Detection of clinically significant prostate cancer: Short dual-pulse sequence versus standard multiparametric MR Imaging - A multireader study Journal Article


Authors: Barth, B. K.; De Visschere, P. J. L.; Cornelius, A.; Nicolau, C.; Vargas, H. A.; Eberli, D.; Donati, O. F.
Article Title: Detection of clinically significant prostate cancer: Short dual-pulse sequence versus standard multiparametric MR Imaging - A multireader study
Abstract: Purpose: To compare the diagnostic performance of a short dual-pulse sequence magnetic resonance (MR) imaging protocol versus a standard six-pulse sequence multiparametric MR imaging protocol for detection of clinically significant prostate cancer. Materials and Methods: This HIPAA-compliant study was approved by the regional ethics committee. Between July 2013 and March 2015, 63 patients from a prospectively accrued study population who underwent MR imaging of the prostate including transverse T1-weighted; transverse, coronal, and sagittal T2-weighted; diffusion- weighted; and dynamic contrast material-enhanced MR imaging with a 3-T imager at a single institution were included in this retrospective study. The short MR imaging protocol image set consisted of transverse T2-weighted and diffusion-weighted images only. The standard MR imaging protocol image set contained images from all six pulse sequences. Three expert readers from different institutions assessed the likelihood of prostate cancer on a five-point scale. Diagnostic performance on a quadrant basis was assessed by using areas under the receiver operating characteristic curves, and differences were evaluated by using 83.8% confidence intervals. Intra- and interreader agreement was assessed by using the intraclass correlation coefficient. Transperineal template saturation biopsy served as the standard of reference. Results: At histopathologic evaluation, 84 of 252 (33%) quadrants were positive for cancer in 38 of 63 (60%) men. There was no significant difference in detection of tumors larger than or equal to 0.5 mL for any of the readers of the short MR imaging protocol, with areas under the curve in the range of 0.74-0.81 (83.8% confidence interval [CI]: 0.64, 0.89), and for readers of the standard MR imaging protocol, areas under the curve were 0.71-0.77 (83.8% CI: 0.62, 0.86). Ranges for sensitivity were 0.76-0.95 (95% CI: 0.53, 0.99) and 0.76-0.86 (95% CI: 0.53, 0.97) and those for specificity were 0.84-0.90 (95% CI: 0.79, 0.94) and 0.82-0.90 (95% CI: 0.77, 0.94) for the short and standard MR protocols, respectively. Ranges for interreader agreement were 0.48-0.60 (83.8% CI: 0.41, 0.66) and 0.49-0.63 (83.8% CI: 0.42, 0.68) for the short and standard MR imaging protocols. Conclusion: For the detection of clinically significant prostate cancer, no difference was found in the diagnostic performance of the short MR imaging protocol consisting of only transverse T2-weighted and diffusion- weighted imaging pulse sequences compared with that of a standard multiparametric MR imaging protocol. © RSNA, 2017.
Journal Title: Radiology
Volume: 284
Issue: 3
ISSN: 0033-8419
Publisher: Radiological Society of North America, Inc.  
Date Published: 2017-09-01
Start Page: 725
End Page: 736
Language: English
DOI: 10.1148/radiol.2017162020
PROVIDER: scopus
PUBMED: 28346073
PMCID: PMC5584651
DOI/URL:
Notes: Article -- Export Date: 5 September 2017 -- Source: Scopus
Altmetric
Citation Impact
MSK Authors