Alveolar dead space as a predictor of severity of pulmonary embolism

Abstract

Objective: To determine whether the alveolar dead space volume (V(D)alv), expressed as a percentage of the alveolar tidal volume (V(D)alv/V(T)alv), can predict the degree of vascular occlusion caused by pulmonary embolism (PE). Methods: Fifty-three subjects with suspected PE were prospectively studied. Pulmonary embolism was diagnosed in 33 by high- probability ventilation/perfusion (V/Q) scan (n = 19) or by pulmonary arteriography (PAG, n = 14). Pulmonary embolism was excluded by PAG in 20 subjects. The V(D)alv/V(T)alv was determined from volumetric capnography and arterial blood gas analysis, which permits measurement of the physiologic dead space, V(D)phys (mL) = [(PaCO2 - PeCO2)/PaCO2] · tidal volume. Airway dead space (V(D)aw) was subtracted to yield the alveolar dead space [(V(D)phys - V(D)aw) = V(D)alv (mL)]; the percentage of alveolar volume occupied by alveolar dead space per breath = V(D)alv/V(T)alv · 100%. Percentage perfusion defect was determined from V/Q scans by a radiologist blinded to other data. Regression analysis was performed to show correlation between V(D)alv/V(T)alv and defect on V/Q scan or systolic pulmonary arterial pressure (SPAP). Results: For subjects with PE, the mean perfusion defect on lung scan was 38 + 22%; the mean V(D)alv = 208 ± 115 mL, V(T)alv = 452 ± 251 mL, and V(D)alv/V(T)alv = 43 ± 18%. Regression of V(D)alv/V(T)alv vs perfusion defect yielded r2 = 0.41. Regression of V(D)alv/V(T)alv vs pulmonary artery pressures yielded r2 = 0.59. For subjects without PE, V(D)alv/V(T)alv = 27 ± 14% and V(D)alv = 89 ± 66 mL. Conclusions: The V(D)alv/V(T)alv correlates with the lung perfusion defect and the pulmonary artery pressures in subjects with PE. These findings show the potential for V(D)alv/V(T)alv to quantify the embolic burden of PE.