Can conventional magnetic resonance imaging, prostate needle biopsy, or their combination predict the laterality of clinically localized prostate cancer?

To evaluate conventional magnetic resonance imaging (MRI), prostate needle biopsy (PBx), and the combination of both tests in predicting the laterality of final specimen pathology after radical prostatectomy.

A total of 574 radical prostatectomy cases that had PBx with at least 12 cores and preoperative prostate MRI with pelvic coil were included. We analyzed the clinicopathologic data with laterality based on PBx and MRI. Unilateral disease in combination was defined as unilateral cancer in PBx and at the same time MRI findings of undetectable or ipsilateral disease. Cohen's kappa (κ) was used to measure agreement between the laterality data.

There were a total of 316 (55.1%) unilateral cancers detected by PBx, whereas there were 139 (24.2%) cases in the final specimen pathology. MRI resulted in 119 (20.7%) undetectable and 205 (35.7%) unilateral cancers. Cancer laterality based on final specimen pathology had only fair agreements with PBx (κ = 0.286), MRI (κ = 0.200), and their combination (k = .291). The positive predictive values to predict pathologic concurrent unilaterality were only 30.4% (96/316), 25.9% (53/205), and 34.8% (72/207), respectively. These trends were similar in low-risk cases.

Preoperative PBx, MRI, and the combination of both methods had only a fair correlation with the laterality of prostate cancer (PC), even in low-risk cases. Approximately two thirds of cases diagnosed as unilateral disease by contemporary PBx, MRI, or their combination were not concurrent unilateral disease in final pathology. This should be recognized when planning nerve-sparing surgery and potentially for candidate selection for focal therapy to treat PC.