FGFR3 and TP53 mutations in a prospective cohort of Belarusian bladder cancer patients
Summary. Aim: The aim of this study was to determine the frequencies of FGFR3 and TP53 mutations in a prospective cohort of 150 bladder cancer patients and to assess the relationship between their mutational status and clinicopathological variables. Materials and Methods: The FGFR3 and TP53 mutations were detected by the SNaPshot method and PCR-single-strand conformational polymorphism analysis followed by DNA sequencing. Results: The activating FGFR3 mutations were found in 71 (47.3%) whereas TP53 mutations were observed in 31 (20.7%) urothelial carcinomas. FGFR3-mutant tumors significantly correlated with lower tumor stage and grade, papillary form of bladder cancer and the absence of metastases while TP53-mutant tumors were strongly associated with higher tumor stage and grade as well as the presence of metastasis. We also found significant inverse correlation between FGFR3 mutations and TP53 alterations in urothelial carcinomas (p=0.03). Four possible genotypes were observed in the whole studied cohort, namely FGFR3mut/TP53wt (41.3%), FGFR3wt/TP53wt (38%), FGFR3wt/TP53mut (14.7%), and FGFR3mut/TP53mut (6%). Tumors with FGFR3wt/TP53wt genotype comprised the subgroup, in which all stages and grades were equally distributed. Conclusions: Our findings confirm the alternative role of FGFR3 and TP53 mutations in the development of bladder cancer. Together these two genetic markers are attributed to 62% of the tumors studied. Tumors with both wild type genes included urothelial carcinomas of all stages and grades and may develop through another genetic pathway. To elucidate complete molecular profile of bladder tumors further additional studies are needed.
Submitted: September 09, 2014.
*Correspondence: E-mail: R.Goncharova@igc.bas-net.by
Abbreviations used: MIBC — muscle invasive bladder cancer; NMIBC — non-muscle invasive bladder cancer; PCR — polymerase chain reaction; SSCP — single-strand conformation polymorphism; UCC — urothelial cell carcinoma; wt/mut — wild type/mutant.
Bladder carcinoma is the most common malignancy of urinary tract. In Europe, as well as in Belarus, it is the 7th most frequent cancer in men [1, 2]. Urothelial cell carcinoma (UCC) comprises 90% of all bladder cancer cases. Bladder tumors are morphologically heterogeneous and can be categorized into two groups: non-muscle invasive and muscle invasive. At the time of diagnosis more than 75% of bladder tumors are non-muscle invasive Ta, CIS, and T1 neoplasms . The remaining percentage of patients (i.e. 20–25%) is diagnosed with muscle invasive bladder cancer (MIBC) ≥Т2 who undergo cystectomy and have a poor prognosis . After transurethral resection, most patients with non-muscle invasive bladder cancer (NMIBC) (70–80%) develop recurrences within 5 years, and some of them (10–20%) show progression into MIBC which is considered potentially lethal. It provides grounds to assume that NMIBC is a heterogeneous group of cancers with different prognosis. A number of researchers proposed the hypothesis that clinical and morphological heterogeneity of bladder tumors is determined by different molecular pathways of urothelial carcinoma pathogenesis [4–7]. Non-muscle invasive bladder tumors are mainly characterized by high frequency of activating FGFR3 missense mutations, opposed to high frequency of inactivating TP53 alterations in MIBC. The prevalence of FGFR3 mutations in non-muscle invasive low grade urothelial carcinomas and their correlation with favourable prognosis suggested FGFR3 mutation to be a potential molecular marker for prognosis of NMIBC [8, 9]. Further molecular studies of bladder cancer not only proved the existence of two alternative genetic pathways of urothelial carcinoma pathogenesis but also revealed a more complex picture of bladder tumor mutability and genome instability [10–12].
In our previous study of a prospective cohort of Belarusian patients with NMIBC and MIBC, we found a high frequency (45.5%) of activating FGFR3 gene mutations that was significantly associated with low stage and low grade tumors . To understand completely the genetic basis of urothelial carcinoma, there is a need to study the mutational status of other genes involved in bladder cancer development. The aim of the present study was to determine the frequencies of FGFR3 and TP53 mutations in a prospective cohort of bladder cancer patients and to find the relationship between clinicopathological parameters of the disease and tumor genotypes according to mutations in both genes.
MATERIALS AND METHODS
Рatients and tissue samples. We analyzed the tumor tissues of 150 patients (115 men and 35 women) with histologically confirmed UCC who were treated at the Department of Urology of N.N. Alexandrov National Cancer Center of Belarus between 2010 and 2013. The median age of the patients was 67.5 years. Written informed consent was obtained from each patient. Fresh tissue samples and matched formalin-fixed paraffin-embedded (FFPE) tissue samples were used as biological material for analysis. Tumors were graded according to the 1973 and 2004 WHO classification and staged according to the tumor-node-metastasis classification guidelines. Sociodemographic and clinical information is presented in Table 1.
Table 1. Characteristics of the patients and tumor samples
DNA extraction. Genomic DNA was isolated from both fresh tissue samples and deparaffinized FFPE tissue sections (5–10 µm) by using a standard phenol-chlorophorm method.
FGFR3 mutation analysis. FGFR3 mutational status was analyzed using a previously described highly sensitive SNaPshot assay  that allows screening for 11 point mutations (R248C, S249C, G372C, S373C, Y375C, G382R, A393E, K652E, K652M, K652Q, and K652T). Briefly, three regions comprising all known FGFR3 mutations were amplified in one multiplex polymerase chain reaction (PCR), followed by extension of internal primers for each mutation with a labeled dideoxynucleotide. Extended primers were separated by capillary electrophoresis in an automatic sequencer ABI Prism 3500 (Applied Biosystems), and the presence or absence of a mutation was defined by the incorporated dideoxynucleotide.
TP53 mutation analysis. The regions comprising exons 5–8 of the TP53 gene were amplified by PCR using the primers described by Thongsuksai et al. . PCRs were performed in a total volume of 15 µl containing 100 ng genomic DNA, 1× buffer (100 mM Tris-HCl, pH 8.3, 500 mM KCl), 0.2 µM of each primer, 2 mM MgCl2, 0.2 mM dNTPs, and 0.5 U Dream Taq DNA Polymerase (Thermo Scientific). The cycling conditions were as follows: denaturation at 94 °C for 5 min, followed by 35 cycles of denaturation at 94 °C for 30 s, annealing at 56 °C for 30 s, extension at 72 °C for 50 s, with a final extension at 72 °C for 10 min on C1000 thermal cycler (Bio-Rad).
For single-strand conformation polymorphism (SSCP) analysis, 7 µl of PCR product were added to 10 µl of a solution containing 95% formamide, 10 mM NaOH, 20 mM EDTA, 0.05% of xylene cyanol, and 0.05% of bromophenol blue. The mixture was denatured at 100 °C for 7 min then immediately cooled on ice and loaded onto a 10% acrylamide gel (20 x 20 cm plate) containing 5% glycerol. Electrophoresis was performed at 280 V constant power and run for 7 h at 8 °C. After electrophoresis the gel was stained with ethidium bromide.
Samples that showed band shift on SSCP gel were subjected to direct sequencing, which was performed using the Big Dye Terminator v 3.1 cycle sequencing kit (Applied Biosystems) in accordance with the manufacturer’s instructions. All mutations were confirmed by sequencing both DNA strands.
Statistical analysis. To assess the association between FGFR3 and TP53 genetic alterations and clinicopathological parameters of tumors Chi-square test, Fisher’s exact test, or Mann — Whitney U test were used when appropriate, and values of p < 0.05 were considered statistically significant.
Frequency and spectrum of FGFR3 gene mutations. Somatic point FGFR3 mutations were detected in 71 (47.3%) of 150 UCCs, double mutations being found in 6 tumors. Seven different missense mutations, namely S249C, R248C, Y375C, G372C, A393E, S373C, and К652М were identified. Among 6 cases with double mutation, three tumors harbored S249C/R248C alterations, while the other three had A393E/R248C, S249C/Y375C and К652М/A393E combinations. The frequency and distribution of all identified FGFR3 mutations are shown in Table 2. The most common changes were S249C (68.8%) and Y375C (16.9%). The overall frequency of the mutations in codons 248, 249, and 375 — the major hot spots in bladder tumors — was 93.5% of total mutations in FGFR3.
Table 2. Frequencies and spectrum of FGFR3 gene mutations
Frequency and spectrum of TP53 gene mutations. TP53 alterations were found in 31 (20.7%) tumors, out of which 37.8% were detected in exon 8, followed by 35.2% in exon 7, 21.6% in exon 5 and 5.4% in exon 6. Twenty five patients had only one mutation, double TP53 mutation was observed in 6 (4%) cases. Distribution of the TP53 alterations over 27 distinct codons is shown in Table 3. All molecular changes of the TP53 gene were presented by missense point mutations except one patient, who had a silent mutation. 48.6% of the gene alterations were transitions and 51.4% were transversions. No mutation leading to a premature termination of the protein was found. The codons 248, 175, and 285 were mutational hot spots in bladder tumors.
Table 3. ТР53 gene mutations detected in bladder tumors
Association of FGFR3 and TP53 mutations with clinicopathological parameters of tumors. Sociodemographical variables and morphologic features of tumors according to FGFR3 and TP53 mutational status are shown in Table 4. No correlation was found between FGFR3 changes and various characteristics like age, gender, smoking, multifocality and tumor size. At the same time, statistically significant association of FGFR3 mutations with papillary form of bladder cancer (р < 0.001) and non-metastatic disease (р = 0.02) was revealed.
The frequency of FGFR3 mutations in non-muscle invasive tumors was 57.7% against only 28.3% in muscle invasive T2-T4 UCCs (p < 0.001). We found strong correlation between FGFR3 mutations and tumor stage (р < 0.001). FGFR3 changes had the highest frequency (62.5%) in Tа tumors and the lowest (13.3%) in T4 carcinomas. Mutational status of the FGFR3 gene was also significantly associated with tumor grade (р < 0.001). The distribution of the frequency of FGFR3 alterations according to tumor grade was as follows: 68.5% in G1, 47.7% in G2 and 9.7% in G3 tumors. Thus, non-muscle invasive papillary low grade UCCs are characterized by high frequency of FGFR3 alterations.
As to TP53 gene, no statistically significant difference was noted in the distribution of TP53 mutations on the basis of gender, age and smoking. However, mutations were strongly related to non-papillary form of bladder cancer, large tumor size, solitarity and the presence of metastasis (Table 4). The frequency of molecular changes of this gene was substantially higher in MIBC (43.4%) than NMIBC (8.2%). A strong association was found between TP53 mutations and high tumor stage (р < 0.001). No TP53 mutation was detected in Tа tumors, the frequency of the gene changes increased from about 10% in T1 to 60% in T4 tumors. Mutational status of the TP53 gene was significantly related to tumor grade: a higher rate of TP53 alterations was observed in the group of G3 (45.2%) tumors compared to G2 (20%) and G1 (7.4%) tumors. Hence, high frequency of TP53 genetic changes characterizes high grade muscle invasive tumors of large size.
Table 4. Clinicopathological variables according to TP53 and FGFR3 mutational status
The strong associations between FGFR3 and TP53 mutational status and tumor stage and grade are clearly depicted in Fig. 1. An inverse relation was found between FGFR3 mutations and TP53 alterations (p = 0.037). Stage and grade information taken together showed a high proportion of FGFR3 mutations and TP53 alterations in tumors with low and high malignant potential, respectively.
Analysis of tumor genotypes revealed that genotype FGFR3mut/TP53wt was the most prevalent, accounting for 41.3% of UCCs; tumors FGFR3wt/TP53wt lacking alteration in both genes were observed in 38% of cases, and frequencies of genotypes FGFR3wt/TP53mut and FGFR3mut/TP53mut were 14.7% and 6%, respectively. Interestingly, FGFR3 and ТP53 alterations were almost mutually exclusive, and they coincided in only 9 tumors of 150 (6%).
Fig. 1 FGFR3 and TP53 mutations according to tumor stage (a) and grade (b)
Tumor genotypes at various stages and grades are shown in Fig. 2.
Fig. 2 Tumor genotypes according to mutations in both FGFR3 and TP53 at various stages (a) and grades (b)
In Tа tumors, more commonly observed genotypes were FGFR3mut/TP53wt (62.5%) and FGFR3wt/TP53wt (37.5%). In the group of T1 carcinomas, the frequencies of various genotypes were as under: FGFR3mut/TP53wt — 50.6%, FGFR3wt/TP53wt — 39.5%, FGFR3mut/TP53mut — 6.2% and FGFR3wt/TP53mut — 3.7%. In invasive T2–4 tumors, genotypes FGFR3wt/TP53mut (35.8%) and FGFR3wt/TP53wt (35.8%) were the most prevalent, followed by genotypes FGFR3mut/TP53wt (20.8%) and FGFR3mut/TP53mut (7.5%).
Distribution of four genotypes according to tumor grade showed that the majority of low grade UCCs had FGFR3mut/TP53wt and FGFR3wt/TP53wt genotypes, while in high grade carcinomas, genotypes FGFR3wt/TP53wt and FGFR3wt/TP53mut were more common. On the whole the data indicate that there is a strong association of FGFR3 mutations with favorable disease parameters (low grade, non-muscle invasive), and TP53 alterations with unfavorable tumor features (high grade, muscle invasive). Patients carrying either FGFR3 or TP53 mutation accounted for 62% of the whole cohort studied. The group of UCCs that were wild type for both genes represented a considerable part (more than 35%) of tumors of all stages and grades.
To our knowledge, the first two pathway model of urothelial carcinoma pathogenesis was proposed by Spruck et al. . Stratification of tumors into groups was based on the differences in the frequencies of TP53 mutations and chromosome 9 deletions. Later, this model was modified by van Rhijn et al.  by adding FGFR3 marker. The authors found an inverse relationship between FGFR3 and ТР53 mutability and clinical morphological parameters of tumors and/or disease-specific survival (a higher frequency of FGFR3 mutations is indicative of high disease-specific survival while elevated rate of TP53 alterations indicates low disease-specific survival). This finding suggested that mutations of corresponding genes may serve as the molecular markers of good and poor prognosis in bladder cancer patients.
We also have demonstrated an inverse relation between FGFR3 and TP53 mutations in Belarusian bladder cancer patients. In non-muscle invasive papillary low grade tumors, we found high percentage of FGFR3 mutations which is in accordance with previous studies in European [16, 17] and Belarusian populations . In the present investigation, like in other reports [18, 19], high frequency of TP53 alterations was observed in advanced tumors.
Moreover, we analyzed the frequencies of genotypes according to FGFR3 and TP53 mutational status in the groups of tumors of different stages and grades. In Ta group, no alteration of the TP53 gene was found. However, a number of studies reported that TP53 mutations may occur at this tumor stage as well, though quite rarely [19, 20, 21].
In non-muscle invasive T1 tumors, we found the high frequency of FGFR3mut/TP53wt genotype accounting for more than 50%, while UCCs with FGFR3wt/TP53mut genotype were observed in 3.7%. Conversely, the previous studies reported that <27% and >20% of T1 tumors had FGFR3mut/TP53wt and FGFR3wt/TP53mut genotypes, respectively [5, 19]. Such differences in the frequencies of T1 tumor genotypes in Belarus compared to Western Europe may be explained by high percentage of muscle invasive disease at first diagnosis in Belarus and further by prevalence of T1 UCCs among the majority of patients with NMIBC. T1 group comprises tumors with different genotypes suggesting their genetic heterogeneity which is reflected in the clinical picture of the disease.
The genetic nature of tumors which are wild type for both FGFR3 and TP53 genes (38% of UCCs in our study) remains unclear. The tumors of this subgroup can develop either through the third pathway of bladder cancer pathogenesis, supposed by van Rhijn et al. , or due to hyperexpression of the normal FGFR3 gene observed by Tomlinson et al. in 42% of UCCs with wild type FGFR3 . Recent investigations on bladder cancer revealed that its formation and development may not be referred to just two main pathways based on FGFR3 and TP53 alterations, respectively .
Whole-genome studies of different cancers found a considerable number of somatic mutations in the tumor genome  and it raised a problem of defining the key drivers. The key mutations rather than accompanying (passenger) ones could be used in bladder cancer diagnosis and prognosis. Detailed molecular genetic analyses conducted by Lindgren et al. and Guo et al. [10, 11] validated the role of FGFR3 and TP53 mutations in UCC pathogenesis. Besides, these analyses along with other data established the pathogenetic role of mutations in several other genes, particularly RB1, PIK3CA, KRAS, HRAS, NRAS, CDKN2A, and TSC1. However, the distribution of RAS [10, 24, 25], CDKN2A and TSC1  mutations is independent of tumor stage and grade, and thus, they are not of high prognostic value.
According to whole-genome analysis, UCC was divided into two molecular subtypes (MS1 and MS2), differentiated by the mutational status and patterns of gene expression . MS1 tumors are characterized by high frequency of activating FGFR3 mutations which often coincide with PIK3CA alterations, whereas TP53 and RB1 changes are more common in MS2 tumors. When comparing mutational profiles of MS1 and MS2 carcinomas with clinicopathological parameters, the authors distinguished the subgroups of UCCs according to the level of genome instability. Tumors with FGFR3 mutations were found genetically more stable than those with TP53 alterations, characterized by high proportion of chromosome aberrations [10, 17].
Whole-genome and whole-exome sequencing of 99 bladder tumors  showed alterations in 37 genes including well known FGFR3, TP53, RB1, PIK3CA, KRAS, HRAS, and TSC1. Moreover, the authors identified the essential role of mutations in the genes involved in sister chromatid cohesion and segregation process. The frequency of genetic alterations affecting this process was 32% and related to more aggressive tumor phenotype. Besides, the high frequency of mutations in the genes involved in cell cycle control, DNA repair and chromatin post-translational modifications was found. The role of these genes in determination of molecular subtypes of bladder tumors will be elucidated by studying the association of mutations with genome instability and clinicopathological parameters.
The data mentioned indicate a complex genetic nature of UCCs. Our findings confirm the alternative role of FGFR3 and TP53 mutations in non-muscle and muscle invasive bladder tumors as well as their genetic heterogeneity. Our data support the idea that tumors with FGFR3wt/TP53wt genotype may represent the third pathway of bladder cancer pathogenesis. Further study of the molecular profile of UCC is needed to define molecular subtypes of bladder tumors, which are important to improve bladder cancer diagnosis and prognosis.
1. Babjuk M, Burger M, Zigeuner R, et al. EAU guidelines on non-muscle-invasive urothelial carcinoma of the bladder: update 2013. Eur Urol J 2013; 64: 639–53.
No Comments » Add comments