Detecting positive surgical margins: Utilisation of light-reflectance spectroscopy on ex vivo prostate specimens

Aaron H. Lay, Xinlong Wang, Monica S C Morgan, Payal Kapur, Hanli Liu, Claus Roehrborn, Jeffrey A Cadeddu

Research output: Contribution to journalArticle

5 Scopus citations

Abstract

Objective: To assess the efficacy of light-reflectance spectroscopy (LRS) to detect positive surgical margins (PSMs) on ex vivo radical prostatectomy (RP) specimens. Materials and Methods: A prospective evaluation of ex vivo RP specimens using LRS was performed at a single institution from June 2013 to September 2014. LRS measurements were performed on selected sites on the prostate capsule, marked with ink, and correlated with pathological analysis. Significant features on LRS curves differentiating malignant tissue from benign tissue were determined using a forward sequential selection algorithm. A logistic regression model was built and randomised cross-validation was performed. The sensitivity, specificity, accuracy, negative predictive value (NPV), positive predictive value (PPV), and area under the receiver operating characteristic curve (AUC) for LRS predicting PSM were calculated. Results: In all, 50 RP specimens were evaluated using LRS. The LRS sensitivity for Gleason score ≥7 PSMs was 91.3%, specificity 92.8%, accuracy 92.5%, PPV 73.2%, NPV 99.4%, and the AUC was 0.960. The LRS sensitivity for Gleason score ≥6 PSMs was 65.5%, specificity 88.1%, accuracy 83.3%, PPV 66.2%, NPV 90.7%, and the AUC was 0.858. Conclusions: LRS can reliably detect PSMs for Gleason score ≥7 prostate cancer in ex vivo RP specimens.

Original languageEnglish (US)
JournalBJU International
DOIs
StateAccepted/In press - 2016

Keywords

  • Positive surgical margins
  • Prostate cancer
  • Prostatectomy
  • Spectroscopy

ASJC Scopus subject areas

  • Urology

Fingerprint Dive into the research topics of 'Detecting positive surgical margins: Utilisation of light-reflectance spectroscopy on ex vivo prostate specimens'. Together they form a unique fingerprint.

  • Cite this