FOXM1 and β-catenin are potential prognostic markers in muscle-invasive bladder carcinoma

Background Forkhead box protein M1 (FOXM1) and β-catenin were conrmed to associate with numerous cancers, which attracted more attention in recent years. Our research investigated the expression of FOXM1 and β-catenin and their effects on prognosis of patients with muscle-invasive bladder cancer (MIBC). Methods In this study, FOXM1 and β-catenin expression was detected by immunohistochemistry in a study cohort including 121 MIBC patients. Results The results showed signicant correlations of FOXM1 and β-catenin expression with tumor stage, tumor grade, and lymph node metastases in MIBC patients. Univariate analysis showed that patients with high FOXM1 (HR = 2.986; P = 0.011) and low β-catenin (HR = 2.623; P = 0.001) expression levels, advanced tumor stage (HR = 2.325; P = 0.002), advanced tumor grade (HR = 2.790; P < 0.001), tumor size (HR = 2.080; P = 0.020), lymph node metastases (HR = 3.392; P < 0.001), or recurrence status (HR = 2.016; P = 0.011) had a signicantly higher risk of worse overall survival (OS). In the multivariate analysis, tumor stage (HR = 2.095; P = 0.009), tumor grade (HR = 1.962; P = 0.019), FOXM1 expression (HR = 2.196; P = 0.017) and lymph node metastases (HR = 2.136; P = 0.015) predicted worse OS. Therefore, FOXM1 and β-catenin expression was relevant to MIBC incidence, tumor stage, tumor grade, metastasis, and survival and might be a reliable index to judge the prognosis of patients with MIBC.


Introduction
Bladder carcinoma is a common cancer of the urinary tract, with a poor prognosis in 2020 worldwide [1].
About three-quarter of patients had non muscle invasive bladder carcinoma (NMIBC), including papillary carcinoma con ned to the mucosa (pTa) or lamina propria (pT1), and the rest of patients had muscle invasive bladder carcinoma (MIBC; T2-T4) at the time of diagnosis [2].
According to the characteristics of pathology and clinical prognosis, transurethral resection of bladder (TURB) was performed in NMIBC patients, and then intravesical instillation of Bacille Calmette-Guérin vaccine or mitomycin C was performed [3]. Although radical cystectomy was effective in patients with MIBC, approximately one-half of patients would have metastasis and recurrence and eventually die of the disease [4]. To date, there were no established biomarkers that could be able to predict progression. Therefore, early diagnosis and effectual therapy of MIBC required molecular prognostic biomarkers.
Forkhead box protein M1 (FOXM1) was known as a predictive factor for tumor progression, and its expression was relevant to a bad prognosis in numerous cancers, such as prostate carcinoma [5], breast carcinoma [6], hepatic carcinoma [7], renal carcinomas [8], and melanoma [9]. However, the correlation between the FOXM1 expression pattern in different human MIBC tissues and its role in carcinoma development was still unclear. In addition, other studies reported that FOXM1 promoted the nuclear localization of β-catenin nuclear and controlled the expression of Wnt target gene and tumorigenesis [10].
The purposes of the present work were rst to study, by immunohistochemistry (IHC), FOXM1 and βcatenin expression according to clinical and pathological parameters and second to evaluate whether FOXM1 and β-catenin could help predict long-term survival outcomes in MIBC patients.

Materials And Methods
Patients 121 patients were enrolled totally in our study, who were rst diagnosed with MIBC (stage pT2 and pT4) at a single academic center from 2011 to 2016. The inclusion criteria were as follows: all patients were con rmed by adequate imaging and clinical examination, nally diagnosed as MIBC by pathological examination, and then discharged patients were followed up or examined for one or more times. We excluded patients who had previously been diagnosed with any stage of MIBC, received radiotherapy, chemotherapy or immunotherapy, and suffered from other cancers. A pathologist working in uropathology graded and reassessed histopathological variables of the patients included totally in our study on the basis of the World Health Organization (WHO) classi cation criteria of 1973 and 2016. Our research was approved by the Second Hospital of Lanzhou University and then followed up retrospectively.

Immunohistochemical staining
Performing IHC as previous description. Tissues were incubated with anti-FOXM1 and β-catenin primary antibodies overnight at 4°C (diluted at 1:50 and 1:100 respectively). We got the FOXM1 and β-catenin primary and the secondary antibodies from Cell Signaling Technology (Beverly, MA, USA).

Evaluation of IHC
The slides were reviewed and scored parallel by two researchers without knowledge of the clinical data from patients, the product of the proportion of positive tumor cells and the staining intensity was used as standards of grading. The proportion of positive tumor cells was divided into 4 categories: 1 (< 25%), 2 (25-50%), 3 (50-75%), and 4 (> 75%). The corresponding scores of staining intensity for no, shallow, medium and corresponding deep staining were 0, 1, 2, 3. The ultimate immunostaining score for each protein was calculated as these two values, providing an allred total score ranging from 0 to 12. So the ultimate staining index is 0-3 for low protein expression, while the ultimate staining index is 4-12 for high protein expression.

Statistical analysis
Chi-square (χ2) tests was used to test the relation between the expression of FOXM1 and β-catenin and characteristics of clinicopathology. Overall survival (OS) was analyzed by Kaplan Meier method and Log rank test. Univariate and multivariate survival analysis was performed by Cox regression analysis. For whole statistical tests, P value was bilateral, P < 0.05 was con rmed to be statistically signi cant. SPSS version 22.0 software was used for statistical analysis (SPSS Inc.).

Results
The positive expression of FOXM1 and β-catenin in tissues of MIBC and their relationship with clinic pathology Table 1 showed the clinicopathological variables of all patients included. We enrolled 121 MIBC patients (median age 55.5 years; range 31-91 years), including 80 men and 41 women in this study. The median follow-up period was 29.5 months (interquartile range 1-96 months).
Potential relationship between FOXM1 or β-catenin and other clinicopathological factors was explored (Table 1). Our results suggested that elevated expression of FOXM1 or β-catenin was signi cantly related to high tumor stage, recurrence status, tumor size, and lymph node metastases (LNM) in MIBC (all P < 0.05). In addition, the expression of β-catenin had relevance to tumor grade (P < 0.01). Moreover, there was a signi cant relation between high FOXM1 and low β-catenin levels (P < 0.001) by using Spearman rank correlation coe cient analysis.

FOXM1 and β-catenin expression in MIBC and normal bladder tissues
The expression of FOXM1 protein was assessed in the nuclei of primary bladder cancer cells, and the expression of β-catenin protein in tumor cells had different degrees of cytoplasmic and membrane staining characteristics, as shown in Figure 1A and B. IHC was performed to test FOXM1 and β-catenin expression in 60 patients with MIBC (Table 2), and we found high FOXM1 protein expression in 44 (73.3%) patient tissues and in 24 (40.0%) normal bladder tissues. The difference between these two groups had statistical signi cance (P=0.031). Among 60 MIBC patient tissues, 45 (75.0%) had high βcatenin protein-expression, which was signi cantly higher than that in the adjacent normal control tissues (25, 41.7%) (P<0.047).

Correlation between FOXM1 and β-catenin expression with OS in MIBC
We found that upregulate of FOXM1 and β-catenin expression (Fig 1C and D) were relevant to poor OS in MIBC patients by the Kaplan-Meier analyses (P < 0.05). The median survival for 121 patients from our research was 66 months. The low-FOXM1 group exhibited signi cantly longer OS (median OS = 74.5 months; mean OS = 72.74 ± 16.19 months) than the high-FOXM1 group (median OS = 64 months; mean OS = 60.71 ± 21.69 months) (P < 0.001). The high β-catenin group exhibited signi cantly shorter OS (median OS = 59 months; mean = 59.54 ± 20.31 months) than the low β-catenin group (median OS = 75.50; mean = 72.41 ± 18.73 months) (P < 0.001). Cox proportional hazard regression model was used to evaluate the effect of FOXM1 and β-catenin expression on OS, as shown in Table 3 Discussion FOXM1 activated genes that regulated tumor cell cycle progression, and the over-expression of FOXM1 was relevant to unfavourable survival outcomes of numerous tumors [11]. Yang et al. [12] reported that in bladder carcinoma cells, downregulation of FOXM1 suppress cell migration and invasion. Roh et al. [13] reported that FOXM1 can be used as a useful tumor marker that are associated with anticancer drug resistance properties in NMIBC patients. Chen et al. [14] con rmed upregulation of FOXM1 was related to bad progression free survival of NMIBC patients. However, the role and clinical value of FOXM1 expression in MIBC were still not investigated in depth.
Our study results showed the following: rst, high expression of FOXM1 predicted a shorter survival time in patients with MIBC; second, upregulation of FOXM1 was relevant to higher tumor stage, tumor grade and lymph node metastases for MIBC patients but was not related to age, tumor size, histologic type or recurrence status; third, compared with normal tissues, the expression of FOXM1 was increased in MIBC tissues, and was positively relevant to the expression of β-catenin. The effect of FOXM1 on the OS of MIBC patients was investigated in this retrospective cohort study. To explore the transformation e ciency of them, the clinicopathological characteristics of FOXM1 and β-catenin were compared in multivariate analysis. FOXM1 promoted the nuclear localization of β-catenin and cancer cell tumorigenesis [10]. High FOXM1 transcription was relevant to higher tumor stage and tumor grade. However, FOXM1 was demonstrated to be an independent OS predictor in our study, which was better than β-catenin in MIBC. FOXM1 protein expression in bladder carcinoma tissues was detected by IHC and was highly expressed in 73.3% of 60 patients with MIBC, which was signi cantly higher than that in normal bladder tissues (40.0%). This result indicated that FOXM1 protein abnormal expression was involved in the occurrence of MIBC, an observation that was entirely consistent with the ndings of Zhao et al [15] in muscleinvasive bladder cancer. We found that high FOXM1 protein expression was related to the tumor stage, degree of differentiation, LNM, and survival outcome of MIBC patients. This result agreed with the study by Gong et al [16], and the present study also demonstrated that the upstream and downstream regulatory factors of FOXM1 (such as FOXO3, AKT and PI3k) might be valuable drug targets, which was worthy of further study for bladder cancer. Given these ndings, FOXM1 could be considered an oncogene, and its activation may contribute to MIBC progression and a poor prognosis. The biological mechanism of FOXM1 explained its prognostic signi cance in NMIBC. FOXM1 was known as an important transcriptional factor, which provided a balanced transcription program to guarantee suitable growth and maturation during embryonic development, and provided the basis for better diagnosis and treatment of tumor, congenital diseases and other developmental diseases [17]. FOXM1 was suspected to play a role in determining phenotype and promoting invasiveness in the development of molecular subtypes of BC [18].
The Wnt/β-catenin pathway has been related to the development and progression of BC [18,19]. In this research, we found that the high expression of β-catenin protein was apparently higher in MIBC tissues than in pair-matched adjacent normal bladder tissues. Moreover, the high expression of β-catenin was relevant to tumor stage, tumor grade, LNM, and a poor prognosis in MIBC. Our ndings agreed with those of Elsherif et al [20], who reported that β-catenin expression was positively correlated with tumor stage, LNM, and poor OS. These results indicated that abnormal β-catenin expression was involved in the occurrence and metastasis of NMIBC, in agreement with the report of Senol et al [21]. This study also demonstrated a signi cant inverse relationship between the expression of FOXM1 and β-catenin from MIBC, raising the possibility that β-catenin plays a role in MIBC progression and prognosis.
Our study had several limitations, including the small sample size and the proportion of patients who were lost to follow up due to many socioeconomic factors. Second, we performed correlation analyses using prognostic information, but owing to these data were not available, we could not consider indicators such as recurrence free survival rate and progression free survival rate. Third, this study was a retrospective single center research. Therefore, more comprehensive researches are needed including larger patient population to assess IHC markers.
In conclusion, this was the rst study to analyze the prognostic signi cance of two important cancer biomarkers in a large cohort of MIBC patients. We showed that FOXM1 is a better predictor of OS in MIBC patients than β-catenin. The ndings in our study highlighted the potential effect of FOXM1 and β-catenin on prognosis for MIBC, and suggested that they may be useful molecular markers of progression in MIBC. According to these results in our research and others, FOXM1 and β-catenin may be predictive factors in MIBC and serve as novel targets for clinical treatment. Identi cation of new molecular markers may also be the rst step in precisely de ning the high-risk molecular pro ling of MIBC.