Absence of detrusor muscle in TUR-BT specimen – can we predict who is at highest risk?

The current study is based on a retrospective dataset of consecutive TUR-BT between 03/2009 and 02/2021 at our tertiary referral center. Data included baseline demographic, clinical and pathological variables. Clinical and surgery-specific variables such as “tumor configuration”, “tumor size” or “location” were obtained from the respective surgical report. Demographic variables such as “age”, “sex”, “comorbidities” or “previous TUR-BTs” were collected from the medical record.

Oncological follow-up was performed by either re-TUR-BTs or cystoscopies at our institution.

Pathological assessment was performed according to the WHO classification from 2004/2016 and 1973. Information regarding the inclusion of DM was retrieved from the histopathological report.

The primary objective of the study is to investigate predictive factors for the presence of DM. We included tumor specific variables defining the complexity of the operation including tumor size (< / ≥ 3 cm), tumor location (bladder-wall, bladder-roof, posterior bladder-wall and proximity to ureteral orifice), the number of previous TUR-BTs, tumor configuration (papillary/flat vs. solid), multiplicity of tumors and the presence bladder diverticula/trabecula. Furthermore, we evaluated parameters which can be influenced including the use of relaxation, the experience of the surgeon (resident in training vs. board-certified urologist), the duration of the surgery as well as the timing of surgery (morning vs. afternoon). At last, we included patient-related variables (age, BMI and gender). For the primary objective, subgroup analyses were performed for the following groups: all patients, primary diagnosis only, high-grade urothelial cell carcinoma only, patients with NMIBC only, re-resection within 6 weeks, and patients with stroma-invasive tumors (pT1) only.

The secondary objective was to define prognostic parameters (including the presence of DM) for the oncological outcome were assessed. Only patients with the primary diagnosis of BC were included in this analysis. Potential prognostic parameters included were sex, smoking history, tumor multiplicity, tumor size, tumor configuration, the use of PDD (photodynamic diagnostis), the presence of bladder diverticula/trabecula, stroma-invasive tumor stage (pT1) and tumor location (bladder wall, bladder roof, posterior bladder wall and proximity to ureteral orifice). Recurrence-free survival was employed as endpoint and was defined as histopathological evidence of a low-grade and/or high-grade BC. For the secondary objective, we analyzed all patients, patients with NMIBC only, patients with stroma-invasive tumors (pT1) only and patients with high-grade tumors only.

Screening identified 3273 TUR-BTs. Overall, we included 2058 TUR-BTs/1472 patients (primary objective) and 472 patients (secondary objective) into the final analyses (Fig. 1). 198 patients were excluded due to missing information about DM and 757 patients were excluded due to no visible tumor or carcinoma-in-situ suspicion only. DM was present in 67.6% (1391/2058) and absent in 32.4% cases (667/2058). The median age of the total cohort was 72 years (IQR: 63–79) and 78.7% (n = 1158/1472) and 21.3% (314/1472) were male and female, respectively. Median follow-up time was 56 weeks (IQR 11–149 weeks). TUR-BTs were performed by 64 urologists. Patients who underwent obturatorius relaxation during their surgery were more likely to have DM present (73.4% vs. 62.7%, p < 0.001). TUR-BTs with DM lasted significantly longer (44 vs. 36 min, p < 0.001). Regarding tumor specific variables, multiple vs. unifocal tumors (70.7% vs 65.0%; p = 0.006) and high-grade vs. low-grade tumors (75.2% vs. 65.9%; p < 0.001) had a higher likelihood of DM in the histological specimen. Also, we have found DM more frequently in larger tumors and with increasing tumor depth. Interestingly, there was no difference between junior and senior surgeons regarding the presence of DM (44.8% vs. 47.5%, p = 0.258). In 63 TUR-BTs there was no malignancy (pT0) in the histological specimen. However, these were all re-resections. Regarding previous intravesical therapies we did not find any significant difference in the presence of DM for BCG (p = 0.054) and Mitomycin C (p = 0.726). Further key patient characteristics are displayed in Table 1 and Fig. 2.

The primary endpoint of the study was to establish predictive factors for the absence/presence of DM. Several patient- and tumor-specific parameters, which either characterize the degree of difficulty of the TUR-BT or can be influenced by the urologist were included. Multivariate analysis for the total cohort revealed age (OR 0.98; 95%CI 0.97–0.99; p = 0.014, B -0,017), surgery duration (minutes, continuous; OR 0.99; 95%CI 0.98–0.99; p = 0.001, B -0.012) and papillary vs. solid tumor configuration (OR 1.99; 95%CI 1.22–3.27; p = 0.006) as independent predictors for the absence of DM (Table 2). Similar results were found in patients with first diagnosis only (duration: OR 0.98; 95%CI 0.97–0.99; p = 0.002, B -0.016; configuration: OR 1.91; 95%CI 1.02–3.61; p = 0.043). In line with the previous results, we found increasing duration of surgery (OR 0.99; 95%CI 0.97–0.99; p = 0.002, B -0.012) and tumor configuration (OR 2.23, 95%CI 1.18–4.20, p = 0.013) to be predictors for finding DM in high-grade urothelial cell carcinoma only. For re-resections we found location (bladder-roof) and number of previous TUR-BTs to influence the absence of DM independently (OR 4.60; 95%CI 1.50–24.70; p = 0.012 and OR 4.99; 95%CI 1.28–19.44); p = 0.021, respectively). In the subgroups NMIBC tumors only and stroma-invasive (pT1) tumors only we did not find independent predictors for DM.

We next investigated, if the presence of DM (Supp. Figure 1) next to other variables has a major and independent impact on the oncological prognosis in the different clinical subgroups.

For the subgroup of high-grade tumors, RFS was significantly longer when DM was present (p = 0.042). Median RFS for patients with DM was 219 weeks and for patients without DM 38 weeks. The presence of DM did not lead to a difference in RFS in the total cohort, in the non-muscle invasive BC group and the subcohort of stroma-invasive BC (Fig. 3). In the multivariate analysis for RFS absence of DM was an independent prognostic factor in patients with high-grade tumors (HR 1.96; 95%CI 1.02–3.79; p = 0.045). Other variables such as sex, tumor location and bladder diverticula did not impact RFS (Table 3). In sum, the absence of DM significantly decreases RFS in a very delicate group of patients.

In the current cohort DM was found in 67.7% of the cases, which corresponds with similar studies, reporting a DM positivity rate of 64% [15], 69.1% [23], 74.2% [17], and 90.4% [22]. In the univariate analysis, the likelihood of DM was significantly higher for larger tumors, high-grade tumors and multiple tumors. Presence of DM also increased with invasion depth. The current results confirm the findings made by previous groups [3, 8, 9, 14]. Shoshany et al. included 332 patients undergoing TUR-BT and found larger tumor size, multifocal BC, higher tumor grade, and solid tumors associated with a higher rate of DM [3]. In a cohort of 356 patients Mariappan et al. [8] demonstrated tumor size and high-grade tumors to be predictive for DM. In a larger cohort, the same group reported increased hospital case volume next to multiple tumors correlated with DM [14].

Interestingly, several studies showed that urologists have a good accuracy in predicting stage and grade of BC during cystoscopy [24, 25]. Therefore, urologists might be more ambitious to resect DM if they suspect invasion of the lamina propria or muscle-invasive disease. On the other side, if a low-risk NMIBC is probable, the surgeon might weigh the prevention of a bladder perforation higher compared to capturing DM. Furthermore, in solid tumors the surgeons may act more confident, as a perforation is less likely. Interestingly, our study revealed no difference between “junior” and “senior” surgeons, which is on the contrary to results of Mariappan et al. [5, 8] and Roupret et al. [20]. On the other side, several studies did not show a significant impact of the skill level on the likelihood of finding DM [3, 22]. This seeming discrepancy might be due to different endourological training programs and supervision standards. Importantly, patients with DM present underwent intraoperative relaxation more frequently. This is of clinical interest, as urologists might be more confident to perform deep resections in vulnerable regions like the lateral bladder walls. In multivariate analyses, the number of previous TUR-BTs and tumor location at the bladder roof were independent predictors for an inferior rate of DM presence in patients undergoing re-resection. To our knowledge, this is the first time that a study analyzed the DM status in re-resections. Our finding can easily be explained by the higher risk of bladder perforations in these cases, and the consequences in the case of intraperitoneal perforation. Consequently, cases with such a high-risk situation should be preserved for urologists with excellent endourological skills. Of note, longer duration of the TUR-BT was independently associated with DM in the complete cohort, patients with first diagnosis and high-grade BC. To our knowledge, this is the first study investigating the impact of operation time on the quality of the TUR-BT and therefore is a key finding of our study. For patients undergoing radical cystectomy, the duration of time also seems to have an impact on outcome parameters [26]. However, it is evident that there is no linear relation between time and outcome. We propose that a TUR-BT needs sufficient time and should be performed with the same diligence as’major’ urological operation like cystectomy. In a time of increasing economic pressure on healthcare providers, our study provides important arguments against a’faster is better’ paradigm, as low-quality TUR-BT will lead to more recurrences leading to higher long-term costs. Taken together, tumor configuration and location, but also the diligence in the sense of time spent for the operation independently define the risk of DM absence in patients undergoing TUR-BT.

The secondary objective of this study was the investigation of DM status on the prognosis. The current study was able to show that absence of DM was an independent prognosticator for reduced RFS in high-grade tumors. This association was not seen for the complete cohort, and the subgroups of NMIBC and pT1, although a non-significant trend was seen. Several other studies positively correlated DM in the TUR-BT specimen with improved RFS [3, 5, 8, 20], progression [9], and lower rates of residual disease [8, 14]. Of note, the other studies investigating recurrence included smaller cohorts and significantly less parameters. The rates of RFS in high-grade tumors in the current study seem long whenever DM is present. However, the results are well comparable to previous literature [3, 22, 27]. In line with our results, a retrospective analysis did not find a correlation between DM status and recurrence in low-grade pTa tumors [28]. In contrast to our study two other investigations demonstrated DM to be predictive of recurrence in patients with pT1 BC [3, 20]. This discrepancy might be explained by differences in the cohorts and follow-up time. Taken together, there is broad evidence that DM should be seen as a key quality indicator for high-grade BC.

To improve the quality of TURB-BTs it seems to be crucial to develop strategies that consistently lead to a higher frequency of DM present. Other approaches involving the introduction of scorecards/checklists also lead to a higher proportion of DM present and therefore a higher-quality TUR-BT [15‐17]. The type of resection method used has also been shown to improve DM rates. Especially en-bloc resection and bipolar resections methods seem to yield better DM resection rates [18, 29].

The retrospective design is among the main limitations of this study and ideally demands prospective validation. Furthermore, inconsistencies of surveillance might mitigate the analyses regarding tumor recurrences. Also, we did not evaluate the occurrence of intraoperative/postoperative complications and therefore we cannot say whether the presence of DM might lead to a higher complication rate in specific tumor locations. Last, due to the relatively short follow-up time, we were not able to show a relevant number of progressions and therefore assess the impact of DM onto progression-free survival.