Nucleolar Organiser Regions: New Prognostic Variable in Breast Carcinomas

390 J Clin Pathol 1990;43:390-392 Nucleolar organiser regions: new prognostic variable in breast carcinomas

E Sivridis, B Sims

Abstract specimens were selected at random from the Nucleolar organiser regions (NORs), files of the Department of Histopathology, which are important for regulating General Hospital, Alexandroupolis. All protein synthesis, were identified in 20 specimens had been fixed in 10% formol-saline breast carcinomas by means of a silver and processed routinely through absolute alco- (Ag) staining technique. Infiltrating neo- hol and xylene before embedding to paraffin plasms with metastases in four or more wax (Paraplast). There were 14 infiltrating axillary lymph nodes possessed, on ductal carcinomas (of scirrhous, medullary, average, a greater number of AgNORs and mucinous variants), two infiltrating per cell nucleus compared with neo- lobular carcinomas, and four pure intraductol plasms without nodal disease, or with one cartinomas. to three positive lymph nodes. The size, The breast tumours were divided into two morphology, and distribution ofAgNORs groups depending on the number of positive within the nucleus were also different in lymph nodes at the time of surgery. The first the two study groups. Overall, these find- group comprised carcinomas in situ and infil- ings suggest that breast carcinomas with trating carcinomas without nodal disease or multiple, irregular, and widely dispersed with metastases in up to three axillary lymph AgNORs tend to be of high grade malig- nodes (eight cases). The second group consis- nancy. ted of infiltrating breast neoplasms with metastases in four or more lymph nodes (12 cases). In every case at least 10 axillary nodes were Nucleolar organiser regions (NORs) are loops examined histologically. of DNA that are responsible for ribosomal The modified AgNOR technique, as des- RNA (rRNA) transcription.' They are located cribed by Ploton et al,6 was applied to paraffin in the nucleoli of cells and in the chromosomes wax sections of 3 ,im. AgNOR counts were 13-15 and 21, 22 in association with proteins.2' performed blind and independently by two As rRNA molecules are the main sites of observers, using a standard procedure'8: in this, protein synthesis it follows that the number of clusters of black dots within nucleoli are NORs in each cell nucleus reflects cellular counted as one AgNOR; black dots dispersed activity." throughout the nucleus are counted as discrete Nucleolar organiser regions can be readily AgNORs. The mean numbers of AgNORs per identified in paraffin wax sections by means of cell nucleus (AgNOR clusters, dispersed a silver (Ag) staining technique (AgNOR)6: AgNORs, and totals) was determined for each they are visualised in the nuclei of cells as case after counting 100 nuclei directly on to brown/black dots by virtue of the argyrophilia tissue sections at x 400 magnification. Cells of NOR-associated proteins.7 These were were examined at random. thought to include proteins B238 and C23,9 AgNOR counting was repeated subse- subunits of RNA polymerase 1 ,10 and probably quently, along the same lines, using projected other phosphoproteins. The function of these transparencies; microphotographs had been molecules is still far from clear; they may partly taken randomly at magnifications of x 200 and reflect ploidy," though an increase in ploidy x 500. Interobserver and intraobserver error does not necessarily lead to an increase in were calculated. AgNORs in human tumours.'2'14 The significance of results (mean numbers of For years the AgNOR technique has been AgNOR clusters plus dispersed AgNORs) was used by cytogeneticists for the investigation of assessed by the Wilcoxon rank sum test. Department of certain genetic disorders, notably that of Histopathology, General Hospital, chromosome 21 .' The recent modification of 68100 the original method,6 however, became Results Alexandroupolis, increasingly applied in histopathology research AgNORs were clearly visible on light micro- Greece as black silver E Sivridis following suggestions that there may be a scopical examination binding association between dots within nuclei stained orange. There were, Department of possible high AgNOR Histopathology, counts and malignant transformation.461617 In however, well defined differences between the General Hospital, this study we investigated the usefulness of the two study groups of breast carcinomas. Thus Birmingham technique in assessing the aggressive- malignant neoplasms without nodal disease or B Sims AgNOR ness in breast carcinomas. with metastases in less than four lymph nodes Correspondence to: on two to three Dr E Sivridis contained, average, AgNORs SD Accepted for publication Methods per cell nucleus (mean 2-81; median 2-65; 12 January 1990 Tumour tissue blocks from 20 mastectomy 0-61; 95%O CI 1-59-403): these were uniform, AgNORs in breast carcinomas 391

Patterns of AgNOR staining in breast carcinomas

112 Mean AgNOR count ulo Lymph node state No of cases AgNOR clusters Dispersed AgNORs Totals

- nL 10-~~~~ Negative nodes or one to three positive nodes 8 1 97 0-84 2 81 8- z 0 Four or more positive nodes 12 2 24 6-33 8-57 '6-

0 round, and dense. By contrast, AgNORs in tial use in predicting behaviour in breast car- breast carcinomas with four or more positive cinomas. This would be very convenient for 2 £ nodes were usually, though not invariably, of in small Group A Group B assessing aggressiveness incisional (n481 (na12) bizarre morphology and more numerous: up to biopsy specimens, needle aspiration cytology, Scattergram showing mean three large, often irregular, dense structures or imprint cytology. The validity of such an total AgNOR numbersl and multiple smaller argyrophilic dots, less assumption, however, could only be verified by cell nucleusfor each distinct and widely dispersed within the a large series of patients participating in long specimen of breast carcinoma examined. nucleus (mean 8 57 AgNORs/nucleus; median term follow up studies.30 8 00; SD 2-63; 95O" CI 3-3-13-72). The In other tissues and disease states claims that results are expressed as a scattergram in the the AgNOR activity is ofprognostic value have figure; the difference between the two study not been warranted by the rather scanty and groups was significant (p < 0-001), although contradictory reports.'83' Equally, suggestions the numbers are too small to be able to detect that the method may be helpful in discrim- clinically important differences. The intra- and inating benign from malignant disease or in inter-observer error was between 4 1 and distinguishing between low and high grade 6 30o. The correlation between AgNOR malignancy have been found to be the case in counts by the two methods of enumeration was some studies, 4 5 16 32 33 but not in others.3 37 good: r = 0-8. The table shows the patterns of Further studies are clearly needed before the AgNOR staining and their mean values per clinical importance of the AgNOR staining tumour cell nucleus. AgNOR counts, though method can be fully appreciated. highly reproducible by both methods, were more easily accomplished in projected photo- We thank Mr P Smith, Chief MLSO, Depart- graphic transparencies. ment of Histopathology, East Birmingham Hospital, for his invaluable technical assistance in respect to the AgNOR silver staining tech- Discussion nique, and Mrs C Price for her secretarial The prognosis of primary breast carcinoma assistance. depends on several morphological and endo- crinological variables. These include: tumour 1 Fakan S, Hernandez-Verdun D. The nucleolus and the size; type of tumour; the degree of differentia- nucleolar organizer regions. Biol Cell 1986;56:189-206. 2 Ferguson-Smith MA, Handmaker SD. Observations on the tion into glandular or tubular structures; the satellited human chromosomes. Lancet 1961;i:638-40. presence, degree, and prevalence of nuclei 3 Lewin B. Gene expression. 2. In: Eucaryotic chromosomes. 2nd ed. New York: John Wiley, 1980:875-8. showing hyperchromatism, pleomorphism, 4 Crocker J, Nar P. Nucleolar organizer regions in lym- and mitotic activity; the degree of lymphocytic phomas. J Pathol 1987;151:111-8. 5 Smith R, Crocker J. Evaluation of nucleolar organizer response; the presence of local invasion and region-associated proteins in breast malignancy. Histo- lymph node metastases; and the tumour recep- pathology 1988;12:113-25. 6 Ploton D, Menager M, Jeanesson P, Himber G, Pigeon F, tor state.' 122 Although these various factors Adnet J-J. Improvement in the staining and in the are evaluated in combination, in practice the visualization of the argyrophilic proteins of the nucleolar organizer region at the optical level. Histochem J 1986; most important single factor in determining 18:5-14. prognosis seems to be the extent of disease 7 Hubbell HR, Rothblum LI, Hsu TC. Identification of a silver binding protein associated with the cytologicaf silver at the time of operation.2>25 Microscopic staining of actively transcribing nucleolar regions. Cell evidence of metastases in four or more axillary Biol Int Rep 1979;3:615-22. nodes is associated with a per- 8 Lischwe MA, Smetana K, Olson MOJ, Busch H. Proteins lymph higher C23 and B23 are the major nucleolar silver staining centage of treatment failure and a lower in- proteins. Life Sci 1979;25:701-8. 9 Ochs R, Lischwe M, O'Leary P, Busch H. Localization of cidence of five year survival, compared with nucleolar phosphoproteins B23 and C23 during mitosis. cases where nought to three lymph nodes are Exp Cell Res 1983;146:139-49. 10 Williams MA, Kleinschmidt JA, Krohne G, Franke WW. affected.923 2628 Argyrophilic nuclear and nucleolar proteins of Xenopus The morphological expression of this prog- laevis oocytes identified by gel electrophoresis. Exp Cell Res 1982;137:341-51. nostic determinant may well be the silver-bind- 11 Trent JM, Carlin DA, Davies JR. Expression of silver ing NOR-associated proteins of the cell stained nucleolar organizing regions (AgNORs) in human as cancer. Cytogenet Cell Genet 1981;30:31-8. nucleus. Thus the present study suggests, 12 Giri DD, Nottingham JF, Lawry J, Dundas SAC, Under- there seems to be a relation between AgNOR wood JCE. Silver-binding nucleolar organizer regions counts (Ag-NORs) in benign and malignant breast lesions: and the number ofpositive lymph nodes correlations with ploidy and growth phase by DNA flow in breast carcinomas: the higher the numbers cytometry. J Pathol 1986;57:307-13. and more 13 Derenzini M, Pession A, Farabegoli F, Trere D, Badiali M, the bizarre the morphology of Dehan P. Relationship between interphasic nucleolar AgNORs in the primary tumour, the more organizer regions and growth rate in two neuroblastoma extensive the nodal disease. The reasons for cell lines. Am J Pathol 1989;134:925-32. 14 Crocker J, Janmohamed RMI, Armstrong SJ, Hulten M, this tendency are obscure but may be related to Leyland MJ. The relationship between numbers of inter- an increased nuclear phase NORs and NOR-bearing chromosomes in non- activity.4529 Hodgkin's lymphoma. JPathol 1989;157:166A. It is tempting to speculate, therefore, that the 15 de la Cruz FF, Gerald PS. Trisomy 21. Baltimore: University numbers, the shape, and the distribution of Park Press, 1981:165-7. 16 Croker J, Skilbeck N. Nucleolar organizer region associated AgNORs within the nucleus might be of poten- proteins in cutaneous melanotic lesions: a quantitative 392 Sivridis, Sims

study.]J Clin Pat/aol 1987;40:885-9. 27 Primary Therapy ofBreast Cancer Study Group. Identifica- 17 Fallowfield ME, Dodson AR, Cook MG. Nucleolar organ- tion of breast cancer patients with high risk of early izer regions in melanocytic dysplasia and melanoma. recurrence after radical mastectomy. II. Clinical and Histopathology 1988;13:95-9. pathological correlation. Cancer 1978;42:2809-26. 18 Howat AJ, Giri DD, Wright AL, Underwood JCE. Silver- 28 Carter CL, Allen C, Henson DE. Relation of tumour size, stained nucleoli and nucleolar organizer region counts are lymph node status, and survival in 24,740 breast cancer of no prognostic value in thick cutaneous malignant cases. Cancer 1989;63:181-7. melanoma. J Pathol 1988;156:227-32. 29 Crocker J, Macartney JC, Smith PJ. Correlation between 19 Russo J, Frederick J, Ownby HE. Predictors of recurrence DNA flow cytometric and nucleolar organizer region data and survival of patients with breast cancer. Am J Clin in non-Hodgkin's lymphomas. J Pathol 1988;154:151-6. Pathol 1987;88:123-31. 30 Brown RA, Beck JS. Statistics on microcomputers: A non- 20 Martin SA, Perez-Reyes N, Mendelsohn G. Angioinvasion algebraic guide to the appropriate use of statistical pack- in breast carcinoma. An immunohistochemical study of ages in biomedical research and pathology laboratory factor VIII-related antigen. Cancer 1987;59:1918-22. practice. 6 Statistical methods for diagnostic tests. J Clin 21 Rank F, Dombernowsky P, Jespersen NCB, et al. Histologic Pathol 1989;42:225-30. malignancy grading of invasive ductal breast carcinoma. 31 Egan M, Raafat F, Crocker J, Williams D. Comparative Cancer 1987;60:1299-305. study ofthe degree ofdifferentiation ofneuroblastoma and 22 Alexieva-Figusch J, van Putten WLJ, Blankestein MA, et al. mean numbers of nucleolar organiser regions. J Clin The prognostic value and relationships of patient charac- Pathol 1988;41:527-31. teristics, estrogen and progestin receptors, and site of 32 Egan M, Raafat F, Crocker J, Smith K. Nucleolar organizer relapse in primary breast cancer. Cancer 1988;61:758-68. regions in fibrous proliferations ofchildhood and infantile 23 Fisher B, Glass A, Redmond C, et al. L-Phenylalanine fibrosarcoma. J Clin Pathol 1988;41:31-3. mustard (L-PAM) in the management of primary breast 33 Egan MJ, Crocker J. Nucleolar organizer regions in cuta- cancer: An update of earlier findings and a comparison of neous tumours. J Pathol 1988;154:247-53. the utilizing L-PAM plus 5 fluorouracil (5 FU). Cancer 34 Nairn ER, Crocker J, McGovern J. Limited value of 1977;39:2883-903. AgNOR enumeration in the assessment of thyroid neo- 24 Nemoto T, Vana J, Bedwani RN, et al. Management and plasms. J Clin Pathol 1988;41:1136. survival of female breast cancer: Results of a national 35 Howat AJ, Giri DD, Cotton DWK, Slater DN. Nucleolar survey by the American College ofSurgeons. Cancer 1980; organizer regions in Spitz nevi and malignant melanomas. 45:2917-24. Cancer 1989;63:474-8. 25 Fisher B, Bauer M, Wickerham I, Redmon CK, Fisher ER. 36 Giri DD, Dundas SAC, Sanderson PR, Howat AJ. Silver- Relation ofnumber ofpositive axillary nodes to prognosis binding nucleoli and nucleolar organizer regions in fine of patients with primary breast cancer. Cancer 1980; needle aspiration cytology of the breast. Acta Cytol 52:1551-7. 1989;33: 173-5. 26 Fisher ER, Gregorio R, Redmond C. Pathologic findings 37 Giri DD, Nottingham JF, Lawry J, Dundas SAC, Under- from the National Surgical Adjuvant Breast Project wood JCE. Silver-binding nucleolar organizer regions (Protocol N. 4). II. The significance of regional node (AgNORs) in benign and malignant breast lesions: histology other than sinus histiocytosis in invasive correlations with ploidy and growth phase by DNA flow mammary cancer. Am J Clin Pathol 1976;65:21-30. cytometry. J Pathol 1989;157:307-13.