Biparametric MRI noninferior to multiparametric MRI for detecting clinically significant prostate cancerBiparametric MRI shows similar accuracy to multiparametric MRI for prostate cancer

This meta-analysis evaluated the diagnostic accuracy of biparametric magnetic resonance imaging (bpMRI) compared with multiparametric magnetic resonance imaging (mpMRI) for detecting clinically significant prostate cancer (csPC) in men with suspected prostate cancer. The analysis included 9403 patients across multiple studies, though the specific number of studies and their settings were not reported. The population consisted of men with suspected prostate cancer, but no further demographic details were provided.

The intervention was bpMRI, which omits dynamic contrast-enhanced imaging, while the comparator was mpMRI, the standard full protocol including contrast. Dosing or protocol specifics were not reported. The primary outcome was diagnostic accuracy for csPC at both per-patient and per-lesion levels. Secondary outcomes included detection of any prostate cancer.

At the per-patient level, bpMRI demonstrated noninferiority to mpMRI for sensitivity, with a difference of -2.3% (95% CI -4.1% to -0.5%), meaning bpMRI had slightly lower sensitivity but within the noninferiority margin. Specificity was higher for bpMRI by +1.8% (95% CI -0.4% to +4.0%), also within noninferiority bounds. At the per-lesion level, specificity was noninferior, but sensitivity was not noninferior; exact effect sizes and confidence intervals were not reported for these outcomes.

Safety and tolerability data were not reported, including adverse events, serious adverse events, or discontinuations. The main limitation noted was fewer studies and greater heterogeneity at the lesion level for sensitivity, which prevented establishing noninferiority for that endpoint.

Compared to prior evidence, this meta-analysis confirms that bpMRI can be a viable alternative to mpMRI for patient-level diagnosis, potentially reducing scan time and avoiding contrast administration. However, the lack of noninferiority at the lesion level for sensitivity means that for lesion characterization, mpMRI may still be preferred. This aligns with previous studies suggesting bpMRI's utility but highlights the need for more data on lesion-level accuracy.

Key limitations include the absence of reported study settings, follow-up duration, and funding sources. The heterogeneity at the lesion level for sensitivity weakens the certainty of the per-lesion findings. Additionally, the noninferiority margin was not explicitly defined in the available data, which is critical for interpreting these results.

Clinically, these results support using bpMRI as a first-line imaging tool for detecting csPC at the patient level, potentially reducing costs and contrast-related risks. However, clinicians should be cautious when relying on bpMRI for lesion-level decisions, as noninferiority was not confirmed. Further research is needed to standardize bpMRI protocols and validate lesion-level accuracy.

Unanswered questions include the optimal noninferiority margin, the impact of bpMRI on biopsy decision-making, and long-term outcomes. The role of bpMRI in active surveillance or follow-up also remains unclear.