A retrospective study of 384 men biopsied at a single Chinese hospital has found that prostate cancer detection was higher in 2022, after prebiopsy MRI became routine, than in 2017, when MRI was not used. The researchers, publishing in BMC Urology, are careful about what this means: because the same systematic biopsy template was used throughout and operators could see the MRI reports, the increase can only be described as era-associated, not attributed to MRI itself.
The study compared two cohorts of men undergoing systematic ultrasound-guided biopsy, one from 2017 and one from 2022. After removing duplicates and biopsy episodes without a recorded outcome, 384 men were analysed: 271 from 2017 and 113 from 2022. No targeted or fusion-guided cores were used in either era, so any change in what the needles found could not be explained by a switch to MRI-targeted sampling.
MRI was performed in 111 of 113 men, or 98.2 percent, in the 2022 cohort, and was absent in 2017. Crude cancer detection did not differ between the two groups, but men in 2022 presented with lower prostate-specific antigen levels, a difference that was statistically significant at P = 0.008. This shift in the profile of men reaching biopsy is itself a plausible consequence of MRI-based selection changing who gets referred.
When the researchers standardised for age and PSA using logistic regression with marginal standardisation, detection was higher in 2022 across all three endpoints. Any cancer rose by 14.4 percentage points, with a 95 percent confidence interval of 4.6 to 23.5. Clinically significant prostate cancer, defined as Gleason grade 7 or higher, rose by 10.4 points, from 0.9 to 19.7. High-grade cancer, Gleason 8 or above, rose by 12.2 points, from 3.2 to 20.9. Spline and imputed analyses produced similar estimates, suggesting the findings were not driven by how missing data or the PSA distribution were handled.
Biopsy volume fell sharply over the same interval, dropping by 59 percent. This pattern, fewer biopsies with a higher yield of significant cancer among those performed, is consistent with MRI acting as a gatekeeper that filters out men unlikely to benefit. But the study design cannot confirm that mechanism, because the operators placing the biopsy needles were not blinded to the MRI findings.
The authors highlight this as the study’s central limitation. Prebiopsy MRI could raise detection through two channels that are hard to separate: by selecting which men undergo biopsy at all, and by influencing where the cores of a nominally systematic template are placed. Since sampling stayed systematic but operators saw the MRI, the observed difference reflects the combined effect of the whole era of practice change, not an isolated MRI effect.
A second part of the study examined whether an apparent shift in where tumours were found within the prostate was real or an artefact of reporting. In 160 positive biopsies, transition-zone involvement appeared to rise from 27.5 percent to 84.3 percent when judged by the coded field in the pathology reports. When the same biopsies were re-coded from the free text of the reports, the figures were 66.1 percent versus 78.4 percent, a much smaller gap that was not statistically significant at P = 0.71.
A blinded second reader validated the coding rules. The reader agreed with the text-based rule with a kappa of 0.78, far more than with the field-based rule, where kappa was 0.29 overall and just 0.09 in the 2017 cohort. The authors conclude that the zone-involvement change was sensitive to coding and should not be read as a biological shift in tumour location.
The MRI reports themselves were also coded two ways. The blinded reader’s classification agreed with the automated coding in 77.5 percent of cases, with a kappa of 0.67, and detection showed a graded association across MRI categories: 25.0 percent, 27.5 percent and 79.5 percent, with P less than 0.001. This graded pattern is striking, but as an observational finding from a single centre it describes an association rather than proving that MRI visibility caused the higher detection.
The study is retrospective, covers one hospital, and spans only two calendar years, so its results may not generalise to other settings or to centres that adopted MRI-targeted fusion biopsy. The authors state plainly that the difference between eras cannot be decomposed into selection and sampling effects. What the study does establish is a measurable rise in detection, including clinically significant and high-grade disease, alongside a large fall in biopsy numbers, during an interval in which prebiopsy MRI went from absent to nearly universal at this centre.
Subject of Research: Retrospective comparison of prostate cancer detection before and after prebiopsy MRI was introduced at a single centre.
Article Title: Prostate cancer detection before and after the introduction of prebiopsy MRI at a single centre: a retrospective comparison of 384 men undergoing systematic ultrasound-guided biopsy
Article References: Xiao, F., Chen, H., Tang, L., You, T., & Cao, B. (2026). Prostate cancer detection before and after the introduction of prebiopsy MRI at a single centre: a retrospective comparison of 384 men undergoing systematic ultrasound-guided biopsy. BMC Urology. https://doi.org/10.1186/s12894-026-02433-6
Image Credits: AI Generated
DOI: 10.1186/s12894-026-02433-6
Keywords: prostate cancer, prebiopsy MRI, systematic biopsy, clinically significant prostate cancer, retrospective cohort, PSA, Gleason grade, biopsy detection rate, transition zone, single-centre study, Prostate, cancer
