Specific Aims

Myelofibrosis (MF) is a chronic, ultimately fatal myeloproliferative neoplasm caused by genetic mutations in hematopoietic stem cells that cause systemic inflammation and progressive fibrosis of bone marrow, disrupting normal architecture and composition of cells. Biopsy remains the only method to assess bone marrow, the primary site of disease in MF. Bone marrow biopsies inherently suffer from sampling error, as the technique only analyzes millimeter amounts of tissue from one anatomic site (iliac crest). Biopsy cannot assess anatomic heterogeneity of disease, a common feature of MF based on autopsies. In patients with extensive fibrosis, bone marrow biopsies frequently recover no tissue (“dry tap”), leaving physicians without critical information about bone marrow composition and severity of disease needed to assess response to therapy. Beyond providing more comprehensive information about bone marrow throughout the skeleton, quantitative imaging biomarkers have the exciting potential to replace an invasive, painful procedure with a non-invasive test.

Oncologists rarely use imaging to analyze bone marrow in MF and other hematologic malignancies because current methods produce largely qualitative rather than quantitative, reproducible data for bone marrow composition. Deficiencies in current methods essentially preclude development of standardized imaging biomarkers that facilitate drug testing in pre-clinical models and clinical trials. Our resource will validate MRI methods that directly address key parameters about disease status and response to therapy in MF: 1) bone marrow composition (quantitative Dixon technique for proton density fat-fraction (PDFF)); 2) cellularity and replacement of normal bone marrow cells and bone trabecula (mobility of water (diffusion, DWI); 3) fibrosis by magnetization transfer (MT); and 4) spleen volume by abdominal 3D imaging. These clinically-approved MRI sequences are offered on scanners from all major vendors for pre-clinical and clinical imaging. After rigorously validating scanning protocols using appropriate phantoms, test/retest procedures, and histology, we will analyze bone marrow composition and architecture in two co-clinical trials: 1) ruxolitinib, the only FDA-approved drug for MF; and 2) a promising investigational therapy from our active clinical trials pipeline. For both mouse and human trials, we will further validate imaging data with clinical parameters, genomics, and histology.

Building on our experience in the Quantitative Imaging Network, we will establish standard operating procedures (SOPs) for mouse models, image acquisition, validation with phantoms, and image analysis. To disseminate these methods, we will make all SOPs and image processing code available through our Resource website. We also will deposit de-identified imaging data sets into the TCIA to advance quantitative bone marrow MRI for co-clinical trials. While our proposed Resource centers on MF, we anticipate our procedures for quantitative MRI of bone marrow will extend to co-clinical trials in multiple other hematologic cancers.

Aim 1. Establish optimized SOPs for quantitative MRI of bone marrow. We will establish standardized workflows for acquisition and quantitative analysis of MRI data needed to analyze critical markers of bone marrow disease in MF: 1) Dixon technique (proton density fat fraction for bone marrow composition); 2) Diffusion (DWI) (cellularity and replacement of normal bone marrow); and 3) MT (bone marrow fibrosis). We also will standardize the conventional measurement of spleen volume. To optimize reproducibility of imaging data, we will validate each workflow with imaging phantoms and test/retest studies in mice and patients. From these studies, we will formulate requirements for maximal accuracy and precision of MRI methods and quantitative data outputs. We will validate imaging data with histology in mouse models that reproduce genetic drivers in patients with MF.

Aim 2. Implement optimized imaging protocols in co-clinical trials for MF. We will quantify response to treatment in two separate co-clinical trials. Trial I will use the JAK1/2 inhibitor ruxolitinib, the standard-of-care drug for MF. In Trial II, we will use the most promising investigational compound available in our pipeline of clinical trials in MF. We will use a validated mouse model in which we transduce hematopoietic stem cells with main driver mutations present in patients with MF. We will quantify changes in bone marrow composition and architecture over the course of therapy, using the same MRI techniques for both mice and humans. We will correlate imaging data with genomics, established clinical parameters, and histology. From these studies, we will determine to what extent the pre-clinical trial predicts response to therapy in patients.

Aim 3. Establish an internet-accessible research resource with workflow and procedures for imaging protocols, image processing, and quantitative datasets from mouse and human studies. A central objective our Resource is to disseminate bone marrow MRI as a quantitative imaging method for co-clinical trials at multiple sites. We will accomplish this goal through publications and a user-friendly website with SOPs for imaging protocols, analysis code, and animal procedures. To make imaging data available to the research community, we will deposit curated datasets in the TCIA.