What is the most effective tool for detecting prostate cancer using a standard MR scanner?

Abstract

Purpose: We aimed to determine which of the following magnetic resonance (MR) imaging sequences is most effective for detecting prostate cancer: T2-weighted (T2W), dynamic contrast-enhanced (DCE) T1-weighted (T1W), or diffusion-weighted (DWI) imaging or apparent diffusion coefficient (ADC) mapping. Materials and Methods: We included 37 male patients with prostate cancer who underwent MR imaging before radical prostatectomy in this retrospective study. Sixty-four foci (>5 mm in size; 35 in the peripheral zone [PZ], 29 in the transitional zone [TZ]) were histopathologically determined to be prostate cancer. We determined the capacity of T2W, DCE-T1W, DWI, ADC mapping alone, and the combination of ADC mapping with DWI, and conventional MR sequences to detect prostate cancer, including their sensitivity and positive predictive value (PPV), with reference to the results obtained in histopathological examinations of whole-mount sections. Results: In the PZ, sensitivities were 31.4% (T2W), 37.1% (DCE-T1W), 51.4% (DWI), and 71.4% (ADC mapping); PPVs were 78.6% (T2W), 92.9% (DCE-T1W), 94.7% (DWI), and 96.0% (ADC mapping). Sensitivity was significantly higher of ADC mapping than other sequences. In the TZ, sensitivities were 55.1% (T2W), 44.8% (DCE-T1W), 82.8% (DWI), and 89.7% (ADC mapping); PPVs were 64.0% (T2W), 46.4% (DCE-T1W), 70.6% (DWI), and 72.2% (ADC mapping). Sensitivity was significantly higher of ADC mapping and DWI than conventional MR imaging, but there was no significant correlation between DWI/ADC mapping and T2W/DCE-T1W with respect to PPVs. Combining sequences did not improve sensitivity; only the PPV in the TZ improved when ADC mapping was combined with DCE-T1W. Conclusion: ADC mapping is the most effective standard MR imaging tool for detecting prostate cancer. The addition of DCE-T1W may improve the PPV of ADC mapping for diagnosing cancer in the TZ. © 2013 Japanese Society for Magnetic Resonance in Medicine.