Sputum basophils and asthma diagnosis: dawn of a new era?

More than 40 years ago, Kimura et al.1 discovered that basophils are enriched in sputum samples of asthma patients. Since then, various scientific approaches have attempted to elucidate the role of basophils in the pathophysiology of asthma. The scarcity of basophils and technical limitations of earlier studies have made it difficult to increase our understanding how basophils affect the disease course in asthma. Moreover, asthma is a highly heterogeneous inflammatory disorder, and the role of basophils may vary between different disease phenotypes. In this issue of Allergy, reports of Brooks et al.2 and Suzuki et al.3 analysed the numbers of basophils in induced sputum obtained from patients with different asthma phenotypes. Both studies are remarkable for several reasons: First, both reports quantified sputum basophils by flow cytometry. In previous studies, the number of sputum basophils was determined by differential cell counts using light microscopy. Technical advances in flow cytometry now enable the identification of rare cell populations in low volumes of complex cell suspensions such as sputum. Importantly, it has been shown that leucocyte numbers determined by flow cytometry correlated with those obtained by microscopy,4, 5 a finding now further confirmed in the present study by Suzuki et al.3 To date, only a few studies have compared leucocyte populations between induced sputum and bronchoalveolar lavage (BAL; summarized in ref. 2), demonstrating that the cellular composition of the two sample types is similar, yet with differing frequencies of cell populations. This observation is not surprising as BAL and induced sputum contain different fractions of contaminating cells and sample cell populations of different origin within the respiratory tract, that is major conducting airways in sputum versus alveolar space in BAL. Also, there may be phenotypic and functional differences between specific cell populations obtained from induced sputum and those collected from BAL, as has been shown for macrophages.6 As a direct comparison between sputum and BAL basophils is currently lacking, it remains to be investigated whether such differences are also observed between sputum basophils and those obtained from BAL. Nevertheless, the work of Freeman et al.7 demonstrates the feasibility of such an approach, even in a multicentre setting. As a second key finding, both reports observed an increased frequency of basophils in sputum of asthma patients vs healthy controls, and a positive correlation between sputum basophil and eosinophil counts. In addition, Suzuki et al. demonstrated that surface marker expression of sputum basophils differed from blood basophils.2, 3 These observations shed some light on the unresolved question regarding the role of human basophils in perpetrating and maintaining the pathogenesis of asthma. In particular, it has not been fully clarified whether basophils are merely innocent bystanders, or active players that affect the course of asthma. The presence of basophils in lung tissue of patients with fatal asthma and the correlation of basophil numbers with airway responsiveness was already reported in the early nineties.8, 9 More recent studies in humans and mice confirm that basophils infiltrate lung tissue upon allergen challenge, which is concordant with increased basophil counts in bronchial biopsies obtained from asthma patients.10 These findings support the notion that basophils play an active role by being in the right place at the right time. The question whether or not basophils provide a decisive functionality in this context is more difficult to answer. In support of such functionality, it has been demonstrated that human basophils are the major early source of IL-4 and IL-13 during allergic asthma,11, 12 which licences them as essential regulatory cells. An immunoregulatory role of human basophils is further suggested by numerous mouse studies (reviewed in ref. 13). Also, the present study by Suzuki et al.3 indirectly endorses the view that human basophils affect the pathogenesis of asthma. Specifically, the analysed surface markers are indicative of both effector and immunoregulatory roles of human basophils. For instance, the study revealed that sputum basophils express lower levels of IL-3R alpha-chain compared to blood basophils. In vitro studies have shown that the expression of IL-3R alpha-chain remains high and stable over a period of 24 hours upon IL-3 addition.14 Hence, one may speculate that sputum basophils from asthma patients were continuously exposed to IL-3 over a prolonged period, resulting in downregulation of IL-3R alpha-chain. IL-3 is known as the major priming factor of human basophils to induce long-term phenotypic and functional changes. Importantly, previous studies have shown that IL-3 promotes IL-4, IL-13 and IL-8 secretion both independently and in synergy with other known basophil stimuli (eg, FcεRI cross-linking, IL-33).15-17 Basophil-derived-IL-4, on the other hand, has been shown to regulate infiltration of eosinophils.18 Although the regulation of eosinophilic inflammation likely depends on multiple factors, the present work by Brooks et al.2 and Suzuki et al.3 thus supports the hypothesis that basophils are essential for the infiltration of eosinophils to airway inflammation. Hence, human basophils may indeed be crucial players in the pathogenesis of asthma. Proposing that basophils affect every asthma phenotype would, however, constitute far-fetched claim. Nevertheless, in this context, the studies of Brooks et al. and Suzuki et al. provide a third highly relevant novel insight by successfully associating sputum basophils to eosinophilic asthma.2, 3 This is a key finding since asthma, historically characterized “by the history of respiratory symptoms such as wheeze, shortness of breath, chest tightness and cough that vary over time and in intensity, together with variable expiratory airflow limitation”19 is now known to be triggered by various immunopathologies. Therefore, asthma is increasingly recognized as a syndrome rather than a unique disease entity, which has resulted in efforts to categorize asthma patients into different disease phenotypes. In particular, varying responses to escalating inhaled and/or systemic corticosteroids between subgroups of severe asthma patients highlight the need to individualize therapies based on disease phenotype. Similarly, targeted therapies such as anti-IL-5 (eg, mepolizumab) or anti-IL-13 (eg, lebrikizumab) are ineffective when patients are not selected according to asthma phenotype (reviewed in ref. 20). One attempt to stratify asthma patients was based on the proportion of eosinophils and neutrophils in induced sputum samples, defining four inflammatory asthma groups: eosinophilic, neutrophilic, mixed granulocytic and paucigranulocytic asthma.21 Further patient stratification may be achieved by combining multiple clinical (eg, age, gender, time of onset, lung function) and molecular (eg, Th2) characteristics. To date, the presence of human basophils has not been incorporated to define asthma phenotypes. The studies by Brook et al. and Suzuki et al. presented in this issue provide the first evidence associating human basophils to a particular asthma phenotype.2, 3 Both studies convincingly connect a higher percentage of sputum basophils to eosinophilic asthma, rather than to noneosinophilic asthma (Figure 1). Moreover, Suzuki et al. demonstrated a higher sensitivity, specificity, positive predictive value and negative predictive value of sputum basophils compared to blood eosinophils and exhaled nitric oxide (FENO), which are both considered good surrogate markers for eosinophilic asthma. Similarly, the areas under the receiver operating characteristics curves were greater for sputum basophils than for blood eosinophils and FENO. This highlights the utility of sputum basophil counts for the discrimination of asthma phenotypes. Given these promising findings, it is tempting to speculate that sputum basophilia may not only serve as a marker for asthma phenotypes, but may also provide a useful means to delineate asthma endotypes. Asthma endotypes group asthma patients according to underlying molecular mechanisms: one particular endotype may contain several phenotypes, while a particular phenotype may also be allocated to several endotypes. A commission of experts from the European Academy of Allergy and Clinical Immunology and the American Academy of Allergy, Asthma & Immunology proposed a tentative classification of seven asthma endotypes described in the PRACTALL (PRACtical ALLergy) consensus.22 Of particular note, the phenotype “eosinophilic asthma” was associated with four different endotypes: allergic asthma (adult), aspirin-sensitive asthma, severe late-onset hypereosinophilic asthma and allergic bronchopulmonary mycosis (Figure 1). With the number of sputum basophils being highly predictive of eosinophilic asthma, a logical next step is to analyse activation status of sputum basophils from patients with asthma categorized as one of the four mentioned endotypes. This could be achieved by analysis in the same manner as presented in the current work by Suzuki et al.,3 combined with a determination of intracellular cytokine profiles23 (Figure 1). Taken together, flow cytometric quantitative, phenotypic and functional analysis of sputum basophils finally seems ready to advance from its pioneering experimental era to implementation as a cutting-edge clinical application empowering individualized asthma management. The two studies in this issue of Allergy provide an important step in this direction.2, 3 The authors would like to thanks Dr. Ursula Amstutz for her help in proofreading the manuscript. The authors declare that they have no conflicts of interest. Both authors contributed equally to the ideas of this editorial. Michaela Fux has provided the draft of the manuscript and the illustration. The authors edited and finalized this editorial in close cooperation.