In our study, the detrusor muscle sampling rate of the TR group (70.7%) was significantly higher than that of the CR group (55.6%); the recurrence rate of the TR group (29.8%) was significantly lower than that of the CR group (48.1%). Both univariate and multivariate logistic regression showed that TR was associated with lower odds of recurrence. Besides, both univariate and multivariate Cox proportional hazards regression also showed that TR was associated with better RFS.
Recurrence rates of NMIBC within 2 years, 5 years, and 10 years after the initial diagnosis were 61.1%, 69.5%, and 74.3%, respectively (14). TURBT is the standard treatment for NMIBC. Current consensus has been made that an adequate TURBT must include detrusor muscle in the specimen, except for Ta low-grade diseases. The lack of detrusor muscle in the specimen is also associated with higher risks of disease understaging, residual tumor, and recurrence (2, 15–17). Besides, Mariappan et al (2010) have reported that unskilled surgeries resulted in lower detrusor muscle sampling and higher recurrence rates (detrusor muscle sampling rates 56.8% vs 72.6%, early recurrence 39.3% vs 24.8%,) (2). Therefore, finding out a method to improve detrusor muscle sampling is important.
Complete resection either by fractioned or en bloc resection has been recommended in EAU guidelines. Detrusor muscle sampling rates for en bloc resection are nearly 96%-100% in past researches (6–8, 10, 18). By contrast, detrusor muscle sampling rates varied between 54%-90% for conventional TURBT (3, 5, 9, 11). As for oncological outcomes, Sureka et al (2014) have reported a lower recurrence rate in patients receiving en bloc resections, compared with patients receiving conventional TURBT (28.6% vs 62.5%) (19). However, two larger RCTs showed no significant difference in terms of recurrence between the two methods (4, 20). Despite potential advantages from en bloc resection, there is no strong evidence proving its superiority to conventional TURBT. Moreover, en bloc technique is skill-intensive and can be expensive if the laser is used for resection.
Richterstetter et al. (2012) have reported an extended TURBT protocol, which consisted of complete resections in fractions of the main tumor, bladder ground, and normal-appearing margins(12). It provides good information about the vertical and horizontal extents of the tumor and has no association with focality of the tumor and the experience of the surgeon. Surprisingly, the recurrence rate of all unifocal primary disease was only 14.4%. Nevertheless, no data regarding the detrusor muscle sampling rate has been reported. The drawbacks of extended TURBT protocol are time-consuming and skill-intensive.
For our two-step TURBT, surgeons only need to resect and retrieve tumors and tumor bases separately. This procedure is easier, more convenient, and more cost-effective than extended TURBT protocol and laser en block methods. The detrusor muscle sampling rate was also higher in the TR group, indicating a better quality of surgery. Patients who received TR were significantly associated with lower recurrence and better RFS, compared with patients who received CR. Despite that the characteristics between the two groups were not similar in terms of DM and initially high-grade cancer, the multivariate Cox proportional hazards regression has confirmed that TR is an independent factor for lower recurrence. The detrusor muscle sampling in our study was not a predictor for outcomes. We speculate the reason to be the majority of Ta disease in our study population and the exclusion of MIBC, suggesting tumor eradication could be obtained without a resection to the detrusor muscle. Furthermore, better outcomes in the TR group might be due to deeper resections of tumor bases that conduced to better tumor eradication. The subgroup analyses of Ta diseases and T1 diseases (Supplementary data) also supported our assumption.
There were some limitations in our study. First, the sample size in the CR group was smaller, and the distribution of patient characteristics was not similar between the two groups. Second, techniques from different surgeons might have led to potential bias. Third, it was not possible to distinguish residual diseases from recurrent diseases. On the other hand, there are several advantages of our study. First, we have a strict follow-up protocol, and all data were collected prospectively. In addition, second TURBT were performed routinely for all patients, and additional mapping biopsies were taken from patients with CIS. These managements reduced the chance of misdiagnosis and improved the detection of recurrence and progression. Second, as more than 80% of recurrences would occur within the first two years (14), we included patients with a follow-up period of more than 2 years. Third, multivariate analysis was used to adjust potential confounders between the TR and CR groups.