Chronic Rhinosinusitis: More Than Just “Asthma of the Upper Airway”

Traditionally, rhinosinusitis (RS) was divided into three distinct conditions according to duration, with chronic RS (CRS) defined as disease present for longer than 12 weeks. In addition, whereas acute and subacute sinusitis were defined as infectious processes driven by pyogenic organisms such as Streptococcus, Haemophilus, and Moraxella, CRS was contrasted by the presence of anaerobic, gram-negative, and Staphylococcus infections. Although more acute forms of RS are still primarily considered to be infectious diseases (although now they are most commonly recognized as being caused by respiratory viruses), there has been a comprehensive rethinking of the role of infection in the pathogenesis of CRS. Although bacteria are present and influential in exacerbating the presence and severity of sinus inflammation, CRS is no longer considered to be primarily an infectious disease but, instead, is a disease of chronic, endogenously driven inflammation. What has been enigmatic, however, is the nature of that inflammation.

In this issue of the Journal, the study by Stevens and colleagues (pp. 682–694), as such, is an important advance in our understanding of the nature of CRS or, rather, the nature of what will ultimately prove to be the numerous distinct diseases that make up the cluster of conditions we currently call CRS (1). More recent thinking has begun to address the need to phenotype CRS by dividing these conditions into those not associated with the tendency to develop nasal polyps (CRSsNP) and those that do (CRSwNP) (2), with the latter group further subdivided according to those who do and do not tolerate aspirin and other cyclooxygenase inhibitors (termed aspirin-exacerbated respiratory disease [AERD]).

In addition to the (relative) ease of being able to use NPs to define the CRS phenotype, the driving feature behind the decision to give polyp presence such prominence in phenotyping these disorders was that the presence of polyps typically (although far from universally) defines a condition whose central feature is infiltration with eosinophils (3, 4). CRSwNP is also highly associated with asthma and has an appearance remarkably similar to asthma (5), a condition more often than not present in these patients. Together, these observations suggest the concept that these are plausibly the same disease process of “allergic” inflammation simultaneously afflicting respiratory epithelium of the upper (sinus) and lower (bronchial) airways. In addition, just as asthma is most commonly associated with eosinophilic inflammation and a Th2 cytokine signature, one of the central outcomes of the current study is the authors’ demonstration that CRSwNP is similarly defined by the robust expression of IL-4, IL-5, and IL-13. In addition to this “Th2” cytokine signature, the investigators also demonstrated robust expression of chemokines associated with eosinophil recruitment, including monocyte chemoattractant protein 4 (MCP-4) (CCL13 [chemokine (C-C motif) ligand 13]) and the eotaxins (C-C chemokine receptor type 11 [CCR11], CCR24, and CCR26).

These similarities of CRSwNP to asthma suggest the question of whether, similar to asthma, it may comprise unique subtypes, such as the “high Th2” and “low Th2” disorders described for asthma (6). That is certainly consistent with the wide range of cytokine, and especially eosinophil cationic protein (ECP), concentrations observed in the study by Stevens and colleagues, with individuals in their Figure 1A readily falling into “high” ECP and “low” ECP cohorts. Similarly, this concept may also be suggested by the overlap of CRSwNP subjects in the principal component analyses (Stevens and colleagues' Figure 4), with individuals who are “more AERD-like,” but also with those who appear “more CRSsNP-like.” Such distinctions will prove important in future studies, as they are likely to analogously predict a steroid nonresponsive group (something not yet properly studied in CRS), but would certainly also explain the wide variability of responsiveness of CRSwNP subjects to treatment with IL-5 antagonists (7).

What is less likely is that CRSsNP comprises the upper airway manifestation of a “low Th2” airway disease. This is less likely, as CRSsNP is not often associated with asthma or any other lung pathology. So what, then, is CRSsNP? Although previous studies have suggested that CRSsNP is an adaptive Th1-induced disease process (8–10), that concept was not confirmed in the current studies. This current investigation is much more extensive than previous work and includes complementary studies simultaneously addressing both tissue and nasal secretions, along with using both proteomics and quantitative transcript expression. Most important, these investigators focused on sinus-associated uncinate tissue, not nasal tissue, for their investigations. As such, the absence of IFN-γ protein or transcripts or any other Th1-associated cytokines in the current study seem fairly compelling, along with the observation that the pathology of CRSsNP is not typical of other “high” IFN-γ conditions; that is, it is not associated with extensive infiltration with activated macrophages.

So if not a Th1 disease, what is CRSsNP? This is not something that can be readily discerned from this study, except from the authors’ conclusion that it does not suggest an adaptive T effector cell–driven disorder. Perhaps the best explanation lies in other investigations suggesting that CRSsNP is a disease of breakdown of innate immunity leading to the spectrum of frequent recurrent episodes of acute rhinosinusitis, bacterial colonization, and biofilm formation, and ultimately the extensive tissue remodeling that characterizes these patients (11, 12).

Although AERD shares many features with CRSwNP, including the prominence of the eosinophilia and polyp formation, there are many distinct features that suggest it is a unique endotype, and not just different spectrums of the same disease process. These include the much more prominent role of eosinophils, as previously published (13) and demonstrated in this study by the fivefold elevation in ECP concentrations (Stevens and colleagues’ Figure 3A). More than that, AERD is distinguished not only by the sensitivity of aspirin and other nonselective cyclooxygenase inhibitors but also by a distinct natural history (13), the unique overexpression of cysteinyl leukotrienes and their receptors (14, 15), and overexpression of PgD₂ (16). Thus, it is intriguing that the current study, although confirming the presence of a Th2 signature, does demonstrate a hint at other distinctive features of AERD. Although excessive granulocyte–macrophage colony–stimulating factor and MCP-1 may support a more robust Th2 process, the dearth of tissue plasminogen activator is quite intriguing, suggesting a role for the procoagulation pathway and fibrin deposition that perhaps drives some of the aggressive remodeling that defines this disorder. Activation of the coagulation pathway might also be consistent with recent studies demonstrating a role for platelet-associated inflammation in AERD (17), which is another area of research largely unexplored in these diseases and inviting further research.