FGFR3 and Tp53 Mutations in T1G3 Transitional Bladder Carcinomas: Independent Distribution and Lack of Association with Prognosis Free

FGFR3 and Tp53 mutations have been proposed as defining two alternative pathways in the pathogenesis of transitional bladder cancer. FGFR3 mutations are associated with low-grade tumors and a favorable prognosis. Tp53 alterations are associated with advanced tumors and, possibly, with a poor prognosis. We focus here on the subgroup of T1G3 superficial tumors because they are a major clinical challenge. Patients (n = 119) were identified from a prospective study of 1,356 cases. Mutations in FGFR3 (exons 7, 10, and 15) and Tp53 (exons 4-9) were analyzed using PCR and direct sequencing. All cases were followed for recurrence and death. Survival was analyzed using Kaplan-Meier curves and multivariable Cox regression. FGFR3 mutations were detected in 20 (16.8%) tumors; 100 mutations in Tp53 were found in tumors from 78 (65.5%) cases. Multiple alterations in Tp53 were present in 19 tumors (16%). Inactivating mutations were present in 58% of tumors. The combined mutation distribution (FGFR3/Tp53) was: wt/wt (34.5%), mut/wt (7.6%), wt/mut (48.7%), and mut/mut (9.2%), indicating that the presence of either mutation did not depend on the other (P value = 0.767). FGFR3 and Tp53 mutations were not associated with clinicopathologic characteristics of patients and did not predict, alone or in combination, recurrence or survival. Taking the risk of the wt/wt group as reference, the mutation-associated risks of cancer-specific mortality were: mut/wt 1.42 (0.15-13.75), wt/mut 0.67 (0.19-2.31), mut/mut 1.62 (0.27-9.59). These molecular features support the notion that T1G3 tumors are at the crossroads of the two main molecular pathways proposed for bladder cancer development and progression.

Bladder cancer is a common malignancy characterized by a diverse clinical course. Transitional cell carcinomas represent >90% of bladder tumors and are classified into superficial (pTa and pT1) and muscle invasive (≥pT2). Superficial tumors are characterized by frequent recurrences and, in a lower proportion of cases, progression (1). A large number of studies have proposed that superficial/papillary and invasive tumors progress via different molecular pathways, with deletions of chromosome 9 more commonly found in superficial/papillary tumors and Tp53 mutations and loss of heterozygosity on chromosome 17 more frequent in carcinoma in situ and invasive tumors (1–5). Tp53 alterations—most commonly evaluated using nuclear overexpression of p53 as a surrogate marker for gene mutation—have been proposed to be associated with poor prognosis (6), but this remains a controversial issue.

Several recent reports have proposed that FGFR3 mutations are associated with a low risk of recurrence of superficial tumors (7, 8). FGFR3 belongs to the tyrosine kinase receptor family. FGFR subfamily members differ in tissue expression, ligand specificity, signal pathway activation, and biological effects (9, 10). These receptors regulate diverse cellular processes including cell growth, differentiation, and angiogenesis. Gain-of-function mutations have been identified in different autosomal dominant human skeletal disorders such as hypochondroplasia, achondroplasia, and thanatophoric dysplasia (11, 12). FGFR3 mutations identical to those identified in these disorders have been reported in multiple myeloma, cervix, and bladder cancers (7, 8, 13, 14). Unlike Tp53 alterations, FGFR3 mutations occur predominantly in superficial/papillary transitional cell carcinomas and are reported to be associated with low-grade tumors and with a favorable prognosis (7, 8, 15, 16). Two recent studies of transitional cell carcinomas dealing with tumors of a broad range of stages and grades report that FGFR3 and Tp53 mutations show a mutually exclusive distribution, suggesting that the two alterations define different routes of transitional cell carcinoma development (15, 16).

In this work, we have studied the prevalence of genetic alterations in these two genes in a prospectively collected series of 119 T1G3 transitional cell carcinomas, as well as their ability to predict outcome. We have focused on this subgroup of superficial tumors because of their high risk of progression, the lack of prognostic markers, and because they constitute only 10% of all transitional cell carcinomas and, therefore, are poorly represented in previously published studies (7, 8, 13–16). In this group of tumors, we find a low prevalence of FGFR3 mutations and a high frequency of Tp53 mutations. In addition, and unlike the findings reported in less homogeneous patient series, mutations in both genes show an independent distribution, are not mutually exclusive, and are not able to predict patient survival.

Tumor samples and patients. Cases were drawn from the EPICURO study which is comprised of 1,356 consecutive patients with incident bladder cancer recruited prospectively in 18 general hospitals in Spain between 1997 and 2001. Sociodemographic and clinical information was retrieved from hospital records. Tumors were staged and graded according to the criteria of the tumor-node-metastasis classification and the WHO-International Society of Urological Pathology (17). Diagnostic slides from each case were reviewed by a panel of expert study pathologists (R. Jaramillo and J. Lloreta) to confirm diagnosis and ensure uniformity of classification criteria across all cases. A total of 119 T1G3 cases were analyzed and are the subject of this report. Table 1 summarizes the characteristics of the cases included in the study.

Cases were prospectively followed-up both through hospital records and by telephone interviews, either to the patients or a next-of-kin when the former was not reachable or was deceased. Recurrence was defined as the appearance of a new tumor after administering treatment for the primary disease. All deaths were recorded but only bladder cancer–related death was considered for the survival analysis. As of November 2004, 48 (40.3%) subjects were alive and free of disease with a median follow-up of 54 months (range 38-72). Out of 40 deaths, 16 were due to bladder cancer and 24 were due to other causes. The latter cases were censored at the time of death for the analysis. Survival was computed as the period comprised between cancer diagnosis and death or last control. There were no cases lost to follow-up. Written informed consent was obtained from all patients. The study was approved by the Ethics Committees of all participating institutions.

FGFR3 and Tp53 mutation analysis. The tumor block that was most representative of the TG staging was selected for analysis. Areas containing >50% tumor cells were manually microdissected from sections of formalin-fixed, paraffin-embedded tissues: three to five consecutive 10-μm sections were deparafinized and DNA was extracted using the DNeasy tissue kit (Qiagen GmbH, Hilden, Germany). Exons 7, 10, and 15 of FGFR3 were amplified and sequenced using tumor DNA. Exons 5 to 8 of Tp53 were amplified and sequenced using tumor DNA; exons 9 and 4 were analyzed when no alterations were found in exons 5 to 8. In nine cases, including all tumors in which a silent mutation was identified, autologous normal leukocyte DNA was used to confirm the somatic nature of the mutations. The sequence of primers used is shown in Table 2. PCR reactions were done in a 50 μL volume using 10 to 50 ng of DNA, 0.2 μmol/L of each primer, 200 μmol/L deoxynucleotide triphosphates, 3.5 mmol/L MgCl₂, 1× PCR II buffer, and 1.5 units of Amplitaq Gold DNA Polymerase (Applied Biosystems, Foster City, CA). PCR conditions were as follows: 95°C (10 minutes) for 1 cycle, 95°C (40 seconds), 63°C (40 seconds; for exons 4, 5, and 7-9 of Tp53 and exons 7 and 10 of FGFR3) or 67°C (40 seconds; for exon 6 of Tp53), 72°C (40 seconds) for 40 cycles, and a final extension step of 72°C (10 minutes). PCR products were separated by electrophoresis and visualized with ethidium bromide. Samples without DNA template were included in all assays as negative controls. PCR products were purified using Qiagen PCR purification kit (Qiagen, Crawley, United Kingdom) according to the manufacturer's protocol.

Mutational analysis was done by direct sequencing of purified PCR products with the Big Dye Terminator kit v 3.1 using an ABIPRISM 377 instrument (Perkin-Elmer Applied Biosystems). Each PCR product was sequenced in both the forward and the reverse directions. The same sets of primers used for the PCR amplification were used for sequencing. For 30% of cases, mutational results were verified by sequencing independent PCR products amplified using DNA from the same tumor block.

Statistical analyses. Mutational status (wild-type versus mutated), type of mutation (transition versus transversion), location of the mutation within the gene (exon versus intron), resulting change at the protein level (amino acid substitution versus premature stop codon), and number of mutations were considered as variables for analysis. The association between Tp53 and FGFR3 mutational status and clinical and pathologic characteristics of tumors were assessed by applying χ², Fisher's exact, Student's, or Mann-Whitney U tests, as appropriate (18).

Kaplan-Meier survival curves were computed by each category of the potential prognostic factors and the log-rank and Breslow tests were applied to compare curves (19). Cox proportional hazard ratios were estimated to obtain risks of recurrence and death for cases in each molecular factor stratum, after adjusting for other confounder variables (20). The assumption of proportional hazards was checked for each variable. The final predictive models were fitted after forcing the FGFR3 and/or Tp53 molecular status variables at all the steps.

Results were considered significant at the two-sided P of 0.05 level. Statistical analysis was done using version 12.0 SPSS statistical package (SPSS, Inc., Chicago, IL, 2003).

Mean age of cases was 67 years. Eighty-seven percent of cases were males, yielding a male/female ratio of 6.4. Most cases presented with a single tumor (66%; Table 1).

Prevalence of FGFR3 and Tp53 mutations in T1G3 bladder tumors.Table 3 shows a summary of the molecular results for FGFR3 and Tp53. Twenty tumors harbored a mutation in FGFR3 (16.8%). The most frequent mutation was S249C (n = 15; 75%; exon 7), followed by Y375C (n = 3; 15%; exon 10) and R248C (n = 2; 10%; exon 7). Exon 15 was sequenced in 86 tumors and no mutations were found. A total of 100 mutations in Tp53 were identified in 78 tumors, yielding a prevalence of 65.5%. There were 70 missense mutations, 9 silent mutations, and 16 mutations leading to a premature stop codon due either to the appearance of stop codon (n = 9) or a frameshift (n = 7). In addition, there were four mutations at intron-exon boundaries and a 30 bp in-frame deletion. Fifty-nine tumors (49.6%) had a single mutation, whereas multiple alterations were present in 19 tumors (16%). All mutations considered silent were verified and found to be absent from normal leukocyte DNA. The following mutation distribution was found: exon 4 (n = 16), exon 5 (n = 28), exon 6 (n = 8), exon 7 (n = 13), exon 8 (n = 24), exon 9 (n = 7), intronic mutations (n = 4). The nine silent mutations identified were found in exons 4 (n = 3), 8 (n = 2), and 9 (n = 4). Cases harboring exclusively silent (n = 6) or intronic mutations (n = 2), and the case with the 10 bp in-frame deletion, were considered as “wild-type” for the analyses; the remaining cases were considered to harbor p53 mutant alleles. Overall, there were 69 tumors (58%) harboring inactivating mutations. In 30% of the tumors harboring pathogenic mutations, the wild-type allele was undetected by sequencing, suggesting loss of one allele. p53 immunohistochemistry results were consistent with the findings of the mutational analysis: 49 of 53 tumors with one or more missense mutations showed p53 nuclear overexpression versus 1 of 8 tumors with a mutation leading to a premature stop codon.⁸

To determine whether FGFR3 and Tp53 mutations displayed a mutually exclusive distribution, the combined genotype was examined. The following mutational patterns were found: FGFR3^wt/Tp53^wt (n = 41; 34.5%), FGFR3^mut/Tp53^wt (n = 9; 7.6%), FGFR3^wt/Tp53^mut (n = 58; 48.7%), and FGFR3^mut/Tp53^mut (n = 11; 9.2%). Therefore, the proportion of cases harboring a Tp53 mutation was similar among FGFR3 wild-type (58.6%) and mutated (55%) tumors, and vice versa (χ² P value = 0.767).

Association of FGFR3 and Tp53 mutations with clinicopathologic characteristics and patient outcome. There were no significant differences among groups defined according to FGFR3 and Tp53 mutational status regarding age, sex, number, size, location of neoplasms, and pathologic characteristics of tumors (papillary versus solid growth pattern, size, multiplicity, and location within the bladder). Treatment received was similar for patients with tumors in the four genotype groups (Table 1 and data not shown). Recurrence and mortality curves were similar for all patient groups, regardless of the tumor FGFR3 and Tp53 mutational status (Figs. 1 and 2). There was a slight difference in the recurrence curves in relationship to FGFR3 status during the first year of follow-up (Fig. 1B). However, this difference was not statistically significant (P = 0.432). Regarding Tp53, type of mutation, location of the mutation within the gene, resulting change at the protein level, and number of mutations were not significantly associated with recurrence or survival in a univariate model. Several multivariate Cox analyses were done to fit the final equation with the highest likelihood ratio test, always including in the model Tp53 and/or FGFR3 mutations. The adjusting variables retained in the models were number of primary tumors, treatment, and number of recurrences. None of the models showed a significant predictive value for recurrence or mortality for any of the FGFR3 and Tp53 genotypes (Table 4). Taking the patients whose tumors that had a wild-type sequence for both genes as a reference, the risk of death for patients whose tumors harbored mutations were: FGFR3^mut/Tp53^wt [1.42; 95% confidence intervals (CI), 0.15-13.75], FGFR3^wt/Tp53^mut (0.67; 95% CI, 0.19-2.31), and FGFR3^mut/Tp53^mut (1.62; 95% CI, 0.27-9.59), P = 0.735.

Superficial bladder tumors are a heterogeneous group of neoplasms with a markedly diverse prognosis. T1G3 tumors represent a particularly important clinical challenge: ∼50% of them will progress, implying that patients require close follow-up. Molecular markers identifying a subset of patients at particular high risk for progression could be used to select candidates for radical cystectomy whereas conservative management could be used for individuals at lower risk (21). The recent identification of FGFR3 mutations as a good prognosis marker in superficial transitional cell carcinomas prompted this study.

We have analyzed T1G3 tumors from a large series of patients included in a multicenter study in Spain. In addition to the large number of cases, several other characteristics of the study are noteworthy: (a) it was initially conceived to examine molecular alterations in genes involved in bladder cancer as markers of prognosis; (b) it prospectively included consecutive incident cases in general hospitals, thus avoiding referral selection bias; (c) a very high participation agreement rate was achieved (>85%); and (d) both active and passive follow-up strategies were used to accurately assess patient's outcome, with no patients being lost to follow-up.

The finding that FGFR3 mutations occur only in 16.8% of T1G3 tumors, whereas inactivating Tp53 mutations occur in 58% of tumors, supports the notion that these tumors resemble invasive bladder cancers at the molecular level. Indeed, nuclear accumulation of p53 was observed in 84% of tumors in this series, including cases that were found to be wild-type upon sequencing,⁹

as in a recent report by Kelsey et al. of a large series of bladder cancer cases (22). The high prevalence of Tp53 mutations found in our study may be related to the strict criteria used by the expert pathologists for tumor staging. In the largest published series, FGFR3 mutations were found in 141 of 182 (77%) Ta tumors and in 23 of 63 (36%) T1 tumors (8). In two reports including a broad spectrum of tumors, FGFR3 mutations were associated with low-stage and low-grade tumors and were found to be associated with a favorable clinical course (15, 16). In addition, FGFR3 and Tp53 mutations were found to occur in a mutually exclusive pattern (15, 16). In our study of T1G3 tumors, mutations in both genes showed an independent—and not mutually exclusive—distribution. In this subgroup of tumors, a more complex pattern of mutations in genes involved in the two prototypical progression pathways of transitional cell carcinomas emerges. Our findings are not in contradiction with prior studies because the latter have reported only on a small number of T1G3 tumors. Rather, they extend and complement earlier work, reflecting that the involvement of these two pathways in bladder cancer is more complex than previously proposed. The mutually exclusive pattern reported previously likely reflects the fact that the majority of cases included in those studies were either low-stage/low-grade or invasive tumors.

FGFR3 mutations did not predict recurrence or survival in patients with T1G3 tumors. This observation suggests that other genetic changes override the “good prognosis” associated with FGFR3 mutations in tumors of lower stage-grade. In chondrocytes, mutational activation of FGFR3 has been shown to induce the phosphorylation of STATs, the nuclear translocation of STAT-1, the constitutive activation of the extracellular signal-regulated kinase-1/2/mitogen-activated protein kinase pathway, and the overexpression of cyclin-dependent kinase inhibitors such as p16, p18, p19, and p21^CIP/KIP, leading to cell cycle exit (23–25). Tp53 mutations and p16^INK4A inactivation occur frequently in transitional cell carcinomas and could contribute to the abrogation of the “favorable” effects of FGFR3 mutations in these tumors. However, other genes are also likely to participate because we did not observe an association between Tp53 mutations and patient outcome.

In this regard, it will be important to identify the molecular alterations of low stage-grade tumors that progress and to determine whether such neoplasms (i.e., “progressor” T1G3 tumors) indeed resemble “incident” T1G3 tumors. Until now, there is little information on this issue. In our series, 34.5% of T1G3 tumors lacked mutations in both genes, supporting the notion that a substantial proportion of T1G3 tumors arise through the acquisition of alternative/additional molecular alterations.

We conclude that, in T1G3 bladder tumors, FGFR3 and Tp53 mutations do not display a mutually exclusive pattern and do not predict patient outcome. Our findings suggest that T1G3 tumors are at the crossroads of the two main molecular pathways proposed to be involved in bladder cancer development and progression.

Institut Municipal d'Investigació Mèdica, Universitat Pompeu Fabra (Barcelona; coordinating center): M. Kogevinas, N. Malats, F.X. Real, M. Sala, G. Castaño, M. Torà, D. Puente, C. Villanueva, C. Murta, J. Fortuny, E. López, S. Hernández, R. Jaramillo.¹⁰

 Hospital del Mar, Universitat Autònoma de Barcelona (Barcelona): J. Lloreta,¹⁰ S. Serrano,¹⁰ L. Ferrer,¹⁰ A. Gelabert, J. Carles, O. Bielsa, K. Villadiego. Hospital Germans Tries i Pujol (Badalona, Barcelona): L. Cecchini, J.M. Saladié, L. Ibarz. Hospital de Sant Boi (Sant Boi, Barcelona): M. Céspedes. Centre Hospitalari Parc Taulí (Sabadell, Barcelona): C. Serra, D. García, J. Pujadas, R. Hernando, A. Cabezuelo, C. Abad, A. Prera, J. Prat. Centre Hospitalari i Cardiològic (Manresa, Barcelona): M. Domènech, J. Badal, J. Malet. Hospital Universitario (La Laguna, Tenerife): R. García-Closas, J. Rodríguez de Vera, A.I. Martín. Hospital La Candelaria (Santa Cruz, Tenerife): J. Taño, F. Cáceres. Hospital General Universitario de Elche, Universidad Miguel Hernández (Elche, Alicante): A. Carrato, F. García-López, M. Ull, A. Teruel, E. Andrada, A. Bustos, A. Castillejo, J.L. Soto. Universidad de Oviedo (Oviedo, Asturias): A. Tardón. Hospital San Agustín (Avilés, Asturias): J.L. Guate, J.M. Lanzas, J. Velasco. Hospital Central Covadonga (Oviedo, Asturias): J.M. Fernández, J.J. Rodríguez, A. Herrero. Hospital Central General (Oviedo, Asturias): R. Abascal, C. Manzano, T. Miralles. Hospital de Cabueñes (Gijón, Asturias): M. Rivas, M. Arguelles. Hospital de Jove (Gijón, Asturias): M. Díaz, J. Sánchez, O. González. Hospital de Cruz Roja (Gijón, Asturias): A. Mateos, V. Frade. Hospital Alvarez-Buylla (Mieres, Asturias): P. Muntañola, C. Pravia. Hospital Jarrio (Coaña, Asturias): A.M. Huescar, F. Huergo. Hospital Carmen y Severo Ochoa (Cangas, Asturias): J. Mosquera. Centro Nacional de Investigaciones Oncológicas (Madrid): M. Esteller. Universitat Autònoma de Barcelona (Bellaterra): R. Miró, R. Marcos. Progenika (Derio, Bizkaia): A. Martínez.

We thank S. Mancilla, T. Lobato, A. Alfaro, G. Carretero, P. Fernández, I. López, and the many clinicians, investigators, nurses, technicians, and patients participating in the study, as well as the members of the EPICUR-Red. The authors also wish to thank M. Dosemeci, D. Silverman, and N. Rothman (National Cancer Institute, Bethesda, MD) for their participation in the etiological component of the Spanish Bladder Cancer Study. E.L. was supported by a Predoctoral Fellowship of the Ramón Areces Foundation, Madrid.