Abstract
Background: The aim of this study was to evaluate the prognosis of patients with T1N0M0 urothelial carcinoma with squamous differentiation (UCSD) or glandular differentiation (UCGD) because it has not been determined whether these variant histologies behave more aggressively than pure urothelial carcinoma (PUC). Patients and Methods: Ninety-nine patients diagnosed with pT1N0M0 bladder cancer and treated conservatively with transurethral resection of bladder tumor at Kanazawa University Hospital between 2007 and 2019 were included in this study. The overall survival, cancer-specific survival (CSS), and recurrence-free survival of the variant histology and PUC groups were evaluated and compared. Results: The variant histology group had significantly lower overall survival (p=0.006) and CSS (p=0.0095) than the PUC group did. Variant histology was found to be an independent prognostic factor in univariate and multivariate analyses for overall survival and CSS. On the other hand, no significant difference in progression-free survival was observed between the two groups (p=0.439). However, the variant histology group had significantly lower overall survival (p=0.004) and CSS (p=0.004) after progression. Conclusion: The prognosis for patients with pT1 bladder cancer with UCSD or UCGD treated conservatively with transurethral resection of bladder tumor was poor. Considering the worse prognosis of these patients after stage progression, early radical cystectomy could be recommended.
- Urothelial carcinoma
- squamous differentiation
- glandular differentiation
- non-muscle-invasive bladder cancer
While most histological types of invasive bladder cancer (BC) are pure urothelial carcinoma (PUC), about 25% are of variant histology (VH). (1). The most frequent VHs in BC are urothelial carcinoma with squamous differentiation (UCSD) and glandular differentiation (UCGD), followed by sarcomatoid differentiation, micropapillary differentiation, and neuroendocrine differentiation (2, 3). It has not been determined whether VH behaves more aggressively than PUC. Although it has been found that VH urothelial carcinoma does not affect the prognosis of patients (4, 5), several reports have indicated that the presence of VH is a poor prognostic factor. Jozwicki et al. reported that a VH occupancy of more than 20% or the presence of more than two types of VH was associated with worse prognosis (6). Mitra et al. found that UCSD and UCGD have a significantly more advanced clinical stage upon diagnosis compared with PUC and are associated with poor prognosis (7). For chemotherapy, there are both reports of similar response to PUC and reports showing worse response (3, 8, 9). Although many studies on the prognosis of muscle-invasive BC with VH have been conducted, only a few have focused on pathological T1 (pT1) cases of invasion of the lamina propria. In this study, we focused on pT1 cases with UCSD and UCGD, the most frequent types of VH, to investigate their prognosis and compare them with PUC.
Patients and Methods
Among patients with BC receiving treatment at Kanazawa University Hospital from 2007 to 2019, those who were initially diagnosed with T1N0M0 and treated conservatively with transurethral resection of bladder tumor (TURBT) were selected. Patients who underwent cystectomy or radiation therapy were excluded. Patients with pathologically diagnosed PUC, UCSD, or UCGD were included in this study. We retrospectively investigated the overall (OS), cancer-specific (CSS), and progression-free (PFS) survival. The OS was measured from the diagnosis of BC until death or at last follow-up. The CSS was measured from the diagnosis of BC until death due to BC or last follow-up. The PFS was measured from the diagnosis of BC until clinical stage progression or last follow-up. Stage progression was defined as muscular invasion and remote and lymph node metastasis during the follow-up period. Follow-up was terminated on March 1, 2022. We also analyzed the factors associated with OS, CSS, and PFS via univariate and multivariate analyses. The clinical stage was determined based on the eighth edition of the Union for International Cancer Control Tumor, Node, Metastasis classification published in 2017 (10).
The OS, CSS, and PFS were estimated using the Kaplan-Meier method, and the differences were compared using log-rank tests. We evaluated the predictive impact of several potential factors on the OS, CSS, and PFS using the Cox proportional hazards model. The hazard ratio (HR) and 95% confidence intervals (CI) were calculated. Comparisons between different groups were performed using the chi-squared test and the Mann-Whitney U-test. Statistical analyses were performed using the commercially available software Prism 8 (GraphPad, San Diego, CA, USA) and SPSS ver. 25.0 (IBM, Armonk, NY, USA), with p-values of less than 0.05 indicating statistical significance. This study was approved by the Institutional Review Board of Kanazawa University Hospital (2022-299).
Results
Table I summarizes the characteristics of the 99 patients. Seventy-six patients with BC were diagnosed with PUC and 23 with UCSD or UCGD. Of the 23 patients in the VH group, most (n=16) had UCGD. No significant differences in age, sex, pathological grade, Bacille Calmette-Guerin (BCG) treatment history, or clinical stage progression rate were observed between the two groups. The median follow-up duration was 64.8 months for the PUC group and 73.6 months for the VH group. Stage progression during the follow-up period was observed in 16 patients in the PUC group and seven in the VH group. No significant difference in treatment after progression was observed between the two groups.
Immunoreactivity for matrix metalloproteases (MMPs). Immunohistochemical reactivity: (−) negative, (+) positive, (++) strongly positive.
The OS (HR=4.459, 95% CI=1.050-18.93; p=0.006) and CSS (HR=5.573, 95% CI=0.853-36.40; p=0.0095) were significantly lower in the VH group than in the PUC group (Figure 1A and B). However, no significant difference in the PFS was observed between the two groups (HR=1.42, 95% CI=0.541-3.709; p=0.439 (Figure 1C).
Kaplan-Meier curve for overall (A), cancer-specific (B) and progression-free (C) survival of patients according to type of carcinoma, pure urothelial carcinoma (PUC) or variant histology (VH). CI: Confidence interval; HR: hazard ratio.
The univariate analysis identified VH (HR=4.970, 95% CI=1.397-17.676; p=0.013) as significant prognostic factors for OS, although age, sex, BCG treatment history, and pathological grade were not in this cohort (Table II). The multivariate analysis showed that VH (HR=4.536, 95% CI=1.067-19.284; p=0.041) was independent factor affecting the OS. Furthermore, the univariate analysis identified VH (HR=6.975, 95% CI=1.269-38.329; p=0.025) as significant prognostic factors for CSS (Table II). The multivariate analysis indicated that only VH was an independent predictor for prognosis (HR=6.007, 95% CI=1.021-35.328; p=0.047) for CSS. For the PFS, only pathological grade 3 (HR=2.472, 95% CI=1.013-6.036; p=0.047) was a predictor in the univariate analysis (Table II) as well as being an independent factor in the multivariate analysis (HR=2.958, 95%CI=1.120-7.230; p=0.017).
Univariate and multivariate analyses of the prognostic factors for overall (OS), cancer-specific (CSS) and progression-free (PFS) survival.
The OS after clinical progression was significantly lower in the VH group (HR=5.434, 95% CI=0.802-36.79; p=0.004) than in the PUC group (Figure 2). Moreover, the CSS after progression was significantly lower in the VH group (HR=6.17, 95% CI=0.889-42.88; p=0.004) than in the PUC group (Figure 2).
Kaplan-Meier curve for overall (A) and cancer-specific (B) survival of patients after clinical stage progression of their disease according to type of carcinoma, pure urothelial carcinoma (PUC) or variant histology (VH). CI: Confidence interval; HR: hazard ratio.
Discussion
In the present study, we found that patients with pT1 BC diagnosed with UCSD or UCGD who received conservative treatment with TURBT had a significantly lower OS and CSS compared with patients with PUC. VH was also an independent prognostic factor for OS and CSS in both the univariate and multivariate analyses. Antunes et al. reported that UCSD was an independent and significant prognostic factor in CSS of patients undergoing radical cystectomy (11). Although several studies have reported that VH is a prognostic factor in muscle-invasive BC, we found that VH is also a prognostic factor in OS and CSS in pT1 cases. However, no significant difference in the PFS was observed between the two groups. Xu et al. found that in non-muscle-invasive BC, UCSD and UCGD had an impact on recurrence rates, but no significant difference in the PFS was observed when compared with PUC (12). Zhao et al. studied the prognosis of pT1 cases diagnosed with UCGD and reported that UCGD is a significant factor influencing PFS (13). Moreover, they also reported that patients with UCGD had a poorer 5-year OS than those with PUC, while there was no significant difference in CSS between two groups. However, some differences can be noted between that study and the present study which may have resulted in different results. Firstly, their study included only patients with high-grade tumors. Secondly, their patients were not treated with BCG.
In the present study, BCG treatment was not an independent factor influencing OS, CSS, or PFS. Although BCG therapy is recommended for pT1 cases (14), the efficacy of BCG treatment for UCSD and UCGD is unknown. Yorozuya et al. reported a study of BCG therapy for patients with non-muscle-invasive BC diagnosed with UCSD or UCGD and found a significantly longer recurrence-free survival, PFS, and CSS in such patients compared with patients who did not receive additional therapy (15). Gofrit et al. reported that patients with VH BC had a significantly worse prognosis than patients with PUC treated with BCG (16). We suspect that this difference in results may be due to the fact that BCG maintenance therapy was used at their facility, whereas maintenance therapy is rarely used at ours.
The OS and CSS after clinical stage progression were significantly lower in the VH group in our study. On the other hand, no significant difference in the PFS was observed between the two groups. These results suggest that the VH group may have a worse prognosis if the disease progresses. The National Comprehensive Cancer Network guidelines list the presence of VH as one of the highest risk factors among patients with pT1 tumors (14). These patients are recommended to undergo early cystectomy because of the high risk for progression to a more advanced stage (17). However, the guidelines list micropapillary, plasmacytoid, and nested variants as VHs that fall under the recommendation for early radical cystectomy, but there is no mention of UCSD or UCGD. Since the prognosis after stage progression in UCSD and UCGD is significantly poorer than that in PUC, patients with these VHs may also require early cystectomy.
However, this study has several limitations. Firstly, it was retrospective in design, and the evaluation of the treatment effects was left to the individual physicians, which may have resulted in bias. Secondly, although second TURBT is recommended for patients with T1 tumors, it was not performed for most of the patients in this study. Thirdly, treatment selection after stage progression was at the physician’s discretion. Finally, patients treated with BCG were included, especially those who were unable to receive a sufficient number of BCG intravesical injections (six or more times) due to severe symptoms, such as fever and pain during urination. Moreover, maintenance therapy was not administered to most patients. These limitations should be considered when interpreting the results. Furthermore, the use of 5-aminolevulinic acid has recently been incorporated into the photodynamic diagnosis, which may further improve the prognosis of non-muscle-invasive BC, and future developments are expected (18).
Conclusion
The OS and CSS of patients diagnosed with T1N0M0 BC and treated conservatively with TURBT were significantly lower in those with UCSD or UCGD compared with those with PUC. VH was subsequently found to be an independent prognostic factor in univariate and multivariate analyses. Moreover, although no significant difference in the PFS was observed, the OS and CSS after BC progression were significantly lower for the VH group. Since UCSD and UCGD have a poor prognosis once they undergo clinical stage progression, it may be desirable to perform early cystectomy for patients with these VHs of BC, even if they have pT1 tumors.
Footnotes
Authors’ Contributions
Conceptualization: HI and RN; methodology: HI and RN; formal analysis: RN, TM, SK and HY; writing-review and editing: RN, HI, KS, KI, YK and AM. All Authors read and agreed to the published version of the article.
Conflicts of Interest
All Authors declare that there are no potential conflicts of interest relevant to this article.
- Received April 21, 2022.
- Revision received June 5, 2022.
- Accepted June 8, 2022.
- Copyright © 2022 The Author(s). Published by the International Institute of Anticancer Research.
This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY-NC-ND) 4.0 international license (https://creativecommons.org/licenses/by-nc-nd/4.0).
