Evidence supporting the role of fungus in CRS

Fungus has been found in the sinuses and/or nasal cavities of CRS patients as well as controls. An initial report described highly sensitive fungal culture techniques that demonstrated fungi in nearly all CRS patients—and in all controls. The methods and results were subsequently duplicated in Europe, and staining of fungal-specific chitin again demonstrated the presence of fungi in all CRS patients. Others have shown the presence of Alternaria DNA in 100% of CRS patients’ sinuses and in 67% of healthy controls’ sinuses. Inasmuch as these reports documented the presence of fungi in the sinuses not only in CRS patients but also in controls, fungi themselves do not appear to be the causative factor.

Much attention has focused on the reaction to fungus, especially the eosinophil and its relationship to fungus and local cytokines. Eosinophils are present in the late phase of allergic rhinitis and are also seen in many forms of CRS, particularly CRS with polyps. Eosinophils are also prominent in the reaction to parasitic infections. The mucin from CRS patients has been found to be more chemotactic for eosinophils compared with the mucin of healthy controls. Mucin from CRS patients has been further found to heterogeneously contain clusters of eosinophils with high levels of eosinophilic granule major basic protein, a cationic protein toxic to extracellular microorganisms but also to respiratory mucosa. Eosinophils are known to migrate into respiratory mucosa by interleukin (IL)-13–induced expression of adhesion molecules in the microvasculature, with subsequent migration out of the vessels and into the tissues. Another cytokine, IL-5, promotes eosinophil differentiation, activation, and survival in the tissues by inhibiting apoptosis. Shin and colleagues exposed peripheral blood mononuclear cells (PBMCs) to fungal antigens in vitro, and reported increased IL-5 and IL-13 production in 89% of CRS patients but not controls. The response was particularly brisk with exposure to Alternaria . The IL-5 production was found to be independent of fungal-specific IgE and instead correlated with fungal-specific IgG, implying a nonallergic mechanism.

These findings have led to a hypothesis wherein fungi on the sinus mucosal surface induce production of cytokines, which promote eosinophil migration through the epithelium toward the mucin. These eosinophils arrive at the fungal-containing mucin and release cationic proteins to destroy the fungi, but in so doing perpetuate and potentially worsen the mucosal inflammation seen in CRS. Some have even described this fungal-driven paradigm as the universal explanation for CRS, stating that “most, if not all, chronic rhinosinusitis conditions have a fungal etiology.”

Evidence questioning the role of fungus in CRS

With initial in vitro evidence showing fungus as a potential etiologic agent in CRS, clinical trials with antifungal therapy soon followed. In 2002, 2 nonblinded uncontrolled studies were published examining topical amphotericin B in the treatment of CRS. Ponikau and colleagues noted symptomatic improvement in 75% of 51 CRS patients treated with amphotericin B irrigations. Endoscopic and radiologic improvements were also reported. Ricchetti and colleagues similarly treated 74 nasal polyposis patients with amphotericin B nasal irrigations and reported resolution of polyps in 39%.

To minimize the potential for patient and observer bias, these open-label pilot studies were followed by randomized, double-blind, placebo-controlled trials of antifungal irrigation. Unfortunately, all of these studies failed to show a substantial clinical effect. Weschta and colleagues compared amphotericin B spray to saline spray, and failed to see an effect in multiple patient parameters including symptom scores, endoscopy, radiologic evaluation, and overall quality of life. Presence of fungus in nasal lavage before or after treatment did not correlate with any outcome parameters. Subsequent work demonstrated that amphotericin B treatment had no effect on the eosinophilic markers tryptase and eosinophil cationic protein, and that these markers did not correlate with the success of fungal eradication.

Ponikau and colleagues examined amphotericin B irrigation, and reported an improvement in radiologic and endoscopic scores. The improvement was statistically significant but the clinical significance was questionable. Moreover, there were no significant improvements in symptom scoring, presence of Alternaria , IL-5 concentration in mucus, or blood eosinophil levels. In this study, 2 of 15 patients (13%) treated with amphotericin B experienced an adverse effect with the medication and discontinued it.

Ebbens and colleagues exhaustively examined numerous symptom parameters and endoscopy scores in patients treated with amphotericin B, and found it to have no effect compared with placebo saline irrigation. A follow-up study examined 24 cytokines, chemokines, and growth factors in CRS nasal lavage samples, and found 13 weeks of topical amphotericin B treatment to affect none of them. Shin and Ye also failed to demonstrate an effect of amphotericin B on proinflammatory markers in patients with nasal polyps as compared with saline.

While these studies cast serious doubt on the fungal hypothesis, it is possible that lack of sufficient delivery could explain the failure of topical therapy. It is known that topical treatments have limited ability to access the sinuses and are instead predominantly delivered to the nasal cavity, even in patients who have previously undergone sinus surgery. Unfortunately, systemic therapy has not been shown to be effective either. Kennedy and colleagues examined the effect of oral systemic terbinafine and found no radiologic improvement in CRS patients compared with untreated controls, even in patients who were positive for fungus in pretreatment cultures. Nevertheless, with this study the possibility remained of inadequate systemic delivery to the sinus mucin. With systemic antifungal therapy, the risk of hepatic complications with systemic therapy must always be borne in mind.

Amphotericin B irrigations, though initially promising in nonblinded pilot studies, have failed to show clinical efficacy in more rigorous trials. Even in those instances where a small effect was noted, an immunomodulatory or cell-permeability effect seems to be just as likely or more likely to be the explanation. Amphotericin B works by binding ergosterol in the fungal cell wall, increasing its permeability to the point of cell death. Amphotericin B also binds cholesterol in the mammalian cell membrane and alters the cells’ permeability, though to a lesser degree. It is therefore possible that the very limited clinical effect seen in one of these studies can be attributed to a direct reduction in mucosal edema, irrespective of the solution’s antifungal activity. Kanda and colleagues further demonstrated that azole antifungals, such as itraconazole, suppress IL-4 and IL-5 expression by T lymphocytes, again questioning the relative importance of the antifungal versus an immunomodulatory effect of these oral medications. Additional in vitro work questions the importance of the antifungal effect in any of these studies. Shirazi and colleagues examined the antifungal efficacy of the amphotericin B concentration commonly used in nasal lavage and found it to be ineffective, whereas 2- and 3-fold increases were effective. Unfortunately, these higher doses are poorly tolerated in the nasal cavity.

How does one reconcile the rather persuasive in vitro data with the paucity of evidence of antifungals’ clinical effectiveness? Can the failure of antifungal medication simply be due to lack of sufficient delivery? Is fungus truly an appropriate universal target for therapy in CRS patients?

With the in vitro data appearing to be at odds with a large number of negative clinical reports, it is important to take a closer look at these laboratory data and, where possible, attempt to replicate them. Specifically, the finding of Shin and colleagues that 89% of CRS patients had non-IgE–mediated elevated IL-5 responses to Alternaria alternata extract has remained compelling evidence that fungus may indeed be an etiologic factor in CRS. By painstakingly replicating the methods as far as possible, Orlandi and colleagues recently examined this relationship in a more heterogeneous group of CRS patients. Patients at multiple points on the disease severity spectrum and residing in different areas of the United States (with sharply differing climates) were purposely chosen to test the universality of the fungal hypothesis. Not only did this subsequent study fail to produce the results of Shin and colleagues in more heterogeneous patients, but some of these results were directly opposite of the original findings.

Orlandi and colleagues found that IL-5 was produced following Alternaria exposure by PBMCs of patients but also by those of controls, and the response was heterogeneous and did not correlate with the presence of CRS. In addition, Alternaria -induced levels of IL-13, the principal chemoattractant for eosinophils, did not differ between CRS patients and controls. Moreover, IL-5 levels correlated strongly with fungal-specific IgE and did not correlate with fungal-specific IgG, contrary to the results seen by Shin and colleagues.