This study is divided in two different parts: clinical effects of BT and evaluation of gene expression levels associated to clinical outcomes. Our results confirmed the efficacy of BT in the treatment of severe asthma, as we observed 81% of responders, in line with recent results published by Langton’s group [26].
Notably, our research was conducted in real life and it follows that patients who underwent BT had more severe asthma than patients included in earlier trials in term of rate of exacerbations (6 vs 0.7), AQLQ score (2.5 vs 4.7) and OCS maintenance dose (89% vs 41%) [8, 9].
Regarding clinical outcomes, a significant improvement in asthma symptoms in terms of ACQ scores, AQLQ scores and rate of exacerbations was observed. These results confirm the clinical benefits of BT in line with previous researches [20, 21, 26]. AQLQ improvement at 12 months was statistically significant (p< 0.0001) which is different from what was presented in AIR2 Trial, where AQLQ was statistically borderline and questioned [12, 25, 27–29]. Moreover, we observed a significant reduction in maintenance doses of OCS.
In the scenario of multiple therapeutic approaches for severe asthma treatments, BT still remains the most controversial. One of the biggest objections is the absence of a specific biomarker able to define target population for BT [12]. Currently, BT is used primary in non-eosinophilic phenotype, according to guidelines as there is no usable biological therapy [25]. In our site, BT started in 2012, when the only available biological treatment was omalizumab. Based on clinical criteria for responder definition described in Table 2, we observed 86% of responders in the T2-low cohort and 77% of responders in the T2-high cohort. This finding confirms that BT is a good therapeutic option in T2-low asthma patients. Moreover, it may be effective also in T2-high asthma patients (18% of patients were switched from omalizumab). Thus, our data emerge in contrast with the conception on BT being a treatment only in case of T2-low endotypes or when biological drugs fails [25]. In the real practice, BT represents a therapeutic option for patients with severe asthma despite of asthma endotypes.
The study expands the knowledge regarding the mechanisms of action of BT using gene expression profiling in bronchial biopsy samples. Genes to be investigated were selected through a hypothesis-driven approach based on and expanding data available in the literature [6, 13–17, 20–24].
BT induced a reduction of ACTA2 and an increase of CD68, FAP, COL1A1, COL1A2 mRNAs in the majority of patients suggesting tissue remodelling with a likely decrease in smooth muscle cells and an increase in monocytes/macrophages/dendritic cells and connective tissue. However, these changes seemed to occur irrespective of clinical outcomes, with the exception of the increase in CD68 mRNA levels which resulted associated with higher numbers of exacerbations post-BT.
ACTA2 gene (also known as α-SMA) encodes for smooth muscle actin. Its expression is characteristic of smooth muscle cells and myofibroblasts. There are some reports showing that one of the effects of BT treatment is the reduction of airway α-SMA [6, 13–15, 20, 21, 30, 31]. Herein data confirm this finding at mRNA level in bronchial biopsies from a wide cohort of asthmatic patients after BT treatment.
CD68 is mainly expressed by monocytes, macrophages (both M1 and M2 macrophages) and dendritic cells. The increase of CD68 positive cells in bronchial tissues from asthmatic patients compared with healthy controls is known [32]. Moreover, CD68 expression has been previously reported as increased after BT treatment at protein level in bronchial biopsies from 12 asthmatic patients by our group [6], herein confirmed at mRNA level by real-time PCR in a larger cohort of patients.
FAP expression is low in most adult tissues under physiological conditions but it is usually high in reactive stromal fibroblasts. FAP is thought to be involved in the control of fibroblast growth and epithelial-mesenchymal transitions during development, tissue repair, and carcinogenesis. In our knowledge, this is the first report that documented the up-regulation of FAP following BT treatment.
COL1A1 and COL1A2 mRNAs encode pro-alpha I and pro-alpha II chains of type I collagen. Chakir J. and collaborators have reported a reduction in type I collagen thickness below the basement membrane by immunohistochemistry in 9 bronchial biopsies collected from patients at third BT session in comparison to baseline [15]. This alteration persisted after 27 months of follow up [16]. In contrast, we found a huge increase of mRNAs for type I collagen in 89% (COL1A1) and 63% (COL1A2) of bronchial biopsies during BT treatment. Our results at mRNA levels agree with the observations of Pretolani and collaborators showing significantly larger and more intense area stained for collagen in bronchial biopsies at the last BT session versus baseline [20]. The explanation for the discordant results could be found in the fact that Chakir J et al [15]. and Salem J.H. et al [16]. focused the attention only on the collagen localized under the reticular layer of the basement membrane, while we analyzed the total content of collagen, as well as Pretolani et al did [20].
Asthma is characterized by airway hyperresponsiveness and it is known that there is a direct correlation with airway remodelling [33]. At cellular level, airway remodelling consists in smooth muscle hyperplasia characterized by a thicker α-SMA layer, epithelial damage and collagen deposition in the subepithelial basement membrane [21, 33–35]. The emerging idea is that the deposition of connective tissue around the airways can determine tissue stiffness, dampening the contractility of α-SMA layer [35, 36]. Based on the present data we speculate that BT treatment could reduce airway hyperresponsiveness decreasing α-SMA layer and increasing fibroblasts with collagen production.
PGP9.5 is a protein expressed in the innervation of mucosal glands, smooth muscle and mucosal blood vessels of the lungs [37]. Previous studies, which used PGP9.5 as a marker for the identification and count of nerve C fibers by immunohistochemistry, have reported a reduction of neural innervation after BT treatment [6, 20, 38]. Herein, changes in PGP9.5 mRNA levels during BT treatment were heterogeneous among patients. However, a higher reduction in PGP9.5 mRNA levels correlated with better patients reported outcome in terms of ΔAQLQ, suggesting that a reduction of innervation might be linked with BT efficacy.
To identify predictors of BT efficacy we associated normalized gene expression levels at baseline and T2 as well as changes in gene expression at T2 versus baseline with clinical parameters 12 months after BT. At baseline, lower mRNA levels of OCLN and COL1A2 resulted associated with parameters of better outcome: lower numbers of exacerbations and lower ACQ values post-BT. Moreover, it appeared valuable to quantify gene expression during BT, at the third session of BT. Indeed, lower mRNA levels of OCLN, CD68, CTGF and higher mRNA levels of SLPI at T2, lower changes in CD68 and CTGF mRNA levels at T2 versus T0 resulted associated with fewer exacerbations post-BT. We speculate that the quantification of OCLN, CD68, CTGF and SLPI mRNA could be useful to identify patients at higher risk of exacerbations following BT treatment.
OCLN is a key component of tight junctions, it is involved in the creation of a barrier in airway epithelial cells reducing the transport from apical to basolateral surface. Reduction in OCLN levels has been associated with impaired barrier function in asthma patients [39, 40]. OCLN is also expressed in tight junctions by myelinating Schwann cells protecting peripheral nerves and promoting neurons communication [41]. This is the first study that investigated OCLN in the context of BT treatment and further in-depth analyses will be necessary to elucidate the relationship between tight junction functionality and development of exacerbations.
CTGF gene expression has been linked to airway smooth muscle cells [42] and fibroblast to myofibroblast transition [43]. Blocking CTGF has been suggested as a rationale option in patients with asthma [44].
SLPI gene expression has emerged as the most promising marker at baseline to predict response to BT in the work by Ano S et al [24]. SLPI can inhibit leukocyte elastase, cathepsin G, trypsin and mast cell chymase. Based on Figure 3 shown in that manuscript [24], it seems even possible to define a cut-off in SLPI expression in bronchial specimens which could distinguish responders from non-responders to BT [24]. We were not able to confirm the usefulness of the quantification of SLPI gene expression at baseline to predict response to BT. However, the quantification of SLPI gene expression levels during BT resulted the best marker associated with the numbers of exacerbations post-BT, thus confirming a potential value for SLPI.
We did not detect any expression of IL-4, IL-13 and IL-17A mRNAs and a limited expression of IL-5 mRNA in bronchial samples. These cytokines, markers of Th2 and Th17 lymphocytes, have been reported as expressed in samples from asthmatic patients (bronchial biopsies, epithelial brushings, sputum, bronchoalveolar lavage) [2] but usually by means of different techniques: immunohistochemistry and in situ hybridization. One study has investigated the expression of IL-4, IL-5 and IL-13 by PCR: no expression has been detected unless tissue fragments were stimulated ex vivo [45]. Recently, Dr Pretolani and collaborators have reported a decrease in IL-13-positive cells per mm2 by immunohistochemistry in bronchial biopsies after BT [31]. Differences in the results can be explained hypothesizing a scarce correlation between mRNA and protein expression and/or a different ability to detect these markers in tissue samples between PCR and immunohistochemistry.
The limit of the present study is the hypothesis-driven selection of the investigated genes. However, the results are in part consistent with those of previous studies, in part open up new possibilities. A high-throughput approach would allow to discover additional mechanisms of BT treatment. Analysis of the transcriptome has been recently performed on samples from really small cohorts of patients who underwent BT treatment: on cytology brushes from tracheal walls of 5 patients [46], and on bronchial biopsies from 8 patients [24]. Furthermore, Sun Q. et al. used a high-throughput approach on 9 primary airway epithelial cell lines obtained from patients before and after BT treatment and cultured in vitro [47]. In our knowledge this study analyzed the largest cohort of patients who underwent BT, coupling molecular and clinical data at baseline, at the last BT session and after 1 year of follow up. Moreover, the use of real-time PCR allowed us to analyze in each sample a higher number of markers compared with previous studies [6, 13–17, 20, 21, 30, 31].