Background
Non-muscle invasive tumors represent 75% of patients diagnosed with transitional cell carcinoma. Transurethral resection of bladder tumors (TURB), intravesical mitomycin C and bacillus Calmette–Guerin (BCG) instillation represent the current standard treatments for non-muscle invasive bladder cancer (NMIBC) (Babjuk et al.
2017). As 50–80% of pTa NMIBC have cancer recurrence and 10–30% of pT1 and CIS patients show disease progression, clinicopathologic parameters are insufficient for disease prediction (Prout et al.
1992; Ark et al.
2014). Given the biologic heterogeneity of bladder cancer, survival and progression varies even within the same stage. Thus molecular biomarkers are needed to improve prediction of treatment response or even for clinical decision-making in the sense of preemptive biomarkers (Youssef and Lotan
2011; Kluth et al.
2015). Based on recent data, the most promising results were provided by gene expression signatures of TURB samples and by liquid biopsies from urine or blood (Contreras-Sanz et al.
2017; Robertson et al.
2017). The elucidation of valuable invasive or non-invasive biomarkers or drug targets is still at its beginning.
The role of the proto-oncogene forkhead-box M1 (FOXM1) in carcinogenesis and drug resistance development is already well established and has been validated in many cancer types (Dai et al.
2015). FOXM1 originates from the forkhead gene family, was first identified in Drosophila and is characterized by a conserved 100-amino acid DNA-binding motif. It is involved as regulator in embryogenesis and numerous developmental processes (Ye et al.
1997). In adult organs, FOXM1 is mainly expressed in proliferating cells or induced by growth factor release. In this context, it is not surprising that FOXM1 serves as proto-oncogene in most cancers and aberrant expression or mutations constitute the origin of many treatment resistance mechanisms (Kwok et al.
2010; Kalin et al.
2011; Saba et al.
2016). Although FOXM1 is exclusively expressed in dividing cells, its targeting may result in many side effects given its involvement in angiogenesis, cell migration and epithelial–mesenchymal transition (Halasi and Gartel
2013). We have recently shown that FOXM1 is a predictor for overall and disease-specific survival in muscle invasive bladder cancer (MIBC) superior to the TNM staging system and MKI67 (Rinaldetti et al.
2017). A recent TCGA study further underlined the role of FOXM1 as regulator in MIBC (Robertson et al.
2017). As MKI67 is considered as the gold standard biomarker for proliferation and prognosis, the impact of FOXM1 needs yet again to be compared with the later (Rodríguez-Alonso et al.
2002).
Recent findings showed that bladder cancer can be subclassified in molecular subtypes with some similarities to breast cancer subtypes (Choi et al.
2014; Hedegaard et al.
2016; Robertson et al.
2017). These findings open the doors for personalized treatment concepts similar to those in breast cancer. That is why the subtype-specific expression of FOXM1 will be analyzed in this study. As the FOXM1 signaling network represents a valuable and promising target for further cancer treatment personalization, we here investigate its clinical impact in three cohorts with a total of 737 NMIBC patients.
Discussion
In this study, the prognostic and clinical impact of FOXM1 has been investigated retrospectively in a multicenter study and has been validated in published datasets. In order to evaluate its translational benefit, FOXM1 has been compared with MKI67 and relevant clinicopathologic parameters in multivariable analyses. Validation was performed by in silico data based on the Lund and Chungbuk cohort. FOXM1 showed a high correlation with MKI67, underlining its known role in cell proliferation and migration (Hamurcu et al.
2016). High FOXM1 transcript levels correlated with high tumor stage and grade. However, we managed to demonstrate that FOXM1 is an independent predictor for PFS superior to the TNM staging system and MKI67. Indeed, patients of the high-risk group characterized with high FOXM1 expression showed a 6- to 8-fold higher risk of progression (
p < 0.001).
The UROMOL low-risk group comprised 75% of the patients, whereas in the Chungbuk cohort 58% of patients were included. This can be explained by the significantly higher portion of T1 NMIBCs of the Chungbuk cohort.
FOXM1 showed a distinct subtype-specific overexpression in the Class 2, genomically unstable and SCC-like subtypes (
p < 0.001, Figs.
1b,
3). As published previously, the Class 2 subtype from Hedegaard at al. overlapped with the Lund subtypes ‘genomically unstable’ and ‘SCC-like’ (Hedegaard et al.
2016). FOXM1 is suspected to play a phenotype-determining role in the development of the molecular bladder cancer subtypes and promote its aggressiveness (Eriksson et al.
2015). The Class 2 subtype, characterized by an APOBEC-related mutational signature and the upregulation of the ERBB gene family, is known to have poorest PFS rate in NMIBCs (Hedegaard et al.
2016). Thus the overexpression of FOXM1 in this specific subtype seems plausible. FOXM1 is well known for its resistance development against many chemotherapies especially cisplatin. Thus, patients with progression of Class 2 tumors may not benefit from platinum-based chemotherapies. Patients with high FOXM1 expression seemed to profit most from BCG instillation therapy. However, only 8 patients of 346 patients from the low-risk FOXM1 group experienced progression, limiting the statistical power for the low-risk group. Thus, further studies for investigating the impact of BCG instillation therapy on patients with low FOXM1 expression are required.
In recent data, FOXM1 has also been shown to be an independent predictor for OS and disease-specific survival in muscle-invasive bladder cancer, with a subtype-specific expression in the luminal subtype (Rinaldetti et al.
2017). Thus FOXM1 has global impact on bladder cancer given its role in both muscle invasive and non-muscle invasive tumors. Up- and downstream FOXM1 regulators (e.g., FOXO3, PI3k, AKT) may be valuable drug targets and should be further explored also in bladder tumors (Yao et al.
2017).
This study has an exploratory character. FOXM1 needs to be further validated in prospective clinical studies in order to evaluate its impact in MIBC resistance development against platinum-based chemotherapies and to validate its prognostic role in NMIBC BCG instillation treatment. In this study two different quantification platforms (RNA-Seq vs microarray) allowed a distinct FOXM1 risk stratification. In order to translate these findings into clinics, a standardizable FOXM1 qPCR screening is needed for future studies (Rinaldetti et al.
2017).
The impact on MIBC and NMIBC as prognostic biomarker superior to clinicopathologic parameters and MKI67, raises FOXM1 to a crucial biomarker for molecular grading and to valuable drug target in bladder cancer (Radhakrishnan and Gartel
2008).
Conclusions
FOXM1 is a highly prognostic marker for bladder cancer disease progression. It is mainly expressed in Class 2 and genomically unstable molecular bladder cancer subtype. As FOXM1 is a druggable proto-oncogene, the elucidation of its impact on bladder cancer survival may contribute to a further personalization of future NMIBC or MIBC therapy.
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (
http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.