- Email: [email protected]
The nature of the cardiac myxoma
1ntcrnationrlJournrl
of
cardiology ELSEVIER
InternationalJournal of Cardiology 57 (1996) 211-216
The nature of the cardiac myxoma S.K. Suvarnaa’*, J.A. Roydsb ‘Department of Histopathology, Northern General Hospital, Herries Road, Sheffield S5 7AU, UK bDepa@ment of Pathology, University of Shefield, Beech Hill Road, Shefield SlO ZRX, UK
Received26 June 1996;accepted4 September1996
Abstract Ten archive cases of cardiac myxoma were evaluated for proliferative activity, metastatic potential and expression of oncogene/tumour suppressorgene products by meansof PCNA, h4IB1, nm23, ~53, Bcl-2 and Rb-1 immunohistochemistry. The myxomas showed variable proliferative activity (PCNA O-41%, average 12.6%, MIBl O-138, average 3.2%) contrasting with the absenceof mitotic activity histologically. All the myxomas showed nm23 staining. None showed p53 reactivity. Eight caseswere negative for Bcl-2 expression,with two casesgiving weak cytoplasmic staining. Rb-1 reactivity showed a variable pattern (staining indices O-868) paralleling the cases’proliferativeactivity. The cardiacmyxomais interpreted as a weakly proliferative lesion with little metastatic potential and no modulation of oncogene/oncogene
suppressorproducts.Whilst not excluding a neoplasticaetiology, the results are consideredmore in keeping with a reactive/hamartomatous process. Keywords: cardiac myxoma; Immunohistochemistry; Oncogene;Proliferation index; Tumour
1. Intruduction Cardiac myxomas comprise 50% of all cardiac tumours [1,2], occurring sporadically and within familial syndromes (Camey’s complex, NAME syndrome, LAM3 syndrome) [3-51. The histological and ultrastructural features are well known 16-81. However, controversy regarding the histogenesis of the myxoma has existed for most of this century. Although initially considered to be an exuberant thrombus, the neoplastic interpretation is now favoured [g-12]. Given that neoplastic lesions may show high proliferative indices, metastatic activity and modulation of oncogeneI oncogene suppressor *Correspondingauthor.
products [ 131, we chose to evaluate the cardiac myxoma with antibodies appropriate to these features. Ki-67 is a nuclear antigen, recognised by MIB 1, that is presentin proliferating cells and absent in quiescent cells [14,15]. Assessmentof cell proliferation by Ki-67 detection has been shown to correlate with tumour grade and prognosis in neoplastic cells. PCNA (a 36 kDa protein subunit of DNA polymerase delta) is one of the late response genesduring mitogenic stimulation, DNA replication and repair [16]. It has a long half-life, being found in cycling and non-cycling cells. ~53 missense mutations lead to stabilisation and visualisation of ~53 by immunohistochemistry [17]. These are the most common genetic abnormalities in human cancer.Wild-type ~53 expression and stabili-
0167-5273/96/$15.00 Copyright 0 1996ElsevierScienceIrelandLtd. AU rights reserved PII
SO167-5273(96)02827-6
supprtxsorgene;Hmcnna
S.K. Suvama, J.A. Royds I International
212
sation is also increased in responseto DNA damage resulting in cell cycle arrest, repair or apoptosis [18]. The product of the retinoblastoma gene Rb-1 (pRb) forms part of the cell cycle control [19,20]. pRb is necessary for cell cycle progression and mitosis, and its function is also implicated in the maintenanceof differentiated cell type. Expression of Bcl-2 is associatedwith rescuefrom apoptosis, particularly in proliferating cells and is a mechanism contributing to malignancy, especially in low grade non-Hodgkin’s lymphomas [21,22]. Loss of nm23, especially in association with high cell turnover, generally correlates with the metastatic potential of many tumours [23-251.
2. Methods
All caseswere examined by routine haematoxylin/ eosin and toluidine blue histochemistry. Five p,m sections were stained with the following antibodies according to manufacturers recommendations or as previously described with modifications as listed
Journal of Cardiology 57 (1996) 211-216
count on the section or up to 200 cells, with results being expressedas a percentageof the total. 3. Results
Ten cases of formalin-fixed, paraffin-embedded cardiac myxoma were retrieved from the archive files. Eight were resected surgically, one was identified within an arterial embolus and one case was diagnosed at autopsy in a case of metastatic breast carcinoma. Review of the case records indicated no one with evidence of the Carney complex. All the casesshowed classic featuresof the cardiac myxoma (Fig. 1) without mitotic activity or cellular atypia. However, the autopsy case showed the cardiac myxoma arising in direct proximity to a focus of metastatic breast carcinoma. Despite the absenceof mitotic activity there was immunohistochemical evidence of proliferative ac-
[25,26].
PCNA, monoclonal PC10 (Novocastra Ltd), 1:200, 1 h room temperature. MB31 monoclonal (The Binding Site Ltd, UK), l:lOO, 4°C overnight. nm23, immunopurified rabbit antibody raised against synthetic peptide 11 (this was a generous gift from Dr P Steeg). ~53, monoclonal DO7 (Novocastra Ltd), 1:200, 1 h room temperature. bcl-2, monoclonal (Dako Ltd) 1:200, 20 h 4°C. Rb-1 monoclonal G3-349 (Pharmingen Ltd, Cambridge Bioscience, UK) 1:150, 4°C overnight. The sections were treated with the relevant ABC kit (Vector Labs, UK), with antigen visualisation by diaminobenzidine. Antigen retrieval was achieved by microwaving sections in 0.01 M sodium citrate buffer for 10 min at 650 W prior to primary antibody for Bcl-2, MIB 1, ~53 and Rb 1. Positive and negative controls were included. The proliferation indices involved the total cell
Fig. 1. The surface of a myxoma showing abundant myxoid matrix, scattered stellate/spindle cells and a characteristic vascular channel.
S.K. Suvama, J.A. Royds I lntemational
Fig. 2. PCNA-stained section showing some positively nuclei (arrowed). Other cell nuclei are negative.
stained
tivity. The mean proliferation index of MIBl was 3.2% (range O-13%) and PCNA 12.6% (range O41%) (Fig. 2). It was noted that those turnouts with no/very low proliferation scores were identical in both assays. Anti-nm23 antibody revealed uniform positive staining, although variable in reaction intensity (5 strong, 4 moderate, 1 weak). Contrastingly, p53 showed no reaction product in any myxoma (Fig. 3 and Fig. 4). Eight cases revealed no Bcl-2 staining, although two showed borderline cytoplasmic positivity. This latter feature was present in all cells, contrasting with the strong positivity of the stromal lymphocytes. Rb-1 staining indices varied between 0 and 86% (average 19.2%). In general, the positive scores paralleled the proliferation activity scores seen with MIBl and PCNA. The autopsy sample showed focal carcinoma with loss of nm23 staining, patchy reactivity with Rb-1 and no staining with ~53 or Bcl-2. The carcinoma
Journal of Cardiology 57 (1996) 211-216
213
Fig. 3. p53-stained section of a myxoma showing no immunohistochemical reaction product, contrasting with the control section (Fig. 4).
showed proliferative activity (seen with PCNA and MIBl) but the tissue available was too small for formal quantification. The individual results are expressed in group format (Table 1). No relationship was identified in relation to the age, sex or size of the myxoma. 4. Discussion Proof of the myxomas histogenesis has been elusive since its identification. The thrombotic theory appearsto have been discredited. The majority view of neoplasia is favoured although a hamartomatous derivation has been considered. Attempts to resolve the histogenesis question have addressedthe ultrastructural and immunohistochemical phenotype of the myxoma and have demonstratedvariable phenotypic differentiation (endothelial, fibroblastic, muscle and epithelial patterns) [6,10,11,27]. This at first
S.K. Suvama, J.A. Royds I International
214
Fig. 4. A colonic tubular adenoma showing positive p53 nuclear staining in the dysplastic glandular epithelium (arrowed).
sight is confusing and derivation from primitive multipotential mesenchymal cells has been invoked as the explanation. In addition, recurrence following incomplete surgery [28,29], embolisation being mistaken for
Journal of Cardiology 57 (1996) 211-216
tumour metastasisand misdiagnosis of cardiac sarcomas showing myxoid features [ 1,2] must be borne in mind when considering the nature of the cardiac myxoma and to avoid ascribing the myxoma with features that are not appropriate. Some studies have shown myxoma cell cultures to have chromosome aberrations [30,31]. However, this is at variance with previously published ploidy data and contrasts [32,33] with other investigators’ difficulty in establishing cell cultures. The results of the present study throw light on the nature of the myxoma. It is noted that none of the myxomas showed ~53 reactivity. Although caution is required regarding uncritical appraisal of ~53 staining [ 171, the absence of such a common gene mutational event does not support a role for neoplasia in the genesis of the cardiac myxoma. Similar uniformity is seen with all the tumours demonstrating nm23 reactivity. This contrasts with the metastatic carcinoma, seen in the autopsy case which had focal nm23 loss. Consequently it is inferred that the cardiac myxoma has little potential for metastasis. High nm23 expression was most likely to be accompanied by moderate or high expressionof cell cycle related genes(Ki-67, PCNA/ pRb). Low nm23 expressionwas uniform and associated with low proliferation and was therefore not likely to signify any metastatic potential. The cellular proliferation markers (MIBl and PCNA) demonstratea wide range of reactivity. It is noted that the proliferation reactivity, where positive,
Table 1 Individual results Case 1 2 3 4 5 6 I 8 9 10
Age/sex 25/F 39/M 30/F 36/F 55/M 51/M 56/F 55/M 49/M 51/F
nm23
ff+ ++ + +++ +++ ++ ++ +++ +++ ++
P53
-
%MIB 1 3 0 0.5 13 0.5 0 2 6 1 0
%FCNA 5 0 0.5 39 20 0.5 5 41 8 0.5
All numerical values expressed as percentage of total cells. Staining grades: -I + I + + I + + + =negative/weak/moderate/strong. Source/origin of the pathological material: A, autopsy; S, surgery; E, embolectomy, Site: RA, right atrium; LA, left atrium; AE, arterial embolus.
%Rb-1
Bcl-2
Source
Site
1 0 8 19 0 19 3 86 76 0
-
S S S S S S E S S A
RA LA LA LA LA LA AE LA LA LA
+ + -
S.K. Suvarna, J.A. Royds I International
showed a similar range of results for both markers indicating the validity of the observations. However, careful scrutiny of the myxoma cases failed to indicate a reason for the range of results. Specifically the myxomas were assessed for the degree of cellularity, amount of myxoid matrix, degree of haemorrhage and amount of blood vessels. Whilst there was fluctuation of thesecomponentparts within the myxomas studied no association was identified with the proliferation markers. Moreover, it is stressedthat no mitotic figures or nuclear atypia was seen histologically. Therefore, we are left with the conclusion that the myxomas do show inherent variation in proliferation activity, perhaps reflecting the stage of growth of the lesion studied. Since the time from initiation to sampling cannot be determined this interpretation cannot be validated. Rb-1 staining broadly paralleled the proliferation activity and was similarly shown to have no relationship to vascular&y, myxoid matrix, cellularity or the degree of haemoirhage. Furthermore, since no other tissue metaplasia (glandular/muscular differentiation) was identified the Rb-1 staining does not parallel any morphological parameter within the myxomas. Bcl-2 is associated with rescue from apoptosis, particularly in proliferating cells and its aberrant expressionis a recognisedmechanismcontributing to malignancy. It is noted that the majority of the myxomas showed no reactivity for this immunostain. That two casesdid show reactivity is intriguing since both these cases are seen to be ‘proliferating’ myxomas. As with the proliferation markers (MIBl and PCNA) it is possible that the reactivity identified representsa stage in myxoma evolution/growth. Taken together, these results infer the cardiac myxoma to be a lesion of low proliferative activity without evidence of mutation/stabilisation of tumour suppressor or oncogene products (~53 and Bcl-2 respectively). The nm23 staining pattern indicates a tumour with low metastatic potential. Rb-1 staining is open to debatebut may simply reflect proliferative activity. Whilst not excluding a neoplastic interpretation, we consider the results to be more in keeping with a benign/reactive process. The possibility of a reactive phenomenon [34] or a hamartomatousprocess cannot be similarly excluded [35]. Further research is indicated into this intriguing entity.
Journal of Cardiology 57 (19%) 211-216
215
Acknowledgments
We wish to thank Dr. P Steegfor the generousgift of the nm23 antibody. We also thank Mr. K. Corke for technical assistance.J. A. Royds is supported by the Yorkshire Cancer ResearchCampaign.
References Ul McAllister HA, Fenoglio JJ. Tumors of the cardiovascular system. Washington DC: Armed Forces Institute of Pathology, 1996; 21-46. I21Rosai .J. Ackerman’s Surgical Pathology, 7th ed. St. Louis: CV Mosby, 1996; 2182-2184. [31 Carney JA. Differences between nonfamilial and familial cardiac myxoma. Am J Surg Path01 1985; 9: 53-55. [41 McCarthy PM, Piehler JM, Schaff HV, Pluth JR, Orszulak TA,Vidaillet HJ et al. The significance of multiple, recurrent, and ‘complex’ cardiac myxomas. J Thorac Cardiovasc Surg 1986; 91: 389-396. I51Carney JA, Hruska LS, Beauchamp GD, Gordon H. Dominant inheritance of the complex of myxomas, spotty pigmentation and endocrine overactivity. Mayo Clin Proc 1986; 61: 165-172. 161Ferrans VJ, Roberts WC. Structural features of cardiac myxomas. Hum Path01 1973; 4: 111-146. [71 Feldman PS, Horvath E, Kovaks K. An ultrastructural study of 7 cardiac myxomas. Cancer 1977; 40: 2216-2232. PI Wold LE, Lie JT. Scanning electron microscopy of intracardiac myxoma. Mayo Clin Proc 1981; 56: 198-200. [91 Lie JT. The identity and histogenesis of cardiac myxomas. Arch Path01 Lab Med 1989; 113: 724-726. IlO1Landon G, Ordonez NG, Guarda LA. Cardiac myxomas. An immunohistochemical study using endothelial, histiocytic and smooth muscle markers, Arch Path01 Lab Med 1986; 110: 116-120. 1111Krikler DM, Rode J, Davies MJ, Woolf N, Moss E. Atria1 myxoma: a tumour in search of its origins. Br Heart J 1992; 67: 89-91. WI Salyer WR, Page DL, Hutchins GM. The development of cardiac myxomas and papillary endocardial lesions from mural thrombus. Am Heart J 1975; 89: 4-17. [131 Cotran RS, Kumar V, Robbins SL. Robbins Pathologic Basis of Disease, 4th ed. Philadelphia: WB Saunders, 1989; 239305. u41 Gerdes J, Li L, Schlueter C, Duchrow M, Wohlenberg C, Gerlach C et al. Immunobiochemical and molecular biologic characterisation of the cell proliferation-associated nuclear that is defined by monoclonal antibody Ki67. Am J Path01 1991; 138: 867-873. r151 McCormick D, Yu C, Hobbs C, Hall PA. The relevance of antibody concentration to me immunohistochemical quantification of cell proliferation associated antigens. Histopathology 1993; 22: 543-547.
216
S.K. Suvamu, J.A. Royds I International
[16] McCormick D, Hall PA. The complexities of proliferating cell nuclear antigen. Histopathology 1992; 21: 591-594. [17] Wynford-Thomas D. p53 in tumour pathology: can we trust immunocytochemistry? J Path01 1992; 166: 329-330. [18] Camnan CE, Kastan MB. Induction of apoptosis by tumour suppressor genes and oncogenes. Cancer Biol 1995; 6: 1725. [19] Royds JA, Webley K, Karcioglu ZA, Parsons MA. Apoptosis-related gene expression in retinoblastomas. Br J Cancer, 1996; 44: 73s. [20] Jacks T, Fazeli A, Schmitt AM, Bronson RT, Goode11 MA, Weinberg RA. Effects of an Rb mutation in the mouse. Nature 1992; 359: 295-300. [21] Tsujimoto Y, Ikegaki N, Croce CM. Characterisation of the protein product of bcl-2, the gene involved in human follicular lymphoma. Oncogene 1987; 2: 3-7. [22] Nathan B, Anbazhagan R, Dyer M, Ebbs SR, Jayatilake H, Gusterson BA. Expression of bcl-2-like reactivity in the normal breast and in breast cancer. The Breast 1992; 2: 134-137. [23] Leone A, Flatow U, VanHoutte K, Steeg PS. Transfection of human nm23-Hl into the human MDA-MB-435 breast carcinoma cell line: effects on the tumor metastatic potential, colonisation and enzymic activity. Oncogene 1993; 8: 23252333. [24] Steeg PS, Bevilacqua G, Kopper L, Thorgeirsson UP, Talmadge JE, Liotta LA, Sobel ME. Evidence for a novel gene associated with low metastatic potential. J Nat1 Cancer Inst 1988; 80: 200-204. [25] Royds JA, Stephenson TJ, Rees RC, Shorthouse AJ, Silcocks PB. mn23 product expression in ductal in situ and invasive human breast carcinoma. J Natl Cancer Inst 1993; 85: 727731.
Jownal of Cardiology 57 (1996) 211-216 [26] Stephenson TJ, Royds JA, Silcocks PB, Brammer HJ, Shorthouse AJ, Underwood JCE. Diagnostic association of p53 immunostaining in fine needle aspiration cytology of the breast. Cytopathology 1994; 5: 145-153. [27] Tanimura A, Masatomo K, Nagayama N, Tanaka S, Kosuga K. Cardiac myxoma. Morphological histochemical and tissue culture studies. Hum Path01 1988; 19: 316-322. [28] Gebode F, Kerth WJ, Hill JD. Surgical management of tumors of the heart. Surgery 1961; 61: 94-101. [29] Read RC, White HJ, Murphy ML, Williams D, Sun CN, Flanagan WH. The malignant potentiality of left atria1 myxoma. J Thorac Cardiovasc Surg 1974; 68: 857-868. [30] Richkind KE, Wason D, Vidaillet HJ. Cardiac myxoma characterized by clonal telomeric association. Genes Chromosomes Cancer 1994; 9: 68-71. [31] Dewald GW, Dahl RJ, Spurbek JL, Camey JA, Gordon H. Chromosomally abnormal clones and nonrandum telomeric translocations in cardiac myxomas. Mayo Clin Proc 1987; 62: 558-567. [32] McCarthy PM, Schaff HV, Winkler HZ, Lieber MM, Carney JA. Deoxyribonucleic acid ploidy pattern of cardiac myxomas. J Thorac Cardiovasc Surg 1989; 98: 1083-1086. [33] Kotylo PK, Kennedy JE, Waller BF, Sample RB. DNA analysis of atrial myxomas. Chest 1991; 99: 1203-1207. [34] Nolan J, Carder PJ, Bloomfield P. Atrial myxoma: tumour or trauma? Br Heart J 1992; 67: 406408. [35] Fletcher CD (Ed). Diagnostic Histopathology of Tumors. Edinburgh: Churchill Livingstone, 1995; 7-41.
of
cardiology ELSEVIER
InternationalJournal of Cardiology 57 (1996) 211-216
The nature of the cardiac myxoma S.K. Suvarnaa’*, J.A. Roydsb ‘Department of Histopathology, Northern General Hospital, Herries Road, Sheffield S5 7AU, UK bDepa@ment of Pathology, University of Shefield, Beech Hill Road, Shefield SlO ZRX, UK
Received26 June 1996;accepted4 September1996
Abstract Ten archive cases of cardiac myxoma were evaluated for proliferative activity, metastatic potential and expression of oncogene/tumour suppressorgene products by meansof PCNA, h4IB1, nm23, ~53, Bcl-2 and Rb-1 immunohistochemistry. The myxomas showed variable proliferative activity (PCNA O-41%, average 12.6%, MIBl O-138, average 3.2%) contrasting with the absenceof mitotic activity histologically. All the myxomas showed nm23 staining. None showed p53 reactivity. Eight caseswere negative for Bcl-2 expression,with two casesgiving weak cytoplasmic staining. Rb-1 reactivity showed a variable pattern (staining indices O-868) paralleling the cases’proliferativeactivity. The cardiacmyxomais interpreted as a weakly proliferative lesion with little metastatic potential and no modulation of oncogene/oncogene
suppressorproducts.Whilst not excluding a neoplasticaetiology, the results are consideredmore in keeping with a reactive/hamartomatous process. Keywords: cardiac myxoma; Immunohistochemistry; Oncogene;Proliferation index; Tumour
1. Intruduction Cardiac myxomas comprise 50% of all cardiac tumours [1,2], occurring sporadically and within familial syndromes (Camey’s complex, NAME syndrome, LAM3 syndrome) [3-51. The histological and ultrastructural features are well known 16-81. However, controversy regarding the histogenesis of the myxoma has existed for most of this century. Although initially considered to be an exuberant thrombus, the neoplastic interpretation is now favoured [g-12]. Given that neoplastic lesions may show high proliferative indices, metastatic activity and modulation of oncogeneI oncogene suppressor *Correspondingauthor.
products [ 131, we chose to evaluate the cardiac myxoma with antibodies appropriate to these features. Ki-67 is a nuclear antigen, recognised by MIB 1, that is presentin proliferating cells and absent in quiescent cells [14,15]. Assessmentof cell proliferation by Ki-67 detection has been shown to correlate with tumour grade and prognosis in neoplastic cells. PCNA (a 36 kDa protein subunit of DNA polymerase delta) is one of the late response genesduring mitogenic stimulation, DNA replication and repair [16]. It has a long half-life, being found in cycling and non-cycling cells. ~53 missense mutations lead to stabilisation and visualisation of ~53 by immunohistochemistry [17]. These are the most common genetic abnormalities in human cancer.Wild-type ~53 expression and stabili-
0167-5273/96/$15.00 Copyright 0 1996ElsevierScienceIrelandLtd. AU rights reserved PII
SO167-5273(96)02827-6
supprtxsorgene;Hmcnna
S.K. Suvama, J.A. Royds I International
212
sation is also increased in responseto DNA damage resulting in cell cycle arrest, repair or apoptosis [18]. The product of the retinoblastoma gene Rb-1 (pRb) forms part of the cell cycle control [19,20]. pRb is necessary for cell cycle progression and mitosis, and its function is also implicated in the maintenanceof differentiated cell type. Expression of Bcl-2 is associatedwith rescuefrom apoptosis, particularly in proliferating cells and is a mechanism contributing to malignancy, especially in low grade non-Hodgkin’s lymphomas [21,22]. Loss of nm23, especially in association with high cell turnover, generally correlates with the metastatic potential of many tumours [23-251.
2. Methods
All caseswere examined by routine haematoxylin/ eosin and toluidine blue histochemistry. Five p,m sections were stained with the following antibodies according to manufacturers recommendations or as previously described with modifications as listed
Journal of Cardiology 57 (1996) 211-216
count on the section or up to 200 cells, with results being expressedas a percentageof the total. 3. Results
Ten cases of formalin-fixed, paraffin-embedded cardiac myxoma were retrieved from the archive files. Eight were resected surgically, one was identified within an arterial embolus and one case was diagnosed at autopsy in a case of metastatic breast carcinoma. Review of the case records indicated no one with evidence of the Carney complex. All the casesshowed classic featuresof the cardiac myxoma (Fig. 1) without mitotic activity or cellular atypia. However, the autopsy case showed the cardiac myxoma arising in direct proximity to a focus of metastatic breast carcinoma. Despite the absenceof mitotic activity there was immunohistochemical evidence of proliferative ac-
[25,26].
PCNA, monoclonal PC10 (Novocastra Ltd), 1:200, 1 h room temperature. MB31 monoclonal (The Binding Site Ltd, UK), l:lOO, 4°C overnight. nm23, immunopurified rabbit antibody raised against synthetic peptide 11 (this was a generous gift from Dr P Steeg). ~53, monoclonal DO7 (Novocastra Ltd), 1:200, 1 h room temperature. bcl-2, monoclonal (Dako Ltd) 1:200, 20 h 4°C. Rb-1 monoclonal G3-349 (Pharmingen Ltd, Cambridge Bioscience, UK) 1:150, 4°C overnight. The sections were treated with the relevant ABC kit (Vector Labs, UK), with antigen visualisation by diaminobenzidine. Antigen retrieval was achieved by microwaving sections in 0.01 M sodium citrate buffer for 10 min at 650 W prior to primary antibody for Bcl-2, MIB 1, ~53 and Rb 1. Positive and negative controls were included. The proliferation indices involved the total cell
Fig. 1. The surface of a myxoma showing abundant myxoid matrix, scattered stellate/spindle cells and a characteristic vascular channel.
S.K. Suvama, J.A. Royds I lntemational
Fig. 2. PCNA-stained section showing some positively nuclei (arrowed). Other cell nuclei are negative.
stained
tivity. The mean proliferation index of MIBl was 3.2% (range O-13%) and PCNA 12.6% (range O41%) (Fig. 2). It was noted that those turnouts with no/very low proliferation scores were identical in both assays. Anti-nm23 antibody revealed uniform positive staining, although variable in reaction intensity (5 strong, 4 moderate, 1 weak). Contrastingly, p53 showed no reaction product in any myxoma (Fig. 3 and Fig. 4). Eight cases revealed no Bcl-2 staining, although two showed borderline cytoplasmic positivity. This latter feature was present in all cells, contrasting with the strong positivity of the stromal lymphocytes. Rb-1 staining indices varied between 0 and 86% (average 19.2%). In general, the positive scores paralleled the proliferation activity scores seen with MIBl and PCNA. The autopsy sample showed focal carcinoma with loss of nm23 staining, patchy reactivity with Rb-1 and no staining with ~53 or Bcl-2. The carcinoma
Journal of Cardiology 57 (1996) 211-216
213
Fig. 3. p53-stained section of a myxoma showing no immunohistochemical reaction product, contrasting with the control section (Fig. 4).
showed proliferative activity (seen with PCNA and MIBl) but the tissue available was too small for formal quantification. The individual results are expressed in group format (Table 1). No relationship was identified in relation to the age, sex or size of the myxoma. 4. Discussion Proof of the myxomas histogenesis has been elusive since its identification. The thrombotic theory appearsto have been discredited. The majority view of neoplasia is favoured although a hamartomatous derivation has been considered. Attempts to resolve the histogenesis question have addressedthe ultrastructural and immunohistochemical phenotype of the myxoma and have demonstratedvariable phenotypic differentiation (endothelial, fibroblastic, muscle and epithelial patterns) [6,10,11,27]. This at first
S.K. Suvama, J.A. Royds I International
214
Fig. 4. A colonic tubular adenoma showing positive p53 nuclear staining in the dysplastic glandular epithelium (arrowed).
sight is confusing and derivation from primitive multipotential mesenchymal cells has been invoked as the explanation. In addition, recurrence following incomplete surgery [28,29], embolisation being mistaken for
Journal of Cardiology 57 (1996) 211-216
tumour metastasisand misdiagnosis of cardiac sarcomas showing myxoid features [ 1,2] must be borne in mind when considering the nature of the cardiac myxoma and to avoid ascribing the myxoma with features that are not appropriate. Some studies have shown myxoma cell cultures to have chromosome aberrations [30,31]. However, this is at variance with previously published ploidy data and contrasts [32,33] with other investigators’ difficulty in establishing cell cultures. The results of the present study throw light on the nature of the myxoma. It is noted that none of the myxomas showed ~53 reactivity. Although caution is required regarding uncritical appraisal of ~53 staining [ 171, the absence of such a common gene mutational event does not support a role for neoplasia in the genesis of the cardiac myxoma. Similar uniformity is seen with all the tumours demonstrating nm23 reactivity. This contrasts with the metastatic carcinoma, seen in the autopsy case which had focal nm23 loss. Consequently it is inferred that the cardiac myxoma has little potential for metastasis. High nm23 expression was most likely to be accompanied by moderate or high expressionof cell cycle related genes(Ki-67, PCNA/ pRb). Low nm23 expressionwas uniform and associated with low proliferation and was therefore not likely to signify any metastatic potential. The cellular proliferation markers (MIBl and PCNA) demonstratea wide range of reactivity. It is noted that the proliferation reactivity, where positive,
Table 1 Individual results Case 1 2 3 4 5 6 I 8 9 10
Age/sex 25/F 39/M 30/F 36/F 55/M 51/M 56/F 55/M 49/M 51/F
nm23
ff+ ++ + +++ +++ ++ ++ +++ +++ ++
P53
-
%MIB 1 3 0 0.5 13 0.5 0 2 6 1 0
%FCNA 5 0 0.5 39 20 0.5 5 41 8 0.5
All numerical values expressed as percentage of total cells. Staining grades: -I + I + + I + + + =negative/weak/moderate/strong. Source/origin of the pathological material: A, autopsy; S, surgery; E, embolectomy, Site: RA, right atrium; LA, left atrium; AE, arterial embolus.
%Rb-1
Bcl-2
Source
Site
1 0 8 19 0 19 3 86 76 0
-
S S S S S S E S S A
RA LA LA LA LA LA AE LA LA LA
+ + -
S.K. Suvarna, J.A. Royds I International
showed a similar range of results for both markers indicating the validity of the observations. However, careful scrutiny of the myxoma cases failed to indicate a reason for the range of results. Specifically the myxomas were assessed for the degree of cellularity, amount of myxoid matrix, degree of haemorrhage and amount of blood vessels. Whilst there was fluctuation of thesecomponentparts within the myxomas studied no association was identified with the proliferation markers. Moreover, it is stressedthat no mitotic figures or nuclear atypia was seen histologically. Therefore, we are left with the conclusion that the myxomas do show inherent variation in proliferation activity, perhaps reflecting the stage of growth of the lesion studied. Since the time from initiation to sampling cannot be determined this interpretation cannot be validated. Rb-1 staining broadly paralleled the proliferation activity and was similarly shown to have no relationship to vascular&y, myxoid matrix, cellularity or the degree of haemoirhage. Furthermore, since no other tissue metaplasia (glandular/muscular differentiation) was identified the Rb-1 staining does not parallel any morphological parameter within the myxomas. Bcl-2 is associated with rescue from apoptosis, particularly in proliferating cells and its aberrant expressionis a recognisedmechanismcontributing to malignancy. It is noted that the majority of the myxomas showed no reactivity for this immunostain. That two casesdid show reactivity is intriguing since both these cases are seen to be ‘proliferating’ myxomas. As with the proliferation markers (MIBl and PCNA) it is possible that the reactivity identified representsa stage in myxoma evolution/growth. Taken together, these results infer the cardiac myxoma to be a lesion of low proliferative activity without evidence of mutation/stabilisation of tumour suppressor or oncogene products (~53 and Bcl-2 respectively). The nm23 staining pattern indicates a tumour with low metastatic potential. Rb-1 staining is open to debatebut may simply reflect proliferative activity. Whilst not excluding a neoplastic interpretation, we consider the results to be more in keeping with a benign/reactive process. The possibility of a reactive phenomenon [34] or a hamartomatousprocess cannot be similarly excluded [35]. Further research is indicated into this intriguing entity.
Journal of Cardiology 57 (19%) 211-216
215
Acknowledgments
We wish to thank Dr. P Steegfor the generousgift of the nm23 antibody. We also thank Mr. K. Corke for technical assistance.J. A. Royds is supported by the Yorkshire Cancer ResearchCampaign.
References Ul McAllister HA, Fenoglio JJ. Tumors of the cardiovascular system. Washington DC: Armed Forces Institute of Pathology, 1996; 21-46. I21Rosai .J. Ackerman’s Surgical Pathology, 7th ed. St. Louis: CV Mosby, 1996; 2182-2184. [31 Carney JA. Differences between nonfamilial and familial cardiac myxoma. Am J Surg Path01 1985; 9: 53-55. [41 McCarthy PM, Piehler JM, Schaff HV, Pluth JR, Orszulak TA,Vidaillet HJ et al. The significance of multiple, recurrent, and ‘complex’ cardiac myxomas. J Thorac Cardiovasc Surg 1986; 91: 389-396. I51Carney JA, Hruska LS, Beauchamp GD, Gordon H. Dominant inheritance of the complex of myxomas, spotty pigmentation and endocrine overactivity. Mayo Clin Proc 1986; 61: 165-172. 161Ferrans VJ, Roberts WC. Structural features of cardiac myxomas. Hum Path01 1973; 4: 111-146. [71 Feldman PS, Horvath E, Kovaks K. An ultrastructural study of 7 cardiac myxomas. Cancer 1977; 40: 2216-2232. PI Wold LE, Lie JT. Scanning electron microscopy of intracardiac myxoma. Mayo Clin Proc 1981; 56: 198-200. [91 Lie JT. The identity and histogenesis of cardiac myxomas. Arch Path01 Lab Med 1989; 113: 724-726. IlO1Landon G, Ordonez NG, Guarda LA. Cardiac myxomas. An immunohistochemical study using endothelial, histiocytic and smooth muscle markers, Arch Path01 Lab Med 1986; 110: 116-120. 1111Krikler DM, Rode J, Davies MJ, Woolf N, Moss E. Atria1 myxoma: a tumour in search of its origins. Br Heart J 1992; 67: 89-91. WI Salyer WR, Page DL, Hutchins GM. The development of cardiac myxomas and papillary endocardial lesions from mural thrombus. Am Heart J 1975; 89: 4-17. [131 Cotran RS, Kumar V, Robbins SL. Robbins Pathologic Basis of Disease, 4th ed. Philadelphia: WB Saunders, 1989; 239305. u41 Gerdes J, Li L, Schlueter C, Duchrow M, Wohlenberg C, Gerlach C et al. Immunobiochemical and molecular biologic characterisation of the cell proliferation-associated nuclear that is defined by monoclonal antibody Ki67. Am J Path01 1991; 138: 867-873. r151 McCormick D, Yu C, Hobbs C, Hall PA. The relevance of antibody concentration to me immunohistochemical quantification of cell proliferation associated antigens. Histopathology 1993; 22: 543-547.
216
S.K. Suvamu, J.A. Royds I International
[16] McCormick D, Hall PA. The complexities of proliferating cell nuclear antigen. Histopathology 1992; 21: 591-594. [17] Wynford-Thomas D. p53 in tumour pathology: can we trust immunocytochemistry? J Path01 1992; 166: 329-330. [18] Camnan CE, Kastan MB. Induction of apoptosis by tumour suppressor genes and oncogenes. Cancer Biol 1995; 6: 1725. [19] Royds JA, Webley K, Karcioglu ZA, Parsons MA. Apoptosis-related gene expression in retinoblastomas. Br J Cancer, 1996; 44: 73s. [20] Jacks T, Fazeli A, Schmitt AM, Bronson RT, Goode11 MA, Weinberg RA. Effects of an Rb mutation in the mouse. Nature 1992; 359: 295-300. [21] Tsujimoto Y, Ikegaki N, Croce CM. Characterisation of the protein product of bcl-2, the gene involved in human follicular lymphoma. Oncogene 1987; 2: 3-7. [22] Nathan B, Anbazhagan R, Dyer M, Ebbs SR, Jayatilake H, Gusterson BA. Expression of bcl-2-like reactivity in the normal breast and in breast cancer. The Breast 1992; 2: 134-137. [23] Leone A, Flatow U, VanHoutte K, Steeg PS. Transfection of human nm23-Hl into the human MDA-MB-435 breast carcinoma cell line: effects on the tumor metastatic potential, colonisation and enzymic activity. Oncogene 1993; 8: 23252333. [24] Steeg PS, Bevilacqua G, Kopper L, Thorgeirsson UP, Talmadge JE, Liotta LA, Sobel ME. Evidence for a novel gene associated with low metastatic potential. J Nat1 Cancer Inst 1988; 80: 200-204. [25] Royds JA, Stephenson TJ, Rees RC, Shorthouse AJ, Silcocks PB. mn23 product expression in ductal in situ and invasive human breast carcinoma. J Natl Cancer Inst 1993; 85: 727731.
Jownal of Cardiology 57 (1996) 211-216 [26] Stephenson TJ, Royds JA, Silcocks PB, Brammer HJ, Shorthouse AJ, Underwood JCE. Diagnostic association of p53 immunostaining in fine needle aspiration cytology of the breast. Cytopathology 1994; 5: 145-153. [27] Tanimura A, Masatomo K, Nagayama N, Tanaka S, Kosuga K. Cardiac myxoma. Morphological histochemical and tissue culture studies. Hum Path01 1988; 19: 316-322. [28] Gebode F, Kerth WJ, Hill JD. Surgical management of tumors of the heart. Surgery 1961; 61: 94-101. [29] Read RC, White HJ, Murphy ML, Williams D, Sun CN, Flanagan WH. The malignant potentiality of left atria1 myxoma. J Thorac Cardiovasc Surg 1974; 68: 857-868. [30] Richkind KE, Wason D, Vidaillet HJ. Cardiac myxoma characterized by clonal telomeric association. Genes Chromosomes Cancer 1994; 9: 68-71. [31] Dewald GW, Dahl RJ, Spurbek JL, Camey JA, Gordon H. Chromosomally abnormal clones and nonrandum telomeric translocations in cardiac myxomas. Mayo Clin Proc 1987; 62: 558-567. [32] McCarthy PM, Schaff HV, Winkler HZ, Lieber MM, Carney JA. Deoxyribonucleic acid ploidy pattern of cardiac myxomas. J Thorac Cardiovasc Surg 1989; 98: 1083-1086. [33] Kotylo PK, Kennedy JE, Waller BF, Sample RB. DNA analysis of atrial myxomas. Chest 1991; 99: 1203-1207. [34] Nolan J, Carder PJ, Bloomfield P. Atrial myxoma: tumour or trauma? Br Heart J 1992; 67: 406408. [35] Fletcher CD (Ed). Diagnostic Histopathology of Tumors. Edinburgh: Churchill Livingstone, 1995; 7-41.