Presentation Authors: Jay Fuletra*, Matthew Kaag, Andrew Watts, Erik Lehman, Jay Raman, Suzanne Merrill, Hershey, PA
Introduction: Staging imaging characteristics in bladder cancer (BC), such as hydronephrosis, are predictive of â‰¥pT3 disease at time of radical cystectomy (RC) and can be useful adjuncts in preoperative medical decision making. However, the predictive capacity of other imaging findings, such as perivesical stranding (PS), remains unclear. Thus, we investigated whether PS prior to transurethral resection of bladder tumor (TURBT) was associated with â‰¥pT3 BC following RC in patients who did not receive neoadjuvant chemotherapy (NAC).
Methods: We identified 433 patients with BC who underwent RC from 2003 to 2018 of which 128 did not receive NAC. Evidence of PS on pre-TURBT imaging was determined by radiologist review and a stranding grading system (minimal/moderate/severe) was created. Wilcoxon rank sum and two-sample t-tests identified associated factors with PS. Multivariable logistic regressions were used to evaluate the capacity of PS as a reliable preoperative predictor for â‰¥pT3 BC.
Results: Of the 128 patients who did not receive NAC, 48 (37.5%) had pT3 and 12 (9.4%) had pT4 BC at RC. Preoperative imaging consisted of CT in 125 (97.7%) patients and MRI in 3 (2.3%). PS and hydronephrosis on imaging was identified in 19 (14.8%) and 45 (35.2%) patients, respectively. On univariate analysis, PS on imaging was not significantly associated with imaging type (p =0.383), greater BMI (p =0.176), or higher pathologic T stage (p=0.243). Additionally, grading of stranding did not improve the association with pathologic T stage (p=0.208). However, hydronephrosis on imaging was seen more frequently in higher pathologic T stage (p=0.034; 22 (45.8%) in pT3 and 7 (58.3%) in pT4). Multivariable analysis noted that PS on imaging was not predictive of â‰¥pT3 BC following RC (OR 1.5, 95% CI [0.5, 4.1], p=0.457). Stratification by grade of stranding did not improve the predictive capacity of PS (p=0.667 for severe vs. no stranding). Hydronephrosis, on the other hand, remained a predictive imaging indicator of â‰¥pT3 BC following RC (OR 3.3, 95% CI [1.5, 7.2], p=0.003).
Conclusions: Perivesical stranding on preoperative staging imaging failed to be a reliable predictor of â‰¥pT3 stage in patients with BC. However, hydronephrosis continued to act as a strong imaging indicator of higher pathological stage. This data should give pause in using perivesical stranding identified on preoperative imaging to guide clinical decision making in BC patients until further investigations can be explored in larger cohorts.