Multiparametric 3T MRI in the evaluation of intraglandular prostate cancer: correlation with histopathology

Prostate cancer is common and may be treated immediately or managed conservatively by observation. We sought to determine how reliable multiparametric MRI is in the detection of intraprostatic prostate cancer and what role it has in risk stratification.

The histology from 38 whole mount prostate specimens was compared with preoperative multiparametric 3T MRI studies with an endorectal receiver coil in place. T1-weighted, T2-weighted, diffusion (b values 50 400 800), perfusion (Ve , Kep , Ktrans , area under the curve) and proton spectroscopic sequences were used.

For cancers greater than 0.5 cc, the detection rate for combined T2-weighted imaging and diffusion-weighted imaging (DWI) was 85%. For cancers 0.1 cc-0.5 cc, the sensitivity was 52%.Per patient, false positive rate was 50% for combined T2-weighted imaging and DWI. Perfusion imaging had a sensitivity of 70% for tumours greater than 0.5 cc but had a per patient false positive rate of 80% influenced by benign prostatic hypertrophy. In only 15 patients could a satisfactory spectroscopy study be obtained. Weak correlation was found between the Gleason score and tumour size (r = 0.51), apparent diffusion coefficient (ADC) (r = -0.30) and (choline + creatine)/citrate ratio (r = 0.41).

T2-weighted imaging and DWI in combination were the best strategy for detecting prostate cancer and had a sensitivity of 85% for detecting lesions greater than 0.5 cc. At 3T, an ADC threshold of between 1100-1200.10(-6)  mm(2) /s was optimal for diagnosing prostate cancer. There are significant limitations in the use of perfusion and spectroscopy to detect prostate cancer. Magnetic resonance imaging-targeted or guided biopsy post-MRI imaging is likely to be needed in some patients to assist risk stratification.