The Role of Quantitative CT and Radiomic Biomarkers for Precision Medicine in Pulmonary Fibrosis

Description

Idiopathic pulmonary fibrosis (IPF) remains deadly despite two FDA-approved therapies. Forced vital capacity (FVC), a one-dimensional assessment of lung function that requires three effort-dependent and error-prone maneuvers, is the standard for evaluating disease severity and monitoring progression. FVC indirectly measures disease activity and is thus insensitive to subtle change. These limitations hamper therapeutic trials. The Gender, Age, and Physiology (GAP) score improves on FVC alone and is the most used scoring model for prognostication, but gender and age aren’t influenced by treatment. Modifiable intermediate molecular markers and other metrics for assessing disease severity and progression remain unmet needs for aiding drug development and clinical decision-making. Computed tomography (CT) captures morphologic patterns and the extent of fibrosis noninvasively. Advances in quantitative CT enable objective detection and quantitation of anatomy, and highly dimensional image features, often termed radiomic, can identify sub-visual characteristics. The investigators seek to evaluate radiomic features alone and in conjunction with other disease dimensions for prognostication and response to treatment in IPF.

The investigators overall objectives are to identify and validate radiologic features, such as total extent of lung fibrosis, for disease activity and intermediate response to therapy, and understand where to position these powerful markers. The investigators hypothesize that DTA scores will contribute to prediction of disease progression and that molecular markers will enhance that performance.

Aim 1: The investigators will validate quantitative CT and radiomic markers for disease progression by independent replication in separate cohorts. The investigators hypothesize that quantitative CT markers will predict disease progression in UVA/Chicago cohorts. Baseline and subsequent CT scans have been voluntarily collected in many PFF-PR cases. The investigators propose collection of 1-year HRCTs in UVA/Chicago participants to evaluate: a) the prognostic value of baseline quantitative CT and radiomic markers (i.e. DTA) in predicting time to progression defined as either 10% relative decline in FVC, lung transplant, or death from any cause, b) associations between changes in CT biomarkers on sequential CT and changes in 1-year FVC and DLCO, and c) change in CT associated with drug treatment. This aim will establish the relative and synergistic value of CT to established physiologic markers.

Aim 2: The investigators will determine if candidate genetic variants for IPF susceptibility and survival are associated with the DTA score and improve predictive performance for survival. The investigators hypothesize that variants in MUC5B, TOLLIP, and Telomere lengths (TL) will enhance DTA fibrosis score associations with progression-free survival in IPF. The investigators will perform a cross-sectional analysis of PFF-PR cases comparing quantitative CT and radiomic markers at baseline with and without “at risk” genotypes for association with severity and progression (decline in FVC over time). This will ascertain what markers improve performance of the DTA fibrosis extent scores using Cox regression analysis and accuracy metrics from Aim 1. Findings will be replicated in UVA/Chicago cohort and in the prospective PRECISIONS cohort. This aim will establish the additive value of genetic markers.

Aim 3: The investigators will assess whether DTA and radiomic markers are additive/synergistic with plasma protein and blood transcriptome markers for disease progression. The investigators hypothesize that selected protein and transcriptomic markers will prove additive to DTA fibrosis extent for prediction of progression-free survival whereas other markers correlated with DTA will not. The investigators have chosen published markers from a 4-protein panel signature, along with CCL18, as examples, given their current level of replication and promise. The investigators will also include a 25-gene FVC predictor for disease progression. Similar analyses, as outlined in Aims 1 and 2, will determine their additive information value.