cardiac tumors [2
, 32
, 33
].Treatment at academic hospitals
The present study showed that patients with primary cardiac tumors were more likely to undergo surgery if they were admitted to an academic hospital than a non-academic hospital. This was likely related to the fact that academic hospitals receive many patients referred for surgery by non-academic hospitals. This study also showed that treatment at an academic hospital was significantly associated with better overall in-hospital prognosis than treatment at a non-academic hospital, even after multivariable adjustment, including performance of surgery. This association indicates that treatment at hospitals with more experienced staff may improve in-hospital prognosis of patients with primary cardiac tumors. Thus, our results may suggest the need for centralization of cases with primary cardiac tumors to improve prognosis.
Study limitations
The present study had several limitations. First, data regarding the reasons for undergoing or not undergoing surgery were unavailable in the database. Second, the DPC database does not include morphological or anatomical tumor data apart from the location of cardiac myxomas. These unmeasured factors may have affected the choice of treatment. Third, diagnoses of cardiac tumors recorded in retrospective studies are generally less well validated than those in prospective studies. In particular, the diagnoses of cardiac tumors in nonsurgically treated patients were less accurate than those in surgically treated patients because the diagnoses were presumably based on clinical studies (e.g. diagnostic imaging) rather than histological examination. Furthermore, diagnoses of comorbidities in the present study were also less well validated than those in prospective studies because the DPC database does not have clear diagnostic criteria for recording comorbidities, similar to other administrative database studies [
34
, - Deshmukh A.
- Patel N.J.
- Pant S.
- Shah N.
- Chothani A.
- Mehta K.
- Grover P.
- Singh V.
- Vallurupalli S.
- Savani G.T.
- Badheka A.
- Tuliani T.
- Dabhadkar K.
- Dibu G.
- Reddy Y.M.
- et al.
In-hospital complications associated with catheter ablation of atrial fibrillation in the United States between 2000 and 2010: analysis of 93 801 procedures.
Circulation. 2013; 128: 2104-2112
35
]. Fourth, the present study was not able to evaluate the prognosis of patients out of hospital after discharge because the DPC database only includes in-hospital data. Fifth, in surgically treated patients with malignant tumors, information on the extent of the surgical resection (from partial resection to complete resection with microscopically negative margins) was not included in the DPC database. Finally, although our study population was relatively large, population representativeness was limited because participation in the DPC study is voluntary.Conclusions
This retrospective cohort study using a large-scale national database found that several background factors were associated with treatment choice and in-hospital prognosis in patients with primary cardiac tumors. Patients with a lower Barthel index or a coexisting primary or metastatic extracardiac malignancy were less likely to undergo surgery, while those with coexisting cerebral infarction and admission to an academic hospital were more likely to undergo surgery. Older patients with a lower consciousness level, a poor activity level, coexisting metastatic extracardiac malignancy, or sarcoma had higher overall in-hospital mortality, while those treated at an academic hospital or who underwent surgical resection had lower overall in-hospital mortality.
Funding
This work was supported by grants from the Ministry of Health, Labour and Welfare, Japan (Grant Numbers: H26-Policy-011 and H26-Special-047). The sponsors had no roles in the study.
Conflict of interest
The authors declare that there is no conflict of interest.
Appendix A. Supplementary data
The following are the supplementary data to this article:
- Supplementary Table 1
Diagnoses.
References
- Cardiac tumours: diagnosis and management.Heart. 2011; 97: 151-160
- Cardiac tumours: diagnosis and management.Lancet Oncol. 2005; 6: 219-228
- Tumors of the heart. A 20-year experience with a review of 12,485 consecutive autopsies.Arch Pathol Lab Med. 1993; 117: 1027-1031
- Frequency of primary tumors of the heart.Am J Cardiol. 1996; 77: 107
- Primary cardiac tumors.Ann Surg. 1980; 191: 127-138
- Primary cardiac tumors: experience at the University of Minnesota.Thorac Cardiovasc Surg. 1990; 38: 183-191
- Characteristics of cardiac myxoma with constitutional signs: a multicenter study in Japan.Clin Cardiol. 2002; 25: 367-370
- Primary cardiac tumors.Int J Cardiol. 2009; 133: 152-156
- Primary cardiac sarcoma.Eur J Cardiothorac Surg. 2006; 29: 925-932
- Cardiac myxomas.N Engl J Med. 1995; 333: 1610-1617
- Survival after resection of primary cardiac tumors: a 48-year experience.Circulation. 2008; 118: S7-S15
- Cardiac MRI is complementary to echocardiography in the assessment of cardiac masses.Echocardiography. 2007; 24: 286-300
- Cardiac tumours: an update: cardiac tumours.Heart. 2008; 94: 117-123
- Characterization of cardiac tumors in children by cardiovascular magnetic resonance imaging: a multicenter experience.J Am Coll Cardiol. 2011; 58: 1044-1054
- Clinical classification of cardiovascular tumors and tumor-like lesions, and its incidences.Gen Thorac Cardiovasc Surg. 2013; 61: 435-447
- Thoracic and cardiovascular surgery in Japan during 2013: annual report by The Japanese Association for Thoracic Surgery.Gen Thorac Cardiovasc Surg. 2015; 63: 670-701
- Thoracic and cardiovascular surgery in Japan during 2012.Gen Thorac Cardiovasc Surg. 2014; 62: 734-764
- Thoracic and cardiovascular surgery in Japan during 2011: annual report by The Japanese Association for Thoracic Surgery.Gen Thorac Cardiovasc Surg. 2013; 61: 578-607
- Thoracic and cardiovascular surgery in Japan during 2010.Gen Thorac Cardiovasc Surg. 2012; 60: 680-708
- Malignant primary cardiac tumors: review of a single institution experience.Cancer. 2008; 112: 2440-2446
- Primary cardiac sarcoma: 25-year Cleveland clinic experience.Am J Clin Oncol. 2014; ([Epub ahead of print])
- Prognostic analysis for survival after resections of localized primary cardiac sarcomas: a single-institution experience.Ann Thorac Surg. 2014; 97: 1379-1385
- Surgical treatment of primary cardiac tumors: 28 years’ experience in Kanazawa University Hospital.Jpn Circ J. 2001; 65: 315-319
- Primary cardiac sarcomas: a retrospective study of the French Sarcoma Group.Eur J Cancer. 2014; 50: 128-136
- Intimal sarcoma is the most frequent primary cardiac sarcoma: clinicopathologic and molecular retrospective analysis of 100 primary cardiac sarcomas.Am J Surg Pathol. 2014; 38: 461-469
- Surgery for primary cardiac tumors in children: early and late results in a multicenter European Congenital Heart Surgeons Association (ECHSA) study.Circulation. 2012; 126: 22-30
- Primary cardiac sarcomas: an immunohistochemical and grading study with long-term follow-up of 24 cases.Histopathology. 1999; 34: 295-304
- Prognostic and bioepidemiologic implications of papillary fibroelastomas.J Am Coll Cardiol. 2015; 65: 2420-2429
- Relationship between hospital volume and major cardiac complications of rotational atherectomy: a nationwide retrospective cohort study in Japan.J Cardiol. 2016; 67: 442-448
- Out-of-hospital versus in-hospital Takotsubo cardiomyopathy: analysis of 3719 patients in the Diagnosis Procedure Combination database in Japan.Int J Cardiol. 2014; 176: 413-417
- Frailty and other geriatric conditions for risk stratification of older patients with acute coronary syndrome.Am Heart J. 2014; 168: 784-791
- Primary cardiac lymphoma.Int J Cardiol. 2011; 149: 358-363
- Clinical perspectives of primary cardiac lymphoma.Angiology. 2003; 54: 599-604
- In-hospital complications associated with catheter ablation of atrial fibrillation in the United States between 2000 and 2010: analysis of 93 801 procedures.Circulation. 2013; 128: 2104-2112
- Trends in hospitalizations for adults with congenital heart disease in the U.S..J Am Coll Cardiol. 2009; 54: 460-467
Article info
Publication history
Published online: June 21, 2016
Accepted:
May 16,
2016
Received in revised form:
April 29,
2016
Received:
December 25,
2015
Identification
Copyright
© 2016 Japanese College of Cardiology. Published by Elsevier Ltd. All rights reserved.
User license
Elsevier user license | How you can reuse 
Elsevier's open access license policy
Elsevier user license
Permitted
For non-commercial purposes:
- Read, print & download
- Text & data mine
- Translate the article
Not Permitted
- Reuse portions or extracts from the article in other works
- Redistribute or republish the final article
- Sell or re-use for commercial purposes
Elsevier's open access license policy
