Endotyping Difficult-to-Treat Chronic Rhinosinusitis with Nasal Polyps by Structured Histopathology Open Access

Chronic rhinosinusitis (CRS), affecting approximately 14% of adults in the USA, 10.9% in Europe, and 8% in China, represents a chronic inflammatory condition of the nose and paranasal sinus cavities with significant morbidity and major impact on the quality of life of affected subjects [1]. Based on the presence or absence of nasal polyps (NPs), CRS is usually classified into two phenotypes: CRS without nasal polyps (CRSsNP) and CRS with nasal polyps (CRSwNPs). Although only ∼25–30% of patients with CRS develop NPs, patients with CRSwNP often have greater severity of clinical disease, impairment of quality of life, and cost burden compared to those with CRSsNP [2‒4], making it more clinically challenging to otolaryngologist.

Endotypes of CRSwNP refer to the distinct subtypes of the disease based on underlying molecular and immunological mechanisms. Identifying these endotypes is crucial for selecting appropriate targeted therapies and improving patient outcomes. CRSwNP endotypes are primarily classified according to the inflammatory pathways involved, such as T1, T2, and T3 inflammation. Eosinophilic inflammation, which is commonly associated with T2 responses, is a key feature of CRSwNP in many cases. The prevalence of T2 signatures in CRSwNP patients varies across different regions, with higher proportions observed in Western countries compared to Asian countries [5, 6].

Currently, the mainstays of treatment for patients with CRSwNP have involved a combination of endoscopic sinus surgery (ESS) and medical management. Based on current international guidelines, the goal of CRSwNP management is to control clinical symptoms rather than cure the disease [1, 7, 8]. As such, outcome measures are particularly important for improving patient-centered care in order to provide better health care delivery. According to the European Position Paper on CRS (EPOS) guideline, difficult-to-treat CRS is defined as a patient who do not achieve an acceptable level of disease control despite adequate surgery, intranasal corticosteroid treatment, and up to 2 short courses of systemic corticosteroids in the last year [1, 7]. Based on this definition, we and others have revealed that around 30∼40% of patients were difficult-to-treat CRSwNP [9‒12], reflecting our limited knowledge of CRSwNP pathogenesis.

Numerous studies have attempted to identify the risk factors associated with CRSwNP poor treatment outcome. For example, clinical factors, such as asthma, aspirin intolerance, increased blood eosinophils, and high sinus computed tomography (CT) scores, were reported to be associated with an increased risk of recurrence or refractoriness in CRSwNP [11‒13]. Inflammatory biomarkers, such as interleukin-5, immunoglobulin E, cystatin SN, and Charcot-Leyden crystals (CLC), were also found to link to poor treatment outcomes of CRSwNP [11, 12, 14, 15, 16, 17]. Histopathologically, CRSwNP with poor treatment outcome has been shown to be associated with several inflammatory and remodeling parameters in polyp tissue, such as tissue eosinophilia, thickened epithelial basement membrane, goblet cell hyperplasia, and total inflammatory burden [18‒21]. However, most of the current evidence is limited by their retrospective designs and small sample size, as well as inconsistent definitions of poor treatment outcome. Structured histopathology is a systematic approach to analyzing tissue samples, using standardized templates for consistent reporting. This method enhances diagnostic accuracy, clinical communication, and treatment decision-making [22]. Increasingly, use of structured histopathology reporting in CRSwNP has been suggested to better understand the unique pathophysiology and inflammatory milieu of a given CRSwNP patient in order to individualize care [21, 22]. However, literature comprehensively exploring the utility of structured histopathology in characterizing difficult-to-treat CRSwNP remains scarce. In this study, we investigate the differences in structured histopathologic parameters between difficult-to-treat and nondifficult-to-treat CRSwNP. Through this work, we seek to further define histopathological characteristics of difficult-to-treat CRSwNP disease. These findings may help individualized disease management, ultimately optimizing patient outcomes.

Patients (≥16 years) with diffuse (bilateral) CRSwNP who underwent ESS in the First Affiliated Hospital of Sun Yat-sen University were prospectively recruited. CRSwNP was diagnosed according to the EPOS guideline [1, 7].

A prospective study was conducted. A study flowchart is presented in Figure 1. The study was part of clinical trial no. ChiCTR-TRC-14004375 and was approved by the Ethics Committee of the First Affiliated Hospital of Sun Yat-sen University. All patients provided informed consent before enrollment. Patients with bilateral CRSwNP undergoing ESS were recruited consecutively from the First Affiliated Hospital of Sun Yat-sen University, which is a tertiary academic hospital. CRSwNP patients were defined based on the criteria set out by the EPOS guideline [1, 7]. Patients with the following items were excluded: (1) Patients receiving systemic and intranasal glucocorticoid treatment within 3 months or 1 month before surgery, respectively; (2) patients with unilateral CRSwNP, fungal rhinosinusitis, cystic fibrosis, antrochoanal polyps, primary ciliary dyskinesia, gastroesophageal reflux disease, immunodeficiency, or sinonasal tumors.

All patients underwent ESS after failed medical therapy. ESS was performed under general anesthesia. Briefly, the extent of surgery was tailored to the extent of the sinus disease based on the findings by preoperative CT. Basically, polyp removal and anterior ethmoidectomy was mandatory, but no instructions regarding the further extent of ESS, such as full-house ESS, were given. Other procedures included posterior ethmoidectomy, sphenoidotomy, Draf IIa/b, septoplasty, and turbinate reduction.

At baseline, the atopic status was evaluated by using assays for specific IgE (HOB Biotech Group, Suzhou, China) against the local common inhalant allergens. Specific IgE concentrations above 0.35 IU/mL were considered positive. The blood eosinophil numbers in total white blood cells were detected by blood routine test. The diagnosis of allergic rhinitis (AR) was made according to the Allergic Rhinitis and its Impact on Asthma guideline [23]. The diagnosis of asthma was performed by a specialist physician and was established according to the Global Initiative for Asthma 2006 guideline [24]. As previously described, subjective symptoms including nasal congestion, anterior rhinorrhea, postnasal drip, and loss of smell, each evaluated using a scale of 0 = none, 1 = mild, 2 = moderate, or 3 = severe, were scored and calculated (range: 0–12) by adding up based on the sum of these scores (total nasal symptom score, TNSS) [1, 7]. Endoscopic findings were evaluated with the modified Lund-Kennedy (L-K) scoring system, which grades visual pathologic states within the sinuses including polyps, discharge, and edema (range: 0–12) [25]. CT scans were scored (range: 0–24) according to Lund-Mackay (L-M) CT scoring system and total bilateral CT scores were calculated by the sum of the scores of each side [26].

The postoperative regimen mainly included saline irrigations, intranasal and systemic corticosteroids, and antibiotics needed to control disease. The length of treatment after ESS was individualized based on patient’s symptoms and post-surgical endoscopic appearance. A duration of 3 months treatment was essential for disease control after surgery. For those with persistent symptoms and inflammation under endoscopy, a prolonged medication, usually 6 months or more, was prescribed.

All patients were followed up at 12–15 months after ESS to evaluate the treatment outcomes, including TNSS, nasal endoscopy for E-score [27], and clinical disease control level. The disease control of CRS was evaluated based on the classification criteria of EPOS 2012 [1], which was divided into three levels: controlled, partly controlled, and uncontrolled. Controlled CRS was defined as no bothersome symptoms, with healthy or almost healthy mucosa, and no need for systemic medicine to control the disease. Partly controlled patients experienced less than 2 of the following items: persistent nasal blockage, mucopurulent rhinorrhea/postnasal drip, facial pain, impaired smell, sleep disturbance/fatigue, disease mucosa under nasal endoscopy, and the need for a course of antibiotics/systemic corticosteroids in the last months. Uncontrolled CRS was defined as three or more features of partly controlled CRS. Patients who did not reach an acceptable level of control (controlled or partly controlled) despite ESS, intranasal corticosteroid treatment, and up to two short courses of antibiotics or systemic corticosteroids in the last year were considered to have difficult-to-treat CRSwNP.

NP tissues of all participants were obtained during sinus surgery, fixed in 10% formalin, and embedded in paraffin. Samples were cut into 4-μm thick sections and routinely stained with hematoxylin-eosin. The stained sections were observed under a microscope (Leica DM4 B; Leica, Wetzlar, Germany) in a blinded fashion concerning all clinical data by two independent observers. Five areas were selected for each specimen for observation: the top 5 densest cellular infiltrate regions of the subepithelial layer were first chosen under a low power field (100×). Then the general structural characteristics of the tissues (200×) were observed in the selected regions. Finally, the cell counting (400×) was performed.

In line with previously established standards for structured histopathology [21, 22, 28, 29], we assessed the following histopathological parameters: epithelial hyperplasia, goblet cell hyperplasia, squamous metaplasia, basement membrane thickening, subepithelial edema, fibrosis, gland number, total number of inflammatory cells, eosinophil infiltration, eosinophil aggregates, and CLC formation. According to the number of layers of the epithelial layer, epithelial hyperplasia was divided into none (<3 layers), mild (3–5 layers), moderate (5–8 layers), and severe (>8 layers) (Fig. 2a–d). According to the proportion of squamous metaplasia cells occupying the layer of epithelial cells, squamous metaplasia was divided into none (no squamous cell metaplasia), mild (degree of squamous cell metaplasia <50%), moderate (degree of squamous cell metaplasia 50–75%), and severe (degree of squamous cell metaplasia >75%) (Fig. 2e–h). According to the degree of subepithelial edema, it was divided into none, mild (focal edema, degree <10%), moderate (degree of edema 10–50%), and severe (diffuse edema, degree >50%, or polypoid change) (Fig. 2i–l). According to the degree of interstitial fibrin deposition, fibrosis is divided into mild (fibrosis <50%), moderate (fibrosis 50–90%), and severe (fibrosis >90%) (Fig. 2m–o). The above indicators used a magnification of ×200. At a magnification of ×400, goblet cells (Fig. 3a), the total number of inflammatory cells, eosinophils, and glands were counted (Fig. 3a, b). Intraepithelial inflammatory cells and those inside vessels were not counted. Eosinophil aggregates were defined as one or more groups of >20 eosinophils/high power field within the lamina propria [21] (Fig. 3c). Eosinophil aggregates and CLC formation (Fig. 3d) were categorized as presence or absence. Except for eosinophil aggregates and CLC, all other parameters were analyzed by averaging five selected fields for final statistical analysis.

The prevalence of chronic sinusitis in China is about 8% using PASS 15 software for sample size estimation. A sample size of 133 produces a two-sided 95% confidence interval with a width equal to 0.09979 when the sample proportion is 0.08.

Statistical analyses were performed with SPSS version 26.0 statistical software (IBM SPSS, Armonk, NY, USA). T tests were used for continuous variables with normal distribution. Mann-Whitney U tests were used for continuous non-normally distributed variables. χ² tests were used to assess differences in categorical variables. A multiple logistic regression models were established to estimate the association between histopathological parameters and the uncontrolled outcome. A p value of less than 0.05 was considered statistically significant.

Between January 2015 and December 2018, a total of 265 subjects were diagnosed with CRSwNP according to the EPOS guideline and participated into the study. Seventy patients were excluded after inclusion. At the end of the study, twenty-one patients were lost to follow-up for geographical reasons. As a result, a total of 174 subjects were included in the per-protocol analysis. The general characteristics of the cohort are presented in online supplementary Table S1 (see online suppl. material at https://doi.org/10.1159/000530864). The mean age was 39.0 ± 13.4 years. One hundred and fifteen (66.1%) patients were male. Seventeen (9.7%) patients were current smokers. Eighteen (12.0%) patients had AR, twenty one (10.3%) had asthma, and fifty one (31.0%) had revision surgery. The preoperative L-K endoscopic scores and L-M CT scores were 9.0 (interquartile range [IQR], 3.0) and 15.0 (IQR, 8.5), respectively. The preoperative peripheral blood eosinophil count and percentage were 0.23 × 10⁹/L (IQR, 0.3 × 10⁹/L) and 3.0% (IQR, 4.73%), respectively. The endoscopic E-score and TNSS at 12–15 months after ESS were 8.0 (IQR, 5.0) and 2.0 (IQR, 3.0), respectively.

Based on the EPOS criteria for the definition of difficult-to-treat CRSwNP, 49 of 174 CRSwNP patients (accounting for 28.2%) were diagnosed as difficult-to-treat CRSwNP. Baseline clinical characteristics of patients with and without difficult-to-treat CRSwNP were shown in Table 1. There were no significant differences in terms of gender, age, proportions of current smoker, and prior sinus surgery between the patients with or without difficult-to-treat CRSwNP. Compared to the nondifficult-to-treat CRSwNP, the patients with difficult-to-treat CRSwNP had higher proportions of patients with AR and asthma. In addition, the preoperative L-M CT scores, peripheral blood eosinophil numbers, and post-operative TNSS were significantly higher in the uncontrolled group. Interestingly, we did not observe significant differences in preoperative L-K endoscopic score and post-operative E-score between the two groups.

Compared with the nondifficult-to-treat CRSwNP, the total number of inflammatory cells, the number and proportion of tissue eosinophils, percentages of eosinophil aggregates, and CLC formation was significantly increased, but the number of interstitial glands was decreased, in the difficult-to-treat CRSwNP (Table 2). In addition, the number of goblet cells in the difficult-to-treat CRSwNP also showed an increase trend compared to the nondifficult-to-treat CRSwNP (p = 0.095). There were no significant differences in epithelial hyperplasia, squamous metaplasia, basement membrane thickening, and subepithelial edema between difficult-to-treat and nondifficult-to-treat CRSwNP.

We further analyzed the association between histopathological parameters and the difficult-to-treat CRSwNP by establishing a multiple logistic regression model. The results showed that increased inflammatory cell infiltration, tissue eosinophil count, eosinophil aggregates, and CLC formation were independent risk factors for the uncontrolled disease status (Table 3). Moreover, the CRSwNP patients with tissue eosinophil aggregation and CLC formation had an increasingly higher likelihood of uncontrolled disease versus those with tissue eosinophilia (Table 3, p for trend <0.001). These results remained statistically significant after being introduced simultaneously and adjusted for age, gender, smoking, prior sinus surgery, AR, and asthma (Table 3).

Based on our observation, along with the above-structured histopathological analysis, we found that, as shown in Figure 4, the difficult-to-treat CRSwNP was histopathologically characterized by intense inflammatory infiltrates (Fig. 4a), abundance of eosinophils (Fig. 4b), and eosinophil aggregates (Fig. 4c), as well as CLC formation (Fig. 4d), whereas the nondifficult-to-treat CRSwNP was low inflammatory infiltration with absence of eosinophil aggregation and CLC formation (Fig. 4e) and increased gland abundance (Fig. 4f).

Endotyping allowing for the identification of CRS patients with a high likelihood of unsuccessful treatment has gained increased importance and therefore this area is now under active investigation [30‒32]. Prior studies and our current study have consistently demonstrated that approximately 30∼40% of patients were difficult-to-treat CRSwNP according to the EPOS criteria despite guideline-based therapies [9‒12], suggesting the endotype of the difficult-to-treat CRSwNP needs be better described.

Numerous previous studies have explored the clinical and immunopathological features associated with difficult-to-treat CRSwNP [33‒35]. However, few studies have looked at the relationship between structured histopathological profiling and EPOS-defined difficult-to-treat CRSwNP. In the present study, we performed a large prospective cohort study to analyze the association of structured histopathological parameters measured in CRSwNP tissues obtained during ESS with the EPOS-based disease state at 1-year post-operation and found that difficult-to-treat CRSwNP was characterized by high inflammatory load, abundance of eosinophils and eosinophil aggregates, and presence of CLC formation. These results corroborate the findings of previous studies [19, 36] and suggest that measurement of structured histopathological profiling in CRSwNP tissues might add prognostic and predictive values to traditional clinical risk factors in evaluation for CRSwNP patients. To our knowledge, this is the largest study that has prospectively examined the association between the structured histopathological profiling and the treatment outcomes in patients with CRSwNP and is the first study to examine these associations in relation to the EPOS-defined difficult-to-treat CRSwNP.

It has been recognized that tissue or blood eosinophilia and presence of comorbid asthma are major phenotypes associated with poorly controlled disease and polyp recurrence in patients with CRSwNP [10, 13, 37]. Soler et al. [38] have shown that tissue eosinophilia correlated with worse disease severity on CT, endoscopic findings, and smell function test and was associated with less improvement in both disease-specific and general quality of life [39]. Nakayama et al. [40] and Lou et al. [15] presented a strong association between tissue eosinophil numbers and polyp recurrence. Recently, we further demonstrated that blood and tissue eosinophilia are independent risk factors for poor disease control in patients with CRSwNP receiving current standard-of-care therapy after adjusting for many demographics, comorbidities, and characteristics previously shown to relate to poor treatment outcome [12]. In the current study, our analysis extended previous studies and showed that tissue eosinophilia (adjusted OR: 1.005), eosinophil aggregation (adjusted OR: 3.536), and CLC formation (adjusted OR: 6.972) were differentially and independently associated with the difficult-to-treat outcome in patients with CRSwNP. Moreover, it has been reported that CLC formation is attributed to activated eosinophils undergoing extracellular trap cell death and loss of galectin-10 cytoplasmic localization, which is closely related to the development of allergic and eosinophilic inflammation [41, 42]. In our study, eosinophil activation was demonstrated by the histopathological presence of eosinophilic aggregates and CLC formation. Therefore, our findings showing the risk of difficult-to-treat outcome exhibited an increasing trend in accordance with the degree of eosinophil activation/degranulation (Table 3, p < 0.001 for trend) suggest that patients with tissue eosinophilic aggregates and CLC formation might represent a more severe eosinophilic endotype requiring closer follow-up and adjunctive therapy to control disease. By contrast, patients with tissue eosinophilia but without eosinophilic aggregates and CLC formation may likely represent a less severe form of eosinophilic inflammation that does not require aggressive therapy and follow-up.

Our study also confirms prior reports showing that the tissue inflammatory load indicated by the increased total number of inflammatory cells is closely related to the uncontrolled disease state in patients with CRSwNP [19, 43]. This result suggests that, in addition to tissue eosinophilic inflammation, the degree of total inflammatory load would provide additional prognostic information and therefore should be evaluated when examining the histopathology of CRSwNP.

Basement membrane thickening is a sign of epithelial remodeling. Previous studies have reported that CRSwNP had significant basement membrane thickening than CRSsNP [44]. In addition, the patients with basement membrane thickening had more impaired quality of life [45] and more relied on topical corticosteroids [46]. These reports suggest that clinical symptoms or disease severity may be worse in CRSwNP patients undergoing epithelial remodeling. However, whether basement membrane thickening is related to treatment outcomes remains unclear. In this study, we did not observe significant difference in basement membrane thickness between patients with and without difficult-to-treat CRSwNP, suggesting that basement membrane thickening may not be related to the EPOS-defined prognosis of CRSwNP. However, we did not include patients with CRSsNP in the cohort, which may result in selection bias. Further studies with sufficiently heterogenous CRS patients will need to be evaluated in future work.

Recently, artificial intelligence, particularly deep learning algorithms, has shown great promise in pathology image analysis. As shown in our recent studies [47, 48], deep learning-based networks, such as the artificial intelligence CRS evaluation platform, can be less expensive and faster than other histopathological and laboratory assays for providing prognostic information and therefore has the potential to improve clinical decision-making for the treatment of CRSwNP patients.

There are several limitations in this study. First, the patients in the current study were all recruited from a tertiary academic hospital and may not be representative of other medical care settings. Second, our study primarily focused on Chinese patients and the conclusions cannot be generalized to the Western population, particularly given the fact that the prevalence of type 2 inflammation in CRSwNP is higher in the Western world compared to the Chinese population [5, 6]. Third, as stated above, we did not include patients with CRSsNP in the study. It is still unclear whether our findings can be extrapolated to the general CRS patients. A well-designed prospective study with various types of CRS patients from diverse populations should be performed to further validate the findings of this study. Fourth, the analyses were based on a single time-point measurement (12–15 months post-operation). It is important to acknowledge that the prevalence of difficult-to-treat CRSwNP might change as the effectiveness of ESS could potentially decrease over time [49, 50]. Therefore, further studies with longer follow-up times are warranted. Fifth, the primary outcome measure in our study focused on evaluating the level of disease control in accordance with the EPOS criteria. Further study incorporating patient-reported outcome measures, such as the validated 22-item Sinonasal Outcome Test score, would provide a more comprehensive understanding of difficult-to-treat CRSwNP. Finally, while our findings demonstrate the value of understanding the histopathologic characteristics of difficult-to-treat CRSwNP, it is important to note that our study focused on polyp specimens collected during surgery. Our current study did not address the feasibility or the benefits of preoperative biopsy in the clinic. Performing biopsies in the clinic before surgery may involve additional risks and challenges, such as patient discomfort, invasiveness, and the availability of appropriate facilities. Further studies are needed to explore the feasibility and benefits of preoperative biopsy in the clinic.

In summary, we have shown significant differences in structured histopathological characteristics between patients with and without difficult-to-treat CRSwNP. The EPOS-defined difficult-to-treat CRSwNP appears to be characterized by increased total inflammatory infiltrates, tissue eosinophilia, eosinophil aggregates, and CLC formation in structured histopathology. These results are expected to serve as a basis for assessing the CRSwNP prognosis of using structured histopathology, thereby facilitating the development of individualized treatment regimens for patients.

The study was part of clinical trial no. ChiCTR-TRC-14004375 and the study protocol was reviewed and approved by the Ethics Committee of the First Affiliated Hospital of Sun Yat-sen University, approval number [2014]31. Written informed consent was obtained from all participants. For the participants under age of 18, written informed consent was obtained from the participants’ parent to participate in the study.

All authors declare no competing financial interests.

This study was supported by grants from the National Natural Science Foundation of China (81873691, 81970854, and 82171105), Shenzhen Fundamental Research Program (JCYJ20220530144805011), and Guangdong Basic and Applied Basic Research Foundation (2023A1515012497).

All authors were involved in the study. Yueming Cui and Kanghua Wang: case collection, histopathological analysis, patient follow-up, data entry, statistical analysis, and paper writing; Kanghua Wang: assisting in case collection; and Jianbo Shi and Yueqi Sun: project leader, financial support, and manuscript revision.

Yueming Cui and Kanghua Wang contributed equally to this work. Edited by: H.-U. Simon, Bern.

All data analyzed during this study are included in this article and supplementary materials. Further inquiries can be directed to the corresponding author upon reasonable request.