Another predictor of biologically unusual myxomas

J THORAC

CARDIOVASC SURG

1989;98: 1083-6

Deoxyribonucleic acid ploidy pattern of cardiac myxomas Another predictor of biologically unusual myxomas A group of patients with cardiac myxoma who have a heritable syndrome involving skin myxomas, endocrine tumors, and lentiginosis-the complex of myxomas, spotty pigmentation, and endocrine overactivity-has been described previously. Patients with the complex had cardiac myxomas at an early age (average, 26 years) with frequent multiple myxomas (53%) and recurrent cardiac myxomas (22%); however, no histologic differences were noted when these tumors were compared with sporadic cardiac myxomas. In the present study, deoxyribonucleic acid flow cytometric analyses of 35 cardiac myxoma specimens were correlated with clinical findings (mean duration of foUow-up, 13 years). Among 30 patients with sporadic (nonfamilial) cardiac myxoma, 24 (80%) had a normal (deoxyribonucleic acid diploid) ploidy pattern, and six (20%) had an abnormal (deoxyribonucleic acid tetraploid) pattern. Specimens from each of the five patients with the complex had abnormal deoxyribonucleic acid tetraploid patterns (p = 0.002 compared with the sporadic myxoma group). Further, aU four patients who had recurrent cardiac myxoma had an abnormal deoxyribonucleic acid ploidy pattern (p = 0.007 compared with patients with nonrecurrent myxomas). Unlike conventional histologic examination, the ploidy pattern of cardiac myxomas seems to be sensitive for detecting biologicaUy unusual tumors, and a deoxyribonucleic acid tetraploid pattern suggests a high risk of recurrence.

Patrick M. McCarthy, MD, Hartzell V. Schaff, MD, Harry Z. Winkler, MD, Michael M. Lieber, MD, and J. Aidan Carney, MD, Rochester. Minn.

Cardiac myxomas are rare tumors that occasionally pursue an aggressive clinical course. The most common cardiac myxoma, typically a single left atrial tumor in an elderly woman, has been called a "sporadic" myxoma, and it has a low risk for recurrence. However, recently we identified subsets of patients who are at high risk for multicentric and recurrent tumors. They include patients with familial occurrence of cardiac myxomas who tend to have myxomas at an early age and in unusual locations. I In addition, some patients with cardiac' myxomas have a

From the Division of Thoracic and Cardiovascular Surgery. the Department of Urology, and the Division of Pathology. Mayo Clinic and Mayo Foundation. Rochester. Minn. Received for publication July 29. 1988. Accepted for publication March 2. 1989. Address for reprints: Hartzell V. Schaff. MD. Mayo Clinic. 200 First St., SW, Rochester, MN 55905.

12/1/13097

heritable syndrome-the complex of myxomas, spotty pigmentation, and endocrine overactivity-': Cardiac myxomas Cutaneous myxomas Myxoid fibroadenoma of the breast Lentiginosis (spotty pigmentation) Pituitary tumors Cushing's syndrome (primary pigmented nodular adrenocortical disease) Testicular tumor (Sertoli-Leydig cell tumor)

Cardiac myxomas in patients with the complex also have an atypical biologic behavior with early development of myxomas, atypical locations, multiple tumors, and a high rate of recurrent myxomas. Rarely, patients with otherwise sporadic myxomas (not familial, not part of the complex) have recurrent myxomas, and these patients are usually younger." Because of the sometimes aggressive course of recurrent myxomas and case reports of subsequent growth of embolic fragments, cardiac myxomas have been labeled "malignant.">" However, with the exception of true myxosarcomas, there are no clearly recog-

1083

The Journal of

1 084

Thoracic and Cardiovascular Surgery

McCarthy et al.

2,000

2,000

DNA diploid

1,500

Cells, no.

Cells, no.

1,000

1,000

500

500

o

DNA tetraploid

1,500

2C

4C

2C

8C

4C

8C

Relative DNA count

Relative DNA count Fig. 1. DNA histogram of normal cardiac myxoma.

Fig. 2. DNA histogram of cardiac myxomas with more than 10% nuclei in the 4C peak (DNA tetraploid).

nized histologic differences between these tumors and sporadic myxomas. Most of these "malignant" myxomas occur in nonsporadic myxomas.t" Conventional histologic evaluation has not been useful for differentiating sporadic from nonsporadic myxomas or for predicting risk for recurrence; however, deoxyribonucleic acid (DNA) analysis of cardiac myxomas has not been studied previously. Therefore, the purpose of the present study was to analyze the nuclear DNA pattern from specimens of cardiac myxoma and to correlate the findings with the clinical course in an effort to determine whether tumor recurrence might be predicted,

stained with propidium iodide by using the method of Vindel<;iv, Christensen, and Nissen.!" The sample was filtered to provide single nuclei and to eliminate nuclear clumps. Samples were run on the flow cytometer within 30 minutes after the addition of propidium iodide. Specific details of the method used have been described.'! Nuclear DNA content was measured on a FACS IV (fluorescence-activated cell sorter) flow cytometer (Becton Dickinson, Sunnyvale, Calif.) equipped with a 5 W argon ion laser run at a wavelength of 514 nm. Histograms of 20,000 nuclei for each sample were recorded at a maximal scanning flow rate of 1000 nuclei per second. Cell-cycle evaluation of the DNA histogram and the coefficient of variation of the diploid Go/G I peak derived by flow cytometry were obtained by means of a computer program for Dean and Jett mathematical analysis.'? The flow cytometricdata were compared statistically by the adjusted x 2 test for the two-by-two contingency table. For quantitating the number of nuclei normally found in the nontumor 4C or G 2 peak, 15 different normal endocardial biopsy specimens were studied. Nuclei extracted from these 15 formalin-fixed and paraffin-embedded samples showed that the mean percentage ± standard deviation of nuclei in the 4C or G 2 peak was 5.72% ± 1.42%. On the basis of these normal tissue control data, an upper limit of 10% of nuclei in the 4C peak (3 standard deviations above the mean) was considered normal. Tumors containing a significant increase in G2 (4C) peak (those having more than 10% nuclei in the 4C peak) were categorized as DNA tetraploid. Samples with a definite 8C peak on the DNA histogram, but without a 6C peak, were also classified as DNA tetraploid. Tumor histograms displaying fewer than 10% nuclei in the G 2 (4C) peak were designated as DNA diploid. Histograms representing the number of nuclei and their DNA content were constructed; they were classified as DNA diploid or DNA tetraploid without knowledge of the patient's clinical characteristics.

Methods Patients. DNA flow cytometry was performed on portions of cardiac myxoma specimens from 35 Mayo Clinic patients. Patient records were reviewed with particular attention to the presence or absence of components of the complex of myxomas, spotty pigmentation, and endocrine overactivity. A family history of cardiac myxoma or of the complex, tumor focus (unifocal or rnultifocal), and results of treatment were recorded, and current follow-up data were available for all patients. Five patients with clinical findings of the complex have been described by us previously." Thirty patients had apparently sporadic myxomas, and these patients were chosen from our earlier surgical group so that long-term follow-up data (mean, 13 years) were available. Subclinical recurrences could not be excluded in all patients. Of the 30 patients with sporadic myxomas, one had a recurrent cardiac myxoma (the only recurrent sporadic myxoma in our experience). Flow cytometric technique. Nuclear suspensions from paraffin-embedded tissue blocks were prepared by using the technique of Hedley and associates.? Briefly, the procedure consisted of dewaxing three 40 ,urn thick sections and then washing the tissue and suspending it in 0.5% pepsin. The resulting nuclear suspension was centrifuged to form a nuclear pellet, and the pepsin supernatant was removed. The isolated nuclei were

Results A total of 35 paraffin-embedded specimens of cardiac myxoma were available for processing and DNA ploidy

Volume 98

DNA ploidy pattern of cardiac myxomas

Number 6 December 1989

1085

Table I. Distribution of nuclear DNA ploidy patterns in 35 patients with cardiac myxomas DNA histogram pal/ern Normal DNA diploid Myxoma

No. of patients

Complex

5 .JQ

Sporadic

No.

No.

%

o

o

24 24

35

Abnormal DNA tetraploid

80 69

%

...Q

100· 20

II

31

5

'p = 0.002 compared with the sporadic group.

Table II. DNA ploidy and tumor recurrence in 35 patients with cardiac myxomas DNA histogram pal/ern Normal DNA diploid Recurrence No recurrence

Abnormal DNA tetraploid"

No. of patients

No.

%

No.

%

4

0

lOOt

30 34

24

0 80 71

4

...Q 10

20 29

24

'One patient was omitted (autopsy specimen). 0.007 compared with group with normal pattern.

'» =

analysis by flow cytometry. High-quality DNA histograms were obtained. No patients exhibited a DNA aneuploid pattern. Of the 30 sporadic cardiac myxomas, 24 (80%) exhibited a normal DNA diploid pattern that closely resembled the DNA histograms observed for normal endocardial biopsysamples (Fig. I), and six specimens (20%) showed a significant increase (more than 10% of the total nuclei) in the4C peak and were designated DNA tetraploid (Fig. 2). The mean G2 peak percentage was 4.5% ± 3.1% for the 24 DNA diploid tumors and 13.4% ± 3.2% for the six DNA tetraploid tumors. All five patients with complex cardiac myxomas had abnormal DNA tetraploid patterns. The mean G 2 peak was 14.7% ± 2.1%. The difference in DNA ploidy distribution between the sporadic and complex cardiac myxoma groups was highly significant (p =:' 0.(02) (Table I). During a mean follow-up of 13 years, recurrent tumor developed in only one (3%) of the 30 patients with sporadic cardiac myxomas but in 75% ofthe patients with complexcardiac myxoma. (In one of the fivepatients with the complex, the cardiac myxoma was found at autopsy; three [75%] of the four surgically resected tumors "recurred.") The difference in incidence of recurrence between the cardiac myxoma groups was highly significant (p < 0.001, log-rank test). The relationship between DNA ploidy and the biologic behavior of the tumor is presented in Table II. None of the tumors with a normal DNA ploidy pattern recurred.

In contrast, 40% of the tumors exhibiting abnormal DNA patterns recurred. This difference was statistically significant (p = 0.(07). In fact, all tumors that recurred showed abnormal DNA tetraploid patterns (p = 0.007 compared with the nonrecurrent tumors).

Discussion The controversy regarding the origin of cardiac myxomas, whether they represent true neoplasms or originate from intracardiac thrombi, 13 has been largely laid to rest. Recent identification of distinct clinical patterns associated with cardiac myxomas.l" as well as the finding of chromosome abnormalities in cardiac myxomas, 14firmly established the neoplastic nature of these tumors. Indeed, there is evidence that the clinical syndrome that involves cardiac myxomas has a pattern of mendelian dominant inheritance.' The findings of this study provide additional evidence that cardiac myxomas are neoplastic and also suggest that the clinical behavior of the cardiac myxoma may be predicted by the DNA ploidy histogram. All cardiac myxomas occurring as part of the complex of myxomas, spotty pigmentation, and endocrine overactivity and all the recurrent cardiac myxomas had abnormal DNA ploidy patterns. Only six (20%) of the patients with clinically sporadic cardiac myxomas had DNA tetraploid patterns, and one of these patients had a recurrent myxoma. In this small sample, the DNA ploidy pattern seems to be very sensitive (lOO%) for identifying patients with clinically

The Journal of

1 086

McCarthy et al.

unusual cardiac myxomas. An abnormal ploidy pattern is not a certain predictor of tumor recurrence, complex cardiac myxoma, or malignant clinical behavior (specificity, 80%); however, no previously described pathologic criteria have been able to predict recurrence. The therapeutic implications of this and the previous studies are as follows: (I) Patients with cardiac myxoma who are at high risk for recurrence of the tumor are those with familial myxomas, those with features of the complex described by Carney and associates.!" those with multiple (synchronous) myxomas, and those with abnormal DNA ploidy patterns; (2) close follow-up, including echocardiography, of patients at high risk is important; and (3) first-degree relatives of patients with complex cardiac myxoma should be screened, as described previously;' for cardiac myxoma and other tumors of the syndrome. REFERENCES 1. Carney JA. Differences between nonfami1ial and familial cardiac myxoma. Am J Surg Pathol 1985;9:53-5. 2. Carney JA, Gordon H, Carpenter PC, Shenoy BV, Go VL W. The complex of myxomas, spotty pigmentation, and endocrine overactivity. Medicine (Baltimore) 1985;64:27083. 3. Carney JA, Hruska LS, Beauchamp GD, Gordon H. Dominant inheritance of the complex of myxomas, spotty pigmentation, and endocrine overactivity. Mayo Clin Proc 1986;61: 165-72. 4. McCarthy PM, Piehler JM, Schaff HV, et al. The significance of multiple, recurrent, and "complex" cardiac myxomas. J THORAC CARDIOVASC SURG 1986;91:389-96.

Thoracic and Cardiovascular Surgery

5. Read RC, White HJ, Murphy ML, Williams D, Sun eN, Flanagan WHo The malignant potentiality of left atrial myxoma. J THORAC CARDIOYASC SURG 1974;68:857-67. 6. Attum AA, Johnson GS, Masri Z, Girardet R, Lansing AM. Malignant clinical behavior of cardiac myxomas and "myxoid imitators." Ann Thorac Surg 1987;44:217-22. 7. DeSousaAL, Muller J, Campbell RL, BatnitzkyS, Rankin L. Atrial myxoma: a review of the neurological complieations, metastases, and recurrences. J Neurol Neurosurg Psychiatry 1978;41: 1119-24. 8. Budzilovich G, Aleksic S, Greco A, Fernandez J, HarrisJ, Finegold M. Malignant cardiac myxoma with cerebral metastases. Surg N eurol 1979;11:461-9. 9. Hedley DW, Friedlander ML, Taylor IW, Rugg CA, Musgrove EA. Method for analysis of cellular DNA content of paraffin-embedded pathological material using flow cytometry. J Histochem Cytochem 1983;31:1333-5. 10. Vindelev LL, ChristcnsenLl, Nissen Nl. A detergenttrypsin method for the preparation of nuclei for flow cytometric DNA analysis. Cytometry 1983;3:323-7. 11. Winkler HZ, Rainwater LM, Myers RP, et al. Stage Dl prostatic adenocarcinoma: significance of nuclear DNA ploidy patterns studied by flow cytometry. Mayo Clin Proc 1988;63: 103-12. 12. Dean PN, Jett JH. Mathematical analysis of DNA distributions derived from flow microfluorometry. J Cell BioI 1974;60:523-7. 13. Fisher ER, Hellstrom HR. Evidence in support of the neoplastic nature of cardiac myxoma. Am HeartJ I 960;60:63()' 40. 14. Dewald GW, Dahl RJ, Spurbeck JL, Carney JA, Gordon H. Chromosomally abnormal clones and nonrandom telomeric translocations in cardiac myxomas. Mayo Clin Proc 1987;62:558-67.