(Circulation. 2001;103:2687.)
© 2001 American Heart Association, Inc.
Clinical Investigation and Reports |
From the Cardiovascular Imaging Center, Department of Cardiology, the Cleveland Clinic Foundation, Cleveland, Ohio.
Correspondence to James D. Thomas, MD, Department of Cardiology, Desk F 15, Cleveland Clinic Foundation, 9500 Euclid Avenue, Cleveland, OH 44195. E-mail thomasj{at}ccf.org
| Abstract |
|---|
|
|
|---|
Methods and ResultsIn a 16-year period, we identified patients with CPF from our pathology and echocardiography databases. A total of 162 patients had pathologically confirmed CPF. Echocardiography was performed in 141 patients with 158 CPFs, and 48 patients had CPFs that were not visible by echocardiography (<0.2 cm), leaving an echocardiographic subgroup of 93 patients with 110 CPFs. An additional 45 patients with a presumed diagnosis of CPF were identified. The mean age of the patients was 60±16 years of age, and 46.1% were male. Echocardiographically, the mean size of the CPFs was 9±4.6 mm; 82.7% occurred on valves (aortic more than mitral), 43.6% were mobile, and 91.4% were single. During a follow-up period of 11±22 months, 23 of 26 patients with a prospective diagnosis of CPF that was confirmed by pathological examination had symptoms that could be attributable to embolization. In the group of 45 patients with a presumed diagnosis of CPF, 3 patients had symptoms that were likely due to embolization (incidence, 6.6%) during a follow-up period of 552±706 days.
ConclusionsCPFs are generally small and single, occur most often on valvular surfaces, and may be mobile, resulting in embolization. Because of the potential for embolic events, symptomatic patients, patients undergoing cardiac surgery for other lesions, and those with highly mobile and large CPFs should be considered for surgical excision.
Key Words: fibroelastoma echocardiography cardiovascular event
| Introduction |
|---|
|
|
|---|
With improved echocardiographic resolution due to higher frequency transducers and new imaging modalities, small, ill-defined valvular lesions are increasingly recognized. Clinicians must be able to decide how to manage such patients with either incidental echocardiographic findings or symptoms that may be attributable to these masses. Therefore, the purpose of this study was to (1) confirm the clinical, pathological, and echocardiographic characteristics of CPF; (2) report the risk of embolic and other complications among patients with presumed CPF; and (3) develop a management strategy for patients with presumed or a definite diagnosis of CPF.
| Methods |
|---|
|
|
|---|
Because all cases were identified retrospectively through the pathology and echocardiography databases, the term prospective refers to the presence of a lesion consistent with CPF identified by echocardiography before surgical confirmation. In this group, CPF may have been identified before or after an embolic event. The retrospective group consists of patients with CPF in whom presurgical echocardiography did not detect a lesion before surgical detection.
A total of 48 patients had CPFs confirmed by pathological examination that were not visible on echocardiograms because these CPFs were extremely small (<0.2 cm). An additional 45 patients with presumed CPF were identified by the echocardiography database and followed for symptoms that may be attributable to CPF. During the study period of >16 years, 109 502 patients had echocardiograms recorded in our database.
Study Methods
Clinical information was obtained from the patients
medical records, which included data from cardiac
catheterization and surgical and pathology reports.
Clinical events assessed included transient ischemic attacks,
stroke, myocardial infarctions, angina (typical chest pain), and
dyspnea. Transient ischemic attacks were attributed to the CPF
if patients had no significant carotid or aortic atheroma
and no atrial fibrillation or valvular heart disease.
Myocardial infarctions, angina, or both were attributed to the CPF if
the patient had no significant coronary
obstruction.
Echocardiographic studies were reviewed retrospectively from stored VHS videotapes by 2 experienced echocardiographers who were blinded to the presence and location of the tumors. The largest dimension of the tumor, its location, and the length and mobility of a stalk were measured using off-line measurement calipers on an echocardiographic machine (Hewlett Packard Sonos 1500). Associated valvular abnormalities were also assessed.
Clinical follow-up data were obtained from clinical visits, mail surveys, or telephone interviews of patients or their families. The primary clinical end points were embolic events, myocardial infarction, and death over the follow-up period.
A separate case-control study was undertaken to determine the accuracy of echocardiography for the diagnosis of CPF. This group consisted of 86 patients matched by age and valve type that included 41 control patients with no suspicion of CPF and 45 patients with pathological confirmation of CPF (from the group of 93 patients with echocardiographically detectable CPF). All patients underwent surgery and had pathological inspection of their valves. An experienced, blinded echocardiographer reviewed these studies to determine the presence and location of CPF.
Statistical Methods
Data are presented as mean±SD.
2 testing was used to compare differences
in the location and mobility of CPFs. ANOVA was used to compare the
size of the tumors and their location. Linear regression was used to
determine the size of tumors measured by
echocardiography and by pathology.
P<0.05 was considered
statistically significant.
| Results |
|---|
|
|
|---|
|
Histology
The tumor is covered by endothelium
that surrounds a layer of acid mucopolysaccharide and an inner
vascular core of connective tissue
(Figure 2
). The amount of collagen, smooth muscle cells, and
elastic fibers is variable within the connective tissue
matrix.
|
Concomitant Lesions
Of the 141 patients with proven CPFs by
echocardiographic studies, 98 (69.5%) had CPFs
associated with cardiac valvular disease; among them, 37
patients (37.8%) had rheumatic valvulitis, and 61 (62.2%) had
fibrosis and/or calcification. Twelve patients had 19 tumors in the
chambers, which were isolated lesions; 31 patients had associated
hypertrophic cardiomyopathy, aortic
aneurysm, or congenital heart disease.
Echocardiographic
Characteristics
Morphology/Appearnace
Tumors appeared round, oval, or irregular on
echocardiography but were generally well-demarcated
and homogenous in appearance. When image quality was optimal, a
"speckled appearance" with "stippling" around the perimeter
could be
seen.2
Accuracy
From the case-control study, the sensitivity and
specificity of TTE were 88.9% and 87.8%, respectively, with an
overall accuracy of 88.4% for the detection of CPF
0.2 cm. A
positive diagnosis for the presence and location of a tumor was correct
in 40 of 45 patients. A negative diagnosis was consistent with
pathology data in 36 of 41 patients. When CPF
0.2 cm were included in
the analysis, the overall sensitivity of TTE was 61.9% and
that for TEE was 76.6%.
Location
Among the 110 CPFs seen by
echocardiography, 49 (44.5%) were on the aortic
cusp (24 on the right, 6 on the left, and 19 on the noncoronary
cusp), with 40 tumors present on the aortic side of the valve and 9
on the ventricular side. Forty tumors (36.4%) were on the
mitral leaflets (23 on the anterior and 17 on the posterior leaflet),
with 32 on the left atrial surface and 8 on the left
ventricular surface
(Table 1
). Significantly more tumors occurred on the valves
than in the chambers (91 of 110, 82.7%, versus 19 of 110, 17.3%;
P<0.001).
|
Size
The size of the tumors ranged from 2 to 28 mm
(mean, 9±4.6 mm; median, 8 mm) for the largest dimension. A
total of 99% were <20 mm. The masses in the cardiac chambers
were larger than those on the aortic or mitral valves (12.0±4.6
mm versus 8.5±4.4 mm in diameter, respectively;
P<0.001). Tumors diagnosed
prospectively were larger than those diagnosed retrospectively.
Examples of CPFs are shown in
Figures 3
and 4
. The largest diameters of CPF were compared
in a random sample of 45 patients by pathology and
echocardiography with good correlation
(r=0.87,
P<0.001;
Figure 5
).
|
|
|
Stalks and Mobility
Of 110 tumors, 48 (43.6%) had a stalk from 1 to 3
mm in length, and all of these were mobile. No tumors without stalks
were mobile. All 19 CPF in the chambers were mobile, as were 29 of the
91 on a valvular surface (31.9%).
Numbers
Single lesions were detected by
echocardiography in 85 patients (91.4%). Multiple
CPFs (range, 2 to 8) were detected in 8 patients (8.6%). One patient
had 8 tumors observed in various locations on the right and left sides
of the heart.
Valvular Function
Of the 26 patients with a prospective diagnosis and
pathological confirmation, 23 had CPFs on valves with normal function.
In no patient was the CPF thought to be responsible for
valvular dysfunction. Of the 67 patients with a retrospective
diagnosis and pathological confirmation, 62 had severe valvular
disease, and all of these patients had CPFs on the dysfunctional valve.
The CPFs in this latter group were smaller and tended not to be mobile
(Table 1
).
Clinical Features
Table 2
lists the clinical features of patients with
pathologically diagnosed CPF in comparison with the group with presumed
CPF. During a follow-up period of 11±22 months, among 26 patients with
a prospective diagnosis of pathologically confirmed CPF, 23 patients
had symptoms that could be attributable to embolization. Three of 45
patients with presumed CPF had symptoms that could be attributed to
possible embolization of the tumor (incidence, 6.6%). The mass could
no longer be detected in 2 patients at follow-up, and these patients
remained asymptomatic. Four patients were lost to
follow-up.
|
Follow-Up Data
Echocardiographic follow-up data were
available for 64 of 141 patients (45.4%) after surgical excision. The
average follow-up time was 630±903 days (range, 10 to 3639 days;
median, 130 days). No mass was detected by
echocardiography in any patient during the
follow-up period.
Clinical follow-up data were available for 110 of the 141 patients (78%). The average follow-up time was 2269±893 days (range, 10 to 4176 days; median, 1221 days). Thirty-nine patients died of disease processes unrelated to CPF (36 died of congestive heart failure, 2 of cancer, and 1 of pneumonia). The remaining 71 patients were in stable condition, without symptoms related to emboli, during the follow-up period.
| Discussion |
|---|
|
|
|---|
Tumor Locations
Approximately 90% of the CPFs reported in the
literature were attached to
valves,3 4 and the
majority were on the aortic
valve.3 For the aortic valve,
no predilection for the tumor to appear on the aortic or the
ventricular side has been
reported.4 In the present
study, 49 of the 110 CPFs (44.5%) were attached to the aortic valve,
predominantly on the aortic side. This common location of the tumor
suggests a potential for dynamic coronary ostial obstruction
leading to myocardial ischemia.
The mitral valve was the next most common location of involvement in published data, with tumors occurring on the anterior or posterior leaflets, the chordae, and the papillary muscles.1 4 10 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 When an atrioventricular valve is involved, the tumor was most often on the atrial side of the valve, as found in our study.4 Along the valve leaflet, the most common site of occurrence was in the midportion, well away from the free edge or the annulus.4 There are also occasional reports of tricuspid37 38 39 40 41 and pulmonic valvular CPF,42 43 and our study confirmed this unusual side of involvement. In addition, nonvalvular sites of attachment have been reported, with left ventricular masses documented on the septum and the outflow tract.6 29 41 43 44 45 46 The masses have also been seen in the right ventricle near the papillary muscle origins.8 11 47 Tumors arising from the right atrium are described in only 3 cases.3 48 It is readily apparent that the lower rate of right-sided detection is likely due to a lack of symptoms from right-sided embolization and under-reporting due to uncommon excision of right-sided valves or entry to the right heart.
Size, Number, and Mobility
As previously described, we found that CPFs were
usually <20 mm in their largest diameter. The largest reported
CPF is 40 mm.49 The
mobility of CPFs has also been a typical feature, with 17 of 18 tumors
having independent mobility and an identifiable stalk in one
analysis.2 In our
study, all of the masses in chambers were mobile. Stalks and mobility
were present in most patients with a prospective diagnosis of CPF;
therefore, as expected, these traits may be associated with the
likelihood of embolization.
Clinical Features
The findings of our study were consistent with
prior studies with regard to the clinical profile and
presentation of patients with CPF. As in other reports that
have detected CPF in neonates and patients as old as 92
years,3 11 30 50
we found a wide distribution of age. Although CPFs are often diagnosed
incidentally, neurological
events,7 21 26 51 52
sudden
death,53 54
angina,8 acute myocardial
infarction,5 14 15 23 54
pulmonary emboli,9
and retinal artery
embolism10 related to CPF
have been reported. In the present study, CPFs were diagnosed
incidentally in many patients with another underlying
cardiovascular disease who were
asymptomatic. Among the patients who had prospectively
diagnosed CPF, underlying heart disease was uncommon, and detection
occurred related to a search for the cause of the symptoms. This fact
suggests that isolated tumors may easily "shed into the blood
stream," whereas tumors fixed by a combined lesion are not easily
shed.
The potential for suspected CPF to cause symptomatic embolization was also demonstrated in this study. After follow-up of 45 patients with the echocardiographic diagnosis of CPF, stroke occurred in one patient (48-year-old woman) who had a mobile CPF on the mitral anterior leaflet and no other cardiovascular disease. Transient ischemic attacks occurred in 2 patients: one of them a (34-year-old woman) had a mobile CPF on aortic noncoronary cups and no other valvular disease.
Diagnosis
With the increased use of 2D TTE, CPF are detected
during life and are occasionally found in patients without
symptoms.4 33
Echocardiography is a convenient and noninvasive
diagnostic technique and should be the first choice of
tests to search for
CPFs.43 52 TEE is
an important tool for delineating the extent and anatomic attachment of
these small tumors because only this technique allows optimal
high-resolution imaging. However, many CPF go undetected by
echocardiography. The reasons
echocardiography may fail to diagnose tumors
include the following: (1) the tumor was masked by an associated
lesion; (2) the tumor was too small to be seen; (3) the examination was
not done carefully with a sufficient index of suspicion; or (4) there
were no significant characteristics to differentiate the CPF from the
degenerative valve disease.
Postsurgical Follow-Up
To our knowledge, there is no information regarding
recurrence of CPF after surgical excision as detected by
echocardiography. In the present study, 64
patients (42.4%) had echocardiographic follow-up at an
average 630 days after surgery. No patients were found to have
recurrent CPF, although most studies used TTE.
Limitations
The presence of a patent foramen ovale in patients with
CPF who had an embolic event was not determined. Because paradoxical
embolism through a patent foramen ovale cannot be excluded, the
incidence of embolism due to CPF may be
overestimated.
Management
On the basis of our findings and a review of the
literature, we recommend the following guidelines for the assessment
and management of patients with CPF. Patients with events that may be
embolic in nature and are not explained by other
cardiovascular or neurological diseases should undergo
TTE and TEE if necessary to exclude cardiac sources of emboli,
including CPF. A mass seen by echocardiography
should be characterized by size, shape, location of attachment,
mobility, presence of a stalk, and multiplicity. Although the
differential diagnosis may still include vegetations (infective or
noninfective), thrombi, degenerative valve tissue, and other benign
tumors, these lesions can often be differentiated by clinical
information, blood cultures, and laboratory tests. Because the presence
of a stalk and associated mobility is a significant predictor of
embolic risk, patients with presumed CPF, especially if left-sided,
should undergo TEE to determine if a stalk is present.
Decisions regarding the primary surgical excision of CPF
depend on the size, location, mobility, and potential or strength of
association of the tumor with symptoms. Excision of isolated
right-sided CPFs is indicated only for large mobile tumors, including
those that result in obstruction or embolization that is
hemodynamically significant. The presence of a patent
foramen ovale with a sizeable right-to-left shunt is an additional
consideration for management of right-sided CPF.
Asymptomatic patients with small, left-sided, nonmobile (no
stalk) CPFs are usually observed. However, larger (
1 cm) CPF,
especially if mobile, should be considered for excision, especially if
other cardiovascular disease is detected or the patient
is young, with low risk of surgery and a high cumulative risk for
embolization. Patients with residual tumors who have had an embolic
event should similarly be considered for excision, depending on the
risks of surgery and other cardiovascular indications.
Isolated CPF excision of aortic or mitral valve lesions can often be
performed through a minimally invasive approach, with no damage to the
valve. Incidentally detected CPF in patients undergoing cardiac surgery
should generally be removed unless they add substantial time and risk
to the operation that cannot be justified based on size, location, and
mobility. No data exist to evaluate the efficacy of anticoagulation or
antiplatelet therapy for patients with CPF, although it is
speculated that deposition of thrombotic material on the tumors may add
to the risk of microembolization.
| Summary |
|---|
|
|
|---|
| Acknowledgments |
|---|
Received November 9, 2000; revision received March 19, 2001; accepted March 21, 2001.
| References |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
C. Jahnke, A. Hamdan, E. Fleck, and I. Paetsch Tissue Characterization of a Suspected Aortic Valve Fibroelastoma With Cardiac Magnetic Resonance Imaging Circ Cardiovasc Imaging, July 1, 2008; 1(1): 87 - 88. [Full Text] [PDF] |
||||
![]() |
S. Moustafa, C. Sauve, P. Page, and K. Serri Incidental finding of a papillary fibroelastoma of the mitral valve chordae Eur J Echocardiogr, May 20, 2008; (2008) jen171v1. [Abstract] [Full Text] [PDF] |
||||
![]() |
A. Basoglu, A. De Wolf, P. Dendale, M. Vanbockrijck, M. Hendrikx, and P. Vranckx Echocardiography remains the standard for the diagnostic evaluation of left ventricular tumors: A case report with anatomical correlation Eur J Echocardiogr, October 1, 2007; 8(5): 401 - 403. [Abstract] [Full Text] [PDF] |
||||
![]() |
C. A. Baker Apical Papillary Fibroelastoma Journal of Diagnostic Medical Sonography, November 1, 2006; 22(6): 382 - 386. [Abstract] [PDF] |
||||
![]() |
K. Gray, E. D. Chan, and J. Alvarado A 48-year-old woman with multiple pulmonary nodules and a right ventricular mass. Chest, October 1, 2006; 130(4): 1248 - 1251. [Full Text] [PDF] |
||||
![]() |
G. M Novaro, R. E Cabrales, and W D. Boyd Mitral Valve Papillary Fibroelastoma: Surgical Considerations Asian Cardiovasc Thorac Ann, June 1, 2006; 14(3): e58 - e59. [Abstract] [Full Text] [PDF] |
||||
![]() |
M. Daccarett, P. Burke, and S. Saba Incidental finding of a large pulmonary valve fibroelastoma: A case report Eur J Echocardiogr, June 1, 2006; 7(3): 253 - 256. [Abstract] [Full Text] [PDF] |
||||
![]() |
P. A. Vagefi, N. R. Bates, L. M. Ptaszek, M. S. Adams, and J. D. Walker Valve-sparing excision of aortic valve papillary fibroelastoma J. Thorac. Cardiovasc. Surg., May 1, 2006; 131(5): 1177 - 1178. [Full Text] [PDF] |
||||
![]() |
M. M. Ragland and T. Tak The role of echocardiography in diagnosing space-occupying lesions of the heart. Clin. Med. Res., March 1, 2006; 4(1): 22 - 32. [Abstract] [Full Text] [PDF] |
||||
![]() |
D. L. Ngaage, C. J. Mullany, R. C. Daly, J. A. Dearani, W. D. Edwards, H. D. Tazelaar, C. G.A. McGregor, T. A. Orszulak, F. J. Puga, H. V. Schaff, et al. Surgical Treatment of Cardiac Papillary Fibroelastoma: A Single Center Experience With Eighty-Eight Patients Ann. Thorac. Surg., November 1, 2005; 80(5): 1712 - 1718. [Abstract] [Full Text] [PDF] |
||||
![]() |
P. J. Sparrow, J. B. Kurian, T. R. Jones, and M. U. Sivananthan MR Imaging of Cardiac Tumors RadioGraphics, September 1, 2005; 25(5): 1255 - 1276. [Abstract] [Full Text] [PDF] |
||||
![]() |
S. Sumino and H. S. Paterson No Regrowth After Incomplete Papillary Fibroelastoma Excision Ann. Thorac. Surg., January 1, 2005; 79(1): e3 - e4. [Abstract] [Full Text] [PDF] |
||||
![]() |
L. Potenza, M. Luppi, M. Morselli, G. Riva, A. Saviola, A. Ferrari, M. De Santis, R. Rossi, and G. Torelli Cardiac Involvement in Malignancies: CASE 2. Right Ventricular Lesion As Presenting Feature of Acute Promyelocytic Leukemia J. Clin. Oncol., July 1, 2004; 22(13): 2742 - 2744. [Full Text] [PDF] |
||||
![]() |
Z. Huang, L. Sun, M. Du, Y. Ruan, and H. Wang Primary cardiac valve tumors: early and late results of surgical treatment in 10 patients Ann. Thorac. Surg., November 1, 2003; 76(5): 1609 - 1613. [Abstract] [Full Text] [PDF] |
||||
![]() |
C Basso, T Bottio, M Valente, R Bonato, D Casarotto, and G Thiene Primary cardiac valve tumours Heart, October 1, 2003; 89(10): 1259 - 1260. [Full Text] [PDF] |
||||
![]() |
E. Gologorsky and A. Gologorsky Aortic Valve Fibroelastomas as an Incidental Intraoperative Transesophageal Echocardiographic Finding Anesth. Analg., November 1, 2002; 95(5): 1198 - 1199. [Abstract] [Full Text] [PDF] |
||||
![]() |
G. D. Touati, D. Carmi, H. Sevestre, and H. Poulain Multiple aortic valve papillary fibroelastoma: do not miss the other one Eur. J. Cardiothorac. Surg., March 1, 2002; 21(3): 596 - 597. [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||