Parametric Response Mapping (PRM) for the Detection of Chronic Lung Injury in Hematopoietic Cell Transplant Recipients

The time to develop CLD or death over 12 months of follow-up in both pediatric and adult subjects will be determined. we will classify subjects into comparison groups of (1) high PRMfSAD (>30%), (2) high PRMPD (>40%) or (3) neither high PRMfSAD nor high PRMPD (PRMNorm group). The two primary analyses are to compare 1-year CLD-free survival between the high PRMfSAD and PRMNorm groups and separately between the high PRMPD and PRMNorm groups, using two-sided, two-sample logrank tests with an overall 5% type I error, adjusted for 2 comparisons using the Bonferroni method

Trajectory of FEV1 decline over a 12-month period in patients with newly diagnosed CLD. To evaluate PRM as a predictor of "lung function decline" in patients with an established diagnosis of CLD.

Pediatric subjects will be enrolled into one of 2 strata (PFT+ vs PFTnull), PRM imaging with whole lung volumetric CT scans will be performed at baseline, using low dose CT techniques in children. Strata 1 (PFT+), the occurrence of CLD will be defined using NIH Consensus Criteria for BOS, or ISHLT criteria for RLD. Stratum 2 (PFTnull), radiographic findings of obstructive or restrictive lung disease on CT, plus clinical symptoms (dyspnea, or hypoxia) will be required to establish a diagnosis of CLD, in lieu of PFTs. PRM profile will be reported as percent functional small airway disease (PRMfSAD), emphysema (PRMEmph), parenchymal lung disease (PRMPD), and normal lung parenchyma (PRMNorm), in relation to the total lung volume.

Subjects will be enrolled into one of 2 strata (PFT+ vs PFTnull), PRM imaging with whole lung volumetric CT scans will be performed at baseline.. For patients in strata 1 (PFT+), the occurrence of CLD will be defined using NIH Consensus Criteria for BOS, or ISHLT criteria for RLD. For patients in stratum 2 (PFTnull), radiographic findings of obstructive or restrictive lung disease (i.e. air trapping, bronchiectactic changes, interstitial fibrosis) on CT, plus clinical symptoms (dyspnea, or hypoxia) will be required to establish a diagnosis of CLD, in lieu of PFTs. Histograms of PRMfSAD, PRMPD, PRMEmph, and PRMNorm values will be displayed by pediatric stratum and overall, with means ± standard errors superimposed on the plots. PRM profile will be reported as percent functional small airway disease (PRMfSAD), emphysema (PRMEmph), parenchymal lung disease (PRMPD), and normal lung parenchyma (PRMNorm), in relation to the total lung volume.

We will have approximately 90% power to detect correlations > 0.47 using 2-sided 1-sample correlation tests. This is a conservative estimate involving only the adult participants, with power increasing once pediatric participants are included in the analysis. In a secondary analysis we will use linear mixed effects models to study FEV1 predicted trajectories after CLD onset, as a function of PRM values. We will explore interactions between PRM values and time post-CLD when modeling FEV1 predicted values over time. PRM profile will be reported as percent functional small airway disease (PRMfSAD), emphysema (PRMEmph), parenchymal lung disease (PRMPD), and normal lung parenchyma (PRMNorm), in relation to the total lung volume.

We will have approximately 90% power to detect correlations > 0.47 using 2-sided 1-sample correlation tests. This is a conservative estimate involving only the adult participants, with power increasing once pediatric participants are included in the analysis. In a secondary analysis we will use linear mixed effects models to study FEV1 predicted trajectories after CLD onset, as a function of PRM values. We will explore interactions between PRM values and time post-CLD when modeling FEV1 predicted values over time. PRM profile will be reported as percent functional small airway disease (PRMfSAD), emphysema (PRMEmph), parenchymal lung disease (PRMPD), and normal lung parenchyma (PRMNorm), in relation to the total lung volume.