The majority of bladder neoplasms are noninvasive (pTa). Only approximately 30-40% of the tumors show invasion at the time of diagnosis. Invasion depth is of critical clinical importance. Prognosis is clearly worse if cancer cells reach the muscle bladder wall (stage pT2-4) compared to minimally invasion (stage pT1) where cancer cells do not go beyond lamina propria of the urinary bladder. Increased genomic instability is suspected to increase the malignant potential of cancer cells.

Table 3. Number of aberrations and tumor grade

a Aberrations per tumor (mean±SD)
b P=0.8970
c P=0.8410
d P=0.8210

In the present study we performed a comparative genomic hybridization analysis for whole genome scan in pT1 and pT2-4 tumors.

Our results did not show significant differences in the total number of chromosomal aberrations neither between pT1 and pT2-4, nor between grade 2 and grade 3 tumors which is in agreement with the study of Rihter et al [9]. The authors have found remarcable genetic similarities between minimally and deeply invasive tumors. Our findings are consistent also with models suggesting two different entities of bladder neoplasms, one (pTaG1/G2) being genetically stable with low risk of progression, and the other one (pTa G3, pT1-4) with a high degree of genetic instability and high risk of progression (G. Sauter, personal communication).

Interestingly, some of the genetic alterations were significantly linked to the tumor stage (P£0.05). Those were gains at 5p, 17q and 20q (P=0.0198, P=0360 and P=0.0336). Gain at 5p has been reported to be associated with muscle invasion also in the study of Rihter et al, while we did not confirm the involvement of deletions at 5q, 7p, and 15q, neither gains at 7p and Xq.

Overrepresentations of 5p (whole arm) are well known in bladder cancer. Cytogenetic studies have suggested that isochromosomes at 5p may represent primary alterations in urinary bladder cancer [11]. Currently there is no strong candidate oncogene at 5p, although 5p DNA copy number gains and amplifications have been frequently found by CGH in various tumor types, such as carcinomas of the lung, head, and neck, uterine cervix as well as osteosarcomas [12] and therefore, it is likely this region to carry at least one oncogene. In the study of Rihter et al. [9], 5p has been also shown to be significantly more frequent in pT2-4 than in pT1 bladder cancer (P=0.004).

Fig. 2. Summary of all relevant DNA copy number changes detected in 16 pT1 bladder cancer tumors. The vertical lines on the right side of the chromosome idiograms indicate genetic gains, those on the left losses of coresponding genetic material. *1p, 16p, 19, 22, X and Y were not included in the analysis.

Gains at 17q and 20q have not been linked to the tumor stage so far, althogh both aberrations are one of the most frequent in bladder cancer [3-9]. Cytogenetic findings including 17q+ and 20q+ have been found frequently also in bladder carcinomas in situ [13, 14]. A comparative study by CGH between pTa and pT1 tumors has revealed no association between gains at 17q and 20q, and tumor invasion, although the aberrations’ frequency increased from pTa to pT1 [3, 4]. There is no significant difference in the genome instability between primary tumors and their metastases [6] while primary tumors with metastases have shown significantly more changes than primary tumors without metastases suggesting increased genomic instability in metastatic tumors [15]. The most frequently affected by gains regions in the tumors with metastases and not in primary tumors without metastases were 17q11-12 (ErbB-2) as well as 10q22-23 and 20q11 which have been proposed by the authors to contain oncogenes involved in bladder tumors metastasis. In the present study for the first time 17q and 20q have been shown to be linked to muscle bledder tumor invasion.

In summary, our study demonstrated that not the total genome instability but specific chromosomal aberrations (5p+, 17q+ and 20q+) are associated with muscle invation of urinary bladder cancer.