Can a hybrid MRI point-based system simplify risk stratification for clinically significant prostate cancer (csPCA)?
For the retrospective study, recently published in the Journal of the American College of Radiology (JACR), researchers developed and evaluated a point-based model — based on PI-RADS scoring, prostate-specific antigen density (PSAD) and extraprostatic extension — for predicting csPCa in 960 men who had prostate MRI exams and a biopsy within six months of the MRI exams. All patients in the. Cohort had PSA levels > 4 ng/mL, according to the study.
The study authors noted the point-based model assigned five points for PI-RADS 5, two points for PI-RADS 4, one point for a questionable presence of extraprostatic extension (EPE) and three points for definite detection of EPE. The researchers said the PSAD scoring included three points for 0.05 - < 0.10 ng/mL2, six points for 0.10 - <0.15 ng/mL2, seven points for 0.15 - < 0.20 ng/mL2 and nine points for > 0.20 ng/mL2.
When researchers employed a risk threshold of > five points, they found that the model provided a sensitivity rate of 89.5 percent and may prevent up to 20.4 percent of prostate biopsies.
Determining optimal cutoff points for risk stratification remains an ongoing challenge in this patient population, according to the study authors.
“While we would miss (approximately) 10% of biopsies, we would reduce prostate biopsy by (approximately) 20%. There is a tradeoff between reducing prostate biopsy and increasing sensitivity. While we want to increase sensitivity for screening for csPCa, a lower point cutoff would find more cancers but require more biopsies. A higher point cutoff would delay the detection of some potentially dangerous csPCa,” wrote lead study author Ronilda Lacson, M.D., Ph.D., an associate director of the Center for Evidence-Based Imaging at Brigham and Women’s Hospital in Boston and an associate professor of radiology at Harvard Medical School, and colleagues.
Three Key Takeaways
- Improved risk stratification. The hybrid MRI point-based model, incorporating PI-RADS scoring, prostate-specific antigen density (PSAD), and extraprostatic extension, offers a simplified approach to stratifying risk for clinically significant prostate cancer (csPCa). This model may enhance risk assessment accuracy by combining multiple parameters, potentially aiding clinicians in decision-making regarding biopsy necessity.
- Sensitivity optimization. The study demonstrates the importance of selecting an appropriate risk threshold cutoff for balancing sensitivity and biopsy reduction. By adjusting the cutoff score, clinicians can optimize sensitivity rates, with higher thresholds leading to fewer unnecessary biopsies but potentially missing some csPCa cases, while lower thresholds enhance sensitivity but increase the biopsy rate.
- Clinical utility of MRI results. Integration of PI-RADS score, PSA density, and extraprostatic extension from MRI reports can facilitate csPCa risk estimation using a point-based system. This approach, once prospectively validated, may streamline risk assessment for patients with elevated PSA levels, offering valuable insights for both patients and clinicians in determining the necessity of prostate biopsies.
When the study authors examined use of the model at a risk threshold cutoff score of > two points, they noted a significantly higher sensitivity rate (98.4 percent) and negative predictive value (NPV) in comparison to the aforementioned risk threshold cutoff of > five points (78.6 percent vs. 70.1 percent). However, they also pointed out a lower reduction of prostate biopsies at 4.3 percent with the lower risk threshold cutoff score.
Overall, though, the researchers found that the point-based model for csPCa risk classification exceeded their expectations in validation testing. For 72 patients with risk scores between 10 and 13, the study authors noted an 80.6 percent risk for csPCa (nearly 16 percent higher than expected) as well as an 86.4 percent observed risk among 66 patients with risk scores between 14 and 17.
“This study developed a clinically useful predictive model for identifying csPCa in patients with elevated PSA using prostate MRI results. Once prospectively validated, availability of PI-RADS score, PSA density, and extraprostatic extension of the tumor in MRI reports can potentially be useful for patients and clinicians in performing csPCA risk estimation using a point-based system,” noted Lacson and colleagues.
(Editor’s note: For related content, see “Study: PET/MRI May Prevent Up to 83 Percent of Unnecessary Biopsies in Men with PI-RADS 3 Lesions,” “New Research Evaluates PI-RADS Upgrading Rules in MRI Exams for Prostate Cancer” and “Seven Takeaways from Recent Review on Prostate MRI Imaging Quality.”)
Beyond the inherent limitations of retrospective study, the authors noted inter-reader variability with PI-RADS scoring and acknowledged that the PI-RADS system was updated twice in the five-year study period. They also pointed out that MRI exams were performed after biopsies in a minority of the reviewed cases.
