To the best of our knowledge this is the first study that describes the detectability of recurrent PC using DCE-MRI without EC in the setting of very low PSA-levels.
DCE-MRI without EC detected recurrent PC with an accuracy of 83%. The pre-RT tumor volumes of the 22 patients with positive pre-RT DCE-MRI findings showed high correlation to pre-RT PSA levels. Our results also indicate a correlation between lesion detectability and increasing PSA-level. A PSA cut-off of ≥0.54 ng/mL detected all lesions positive on DCE-MRI as true positive. We also alert radiation oncologists to acknowledge a commonly observed unequivocal contrast enhancement at the anterior aspect of the proximal urethra immediately below the vesicourethral anastomosis, which represents physiological contrast enhancement.
Modern three-dimensional RT-planning enables the delineation of a GTV for targeted dose escalation on RT-planning CT scans. However, recurrent PC is usually missed on RT-planning CT due to tumor densities equal to the surrounding tissues. Therefore, local recurrent PC is managed so far by delivering RT to the entire fossa prostatica, knowing that local recurrences tend to be located around the vesicourethral anastomosis, but usually without having precise information where the recurrence is located and furthermore without having the possibility to fuse a precise image-based information with the RT-planning CT
DCE-MRI in combination with EC was recently validated for the detection of local recurrent prostate cancer. The reported sensitivities and specificities ranged between 71 – 88%, and 94 – 100%, respectively
[7, 8, 18]. Nevertheless, since endorectal coil placement does not allow image fusion with RT-planning CT, we assessed the sensitivity and specificity of DCE-MRI without EC for the detection of local recurrent prostate cancer.
In this study the pre-RT PSA level of the 22 patients with positive DCE-MRI findings averaged 0.74±0.64 ng/mL (median, 0.51 ng/mL, range, 0.11-2.38) and 0.57±0.58 ng/ml (median, 0.34 ng/mL; range, 0.08-2.38 ng/mL) for the entire group. Against the background of lower pre-RT PSA levels compared to previous studies that investigated the sensitivity of DCE-MRI with EC in the detection of PC recurrence (PSA, averaged 0.8 - 1.26 ng/mL), the sensitivity of 67% reported in this study appears to be comparable to the performance of DCE-MRI with EC (sensitivity, 71 – 88%)
[7, 8, 18].
In the normal post-surgical anatomy of the male pelvis following radical prostatectomy variable degrees of post-surgical fibrosis may be present
[19, 20]. However, no enhancement of the prostatic bed in the arterial phase after administration of i.v. gadolinium had been described by Allen et al.
 or had been found in the control group of Sciarra et al.
. In addition to these results we describe a frequent contrast enhancement inferior of the vesicourethral anastomosis, anterior to the proximal urethra in 27 of the 33 post-RP patients (82%) with distinctive morphologic characteristics as described above (Figure
4a and b). This finding would be familiar to radiologists with thorough experience of prostate- and urogenital MRI-imaging representing physiological periurethral vascular elements in the region of the diaphragma urogenitale, which can be seen regularly in contrast enhanced MRI of the male pelvis without prior surgery
[17, 21, 22]. Six of the 33 patients (18%) did not exhibit these characteristic changes. This might be explained by varying resection margins during RP. To avoid false positive readings knowledge of this non-disease-specific structure is crucial for correct image interpretation. Local recurrent tumor nodules tend to be located most frequently around the posterior and lateral aspects of the vesico-urethral anastomosis and rather rarely in the anterior aspects as we found in our cohort which is in concordance with previous studies
Limitations of the study
First, this is a retrospective study and the results need to be confirmed in future prospective clinical trials.
Second, histopathologic confirmation of the DCE-MRI findings was not available. However, detection rates using ultrasound or MRI guided biopsy are reported to be low (range, 30 – 66%), with a positive biopsy of less than 30% in patients with PSA levels <1 ng/mL
[5, 8, 23–25].
We used the post-RT DCE-MRI (which was performed at median 15 months after salvage RT) and complete PSA response as a negative control. This is justifiable, since the combination of a negative post-RT DCE-MRI and unremarkable PSA levels rule out macroscopic disease. Whether the high true negative rate reported in this study stands up to a control group of post-RP patients without local recurrent PC needs further investigation, however, our high true negative rate confirms reports of other authors