Discussion
Previous studies found that mepolizumab decreased symptoms of CRSwNP.15-17 However, those studies did not investigate mepolizumab-induced changes to eosinophilic inflammation in target sinonasal tissue. There is also negligible data about the response of local tissue to mepolizumab in other eosinophilic diseases.12, 13
A key finding in the current study was a reduction in tissue eosinophils with mepolizumab. This study found a reduction in tissue eosinophils of 59% versus a reduction in blood eosinophils of 83%. This finding is consistent with the only other study – an asthma trial – to have assessed the drug’s effect on airway tissue eosinophils.12 The magnitude of the mepolizumab-mediated decrease in airway eosinophil numbers in the present study (59%) was similar to the reduction (55%) in the asthma study.12 The paucity of tissue eosinophils at Week 24 may be largely attributable to reduced eosinophil production in bone marrow and reduced trafficking of the cell to tissue, which is demonstrated by reduced blood eosinophil levels. IL-5 promotes eosinophilopoiesis in bone marrow,26, 27 and introducing an anti-IL-5 agent curtails bone marrow production of the cell.26 Eosinophils in tissue may live up to several weeks,28-30 compared to the lifespan of an eosinophil in circulation, which is approximately 24 h.31 Given the study duration is sufficiently long to account for the difference in half-life between tissue eosinophils and blood eosinophils, factors other than half-life presumably account for fewer circulating eosinophils and a relatively higher number in tissue. Resident tissue eosinophils, with an immunosuppressive and regulatory role in airway tissue, may account for the cells persisting in tissue, despite very low circulating eosinophils.32 However, it remains unclear whether the remaining eosinophils found in sinonasal tissue of treated patients are inflammatory or resident tissue eosinophils.
The finding of a reduced tissue eosinophil count with mepolizumab is clinically salient because high tissue eosinophil counts are strongly associated with polyp recurrence.33
A study component was the investigation of eosinophil degranulation in view of its purported role in eCRS pathophysiology.34, 35 Activation and degranulation of eosinophils involves the extra-cellular release of pre-formed granules, including cytotoxic proteins (major basic protein, EPX and eosinophil cationic protein), which mediate sinonasal pathology through epithelial damage and fibrosis.35-37 In contrast to cationic protein and major basic protein, EPX is an eosinophil-specific granule protein.36, 38, 39 Interestingly, the current study found no change in EPX staining following mepolizumab. In an asthma study, sputum EPX concentration also did not change in responders to mepolizumab, but was elevated if symptoms worsened post-treatment.40
This study was the first to investigate mepolizumab’s effect on cytokine concentrations in any disease-specific tissue. In line with this study’s finding of increased IL-5 in tissue, increased IL-5 after mepolizumab has been found in bronchiolar lavage fluid12, 13 and plasma of asthma patients.41-43 Sputum IL-5 was also elevated in asthma patients who showed inadequate response to mepolizumab.40 IL-5 signalling impacts on key stages of the eosinophil lifecycle, including mobilisation from bone marrow, migration in circulation, and trafficking to and survival in tissue, which generates tissue eosinophilia.30, 44-47 It is possible that the assay in the current study measures IL-5 bound to mepolizumab. A future direction would involve evaluating the concentration of IL-5 bound to mepolizumab when a commercially anti-mepolizumab antibody becomes available. Elevated levels of IL-5 may also suggest inadequate neutralisation of this key eosinophilic cytokine, and may indicate treatment underdosing, which has been raised as a concern in a previous report.48 A recent study40 found elevated IL-5 immunoglobulin-G immune-complex deposition in the sputa of asthma patients who worsened on mepolizumab therapy – the autoimmune-mediated pathology may have sustained the eosinophilic inflammation and prevented symptom resolution.40
There are several possible explanations for the increase in IL-5 in sinonasal tissue in the current study. First, the functional absence of IL-5 (bound to mepolizumab) and/or reduction in eosinophils may trigger a feedback loop, stimulating local tissue cells known to release IL-5, such as ILC2 cells, Th2 cells, and mast cells.49, 50 A similar feedback loop may also occur systemically, given reports of increased plasma IL-5 with mepolizumab in other eosinophilic diseases.41-43, 51 Second, a decreased rate of elimination of IL-5 from tissue, when bound in a mepolizumab-IL-5 complex compared with free ligand, may be responsible.52-54
The study found a significant increase in the tissue concentration of other type-2 cytokines IL-4, IL-13 and GM-CSF with mepolizumab. Similar to the possible mechanism for IL-5 upregulation after mepolizumab, the decrease in airway eosinophils may have stimulated local production of other type-2 cytokines in a feedback pathway. Increased type-2 cytokines after mepolizumab, including IL-4, IL-13 and GM-CSF, may provide another explanation for the lack of complete removal of blood and tissue eosinophils.12, 13, 55, 56 Cytokines other than IL-5 have an ability, albeit weaker than IL-5, to stimulate bone marrow and local tissue eosinophilopoiesis.12, 26, 45, 57-59 For the first time, the present study provides evidence that, after mepolizumab administration, increased type-2 cytokines in tissue may be partly responsible for continued bone marrow eosinophilopoiesis, via distant signalling, and local tissue eosinophilopoiesis. The study also found increased IL-10 in sinonasal tissue. Evidence suggests that IL-10 acts as an anti-inflammatory cytokine through its inhibition of release of cytokines such as GM-CSF and IL-5, thereby reducing eosinophilic inflammation.60
The upregulation of type-2 cytokine concentrations, particularly the large increase in IL-5 concentration, raises concerns that ceasing mepolizumab might exacerbate type-2 inflammation. In a clinical trial of another anti-IL-5 drug (reslizumab), there was a rebound increase in eosinophil levels following treatment discontinuation.61 In patients treated with an anti-IL-4 and anti-IL-13 drug (dupilumab), a transient increase in blood eosinophils was not associated with disease worsening.62 Therefore, while the reduced level of eosinophils found in the current study may be clinically beneficial, this reduction is not the complete answer to disease control. There has been no rebound effect described for mepolizumab in the 12-month period following its stoppage.63 Nevertheless, monitoring of patients after mepolizumab cessation in further drug trials and clinical practice is advisable in light of the potential risk of recurrent disease.
There was no change in non-type-2 cytokines IL-1β and TNF-α after mepolizumab therapy. TNF-a and IL-B are proinflammatory cytokines that enhance inflammation.64 IL-4 has been reported to suppress IL-1B and TNF-a production.65, 66 Elevated levels of IL-4 and IL-10 may have contributed to the lack of change in IL-1β and TNF-α.
Therefore, it is suggested that in patients with inflammatory airway disease, mepolizumab’s antagonism of IL-5-mediated eosinophilopoiesis and eosinophil trafficking to tissue is responsible for reduced sinonasal eosinophils. The novel finding of upregulation of type-2 cytokines, particularly IL-5, in sinonasal tissue indicates that mechanistic factors, such as a possible mepolizumab-induced feedback loop, may explain incomplete response to therapy, highlighting the need for more data to improve current treatments.