Contrast media kinetics in multiparametric magnetic resonance imaging before radical prostatectomy predicts the probability of postoperative incontinence

Abstract

PURPOSE To evaluate the role of preoperative multiparametric magnetic resonance imaging (MRI) as predictor of post-prostatectomy incontinence (PPI). METHODS We analyzed patients who underwent robot-assisted radical prostatectomy for localized prostate cancer at our institution between July 2015 and April 2017. In these patients, we measured the perfusion quality of the pelvic floor with contrast media kinetics in the preoperative MRI of the prostate and compared the levator ani muscle (region of interest) to the surrounding pelvic muscle structures (reference). Prospectively collected questionnaires regarding urinary incontinence were then evaluated 1 year postoperatively. Outcomes were dichotomized into "continent" (ICIQ-Score = 0-5) and "incontinent" (ICIQ-Score 6). In each patient, we determined the perfusion ratio of the levator ani muscle divided by the surrounding pelvic muscle structures and compared them among the groups. RESULTS Forty-two patients were included in the study (n = 22 in "continent", n = 20 in "incontinent" group). The median perfusion ratio from the continent group was significantly higher compared to the incontinent group (1.61 vs. 1.15; 95% CI 0.09-0.81, p = 0.015). The median perfusion ratio in "excellent" (ICIQ-Score = 0) was significantly higher than in "poor" (ICIQ-Score 11) outcomes (1.48 vs. 0.94; 95% CI 0.04-1.03, p = 0.036). Further, a higher perfusion ratio was negatively correlated with ICIQ-Score (r =-0.33; 95% CI-0.58 to 0.03; p = 0.031). CONCLUSIONS Our data demonstrate a promising new strategy to predict PPI through the perfusion quality of pelvic muscle structures with contrast media kinetics. This may facilitate preoperative patient consulting and decision-making.

Anatomical predictors of long‐term urinary incontinence after robot‐assisted laparoscopic prostatectomy: A systematic review

“…On the other hand, some studies also presented the odds ratio (OR) to measure the impact of MUL on UI. In Sadahira's research, 14 the OR of incontinent patients with higher MUL was 0.37 (95% CI: 0.20–0.68) in the multivariate analysis and in Hikita et al 10 it was 0.95 (95% CI: 0.85–1.10); in Schmid's, 15 it was 0.05 (95% CI: 0.00–0.29) in the univariate analysis. In Song's article, 13 the OR of incontinence in patients with MUL ≤ 13.5 was 5.95 ( p = .003) in the multivariate analysis.…”

“…Average MUL in continent patients was 15.9 (±2.6) mm in the article by Song et al, 13 16.1 (95% confidence interval [CI]: 13.9–18.9) in the study by Regis et al, 12 14.5 in the article by Schmid et al, 15 and 12.1 (95% CI: 9.7–14.9) in the article of Sadahira 14 . Conversely, average MUL in incontinent patients was lower, 13.9 (±2.9) mm in Song, 13 10 (95% CI: 8.7–12.1) in Regis, 12 9.3 in Schmidt, 15 and 10.3 (95% CI: 8.7–12.4) in Sadahira 14 …”

“…One aspect that made it difficult to generalize the results was the heterogeneity of MRI resolution. Regis et al, 12 Schmid et al 15 and Kim et al 17 used 3 T MRI, Kitamura 11 et al used 1.5 T MRI, and Sadahira 14 et al used both 3 T and 1.5 T MRI. Furthermore, in the studies by Kitamura 11 and Sadahira 14 MRI measurements were made by a urologist, and in the article by Regis 12 they were analyzed by a urologist but doubtful cases were consulted with a radiologist.…”

“…Patients who used 0 to 1 pad for safety on the expanded prostate cancer index composite (EPIC) questionnaire were considered continent in the studies made by Song et al, 13 Kitamura et al, 11 and Regis et al 12 By contrast, other articles, such as Sadahira et al 14 and Lin et al, 16 were stricter and defined patients as continent only when 0 pads were used on the EPIC scale (excluding the use of a safety pad). On the other hand, Schmid's study 15 considered patients with International Consultarion on Incontinence Questionnaire score less than 6 as continent; therefore, not only were the criteria to define continence more lenient, but they also used a different scale, which made it difficult to compare their outcomes with the results of other works. Moreover, all articles were based on surveys filled out by patients subjectively, and only Kim et al 17 used a 24‐h pad‐test, which is the most specific way to quantify UI.…”

“…When performing this review, we also encountered difficulties in quantifying the magnitude of the effect of MUL on UI (Table 3), considering that Sadahira et al, 14 Song et al, 13 and Schmid et al 15 showed both the OR and the difference of MUL in continent and incontinent patients. By contrast, Regis et al 12 showed their results using the MUL difference but did not use the OR.…”