Squamous Differentiation in pT1 Low-Grade Bladder Urothelial Carcinoma is Associated with High Risk of Recurrence

Background A large number of studies have conrmed that squamous differentiation is closely related to the prognosis of bladder cancer, and studies also showed that it is an independent predictor for the prognosis of such disease. However, the role of squamous differentiation in pT1 low-grade bladder cancer (LGBC) has not been reported yet. This study mainly intends to explore the clinical signicance of squamous differentiation in pT1 LGBC. Methods The clinicopathological data of 582 patients with pT1 LGBC from April 2007 to December 2018 were retrospectively analyzed, including 51 cases pathologically diagnosed as pT1 LGBC with squamous differentiation, and 531 cases as pT1 LGBC without squamous differentiation. The clinicopathological characteristics and prognosis of patients in the two groups were compared. Results 51 cases (8.7%) were pathologically diagnosed as pT1 LGBC with squamous differentiation. As compared with non-squamous differentiation, squamous differentiation was closely related to the age of patients (P=0.009) and postoperative recurrence (P<0.001). Further studies on the pT1 LGBC recurrence group showed that the patients in the squamous differentiation group were more prone to multiple recurrences (P=0.033) and invasive recurrence (P<0.001) after surgery. The Kaplan-Meier analysis indicated that pT1 LGBC patients with squamous differentiation had shorter recurrence-free survival (RFS) time. The Cox proportional hazard analysis conrmed that squamous differentiation was an independent predictor of RFS time in patients with pT1 LGBC. Conclusions Squamous differentiation is closely related to postoperative recurrence of pT1 LGBC, especially multiple recurrences and invasive recurrence after surgery. Squamous differentiation is an the RFS

group showed that the patients in the squamous differentiation group were more prone to multiple recurrences (P=0.033) and invasive recurrence (P<0.001) after surgery. The Kaplan-Meier analysis indicated that pT1 LGBC patients with squamous differentiation had shorter recurrence-free survival (RFS) time. The Cox proportional hazard analysis con rmed that squamous differentiation was an independent predictor of RFS time in patients with pT1 LGBC.
Conclusions Squamous differentiation is closely related to postoperative recurrence of pT1 LGBC, especially multiple recurrences and invasive recurrence after surgery. Squamous differentiation is an independent predictor of the RFS time in patients with pT1 LGBC.

Background
The incidence of bladder tumors ranks rst among urinary tract tumors, which seriously threatens the health of people all over the world [1,2]. The majority of bladder tumors are urothelial carcinomas [3], and tissue-type variations with squamous differentiation have been found in about 20%-40% of cases [4][5][6][7].
The kind of tissue-type variation is closely related to high invasiveness and easy recurrence of tumor [8].
A large number of studies have clari ed that squamous differentiation is an important prognostic factor of urothelial carcinoma [9][10][11][12][13], but there is no relevant research on squamous differentiation in patients with pT1 LGBC. Our study intends to clarify the role of squamous differentiation in pT1 LGBC, and to our knowledge, this is the rst study focusing on the clinical signi cance of squamous differentiation in pT1 LGBC.

Patients and specimens
We retrospectively analyzed the clinical and pathological data of 582 patients pathologically diagnosed as pT1 LGBC in our hospital from April 2007 to December 2018. Patients with the history of urothelial carcinoma or associated urothelial carcinoma were excluded. The patients were divided into two groups based on whether they had squamous differentiation according to the pathological diagnosis. Surgical treatment and intravesical chemotherapy were performed to all patients. The indicators for evaluating the disease progression and prognosis of the patients included age, gender, recurrence, lymphatic vascular invasion and squamous differentiation. Cystoscopy was conducted regularly after the operation.
Pathological biopsy was performed if abnormalities were found, and the recurrence was judged based on the results of the biopsy: the urothelial carcinoma in the pathological report of biopsy is de ned as recurrence, e.g. the second pathological grade higher than the rst pathological grade is de ned as invasive recurrence. RFS (recurrence free survival) is de ned as the time after patients' operation and before the rst recurrence of bladder cancer. The pathological classi cation is based on the WHO 2004 version of the malignant degree classi cation system of bladder urothelial carcinoma.

Pathology
All surgical specimens were completely presented to a pathologist, and the pathologist reviewed each pathological section one by one, in order to fully evaluate the clinical stage and pathological grade of tumors. Then two pathologists selected three representative sections of each specimen for independent review, to nally determine the pathological results and squamous differentiation. Routine hematoxylineosin (H&E) staining was used to diagnose the presence of lymphatic vascular invasion and squamous differentiation. The presence of intercellular bridges or keratinization was a sign of squamous differentiation.

Statistical methods
The study data was statistically analyzed by R (Version 3.6.3). Chi-square test and T-test were used to analyze the relationship between squamous differentiation and clinicopathological characteristics. The Kaplan-Meier method was used to analyze the overall survival time and the RFS time, and the difference analysis was calculated by log-rank test. The Cox proportional hazard analysis was used for the univariate and multivariate analysis. P-value was bidirectional, and when P<0.05, the difference is deemed statistically signi cant.

Clinical characteristics
Clinical and pathological characteristics were presented in Table 1. There were a total of 582 patients pathologically diagnosed as pT1 LGBC, with the average age of 62.9, including 466 males and 116 females, with the ratio of male to female of 4:1. 51 patients (8.7%) were pathologically diagnosed as pT1 LGBC with squamous differentiation. Squamous differentiation had no signi cant relationship with the gender (P=0.159) or lymphatic vascular invasion (P=0.286), but it was correlated with the age (P=0.009).
Elderly patients were more prone to squamous differentiation. As compared with patients of the group without squamous differentiation, patients of the group with squamous differentiation were more prone to tumor recurrence after surgery (35% VS 11%, P<0.001).
During the follow-up period, a total of 78 pT1 LGBC patients experienced postoperative recurrence. Detailed clinical and pathological characteristics were shown in Table 2. As compared with the group without squamous differentiation, the group with squamous differentiation was more prone to multiple recurrences (56% VS 28%, P=0.033) and invasive recurrence (44% VS 18%, P<0.001) after surgery.

Oncological outcome
The median follow-up time was 58 months (6-136 months). About 4.8% (28 cases) patients died during the follow-up period, and about 13.4% (78 cases) patients had tumor recurrence after surgery. The veyear overall survival rate of patients with pT1 LGBC was 97.8%, that of the group of tumor patients with squamous differentiation was 92.2%, and that of the group of tumor patients without squamous differentiation was 98.3%. There was no signi cant difference in the overall survival rate between the two groups (P=0.065, Fig. 1). However, as compared with the group without squamous differentiation, the group with squamous differentiation had a shorter RFS time (P<0.001, Fig. 2) and a shorter median recurrence time (30 VS 37 months). The univariate COX analysis showed squamous differentiation (HR, 2.834; 95% CI, 1.690-4.754; P = 0.001, Table 3). Lymphovascular invasion (HR, 1.79; 95% CI, 1.137-2.817; P = 0.012, Table 3) was an important factor to predict the RFS time of patients with pT1 LGBC. The univariate and multivariate Cox analyses con rmed that squamous differentiation (HR, 2.969; 95% CI, 1.765-4.999; P<0.001; Table 3) was an independent predictor of RFS time, but age and gender were not important factors affecting the RFS time.

Pathology
Histological variations in squamous differentiation were observed by pathologists in all the 51 cases. When an obvious intercellular bridge and/or keratinization were observed in tumor cells, it was considered that the tumor cells had squamous differentiation. Its main histological characteristics are: 1. Clusters or nests of large and middle nuclear in ltrating tumor cells, often with nucleoli; 2. Unclear separation of cytoplasmic amphiphilic or eosinophilic cytoplasmic background. HE stained sections were used to con rm the presence of lymphatic vascular in ltration.

Discussion
Mazzucchelli and Bacchi rst reported squamous differentiation as an indicator of the severity and prognosis of urothelial carcinoma [14]. Subsequently, in studies on bladder urothelial carcinoma, squamous differentiation has attracted researchers' more and more attention. Recently, several studies have shown that squamous differentiation is an independent prognostic factor of urothelial carcinoma [15][16][17], consistent with the results of the study of Li and Yu, who demonstrated by univariate and multivariate analyses that squamous differentiation was an independent prognostic factor of pT1 bladder urothelial carcinoma, and was associated with high risk of recurrence and poor prognosis [10]. However, a large number of literatures have proved that squamous differentiation is only one of the factors affecting the prognosis. For example, Xu et al. analyzed 869 patients with TURBT and found that squamous differentiation was only an independent predictor of postoperative recurrence, which was closely related to tumor recurrence and RFS [17]. Minato et al. retrospectively analyzed the clinical and pathological data of 80 patients who had undergone radical bladder cancer resection, and the results showed that squamous differentiation was an independent predictor of the overall survival and RFS time [18].
There is no relevant research report on the role of squamous differentiation in pT1 LGBC. It was found in our study that elderly patients with bladder cancer (P=0.159) were more prone to squamous differentiation. As compared with the group without squamous differentiation, the group of pT1 LGBC patients with squamous differentiation was more prone to tumor recurrence after surgery (P<0.001). There was no statistical difference between squamous differentiation and the gender of patients (P=0.159) or lymphatic vascular invasion (P=0.286). Further studies of the clinical and pathological data of 78 patients with recurrence showed that the group of patients with squamous differentiation was more prone to multiple recurrences (56% VS 28%, P=0.033) and invasive recurrence (44% VS 18%, P<0.001) after surgery. Therefore, squamous differentiation was closely related to postoperative recurrence of patients with pT1 LGBC, and was an important factor affecting postoperative recurrence of patients.
The study of Li et al. con rmed that squamous differentiation is an important indicator to predict the sensitivity of chemotherapy in the bladder, which is closely related to the insensitivity and poor prognosis of chemotherapy [19]. Lin et al. conducted a meta-analysis of a total of 13,284 patients with LGBC from 19 studies, and the results showed that squamous differentiation is not an independent predictor of the prognosis of bladder cancer, but is associated with poor clinicopathological characteristics, high risk of recurrence and poor prognosis [20]. The results of our study support the above conclusions. The ve-year overall survival rates of the groups of pT1 LGBC patients with squamous differentiation and without squamous differentiation were 92.2% and 98.3% respectively. There was no signi cant difference in the overall survival rate between the two groups (P=0.065, Fig. 1). However, as compared with the group without squamous differentiation, the group with squamous differentiation had a shorter RFS time (P<0.001, Fig. 2) and a median recurrence time (30 VS 37 months). The results of the univariate COX analysis showed squamous differentiation (HR, 2.834; 95%CI, 1.690-4.754; P=0.001; Table 3) is an important factor affecting the RSF time of patients with pT1 LGBC. The Cox proportional hazard model was used for the multi-factor analysis, further con rming the squamous differentiation (HR, 2.969; 95% CI, 1.765-4.999; P<0.001; Table 3) is an independent predictor of RFS time.
Sanderson et al. conducted a follow-up analysis of 1069 bladder cancer patients who had undergone radical bladder cancer resection, and the results showed that the median recurrence time was 39.6 (4.8-112 months) months [21]. Mariappan et al. conducted a 25-year follow-up of 115 patients with bladder cancer, and the results showed that the majority of patients with bladder urothelial carcinoma recurring within 5 years after surgery, and 12% of them had disease progression [22]. Our follow-up results are consistent with the results of the study of Sanderson and Mariappan. Analysis of the follow-up results of 78 patients showed that 64 (82%) patients with pT1 LGBC recurrence within 5 years after surgery. The squamous differentiation group had a shorter median recurrence time (30 vs. 37 months), but 18% patients with pT1 LGBC still had tumor recurrence after 5 years. This prompts urologists that patients with pT1 LGBC may need a longer period of close follow-up after surgery.

Conclusions
This is the rst study focusing on the clinical signi cance of squamous differentiation in patients with pT1 LGBC. Squamous differentiation is closely related to postoperative recurrence of pT1 LGBC, especially multiple recurrences and invasive recurrence after surgery. Squamous differentiation is an independent predictor of the RFS time in patients with pT1 LGBC.

Consent for publication
All authors have reviewed the nal version of the manuscript and agree with its content and submission.

Availability of data and materials
The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.    Kaplan-Meier curve of the recurrence-free survival rates for the squamous differentiation-negative and squamous differentiation-positive (p<0.001). SDN: squamous differentiation-negative; SDP: squamous differentiation-positive.