Persistent discharge or edema after endoscopic sinus surgery in patients with chronic rhinosinusitis is associated with a type 1 or 3 endotype

Abstract

Background

Patients with chronic rhinosinusitis (CRS) may have persistence of polyps, discharge, or edema after endoscopic sinus surgery (ESS). Inflammation in CRS can be classified into three endotypes, with the presence of polyps associated with the type 2 endotype. Here, we evaluate the endotypic underpinnings of discharge or edema without polyps after ESS.

Methods

At a visit 6–12 months post ESS, patients underwent endoscopy and completed the CRS‐PRO and SNOT‐22. Luminex analysis of middle meatal mucus obtained at that visit was performed for IFN‐γ, ECP, and IL‐17a. Type 1, 2, and 3 endotypes were defined as greater than the 90th percentile expression of each marker, respectively, in controls. Wilcoxon rank‐sum and chi‐squared tests were used to compare cytokine levels and endotype prevalence between those with and without endoscopic findings.

Results

A total of 122 CRS patients completed a clinical exam (median: 8.2 months post ESS). Of the 122 patients, 107 did not have polyps on endoscopy. Of these 107 patients, 48 had discharge, 44 had edema, and 46 had neither discharge nor edema. Compared with those patients without any findings, patients with discharge or edema reported significantly worse severity as measured by CRS‐PRO (10.5 vs. 7.0, p = 0.009; 12.0 vs. 7.0, p < 0.001; respectively), and had higher post‐ESS IFN‐γ, ECP, and IL‐17a. Patients with discharge had higher prevalence of only T1 and T3 endotypes, while patients with edema had higher prevalence of only the T3 endotype.

Conclusions

Post‐ESS discharge or edema in the absence of polyps was associated with higher patient‐reported outcome severity and was more strongly associated with type 1 or 3 inflammation.

1. INTRODUCTION

Historically, chronic rhinosinusitis (CRS) has been phenotypically classified based on the presence of nasal polyps: CRS without nasal polyps (CRSsNP) and CRS with nasal polyps (CRSwNP). However, research has shown that this characterization oversimplifies the complex biochemical underpinnings, pathogenesis, and clinical presentations that represent the spectrum of disease of patients with CRS. Recently, there has been movement to elucidate the distinct pathobiological and inflammatory pathways that lead to the formation of CRS in patients, and to categorize CRS into endotypes based upon specific molecular biomarkers derived from these pathways. ¹ , ² , ³ , ⁴ , ⁵

Chronic rhinosinusitis has been endotyped based on three major kinds of cell‐mediated effector immunity: types 1, 2, and 3 (T1, T2, T3, respectively). ¹ , ⁶ The T1 endotype is categorized by elevated expression of the T1 cytokine IFN‐γ; the T2 endotype is categorized by elevated T2 cytokines (IL‐4, IL‐5, and IL‐13) as well as measures of eosinophil density (eosinophilic cation protein [ECP]); and the T3 endotype is categorized by elevated T3 cytokines (IL‐17A and IL‐17F). ¹ , ⁷ With the development of biologic therapies that target specific inflammatory pathways, endotyping patients with CRS has increasing importance as a mechanism to identify patients likely to benefit from biologic treatments. ⁸ , ⁹ , ¹⁰ Furthermore, the increased precision of pharmacotherapeutics necessitates better understanding of phenotypic predictors of endotypes.

Functional endoscopic sinus surgery (ESS), which attempts to restore physiologic sinus ventilation and drainage, is a major treatment modality for patients who do not improve with appropriate medical therapy. ¹¹ ESS has been found to significantly improve subjective disease measures such as patient‐reported outcome measures (PROMs) and objective metrics such as endoscopic severity, in both the short term and long term, after surgery. ¹² , ¹³ , ¹⁴ Nonetheless, complete resolution of endoscopic findings after ESS is uncommon. In an analysis of CRSwNP patients who had undergone ESS, DeConde et al. found that the recurrence of nasal polyposis after ESS was 35% at 6 months, while presence of edema after ESS was 83% at 6 months. ¹⁵ On a follow‐up endoscopy of CRS patients who had undergone ESS in the prior 3 years, Ryan et al. found that 69% had discharge, 57% had edema, and 24% had polyps. ¹⁶

Most prior studies of CRS have utilized tissue, mucus, or serum taken at the time of surgery to evaluate endotypic inflammation. ¹⁷ , ¹⁸ , ¹⁹ , ²⁰ However, ESS has the potential to affect some of the key endoscopic findings upon which phenotypic‐endotypic associations have been established, making it difficult to extrapolate to the implications of postsurgical findings. By utilizing middle meatal mucus, a sample that is easily obtained in clinic, we can perform a practical approach to postsurgical endotyping of patients and the study of phenotypic‐endotypic associations. As prior research has found a strong association between occurrence of polyps and high levels of T2 inflammation, ¹⁷ , ²¹ we limit our study to patients who do not have presence of polyps after ESS. Here, we seek to evaluate the occurrence rate of post‐ESS edema and discharge in patients without postsurgical polyps, and whether discharge and edema are associated with patient perception of greater disease burden as measured by PROMs. Moreover, we seek to evaluate whether post‐ESS edema and/or discharge is associated with specific inflammatory pathways that can potentially be targeted by future directed biologic therapies.

2. METHODS

2.1. Patient enrollment and follow‐up

Patients with CRS who had undergone ESS at Northwestern Memorial Hospital were prospectively invited to participate in a research evaluation 6–12 months post ESS. All CRS patients met the criteria for the International Consensus Statement on Allergy and Rhinology: rhinosinusitis at the time of ESS. ²² We included a mixed population with some patients undergoing primary and some revision ESS. Patients with an established immunodeficiency, pregnancy, or coagulation disorder or diagnosis of classic allergic fungal sinusitis, eosinophilic granulomatous polyangiitis (EGPA or Churg–Strauss syndrome), or cystic fibrosis were excluded from the study. Patients provided separate written informed consent to access previously collected clinical information and undergo a research‐related examination that included a computed tomography (CT) scan, endoscopy, and administration of PROMs. The Northwestern IRB reviewed this study and approved this protocol (IRB STU000202510).

2.2. Patient‐reported outcome measure and clinician‐derived measures of disease severity

At the postoperative visit, all patients underwent nasal endoscopy, and endoscopic severity scores were recorded by treating rhinologists in the form of modified Lund–Kennedy endoscopic severity scores (MLK). The MLK is composed of subscores ranging from 0 to 2 based on edema, discharge, and polyposis, and we considered a score ≥1 within each subscore to be a positive endoscopic finding. ²³ At this postoperative visit, patients also completed two PROMs; 95% completed the SNOT‐22, while 94% completed the CRS‐PRO. The CRS‐PRO ²⁴ , ²⁵ and SNOT‐22 ²⁶ are validated patient reported sinus‐specific quality‐of‐life measures, with scoring ranges of 0–48 and 0–110, respectively, with higher scores designating worse symptoms. The CRS‐PRO was further evaluated using its three validated subdomains, as was the SNOT‐22 that was divided into its five validated subdomains, as previously described. ²⁷ , ²⁸ Any missing at random PROM datapoints (< 1%) were imputed using the median value for the missed question. Lastly, at this postoperative visit, all patients underwent noncontrast CT scan of the sinuses, which were scored using the Lund–Mackay scoring (LMS) system. ²⁹ , ³⁰

2.3. Mucus acquisition and laboratory analysis of inflammatory mediators

At the time of postoperative visit, prepunched 3/8 inch hydroxylated polyvinyl acetate (Medtronic, Inc) sponges were placed in the middle meatus (MM) for mucus collection and kept for 8 min prior to removal. Samples were kept at −80°C prior to and after processing. The polyvinyl alcohol sponges were then centrifuged at 14,000 rpm for 10 min and extracted with a further 100 μl of phosphate‐buffered saline (PBS) and 1% protease inhibitor cocktail to collect nasal secretions. The concentration of eosinophil cationic protein (ECP) in MM mucus was determined by a commercial ELISA kit (MBL, Woburn, MA) following the manufacturer's instructions. The concentrations of IFN‐γ and IL‐17a in MM mucus were measured using EMD Millipore MILLIPLEX MAP human high‐sensitivity T‐cell panel Luminex kits (Burlington, MA) following the manufacturer's recommended protocols. We specifically focused evaluation on these three biomarkers, as they are well characterized to be associated with each of the three endotypes found in CRS. ¹

2.4. Endotyping and statistical analysis

We recruited 38 patients who did not meet the criteria for CRS and were undergoing septoplasty surgery (Table S1). At the time of surgery, middle meatal mucus was collected from these patients in the same way we collected middle meatal mucus from our CRS patients. We quantified the concentration of IFN‐γ, ECP, and IL‐17 in the mucus of these non‐CRS patients utilizing ELISA and Luminex analysis as previously delineated. We defined a CRS patient as having the type 1, 2, or 3 endotype if the concentration of biomarkers in their mucus was above the 90th percentile expression of IFN‐γ, ECP, and IL‐17a, respectively, in the 38 non‐CRS control patients. This approach to defining CRS endotypes as greater than 90th percentile of control patients for these three biomarkers was previously utilized by Stevens et al. ¹⁷ Statistical analyses were performed using R, Version 4.0.5 (R Foundation for Statistical Computing, Vienna, Austria). Descriptive categorical data is presented as frequency counts and percentages. We assessed variables for normality using Shapiro–Wilk tests. Descriptive continuous data is presented as median and interquartile range (IQR) when non‐normally distributed, and mean and standard deviation (SD) when normally distributed. Wilcoxon rank‐sum test and independent t‐tests were used to analyze differences between the medians and means of two groups, respectively. Pearson's chi square was used to compare differences in categorical data between groups. A two‐tailed p < 0.05 was deemed to be statistically significant for the above analyses.

3. RESULTS

A total of 122 patients who underwent ESS for CRS completed a postsurgical clinic exam 6–12 months after surgery. At the time of surgery, 70 patients (57%) were classified as CRSsNP, while 52 patients (43%) were classified as CRSwNP. Median follow‐up time between ESS and post‐ESS clinic visit was 8.2 months (IQR: 6.7, 10.5). On postsurgical exam, 15 patients (12%) had polyps on endoscopy and were excluded from further analysis because the underpinnings of the polyps were well established. Of the remaining 107 patients, 48 patients (45%) had presence of discharge, 44 patients (41%) had presence of edema, and 46 patients (43%) had no endoscopic findings (Figure 1, Table S2).

FIGURE 1.

Endoscopic findings of chronic rhinosinusitis (CRS) patients who did not have polyps on post‐ESS endoscopy. ESS, endoscopic sinus surgery

3.1. Postoperative discharge

When comparing patients with presence of post‐ESS endoscopic discharge to patients without any endoscopic findings, demographic characteristics and post‐ESS medical management did not significantly differ between the two groups (Tables 1 and 2). Patients with discharge had greater presurgical disease burden as measured by the CRS‐PRO, but did not have greater presurgical disease burden as measured by the SNOT‐22 subjective assessment. Compared with patients without any endoscopic findings, patients with discharge had higher post‐ESS total CRS‐PRO scores (10.5 vs. 7.0, p = 0.009) and Lund–Mackay radiographic scores (5.0 vs. 1.5, p < 0.001). The CRS‐PRO Rhinopsychologic domain and the thematically similar SNOT‐22 Rhinologic and Extra‐Rhinologic subdomains were correspondingly increased in patients with discharge (Table 3). Looking at biomarkers, patients with findings of endoscopic discharge had higher levels of post‐ESS IFN‐γ, ECP, and IL‐17a than patients without any endoscopic findings (Figure 2). Patients with discharge also had higher prevalence of type 1 endotype (13% vs .0%, p = 0.027) and type 3 endotype (43% vs. 17%, p = 0.008). Prevalence of type 2 endotype did not differ between patients with discharge and patients without any findings (20% vs. 12%, p = 0.305).

TABLE 1.

Patient characteristics of patients who had recurrence of endoscopic discharge post ESS (left panel), and endoscopic edema post ESS (right panel) compared with patients without any endoscopic findings

Characteristic	No endoscopic findings, N = 46	Post‐op discharge, N = 48	p‐value	No endoscopic findings, N = 46	Post‐op edema, N = 44	p‐value
Age (years, median, IQR)	50.5 (32.0, 60.5)	46.5 (34.0, 56.5)	0.691	50.5 (32.0, 60.5)	46.5 (35.8, 57.0)	0.765
Sex (female, n, %)	20 (43%)	27 (56%)	0.216	20 (43%)	26 (59%)	0.139
Ethnicity			0.434			>0.999
Hispanic or Latino	1 (2.2%)	2 (4.2%)		1 (2.2%)	1 (2.3%)
Not Hispanic or Latino	42 (91%)	39 (81%)		42 (91%)	41 (93%)
Declined to answer/unknown	3 (6.5%)	7 (15%)		3 (6.5%)	2 (4.5%)
Race			0.139			0.181
Asian	2 (4.3%)	5 (10%)		2 (4.3%)	5 (11%)
Black or African American	2 (4.3%)	5 (10%)		2 (4.3%)	5 (11%)
Native American Indian or Alaska Native	0 (0%)	1 (2.1%)		0 (0%)	1 (2.3%)
White or Caucasian	40 (87%)	31 (65%)		40 (87%)	29 (66%)
Declined to answer/unknown	2 (4.3%)	6 (12%)		2 (4.3%)	4 (9.1%)
Prior surgery	21 (46%)	18 (38%)	0.423	21 (46%)	19 (43%)	0.814
AERD	2 (4.3%)	3 (6.2%)	>0.999	2 (4.3%)	1 (2.3%)	>0.999
Allergy status			0.212			0.857
Either never skin tested, self‐reported negative, or formal‐reported negative	27 (59%)	22 (46%)		27 (59%)	25 (57%)
Self‐reported and formal‐reported positive	19 (41%)	26 (54%)		19 (41%)	19 (43%)
Smoker (prior or current)	9 (20%)	13 (27%)	0.389	9 (20%)	10 (23%)	0.713
Polyp status at the time of ESS (CRSwNP)	18 (39%)	16 (33%)	0.559	18 (39%)	18 (41%)	0.863
Presurgery SNOT‐22 score (median, IQR)	42.0 (27.0, 52.0)	49.0 (37.0, 58.0)	0.105	42.0 (27.0, 52.0)	50.0 (37.0, 59.5)	0.072
Presurgery CRS‐PRO score	23.0 (17.0, 29.0)	29.5 (22.3, 32.0)	0.045	23.0 (17.0, 29.0)	30.5 (22.0, 33.5)	0.035

p‐values from Pearson's chi‐squared test for categorical variables; Wilcoxon rank sum test for continuous variables. Abbreviations: AERD, aspirin‐exacerbated respiratory disease, CRSwNP, CRS with nasal polyps, ESS, endoscopic sinus surgery; IQR, interquartile range. p‐values less than 0.05 are marked in bold.

TABLE 2.

Post‐endoscopic sinus surgery (ESS) management of patients who had recurrence of endoscopic discharge post ESS (left panel) and endoscopic edema post ESS (right panel) compared with those patients without any endoscopic findings

Characteristic	No endoscopic findings, N = 46	Post‐op discharge, N = 48	p‐value	No endoscopic findings, N = 46	Post‐op edema, N = 44	p‐value
Months from surgery to follow‐up	8.95 (7.28, 10.91)	7.82 (6.26, 9.42)	0.052	8.95 (7.28, 10.91)	8.06 (6.40, 9.52)	0.104
Post‐ESS oral corticosteroids a	6 (13%)	9 (19%)	0.450	6 (13%)	7 (16%)	0.699
Post‐ESS antibiotics a	8 (17%)	9 (19%)	0.864	8 (17%)	9 (20%)	0.711
Post‐ESS intranasal corticosteroid sprays b	13 (28%)	14 (29%)	0.923	13 (28%)	14 (32%)	0.713
Post‐ESS intranasal budesonide or mometasone drops or rinses b	44 (59%)	27 (56%)	0.726	40 (51%)	31 (70%)	0.039
Post‐ESS montelukast c	7 (15%)	4 (8.3%)	0.299	7 (15%)	7 (16%)	0.928
Post‐ESS nasal antihistamine c	7 (15%)	7 (15%)	0.931	7 (15%)	5 (11%)	0.591
Post‐ESS oral antihistamine c	11 (24%)	9 (19%)	0.541	11 (24%)	11 (25%)	0.905
Post‐ESS biologic therapy d	3 (6.5%)	3 (6.2%)	>0.999	3 (6.5%)	2 (4.5%)	>0.999

n (%) for categorical variables; median (interquartile range [IQR]) for continuous variables. p‐values from Pearson's chi‐squared test for categorical variables; Wilcoxon rank sum test for continuous variables. p‐values less than 0.05 are marked in bold.

^a Used in past 3 months.

^b Currently using.

^c Used in past month.

^d Three patients without post‐ESS endoscopic were on dupilumab. One patient received 7 doses, one patient received 8 doses, and one patient received 10 doses. One patient with just post‐ESS discharge received four doses of benralizamub. One patient with both post‐ESS discharge and post‐ESS edema received three doses of omalizumab, while another patient with both findings received one dose of benralizumab.

TABLE 3.

Post‐endoscopic sinus surgery (ESS) patient‐reported outcome measures and radiographic severity measures in patients who had recurrence of endoscopic discharge post ESS (left panel) and endoscopic edema post ESS (right panel) compared with those patients without any endoscopic findings

Characteristic	No endoscopic findings, N = 46	Post‐op discharge, N = 48	p‐value	No endoscopic findings, N = 46	Post‐op edema, N = 44	p‐value
SNOT‐22 total score	14.5 (4.8, 22.0)	18.0 (8.0, 30.8)	0.074	14.5 (4.8, 22.0)	22.5 (10.3, 38.5)	0.009
Rhinologic domain	3.5 (1.0, 8.0)	6.5 (3.0, 11.0)	0.008	3.5 (1.0, 8.0)	8.0 (3.0, 13.5)	0.002
Extra‐rhinologic domain	2.0 (0.0, 4.0)	4.0 (2.0, 8.0)	0.001	2.0 (0.0, 4.0)	4.0 (2.0, 9.8)	<0.001
Ear facial domain	2.0 (0.0, 4.0)	2.0 (1.0, 5.0)	0.298	2.0 (0.0, 4.0)	2.0 (1.0, 5.8)	0.13
Psychologic domain	2.5 (0.0, 6.3)	3.5 (0.0, 8.0)	0.661	2.5 (0.0, 6.3)	4.0 (1.0, 9.0)	0.204
Sleep domain	3.0 (0.0, 7.0)	3.5 (0.0, 7.8)	0.876	3.0 (0.0, 7.0)	5.0 (1.0, 10.0)	0.123
CRS‐PRO total score	7.0 (3.0, 12.0)	10.5 (6.0, 18.0)	0.009	7.0 (3.0, 12.0)	12.0 (8.0, 20.8)	<0.001
Rhino‐psychologic domain	4.0 (2.5, 7.5)	8.0 (3.2, 12.0)	0.013	4.0 (2.5, 7.5)	9.0 (6.3, 16.5)	<0.001
Cough domain	2.0 (0.0, 2.5)	2.0 (1.0, 4.0)	0.056	2.0 (0.0, 2.5)	3.0 (1.0, 4.8)	0.005
Facial domain	0.0 (0.0, 1.0)	1.0 (0.0, 2.0)	0.146	0.0 (0.0, 1.0)	1.0 (0.0, 2.0)	0.044
Lund–Mackay score	1.5 (0.0, 3.0)	5.0 (3.0, 10.0)	<0.001	1.5 (0.0, 3.0)	6.0 (3.5, 11.0)	<0.001

Median (interquartile range [IQR]) with p‐values from Wilcoxon rank sum test. p‐values less than 0.05 are marked in bold.

FIGURE 2.

Post‐ESS cytokine levels in patients who had endoscopic discharge post‐ESS compared with patients without any endoscopic findings. Endotype defined as > 90th percentile in controls. p‐values from Pearson's chi‐squared test for categorical variables; Wilcoxon rank sum test for continuous variables. ESS, endoscopic sinus surgery

3.2. Postoperative edema

When comparing patients with presence of post‐ESS endoscopic edema with patients without any endoscopic findings, demographic characteristics did not significantly differ between the two groups (Table 1). Patients with edema had greater presurgical disease burden as measured by the CRS‐PRO but did not have greater presurgical disease burden as measured by the SNOT‐22 subjective assessment. A greater proportion of patients with post‐ESS edema received intranasal budesonide or mometasone drops or rinses after surgery compared with patients without any endoscopic findings (70% vs. 51%, p = 0.039) (Table 2). Compared with patients without any endoscopic findings, patients with edema had higher post‐ESS total SNOT‐22 scores (22.5 vs. 14.5, p = 0.009), CRS‐PRO scores (12.0 vs. 7.0, p < 0.001), and Lund–Mackay radiographic scores (6.0 vs. 1.5, p < 0.001). They also had significantly higher post‐ESS SNOT‐22 Rhinologic and Extra‐Rhinologic subdomain scores, as well as higher CRS‐PRO Rhino‐psychologic, Cough, and Facial pain subdomain scores (Table 3). Looking at biomarker data, patients with findings of endoscopic edema had higher levels of post‐ESS IFN‐γ, ECP, and IL‐17a than patients without any endoscopic findings (Figure 3). Patients with edema also had higher prevalence of type 3 endotype (44% vs. 17%, p = 0.007). Prevalence of type 1 and type 2 endotype did not differ between patients with edema and patients without any findings (12% vs. 0%, p = 0.055; 17% vs. 12%, p = 0.562; respectively).

FIGURE 3.

Post‐ESS cytokine levels in patients who had endoscopic edema post‐ESS compared with patients without any endoscopic findings. Endotype defined as > 90th percentile in controls. p‐values from Pearson's chi‐squared test for categorical variables; Wilcoxon rank sum test for continuous variables. ESS, endoscopic sinus surgery

Due to the high prevalence of patients with concomitant endoscopic findings, we also analyzed the n = 31 (29%) patients with both edema and discharge on post‐ESS endoscopy (Table S3). When comparing patients with presence of both post‐ESS endoscopic edema and discharge with patients without any endoscopic findings, demographic characteristics, presurgical disease burden, and post‐ESS medical management did not significantly differ between the two groups. Compared with patients without any endoscopic findings, patients with both edema and discharge had greater disease burden as measured by both the CRS‐PRO and the Lund–Mackay radiographic score. Looking at biomarker data, patients with findings of endoscopic edema and discharge had higher levels of post‐ESS IFN‐γ, ECP, and IL‐17a than patients without any endoscopic findings, but only had greater prevalence of the type 1 and type 3 endotype.

4. DISCUSSION

In this study, we evaluated the prevalence, symptomatic associations, and endotypic underpinnings of post‐ESS edema and discharge in patients without postsurgical polyps. Using a group of 107 CRS patients without polyps on post‐ESS endoscopy, we found that 45% (n = 48) had presence of discharge, 41% (n = 44) had presence of edema, and 43% (n = 46) had no endoscopic findings. The presence of edema and discharge was more frequent than findings of polyps at this interval after surgery. Even without the presence of polyps, patients with presence of discharge and edema still had higher LMS compared with patients without any endoscopic findings. Patients with discharge also had worse patient‐reported disease burden as measured by the CRS‐PRO, while patients with edema had worse disease burden as measured by both the SNOT‐22 and the CRS‐PRO. Patients with discharge and edema did have pan‐elevated inflammatory markers compared with patients without endoscopic findings. However, when utilizing our endotype categorization, we find that patients with discharge had higher prevalence of only the T1 and T3 endotypes, while patients with edema had higher prevalence of only the T3 endotype.

Our finding that patients with discharge have worse disease burden as measured by the LMS aligns with a previous study which found that the Lund–Kennedy discharge subscores in post‐ESS patients correlated with their LMS (r = 0.80, p < 0.001). ¹⁶ In that study, Ryan et al. found that 69% of patients had presence of discharge on post‐ESS endoscopy. The higher prevalence of discharge in that study compared with ours (69% vs. 45%) may be attributable to the greater follow up time (1.2 years vs. 8 months) and the ensuing reemergence of disease. While the authors in that study found poor correlation between the SNOT‐20 PROM and endoscopic findings in general (r = 0.03, p = 0.42), we found that the CRS‐PRO PROM was associated with endoscopic findings, as patients with the presence of discharge or edema had higher CRS‐PRO scores compared with those without any endoscopic findings (10.5 vs. 7.0, p = 0.009; 12.0 vs. 7.0, p < 0.001; respectively). We similarly found that the SNOT‐22 total score was higher among patients with edema. We hypothesize the differences between our study and Ryan et al.’s study are due to the increased representation of rhinologic symptoms and disease‐attributable psychologic items on both the SNOT‐22 and especially the CRS‐PRO compared with the SNOT‐20.

Previously, DeConde et al. reported a post‐ESS edema occurrence rate in CRSwNP patients of 83% at 6‐month, 82% at 12‐month, and 78% at 18‐month follow‐up endoscopy exams. ¹⁵ Even when we limit our study to all those that were diagnosed with CRSwNP at baseline, we still find that edema occurred in only 52% of patients at a median of 8.2 months. In a multivariate regression model, the authors found that only a history of previous ESS was found to be associated with edema recurrence (OR = 4.5, p < 0.001). ¹⁵ It is possible that the lower edema rate found in our study can be attributable to the lower percentage of patients undergoing revision surgery (41% in our population vs. 65% in their population). Alternatively, the lower rate of edema in our cohort may be attributable to the large proportion of patients who received intranasal budesonide or mometasone drops or rinses after surgery, compared with the abovementioned cohort which received nasal saline rinses and medical therapeutics as necessary. Our finding that patients with discharge or edema had higher levels within only two of the five SNOT‐22 subdomain (Rhinologic and Extra‐Rhinologic), aligns with another study by DeConde et al., which found that edema and discharge are most closely correlated with sinonasal‐specific symptoms. ³¹ In that study, they found a significant correlation between edema and discharge and the SNOT‐22 Rhinologic domain (r = 0.28, r = 0.29, respectively), and between discharge and the Extra‐Rhinologic domain (r = 0.19), but did not find that endoscopic findings correlated with any other domain. Interestingly, we found that patients with edema had worse subjective disease burden as measured by the SNOT‐22 and CRS‐PRO than those with discharge and performed worse on the Cough and Facial domains of the CRS‐PRO. Further study is necessary to elucidate how edema may uniquely be contributing to worse subjective perception of disease severity.

Identifying biologic therapies that target specific inflammatory pathways of CRS has become a subject of great research interest given the availability of precision biologics for certain forms of CRS. ¹⁰ As a consequence, there has been movement to categorize CRS into endotypes based upon specific molecular biomarkers derived from the distinct pathobiological and inflammatory pathways that lead to the disease. ¹ , ² , ³ , ⁴ , ⁵ The association between nasal polyps and type 2 inflammation has been well established, and it is known that in western countries at least 80% of CRSwNP patients exhibit a type 2 inflammatory endotype. ² Studies have also shown the T2 endotype and histopathologic eosinophilia are associated with polyp recurrence after ESS. ³² , ³³ , ³⁴ Moreover, there are now three biologics approved for CRSwNP. ³⁵ , ³⁶ , ³⁷ However, our results show that endoscopic findings of discharge and/or edema without the presence of polyps after ESS still are associated with greater disease burden even when these patients do not have postsurgical polyps. Our study intriguingly found that patients with discharge had higher prevalence of T1 and T3 endotypes (p = 0.027, p = 0.008), while patients with edema had higher prevalence of T3 endotypes (p = 0.007). While our T2 biomarker ECP was also significantly elevated in patients with edema and discharge relative to those with no endoscopic findings, we found the degree of elevation not significantly different using our non‐CRS control‐based endotyping method. The primary physiologic function of T1 immunity is protection against intracellular microbes, including viruses, bacteria, and protozoa, while T3 immunity is devoted to protection against extracellular bacteria and fungi. ⁶ Pathologically, both T1 and T3 immunity are believed to play a role in several autoimmune diseases including rheumatoid arthritis and multiple sclerosis. ⁶ While not considered a targeted therapy, corticosteroids have been well studied in CRS and are known to inhibit T1 inflammation. ³⁸ , ³⁹ In regard to T3 inflammation, two inhibitors of IL‐17Rα and IL‐17α, brodalumab and secukinumab, have been approved for treating psoriasis and psoriatic arthritis but have not been well studied in CRS. ¹⁰ Further research is needed on the role that these therapies may play in treating the T1‐ and T3‐driven manifestations of CRS.

We recognize that this study has several limitations. Firstly, utilization of mucus rather than tissue is still a nascent area of practice in CRS research despite potential advantages in its utility for following disease outside the operating room. Prior studies have found that cytokine levels in sinus mucus correlate with levels in sinus tissue, that specific cytokines are elevated in CRS mucus versus control mucus, and that several cytokines found in mucus are associated with objective olfactory function. ⁴⁰ , ⁴¹ At the same time, there remains concern that biomarkers in mucus may not be sensitive enough to properly endotype all patients, or may be more variable even in patients without CRS, thus under‐reporting specific endotypes. This is compounded by the fact that inflammatory mediators were measured only once rather in duplicate or triplicate due to low volumes of mucus. As patients of five different surgeons were analyzed, there may be interoperator variability in endoscopic scoring, although there were strong associations between the endoscopic measures with radiographic, PROM, and biomarker severity despite rating physicians being unaware of the biomarker results at the time of evaluation. Additionally, intra‐ and post‐ESS management was left to the discretion of treating physician, which may confound outcomes, as patients may have been on medical treatment at the time of endoscopy. Finally, our endoscopic and endotyping visit took place at an intermediate term (6–12 months) after ESS, which is shorter than other post‐ESS follow‐up studies that have previously been published. On the other hand, this follow up time is similar in length to many of the studies of biologics for CRS.

5. CONCLUSION

Patients with post‐ESS discharge in the absence of postsurgical polyps had worse disease burden as measured by the LMS and CRS‐PRO and had higher prevalence of the T1 and T3 endotypes. Patients with edema and no polyp occurrence had worse disease burden as measured by the LMS, SNOT‐22, and CRS‐PRO and had higher prevalence of the T3 endotype.

FUNDING

This work was supported by NIH grants: R01 AI134952, R01 DC016645, T32 AI083216, and the Chronic Rhinosinusitis Integrative Studies Program 2 (CRISP2) P01 AI145818.

CONFLICT OF INTEREST

D. B. Conley reports consulting fees for Intersect ENT and XORAN. K. C. Welch reports consultant fees from Baxter, OptiNose, and Acclarent. A. Kato reports a consultant fee from Astellas Pharma and a gift for his research from Lyra Therapeutics. B. K. Tan reports personal fees from Sanofi Regeneron/Genzyme.

Presentation: Accepted for oral presentation at the 2022 American Rhinologic Society Spring Meeting at COSM, Dallas, TX, April 28‐29, 2022.

Supporting information

Supporting Information

Stein E, Schneider AL, Harmon R, et al. Persistent discharge or edema after endoscopic sinus surgery in patients with chronic rhinosinusitis is associated with a type 1 or 3 endotype. Int Forum Allergy Rhinol. 2023;13:15–24. 10.1002/alr.23042