Seminal vesicle invasion (SVI) refers to the spread of prostate cancer into the seminal vesicles, the paired glands that sit just behind the bladder. When cancer has invaded these structures, it is classified as stage T3b, a finding that dramatically changes treatment planning and has a significant negative impact on prognosis.
Accurately staging SVI before surgery or radiotherapy is important because it affects decisions about surgical margins, whether to extend radiation fields, and whether to add hormone therapy. Underestimating invasion can lead to undertreated disease, while overestimating it may push patients toward unnecessarily aggressive treatment.
Multiparametric MRI (mpMRI) has become the standard imaging modality for local prostate cancer staging because it provides complementary information from multiple sequences. However, the specific value of individual sequences for detecting SVI, and how they should best be combined, remained an open clinical question that this study aimed to address.
The study enrolled 131 consecutive patients who underwent preoperative endorectal multiparametric MRI followed by radical prostatectomy. The surgical pathology report served as the gold standard for confirming whether seminal vesicle invasion was present or absent.
Three experienced radiologists independently evaluated the MRI scans under three progressively richer reading conditions: T2-weighted imaging alone, then T2-weighted plus diffusion-weighted imaging (DWI), and finally T2-weighted plus DWI plus dynamic contrast-enhanced (DCE) imaging. Each reader was blinded to both the other conditions and to pathology results during each reading session.
This within-patient design, where the same images were read multiple times under different conditions, allowed a direct comparison of diagnostic performance across imaging protocols. Receiver operating characteristic (ROC) analysis was used to calculate area under the curve (AUC), sensitivity, specificity, and positive and negative predictive values for each condition and each reader.
T2-weighted imaging (T2W) provides detailed anatomical contrast between soft tissues and has been the backbone of prostate staging for decades. It shows the zonal anatomy of the prostate and the seminal vesicles clearly, and radiologists look for direct tumor extension, loss of the fat plane between the prostate and seminal vesicle, and asymmetric low signal intensity within the seminal vesicle as signs of invasion.
Diffusion-weighted imaging (DWI) measures how freely water molecules move through tissue at a microscopic level. Cancer cells have densely packed nuclei that restrict water diffusion, producing a characteristic high signal on DWI and low signal on the calculated apparent diffusion coefficient (ADC) map. This contrast can help highlight tumor within or adjacent to the seminal vesicles that may be subtle on T2W alone.
Dynamic contrast-enhanced (DCE) imaging captures the rapid uptake and washout of intravenous gadolinium contrast agent in tumor tissue, reflecting the abnormal leaky blood vessels that cancers develop. High early enhancement followed by rapid washout is a hallmark of prostate cancer on DCE and was evaluated here for its incremental value in SVI detection beyond DWI.
Adding diffusion-weighted imaging to T2-weighted imaging produced a meaningful and statistically significant improvement in diagnostic performance. Specificity for detecting seminal vesicle invasion improved from 93.1% to 96.6% when DWI was added, and the positive predictive value (PPV) rose from approximately 52% to 70%.
These gains matter clinically: higher specificity means fewer false-positive SVI calls, reducing the risk of inappropriately staging a patient as T3b and steering them toward more aggressive treatment. Higher PPV means that when a radiologist calls SVI positive, there is a much greater likelihood that the pathologist will confirm it.
Adding dynamic contrast-enhanced imaging to the T2W plus DWI combination provided no statistically significant further improvement in any metric for any of the three readers. This finding suggests that DCE does not contribute incremental diagnostic value for SVI detection once DWI information is already available, which has implications for how mpMRI protocols should be designed.
The findings support a T2W plus DWI protocol as the minimum standard for preoperative staging of seminal vesicle invasion. The incremental benefit of DWI over T2W alone is clinically meaningful, particularly in reducing false positives that could lead to overstaging.
The absence of added value from DCE has practical implications. DCE requires intravenous contrast injection, adds scanning time and cost, introduces a small risk of adverse reactions, and generates substantially more image data to review. If it does not improve SVI detection, its routine inclusion in prostate staging protocols could be reconsidered, or it could be reserved for specific clinical scenarios where other diagnostic questions benefit from it.
Radiologists who review prostate MRI for staging purposes should ensure they actively interpret DWI and ADC maps when evaluating the seminal vesicles, rather than relying predominantly on T2-weighted findings. The training and experience of the reader also remains important, as this was a study conducted by subspecialty-trained genitourinary radiologists.
This prospective study with surgical pathology as the gold standard provides level 2 evidence that adding DWI to T2-weighted MRI significantly improves the detection of seminal vesicle invasion in prostate cancer, primarily through gains in specificity and positive predictive value.
The lack of additional benefit from DCE challenges the assumption that more sequences always improve diagnostic performance. These results suggest that protocol optimization requires evidence-based evaluation of each component, and that a streamlined T2W plus DWI protocol may be both simpler and just as accurate for this specific staging task.
The study adds to the growing evidence base supporting multiparametric MRI as the preferred preoperative staging tool for prostate cancer, while helping define which specific sequences are essential for which clinical questions. Future research could evaluate whether these findings hold across different reader experience levels, scanner field strengths, or patient populations.