Evaluation of the value of Preoperative Sialic Acid Levels in Diagnosis and Localization of Urothelial Tumors

Objective: To explore SA levels in the serum of urothelial tumor patients and their correlation with clinical pathological features and localization. Materials and Methods: Our research retrospectively collected data from 591 patients with urothelial tumors between July 2014 and April 2018. The SA levels in the serum of urothelial tumor patients and their correlation with clinical pathological features and localization were investigated. Univariate and multivariate logistic regression analyses were further performed to identify independent associations. Results: The levels of SA were significantly associated with the malignant degree (tumor grade and infiltration) of bladder cancer and tumor localization (all p < 0.05). The multivariate logistic regression model showed that SA levels were independently associated with the presence of high-grade urothelial carcinoma (BUC: HR = 1.941, UTUC: HR = 3.820, all p <0.05) and upper urinary tract urothelial carcinoma (HR = 2.047, p < 0.05). Finally, we validated the diagnosis and localization value of SA in an independent cohort from another institutions. Conclusions: Elevated serum SA levels are an independent predictor of high-grade urothelial carcinoma and upper urinary tract urothelial carcinoma, indicating that SA levels may be a potential biomarker for the diagnosis, prognosis and localization of urothelial tumors.


Introduction
Urothelial carcinoma (UC) is the ninth most common cancer and is the thirteenth most common cause of death due to cancer, accounting for approximately 430,000 new cases and 165,000 mortalities globally [1,2]. In China, there are 80,500 incident cases and 32,900 mortalities caused by UC each year [3]. UC is the most frequent type of cancer of the bladder (BUC) and upper urinary tract [4], while the morbidity and mortality of UC have gradually increased over recent years. Therefore, early detection and diagnose have an important significance to the prognosis and quality of life of patients with UC. Cytological experiments and biopsies under an ureteroscope and a bladder scope are currently the two main means of UC early detection and diagnosis [5]. However, use of the ureteroscope and bladder scope not only is invasive but also has a high rate of missed diagnosis, especially for UC in situ and upper tract urothelial carcinoma (UTUC) [5,6]. Therefore, the search for a specific, highly sensitive marker, that can predict the biological behavior of UC, as an index for predicting tumor progression and metastasis, in order to guide clinical diagnosis and therapy is particularly important.
Sialic acid (SA) is a monosaccharide with a nine-carbon backbone that occupies the terminal position on macromolecules and cell membranes [7]. SA is involved in autoimmune diseases, microbial invasion, virulence pathogenesis and tumor growth [8]. Malignant cells often have an increased concentration of SA on the surface and secrete SA to increase the concentration in blood [9]. In previous Ivyspring International Publisher studies, it was found that the lipid-bound subfraction of SA concentration was of limited value for detecting early stages of genitourinary malignancies [10]. However, small sample sizes, no further analyses of multiple factors, no assessment of localization values and no validated cut-off values existed in these studies. Therefore, the present study aimed to assess SA levels in patients with UC to determine the pathogenesis and development of UC.

Patients
From July 2014 to April 2018, the clinical data of 591 patients with urothelial carcinoma who underwent surgical treatment in the Department of Urology, Qilu Hospital of Shandong University, were retrospectively analyzed. Among 591 UC patients, 406 patients were diagnosed with bladder tumors, 99 with ureteral tumors and 86 with renal pelvic tumors. Additionally, a total of 322 consecutive UC patients (including 185 BUC patients, 77 ureteral carcinoma patients and 60 renal pelvis carcinoma) treated at Shandong Provincial Hospital between July 2017 to Mar 2021 were used for validation set.
The inclusion criteria for the patients enrolled in the study were as follows: • Patients with a pathological diagnosis of urothelial tumors, including bladder tumors, ureteral tumors and renal pelvic tumors, and patients who underwent surgical treatment at Qilu Hospital of Shandong University; • Patients with complete clinical and pathological data, such as preoperative blood parameters, postoperative pathological results and other basic information; • Patients who did not undergo any adjuvant treatments, such as radiotherapy or chemotherapy, before surgery.
The exclusion criteria for the patients enrolled in the study were as follows: • The coexistence of any other malignant tumor; • A pathological diagnosis of nonurothelial tumors; • A history of bladder urothelial tumors and upper tract urothelial tumors; • A history of cerebral infarction and cerebral hemorrhage within the last 1 month or myocardial infarction within the latest 6 months; • The administration of procoagulant or anticoagulant drugs within the past 2 weeks; • The presence of obvious infection or inflammation.

Data collection
Clinical data including patient age at the time of diagnosis, sex, smoking history, routine blood examination results (white blood cell count, platelet count, plasma fibrinogen level, etc.), and tumor characteristics were obtained from the electronic patient records at our institution.

SA measurement
Before any clinical treatment, 5 mL venous blood samples from patients after 8 hours of fasting were collected for the evaluation of SA levels. Blood samples are stored in tubes containing clotting activators and gels. The Roche Cobas 8000 automatic analyzer was used for the determination of SA in serum. SA levels of 45.6-75.4 mg/dL were defined as normal.

Statistical analysis
The Kolmogorov-Smirnov test was adopted to assess whether the level of SA conformed to a normal distribution, and the values were presented as the mean ± standard deviation (SD). Student's t-test and one-way ANOVA were used for normally distributed data; if the data distribution was not normal, the Mann-Whitney U-test and Kruskal-Wallis H-test were employed for the nonparametric analysis. Data were analyzed and processed using the Statistical Package for Social Sciences version 22.0 (SPSS Inc., Chicago, IL, USA). P values < 0.05 in two-tailed tests were considered statistically significant.

Clinicopathological characteristics of patients
Our research collected data from 591 patients with urothelial tumors; 448 (75.80 %) were male, and 143 (24.20%) were female. The median age was 65 years, ranging from 17 to 94 years. The mean serum SA level of all patients was 56.41 ± 9.85 mg/dL and 61.40 ± 12.93 mg/dL for BUC and UTUC respectively. There was a significant correlation between serum SA levels and sex, WBC, PLT, AKP, LDH, PFL, and tumor size (all p < 0.05, Fig. 1 Fig. 2A). The clinical characteristics of the enrolled patients are shown in Table 1.

The role of SA level in the localization diagnosis of UC
Of all patients, 406 (68.70%) had bladder tumors, 99 (16.75%) had ureteral tumors, and 86 (14.55%) had renal pelvic tumors. Our study found that the serum SA level in the renal pelvic tumor and ureteral tumor groups was significantly higher than that in patients with bladder tumors (p < 0.05, Fig. 2D and Table 2), indicating that the serum SA level may have valuable applications in the localization diagnosis of urothelial tumors.

Correlation analyses
Linear correlation analyses were conducted to further explore correlations between SA levels and WBC, PLT, LDH, AKP, and PFL in UC patients. We found that there was a significant positive relationship between serum SA levels and WBC (r = 0.300, p <0.05, Fig. 3A

The value of serum SA level in accurate location and qualitative diagnosis for UC patients
ROC curve analysis for preoperative SA levels was used to evaluate the value of localization and qualitative diagnoses. The optimal threshold values for preoperative SA levels were > 52. 27 Table 4), and was a predictor for the tumor localization (UTUC: HR = 1.660, p < 0.05, Table 4).

Validation in an independent cohort
To find out whether the effect of SA was applicable for other UC cases, we reviewed the medical records of 322 UC patients from another institutions as an independent cohort. Our study validates the serum levels of SA in advanced-stage patients were significantly higher than those in earlystage patients ( Fig. 5B). In addition, serum SA level was obviously elevated in UTUC patients compared to BUC patients (p < 0.05, Fig. 5C). The detailed clinicopathological data was presented in Table 5 and  Table 6.

Discussion
Urothelial carcinoma can be divided mainly into UTUC and BUC, and has a notably high rate of clinical variability, recurrence, progression, and cancer-specific mortality [11,12]. UTUC has symptoms and signs similar to those of BUC [13], whereas UTUC is markedly more aggressive and seems to have a higher burden of nodal metastases at diagnosis than BUC [14,15]. Our current diagnostic and detection methods for UC are invasive, radioactive and expensive procedures [16], such as CT, MRI and pathology. Therefore, there is an urgent need to find an effective, simple and noninvasive method for the diagnosis, orientation and guidance of clinical treatments for UC.      Sialic acids, a family of monosaccharides with negative charges, are typically located at the terminal positions of cell surface glycoproteins and glycolipids [17]. Due to their special physical and chemical properties, such as charge and size, SAs are involved in a variety of biologically important processes and influence both the degree of severity and the progression of disease in a range of illnesses. Moreover, the combinatorial diversity enables SAs to influence the structure and function of glycoproteins and lipids as well as to regulate cell-cell and cellextracellular matrix interactions [18]. For example, sialic acids can mediate pathogen infection, immunogenicity, cell adhesion and migration, vascularization and differentiation and can serve as ligands of sialic acid-binding proteins, such as factor H, selectins and Siglecs [19][20][21].
There are quantitative and qualitative changes in the expression of cellular surface molecules in the canceration course, which is crucial for the unlimited proliferation and malignant behavior of neoplastic cells [22]. Glycoconjugates including glycoproteins and glycolipids, are ubiquitous, essential components of cell membranes and therefore participate in malignant transformation and tumor progression [23]. The aberrant expression of SA facilitates cancer cell migration and metastasis formation and is an important factor in tumor immunological escape and the immunosuppressive microenvironment [24][25][26]. Strikingly, these glycoconjugates are released into the circulation through increased turnover, secretion and/or shedding from malignant cells, which results in increased SA levels in the blood [9,27]. Previous studies have confirmed that there is marked clinical significance for measuring serum SA in the diagnosis and treatment evaluation of several solid tumors, including colorectal cancer [28], osteosarcoma [29], oral squamous cell carcinoma [27], pancreatic cancer [30], melanoma [26], and prostatic cancer [31].
In the present study, we examined preoperative SA levels and other clinical features of 591 patients with newly diagnosed urothelial tumors. We found that the extent of malignancy was significantly correlated with serum SA levels. Meanwhile, we noticed that the SA level in patients with renal pelvis carcinoma and ureteral tumors was higher than that in patients with bladder tumors. To further understand the association between SA levels and other clinical features, linear correlation analyses were performed and showed that serum indicators such as WBC, PLT, LDH, AKP, and PFL were linearly correlated with SA levels. Furthermore, ROC curves further confirmed the practical application value of SA levels in the diagnosis and localization of urothelial tumors. The optimal cut-off values with the maximum Youden index were > 61.60 mg/dL (BUC: high-grade vs. low-grade), 54.35 mg/dL (BUC: MIUC vs. NMIUC), and 60.31 mg/dL (UTUC vs. BUC). Logical regression analysis showed that the SA level was a powerful independent predictor of high-grade UC and UTUC, which supported the potential utility of SA levels in clinical practice as a noninvasive biomarker for the diagnosis and localization of urothelial tumors. Finally, we validated the diagnosis and localization value of SA in an independent cohort from another institution. However, there are still some controversies regarding the quantitative changes in SA occurring in patients with urothelial tumors. In a study performed by Habibi et al. [32], lipid-bound sialic acid (LBSA) and protein-bound sialic acid (PBSA) levels were significantly higher in bladder cancer patients than in normal controls and were significantly correlated with tumor grade, which was in accordance with some earlier reports [10,33]. In contrast, in a report by Lagana et al. [34], the levels of total sialic acid (TSA) and LBSA in patients with bladder tumors were not significantly different from those in normal subjects, which was in agreement with our finding that SA levels in UC patients did not differ significantly from those in patients with papilloma and PUNLMP. In addition, some early studies observed that SA levels were also elevated in benign diseases, such as severe infection [35], type II diabetes [36], rheumatoid arthritis [37] and prostatitis [10], and may serve as a nonspecific acute phase reactant, which may have given rise to bias and disturbed the results of our study. According to the results of our study, we speculate that serum SA levels have strong limitations in the screening of early UC but that they may be useful primarily in the evaluation of the malignant degree of advanced disease and in the localization of urothelial tumors.
However, there were several shortcomings and limitations in this study. First, because the design was analytical and retrospective, the study may have suffered from confounding bias. Besides, the postoperative pathology results collected in clinical practice couldn't provide sufficient information for clinicopathologic stage. Additionally, the gathered clinical information didn't involve patient prognosis, such as relapse-free and overall survival, which would be the focus of our future work.
In conclusion, we believe that SA can be a useful biomarker for the evaluation of the malignant degree of UC and the localization of urothelial tumors. However, further studies are needed to determine whether serum SA levels have applicable clinical value as a single biomarker or as an auxiliary test to traditional approaches in the detection and localization of urothelial tumors.