Background: DOSI is an experimental imaging technique that employs risk-free near-infrared light for quantitative measurements of breast tissue perfusion, metabolism, and composition without exogenous contrast. A multi-center ACRIN 6691 study was designed to evaluate whether changes from baseline to mid-therapy in a DOSI-derived Tissue Optical Index (TOI) could predict pathologic complete response (pCR) in breast cancer neoadjuvant chemotherapy (NAC).

Methods: DOSI instruments were constructed at the University of California, Irvine and delivered to 6 participating sites. Bedside measurements were conducted by scanning a handheld probe over a region of interest (up to ∼10 x 10 cm2) on both breasts. Instruments were standardized and validated using a common set of tissue simulating phantoms and protocols. DOSI-derived near-infrared absorption and scattering spectra (650-1000 nm) were used to calculate the tissue concentration of oxy- and deoxyhemoglobin (ctO2Hb, ctHHb), water (ctH2O), %lipid and the tissue optical index (TOI=ctHHb x ctH2O/%lipid) in each probe location. Baseline to mid-therapy changes in the tumor to normal (T/N) TOI ratio were evaluated from DOSI images as the primary imaging endpoint for predicting clinical outcome (pCR). 60 female breast cancer patients (ages 28-69 years, mean 48.9±11), with locally advanced disease (tumors >2cm) were enrolled across the 6 institutions. DOSI measurements were performed at baseline, during the first week of therapy, at midpoint, and at the completion of NAC. Logistic regression was used to assess the association between pathologic complete response (pCR) and % change in T/N TOI from baseline to mid-therapy. In addition, area under the receiver operating characteristic curve (ROC AUC) and its corresponding 95% confidence interval were calculated.

Results: Of the 34 participants (mean age 48.4 ± 10.7) with complete and evaluable data, 10 (29%) achieved pCR as determined by central pathology review. The % change in TOI ratio ranged from -82% to 321%, with a median of -36%. Using -40% as a threshold, we found that subjects in the group with a 40% or more decrease in T/N TOI were more likely to be pCR (p=0.0586, OR=4.667, 95% CI: 0.945 to 23.038). The % change in TOI ratio from baseline to mid-therapy has an AUC of 0.604 (95% CI: 0.394 to 0.814) to distinguish pCR from non-pCR.

Conclusions: DOSI has been successfully implemented in a multi-center setting and changes in T/N TOI are a promising predictor of NAC clinical outcome (pCR). A larger study population is needed to fully assess the utility of TOI and other DOSI imaging endpoints for guiding therapies and predicting NAC response in individual subjects.

ACRIN receives funding from the NCI through U01 CA079778 and U01 CA080098.

Citation Format: Bruce J Tromberg, Zheng Zhang, Anais Leproux, Thomas D O'Sullivan, Albert E Cerussi, Philip Carpenter, Rita Mehta, Darren Roblyer, Wei Yang, Keith D Paulsen, Brian W Pogue, Shudong Jiang, Peter Kaufman, Arjun G Yodh, So-Hyun Chung, Mitchell Schnall, Brad Snyder, Nola Hylton, David A Boas, Stefan A Carp, Steven J Isakoff, David Mankoff. Predicting pre-surgical neoadjuvant chemotherapy response in breast cancer using diffuse optical spectroscopic imaging (DOSI): Results from the ACRIN 6691 study [abstract]. In: Proceedings of the Thirty-Seventh Annual CTRC-AACR San Antonio Breast Cancer Symposium: 2014 Dec 9-13; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2015;75(9 Suppl):Abstract nr S4-04.