In the current study, we have shown for the first time higher HVEM and IL-5 serum levels in CVA patients compared to CA patients. Although CVA patients with chronic cough as the only respiratory symptom can be distinguished from CA patients, our finding is still important in presenting features of particular inflammatory mediators such as HVEM and IL-5 in CVA. It has been suggested that HVEM and IL-5 are significantly associated with persistent airway inflammation in asthma. Therefore, our study has implied IL-5-induced eosinophilia and HVEM-related inflammation may be responsible for the pathogenesis of CVA.
The present study assessed 64 adult patients who had persistent uncontrolled asthma by the GINA-defined criteria. The CVA group had a higher prevalence of female patients and a shorter duration of disease compared to the CA group, and the results are consistent with the earlier report. Meanwhile, we also found noticed the routine PFT parameters for evaluating mechanical properties of the large, medium-sized, and small airways in CVA group were significantly greater than those in CA group. It may suggest that the decline of pulmonary function is relatively milder in CVA patients. However, previous studies found that the values of MEF50% and MEF25% in CVA patients were lower compared with CA patients, and suggested that CVA is mainly characterized by dysfunctions of small airways [15, 16]. The discrepancy in results may be explained by the different enrolled subjects. Moreover, we still need to monitor the changes of PFT parameters and levels of airway inflammation to prevent CVA from developing into CA.
It has been reported that eosinophilic inflammation may be an underlying mechanism of CVA,  and a blood eosinophil count of > 300 cells/ul was considered an eosinophilic phenotype of asthma in the previous year.[5, 17, 18] In our study, the median blood eosinophil count of two groups were both greater than 300 cells/ul, while blood eosinophil count and eosinophil percent were significantly higher in CVA group compared to those in CA group. We also found that the serum IL-5 level was significantly increased in CVA patients. IL-5 is a major cytokine responsible for the growth and differentiation of eosinophils, and its binding of IL-5 receptor promotes adhesion of eosinophils to airway mucosa, which is a key contributor to airway inflammation in asthma. [13, 19] Therefore, the results indicate that the IL-5/eosinophil pathway may play a pathogenetic role in CVA, and anti-IL-5 therapy may provide an alternative approach for patients with uncontrolled CVA. However, there was no significant difference in FeNO levels between the two groups. Although FeNO could be a useful biomarker for eosinophilic inflammation,  it is highly susceptible to various factors such as atopy, smoking status, pulmonary function and so on. Thus, the FeNO levels between the two groups should be comprehensively evaluated.
HVEM has been implicated to play in airway inflammation and airway remodeling of asthma by combining with TNFSF14. Our study showed that the level of serum HVEM in CVA group was significantly higher than those in CA group. Furthermore, the multivariate logistic regression analysis revealed that HVEM is an independent risk factor for identifying CVA (OR = 1.105, P = 0.015). It may suggest that HVEM-mediated inflammation is uniquely important for the pathogenesis of CVA, but whether this network may also participate in the IL-5/eosinophil pathway of CVA patients is not known. These results provide a new way to explore the pathogenesis of CVA. In addition, the results of ROC curve analysis showed that serum IL-5 and serum HVEM can be used to distinguish CVA from CA. In both biomarkers, HVEM had a higher AUC value of 0.789, and the sensitivity and specificity were 85.0% and 61.1%, respectively, at a better cut-off value of 98.69 pg/ml. Moreover, the combination of serum HVEM and serum IL-5 can improve the AUC value to 0.822 with a specificity of 100%, indicating this combination might be useful to identify the CVA phenotype. While in clinical practice, it would not be commonly used because CVA patients are easy to be identified from CA patients by symptoms. Therefore, increased serum levels of HVEM and IL-5 may implicate elevated airway inflammation in CVA patients and need selected anti-inflammatory drugs to control clinical symptoms. Whereas, this still needs to be confirmed in further studies. Besides, the results of correlation analysis showed that no significant correlation was found between the laboratory indicators (e.g. serum IL-5, serum HVEM, and FeNO) and PFT parameters, suggesting high levels of these inflammatory biomarkers were not associated with the pulmonary function of CVA patients.
There were some limitations in our study. First, our study was from a single center and the sample size was limited, future multicenter studies should be performed to further prove the result. Second, we did not detect the levels of sputum IL-5 and sputum HVEM between the two groups, whether the levels of these inflammatory biomarkers in airways have similar changes are not evaluated. Finally, we did not investigate the mechanisms of HVEM and IL-5/eosinophil pathway in CVA phenotype, and further studies are needed to verify them.