Introduction
We aimed to determine if there is significant variation in early recurrence after transurethral resection (TURBT) surgery between sites taking part in the RESECT study (NCT05154084) after accounting for tumor characteristics.
Methods
We conducted an international, multicenter, observational study. A mixed-effects logistic regression model with tumor size, tumor number, tumor grade, tumor stage as fixed effects, and site as a random effect was fitted. Cases with first, presumed non-muscle-invasive bladder cancer (NMIBC) undergoing TURBT were included. Cases were excluded if first check followup had not been completed. Sites were excluded if they did not have at least 10 cases with first check followup. Local and/or national approvals or ethical exemptions were obtained prior to commencing the study at participating sites.
Results
After exclusions, 186 sites (U.K: 80; Europe: 59; North America: 18; Asia: 17; Africa 7; South America: 3; Oceania: 2) contributing a total 4597 cases (average 25 cases) were included. Median recurrence rate per site was 12% (IQR 0–22) for low-grade tumors and 27% (IQR 13–42) for high-grade tumors (Figure 1). After controlling for tumor size, number, stage, and grade (all significantly and independently associated with early recurrence) (Table 1), there was significant residual variation attributable to site (p<0.0001, intra-class correlation, 0.1). Adjustment for sites improved the regression model from an area under the receiver operating characteristic curve of 0.66 to 0.74. Initial analysis of surgical and perioperative practice showed wide variation; a mean of 75% (IQR 66–92) of cases per site had detrusor muscle resection and 42% (IQR 17–58) had use of single instillation of intravesical chemotherapy. Other differences in operative and perioperative practice were identified through surveys.
Abstract 25. Figure 1.

Median recurrence rate per site.
Abstract 25. Table 1.
Mixed effects logistic regression for early recurrence after first TURBT for NMIBC
| Tumor feature | OR | p | 95% CI |
|---|---|---|---|
|
| |||
| Diameter(cm) | |||
| <1 | 1.00 | ||
| 1–3 | 1.63 | <0.001 | 1.29–2.08 |
| 3–7 | 2.55 | <0.001 | 1.94–3.33 |
| >7 | 2.42 | 0.003 | 1.34–4.38 |
|
| |||
| No. | |||
| 1 | 1.00 | ||
| 2–3 | 1.49 | <0.001 | 1.24–1.81 |
| >3 | 1.92 | <0.001 | 1.56–2.36 |
|
| |||
| Stage | |||
| Ta | 1.00 | ||
| Tis | 2.00 | 0.015 | 1.15–3.50 |
| T1 | 1.90 | <0.001 | 1.57–2.30 |
|
| |||
| Grade | |||
| Low | 1.00 | ||
| High | 1.44 | <0.001 | 1.19–1.73 |
|
| |||
| Constant | 0.07 | <0.001 | 0.06–0.09 |
|
| |||
| Random intercept – site | 0.36 | <0.001 | 0.23–0.57 |
Conclusions
There is significant variation in the early recurrence rate of NMIBC after TURBT surgery between sites that could not be explained by currently understood tumor features. We have identified differences in surgical technique and perioperative practice that may impact this, and further investigation is warranted to understand how these factors impact recurrence rates.
Footnotes
Funding: The RESECT study is supported by unrestricted grants from Karl Storz, Photocure, and Medac Pharma.








