Abstract
Reduced pulmonary diffusing capacity for carbon monoxide (D(LCO)) can be observed in pulmonary arterial hypertension (PAH) and associates with increased mortality. However, the prognostic value of D(LCO) when corrected for haemoglobin (D(LCOc)), an independent modifier of D(LCO), remains understudied. Additionally, the prognostic role of ventilation (V)-perfusion (Q) emission computed tomography (V/Q SPECT) findings in patients with PAH, which may concurrently be performed to rule out chronic thromboembolic pulmonary hypertension, is uncertain. A retrospective cohort study was conducted on 152 patients with PAH referred to a tertiary hospital for evaluation from January 2011 to January 2020. Lung function tests, clinical data and V/Q SPECT were ascertained. Cox regression analysis was performed to evaluate the association between D(LCOc), D(LCO) and V/Q SPECT defects at referral with all-cause mortality. In equally adjusted Cox regression analysis, each percentage increase in D(LCOc) % predicted (%pred) (hazard ratio (HR) 0.97; 95% CI: 0.94-0.99) and D(LCO)%pred (HR 0.97; 95% CI: 0.94-0.99) was similarly associated with all-cause mortality. There was no detectable difference in area under the curve for prediction of all-cause mortality by D(LCOc)%pred and D(LCO)%pred (C-index 0.71 and 0.72, respectively, P = 0.85 for difference). None of the defects noted on V/Q SPECT were significantly associated with mortality, but mismatched defects were associated with lower values of D(LCOc)%pred and D(LCO)%pred. D(LCOc)%pred and D(LCO)%pred perform equally as prognostic markers in PAH, supporting the use of either metric when available for prognostic stratification.