Quantitative CT Imaging Analysis for Chronic Lung Allograft Dysfunction in Single Lung Transplant Recipients

Abstract

Purpose Chronic lung allograft dysfunction (CLAD) limits long term survival after lung transplantation (LTx). Distinction of CLAD subtypes relies on total lung capaciy (TLC) testing which is more cumbersome and less frequently performed compared to FEV1. Also, diagnosing CLAD is more difficult in single compared to double LTx patients due to the contribution of both the native lung and the allograft to FEV1 changes. CT can evaluate the lungs individually and could represent an alternative to TLC testing. Our aim was to investigate the utility of quantitative CT (QCT) in CLAD single LTx patients. Methods Of 138 patients who underwent single LTx 2006-2014, 35 (16 right, 19 left) developed CLAD, were followed for at least 1 year after CLAD onset, and had CT scans at baseline and at CLAD onset. CT lung volume (total volume of both native and allograft lungs) was compared with TLC values. The area under the ROC curve (AUC) was used to compare patients with shorter survial (< 1 year) to those with longer survival (≥ 1 year) after CLAD onset using QCT analysis of allograft lung: volume as a % of baseline, functional principal component analysis (FPCA) using hounsfield units (HU) density histogram at CLAD onset, mean HU at CLAD onset, and HU as a % of baseline. FPCA is a statistical approach that was used for dimension reduction of HU density histogram from QCT data. Results 31 patients had TLC at baseline and 29 at CLAD onset. CT volumes were highly correlated with TLC at both time points (p < 0.0001). Average FEV1(% of baseline) for shorter survival patients (n=12) and longer survival patients (n=23) was similar (74±5%, 75±5%, p=0.67). AUC for differentiating between the 2 groups using QCT on the allograft lung were FPCA = 0.69[95% CI0.51-0.83], p=0.047, mean HU = 0.67[0.50-0.82], p=0.11, volume (% of baseline) = 0.64[0.46-0.80], p=0.24, and HU (% of baseline) = 0.61[0.43-0.77], p=0.33. Conclusion QCT may be a useful radiologic tool for discriminating CLAD prognosis not only for double LTx but also for single LTx. Chronic lung allograft dysfunction (CLAD) limits long term survival after lung transplantation (LTx). Distinction of CLAD subtypes relies on total lung capaciy (TLC) testing which is more cumbersome and less frequently performed compared to FEV1. Also, diagnosing CLAD is more difficult in single compared to double LTx patients due to the contribution of both the native lung and the allograft to FEV1 changes. CT can evaluate the lungs individually and could represent an alternative to TLC testing. Our aim was to investigate the utility of quantitative CT (QCT) in CLAD single LTx patients. Of 138 patients who underwent single LTx 2006-2014, 35 (16 right, 19 left) developed CLAD, were followed for at least 1 year after CLAD onset, and had CT scans at baseline and at CLAD onset. CT lung volume (total volume of both native and allograft lungs) was compared with TLC values. The area under the ROC curve (AUC) was used to compare patients with shorter survial (< 1 year) to those with longer survival (≥ 1 year) after CLAD onset using QCT analysis of allograft lung: volume as a % of baseline, functional principal component analysis (FPCA) using hounsfield units (HU) density histogram at CLAD onset, mean HU at CLAD onset, and HU as a % of baseline. FPCA is a statistical approach that was used for dimension reduction of HU density histogram from QCT data. 31 patients had TLC at baseline and 29 at CLAD onset. CT volumes were highly correlated with TLC at both time points (p < 0.0001). Average FEV1(% of baseline) for shorter survival patients (n=12) and longer survival patients (n=23) was similar (74±5%, 75±5%, p=0.67). AUC for differentiating between the 2 groups using QCT on the allograft lung were FPCA = 0.69[95% CI0.51-0.83], p=0.047, mean HU = 0.67[0.50-0.82], p=0.11, volume (% of baseline) = 0.64[0.46-0.80], p=0.24, and HU (% of baseline) = 0.61[0.43-0.77], p=0.33. QCT may be a useful radiologic tool for discriminating CLAD prognosis not only for double LTx but also for single LTx.