CT-Based Modeling of Bone Micro-Architecture and Fracture-Risk in COPD

The goal of this translational study is to establish a newly emerging CT-based tool for the characterization of changes in bone micro-architecture and assessment of their implications for fracture-risk in a population of COPD patients at risk for osteoporosis. The tool will be suitable and generalizable across emerging CT scanners from different vendors, and it will provide a more structurally-based assessment of osteoporosis and bone loss than is provided by simple bone density measures. The study will characterize the impact of different COPD-related factors on bone structure, and their implications for fracture-risk, leading to the development of a COPD-specific model for assessment of fracture-risk that will utilize patient-specific demographic, clinical and radiographic data, and CT BMD at the spine, as well as bone structural measures at the hip and/or ankle.

This translational study seeks to establish a Chronic Obstructive Pulmonary Disease (COPD)-specific fracture prediction model using the investigators unique computed tomography (CT)-based assessment of peripheral bone micro-architecture. Osteoporosis, a common comorbidity among patients with COPD, accelerates morbidity and mortality. The basis for this comorbidity is poorly understood, thus the need for characterizing the link between COPD-related factors and bone micro-architecture and their association to fracture-risk. Multiple COPD-related factors are associated with osteoporosis. Different COPD-related causes of bone loss may non-uniformly impact cortical and trabecular bone structures with varying mechanical consequences, reflective of divergent COPD-associated fracture-risk in individuals with similar bone mineral density (BMD). Little is known about this linkage, and the study goal is to fill this knowledge gap using a clinically suitable emerging CT-based tool for characterization of bone micro-architecture at peripheral sites. Specifically, this study will-(1) establish the generalizability of the investigators bone micro-architecture assessment applied to emerging low dose / high resolution CT scanners from different vendors; (2) assess its potential as compared to dual energy x-ray absorptiometry (DXA) to explain prevalent fractures and predict incident fractures among patients with COPD; (3) quantify the impact of different COPD-related factors on bone structures and their implications for fracture-risk; (4) identify COPD subtypes with rapid bone structural degeneration; and (5) develop a COPD-specific model for assessment of fracture-risk using patient-specific data. The study will take advantage of-(1) existing COPD patient cohorts with lung characterization at the University of Iowa (UI) and Columbia University (CU) representing a wide demographic range; (2) access to emerging CT scanners at both sites; and (3) unique image processing methodologies for quantifying three-dimensional bone structural metrics. The study will recruit 550 smokers with and without COPD from the UI and CU cohorts of the COPDGene and SPIROMICS studies. Smokers without COPD will comprise the control group for the study. At baseline and 3-year follow-up visits, the study team will collect-(1) data related to risk factors; (2) a lateral spine CT scout scan to assess vertebral fractures; (3) high resolution CT scans of the hip and ankle for computation of bone structural metrics; (4) whole-body, spine and hip DXA scans for evaluation of bone mineral density and body composition; and (5) DXA vertebral fracture assessment. This study will establish an emerging CT-based scanner-independent generalizable tool to assess bone response to different therapeutic interventions aimed at slowing or reversing bone loss, and possibly restoring bone structure, potentially leading to more patient-specific interventions. Also, this study seeks to explain the relationships among various COPD-related factors, bone structural changes and their implications for fracture-risk. Finally, a COPD-specific model for assessment of fracture-risk will be developed that will utilize patient-specific demographic, clinical and radiographic data, and CT BMD at the spine, as well as bone structural measures at the hip and/or ankle.

All subjects will participate in both baseline and three year follow-up visits and undergo identical interventions.

Smokers, defined has having at least a 10 pack-year lifetime history, with and without COPD will participate in the following interventions: Vital Signs Urine Pregnancy Test on woman of child bearing potential Pre- and Post-Bronchodilator Spirometry Questionnaires Blood Test for Vitamin D level, Hemoglobin A1c, and creatinine level Duel-energy X-ray absorptiometry scan (DXA) of the whole body, spine, and hip Duel-energy X-ray absorptiometry scan (DXA) for vertebral fracture assessment Multi-detector computed tomography (MDCT) of the hip and ankle

All subjects who completed a baseline visit will return for a follow-up visit and participate in the following interventions: Vital Signs Urine Pregnancy Test on woman of child bearing potential Pre- and Post-Bronchodilator Spirometry Questionnaires Blood Test for Vitamin D level, Hemoglobin A1c, and creatinine level Duel-energy X-ray absorptiometry scan (DXA) of the whole body, spine, and hip Duel-energy X-ray absorptiometry scan (DXA) for vertebral fracture assessment Multi-detector computed tomography (MDCT) of the hip and ankle

Heart rate, respirations, blood pressure, temperature, oxygen saturation arterial oxygen saturation (SaO2), height and weight

Method: Incident vertebral fracture cases between baseline and three-year follow-up visits will be determined by expert visual reading of baseline and three-year follow-up lateral spine CT scout scans.

Method: Prevalent fracture cases at the baseline visit will be determined using expert visual reading of the baseline lateral spine CT scout scan.

Method: Fragility fracture-risk will be computed using a CT-based model comprising of patient-specific demographic, clinical and radiographic data, CT bone mineral density at the spine, and CT bone micro-structural measures at the ankle.

Method: Fragility fracture-risk will be computed using a DXA-based model comprising of patient-specific demographic, clinical and radiographic data, and whole-body, spine, and hip DXA bone mineral density.

Method: CT measures of micro-structure at the ankle will be computed using a high-resolution ankle CT scan and computerized algorithms.

Method: CT measures of thoracic spine bone mineral density will be computed using a chest CT scan and computerized algorithms.

Inclusion Criteria: Current or former smoker, defined as having at least 10-pack year lifetime history Age: 45-90 Subjects will be stratified into groups based on COPD disease status: smokers with COPD and smokers without COPD. Smokers with COPD will be further stratified into three groups: low emphysema (< 3%), moderate emphysema (between 3 and 10%), and severe emphysema (> 10%). Exclusion Criteria: Pregnant or breast-feeding Metastatic Malignancy Currently receiving dialysis Any lower extremity fracture within the last year Any lower extremity fracture with hardware implant(s) History of bilateral tibia fractures

Clinical exam, questionnaire, and image-derived data will be made available to other researchers who request it formally. The data will be de-identified and linked only by code numbers. To further minimize the risk of breaches of confidentiality, we will establish data-use sharing agreements, and consider employment of electronic firewalls, locked storage facilities, password authentication of users, audit trails, and disaster prevention/recovery plans. We will stipulate that data are used only for research purposes, that proposed research using the data will be reviewed by an Institutional Review Board, and data will not be transferred to other users. The data would be available upon completion of necessary data management and statistical analyses. At that time, data could be provided on appropriate storage media via mail. Data files would be provided in SAS format, with PDF documentation files. We also will provide copies of questionnaires and images from which our data are derived.

The data would be available about 12 months after completion of the grant, with necessary data management and statistical analyses done by then. At that time, data could be provided on the appropriate storage media through the mail, with appropriate variable labels and other necessary documentation.

Data will only be available to researchers who formally request it and who have entered into a data-use agreement with the principal investigators.