Comparison of Pi10 and PiSlope calculation methods and association with lung function

The average wall thickness of a theoretical airway with a lumen perimeter of 10 mm (Pi10) and the slope of the luminal perimeter of airways against their wall thickness (PiSlope) are CT-derived biomarkers of airway remodeling in chronic obstructive pulmonary disease (COPD). However, numerous calculation methods for Pi10 have been used across studies, leading to substantial variability in Pi10 values. This study evaluated reliability among 10 Pi10/PiSlope methodologies and their associations with lung function.

CanCOLD participants underwent full-inspiration CT scans; CT airway segmentation was performed using VIDA software. Ten literature-based methods were used to calculate Pi10 (regression-derived wall thickness at 10 mm perimeter) and PiSlope (slope of wall area/perimeter plot). The 10 methods varied in their approaches: some focused on specific airway sizes (Patel, Van Tho, Nakano, Telenga), others limited measurements to certain airway generations (Gietema, Park, Dudurych) or covered broader airway size ranges and generations (Bhatt, Jobst, Smith). Reliability was assessed via intraclass correlation (ICC); lung function associations were assessed using regression.

Pi10 and PiSlope were calculated for 1351 participants (67 ± 10 years; 550 women). Pi10 and PiSlope  measurement methods with higher airway counts (Patel, Nakano, Smith, Jobst, Bhatt) showed excellent reliability (ICC > 0.90, p < 0.05). Compared to Pi10, PiSlope demonstrated stronger, consistent negative associations with FEV1, FEV1/FVC, and FEF_25-75% across all methods, especially in those with fewer airways.

These results provide insight into which Pi10/PiSlope calculation methods may generate more reliable findings and indicate PiSlope may provide a more comprehensive measure of airway remodeling in COPD.

Question CT Pi10/PiSlope measures airway remodeling in COPD, but calculation methods vary. Are CT Pi10/PiSlope reliable and provide comparable associations with lung function across various methodologies? Findings Pi10/PiSlope calculation methods that included a higher number of CT airways were more reliable, and PiSlope showed stronger associations with lung function than Pi10. Clinical relevance These results provide insight into which methods may generate more reliable findings, aiding interpretation of results between studies, and support the potential of PiSlope as a reliable biomarker for monitoring airway remodeling in COPD.