Risk-based MRI-directed diagnostic pathway outperforms non-risk-based pathways in suspected prostate cancer biopsy-naïve men: a large cohort validation study

• Published in: European radiology Vol. 32; no. 4; pp. 2330 - 2339
• Main Authors: Bittencourt, Leonardo Kayat, Guricova, Karolina, Zucker, Isaac, Durieux, Jared C., Schoots, Ivo G.
• Format: Journal Article
• Language: English
• Published: Berlin/Heidelberg Springer Berlin Heidelberg 01-04-2022; Springer Nature B.V
• Subjects: Avoidance; Biopsy; Chi-square test; Cohort Studies; Decision analysis; Diagnosis; Diagnostic Radiology; Diagnostic systems; Genetics; Health risks; Humans; Image-Guided Biopsy - methods; Imaging; Internal Medicine; Interventional Radiology; Magnetic resonance imaging; Magnetic Resonance Imaging - methods; Male; Maximization; Medical diagnosis; Medicine; Medicine & Public Health; Neuroradiology; Optimization; Prostate cancer; Prostatic Neoplasms - diagnostic imaging; Prostatic Neoplasms - pathology; Radiology; Reduction; Retrospective Studies; Risk analysis; Risk factors; Statistical analysis; Statistical tests; Ultrasound; Urogenital; MRI-directed diagnosis pathway; Prostate biopsy; Prostate cancer; Magnetic resonance
• ISSN: 0938-7994; 1432-1084
• DOI: 10.1007/s00330-021-08407-6

Objectives
To investigate and compare the performance of different proposed diagnostic pathways in a cohort of biopsy-naïve men at risk for prostate cancer (PCa), in terms of biopsy avoidance, accurate diagnosis of clinically significant prostate cancer (csPCa), and reduction in overdiagnosis of clinically insignificant cancer (cisPCa), with particular focus on a recently suggested “risk-based” MRI-directed diagnostic pathway.
Methods
Single-center, retrospective cohort study, including 499 biopsy-naïve men at risk for PCa. All men underwent PI-RADS-compliant prostate MRI, transrectal ultrasound fusion-guided targeted (TBx), and systematic biopsy (SBx). Five diagnostic pathways were retrospectively evaluated and compared for. Outcome measures were biopsy avoidance, combined with missed csPCa and detected cisPCa. csPCa and cisPCa were defined as ISUP grade group ≥ 2 and grade = 1, respectively. Chi-square test was used for statistical analysis. Decision curve analyses were used to compare the benefits of the pathways across a range of biopsy thresholds.
Results
The prevalence (detection-focused [reference] pathway) of csPCa and cisPCa was 52.9% (264/499) and 23.0% (115/499). MRI-focused pathway (no biopsy in PI-RADS 1–2 men) did not significantly reduce ISUP ≥ 2 cancer detection (52.1% (260/499);
p
 = 0.13), but significantly reduced ISUP 1 cancers diagnosed (20.6% (103/499);
p
 < 0.01), and biopsy avoidance was 11.8% (59/499). The risk-based MRI-directed pathway (no biopsy in low-risk PI-RADS 1–3 men) resulted in a small reduction of ISUP ≥ 2 diagnosed (51.7% (258/499);
p
 = 0.04), however non-significant when compared to MRI-focused pathway (
p
 = 0.625). Moreover, the risk-based pathway further reduced detection of ISUP 1 (18.6% (93/499);
p
 < 0.01), and biopsy avoidance was 19.2% (96/499). Decision curve analysis showed maximized net benefit of the risk-based pathway, for the range of threshold probabilities between 6.25 and 65%.
Conclusion
The risk-based MRI-directed pathway for prostate cancer diagnosis was optimal in balancing accurate diagnosis, reducing overdiagnosis, and maximizing biopsy avoidance. This substantial evidence should inform guideline recommendations towards using “risk-based” MRI-directed biopsy decisions in biopsy-naïve men at risk of significant prostate cancer.
Key Points
•
Our study recognizes the added value of prostate MRI and MR-targeted biopsies in order to propose clinical diagnostic pathways for prostate cancer, towards maximizing the potential avoidance of unnecessary biopsies, while maintaining optimal detection rate of clinically significant prostate cancer.
•
The risk-based MRI-directed pathway incorporates risk factors such as PSA density, digital rectal examination, and family history to further refine the initial stratification of patients based on PI-RADS scores.
•
In this study, the risk-based pathway had the most optimal performance in terms of combination of outcomes, with the highest rate of biopsy avoidance (19.2%), while keeping a high detection rate of clinically significant prostate cancer (51.7%), when compared to the reference standard (52.9%).