Parametric Response Mapping (PRM) for the detection of chronic lung injury in hematopoietic cell transplant recipients

ABSTRACT Chronic lung dysfunction (CLD) is a potentially severe complication of bone marrow transplantation (BMT), particularly prevalent in patients who develop chronic graft versus host disease (GVHD) post-BMT. The development of CLD is associated with a 20%-40% long term survival rate, with the majority of patients diagnosed with CLD only after they have developed significant (and potentially irreversible) lung pathology. There is a critical need to identify CLD earlier in its clinical course, at a stage when treatment may prove beneficial to patients. Parametric response mapping (PRM) is a novel computed tomography (CT) based methodology that when applied to standard CT scans is capable of identifying and quantifying a variety of lung disease subtypes in patients. This technology was initially developed at Michigan Medicine over 10 years ago for patients with chronic obstructive pulmonary disease (COPD). In both single center and multicenter trials, PRM has demonstrated efficacy for the identification of CLD in adult BMT patients. Our central hypothesis is that PRM can identify patients with chronic GVHD who are at high risk for the development of CLD. We propose an observational study to examine PRM in adult and pediatric subjects (≥ 3 years in age) with chronic GVHD post- BMT. To test our hypothesis, we have established 3 specific aims. In the first two, we will examine PRM values at two main time points: at the onset of chronic GVHD (Aim 1) and at the onset of CLD (Aim 2), with CLD defined by standard NIH criteria. In Aim 3, we will develop a machine learning model that will incorporate serologic and pulmonary function test (PFT) biomarkers with PRM values to develop a composite biomarker strategy. Upon completion of our current proposal, we will establish PRM as a predictor of CLD in adult BMT patients. In addition, the PRM, PFT, and serologic data obtained from pediatric BMT patients will provide a data bank for future research in this population. Historically, the ability to identify CLD in pediatric patients has been PFT-dependent, with PFTs technically challenging to conduct in young children. The use of PRM as a diagnostic indicator of CLD in pediatric BMT patients will be a major advance of this proposal. Finally, our proposal takes our PRM methodology forward in adults, allowing us to study the trajectory of lung disease in adult BMT recipients with chronic GVHD, and following the onset of CLD.