James Michael Provenzale MD, Presenter: Consultant, Bayer AG Research Grant, General Electric Company Consultant, Biomedical Systems Consultant, Theradex Systems, Inc Shareholder, Amirsys, Inc Consultant, Takeda Pharmaceutical Company Limited Consultant, Shire plc Consultant, Medical Communications Media, Inc

Aaron Mohs PhD, Abstract Co-Author: Consultant, Spectropath, Inc

Michael Mancini PhD, Abstract Co-Author: Employee, Spectropath, Inc

Corey Saba DVM, Abstract Co-Author: Nothing to Disclose

Elizabeth Howerth DVM, Abstract Co-Author: Nothing to Disclose

Karen Cornell DVM, PhD, Abstract Co-Author: Nothing to Disclose

Shuming Nie PhD, Abstract Co-Author: Nothing to Disclose

To assess the feasibility of using a fluorescence imaging and spectroscopic handheld device for intraoperative tumor detection.

Sixteen dogs underwent 17 surgeries for excision of 18 tumors following an intravenous infusion of 220 mcg/kg indocyanine green (ICG) at 24 hours (n=11), 220 mcg/kg ICG 4 hours (n=1) or 71 mcg/kg 4 hours prior to surgery (n=4). Fluorescence spectra were obtained with a hand-held spectroscopic pen device that provides excitation light, collects emitted light from tissue and transfers collected light to a spectrograph that resolves over the range of 800-930 nm. We compared median ICG signal intensity value and 95% confidence interval in normal tissue and tumor using the Mann-Whitney non-parametric test and (when possible) calculated receiver-operator characteristic (ROC) curves.  

We reliably detected ICG signal in 15 masses. In 8 tumors, sufficient amounts of both tissue types allowed full statistical analysis. In 6 tumors, the median values for normal tissue and tumor significantly differed. ROC curve analysis showed area under the curve (AUC) values ranging from 0.82 to 1.00 (mean: 0.93), indicating a high degree of tissue/tumor discrimination. In 2 tumors (in the same patient), median values were not significantly different. Mean sensitivity in 4 tumors was 88.1% (range 75%-100%), mean specificity was 78.9% (range: 64.3%-90%), and mean likelihood ratio was 5.25.

In general, we found good ability to discriminate tumor from normal tissue in canine tissue. Our results indicate promise for the use of this device and an ICG contrast agent for tumor surgery in humans.

Our study was a successful proof-of-principle experiment for future use of indocyanine green contrast material and our handheld imaging device to guide tumor resection in humans.

Provenzale, J, Mohs, A, Mancini, M, Saba, C, Howerth, E, Cornell, K, Nie, S, Evaluation of a Hand-held Optical Imaging Device for Tumor Resection.  Radiological Society of North America 2012 Scientific Assembly and Annual Meeting, November 25 - November 30, 2012 ,Chicago IL. http://archive.rsna.org/2012/12030694.html