CP #1: Diffusion-MRI / PET prediction of breast cancer treatment response

Diffusion-MRI / PET prediction of breast cancer treatment response


High Interstitial Fluid Pressure (IFP), a common characteristic of breast tumors, has been shown to obstruct the delivery of therapeutic drugs to the tumor core [1-6]. While a number of therapies can reduce IFP and normalize tumor vessels, there is a small time window for vascular normalization that allows for effective drug delivery: when anti-angiogenic therapy is sustained for too long, for example, excessive vascular pruning also yields poor drug delivery. In phantom studies [7] and in a preclinical mouse mammary carcinoma study [8], we have been able to demonstrate that the pseudodiffusivity (Dp) associated with Intravoxel Incoherent Motion (IVIM) [9] is inversely correlated with invasively measured IFP. This noninvasive method of measuring IFP can be used to determine the “normalization” window and to monitor treatment response.


We will correlate IVIM parameters and invasively measured IFP in vivo in breast cancer patients. We will also characterize the relationship between proliferation markers derived from diffusion-weighted imaging (DWI) and from 18F-FLT PET in order to develop a model for prediction of breast cancer treatment response.


A cross-sectional analysis will be performed 150 patients with intermediate (ductal carcinoma in situ (DCIS), lobular carcinoma in situ (LCIS)) and malignant​(invasive ductal carcinoma (IDC), invasive lobular carcinoma (ILC)) breast tumor subtypes. Each patient will be imaged during MRI-guided biopsy, at which time IFP will also be sampled with an invasive wick-in-needle (WIN) approach.Tissue diffusivity (Dt), perfusion fraction (fp), and pseudodiffusivity (Dp) data will be generated with IVIM analysis. Dt will be compared with histological analysis from the tissue biopsy and fp will be compared with microvessel density. Another cohort of 50 breast cancer patients will be evaluated in a simultaneous MR/PET system. The IVIM parameters of the breast lesions will be measured in the same fashion and compared with the same histological markers as measured on tissue biopsy. 18F-FLT-PET will be used to map proliferation rate [10] and will be correlated with the Ki-67 histological proliferation marker. The 18F-FLT-PET uptake and Dt will be correlated for the entire group and separately based on histological grade or subtype. This group will undergo three MRI scans: (1) just prior to MR-guided biopsy, (2) following the beginning of treatment, and (3) just prior to final surgery. WIN IFP sampling will be performed at the time of biopsy and at final surgery. Pathologic analysis will be performed on biopsy and surgical specimens.

Push-Pull Relationship with TR&D Projects 1, 2, 3:

Our project team has a strong track record of collaboration and publication with BTRC staff [7,8,11-15].  Figure 1 illustrates new preliminary data acquired since the original​submission, and in that time we have also performed numerous breast imaging studies on our MR-PET scanner.

TR&D#1: Rapid image acquisition approaches developed in TR&D #1 will benefit the quality of DWI data collected, first by allowing single-shot acquisitions (echo-planar imaging (EPI) or turbo spin echo (TSE)[16]) with shorter echo-train lengths and therefore less spatial blurring, and eventually by allowing motion-robust continuous acquisition. Note that Cartesian trajectories and specific diffusion-focused reconstructions are now included in the revised TR&D#1, strengthening the push-pull relationship with this CP.

TR&D#2: Given the complex lesion texture often encountered in malignant breast lesions, high resolution may be vital to proper diagnosis and quantification.  Parallel transmission [17,18] will enable higher resolution by allowing reduced field of view acquisitions with correspondingly lower matrix size and reduced image blurring. This degree of freedom has only recently been considered in breast diffusion imaging [19], and our BTRC core is uniquely poised to apply the breakthroughs in RF control to breast oncological imaging.

TR&D#3: The MR/PET BTRC element will be key to Specific Aim 2 of the present proposal, which will compare markers of tumor cell proliferation from DWI and 18F-FLT-PET. Firstly, the Radiochemistry lab will be a key resource in managing the acquisition, dosage, and sensitivity of the FLT tracer. Also, given our goal of careful cross-correlation in the same lesion, accurate co-registration of the breast lesions in both modalities will be crucial to success, and studies in our breast cancer cohort will serve as a valuable test bed for MR-based PET corrections and eventually truly joint reconstruction of combined datasets.


Latest Updates

05/10/2019 - 21:50
05/10/2019 - 21:23

Philanthropic Support

We gratefully acknowledge generous support for radiology research at NYU Langone Health from:
• The Big George Foundation
• Raymond and Beverly Sackler
• Bernard and Irene Schwartz

Go to top