Predicting breast cancer response to neoadjuvant chemotherapy based on tumor vascular features in needle biopsies
Terisse A. Brocato, Ursa Brown-Glaberman, Zhihui Wang, Reed G. Selwyn, Colin M. Wilson, Edward F. Wyckoff, Lesley C. Lomo, Jennifer L. Saline, Anupama Hooda-Nehra, Renata Pasqualini, Wadih Arap, C. Jeffrey Brinker, Vittorio Cristini
Terisse A. Brocato, Ursa Brown-Glaberman, Zhihui Wang, Reed G. Selwyn, Colin M. Wilson, Edward F. Wyckoff, Lesley C. Lomo, Jennifer L. Saline, Anupama Hooda-Nehra, Renata Pasqualini, Wadih Arap, C. Jeffrey Brinker, Vittorio Cristini
View: Text | PDF
Resource and Technical Advance Oncology

Predicting breast cancer response to neoadjuvant chemotherapy based on tumor vascular features in needle biopsies

Abstract

In clinical breast cancer intervention, selection of the optimal treatment protocol based on predictive biomarkers remains an elusive goal. Here, we present a modeling tool to predict the likelihood of breast cancer response to neoadjuvant chemotherapy using patient-specific tumor vasculature biomarkers. A semiautomated analysis was implemented and performed on 3990 histological images from 48 patients, with 10–208 images analyzed for each patient. We applied a histology-based mathematical model to 30 resected primary breast cancer tumors and then evaluated a cohort of 18 patients undergoing neoadjuvant chemotherapy, collecting pre- and posttreatment pathology specimens and MRI data. We found that core biopsy samples can be used with acceptable accuracy to determine histological parameters representative of the whole tissue region. Analysis of model histology parameters obtained from tumor vasculature measurements, specifically diffusion distance divided by the radius of the drug-delivering blood vessel (L/rb) and blood volume fraction (BVF), provides a statistically significant separation of patients obtaining a pathologic complete response (pCR) from those who do not. With this model, it is feasible to evaluate primary breast tumor vasculature biomarkers in a patient-specific manner, thereby allowing a precision approach to breast cancer treatment.

Authors

Terisse A. Brocato, Ursa Brown-Glaberman, Zhihui Wang, Reed G. Selwyn, Colin M. Wilson, Edward F. Wyckoff, Lesley C. Lomo, Jennifer L. Saline, Anupama Hooda-Nehra, Renata Pasqualini, Wadih Arap, C. Jeffrey Brinker, Vittorio Cristini

×

Figure 6

Diffusion analysis workflow.

Options: View larger image (or click on image) Download as PowerPoint
Diffusion analysis workflow. (A) Original CD34-stained histology grid be...
(A) Original CD34-stained histology grid before any processing. (B) Same tissue region as in A, but with the outer inked portion removed due to the increased likelihood of false positives on the perimeter of core biopsy samples (the University of New Mexico Pathology Department inks tissue cores for quality purposes). (C) Computerized version of B with differentiation between tissue CD34– (blue), vasculature CD34+ (red), and non-tissue regions (gray). (D) Diffusion analysis of image in C, which was performed by code developed in MATLAB. Parameters measured were: blood vessel radius (rb), blood volume fraction (BVF), and diffusion distance (L). Blood vessels are outlined in red, and the total area of blood vessels in a tissue region is the BVF. Blue shows the central long axis of each vessel (multiple vessel radius measurements were taken perpendicular to this axis). An average of all vessel radii in each image analyzed is taken to be rb (μm). The black lines discretize the image into regions defined by having the closet proximity to the enclosed vessel; then the distance from each black boundary to the blood vessel boundary (red) is measured; and all distances averaged is the diffusion penetration distance (L) measured in μm. White is the tumor tissue region, all of which is considered for analysis. Green is the background/non-tissue region not considered for analysis.