Interim analysis of the “shutter speed” MRI model to detect clinically significant prostate cancer.

Abstract

339 Background: Prostate cancer (PCa) MRI can fail to identify high-grade lesions, and mis-grade lesions, up to 30% of the time. Objective imaging techniques are needed to better discern aggressive from indolent PCa. While the dynamic contrast enhanced (DCE) sequence of mpMRI has not played a routine role in identifying clinically significant PCa (csPCa), improvements in DCE may aid in the detection of cancers with weak signals on other MRI sequences. We studied the ability of “Shutter Speed” MRI (SSMRI), a technique based on DCE including objective measurements of perfusion, to detect csPCa lesions and adverse pathology. Methods: We analyzed PCa lesions from patients who underwent standardized MRI (Siemens VIDA 3T, endorectal coil; planned n=124) followed by radical prostatectomy with whole-mount histopathology (n=95). An experienced GU pathologist annotated lesion size and Grade Group (GG). An experienced radiologist reviewed all mpMRI lesions, and the SSDCE parameters were measured post-acquisition using a standardized software package. This paired data analysis compared the ability of SSMRI (SSDCE, T2, DWI) and mpMRI (standard DCE, T2, DWI) to detect csPCa lesions (ISUP ≥ GG 2, ≥5mm) or adverse pathology (extracapsular extension [ECE], seminal vesical invasion [SVI], lymph node invasion [LNI], GG≥4, or their composite [CO]) using logistic regression. We constructed receiver operator characteristic curves (ROC), and AUC was compared for the detection of csPCa (primary endpoint) and adverse pathology (secondary endpoint) (Table). Results: Patient mean age and PSA at diagnosis were 64.6 years and 9.7 ng/mL respectively. 91 lesions were assessed, of which 25 met criteria for csPCa. When comparing ROC for detection of csPCa, SSDCE improves the AUC of DCE alone from 0.72 (95% CI 0.50-0.95) to 0.83 (95% CI 0.71-0.95), and when added to mpMRI improves the AUC from 0.80 (95% CI 0.75-0.86) to 0.94 (95% CI 0.88-1.00) (p <0.0001). While the AUC improved by 0.13 for the detection of SVI when SSDCE was added to mpMRI, this did not reach statistical significance, nor did it for any other adverse feature. Conclusions: SSMRI significantly enhanced the detection of csPCa lesions. A larger cohort is needed to assess SSMRI’s ability to detect adverse pathology. These findings further the efforts toward earlier and more accurate diagnosis of csPCa. Comparison of AUC: Addition of SS to DCE and mpMRI. Methods and AUC (95%CI): DCE SSDCE mpMRI SSMRI P value csPCa 0.72 (0.50-0.95) 0.83 (0.71-0.95) 0.80 (0.75-0.86) 0.94 (0.88-1.00) P<0.0001 ECE 0.64 (0.50-0.79) 0.63 (0.48-0.78) 0.70 (0.59-0.81) 0.72 (0.59-0.86) P=0.66 SVI 0.75 (0.61-0.89) 0.75 (0.60-0.90) 0.71 (0.60-0.83) 0.84 (0.72-0.96) P=0.12 LNI 0.64 (0.48-0.79) 0.59 (0.42-0.76) 0.71 (0.62-0.81) 0.73 (0.57-0.88) P=0.88 GG≥ 8 0.67 (0.50-0.83) 0.63 (0.45-0.82) 0.67 (0.56-0.79) 0.74 (0.57-0.91) P=0.23 CO 0.66 (0.51-0.81) 0.62 (0.46-0.78) 0.70 (0.58-0.81) 0.71 (0.56-0.85) P=0.67