PTU - Polskie Towarzystwo Urologiczne

Artykuł opublikowany w Urologii Polskiej 2000/53/1.


Sławomir Chaberek 1, Agnieszka D. Kolasińska 2, Kerstin Rolfe 2, Marek Żabiński 3, Danuta Shafie 3, Gidon Lieberman 2, Jarosław Ćwikła 1
1 Department of Radiology and Nuclear Medicine, SPSK 2, Otwock
Head of Department: Dr. J. Szawdyn
2 Academic Department of Obstetrics and Gynaecology Royal Free Hospital, London, UK
Head of Department: Prof. R. McLean
3 Department of Pathology and Urology, Departmental Hospital, Olsztyn
Head of Department: Dr. J. Gugała and Dr. M. Żabiński

słowa kluczowe

nerka rak nerki cyklina D1 immunohistochemia analiza obrazu


Cel pracy. W badaniu wstępnym określono zależność między ekspresją cy-
kliny D1 a morfologicznymi czynnikami rokowniczymi oraz czasem przeżycia
chorych na raka nerki (ang. RCC - renal cell carcinoma).
Materiał i metoda. Materiał biologiczny tkanki guzów został pobrany pod-
czas nefrektomii od 12 chorych (średnia wieku 57 lat, SD 5,43; stosunek
mężczyzn do kobiet wynosił 1:1). Wszyscy chorzy mieli potwierdzony w ba-
daniu histologicznym rak nerki. Ekspresję cykliny D1 oznaczono za pomocą
metody immunohistochemicznej (ang. IHC), gdzie powyżej 20% wybarwio-
nych jąder komórkowych w badanym preparacie oznaczało wynik dodatni.
Obliczenia dokonano w sposób ręczny oraz automatyczny. Rezultaty IHC były
następnie korelowane z wybranymi czynnikami rokowniczymi RCC (wiek cho-
rych, wielkość guza, skala oraz skala Robsona), Średni czas obserwacji ba-
danych chorych wynosił 80,2 miesiąca (SD 7,36).
Wyniki. Histologicznie stwierdzono 8 RCC typu jasnokomórkowego, 2 typu
mieszanego oraz 2 raki typu brodawkowego. Ekspresję cykliny D1 stwierdzono
u 9 chorych. Nie stwierdzono różnic w ekspresji cykliny D1 pomiędzy określony-
mi typami histologicznymi RCC, skalą TNM oraz skalą Robsona. Nie stwier-
dzono zależności pomiędzy ekspresją cykliny D1 oraz wybranymi czynnikami
rokowniczymi. Dodatkowo nie stwierdzono różnic w ekspresji cykliny D^mię-
dzy chorymi, którzy zmarli z powodu choroby a tymi, którzy przeżyli (p = 0,29).
Wnioski. Analiza ekspresji cykliny D1 w RCC za pomocą IHC jest prostą i
powtarzalną metodą; obliczenia można dokonywać w sposób ręczny lub au-
tomatyczny. Wstępne badania wskazują, że większość guzów nerki wykazu-
je ekspresję cykliny Dr Nie stwierdzono zależności między ekspresją cykliny
D1 a wybranymi czynnikami rokowniczymi RCC.

Renal cell carcinoma (RCC), which originates from proximal tubular
epithelium, continues to be a one of the most frustrating tumours for
clinicians to manage and treat [8,19]. The biological behaviour of RCC is
unpredictable when evaluated by conventional histology [8,14,19]. The
analysis of molecular factors including proliferation markers, karyotype,
oncogenes, cell adhesion molecules and proteases are areas of intense
investigation and may provide mechanisms for identifying patients who
require more (or less) aggressive treatment [8,14,19].
In order to establish appropriate therapeutic modalities for RCC, it is
essential to collect information about factors affecting cancer cell biology
[14]. A coordination of processes results in a series of changes (phase
transitions) in the activity of intracellular enzymes, which are known as
cyclin-dependent kinases (cdks). The active forms of these enzymes ex-
ist in protein complexes which include a cell-cycle stage-specific protein
known as a cyclin, and other proteins such as proliferating cell nuclear
antigen (PCNA) and a cdk inhibitor [19]. Regulated degradation of the
cyclin proteins also contributes to this control mechanism. Cyclin, when
present, can complex with a cdk, and the cdk in this complex is activated
by regulated kinase activity (CAK, cdk-activating kinase). The activated
cdk then phosphorylates target gene products required for the transition
of the cell to the next phase of the cycle [8,19]. Analysis of cell regulatory
proteins, which trigger cell division and cancer progression, is necessary
to understand tumour biology and has prognostic value in some cancers
[8, 15, 19]. Recent data indicates that tumour proliferation, if assessed
using Ki-67 antigen expression, is inversely associated with survival in
patients with RCC [12]. Currently there are prognostic factors that are
considered to be significant when evaluating patient survival [20]. Most
importantly seems to be microvessel count in patients with low-stage
tumours, macro venous invasion, nuclear grade and tumor size [3, 18,
21]. Therefore, the aim of this study is to evaluate the correlation be-
tween cyclin D1 expression with morphological characteristics and out-
come in patients. Additionally we have evaluated the role of cyclin D1
expression as a tumour proliferation marker compared to the survival
rates of patients who have undergone radical nephrectomy.
Patients. There were 12 patients enrolled in this study. The mean age
of the patients was 57.3 years (SD 5.33, age range of 45 to 66, male to
female ratio of 1:1). Tumours were staged based on the Robson modifica-
tion of the Flocks and Kadesky system [13] and the TNM classitication
[8,14]. The mean follow-up of all patients was 30.2 months (SD 7.36). All
patients had histopathological confirmation of RCC and were treated by
conventional radical nephrectomy. Haematoxilin and eosin staining was
performed in each case to assess tumour type and histological grading.
Immunohistochemistry. Biopsied tissue was remounted from bee’s
wax to paraffin. 5-micron thick sections were then cut and mounted on
3-aminopropylthethoxysdane-coated glass slides. We employed a stan-
dard immuno-peroxidase technique using a streptavidinbiotin/horse-
radish-peroxidase complex (DAKO, CA, USA). 3.3-diaminobenzidine
tetrahydrochloride (DAB) acted as the chromage. Following the block-
ing of non-specific peroxidases antigen Retrieval was performed by mi-
crowaving the sections in a citric buffered solution. The primary anti-
body (anti-CCNDJ was diluted to 1: 20 concentration and the sections
were left to incubate overnight at room temperature. The sections were
then incubated for one hour in the secondary antibody (biotinylated rab-
bit anti-mouse), followed by streptavidin/horseradish peroxidase. Be-
tween each stage the sections were washed in phosphate buffered saline.
DAB was applied as the chromagen and then the sections were counter-
stained in Mayes haemillum and mounted. Sections were incubated in
20% rabbit serum, then in anti-CCND1 diluted 1: 20 in PBS and SAB, fol-
lowed by biotynilated rabbit anti-mouse IgG diluted to 1: 400 concentra-
tion (in 3% normal serum). The streptavidin-biotin method (DAKO; CA,
USA) diluted in 1 in 200 in TRIS (Sigma-Aldrich, MO, USA) followed by
a counterstain (diaminobenzidene) demonstrated cyclin D1. Positive and
negative controls were used to compare biopsied tissue. Positive control
was a breast cancer sample known to stain positively for cyclin D1, and
omitting the primary antibody produced a negative control. In each case
a further control was performed to measure cyclin D] expression in areas
of healthy kidney separate from kidney with tumour mass. All blocks
were stored for up 36 months and there appeared to be no appreciable
deterioration in cyclin D1 immunohistochemistry.
Quantification of staining. Cyclin DT expression was evaluated using
light microscopy at x 400 magnification. Staining for cyclin D1 was scored
according to the braun colour appearance in the nuclei of the stained cell.
The percentage of tumour cells with nuclear staining was measured by
counting at least overall 250 tumour cells in five selected fields that dis-
played the highest staining. Cytoplasmic staining was sometimes present,
however, not assessed. Three readers independently evaluated the sam-
ples. When there was a disagreement a majority decision was taken. Ad-
ditionally, staining was evaluated using an automated technique. Image
analysis was performed using a “MulitiScan TM” system for digital imag-
ing and analysis..Images were converted to a digital format using a stan-
dard Hewlett-Packard scanner, which had a resolution 300 dpi (dots per
inch) – horizontal and vertical, in a true colour (24-bit RGB) bitmap. All
images were corrected using gamma correction and contrast normaliza-
tion. After elimination of artefacts, images were converted to a monochrome
display. Monochrome images were filtered using a median filter. The
prepared image converted to binary to extract all cells from the image.
Cells were counted using an automatic count mode. Results were writ-
ten in a “Muliti Scan TM” report. Braun signet cells were counted. The
MulitiScan TM” report was exported to “Excel” (Microsoft CA; USA).
The mean and SD of the manual and automated techniques were calcu-
Statistic. The Pearson correlation coeffcient and levels of significance
between both manual and automated counting were performed (p < 0.05
significant). The difference between the histological subtypes were per-
formed using the one sided T-test for unequal variance. Non-parametric
statistics (Kendal-tau correlation coeffcient) were used to evaluate the
correlation between cyclin D1 expression and prognostic factors. All sta-
tistics were performed using Statistica 5.1 software package (StatSoft Inc.
The histological subtypes were 8 clear cell, 2 granular and 2 papillary
renal carcinomas. According to the Robson classification there were 2
stage I, 3 stage II, 5 stage III, and 2 stage IV carcinomas. According to the
TNM classification there were 2 stage 1,4 stage II, and 6 stage III carcino-
mas. Cyclin D1 expression was detected in 9 patients, where more than
20% of the nuclei (from 250 cells) were positive (Fig. 1), calculated both
manually and automatically.
Two patients demonstrated an absence of staining and one patient had
weak staining (Fig. 2). The mean value of positive cyclin D: nuclei
evaluated manually was 0.38 (SD 0.3) whilst using the automated tech-
niąue was 0.47 (SD 0.29). There was good correlation between both meth-
ods (Fig. 3) (Pearson correlation coerficient r = 0.832, p = 0.001). 9 control
studies were performed. Sections of uninvolved kidney demonstrated
expression in the proximal tubules but no expression in the glomeruli
(Fig. 4). The clinical, histological and cyclin D1 expression data of all pa-
tients are summarized in Table I.
There was no difference between cyclin D: expression and histological
type of the tumour (p > 0.05, one sided T-test for uneąual variance). There
was no significance between cyclin DT expression in different grades of
RCC tumours according to the TNM or Robson classification (v > 0.05T,
one sided T-test for uneąual variance). A relationship (p > 0.05, Kendal-
tau correlation coefficient) was noted between cyclin D expression and
Prognostic factors of RCC (tumour size, age of patients, TNM or Robson
grading score). There was no difference between cyclin D expression in
patients who died and those who survived (p = 0.29 one sided T-test for
uneąual variance).
In this initial study we have evaluated the correlation between cyclin
D: expression and some prognostic factors in patients with renal cell car-
cinoma (RCC). As previously reported when mitogenic signals stimulate
a cell to enter the cycle from ąuiescence, the expression of D-type cyclins
(DT, D2, and D3) are stimulated and continue through G [19] as long as
growth factor supply is present [8,10,19]. This regulation of D-type cy-
din levels by extracellular signals is unique and is not observed with the
other cyclin proteins. Authors, using immunohistochemistry and west-
ern blot method, have found that 59% of RCC overexpress cyclin D1 [5].
These results are similar to our study, which have indicated that 67% of
RCC express cyclin Dr There is dispute concerning the relationship be-
tween proliferation makers and histological grade. Some authors have
round no relationship between the proliferation rate evaluated by as-
sessment of Ki-67 antigen expression or PCNA (proliferation cell nuclear
antigen) and the histological grade or TNM stage [1,11]. Others indicat-
ed no correlation between PCNA also Ki-67 antigen expression and final
outcome of patients [3, 4].
In contrast Jachum et al. and Tanapfel et al. who have analysed prolifer-
ation rate in RCC evaluated as Ki-67 antigen expression (MIB1 antibody)
have indicated significant correlation between Ki-67 expression and tu-
mour grade and final outcome of patients with RCC [2,7,17]. The tumour-
specific proliferation rate obtained by Ki-67 and PCNA as well labelling
seems to be an independent marker to describe the proliferate activity and
aggressiveness of individual tumours [6,12,16]. Current data has indicat-
ed that well differentiated carcinomas (G I) showed statistically lower pro-
liferate activity than moderate and poorly differentiated carcinomas (G
11/G III) [6,16]. In our initial study in a relatively small population of pa-
tients, we found a tendency toward a positive relationship between cyclin
D1 expression and Robson’s classification (p = 0.09).
Larson et al. using multivariate analysis showed both the S-phase frac-
tion and tumour stage gave significant independent prognostic informa-
tion. Also additional showed that PCNA expression indicated group of
patients with non-metastatic disease, with better prognosis, low PCNA
and worst, those with high PCNA index [9]. There are several possible
explanations for the variability of these results. First there are several
types of mutations and only some of them are associated with increased
tumour proliferation. A single expression of cyclin D1 is not sufficient to
trigger cell proliferation and cancer progression.
The diagnostic and prognostic value of cyclin D1 in RCC has not been
establish. Our initial study showed that the presence of cyclin D: expres-
sion did not correlate with standard prognostic factors of RCC. Never-
theless, the frequent association of cyclin D1 overexpression in RCCC ap-
pears to reflect increased the number of cells entering G1 [5, 8,19].
In conclusion, our results indicate that most of RCC tumours overex-
press cyclin D1, also his study shows that analysis of cyclin D: expression
in RCC by immunohistochemistry is a simple and easily reproducible
technique, which can be both manually and automated evaluated. Ex-
pression of cyclin D1 is not the best indicator for RCC progression alone.
These results suggest that cyclin Dt expression alone is not a completely
reliable prognostic indicator for RCC. Further longer study should be
performed to evaluate role of cyclin D1 as an additional prognostic factor
of patients with RCC.


  1. [1] Cronin, K. J., Williams, N. N., Kerin, M. J., Creagh, T. A., Dervan, P.
  2. A., Smith, J. M., Fitzpatrick, J. M.: Proliferating cell nuclear antigen: a new
  3. prognostic indicator in renal cell carcinoma. J. Urol. 1994,152, 834-836.
  4. [2] Delahunt, B., Bethwaite, P. B., Thornton, A., Ribas, J. L.: Proliferation of
  5. renal cell carcinoma assessed by fixation-resistant polyclonal Ki-67 antibody label-
  6. ing. Correlation with clinical outcome. Cancer 1995, 75, 2714-2719.
  7. [3] Gelb, A. B., Sudilovsky, D., Wu, C. D., Weiss, L. M., Medeiros, L. J.:
  8. Appraisal of intratumoral microvessel density, MIB-1 score, DNA content, andp53
  9. protein expression as prognostic indicators in patients with locally confined renal
  10. cell carcinoma. Cancer 1997, 80,1768-1775.
  11. [4] Grignon, D. J., Abdel-Malak, M., Mertens, W., Koster, J., Keeney, M.,
  12. Sakr, W., Shepherd, R. R.: Prognostic significance of cellular proliferation in
  13. renal cell carcinoma: a comparison of synthesis-phase fraction and proliferating cell
  14. nuclear antigen index. Mod. Pathol. 1995,1,18-24.
  15. [5] Hedberg, Y., Davoodi, E., Roos, G., Ljungberg, B., Lanberg, G.: Cyclin
  16. D1 overexpression and cell cycle defects are common in human renal cell carcinomas.
  17. Proc. Amer. Cancer Soc, Suppl. Cancer Res. 1998, 3013.
  18. [6] Hindermann, W., Berndt, A., Wunderlich, H., Katenkamp, D., Kos-
  19. mehl, H.: Quantitative evaluation of apoptosis and proliferation in renal cell carci-
  20. noma. Correlation to tumour subtype, cytological grade according to thoenes-classi-
  21. fication and the occurrence of metastasis. Pathol. Res. Pract. 1997,193,1-7.
  22. [7] Jochum, W., Schroder, S., al-Taha, R., August, C, Gross, A. J., Berger,
  23. J., Padberg, B.C.: Prognostic significance of nuclear DNA content and prolifer-
  24. ative activity in renal cell carcinomas. A clinicopathologic study of 58 patients us-
  25. ing mitotic count, MIB-1 staining, and DNA cytophotometry. Cancer 1996, 77,
  26. 514-521.
  27. [8] de Kernion, J, B.: Renal tumors,In:PC. Walsh (ed.): Campbell's urology. 5th ed.,
  28. vol. 2, P. Sounders, Philadelphia 1986,1294-1342.
  29. [9] Larsson, P., Roos, G., Stenling, R., Ljungberg, B.: Tumor-cell proliferation
  30. and prognosis in renal-cell carcinoma. Int. J. Cancer 1993, 55, 566-570.
  31. [10] Matsushime, H., Roussel, M. F., Ashmun, R. A., Sherr, C. J.: Colony-
  32. stimulating factor 1 regulates novel cyclin during the G1 phase of the cell cycle. Cell
  33. 1991, 65, 701.
  34. [11] Papadopoulos, I., Weichert-Jacobsen, K., Sprenger, E., Wacker, H. H.:
  35. DNA content and the proliferation marker Ki-67 as prognostic indicators in renal
  36. cell carcinoma. Urol. Int. 1994, 53,181-185.
  37. [12] de Riese, W. T., Crabtree, W. N., Allhoff, E. P., Werner, M., Liedke, S.,
  38. Lenis G., Atzpodien, J., Kirchner, H.: Prognostic significance of Ki-67 im-
  39. munostaining in nonmetastatic renal cell carcinoma. J. Clin. Oncol. 1993,11,1804-
  40. 1808.
  41. 13] Robson, C, Churchill, B. M., Anderson, W.: The results of radical nephrecto-
  42. my for renal cell carcinoma. J. Urol. 1969,101, 297-301.
  43. 14] Stenzl, A., de Kernion, J.: Pathology, biology and clinical staging of renal cell
  44. carcinoma. Semin. Oncol. 1989,16, 3-11.
  45. [15] Tan, P., Cady, B., Wanner, M., Worland, P., Cukor, B., Magi-Galluzzi,
  46. C, Lavin, Ph., Draetta, G., Pagano, Loda, M.: The cell cycle inhibitor p27
  47. is an independent prognostic marker in small (Tlab) invasive breast carcinomas.
  48. Cancer Res. 1997, 57, 1259-1263.
  49. [16] Tanioka, F., Hiroi, M., Yamane, T., Hara, H.: Proliferating cell nuclear an-
  50. tigen (PCNA), immunostaining and fiow cytometric DNA analysis of renal cell
  51. carcinoma. Zentr. Pathol. 1993,139,185-193.
  52. [17] Tannapfel, A., Hahn, H. A., Katalinic, A., Fietkau, R. J., Kuhn, R.,
  53. Wittekind, C. W.: Incidence of apoptosis, cell proliferation andp53 expression in
  54. renal cell carcinomas. Anticancer Res. 1997, (2A) 1155-1162.
  55. [18] Usubutun, A., Ayhan, A., Uygur, M. C, Ozen, H., Toklu, C, Ruacan, S.:
  56. Prognostic factors in renal cell carcinoma. J. Exp. Clin. Cancer Res. 1998,17,77-81.
  57. [19] DeVita, V. T.: Principles and practice of oncology. V. T. DeVita, S. Hellman, S.
  58. A. Rosenberg. Lippincrott (eds.). Philadelphia 1989, 276-300.
  59. [20] Yasunaga, Y., Shin, M., Miki, T., Okuyama, A., Aozasa, K.: Prognostic
  60. factors of renal cell carcinoma: a multivariate analysis. J. Surg. Oncol. 1998, 68,
  61. 11-18.
  62. [21] Yoshino, S., Kato, M., Okada, K.: Evaluation of the prognostic significance of
  63. microvessel count and tumor size in renal cell carcinoma. Int. J. Urol. 1998,5,119-
  64. 123.