Use of quantitative CT in the real-world evaluation of progressive pulmonary fibrosis; the Retrospective Evaluation in the US of e-Lung in PPF (REVISE PPF) multicentre observational study

Rationale

The Retrospective EValuation in the US of e-Lung in PPF (REVISE PPF) study is a multicentre retrospective study designed to evaluate the ability of a commercially available quantitative CT (qCT) algorithm (e-Lung, Brainomix, Oxford) to expedite the diagnosis of progressive pulmonary fibrosis (PPF) and to risk-stratify patients at increased risk of mortality and PPF.

‍

Methods

Patients from 3 specialist interstitial lung disease (ILD) centres were studied: University of Alabama at Birmingham (UAB), Weill Cornell Medicine (WCM) and University of Chicago (UChicago). e-Lung qCT biomarkers of ILD severity evaluated were the weighted reticulovascular score (WRVS), a measure of peripheral fibrosis, and ILDtotal disease extent (TDE). Comparison was made between time to local clinical diagnosis of PPF versus time to surpassing e-Lung thresholds for radiological progression on serial CT (WRVS ≥3% and TDE ≥1.5%). Associations between i) high risk baseline qCT fibrosis scores (WRVS≥15%), ii) eLung defined radiological progression and future PPF diagnosis and mortality were established.

‍

Results

509 patients were included of whom 267 (52%) had a diagnosis of PPF. In patients with serial CT, e-Lung identified radiologic progression earlier than the clinical PPF diagnosis in 78% (UAB), 83% (WCM) and 52% (UChicago) of patients at a mean time of 14 months, 28 months and 8 months respectively. Baseline WRVS≥15% was associated with future development of PPF in all 3 cohorts: UAB odds ratio (OR) 4.09 (95% CI: 2.07-8.11, p<0.0001), WCM OR 2.48 (95% CI: 1.26-4.86, p<0.01)), and UChicago OR 2.30 (95% CI: 1.16-4.57, p=0.017). Patients with WRVS increase ≥3% on serial CT were at increased risk of death: HR 6.92, (95% CI: 2.26 – 21.24, p < 0.001) in pooled UAB / WCM and HR 5.86, (95% CI: 3.45 - 9.93, p < 0.001) in UChicago cohorts. Patients with TDE increase ≥1.5% were also at increased risk of death: HR 16.41 [95% CI 3.36 – 80.20; p<0.001] in pooled UAB / WCM and HR 4.75 [95% CI 2.90 – 7.78; p < 0.0001] in UChicago cohorts

‍

Conclusion

In diverse cohorts of patients with ILD from three specialist US centres, e-Lung was able to stratify patients at risk of future progression from a baseline CT and identified patients with radiologic evidence of progressive pulmonary fibrosis up to 28 months earlier than local clinical diagnoses. e-Lung qCT biomarkers used to identify earlier progression were also linked to increased mortality risk. These data support the prospective evaluation of e-Lung in routine ILD clinical practice.

‍