Elevated mucus interleukin-17A levels are associated with increased prior sinus surgery for chronic rhinosinusitis

Abstract

Background: Recent advances in molecular biology have enabled the identification of potential inflammatory endotypes of chronic rhinosinusitis (CRS), with prior work suggesting differential short-term surgical outcome trajectories based on cytokine signatures. However, there is a paucity of data assessing long-term treatment failure and need for revision surgery based on inflammatory biomarkers.

Methods: Retrospective analysis of prospectively collected cross-sectional data from 231 patients electing surgical therapy for CRS. Intraoperative mucus specimens were quantitatively sampled for inflammatory cytokines using a multiplex flow cytometric bead assay. Univariate Spearman correlations between cytokine levels and prior number of surgeries were assessed. A stepwise adjusted multivariate Poisson regression analysis was used to model patient-reported prior sinus surgery counts as a function of cytokine levels.

Results: Several cytokines (interleukin [IL]-1β, IL-4, IL-5, IL-6, IL-8, IL-10, IL-13, IL-17A, tumor necrosis factor α [TNF-α], interferon γ [IFN-γ], and eotaxin) demonstrated significant positive correlations with number of prior surgeries. However, only higher IL-17A levels were independently associated with a higher number of prior sinus surgeries (β = 0.345, p = 0.0003) after adjusting for the significant covariates of age (β = 0.018, p = 0.0036), Lund-Mackay score (β = -0.046, p = 0.02), history of aspirin-exacerbated respiratory disease (β = 1.01, p < 0.0001) and allergic fungal rhinosinusitis (β = 1.08, p < 0.0001). Higher levels of regulated on activation, normal T-cell expressed and secreted (RANTES) were conversely associated with a lower number of prior surgeries (β = -0.17, p = 0.048).

Conclusion: An IL-17A-predominant cytokine profile is linked to an increased number of prior sinus surgeries. Thus, type 3 inflammatory markers may indicate a particularly difficult-to-treat, recalcitrant CRS endotype.

Keywords: IL-17A; chronic rhinosinusitis; cytokines; endotype; sinus surgery; type 3 inflammation.

Figure 1:. Variable Importance Plot for Random Forest Regression Model

Variable importance plot showing the relative importance of each variable as a predictor of the number of previous sinus surgeries in the random forest model. %IncMSE: increasing % mean squared error; TNF-a: Tumor necrosis factor alpha; IFN-g: interferon-gamma; RANTES: Regulated on Activation, Normal T cell Expressed and Secreted.

Figure 2:. Random Forest Partial Dependence Plots

Partial dependence plots for IL-17A, IL-8, IL-6, and RANTES are shown. Note the relatively linear increase in the predicted number of prior surgeries for IL-17A as compared to the non-linear bimodal estimates for IL-6 and IL-8. In contrast, a high number of prior surgeries was only predicted for extremely low values of RANTES.

Figure 3:. Average Number of Prior Surgeries Increase with IL-17A Levels

Histogram of mean number of surgeries grouped by increasing quintiles of mucus IL-17A. A linear trend is observed with increasing prevalence of past sinus surgery with higher IL-17A levels.