TL;DR
This study investigated the diagnostic accuracy of robot-guided MRI/TRUS fusion biopsy for prostate cancer in men with prior negative biopsies. While the fusion-guided approach required significantly fewer biopsies, it did not demonstrate a superior detection rate for clinically significant prostate cancer compared to systematic biopsies.
What Problem Does This Paper Address?
Prostate cancer (PCa) is a significant health concern, and early detection is crucial for effective treatment. However, current diagnostic methods face challenges. Prostate-Specific Antigen (PSA) tests have low specificity, and traditional transrectal ultrasound (TRUS)-guided biopsies often have low detection rates and can lead to complications like infections. While transperineal approaches and multiparametric MRI (mpMRI) have improved biopsy efficiency, there's ongoing research into optimizing these techniques, particularly MRI/TRUS fusion, to maximize cancer detection while minimizing invasiveness and patient burden. This paper specifically addresses the utility of a novel robot-guided, software-based transperineal MRI/TRUS fusion biopsy system in a high-risk population of men who have already had a negative biopsy, aiming to determine if this advanced technique offers a significant advantage over traditional systematic biopsies.
Methods
The study employed a robot-guided, software-based transperineal MRI/TRUS fusion biopsy system (MonaLisa by Biobot surgical). This system utilizes an "elastic" fusion algorithm to account for tissue deformation during imaging and a robotic arm for precise biopsy placement and depth control. Patients underwent mpMRI prior to the biopsy. Lesions were segmented manually in the software based on PI-RADS v2 classification. Biopsies were then planned using the fusion software, with both targeted biopsies of suspicious lesions and systematic random biopsies performed using the same transperineal approach. A key aspect of the methodology was the use of a systematic, volume-adapted transperineal random biopsy as a control, ensuring a direct comparison under similar procedural conditions. Patients did not receive antibiotic prophylaxis.
Datasets
The study utilized data from 52 patients who underwent MRI/TRUS fusion biopsy between October 2015 and May 2016. These patients had elevated PSA levels or clinical suspicion for PCa and had previously undergone a negative 12-core transrectal ultrasound-guided biopsy. The dataset comprised 519 targeted biopsies from 135 identified lesions and 1561 random biopsies, totaling 2080 biopsies. The PI-RADS v2 classification was used to categorize lesions.
Key Insights and Ideas
- Fusion Biopsy Efficiency: The MRI/TRUS fusion approach significantly reduced the number of biopsies required to detect clinically significant prostate cancer (median of 9 cores for targeted vs. 32 for random biopsy). This has implications for pathologist workload and patient comfort.
- No Superior Detection Rate for Fusion: Despite the efficiency gains, the MRI/TRUS fusion targeted biopsy did not show a statistically significant advantage in detecting clinically significant prostate cancer compared to systematic random biopsies in this specific patient cohort.
- Peripheral vs. Central Lesions: The study found a significantly higher detection rate for both any and clinically significant PCa in peripheral lesions compared to central lesions when using the targeted biopsy approach.
- PI-RADS Limitations: The lesion-based analysis indicated that the PI-RADS scoring system, on its own, has limitations in predicting cancer detection, with lower detection rates for PI-RADS 3, 2, and unclassified lesions.
- Bilateral Tumor Prevalence: A higher rate of bilateral tumors was observed in the random biopsy group compared to the targeted biopsy group, suggesting that systematic sampling might be better at identifying multifocal disease.
- Adverse Events: The study reported a low rate of adverse events, with one case of rectum perforation in a patient with a very large prostate volume, highlighting the need for caution in such cases.
Key Results
- Overall Detection: The overall detection rate for any prostate cancer was 50.0% for targeted biopsy and 59.6% for random biopsy.
- Clinically Significant PCa Detection: The detection rate for clinically significant PCa (Gleason grade ≥ 4) was 44.2% for targeted biopsy and 50.0% for random biopsy.
- Sampling Efficiency: The median number of cores needed to detect clinically significant PCa was 9 for target biopsy (IQR: 6-14.0) and 32 for random biopsy (IQR: 24-32).
- Utility: Targeted biopsy detected 0% of additional clinically significant PCa cases, while random biopsy detected 3.9% (2/52) of additional cases.
- Peripheral vs. Central Detection: Detection rates for both any and significant PCa were significantly higher in peripheral lesions (39.0% and 32.0%, respectively) compared to central lesions (22.0% and 15.0%, respectively) for targeted biopsies.
- Bilateral Tumors: Significantly higher rates of bilateral tumors were found in random biopsies (70.97%) compared to target biopsies (32.26%).
Limitations and Future Work
The authors acknowledge that their study was a retrospective analysis of a relatively small cohort (52 patients). They also note that the PI-RADS scoring system, when applied on a lesion-based analysis, showed poor performance. Future work could focus on refining the PI-RADS classification scheme to improve its accuracy and reduce false positives and negatives. The authors also suggest that further research is needed to develop feedback correlations from target and random biopsies to the original MRI data to enhance radiological classification. They also propose that a larger study is needed to investigate the implications of high-volume prostates and the potential for underestimation of complications. The economic implications of both targeted (mpMRI, software evaluation) and random (higher pathologist workload) biopsies are also noted as areas for consideration.
Summary updated