Fluorescence in Situ Hybridization Identifies More Aggressive Types of Primarily Noninvasive (Stage pTa) Bladder Cancer

00226347/97/15762116$03.00/0

v01.157,2116-2119,

June 1997 Printed in U.S.A.

Ttnr JOURNALOF U R o m Y Capyright 0 1997 by AMERICAN UROLOCICAL ASSOCIATION,INC.

FLUORESCENCE IN SITU HYBRIDIZATION IDENTIFIES MORE AGGRESSIVE TYPES OF PRIMARILY NONINVASIVE (STAGE pTa) BLADDER CANCER ARMIN PYCHA, CHRISTINE MIAN, ANDREA HAITEL, JOHANN HOFBAUER, HELENE WIENER AND MICHAEL MARBERGER From the Departments of Urology and Clinical Pathology, University of Vienna, Vienna, Austria

ABSTRACT

Purpose: We evaluated the genetic changes in cytological specimens of bladder cancer by fluorescence in situ hybridization, and related them to stage and grade of the tumor, ploidy, p53 and Ki-67 expression, and clinical outcome to determine a simple method to identify tumors with a poorer prognosis. Materials and Methods: Using fluorescence in situ hybridization the numerical aberrations of chromosomes 7 , 9 and 17 in barbotages and imprints of 50 patients with transitional cell cancer of the bladder were determined. Of the patients 29 had a primary stage pTa tumor, while 21 with stage pT1 or greater disease formed the control group. Data were compared to ploidy status, and Ki-67 and p53 immunoreactivity. Results: Repeated monosomy 9 and haploid or diploid status on static ploidy determination were found in patients with primary stage pTa tumors without recurrence. Immunoreactivity of pi53 was negative in all of these patients, while there was a low percentage of positive staining for Ki-67. Patients with recurrent and progressive disease had a high incidence of trisomy 7 and 17, aneuploid status and high positivity for both immunological markers. For chromosomes 7 and 17, and ploidy status bivariate analysis showed a significant difference. Conclusions: The evaluation of chromosomal aberrations in barbotage and imprint specimens clearly establishes a relationship between chromosomal defects and aggressiveness of the tumor. The majority of nonaggressive stage pTa transitional cell carcinomas can be distinguished from potentially lethal cases by fluorescence in situ hybridization a t a diagnostic point when the grading is not yet prognostic. KEY WORDS: bladder neoplasms, fluorescence, neoplasm staging, protein p53

Most (70 to 75%) bladder tumors are of low grade and noninvasive. Despite a generally good prognosis 70% of these patients have recurrent disease,' and 6 to 10%have progression finally leading to death.2 This completely different course between cases with a single occurrence and those with lethal progression of initially apparent histologically identical tumors is the reason for ongoing interest in prognostic factors. A starting point in evaluation was cytogenetic analysis soon after genetic aberrations were described in bladder cancer.3.4 Partial deletions of chromosome 9,3-6 and numerous aberrations of chromosomes 7, 94.7.8and 173.8-10were defined as the most important alterations in transitional cell cancer. Today tumor is considered to result from a multistep genetic alteration, the details of which are still controversial. The increased expression of ~ 5 3 ~ . ' O - ' 3and Ki-6714.15 in transitional cell carcinoma is well documented. Most of these parameters were evaluated on histological material, which required surgical intervention and sufficient tissue. We evaluate the genetic changes in cytological material by fluorescence in situ hybridization, and relate them to stage and grade of the tumor, ploidy, p53 and Ki-67 expression of the tumor cells, and clinical outcome. Using fluorescence in situ hybridization on barbotage and intraoperative imprint specimens it should be possible to develop a simple method to identify tumors with a poorer prognosis.

MATERIALS AND METHODS

We studied prospectively 18 women and 32 men 39 to 85 years old (mean age 68) with histologically verified transitional cell carcinoma of the bladder, which was staged and graded according to the 1992 classification of the International Union Against Cancer.16 A mean followup plus or minus standard deviation was 22 t 2 months, and included cystoscopy every 3 months. In 29 patients (group 1) the primary tumor was stage pTa and grade 1in 1 9 , 2 in 9 and 3 in 1. The 2 1 patients (group 2) with stage pT1 or greater disease form the control group. Detailed data are given in the table. All cytogenetic studies were done on intraoperative barbotage specimens and cytological imprints of the resected material. All barbotage specimens were collected by irrigating the bladder with 50 ml. saline solution before transurethral resection with the patient under anesthesia. Before placing the surgical specimens into 37% formaldehyde the resected material was rolled on or softly pressed against a glass slide. The procedure did not negatively influence the histological evaluation of the biopsy. Cytological preparation, fluorescence in situ hybridization analysis, deoxyribonucleic acid (DNA) image analysis and irnmunocytochemistry. Barbotage specimens were spun twice at 350 x gravity for 5 minutes using a sedimentation conus for the second spin. Intraoperatively obtained touch preparations (imprints) were spray fixed immediately after collecAccepted for publication December 6, 1996. Sup rted in part by Grant 1938, Biirgermeisterfonds der Stadt tion. The cell content was examined by staining 1 slide Wien,%enna, Austria. according to the Papanicolaou method. The remaining cyto2116

FLUORESCENCE IN SITU HYBRIDIZATION IN DIAGNOSIS OF NONINVASIVE BLADDER CANCER

2117

Histopathologlcal findings, clinical status, imrnunocytochemical and ploidy data. and chromosomal aberrations in rmprcnts and bladder washrrlgs of 29 patients with resected stage pTa (group I ) and 21 wLth resected stage pTI or greater (group 2) transrtronal cell carcinoma --Ir

Histopathological StageIGrade

Clinical Status

lmmunocytochemist~ P53

pTdl pTdl pTdl pTdl pTa/l pTdl pTdl pTdl pTdl pTdl pTdl pTdl pTdl pTdl pTdl pTd2 pTd2 pTd2 pTd2 pTd2 pTd2 pTd2 pTdl pTd1 pTdl pTdl pTd2 pTd2 pTd3

*

pTlll pT1/1 pT1/2 pTx/l pT2/3 pT112 pT112 pT113 pT213 PW2 pTW3 PW3 pTlll pTU2 pTW1 PW3 pTY3 pTU3 pTU3 pTY3 pTU3 Findings in barbotage

No recurrence No recurrence No recurrence No recurrence No recurrence No recurrence No recurrence No recurrence No recurrence Recurrence Recurrence Recurrence Recurrence Recurrence Recurrence Recurrence Recurrence Recurrence Recurrence Recurrence Recurrence Recurrence Progression Progression Progression Progression Progression Progression Progression Recurrence Recurrence Recurrence Recurrence Recurrence Progression Cystectomy Recurrence Recurrence Cystectomy Cystectomy Cystectomy Recurrence Cystectomy Recurrence Cystectomy Recurrence Progression Recurrence Promession ProGession specimens only.

Ki-67 20 15 15 10 -

20

-

22 10 15

60 55 6 20 8 50 30

17 50 34 28 52

32

60

30

25 24 57

45 55 30 25 54 25 30 14 50 57

-

35 40 50

Numerical Aberrations

ploidy Chromosome 7

Group 1 Haploid Haploid Haploid Diploid Diploid Haploid Haploid Haploid Diploid Haploid Aneuploid Aneuploid Aneuploid Aneuploid Aneuploid Aneuploid Aneuploid Aneuploid Aneuploid Aneuploid Aneuploid Aneuploid Aneuploid’ Aneuploid Aneuploid Aneuploid heuploid* Aneuploid‘ Aneuploid Group 2 heuploid* heuploid* Aneuploid Aneuploid Aneuploid Aneuploid Aneuploid Aneuploid Aneuploid Aneuploid Aneuploid Aneuploid Aneuploid Aneuploid Aneuploid Aneuploid Aneuploid Aneuploid Aneuploid Aneudoid Aneuploid

logical specimens were bases for the determination of ploidy, immunocytochemical evaluation of Ki-67 and p53, and detection of numerical chromosomal aberrations. Using fluorescence in situ hybridization the specimens were evaluated for numerical aberrations of chromosome 7 , 9 and 17. The protocol followed in principle the method developed by Pinkel et al.” A slight modification was made to shorten procedure time. The slides were washed in 50% ethanol and phosphate buffered saline, and denatured in formamide 20 x salindsodiurn citrate a t 75C for 5 minutes. The directly labeled DNA target was dehydrated in an ice-cold ethanol series and air dried. Hybridization was performed for 30 minutes at 42C. Then slides were washed twice in saline./ sodium citrate solutions of different concentrations. Counterstaining was achieved by application of 10 ~ l4’6-diamidino. 2-phenylindole-2 hydrochloride per slide. For each evaluation E p o t s of 200 cells were counted using X600 magnification (oil immersion) and images were captured by a Smart Capture

None None None None None None None None None None +7 +7 None None None +7 +7 +7 +7 +7 None

+I +7* +I* +7 +7 + 7* +7*

+I +7’ i7’ None +7 +7 +7 +7 +7 None +7 +7

+I None None +7 +7 +7 +7 +7 +7

+I

Chromosome 9 -9

-9 -9 -9 -9

-9 -9 -9 -9 -9 None

-9 -9 -9 +9 None -9 -9

-9 -9 -9 -9 -9 -9 None -9 -9 -9 None -9 -9 None -9 -9 -9 None -9 -9 None None None -9

-9 None None None None -9 -9 None

Chromosome 17 None None None None None None None None None None None None + 17 + 17 None None None + 17 + 17 None + 17 + 17 None + 17* + 17 + 17 None None + 17 None None + 17 None + 17 None None + 17 + 17 + 17 + 17 + 17 + 17 + 17 + 17 + 17 None + 17 + 17 + 17 None

System* mounted on a fluorescence microscope. Using lymphocytes of healthy probands as controls the sensitivity of the method was proved by assuring diploidy in the majority of cells. According to previous data the cutoff level for true aneusomy in bladder washings and imprints was established at 20% for chromosome 7, 9 and 17, respectively.? For evaluation of the ploidy status cyto-spin slides and imprints were fixed in Bohm’s solution and stained according to a slightly modified Feulgen-Schiff method. Briefly, hydrolyzation with 5 N. hydrochloric acid was done for 1 hour at room temperature instead of a few minutes in a warm water bath. DNA was measured by image analysis allowing an interactive selection of cells. ARer calibrating with cells from a cyto-spin preparation of a bladder washing from patients without any bladder malignancy, at least 200 urothelial cells were measured. Special attention was paid to avoid degenerative cells and cells partly or totally overlapping. Simple

* Vysis, Inc., Downers Grove, Illinois.

2118

FLUORESCENCE IN SITU HYBRIDIZATION IN DIAGNOSIS OF NONINVASIVE BLADDER CANCER

algorithms were used to describe histograms and to summarize results as diploid or aneuploid. In concordance with previous observations, a slight right shoulder without any additional peak was considered as representing a cycling population, sometimes also including a small peak in 4C and some cells in 8C.19 p53 and Ki-67.Detection of p53 and Ki-67 was based on immunocytochemistry using the peroxidase-antiperoxidase reaction. For the p53 staining protocol specimens were incubated with 1:40 diluted monoclonal antibody to p53 suppressor gene. MIB-1 monoclonal antibody was used for Ki-67 detection. Staining protocol demanded a 1:30 dilution of the antibody. Before reaction endogenous peroxidase was blocked by 3% hydrogen peroxide. Aminoethylcarbazole was used as chromogene and 200 cells were counted by light microscopy at a magnification of X400. Positive results were reported as percent of the cells counted. Statistical evaluation was performed by bivariate analysis using the Pearson chi-square and Kruskal-Wallis tests. RESULTS

Of the 29 patients with primary stage pTa tumors 9 (31.0%) remained free of recurrence, 13 (44.8%)had recurrent transitional cell carcinoma of the same grade and stage, and 7 (24.1%)had progression to stage T1. Of the control patients 11 (52.4%)had recurrent tumor and 4 (19.0%)had progression. Six patients with primarily stage pT2 transitional cell carcinoma underwent cystectomy after the initial transurethral resection. The table shows the results of the cytogenetic analysis of imprint and barbotage specimens from the 29 patients with stage pTa transitional cell carcinoma (group 1). Data are given in relation to clinical followup, ploidy status, and p53 and Ki-67 positive results. All patients who did not have a recurrent tumor within 24 months had grade 1 disease with a consistent finding of singular monosomy 9 in imprint and barbotage specimens (part A of figure). No positivity for p53 was noted in any cytological specimens, and the status was described as diploid or haploid. The immunoreactivity of the proliferation marker Ki-67 ranged from 0 to 22%(mean 12.4). Several numerical chromosomal aberrations were found in patients with recurrent tumors of the same stage and grade. In addition to the predominant findings of monosomy 9 (10 of 13 cases) trisomy of chromosomes 7 and 17 was found in 62 and 46%of the cases, respectively. Bivariate analysis of all parameters revealed significant differences in nonrecurrent and recurrent tumors for ploidy status (p <0.00001), and chromosomes 7 (p = 0.00024) and 17 (p = 0.02857). All

specimens with trisomy 17 were positive for p53 (range 5 to 38%).Image analysis revealed aneuploid histograms in 12 of 13 cases. The proliferation activity ranged from 0 to 60%. Mean Ki-67 positivity was 18.8%,which was greater than in the group without recurrence. Patients with recurrent tumors of higher stage and/or higher grade were considered to have progressive disease. Cytological specimens were characterized particularly by trisomy of chromosome 7 on imprint and barbotage or barbotage specimens only. Trisomy 7 was accompanied by monosomy 9 in 5 cases and by trisomy 17 in 4. Immunocytochemical p53 positivity of 8,lO and 50% was reported for only 3 cases. Aneuploidy was noted in all cases on imprint and barbotage or barbotage specimens only. Proliferation activity as shown by Ki-67 positivity ranged from 0 to 60% (mean 24.9%). The 21 control patients with histological stage pT1 disease or greater (group 2) showed a heterogeneous chromosomal pattern (see table). Evaluation of specimens revealed monosomy of chromosome 9 in 11 patients (52%).The incidence of numerical aberrations in chromosome 9 decreased compared to group 1 and was minimally less than previously reported data.20 The predominant finding was trisomy of chromosome 7 in at least 1of the specimens investigated in 81%of cases and trisomy of chromosome 17 in 67% (part B of figure). Static ploidy determination resulted i n aneuploid histograms in all cases, p53 positivity counted on 200 cells ranged from 0 to 80%(mean 19) and mean Ki-67 immunoreactivity within this group was 29.23%(range 0 to 57%). DISCUSSION

Oncogenesis of transitional cell carcinoma seems to follow 2 different pathways.5 Changes in chromosome 9, on which a tumor suppressor gene is postulated, appear to be responsible for superficial bladder tumors. Stage pTa tumors with aberrations in chromosome 9 are postulated to recur often but seldom to progress. The addition of a p53 defect in the aberration of chromosome 9 triggers progression.5 Our patients with low grade noninvasive transitional cell carcinoma had a highly heterogeneous clinical outcome. Of all stage Ta grade 1tumors analyzed 21%progressed, which is twice the reported progression rates. Patients were not preselected so that the high percentage is unexplainable. Since no recurrence was noted during the first 5 months after transurethral resection there is no reason t o suspect a residual tumor. Nevertheless, clearly all stage pTa tumors cannot be considered harmless. Patients with tumors chromosomally characterized by singular monosomy 9 had a n overall

Fluorescence in situ hybridization analysis of imprint specimen of bladder cancer shows centromere 9 (red), centrornere 7 (green) and hydrochloride counterstaining (blue).A, group of cells with 2 centromere 7 and centromere 17 (yellow) with 4’6-diamidino-2-phenylindole-2 17 si als, respectively, but only 1 centromere 9 signal. B , group of cells with 2 centromere 9 signals but 3 or more centromere 7 and 17 signar.

FLUORESCENCE IN SITU HYBRIDIZATION IN DIAGNOSIS OF NONINVASIW BLADDER CANCER

good prognosis and were recurrence-free until the end of followup a t 24 months. However, if the chromosomal pattern revealed an additional trisomy 7 the behavior of the tumor appeared to change. Among patients with a recurrent tumor of the same stage and grade as the primary trisomy 7 was found in 61% of the primarily resected specimens. The incidence of trisomy 7 increased to 100% in patients whose recurrent tumors were of higher stage and grade compared to the primary lesions. Waldman et a1 suggested that there is a direct correlation between an increased number of chromosome 7 and aggressive tumor behavior.7 In fact, there should be genes expressed on chromosome 7 that promote tumor proliferation insofar as they exist in increased quantities. Actually the stage pTa tumors with a n aberration of chromosome 7 demonstrated increased expression of the proliferation marker Ki-67 and a greater incidence of aneuploidy. The chromosomal and ploidy status determined by image analysis in progressive stage pTa tumors did not differ s i d cantly from the results found in the control group of histological stage pT1 or greater transitional cell cancers. In contrast, immunocytochemical positivity of p53 and the proliferative activity described by Ki-67 positivity vaned among the later groups. CONCLUSIONS

Our data underline the significance of evaluating the chromosomal pattern in superficial low grade transitional cell carcinoma and at the same time confirm the postulated pathway in the oncogenesis to transitional cell carcinoma.6 Evaluation of the chromosomal aberrations in barbotage and imprint specimens clearly establishes a relationship between chromosomal defects and tumor aggressiveness. The majority of nonaggressive stage pTa papillary transitional cell carcinomas can be distinguished from those less frequent cases that are potentially lethal by fluorescence in situ hybridization, a simple and efficient method, at a diagnostic point when the grading is not yet prognostic. With the help of this new method it might be possible to decrease the 6 to 10% rate of potentially dangerous stage Ta transitional cell carcinoma in a lasting manner. REFERENCES

1. Soloway, M.: Managing superficial bladder cancer: a n overview. Urology, suppl., 40:5, 1992. 2. Holmang, S., Hedelin, H., Anderstrom, C. and Johansson, S. L.: The relationship among multiple recurrences, progression and prognosis of patients with stages TA and T 1 transitional cell cancer of the bladder followed for at least 20 years. J. Urol., 153: 1823, 1995. 3. Olumi, A. F., Tsai, Y. C., Nichols, P. W., Skinner, D. G., Cain, D. R., Bender, L. I. and Jones, P. A.: Allelic loss of chromosome 17p distinguishes high grade from low grade transitional cell carcinomas of the bladder. Cancer Res., 50: 7081, 1990. 4. Meloni, A. M., Peier, A. M., Haddad, F. S., Powell, I. J., Block, A. W., Huben, R. P., Todd, I., Potter, W. and Sandberg, A. A.: A new approach in the diagnosis and follow-up of bladder

2119

cancer. FISH analysis of urine, bladder washings, and tumors. Cancer Genet. Cytogen., 71: 105, 1993. 5. Spruck, C. H., 111, Ohneseit, P. F., Gonzalez-Zulueta, M., Esrig, D., Miyao, N., Tsai,Y. C., Lerner, S. P., Schmiitte, C., Yang, A. S., Cote, R.,Dubeau, L., Nichols, P. W., Hermann, G. G., Steven, K, Horn, T., Skinner, D. G. and Jones, P. A.: Two molecular pathways to transitional cell carcinoma of the bladder. Cancer Res., 54: 784, 1994. 6. Ruppert, J. M., Tokino, K. and Sidransky, D.: Evidence for two bladder cancer sumressor loci on human chromosome 9. Cancer Res., 53: 5093: '1993. 7. Waldman. F. M., Carroll, P. R., Kerschmann, R., Cohen, M. B., Field, F. G. and Mayall, B. H.: Centromeric copy number of chromosome 7 is strongly correlated with tumor grade and labeling index in human bladder cancer. Cancer Res., 51: 3807,1991. 8. Wheeless, L. L., Reeder, J. E., Han, R., OConnell, M. J., Frank, I. N., Cockett, A. T. and Hopman, A. H.: Bladder irrigation specimens assayed by fluorescence in situ hybridization to interphase nuclei. Cytometry, 17: 319, 1994. 9. Pauwels, R. P., Schapers, R. F., Smeets, A. W., Debruyne, F. M. and Geraedts, J. P.: Grading in superficial bladder cancer. (1) Morphological criteria. Brit. J. Urol., 61: 129, 1988. 10. Thomas, D. J., Robinson, M. C., Charlton, R., Wilkinson, S., Shenton, B. K and Neal, D. E.: p53 expression, ploidy and progression in pT1 transitional cell carcinoma of the bladder. Brit. J. Urol., 73: 533, 1993. 11. Sauter, G., Deng, G., Moch, H., Kerschmann, R., Matsumura, K, De Vries, S., George, T., Fuentes, J., Carroll, P., Mihatach, M. J. and Waldman, F. M.: Physical deletion of the p53 gene in bladder cancer. Detection by fluorescence in situ hybridization. h e r . J. Path., 144: 756, 1994. 12. Levine, A. J., Momand, J. and Finlay, C. A.: The p53 tumor suppressor gene. Nature, 351: 453, 1991. 13. Mellon, K, Wilkinson, S., Vickers, J., Robinson, M. C., Shenton, B. K and Neal, D. E.: Abnormalities in p53 and DNA content in transitional cell carcinoma of the bladder. Brit. J . Urol., 7 3 522, 1994. 14. Mellon, K, Neal, D. E., Robinson, M. C., Marsh, C. and Wright, C.: Cell cycling in bladder carcinoma determined by monoclonal antibody Ki67. Brit. J. Urol., 66: 281, 1990. 15. Fontana, D., Bellina, M., Gubetta, L., Fasolis, G., Rolle, L., Scoffone, C., Porpiglia, F., Colombo, M., Tarabuzzi, R. and Leonardo, E.: Monoclonal antibody Ki-67 in the study of the proliferative activity of bladder carcinoma. J . Urol., 148: 1149, 1992. 16. Hermanek, P. and Sobin, L. H.: UICC TNM Classification of Malignant Tumours, 4th ed. New York: Springer-Verlag, 1992. 17. Pinkel, D., Straume, T. and Gray, J . W.: Cytogenetic analysis using quantitative, high-sensitivity, fluorescence hybridization. Proc. Nat. Acad. Sci., 83: 2934, 1986. 18. Oud, P. S., Henderik, J. B., Huysmans, A. C., Pahlplatz, M. M., Hermkens, H. G., Tas, J., James, J . and Vooijs, G. P.: The use of Light Green and Orange I1 as quantitative protein stains, and their combination with the Feulgen method for simultaneous determination of protein and DNA. Histoehemistry, 80. 49, 1984. 19. Koss, L. G.: Diagnostic Cytology and its Histopathologic Bases, 4th ed. Philadelphia: J. B. Lippincott Co., vol. 1, chapt. 35, pp. 1596-1600, 1992. 20. Cairns, P., Shaw, M. E. and Knowles, M. A.: Initiation of bladder cancer may involve deletion of a tumour-suppressor gene on chromosome 9. Oncogene, 8 1083, 1993.