SPECT/CT to quantify early small airway disease and its relationship to clinical symptoms in smokers with normal lung function: a pilot study

Introduction: Smokers frequently display respiratory symptoms despite the fact that their pulmonary function tests (PFTs) can be normal. Quantitative lung ventilation single-photon emission computed tomography (SPECT/CT) can provide a quantification of lung ventilatory homogeneity and could prove useful as an early marker of airway disease in smokers. We measured the effects of smoking on regional ventilation distribution in subjects with normal lung function and evaluated whether ventilation distribution in these subjects is related to lung function tests results and clinical symptoms. Methods: Subjects without any history of respiratory disease were prospectively recruited and separated in two groups: active smokers (AS: >= 10 cigarettes/day and history of >= 15 pack-years) and never smokers (NS: lifetime exposure of <5 cigarettes). All subjects performed PFTs (which had to be normal, defined as z-score values of forced expiratory volume in 1 s (FEV1), forced vital capacity (FVC), FEV1/FVC ratio, total lung capacity (TLC) residual volume and diffusion capacity (DLCO) all falling between -1.65 and +1.65) and underwent SPECT/CT with Technegas, which generated subject- specific ventilation heterogeneity maps. The area under the compensated coefficient of variation (CV) density curve for CV values > 40%, (AUC-CV40%) was used as the measure of ventilation heterogeneity. Results: 30 subjects were recruited (15 per group). Subjects in the AS group displayed higher dyspnea levels (1 [1-2] vs. 0 [0-1] units on mMRC scale, p < 0.001). AUC- CV40% was significantly higher in the AS group (0.386 +/- 0.106 vs. 0.293 +/- 0.069, p = 0.004). AUC-CV40% was significantly correlated to FEV1 (rho = -0.47, p = 0.009), DLCO (rho = -0.49, p = 0.006), CAT score (rho = 0.55, p = 0.002) and mMRC score (rho = 0.54, p = 0.002). Subjects with mMRC >0 had higher AUC-CV40% values than those without dyspnea (0.289 +/- 0.071 vs. 0.378 +/- 0.102, p = 0.006), while FEV1 and DLCO were not different between those groups. ROC analyses showed that the AUC for AUC-CV40% in identifying subjects with mMRC score >0 was 0.78 (95%CI 0.61-0.95, p = 0.009), which was significantly higher than that of FEV1 and DLCO. Discussion: In smokers with normal lung function, ventilatory inhomogeneities can be quantified using SPECT/CT. AUC-CV40% values are related to lung function decline and to respiratory symptomatology, suggesting a potential role for this marker in the evaluation of symptomatic smokers.