TL;DR
This study introduces and validates a computer-based method for quantifying pleuroparenchymal fibroelastosis (PPFE) on CT scans in patients with idiopathic pulmonary fibrosis (IPF). This automated approach proves more effective than visual assessment in predicting patient survival and lung function decline, suggesting PPFE is a distinct prognostic marker in IPF.
What Problem Does This Paper Address?
Idiopathic pulmonary fibrosis (IPF) is a devastating lung disease with a poor prognosis. While various imaging patterns are associated with IPF, the prognostic significance of pleuroparenchymal fibroelastosis (PPFE) – a specific fibrotic pattern predominantly affecting the upper lobes – in IPF patients has not been thoroughly investigated. Existing methods for assessing PPFE, primarily visual scoring by radiologists, are subject to inter-observer variability. This limits their reliability for clinical decision-making and patient stratification in clinical trials. The authors aim to address this by developing and validating a computer-based method to quantify PPFE extent and assess its prognostic value in IPF.
Methods
The study employed a retrospective observational design using two independent cohorts of IPF patients: a derivation cohort and a validation cohort.
- Visual CT Assessment (vPPFE): Radiologists visually scored the extent of PPFE in the upper lobes using a 3-point scale (absent, moderate, marked).
- Computer-based CT Evaluation (cPPFE): A novel automated quantification method was developed. This involved lung segmentation followed by threshold-based isolation of PPFE on the visceral pleural surface within the upper lung zones. The method specifically focused on the peripheral lung surface to avoid conflation with typical IPF fibrosis.
- Survival Analysis: Linear mixed-effects (LME) models were used to assess the 1-year decline in forced vital capacity (FVC). Cox regression models were employed to analyze mortality. Multivariable models were adjusted for key clinical variables like age, gender, smoking history, antifibrotic treatment, and diffusion capacity for carbon monoxide (DLco).
- Prognostic Discrimination: Kaplan-Meier curves and receiver operating characteristic (ROC) analyses were used to compare the prognostic performance of visual and computer-based PPFE scores.
Datasets
The study utilized data from three medical centers:
- Derivation Cohort: Patients presenting to Ege University Hospital, Izmir, Turkey (n=142).
- Validation Cohort: Patients presenting to St Antonius Hospital, Nieuwegein, Netherlands (n=145), and University Hospital Southampton NHS Foundation Trust, Southampton, UK (n=122, combined for analysis).
No new benchmark datasets were introduced; the study used existing patient data collected for clinical purposes.
Key Insights and Ideas
- PPFE is a Common and Distinct Endotype: PPFE is present in a significant proportion of IPF patients (49% in the derivation cohort, 72% in the validation cohort). Crucially, its prognostic impact appears independent of the severity of general interstitial lung disease (ILD) on CT, suggesting it represents a distinct pathological process or endotype within IPF.
- Computerized Quantification Outperforms Visual Scoring: The automated cPPFE method identified more patients with severe outcomes than visual scoring. This highlights the potential for objective, reproducible quantification in improving prognostic accuracy.
- PPFE Predicts Worse Outcomes: Both visual and computer-based PPFE assessments were independently associated with accelerated FVC decline and increased mortality in IPF patients.
- Upper Lobe Focus is Sufficient: The observation that PPFE never occurred in the middle or lower lobes without being present in the upper lobes justified focusing the scoring on the upper lung zones, simplifying the assessment.
- Potential for Clinical Trial Stratification: The findings suggest that PPFE quantification could be valuable for stratifying patients in clinical trials, particularly for antifibrotic therapies, to ensure more homogeneous patient groups and clearer interpretation of treatment effects.
Key Results
- Prevalence: PPFE was identified in 49% (derivation) and 72% (validation) of IPF patients.
- FVC Decline: Marked vPPFE was associated with a significant 1-year FVC decline (e.g., 18.3% in the derivation cohort). Continuous cPPFE was also strongly associated with FVC decline.
- Mortality: Marked vPPFE was associated with significantly increased mortality (e.g., hazard ratio [HR] of 7.70 in the derivation cohort). A cPPFE threshold of >2.5% was also a strong predictor of mortality (e.g., HR of 5.26 in the derivation cohort).
- Discriminative Power: The cPPFE threshold of ≥2.5% demonstrated better discrimination of poor outcomes at 2 and 3 years compared to marked vPPFE. Continuous cPPFE scores showed higher AUC values than vPPFE scores in survival prediction.
- Antifibrotic Treatment: In the combined IPF cohort, antifibrotic treatment was associated with a lower risk of mortality, even in patients with PPFE.
Limitations and Future Work
- Single Reviewer for Visual Scoring: The study acknowledges the lack of a second reviewer for visual PPFE scoring, although previous studies have shown good inter-observer agreement for PPFE.
- Lack of Histopathological Confirmation: The study relied on CT imaging for PPFE diagnosis, as obtaining histological confirmation in a large cohort is challenging and carries risks. However, established CT criteria and guidelines were used.
- Advanced IPF Cohort: The study population may have had relatively advanced IPF, and future work should investigate if the prognostic impact of PPFE remains consistent in patients with less severe disease.
- Novel Therapeutic Targets: The authors suggest that the distinct pathophysiological mechanisms potentially involved in PPFE progression may warrant the development of novel therapeutic targets beyond current antifibrotic medications.
- Longitudinal Imaging: The potential for cPPFE to sensitively detect disease progression on longitudinal imaging warrants further investigation.
Summary updated