Your browser doesn't support javascript.
loading
Per-sextant localization and staging of prostate cancer: correlation of imaging findings with whole-mount step section histopathology.
Graser, Anno; Heuck, Andreas; Sommer, Bernhard; Massmann, Joerg; Scheidler, Juergen; Reiser, Maximillian; Mueller-Lisse, Ullrich.
Affiliation
  • Graser A; Department of Clinical Radiology, University of Munich, Marchioninistrasse 15, Munich 81377, Germany.
AJR Am J Roentgenol ; 188(1): 84-90, 2007 Jan.
Article in En | MEDLINE | ID: mdl-17179349
ABSTRACT

OBJECTIVE:

The objective of our study was to determine the diagnostic accuracy and interobserver agreement of 1.5-T prostatic MRI for per-sextant tumor localization and staging of prostate cancer as compared with whole-mount step section histopathology. MATERIALS AND

METHODS:

Combined endorectal-pelvic phased-array prostatic MRI scans obtained at 1.5 T of 106 patients with biopsy-proven prostate cancer who had undergone radical prostatectomy with whole-mount step section histopathology within 28 days of MRI were retrospectively analyzed by three independent abdominal radiologists (reviewers 1, 2, and 3). Sextants of the prostate (right and left base, middle, and apex) were evaluated for the presence of prostate cancer and extracapsular extension (ECE) using a 5-point confidence scale. Data were statistically analyzed using receiver operating characteristic (ROC) analysis. Interobserver variability was assessed by kappa statistics. For calculation of sensitivity and specificity, data from the 5-point confidence scale were dichotomized into negative (score of 1-3) or positive (score of 4 or 5) findings.

RESULTS:

Forty-one patients had ECE (tumor stage T3), and 65 patients had organ-confined disease (stage T2). Of 636 prostatic sextants, 417 were positive for prostate cancer and 135 were positive for ECE at histopathology. For prostate cancer localization, ROC analysis yielded area under the ROC curve (AUC) values ranging from 0.776 +/- 0.023 (SD) to 0.832 +/- 0.027. For the detection of ECE, the AUC values ranged from 0.740 +/- 0.054 to 0.812 +/- 0.045. Interobserver agreement (kappa) ranged from 0.49 to 0.60 for prostate cancer localization and from 0.59 to 0.67 for the detection of ECE.

CONCLUSION:

Using the sextant framework, independent observers reach similar accuracy with moderate to substantial agreement for the localization of prostate cancer and ECE by means of MRI of the prostate.
Subject(s)
Search on Google
Database: MEDLINE Main subject: Prostatic Neoplasms / Magnetic Resonance Imaging / Anatomy, Cross-Sectional / Biopsy, Fine-Needle / Neoplasm Staging Type of study: Diagnostic_studies / Prognostic_studies Limits: Humans / Male / Middle aged Language: En Year: 2007 Type: Article
Search on Google
Database: MEDLINE Main subject: Prostatic Neoplasms / Magnetic Resonance Imaging / Anatomy, Cross-Sectional / Biopsy, Fine-Needle / Neoplasm Staging Type of study: Diagnostic_studies / Prognostic_studies Limits: Humans / Male / Middle aged Language: En Year: 2007 Type: Article