Prognostic Nomogram Incorporating Immunohistochemical Results for the Overall Survival of Patients with Bladder Cancer.

Objectives: In this study, we want to combine GATA3, VEGF, EGFR and Ki67 with clinical information to develop and validate a prognostic nomogram for bladder cancer. Methods: A total of 188 patients with clinical information and immunohistochemistry were enrolled in this study, from 1996 to 2018. Univariable and multivariable cox regression analysis was applied to identify risk factors for nomogram of overall survival (OS). The calibration of the nomogram was performed and the Area Under Curve (AUC) was calculated to assess the performance of the nomogram. Internal validation was performed with the validation cohort., the calibration curve and the AUC were calculated simultaneously. Results: Univariable and multivariable analysis showed that age (HR: 2.229; 95% CI: 1.162-4.274; P=0.016), histology (HR: 0.320; 95% CI: 0.136-0.751; P=0.009), GATA3 (HR: 0.348; 95% CI: 0.171-0.709; P=0.004), VEGF (HR: 2.295; 95% CI: 1.225-4.301; P=0.010) and grade (HR: 4.938; 95% CI: 1.339-18.207; P=0.016) remained as independent risk factors for OS. The age, histology, grade, GATA3 and VEGF were included to build the nomogram. The accuracy of the risk model was further verified with the C-index. The C-index were 0.65 (95% CI, 0.58-0.72) and 0.58 (95% CI, 0.46-0.70) in the training and validation cohort respectively. Conclusions: A combination of clinical variables with immunohistochemical results based nomogram would predict the overall survival of patients with bladder cancer.


Introduction
With increasing incidence and mortality, bladder cancer (BCa) has become one of the most important public health problems. With over 549,000 new cases and 200,000 deaths estimated worldwide, bladder cancer ranking in 10th place for incidence and 14th place for mortality [1].In respect to disease management, surgery, chemotherapy and immunotherapy are widely used clinically [2].
However, prognosis after these therapies is unsatisfactory, with a high incidence of recurrence and mortality. Therefore, the prognostication and risk assessment is crucial for treatment decisions.
The American Joint Committee on Cancer (AJCC) TNM staging system is used widely to predict disease recurrence. To predict the probabilities of disease recurrence and progression, the European Organization for Research and Treatment of Cancer (EORTC) also developed a scoring system and risk tables. What's more, the Bladder Cancer Research Consortium (BCRC) and the International Bladder Cancer Nomogram Consortium (IBCNC) also constructed risk models respectively [3]. On the one hand, these models show advantages in estimating prognosis in a certain population. On the other hand, they never include novel information such as biomarkers, genetic markers and molecular classifications and so on.
Risk models often include p53 [4],SMAD6 [5], MMP11 [6], Ki67 [7] and so on to predict prognosis in recent studies. Although GATA3 has been validated associated with bladder cancer prognosis and shows a protective effect [8,9], which has not been involved in risk models ever. In this study, we combine GATA3, VEGF, EGFR, Ki67 and clinical information to develop and validate a prognostic nomogram for bladder cancer.

Material And Methods Patients
This retrospective study was approved by the Institutional Review Board and the patient informed consent was obtained before surgery. We retrospectively identified 4362 bladder cancer patients in the pathology database from January 1996 to July 2018. A total of 188 patients with clinical information and immunohistochemistry were enrolled in this study finally according to the following inclusion criteria: (1) Pathology confirmed bladder cancer; (2) The pathological results showed immunohistochemical information; (3) Clinical information available; (4) The patients with follow up data. In this study, patients were distributed into training cohort and validation cohort according to a ratio of 7:3. The flowchart is shown in Fig. 1.

Recorded data
We collected clinical information from medical records, pathological reports and telephone follow-up.
Variables which had been widely studied in bladder cancer were included as explanatory variables in our study. These factors include: The superficial stage included pTa, pT1 and TIS and the invasive stage incorporated pT2, pT3 and pT4. Squamous carcinoma, adenocarcinoma, sarcomatoid carcinoma, neuroendocrine carcinoma and so on were defined as other histology. Cases with no, focal or suspicious GATA3 expression were defined as low expression, meanwhile cases with exact GATA3 expression(include "+,++ +++ ++++") were defined as high expression, the same with VEGF, EGFR and Ki67. The study response variable was the overall survival, which was defined as from diagnosis of bladder cancer to death or censoring.

Construction of the clinical factor model
Univariable cox regression analysis was used to identify factors that affect overall survival.
Multivariable cox regression analysis was applied to develop the risk models by using the significant factors (P < 0.1) from the univariable Cox regression analysis as inputs. Hazard ratio (HR) and 95% confidence intervals (CI) were calculated for every dependent variable.

Validation of the nomogram
The calibration of the nomogram was performed and the Area Under Curve (AUC) was calculated to assess the performance of the nomogram. Internal validation was performed with the validation cohort., the calibration curve and the AUC were calculated simultaneously.

Statistical analysis
Statistical analysis was performed with SPSS (version 22.0, IBM). Univariable and multivariable cox regression analysis were used to screen risk factors to build nomogram. Figures were plotted using R statistical software (version 3.5.3, https://www.r-project.org). The nomogram development and calibration plots were performed using the "rms" package and the receiver operating characteristic curves (ROC) was plotted using the "pROC" package. The Kaplan-Meier survival curve was calculated using the "survival" package. The p < 0.1 was considered statistical significance in univariate analysis, while p < 0.05 was considered statistical significance in the multivariate analysis.

Results
Clinicopathological characteristics of patients 188 patients met our inclusion criteria, who were divided into training (124 cases) and validation (64 cases) sets. The clinical variables of the patients in the training and validation cohorts are provided in Table 1. All factors were found no difference between two groups. The mean overall survival was 44.07 months and 36.69 months in the two groups, which show no difference (P = 0.262).  These variables were included in the multivariable analysis, which showed that age (HR: 2.229; 95%  Table 2).
The nomogram construction and assessment The age, histology, grade, GATA3 and VEGF were included to build the nomogram (Fig. 2). The accuracy of the risk model was further verified with the C-index and calibration curve. The C-index

Discussion
The EORTC risk tables and the Club Urológico Españolde Tratamiento Oncológico (CUETO) scoring model are most widely used and validated tools for prognosis prediction in bladder cancer. However, they all overestimate the recurrence and progression risks in patients [10]. There are limited studies that have incorporated biomarkers into predictive models. What's more, GATA3 has never been included in nomogram. Thus, we tried to combine clinical factors and pathological immunohistochemical results to construct a practical predictive model. indicators to optimize the model. We will collect pathological sections from the department of pathology to further fulfill the pathological data. Finally, an external analysis was not performed with an independent validation cohort to further verify the model. In summary, we need to validate and optimize the risk model with a larger, multicenter, prospective study.
In conclusion, we have shown that a combination of clinical variables with immunohistochemical results based nomogram would predict the overall survival of patients with bladder cancer. This kind of model may serve as an effective tool to make clinical decisions. Figure 1 The flowchart of incorporating patients.

Figure 2
Nomogram for the prediction of the overall survival of the patients with bladder cancer, where age, histology, grade, GATA3 and VEGF define the overall survival probability.