Despite the extensive research into genetic alterations of bladder cancer and detailed models which link such changes to tumor initiation and progression 
, there are few reliable markers to distinguish tumors with aggressive characteristics at the time of early diagnosis and we are still looking for the method of election to detect them. In this regard, a recent prospective study has even suggested that cystoscopy alone remains the most cost-effective strategy for detecting recurrence of bladder cancer not invading the muscle 
. However, in contrast to what previously reported by others 
, several authors claimed the same conclusion 
, and the role of Urovysion in suspicious urine specimens remained questionable, especially in view of its high cost.
The development of array-CGH led to the possibility to analyze the whole genome in a single experiment, suggesting its possible application in screening/surveillance programs of cancer patients. In the case of bladder cancer, array-CGH would give the possibility to analyze the DNA from a biopsy of the tumor, while by Urovysion urine specimens are usually analyzed.
The main drawbacks of this technique are that, even if it is specific and sensitive, it is invasive and still expensive. Furthermore, to date there are no sufficient data to support the use of array-CGH in this kind of programs, but it could be interesting to apply this technique for patients' categories with high cancer risk.
The multitarget Urovysion assay has been developed for the detection of TCC in urine specimens 
. The optimal FISH probe set was determined by testing different probes for TCC detection in urine from patients with bladder cancer and selecting those that were either the most sensitive individually or that complemented other probes to enhance the overall sensitivity of the test. The CEP probes and LSI 9p21 were complementary because the CEP probes detect hyperdiploidy, common in carcinoma in situ
and invasive TCC, while the LSI 9p21 probe detects deletions of the 9p21 band, common in non-invasive TCC 
. It has been previously suggested that a false-negative FISH result represents mostly low-grade TCC that do not shed tumor cells into the urine or do not exhibit the chromosomal alterations that are detected by the assay 
. Another limit, and another possible explanation for false-negative FISH results, might be attributed to the low number of neoplastic cells present in the specimens 
In the first step of this study we compared the performance of this multitarget assay for the detection of bladder tumor cells both in FIN, without histological diagnosis and even with a low number of neoplastic cells, and in FFPE tissue. Our analysis evidenced a good correspondence of Urovysion FISH data between FIN and FFPE for LGNI and HGIN tumors; in particular, in the former group, FIN tended to detect a smaller number of signal respect to FFPE, while in the latter group an opposite tendency was appreciated. For HGNI TCCs, significant differences emerged for three targeted probes, but it could be due to the low number of samples of this group. The performance of this targeted test is therefore sufficiently acceptable also on FIN samples; furthermore the same CNAs were faithfully reflected by the analysis on FFPE. It remains to investigate whether it is an efficient method to detect the most representative and effective CNAs of TCCs. For this purpose, in the second step of this study, array-CGH was performed on 10 additional TCCs to dissect the spectrum of alterations in bladder cancer and to identify recurrent aberrations that may contain cancer-related genes.
We detected numerous genetic changes by array-CGH: the most frequent loss involved chromosome 9p-arm while the most frequent gain involved chromosome 20q-arm, as previously reported by others 
. Surprisingly, we didn't find a high percentage of tumors with gain of 6p22.3 and 8q reported in other studies 
. LOH and under-representation of chromosome 9 is the most frequently described genetic alteration in TCC (>50%). The common loss of an entire copy of chromosome 9 indicates the presence of tumor suppressor genes both on 9p and 9q, and candidate genes have been identified in several regions including 9p21 (CDKN2A
), 9q12-13 (PTCH
), 9q32-33 (DBC1
) and 9q34 (TSC1
). In this study, we observed complete or partial loss of 9p and/or 9q in 7/10 tumors, in both HG and LG. Moreover, in some HG we observed a gain for this locus, even if this could be due to chromosome 9 polyploidy (as the sign of chromosomal instability). The most frequent gain is 20q (6 tumors), in accordance with data previously reported in many other cancers, including bladder, colon, ovarian and breast 
. Association of 5p and 20q gains, found in 3 HG tumors, reported by Bruch 
, could be associated with progression. Finally, gain of 17q21 is identified only in HG tumors, suggesting a possible role in tumor progression.
The most interesting point of this study is the comparison of array-CGH data and Urovysion FISH data. Indeed, we evidenced not only a high intra and inter-tumor heterogeneity in FFPE material, as emerged from the analysis of two different tumoral areas of the same tumor; we also found some discrepancies in the two techniques that could be partially ascribed to a possible masking effect from normal cells or to a compensatory effect derived from the great tumor heterogeneity. This heterogeneity has already been described by our group in bladder cancer stem-like cells that are genetically different 
. We can suggest that this diversity generates viable and clonally related subpopulations that become heterogeneous in the same tumor.
The overall array-CGH data stressed once again on the presence of frequent alterations (i.e. 20p and 5p gains) that cannot be detected by Urovysion assay. A further advantage of using an integrated technical approach emerged for 028CR09 sample: the amplification of 3p evidenced by array-CGH was studied by FISH with Urovysion assay and a LSI 3p probe. Through the integration of multiple methods, we were able to discriminate the true amplification from a chromosome 3 polysomy. This locus includes the peroxisome proliferator-activated receptor gamma (PPARG
), a ligand activated transcription factor implicated in the regulation of proliferation and differentiation of urothelium 
In conclusion, considerable effort is still required to define the genes underlying the chromosomal abnormalities to a better understand of the genetic mechanisms in order to develop new therapeutic strategies. Our results confirmed the importance of global genomic screening methods, that is array based CGH, to comprehensively determine the genomic profiles of large series of TCCs tumors. However, this technique has yet some limitations, such as not being able to detect low level mosaicism, or not detecting any change in the number of copies for a kind of compensatory effect due to the presence of high cellular heterogeneity. Thus, it is still advisable to use complementary techniques such array-CGH and FISH, as the former is able to detect alterations at the genome level not excluding any chromosome, but the latter is able to maintain the individual data at the level of single cells, even if it focuses on few genomic regions.