Microsoft Word - 7DumitrescuDiagnostic.doc 64 Romanian Neurosurgery (2010) XVII 1: 64 - 72 Diagnostic value of silver nitrate staining for nucleolar organizer regions in cerebral astrocytic tumors Gabriela – Florenta Dumitrescu1, Anca Indrei2, I. Poeata3, Danisia Haba4, Elena Adinisia Agrigoroae1, Fl. Grămadă3, Dana-Mihaela Turliuc3 1Emergency Clinical Hospital “Prof. Dr. N. Oblu” Iaşi, Department of Pathology; 2University of Medicine and Pharmacy “Gr. T. Popa” Iaşi, Department of Morphological Sciences, Discipline of Anatomy; 3University of Medicine and Pharmacy “Gr. T. Popa” Iaşi, Department of Surgery II, Discipline of Neurosurgery; 4University of Medicine and Pharmacy “Gr. T. Popa” Iaşi, Department of Oro-maxillo-facial Surgery, Discipline of General and Dental Radiology Abstract Aim: to compare the AgNOR's mean number with the histological type and grade of cerebral astrocytic tumors. 16 primary cerebral astrocytic tumors (4 diffuse astrocytomas, 4 anaplastic astrocytomas and 8 glioblastomas) stereotactic biopsied in the Department of Neurosurgery, Clinical Hospital „Prof. Dr. N. Oblu” Iaşi, and histopathologically conventional diagnosed in Department of Neuropathology of the same hospital, were retrospectively identified. Tumor specimens were submitted to a combined staining technique: one – step silver nitrate method for AgNOR protein sites (modified after Ploton et al, 1986) counterstained with periodic acid- Schiff staining for basement membrane of vascular components. The mean AgNOR values (mAgNOR) for tumoral and vascular nuclei were determined. The average values of mean AgNORs/nucleus (mAgNOR/nucleus) presented a linear increase with increasing grade of malignancy from 1.96 for diffuse astrocytoma (GII), 2.34 for anaplastic astrocytoma (GIII), to 3.18 for glioblastoma multiforme (GIV). mAgNOR/tumoral nucleus also showed a linear correlation with the histological tumor grade: 2.27 for diffuse astrocytomas, 2.78 for anaplastic astrocytomas, and 3.35 for glioblastomas multiforme. A distinct difference between the mean values of AgNORs/vascular nucleus was expressed: 1.52 for diffuse astrocytoma (GII), 1.90 for anaplastic astrocytoma (GIII), and 3.18 for glioblastoma multiforme (GIV). There were some overlaps between GIII and GIV astrocytic tumors regarding the mAgNOR/tumoral nucleus: maximum value in anaplastic astrocytomas (GIII) was 2.91 and minimum value in small cells glioblastomas (GIV) was 2.47. Differentiation could be achieved with mAgNORs/vascular nucleus as no extreme value overlapped: maximum value in anaplastic astrocytomas (GIII) was 1.99 and minimum value in small cells glioblastomas (GIV) was 2.54. The malignancy grade of an astrocytic tumor can accurately be establish both on histological features of the conventional stained sample and on the average number, the shape and the distribution of AgNORs within tumoral and vascular nuclei, as AgNORs determinations supplement the histological information in small biopsies. Keywords: astrocytic tumor, nucleor organizer region, tumoral grade In the practice of neurooncology it is extremely important to know the histological grade of a brain tumor and particularly of an astrocytic neoplasia. Objective tumor growth potential is achieved today by immunohistochemical Gabriela – Florenta Dumitrescu et al Diagnostic in cerebral astrocytic tumors 65 methods, but there are many other cell cycle-associated molecules that are potential targets for histochemistry assessment of cell proliferation. The metaphase Nucleolar Organizer Regions (NORs) are chromosomal segments or loops (rDNA) containing ribosomal genes associated with proteins. These genes are clustered in 10 loci of the human acrocentric chromosomes 13, 14, 15, 21, and 22. The transcriptional activity of these intranuclear segments plays a pivotal role in the formation of nucleoli, directing the syntesis of both ribosomes and associated proteins [5]. NORs associated protein [C23 (nucleolin) and B23 (nucleophosmin)] are selectively stained by silver impregnation technique in formalin-fixed paraffin-embedded tissues. After silver-staining, the NORs can be easily identified as black dots exclusively localized throughout the nucleolar area, and are called “AgNORs”. The expression of AgNOR proteins was associated with several biological properties of neoplastic cells: metabolic activity, DNA content, histological grade of differentiation and, especially, the rapidity of cellular proliferation [31, 7, 26, 25]. The specific significance of AgNORs is not well understood. AgNOR quantification was used in tumor pathology as a parameter to distinguish malignant cells from benign or normal cells [11, 10, 9]. In the latest 20 years more than 400 studies was done in order to determine the correlation of AgNORs' quantity and quality with tumor aggressivness and rapidity of proliferation, with variable success rates that depended on tumor type [5]. A large number of papers have shown a linear correlation between AgNOR count and growth fraction in various human malignancies: breast cancers [32, 18], urinary bladder carcinoma [6], colorectal carcinoma [23], bone tumors [10], utererine cancer [12], endocrine cancers [22], skin cancers [2], lung cancers [28], brain cancers [8, 20]. Moreover, the method can be applied to small biopsies, can identify neoplastic clones with different proliferative activities and may stratify patients into different risk groups [25]. Because assessment of AgNORs offers an alternative approach to measure the tumor growth fraction and because there are a small number of articles dealing with AgNORs expression in astrocytic tumors, in this article we analyze possible correlation between AgNOR's morphology and AgNOR’s number with histological tumor grade in order to establish a possible role for it as an objective indicator of tumor biological behavior, just after the resection of a cerebral astrocytic neoplasia. Material and Methods Selection of cases 16 primary cerebral astrocytic tumors stereotactic biopsied in the Department of Neurosurgery, Clinical Hospital „Prof. Dr. N. Oblu” Iaşi, and histopathologically diagnosed in Department of Neuropathology of the same hospital, were retrospectively identified. In all cases, tumor samples were fixed in 10% buffered formalin, included in paraffin, and stained with H&E according to standard procedure. 66 Romanian Neurosurgery (2010) XVII 1: 64 - 72 Table 1 Histological subtypes and grade of malignancy of the 16 investigated astrocytomas Histological subtypes with grade of malignancy No. of cases Diffuse astrocytoma (grade II) 4 Anaplastic astrocytoma (grade III) 4 Glioblastoma multiforme (grade IV) With small cells With giant cells 8 4 4 All 16 cases were histopathological diagnosed and graded according to the WHO Classification of Brain Tumors [19] and divided into groups (Table 1). There were 4 cases of diffuse astrocytoma - grade II (GII), 4 cases of anaplastic astrocytoma - grade III (GIII), and 8 cases of glioblastoma multiforme - grade IV (GIV). AgNOR Staining Technique Five µm thick paraffin sections, obtained from paraffin blocks with the most representative tumoral areas, were selected and submitted to a combined staining technique: one – step silver nitrate method for AgNOR protein sites (modified after Ploton et al, 1986) [27] counterstained with periodic acid-Schiff staining for basement membrane of vascular components. Briefly, slides were deparaffinized in xylene, and hydrated through descending concentrations of ethanol to double distilled, deionized water. The sections were stained with freshly prepared silver colloidal solution for 25 min in a thermostatically controlled environment (37oC). The working solution contained one volume of 2% gelatine in 1% aqueous formic acid and two volume of 50% aqueous silver nitrate solution. Sections were counterstained with periodic acid - Schiff solution, dehydrated through ascending grades of ethanol, cleared in xylene, and mounted in Canada balsam. Quantification of the number of AgNORs Histological sections, stained by the exposed method, were examined by eye, using in each case Olympus microscope, magnification 1000x, an oil-immersion lens, and cedarwood oil. In all cases we focused the image throughout the section thickness using at least 10 different microscopic fields in order to determine the homogeneous AgNOR quantitation throughout the tumor. Quantification of the AgNOR number has always been done in well preserved cells, excluding areas of tumoral necrosis, areas with staining artifacts or overlapping cells. We analysed the dense cellular areas and those with vessels that had maximal endothelial proliferation. Using the average formula we determined the mean AgNOR values for tumoral and vascular cells, taken individually and then together, for every histological subtype and then for every grade of malignancy. Mean number of AgNORs for each case was defined as the ratio between total number of black dots determinated in 100 individual tumor cells and the number of analysed cells. The same ratio was done for 100 individual vascular cells. At each determinated black dot were noted the size (subjective quantified in: small, medium and large), the appearance (granular or homogeneous), and any other morphological features encountered. Gabriela – Florenta Dumitrescu et al Diagnostic in cerebral astrocytic tumors 67 Results The average values of mean AgNORs/nucleus (mAgNORs/nucleus) for each investigated tumor and its correlation with malignancy grades were represented in Table 2. Taking into account both tumoral and vascular cells, the mean values of mAgNORs/nucleus presented a linear increase with increasing grade of malignancy of cerebral astrocytomas from 1.96 for diffuse astrocytoma (GII), 2.34 for anaplastic astrocytoma (GIII), to 3.18 for glioblastoma multiforme (GIV). A distinct difference between the mean values of AgNORs/vascular nucleus was expressed: 1.52 for diffuse astrocytoma (GII), 1.90 for anaplastic astrocytoma (GIII), and 3.18 for glioblastoma multiforme (GIV). The correlation between mAgNORs/tumoral nucleus and histological subtype of glioblastoma multiforme showed a distinct difference between small cells and giant cells variants, as giant cells glioblastoma expressed higher values (3.94) than small cells subtype (2.76). The differential diagnosis between anaplastic astrocytomas and small cells glioblastomas became very difficult to determine based only on mAgNORs/tumoral nucleus because the mean values are almost identical (2.78 versus 2.76) as there were some overlaps of mAgNORs/tumoral nucleus between these two grades of malignancy: maximum value in anaplastic astrocytomas (GIII) was 2.91 and minimum value in small cells glioblastomas (GIV) was 2.47. Differentiation could be achieved by association of mAgNORs/vascular nucleus as there was no overlap between any investigated astrocytic tumor: maximum value in anaplastic astrocytomas (GIII) was 1.99 and minimum value in small cells glioblastomas (GIV) was 2.54 (Table 3). Figure 1 Diffuse astrocytoma (GII): 1 to 5 small black dots in tumoral nuclei (silver impregnation method, magnification x1000) Figure 2 Glioblastoma multiforme (GIV): tumoral nuclei with 2 to 7 medium black dots scattered across nucleus area; one nucleus with bizarre great AgNOR dot (arrow) (silver impregnation method, magnification x1000) 68 Romanian Neurosurgery (2010) XVII 1: 64 - 72 Table 2 Mean values of AgNORs in tumoral and vascular nuclei of cerebral astrocytomas according to their histological subtypes and malignancy grades AVERAGE NUMBER OF AgNORs/NUCLEUS tumoral cells vascular cells Whole tumor MALIGNANCY GRADE Range Group mean Range Group mean Range Group mean±S D Diffuse astrocytoma (G II) 2.14 2.26 2.31 2.38 2.27 1.43 1.45 1.53 1.54 1.52 1.78 1.92 1.96 2.18 1.96 Anaplastic astrocytoma (G III) 2.63 2.74 2.85 2.91 2.78 1.82 1.82 1.98 1.99 1.90 2.28 2.30 2.36 2.42 2.34 Glioblastoma multiforme (G IV) 2.47 2.68 2.79 3.12 3.76 3.83 4.02 4.15 3.35 2.54 2.72 2.72 3.60 2.77 2.86 3.04 4.00 3.03 2.50 2.75 2.92 3.14 3.34 3.59 3.39 3.88 3.18 Table 3 Differential diagnosis between glioblastoma multiforme and anaplastic astrocytoma based on mAgNOR/tumoral nucleus and mAgNOR/vascular nucleus HISTOLOGICAL SUBTYPE Anaplastic astrocytoma Small cells glioblastoma Giant cell glioblastoma NUMBER of AgNORs/ NUCLEUS Range Mean Range Mean Range Mean mAgNOR/ tumoral nucleus 2.63 2.74 2.85 2.91 2.78 2.47 2.68 2.79 3.12 2.76 3.76 3.83 4.02 4.15 3.94 mAgNOR/ vascular nucleus 1.82 1.82 1.98 1.99 1.90 2.54 2.72 2.72 3.60 2.895 2.77 2.86 3.04 4.00 3.167 Gabriela – Florenta Dumitrescu et al Diagnostic in cerebral astrocytic tumors 69 Table 4 Comparison between average values of AgNORs /nucleus in the literature and in the present study MEAN NUMBER OF AgNORs/nucleus Diffuse astrocytoma (GII) Anaplastic astrocytoma (GIII) Glioblastoma multiforme (GIV) Author tumoral nucleus vascular nucleus nucleus tumoral nucleus vascular nucleus nucleus tumoral nucleus vascular nucleus nucleus Dumitrescu G. (2009) 2,27 1,52 1,96 2,78 1,90 2,34 3,35 3,03 3,18 Janczukowic z J. (2003) [17] 1.65 2.11 2.55 Choi YM et al. (1997) [4] 1.2 ±0.26 1.90 ±0.64 1.96± 0.57 Hara A. et al. (1991) [15] 1,52±0,07 1,80± 0,13 1,98± 0,23 2,87 ± 0,50 2,05 ± 0,29 3,13 ± 1,13 Tokunaga Y. et al. (1997) [30] 2.04± 0.54 2.40±0.7 7 2.71 ± 1.13 Haberland C. et al. (1996) [14] 1,73 2,81 4,56 Tokiyoshi K. et al. (1992) [29] 1.68±0.8 7 1,85±1.0 3 2,76±1.2 6 Pedal W.P. et al. (1994) [24] 1,98 2,84 Small cells GM: 3.33 Giant cells GM: 4,24 Discussion Analysis of the AgNORs' mean number is a widely accepted method for diagnostic assessment of a wide range of tumors, both on histological and cytological preparations. In 1986, Ploton et al. [27] have used and improved the silver impregnation method brought in histopathological practice by Goodpasture and Bloom in 1975 [13]. Ploton et al. suggested that the number of AgNORs/cell correlates with cellular activity and may be an indicator of 70 Romanian Neurosurgery (2010) XVII 1: 64 - 72 malignancy, because a great number of AgNORs represents a more active proliferation, and thus a more malignant cell [27]. AgNORs have found to be markers of proliferation in brain astrocytomas because from 1990 onwards some articles showed linear correlations between the number of AgNORs, Proliferating Cell Nuclear Antigen (PCNA), binding index of Ki-67, or mitotic rate and the histological grade [20, 3, 21, 17, 4]. Several studies have been conducted on the possibility of discrimination between astrocytomas with low malignancy and those with high malignancy, using only the mean number of AgNOR dots per nucleus (mAgNOR/nucleus) [16]. The present study was designed to verify the usefulness of AgNORs determinations in discriminating histological malignancy grades of cerebral astrocytic tumors. We found a wide range of AgNORs values within each histological grade, probably reflecting that well known morphological heterogeneity and biological behaviour variations of cerebral astrocytic tumors. Nevertheless there was a linear correlation between mean values of AgNOR/nucleus and histological grade. Literature review on mean number of AgNORs in astrocytic tumors is summarized in Table 5 and data are compared with the results of the present study. It must be noted that the average values of AgNORs obtained in our study are in agreement with previous reports. All the authors obtained a gradual increase spectrum of the average number of AgNOR from diffuse astrocytomas, anaplastic astrocytoma, to glioblastomas multiforme and concluded that the number of AgNORs reflect the grade of malignancy in these neoplasias [17, 4, 15, 30, 14, 29, 24]. Janczukowicz (2003) found a statistically significant relationship between histological grade and expression of AgNOR in astrocytic tumors, but he stressed the presence of a small overlapping of extreme values between astrocytomas GII and GIII and between GIII and GIV [17]. We also observed a remarkable overlapping of the extreme values of the average number of AgNOR/tumoral nucleus between anaplastic astrocytomas (GIII) and small cells glioblastomas (GIV) probably because some of the anaplastic astrocytomas considered samples were in fact infiltration areas of a glioblastoma, the sample was too small and had no visible vessel or necrosis. Thus, AgNOR method may be particularly useful in small specimens or in biopsies from the infiltrating edge of an astrocytic tumor, where the usual histological features of malignancy may be absent. A further distinction between these two grades of malignancy could be made if mAgNOR/vascular nucleus is taken into account because none of the obtained values overlapped when comparing anaplastic astrocytomas with glioblastomas multiforme. Hara et al. (1991) calculated mAgNOR/vascular nucleus of astrocytomas grade II, III and IV (1.80±0.13, 2.87±0.50, and 3.13±1.13) and found it significantly higher than that of vascular nucleus of normal brain Gabriela – Florenta Dumitrescu et al Diagnostic in cerebral astrocytic tumors 71 without neoplastic transformation (1.26 ± 0.05) [15]. It must be noted that in the present study we did not found such great values for mAgNOR/vascular nucleus in anaplastic astrocytoma, probably because Hara et al. used 1979 WHO Histological Typing of Brain Tumors and we used 2000 WHO Brain Classification, where no vascular proliferation is needed to establish grade III of malignancy for these tumors. In the present study, the presence of increased average number of mAgNOR/vascular nucleus in anaplastic astrocytomas (GIII) and, especially, in glioblastomas multiforme (GIV) suggested that proliferative processes of tumor cells and vascular cells are interconnected. All these results showed that proliferative activity, both in tumoral cells and in vascular tumoral cells of the cerebral astrocytomas, increased with increasing grade of malignancy and therefore quantification of AgNORs is a useful marker in assessing the malignancy of these tumors. Conclusion A wide range of AgNOR values has been observed within each histological type and grade, probably reflecting variations in the biological behaviour. The number of AgNORs increased gradually from diffuse astrocytomas (GII) to anaplastic astrocytomas (GIII) and glioblastomas multiforme (GIV). Some overlaps of mAgNOR/tumoral nucleus have been detected between anaplastic astrocytomas (GIII) and glioblastomas multiforme (GIV) but the cause could be a wrong diagnostic of malignancy due to the small size of the stereotactic biopsies. The average number of mAgNORs/vascular nucleus appeared to be a more precious indicator as there was no overlap between any investigated astrocytic tumors. AgNOR method proved to be a useful method for histological grading as it supplemented the information obtained with conventional histological assessment. References 1. Bancroft JD, Gamble M. Theory and practice of histological techniques, Churchill Livingstone Elsevier, 6th edition, 2007, p. 303. 2.Barzilai A, Goldberg I, Yulash M, Pavlotsky F, Zuckerman A, Trau H, Azizi E, Kopolovic J. Silver- stained nucleolar organizer regions (AgNORs) as a prognostic value in malignant melanoma Am J Dermathopathol 1998; 20:473-7 3.Berny W, Weiser A, Markowska-Woyciechowska A, Jarmundowicz W, Zub W, Załuski R. Analysis of PCNA, Ki67, AgNOR and p53 expression in brain glial tumors. Neurol Neurochir Pol. 2004; 38(6):457-463 4.Choi YM, Kim TY, Yoon KJ, Moon HB, Kim JM. A Comparative Study on MIB-1, AgNORs and PCNA Expressions in Astrocytic Tumors. J Korean Neurosurg Soc 1997; 26(4):476-485 5.Crocker J, Murray PG (eds.), Molecular biology in cellular pathology, 2003, Wiley, England , 2nd ed., p. 147-8 6.Cucer N, Imamoglu N, Tozak H, Demirtas H, Sarac F, Atila Tatlısen A, Öztürk F. Two-dimensional agnor evaluation as a prognostic variable in urinary bladder carcinoma: A different approach via total agnor area/nucleus area per cell. Micron. 2007; 38(6):674-679 7.Derenzini M, Trerè D. Silver-stained Nucleolar Organizer Regions (AgNOR). Pathologica. 2001; 93(2):99-105 8.Ducrot P, Joundi A, Diebold MD, Pluot M. Value of the nucleolar organizers (AgNOR) in brain gliomas. Arch Anat Cytol Pathol 1991; 39(3):88-93 9.Elangovan T, Mani NJ, Malathi N. Argyrophilic nucleolar organizer regions in inflammatory, premalignant, and malignant oral lesions: A quantitative and qualitative assessment. Indian J Dent Res 2008;19(2):141-148 10. Eslami M, Baghaee F, Alaeddini M. Diagnostic value of silver–stained Nucleolar Organizer Regions in Osteosarcoma, Fibrous Dyaplasia and Ossifying Fibroma of the Jaws. Acta Medica Iranica, 2005; 43(4):243-248 72 Romanian Neurosurgery (2010) XVII 1: 64 - 72 11.Giuffrè G, Barresi V, Catalano A, Cappiello A, Stagno d'Alcontres F, Tuccari G. Actinic keratosis associated with squamous and basal cell carcinomas: an evaluation of neoplastic progression by standardized AgNOR analysis, Europ J Histochem 2008; 52(1):53-60 12.Giuffrè G, Fulcheri E, Gualco M, Fedele F, Tuccari G. Standardized AgNOR analysis as a prognostic marker in endometrial carcinoma, endometrioid type. Anal Quant Cytol Histol 2001; 23(1):31-9 13. Goodpasture C, Bloom SE. Visualization of Nucleolar Organizer Regions in mammalian chromosomes using silver staining. Chromosoma 1975; 53(1):37-50. 14.Haberland C, Martin H, Guski H, Hufnagl P, Vogel S. AgNOR analysis of astrocytoma in childhood. Zentrallbl Neurochir. 1996; 57(1):5-11 15. Hara A, Sakai N, Yamada H, Hirayama H, Tanaka T, Mori H. Nucleolar organizer regions in vascular and neoplastic cells of human gliomas. Neurosurg 1991; 29(2):211-5 16.Hashmi IU, Bukhari MH, Tayyab M, Qureshi Irfan Zia, Naseer A Chaudrhy, Sabiha Riaz. Evaluation of nucleolar organizer region in normal brain tissues and astrocytic gliomas. Biomedica 2006; 22(1):41-7 17.Janczukowicz J. The prognostic role of proliferation activity in human CNS tumors: the determination of AgNOR, PCNA and Ki-67 expression. Part 1: AgNOR expression in CNS tumors. Folia Neuropathol 2003; 41(2):97-101 18. Kidogawa H, Nanashima A, Yano H, Matsumoto M, Yasutake T, Nagayasu T. Clinical significance of double staining of MIB-1 and AgNORs in primary breast carcinoma Anticancer Res. 2005; 25(6B):3957-62 19.Kleihues P, Cavenee WK (Eds.): World Health Organization Classification of Tumours. Pathology and Genetics of Tumours of the Nervous System. IARC Press: Lyon 2000 20.Korkolopoulou P, Christodoulou P, Papanikolaou A, Thomas-Tsagli E. Proliferating cell nuclear antigen and nucleolar organizer regions in CNS tumors: Correlation with histological type and tumor grade. Am J Surg Pathol 1993; 17:912-9 21.Koyuncuoğlu M, Yücesoy K, Cingöz S, Canda MŞ, Çabuk M, Akta S. Pronostic Parameters in Astrocytoma: p53 and c-erbB2 Immunoreactivity, PCNA, and AgNOR Count. Türkish Neurosurgery 1999; 9: 85-91 22.Lumachi F, Ermani M, Marino F, Poletti A, Basso SMM, Iacobone M, Favia G. Relationship of AgNOR counts and nuclear DNA content to survival in patients with parathyroid carcinoma, Endocr Relat Cancer 2004; 11:563-569 23.Öfner D, Riedmann B, Maier H, Hittmair A, Rumer A, Tötsch M, et al. Standardized staining and analysis argirophilic nucleolar organizer region associated proteins (AgNORs) in radically resected colorectal adenocarcinoma – Correlation with tumor stage and long-term survival. J Pathol 1995; 175:427-441 24.Pedal WP, Warzok RW, Hufnagl P, Roth K. Nucleolar organizer regions (AgNORs) in astrocytic tumors. Zentralblatt Pathol, 1994, 140(1): 89-94 25.Pich A, Chiusa L. Margaria E, Prognostic relevance of AgNORs in tumor pathology. Micron. 2000; 31(2):133-41 26.Pich A, Margaria E, Chiusa L. Significance of the AgNOR in tumor pathology. Pathologica 2002; 94(1):2- 9 27.Ploton D, Visseaux – Coletto B, Canellas JC, Bourzat C, Adnet JJ, Lechki C, Bonnet N. Semiautomatic quantification of silver - stained nucleolar organizer regions in tissue sections and cellular smears. Anal Quant Cytol Histol Cytology 1992, 14(1):14-23 28.Rodriguez OR, Antonangelo L, Yagl N, Minamoto H, Schmidt AF, Capelozzi VL, Goldenberg S, Saldiva PHN. Prognostic significance of Argyrophilic Nucleolar Organizer Region (AgNOR) in Resected Non-small Cell Lung Cancer (NSCLC) Jpn J Clin Oncol 1997; 27(5):298-304 29.Tokiyoshi K, Yoshimine T, Maruno M, Nokagawa H, Iwata Y, Kuroda R, Hayakawa T. The analysis of nuclear organizer regions of astrocytomas with various histologic malignancies, No To Shinkei-Brain and Nerve, 1992, 44(12): 1083-1086 30.Tokunaga Y, Khalid H, Hiura T, Shibata S. Proliferation of Human Intracranial Neoplasm Assessed with Ki-67 (MIB-1) Labeling Index and Argirophilic Nucleolar Organizer Regions. Acta Med Nagasaki 1997; 42:44-50 31.Trerè D. AgNOR staining and quantification. Micron 2000; 31(2):127-31 32.Trerè D, Ceccarelli C, Montanaro L, Tosti E, Derenzini M. Nucleolar Size and Activity Are related to pRb and p53 Status in Human Breast Cancer, J Histochem Cytochem 2004; 52:1601-1607