Discussion
Our study conducted in schooler asthmatic children found significantly lower lung function at baseline in patients compared to healthy controls, either “usual” spirometric parameters or shape indexes, and higher BD reversibility. Shape indexes were found to be slightly more sensitive in identifying asthmatic children at baseline than zFEV1 with the same specificity while their BD reversibility showed high sensitivity and low specificity compared to Δ%FEV1. Conversely, specificity was higher for Δ%FEV1.
Qualitative and quantitative assessment of the descending limb of the FVL and its variability with airway obstruction have already been addressed in past studies, by visual examination or objective measures13,14,23,24. β-angle, constructed to transcribe the curvature of the FVL into a single numerical value 16, is the angle between the first and second half of the FVL at the point of maximum flow point at mid pulmonary volume. β-angle and its simplified estimate, FEF50/PEF, in absolute values or z-scores, were significantly lower at baseline in our school-aged asthmatic children compared to healthy subjects, in keeping with previous reports demonstrating more concave FVL configuration measured by smaller β-angle in asthmatic preschoolers 17 or schoolers18,19,25.
When testing the diagnostic value to separate asthmatics from controls in our population, AUC’s were similar for zβ-angle or zFEF50/PEF and “usual” spirometric parameters, with relatively lower AUC for zFEV1 (figure 2 and table 3 ). zβ-angle and zFEF50/PEF showed slightly higher sensitivity than zFEV1 suggesting that the presence of an early airway obstruction is rather detected by indexes describing FVL curvature than by zFEV1. The specificity was the same for zβ-angle and zFEF50/PEF or zFEV1 (table 3 ). A recent report in schooler asthmatics found large AUC and sensitivity for β-angle and, in contrast to our study, FEV1 sensitivity was also high while β-angle and FEV1’s specificities were very low 18. The different patterns of sensitivity – specificity at baseline indicate that a significant proportion of false diagnosis of asthma would result if solely shape indexes would be taken in consideration to identify airway obstruction.
FEV1, the gold standard parameter for diagnosing an airway obstruction, is obtained in the first second of the forced expiration, situated on the effort-dependent part of FVL. Thus, FEV1 would theoretically be more affected by subject’s effort than distal flows (i.e. FEF50 reported here) located on the second part of forced expiration, considered effort-independent once airway dynamic compression has occurred. Nevertheless, FVL shape is highly conditioned by child’s cooperation and comprehension to perform an optimal, sustained expiratory effort, to obtain a reproducible PEF and a complete expiration to attain the residual volume. This maneuver is easily acquired with appropriate guidance and encouragement from experienced technicians in a pediatric lung function testing lab, but caution must be taken when interpreting a pediatric FVL.
To improve the diagnostic ability of spirometric parameters to detect asthma in children, we investigated whether the combination of shape indexes with “usual” spirometric parameters will improve their discriminating power. AUC’s improved and the best sensitivity was obtained when shape indexes were associated with zFEV1 and the best specificity when they were associated with zFEV1/FVC (table 3 ), in agreement with recent reports demonstrating that the association of quantitative and qualitative parameters may enhance the diagnosis accuracy 18,25.
Our study confirms and extends these observations by testing the BD reversibility of shape indexes and “usual” spirometric parameters and shows larger BD responses for shape indexes in asthmatics than controls, with no difference for Δ%FEV1 (table 1 ). Furthermore, Δ%β-angle and Δ%FEF50/PEF were significantly higher in symptomatic than in asymptomatic asthmatic children (table 2 ), but not BD response of the other spirometric parameters. Δ%β-angle exhibited high sensitivity and low specificity contrary to Δ%FEV1 showing low sensitivity and high specificity (table 3 ). A significant decrease in the upward concavity of the FVL after short-acting BD inhalation have previously shown for other shape indexes in asthmatic children aged 5 to 12y15 and adults 24 or after 8 weeks of inhaled corticosteroids in young adults 14.
The effect of short acting BD in asthma is dependent on complex interactions between airway caliber, ventilation and lung perfusion distribution, ventilation-to-perfusion ratio, and cardiac output26. Short acting BD may have contradictory effects related to its action on distribution of ventilation – perfusion inequalities that may improve, worsen or not change27. BD might act on peripheral airway obstruction with improvement of gas mixing and, consequently, of FVL shape. Contrarily, BD may also worsen gas mixing as partially opens up previously full closed airways, that will become inhomogenously ventilated regions with slow emptying during expiration 26, resulting in no change in FVL curvature.
The aim of our study was to investigate the BD reversibility of shape indexes in the context of a lung function testing lab, procedure routinely used in the diagnosis and follow-up of asthmatic children. Our studied children had less severe asthma with normal baseline lung function in the majority (zFEV1/FVC <-1.64 in less than a fifth) and anti-asthmatic treatment in more than half that might partly underestimate the BD change. Nevertheless, our data suggest a positive effect of BD inhalation on FVL shape, especially in symptomatic asthmatics, with high sensitivity for Δ%β-angle and high specificity for Δ%FEV1 to separate asthmatics from healthy children. Accordingly, considering exclusively Δ%FEV1 as optimal diagnostic tool for reversibility assessment as stated in a number of international guidelines 3,28,29, would result in a large number of false negatives.
In conclusion, our study confirms previous reports that the shape of FVL should be taken in consideration when interpreting a forced spirometry in children. Shape indexes and FEV1 at baseline and their BD reversibility had about the same diagnostic value in detecting airway obstruction but with a different pattern of sensitivity – specificity, being complementary. Quantitative and qualitative assessment of FVL may improve the ability to detect airway obstruction, FEV1 reflecting more proximal while shape indexes peripheral, distal bronchial obstruction. Future perspectives will imply to assess more frequently in practice the association of different lung function parameters by establishing a spirometry score based on shape indexes and “usual” spirometric values, easy to compute and to integrate into spirometers’ software, in order to improve asthma diagnostic or to assess the long-term benefit of anti-asthmatic therapies in pediatric population.