Background and Objective: Patients with IPF show variable survival duration. We assessed the use of various combinations of blood, lung function and imaging measures to predict mortality in IPF.
Methods: In an Oxford IPF cohort (n=75) with CALIPER-compatible CTs we determined annualised change (?) in forced vital capacity (FVC), total CT fibrosis score (TLF%), and CALIPER-measured lung volume (CTvol) over a 3-year period. Different combinations of these metrics were compared against mortality using Cox regression. We compared concordance indexes (c-statistic) of different Cox models to determine the most predictive and discriminative combination for mortality. Multivariate models were adjusted for age, gender, baseline TLF%, antifibrotic use and either included* or excluded? leukocytes.
Results: Combined measurements of ?CTvol, ?FVC and ?TLF% were more predictive of mortality than single measures. Each model was further enhanced by addition of leukocytes (continuous variables, Table 1).
| Measure of progression | HR | p value | C-statistic |
| ?CTvol>10%? | 2.95 | 0.013 | 0.747 |
| ?CTvol>10%* | 3.86 | 0.005 | 0.792 |
| ?TLF>10%? | 2.37 | 0.067 | 0.760 |
| ?TLF>10%* | 2.02 | 0.143 | 0.770 |
| ?FVC>10%? | 6.15 | <0.001 | 0.811 |
| ?FVC>10%* | 10.71 | <0.001 | 0.866 |
| ?FVC>5% or ?CTvol>5% or ?TLF%>5%? | 4.24 | 0.048 | 0.728 |
| ?FVC>5% or ?CTvol>5% or ?TLF%>5%* | 7.09 | 0.014 | 0.833 |
| ?FVC>10% or ?CTvol>10% or ?TLF%>10%? | 7.14 | <0.001 | 0.805 |
| ?FVC>10% or ?CTvol>10% or %?TLF>10%* | 8.22 | <0.001 | 0.851 |
Conclusion: In our IPF cohort, adding blood leukocytes to a composite of disease progression measures enhanced prediction of mortality. The findings also support potential role for monocytes, neutrophils & lymphocytes in pathobiology of IPF.