Sessions: Poster Forum 2 - Monday - 12:00 pm
Date: November 13 - Monday
Time: 12:00 pm - 01:00 pm

Sub Topic | Secondary Topic: Biotherapeutics and Biotechnology - Biomarkers | Measurement

Authors: Ryan Williams, Memorial Sloan Kettering Cancer Center (Main Author, Presenting Author); Christopher Lee, Memorial Sloan Kettering Cancer Center; Thomas Galassi, Weill Cornell Medicine; Jackson Harvey, Weill Cornell Medicine; Rachel Leicher, Weill Cornell Medicine; Maria Sirenko, Memorial Sloan Kettering Cancer Center; Madeline Dorso, Weill Cornell Medicine; Janki Shah, Memorial Sloan Kettering Cancer Center; Narciso Olvera, New York University; Fanny Dao, New York University; Douglas Levine, New York University; Daniel Heller, Memorial Sloan Kettering Cancer Center

Presenting Author: Ryan Williams

Purpose: Ovarian cancer accounts for 238,000 new diagnoses and 151,000 deaths per year worldwide. Late-stage diagnoses occur in over 60% of patients, higher than any other form of cancer. However, five-year survival is 92% in patients diagnosed at early stages of disease. Unfortunately, current screening methods have a high false-positive rate and poor sensitivity for early stages of disease, none of which reduce mortality. Single-walled carbon nanotubes (SWCNT) exhibit optical properties well-suited for biosensing applications. Nanotubes exhibit tissue-penetrant near-infrared (NIR) photoluminescence and unique photostability, allowing for repeated, long-term measurement. SWCNT are also sensitive to their local environment and photoluminescent emission can undergo modulation of the optical bandgap, causing a shift in wavelength. Prior work has also shown nanotube fluorescence in live mice and can detect analytes in complex biological environments. These characteristics allow nanotubes to optically transduce analyte concentrations over time in vivo.

Methods: We engineered an ovarian cancer biomarker sensor by non-covalently attached an antibody to the polymer wrapping of SWCNT. Nanosensor selectivity and specificity were characterized in vitro via NIR spectroscopy. We then measured the biomarker ex vivo in patient serum and ascites samples by NIR hyperspectral microscopy. We developed an implantable device by encapsulating the sensor in a semipermeable, biocompatible membrane. This implantable sensor was initially surgically immobilized in the peritoneal cavity of healthy mice, where we injected control protein or the specific biomarker. Probe-based NIR spectroscopy and whole-animal hyperspectral imaging was used to obtain the sensor response. Finally, mice bearing four orthotopic models of ovarian cancer were implanted with the sensor device and the response of the sensor was measured.

Results: The sensor exhibited exquisite specificity and low-nanomolar selectivity for the ovarian cancer biomarker. It also differentiates the serum and ascites of patients with ovarian cancer versus benign conditions based on biomarker presence. In vivo, we quantitatively measured exogenous biomarker injected into healthy mice while the sensor did not respond to control protein. Finally, we differentiated between two mouse models that produce the specific biomarker versus two that do not, demonstrating the ability to specifically detect relevant disease biomarkers.

Conclusion: We developed an implantable sensor for an ovarian cancer biomarker that quantitatively measures protein injected into mice. Further, this device can detect the biomarker produced by orthotopic ovarian cancer. We expect to further test the in vivo lifetime and range of this sensor device towards clinical translation. We hope to increase early detection rates, enhancing patient prognosis by translation of this device to detect ovarian cancer at the site of early disease. This is an Encore Presentation from Biomedical Engineering Society 2017.

See attached abstract pdf for images.

Abstract Link: