Clinicopathological features of malignant urachal tumor: A hospital‐based cancer registry data in Japan
Abstract
Objectives
To identify the clinicopathological features of malignant urachal tumor in Japan, and analyze the 5‐year overall survival of malignant urachal tumor patients.
Methods
We used the hospital‐based cancer registry data to extract malignant urachal tumor cases that were diagnosed in 2008–2009 and 2012–2015, histologically confirmed, and received the first course of treatment. We analyzed the 5‐year overall survival using the 2008–2009 cohort’s data.
Results
We identified 456 patients, and malignant urachal tumor accounted for 0.4% of all malignant bladder cancers. The median age was 61 years (range 2–97), and 66% were men. The most common histology was adenocarcinoma (80%), followed by urothelial carcinoma (11%) and squamous cell carcinoma (3%). The proportions of patients were: 19% Ta/Tis/T1N0M0, 55% T2–3N0M0, 13% T4/N+ and 13% M+. Regarding the initial treatment, the proportions of surgery alone were 79% and 33% in cases of T3 or less N0M0 and T4/N+, respectively. The proportion of combination therapy including surgery and chemotherapy were 13% and 44% in T2–3N0M0 and T4/N+, respectively. Radiation therapy was not common at any stage. In the 2008–2009 cohort, the 5‐year overall survival rate in Ta/Tis/T1N0M0, T2–3N0M0, T4/N+ and M+ were 60%, 64%, 63% and 12%, respectively.
Conclusions
Malignant urachal tumors are quite rare in Japan, and most of those without metastasis are likely to be treated by surgery alone, even at advanced stages. A standard of care must be established for malignant urachal tumor patients at advanced stages or with metastasis, as the prognosis of these patients can be poor.
Abbreviations & Acronyms
-
- AC
-
- adenocarcinoma
-
- DCCH
-
- designated cancer care hospital
-
- F/U
-
- follow‐up
-
- HBCR
-
- hospital‐based cancer registry
-
- ICD‐O‐3
-
- International Classification of Disease for Oncology, 3rd edition
-
- MUT
-
- malignant urachal tumor
-
- OS
-
- overall survival
-
- SCC
-
- squamous cell carcinoma
-
- SEER
-
- Surveillance Epidemiology and End Result
-
- TNM
-
- tumor–node–metastasis
-
- TURBT
-
- transurethral resection of bladder tumor
-
- UC
-
- urothelial carcinoma
-
- UICC
-
- Union for International Cancer Control
Introduction
MUT, a rare genitourinary tumor derived from the urachus at the dome of the bladder, accounts for 0.1–0.7% of all malignant bladder cancers.1-3 The most common histology is AC;4-7 other subtypes are rare. Patients with MUTs are usually diagnosed at an advanced stage with extravesical extension and lymph node metastasis, and the prognosis is thus generally poor.1, 5 Surgery is a standard of care for localized MUT,5, 8 but the most appropriate care for advanced cases has not been established, as there are few data from case reports/case series. Until recently, there has been only limited information about the clinicopathological features of MUTs from a large population‐based study.4 Herein, to identify the clinicopathological features of MUT cases in Japan, we retrospectively analyzed the cases of the patients newly diagnosed with MUT in 2008–2009 and 2012–2015 using the HBCR from Japan’s nationwide DCCHs and other core cancer care hospitals.
Methods
Data sources
We used the HBCR data from the nationwide DCCHs and other core cancer care hospitals to identify MUT patients diagnosed in 2008–2009 and 2012–2015. The information of patients who are newly diagnosed as having cancer are routinely collected by these hospitals. As described, the data from the HBCR cover approximately 67% of the patients diagnosed with cancer in Japan.9 Well‐trained cancer registrars at each hospital register the details of diagnosed cancer cases based on standardized criteria. The data include the patients’ demographics, tumor characteristics and the first course of treatment. The first course of treatment is defined as a treatment that is planned for improving the cancer prognosis at an initial diagnosis. As treatment modalities, the present data include surgery, radiation therapy and chemotherapy. “Surgery” includes open, laparoscopic and endoscopic surgery. In the cases of the bladder tumors including MUTs, open or laparoscopic surgery generally included partial or radical cystectomy, and endoscopic surgery included TURBT.
We used the data of the 2008–2009 cohort for an analysis of survival, as the data of the 2012–2015 cohort did not include the survival information. In contrast, the 2008–2009 cohort had the survival information 5 years after the diagnosis. In this cohort, the data were obtained from the hospitals that had a >90% F/U rate for all cancer patients. We cannot calculate the median F/U period of these patients, because the exact F/U period of some patients was not available in this database. In the present study, we used the data from 251 DCCHs for the survival analysis (Table S1). For the analysis of the clinicopathological features of the MUTs, we used the data of both the 2008–2009 and 2012–2015 cohorts. The data of the 2012–2015 cohort were collected from 745 hospitals (427 DCCHs and 318 other core cancer center hospitals).
The staging information was based on the 6th UICC TNM Classification in the 2008–2009 cohort, and the 7th UICC TNM Classification was used in the 2012–2015 cohort. As the differences between these two staging systems have no influence on the staging of the MUTs, we analyzed the clinical stage using the data of both cohorts.
Identification of MUT patients
We identified eligible patients from the data by using following inclusion criteria: (i) newly diagnosed with a malignant bladder tumor located at the urachus (C67.7) in the years 2008–2009 or 2012–2015; (ii) received the first course of treatment at a DCCH or other core cancer care hospital; and (iii) had a histologically confirmed tumor with ICD‐O‐3 histology codes. Cases with histology codes 8000 (two in 2008–2009 and 10 in 2012–2015) were excluded, as we could not specify the histological subtype.
Selection of MUT cases for the clinical stage analysis
For the analysis of the patients’ clinical stages, 80 (21%) patients in the 2012–2015 cohort and nine (12%) patients in the 2008–2009 cohort were excluded due to lack of clinical stage information. We identified a total of 367 patients in the two cohorts. We classified the patients into four groups based on the extension of disease: Ta/Tis/T1N0M0, T2–3N0M0, T4/N+ and M+. Using these stratified groups, we analyzed the pattern of the first course of treatment and the 5‐year OS rate.
All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards.
Statistical analysis
The 5‐year OS rate was analyzed by the Kaplan–Meier method and compared between groups by the log–rank test. All statistical comparisons were two‐sided, and P < 0.05 were considered significant. spss 25.0 for Windows (SPSS, Chicago, IL, USA) was used.
Ethical considerations
The study protocol and data processing were approved by the Tsukuba University Hospital Ethical Board (H29‐267). In rare disease research, there are some cases where the number is <10 patients. However, patients’ privacy should be considered for publication, as it could lead to identifying an individual. Therefore, we reported the proportion of those cases in the text or figures according to recommendations of the Ministry of Health, Labor and Welfare.
Results
As shown in Figure 1, among the 10 268 cases from 246 hospitals (2008–2009) and 97 126 cases from 737 hospitals (2012–2015) that were registered as malignant bladder tumor (C67) cases and included a first course of treatment, 78 (0.7%) and 390 (0.4%) cases were registered as a tumor at the urachus. A total of 456 patients (76 in 2008–2009 and 380 in 2012–2015) with confirmed histological subtypes were identified. The annual numbers of patients newly diagnosed from 61 hospitals in the 2‐year 2008–2009 cohort were 37 and 39, respectively. The annual numbers of patients newly diagnosed from 255 hospitals in the 4‐year 2012–2015 cohort were 81, 98, 94 and 107, respectively.
Table 1 summarizes the characteristics of patients with MUT. The median age of all patients was 61 years (range 2–97 years). More than half (66%) of the patients were men. The most common histology was AC (79.6%), followed by UC (10.7%), SCC (3.1%) and other forms (6.6%). “Other forms” included histological subtypes, such as carcinoma (not otherwise specified), sarcoma and neuroendocrine carcinoma. In terms of the distribution of ages stratified by histological subtypes, most of the patients with AC were aged in their 40s to 60s, and the patients with UC or SCC were generally aged in their 70s (Fig. 2).
| All (%) | |
|---|---|
| No. patients | 456 |
| Age, median (range) (years) | |
| 61 (2–97) | |
| Sex | |
| Male | 301 (66.0) |
| Female | 155 (34.0) |
| Stage | |
| Ta/Tis/T1N0M0 | 69 (15.1) |
| T2–3N0M0 | 201 (44.1) |
| T4/N+ | 48 (10.5) |
| M+ | 49 (10.7) |
| Unknown | 89 (19.5) |
| Histology | |
| AC | 363 (79.6) |
| UC | 49 (10.7) |
| SCC | 14 (3.1) |
| Others | 30 (6.6) |
Figure 3 shows the distribution of sex stratified by histological subtypes. The patients with SCC were more likely to be women (58%), and most of the patients with the other subtypes (including AC and UC) were men. Regarding the distribution of clinical stages as shown in Figure 4, the proportions of patients were as follows: Ta/Tis/T1N0M0 (19%), T2–3N0M0 (55%), T4/N+ (13%) and M+ (13%). Among the patients with AC, the proportions of patients in the corresponding groups were 18%, 54%, 13% and 15%, respectively. In contrast, the proportion of Ta/Tis/T1N0M0 among the patients with UC was higher at 27% (vs 18% of the AC patients). Over half of the patients with these three subtypes of histology were T2–3N0M0. The proportions of T4/N+ or M+ among the patients with AC and UC were 28% and 16%, respectively, whereas the proportion among the SCC patients was higher (33%). In terms of the relationship between the age and clinical stages, the distribution of age at each stage was similar, as shown in Figure S1.
Figure 5 shows the patterns of the first course of treatment. For the initial treatment, the proportion of Ta/Tis/T1N0M0, T2–3N0M0 and T4/N+ patients who received surgery alone were 75%, 80% and 33%, respectively. Although open surgery is commonly carried out for MUTs, just 10% of the Ta/Tis/T1N0M0 patients and 7% of the T2–3N0M0 patients were treated with laparoscopic surgery. The proportion of patients who received combination therapy including surgery and chemotherapy were 13% and 44% in T2–3N0M0 and T4/N+, respectively. Radiation therapy was not common at any stage of MUT. Among 15 (4%) patients who received radiation therapy, the proportions of patients who received radiation therapy alone and combination therapy including radiation therapy and surgery or chemotherapy were 0.5% and 3.5%, respectively. Among the M+ patients, the most common treatment was chemotherapy alone (37%), and 16% of patients were treated with both surgery and chemotherapy.
Figure 6 shows the 5‐year OS rates of 67 patients with MUT stratified by clinical stages. A total of 28 patients died of the disease, and seven patients were censored. The 5‐year OS rate in the groups of Ta/Tis/T1N0M0, T2–3N0M0, T4/N+ and M+ patients were 60%, 64%, 63% and 12%, respectively. In an analysis between the patients who underwent surgery alone and those who were treated with a combination of surgery and chemotherapy among all of the patients in the T2–3N0M0 and T4/N+ groups (n = 45), the 5‐year OS rate in the combination surgery + chemotherapy group (n = 10) was higher (90%) compared with the 56% in the surgery alone group (n = 35), but this difference was not significant (P = 0.074).
Discussion
There have been just two large‐scale population‐based studies of MUTs. Bruins et al.10 reported 152 cases in the Netherlands in 2012, and Mylonas et al.4 reported 420 cases in Greece in 2017. Although the report from Greece is the largest population‐based study of MUTs, the present study provides the second‐largest cohort and the first report from Asia at this time. According to the recent literature, MUTs account for approximately 0.1–0.7% of all bladder cancers.1-3 The proportion (0.4%) of MUTs in the present study is similar to that of the studies of Western populations. In contrast, a systematic review of 157 cases from Japan in 1983 concluded that the proportion of MUTs was relatively higher in Japan (0.55–1.2%) than that in Western countries (0.07–0.70%).11 In addition, Mylonas et al. reported that seven of 16 published studies between 1991 and 2016 were carried out in Eastern Asia.4 Although the proportion of MUTs among all bladder cancers might be more frequent in Asia, further investigations including population‐based studies or multi‐institutional studies are necessary in the future.
We observed that the most common histology was AC (80%), followed by UC (11%) and SCC (3%). In a study using the SEER database, the proportion of histological subtypes of MUT are quite similar to the present findings; that is, 81% in AC, 10% in UC and 3% in SCC.4 The present results regarding the distributions of age and sex among the patients with AC were also similar to those in other reports.4-7 However, the ages of most of our non‐AC patients (including those with UC or SCC) were higher (70s) than those of the patients with AC, and over half (58%) of the patients with SCC were women. Although there is no available report of a higher proportion of female patients in SCC, the age peak of the present patients with non‐AC tends to be higher than that (mean 64.1 years) in another report.12
With regard to the staging system, Bruins et al. used the Sheldon staging system.10 In 2012, Siefker‐Radtke recommended that the TNM staging system not be used for bladder cancer, because MUTs arise outside of the bladder and grow into the bladder secondarily.8 In contrast, the TNM staging system is defined based on the growth of tumors that arise inside of the bladder. Therefore, they recommended the use of the Sheldon or Mayo staging system.1 However, the recent population‐based study of MUTs used the TNM staging system.4 In 2015, Dhillon et al. reported that the TNM staging system might also be useful for MUTs, as the mortality rate is associated with the tumor stage in the Sheldon, Mayo and TNM staging systems.13 Therefore, it seems to be acceptable to use the TNM staging system for the MUTs. In the present study, most of the MUT patients presented with T2–3N0M0 (55%), and the proportion of T4/N+ or M+ was 27%. The proportion of patients with non‐muscle‐invasive MUT was just 19%. The present results of the proportions were similar to those from another study using the TNM staging system: 19%, 45% and 31% presenting at stage I, stage II–III and stage IV, respectively.4
In terms of care, the most common modality used for localized MUTs was surgery alone. The recommended surgical procedure is partial cystectomy with en bloc removal of the umbilical ligament and umbilicus, while the benefit of lymph node dissection is unclear.5, 8 In the present study, the proportions of patients treated by the combination therapy of surgery and chemotherapy were 13% in the T2–3N0M0 patients and 44% in the T4/N+ patients. These patients might have received neoadjuvant or adjuvant chemotherapy. No further information about chemotherapy including neoadjuvant and adjuvant chemotherapy was available from the HBCR data. In the present study, radiation therapy was less likely to be used (4%), and this is similar to another study’s finding (10%).4 In terms of chemotherapy, most M+ patients with distant metastasis receive chemotherapy alone. However, there was no way to know what regimens were used, because the regimen information is not available from the HBCR. As MUTs show histopathological and clinical similarities with colorectal AC, a cytotoxic chemotherapy regimen for colorectal cancer in metastatic or unresectable patients, such as 5‐fluorouracil‐based or 5‐fluorouracil and cisplatin, is likely to be used.8, 14 In contrast, there are some case reports that suggest the usefulness of a combination of gemcitabine and cisplatin based on urothelial carcinoma.15, 16 Recently, based on the genomic similarities between MUTs and colorectal AC, the targeted therapy, such as epidermal growth factor receptor inhibitor, suggests the potential effect.17, 18 As another novel treatment for disseminated MUTs, some authors reported positive results on hyperthermic intraperitoneal chemotherapy for patients with peritoneal metastases.19 However, further studies will be required, as these data derive from a small number of patients.
In our analysis of the patients’ prognoses, the 5‐year OS rates in the Ta/Tis/T1N0M0, T2–3N0M0, T4/N+ and M+ groups were 60%, 64%, 63% and 12%, respectively. The recent study using the TNM staging system obtained 5‐year OS rates at stage I, II, III and IV of 73%, 60%, 58% and 20%, respectively.4 Although the staging system is different, the 5‐year OS rate from the present study is similar. The prognosis of M+ patients is quite poor. Previous studies reported the 5‐year OS rate for metastatic MUT patients was <20%.1, 10 Among the present T2–3N0M0 and T4/N+ patients, despite their low numbers, the 5‐year OS rate of the patients treated by the combination of surgery + chemotherapy tended to be higher than that of the patients treated by surgery alone. There is no similar report about the effectiveness of perioperative chemotherapy.
The appropriate care for advanced‐state MUT patients is still uncertain due to the lack of evidence from prospective studies. To establish a standard of care, a prospective study is necessary in the future. The epidemiological characteristics determined by the present study might be helpful for planning the study protocol, especially in terms of treatment for advanced patients.
The present study had potential limitations due to the availability of the data from the HBCR. First, the staging system is based on the TNM classification for bladder cancer. Second, there was no detailed information about surgery excluding open, laparoscopic and endoscopic surgery. There was no way to know whether patients underwent a partial or radical cystectomy. In terms of chemotherapy, there was no available information about the regimens or the setting of neoadjuvant or adjuvant therapy. Third, the pathology did not undergo a central review; it depended on one or more pathologists at the hospitals where the patients were diagnosed. Finally, the sample size of the cohort for an analysis of survival was small. Further study will be necessary to clarify more accurate survival information of MUTs. Despite the limitations related to the HBCR, we present the first report about MUTs based on the largest database in Asia.
In conclusion, MUTs were quite rare. Although the most common histology was AC, there were some differences among the histological subtypes in the distribution of age, sex and clinical stage. The majority of the MUTs without metastasis (even at advanced stages) were likely to be treated by surgery alone. It is an important future task to establish a standard of care for MUT patients at advanced stages or with metastasis, as our present analyses confirmed that the prognosis of these patients is not satisfactory.
Acknowledgments
We thank the tumor registrars in hospitals for their efforts on collecting data that enabled our research analysis.
Conflict of interest
None declared.
