Anterior wall adenocarcinoma of bladder with similar clinicopathological and prognostic characteristics as common bladder carcinomas should not be treated as or classified into urachal adenocarcinomas

Abstract Purpose To discuss whether the dome or anterior wall of bladder adenocarcinoma (BAC) should be classified into urachal carcinoma (UrC) and the relationship of primary tumor location (PTL) as well as treatment with survival. Methods Surveillance, Epidemiology, and End Results 18 database was examined for eligible patients from 1975 to 2016. Patients were classified into adenocarcinoma originating from the urachus (UAC), the dome (D‐BAC), the anterior wall (A‐BAC), and the other sites adenocarcinoma of the bladder (O‐BAC). The clinicopathological features, treatment, and survival were compared among the groups. Results Comparable clinicopathologic features were obtained between UAC and D‐BAC, which were different from those of A‐BAC and O‐BAC; otherwise, the latter two had similar clinicopathologic features. Univariable and multivariable Cox regression analyses indicated that PTL was an independent predictor for survival. O‐BAC conferred the worst prognosis then followed by A‐BAC, D‐BAC, and UAC. For non‐metastatic UAC or D‐BAC, partial cystectomy (with an en bloc resection of the urachus and umbilicus) is optimal for survival. However, the worse survival of non‐metastatic D‐BAC (compared with UAC) suggested different modalities, maybe more intensive surgery approaches, should be considered for D‐BAC. Conclusion This study illustrates that PTL of UAC and BAC was an independent predictor for survival. A‐BAC had comparable characters and prognosis with O‐BAC and should not be classified into and treated as UrC. For non‐metastatic disease, non‐metastatic D‐BAC may need more intensive modality.


| INTRODUCTION
Urachus, arising from the degeneration of allantois during embryogenesis, is a tubular structure that connects the urinary bladder to the umbilicus. When the physiological lumen obliteration fails, the urachal remnant, characterized by a tubular or cystic muscular structure lined by epithelium, persists in a minority of adults. 1,2 Neoplasms originating from the urachal remnant are also defined as urachal carcinoma (UrC). Because it is rare and sometimes tricky to differentiated from bladder adenocarcinoma (BAC), UrC is commonly accounted for bladder carcinoma (accounts for less than 1% of all bladder carcinoma). 1,3,4 However, typical UrC demonstrates clinical and pathologic features that are different from those of bladder carcinomas. [5][6][7][8] Urachal adenocarcinoma (UAC) is the primary pathological type of UrC 9,10 and is less common than BAC. 11 Due to its location relationship with the bladder, the diagnosis of UrC is usually confused, and its diagnosis and treatment also remain a challenge. 12 Several stage classification genres of UrC have been proposed, but the most often applied are the Sheldon and Mayo Staging Systems, 13 including the tumor situated in the bladder dome and/or anterior wall, whose diagnostic criteria are adopted and included in that of World Health Organization (WHO) (2016). The classification of the dome and anterior wall into the UrC might be based on the comprehension of the origin of the urachal ligament, but the rationality of classification is worth further discussion. Definitely, published studies have recently indicated significantly different clinical characteristics and prognoses between the UAC and BAC. 14,15 However, in these studies, the urachus and dome adenocarcinoma of the bladder were usually classified into UAC, showing marked differences in characteristics and prognosis from BAC. Otherwise, cancers from the anterior wall of the bladder were not clearly described. [14][15][16] Up to now, no study has been reported to illustrate the discrepancy of the clinical characteristics and prognosis between the cancers of the dome, anterior wall of the bladder, and urachus separately. The aim of this study, based on the analysis of the clinicopathological features, treatment, and prognostic factors of primary tumor sites of UAC and BAC, is to furtherly discuss whether the dome or/and anterior wall adenocarcinoma of bladder should be classified into UrC.

| Patient selection
The Surveillance, Epidemiology, and End Results (SEER) 18 database, maintained by the National Cancer Institute (Bethesda, MD), was examined for all UAC and BAC patients from 1975 to 2016. The data used and analyzed in our study can be downloaded from the SEER (https://seer.cancer.gov/). Patients who meet the following selection criteria were reviewed and included in this study: (i) tumor located in the dome, anterior wall, other sites of the bladder as well as urachus (the third edition of International Statistical  Classification of Disease for Oncology (ICD-O-3), primary site code C.67.0-C67.9); (ii) diagnosed from the January 1, 1975 to December 31, 2016; (iii) pathologically confirmed tumor; (iv) pathology encoded with adenocarcinoma (mucinous cells, signet ring cells, papillary cells, clear cells, mixed subtypes, intestinal type, and ordinary adenocarcinoma [not specified]). Patients without complete information, death within 30 days after follow-up, and not first tumor were excluded from this study.
Demographics (gender, race, and age at diagnosis), as well as clinical parameters (tumor grade, SEER stage, histologic type, tumor-related surgical methods, chemotherapy, and radiotherapy) of the patients with the diagnosis of urachal and BACs were extracted via the option of "case listing." For analysis purposes, eligible patients were classified into adenocarcinomas arising from urachus (UAC), the dome (D-BAC), the anterior wall (A-BAC) of the bladder, and other sites of the BAC (O-BAC). For D-BAC and A-BAC, they may include cancers originating from urachus or bladder. Complete resection of urachus and navel, as well as partial cystectomy, are currently considered as the standard surgical methods for UAC. However, given that no detailed information regarding the total urachus and navel resection is provided in the database of SEER, we speculated that this surgical approach was classified into partial cystectomy.

| Statistical analysis
Data exported from the latest SEER*Stat 8.3.6 software version were stored in Microsoft Excel 2016 (Microsoft Corporation). Kaplan-Meier survival curves were utilized for both overall survival (OS) and disease-specific survival (DSS). In our study, survival represents the date from cancer diagnosis to death. At the time of the last follow-up, patients presumed alive were censored. Hazards ratios (HR) for the OS and DSS were calculated using the Cox Proportional Hazards Regression Model. Statistical analysis was conducted using the SPSS 25.0 software (SPSS Inc.). Statistical significance was defined as p values less than 0.05 in a twotailed test.

| Baseline characteristics
A higher proportion of males could be observed in each primary location subset. The percentages of males were comparable between A-BAC (62.9%) and O-BAC (64.9%) groups, which were slightly higher than those in UAC (53.2%) and D-BAC (54.2%) groups. UAC or D-BAC patients were significantly younger than A-BAC or O-BAC patients, with

| Survival analysis of all eligible patients
To analyze the clinicopathological characteristics, particularly the impact of tumor sites of adenocarcinoma on survival, univariable and multivariable analyses were performed for all enrolled patients. The age ≤ 60, low tumor grade, mucinous adenocarcinoma, and early SEER stage have been demonstrated independent predictors of good prognosis for OS and DSS in both univariable and multivariable analyses ( Table 2). Statistically significant variables for both 5-year OS and DSS rate calculated by Kaplan-Meier analysis are shown in Figure 1. UAC generally had better survival than BAC (Figure 1g,h). Notably, as far as the impact of primary tumor sites on OS, UAC conferred the best prognosis, then followed by D-BAC  Figure 1). After excluding UAC and D-BAC, A-BAC has a similar prognosis as O-BAC in both the univariable and multivariable analyses (Table S1).

| Analyses in non-metastatic patients
To further understand the impact of treatment on survival of non-metastatic UAC and BAC, we extracted the data from localized and regional diseases and divided the eligible patients into UAC, D-BAC, and BAC (containing A-BAC and O-BAC due to the relatively fewer cases of A-BAC and its similar characteristics to O-BAC). The indicators for OS and DSS were analyzed in the univariable and multivariable Cox regression analyses. The grade and SEER stage were also demonstrated as independent predictors of prognosis for some OS and DSS. Importantly, partial cystectomy seems to be shown the optimal treatment of surgery in patients with UAC or D-BAC, both radical cystectomy and nonstandard surgery approach (other approaches of non-radical cystectomy or absence of cancer-related surgery) would significantly increase the risk of death. Interestingly, although partial cystectomy illustrates numerically benefit for D-BAC, the advantage is not as remarkable as that for UAC. Otherwise, the death risk of radical cystectomy is relatively lower in D-BAC than in UAC (Figure 2.) No groups of nonmetastatic patients were found to benefit from chemotherapy or radiotherapy (Table 3).

| Analyses in metastatic patients
As for the impact of treatment on metastatic diseases, palliative surgery seemed to decrease the risk of death in patients with UAC and BAC in both the univariable and multivariable analyses. Moreover, patients with metastatic A-BAC or O-BAC could significantly benefit from systemic chemotherapy, but radiotherapy did not take any survival benefits to metastatic patients (Table 4). been published. 9,17 Due to its rarity, treatment modalities and prognosis of UrC remain unclear, and prospective trials are still lacking. In accordance with the conclusion of previous studies on UAC, our study demonstrates that the age ≤ 60, low tumor grade, mucinous adenocarcinoma, early tumor stage, and primary tumor site are independent predictors of good prognosis and are also helpful to distinguish UAC from the BAC. 14,16,18,19 In addition, we also demonstrate that UAC is a fundamentally different disease from BAC and generally has better survival than BAC. The definition of UrC generally includes carcinomas from the urachus, the dome, and the anterior wall of the bladder. However, a recent study performing the SEER database has found that UAC and BAC seemed to have discrepant outcomes. In these studies, neoplasms located in urachus or/ and dome were commonly classified as UAC and showed a relatively better outcome than those in other sites of bladder 16 ; otherwise, the A-BAC was not specially classified. In the current study, we separated UAC, D-BAC, A-BAC from all eligible patients and compared the clinicopathologic characteristics and survival of different groups with BAC. This study, for the first time, from the perspective of baseline characteristics, such as gender, age, race, and grading, illustrates that A-BAC has similar characters to O-BAC, while D-BAC is closer to UAC. Moreover, A-BAC and O-BAC have similar survival, while D-BAC has a closer prognosis to UAC. Therefore, we agree that it is reasonable to classify UAC and D-BAC as UrC, while it should be debated to categorize A-BAC into UrC.

| DISCUSSION
Currently, no standard treatment guideline is available for urachal malignant neoplasms. Treatment may be different for localized or metastatic UAC. For localized patients, the backbone therapy is surgery. It is very important for clinicians to differentiate UAC from BAC, as they may require different therapeutic strategies. The approach of surgery for UAC is partial cystectomy with complete resection of the mid-umbilical ligament to the umbilicus. 6 In contrast, BAC is commonly referred to as radical cystectomy for bladder urothelial carcinoma. In the current study, partial cystectomy is confirmed as the optimal treatment of surgery in patients