Endoscopic Sinus Surgery Improves Aspirin Treatment Response in AERD Patients

Abstract

Background

Aspirin desensitization and treatment benefits most patients with aspirin exacerbated respiratory disease (AERD) though some fail therapy. Our objective was to assess whether recent endoscopic sinus surgery (ESS) improved aspirin treatment outcomes in AERD patients who initially failed aspirin therapy.

Methods

Outcomes of aspirin desensitization and treatment in AERD patients prospectively enrolled were assessed pre-operatively and at 4, 12, and 24 weeks after ESS by determining changes in the Asthma Control Test (ACT), Rhinoconjunctivitis Quality of Life Questionnaire (RQLQ) scores, and respiratory function. Biomarkers, including Fractional excretion of Nitric Oxide (FeNO), spirometry, nasal inspiratory peak flow (NPF), IgE, and eosinophil count, were measured.

Results

19 patients who benefited (responders) and 21 patients who failed (non-responders) preoperative aspirin treatment with a distant history of ESS (mean 48 months) were identified. Non-responders were more likely to be African-American (71%, p<0.01) and have higher baseline IgE levels (252 kU/L vs. 87 kU/L in responders, p<0.01). After ESS, all non-responders could tolerate a second round of aspirin desensitization and treatment. While the primary aspirin therapy was associated with a significant increase in IgE in non-responders, the second aspirin desensitization was more successful when attempted 3–4 weeks after ESS and there was no significant increase in IgE.

Conclusions

Antecedent ESS enhances aspirin treatment responses in AERD patients and may convert patients who failed aspirin treatment prior to surgery to a more responsive phenotype after ESS. Patients with higher baseline serum IgE levels may benefit from ESS performed shortly before aspirin desensitization and therapy.

INTRODUCTION

Aspirin Exacerbated Respiratory Disease (AERD), also known as Samter’s Triad, is a chronic condition that manifests as asthma, chronic rhinosinusitis with nasal polyposis, and hypersensitivity reactions such as bronchospasm, rhinorrhea, and conjunctivitis to aspirin and other non-steroid anti-inflammatory drugs (NSAIDs). AERD affects 7% of patients with asthma and 15% of patients with severe asthma¹.

Although the understanding of its pathogenesis is limited, AERD is associated with extensive eosinophilic airway inflammation and mast cell degranulation mediated by a type 2 inflammatory response.^2,3 Endoscopic sinus surgery (ESS) is typically performed 2 to 4 weeks prior to starting aspirin desensitization and therapy to minimize the risk of polyp regrowth during the interval period.⁴ However there is a paucity of data about which patients benefit from ESS prior to aspirin therapy. Regardless of ESS status and timing, aspirin desensitization followed by daily aspirin provides therapeutic benefits to most patients with AERD, including improved asthma control, decreased oral steroid use, and reduced nasal polyp recurrence⁵. Swierczynska-Krepa et al. found that patients with AERD who underwent aspirin desensitization reported improvements in smell and nasal blockade, lower dosages of inhaled corticosteroids, and decreases in SNOT-22 and Asthma Control Questionnaire scores⁶. Aspirin desensitization followed by daily maintenance appears to be safe and effective in improving airway disease and quality of life even after 10 or more years of continuous use⁷.

While ESS enhances aspirin treatment response in patients with AERD when performed 3 to 6 weeks before the initiation of aspirin desensitization and treatment,⁸ there is insufficient research on aspirin desensitization outcomes with and without antecedent ESS. In addition, aspirin seems to benefit some patients even if surgery timing was not accounted for or if ESS was performed longer than 12 months before aspirin treatment start.^6,9,10

The goal of this study is to further elucidate the role of ESS in AERD treatment and to characterize AERD patients who are less likely to respond to aspirin treatment without preceding ESS.

MATERIALS AND METHODS

Study Population

This study was evaluated and approved by the Institutional Review Board at the Albert Einstein College of Medicine. Patients over the age of 18 with physician-diagnosed AERD, defined as a hypersensitivity reaction to aspirin with stable asthma (baseline FEV1 >70% of predicted or 1.5 L absolute volume, and no asthma exacerbations for at least two months before the first challenge) and evidence of nasal polyposis on nasal endoscopy, were prospectively enrolled into an ongoing observational cohort investigation between 2016 and 2018.

Aspirin Desensitization and Treatment Protocol

All patients underwent aspirin desensitization according to a previously published protocol¹¹. For the week prior to aspirin desensitization, patients were instructed to take montelukast to increase the safety of the aspirin challenge¹². Hypersensitivity reactions during the challenge were treated according to the symptoms. The patients who completed the aspirin desensitization protocol were prescribed 650 mg of aspirin twice daily. After 4 weeks of aspirin treatment patients were re-evaluated and if deemed appropriate, patients continued aspirin treatment with 650 mg in the morning and 325 mg in the evening for a total of 975mg of aspirin daily for an additional 8 weeks.

Aspirin Treatment Response

Response to aspirin treatment was defined by the ability to complete the desensitization protocol and by the change in ACT score of at least 3 points¹³ and/or at least 0.5 points¹⁴ in the RQLQ score. Patients who were unable to tolerate aspirin desensitization due to persistent, non-resolving symptoms of the hypersensitivity reaction (e.g. abdominal pain, unresolved bronchospasm, and/or a persistent skin rash), or had a decline in ACT score, or experienced an increase in RQLQ score indicating worsening of respiratory symptoms after four weeks of aspirin treatment were defined as “non-responders”. The remaining cohort of study patients were deemed as “responders”.

Endoscopic sinus surgery (ESS)

Patients were deemed to be ESS candidates if they had medically refractory disease defined as unsuccessful symptom resolution after previous medical therapy with systemic, broad-spectrum or culture-directed antibiotics (≥2 weeks duration) and either topical nasal corticosteroid sprays (≥3 week duration) or at least a 5-day trial of systemic corticosteroid therapy.^15,16 As mentioned above, some of the “non-responder” patients who failed an aspirin treatment attempt and patients who demonstrated an initial response with aspirin therapy but then experienced a subsequent decline in symptom control also underwent ESS on the appropriate involved sinuses.

Surgical procedures consisted of either unilateral or bilateral maxillary antrostomy, partial or total ethmoidectomy, sphenoidotomy, frontal sinus procedures, inferior turbinate submucous resection, septoplasty, and the use of intraoperative stereotactic surgical navigation. All ESS patients had placement of bilateral steroid-releasing sinus implants (Propel™, Intersect ENT, Palo Alto, CA) in the ethmoid sinuses and/or frontal recess during ESS.¹⁷ All patients were seen within 1 to 2 weeks after surgery for postoperative debridement. Postoperative medical therapy consisted of nasal saline irrigations, topical corticosteroid rinses or sprays, oral steroids and, in some cases, antibiotics. Oral corticosteroid tapers were administered to all patients beginning on postoperative day 1 and starting at 30 mg daily before tapering off by 10 mg either every 3 or every 4 days. Thus, the total duration of the corticosteroid taper was either 9 or 12 days in duration.

Postoperative aspirin desensitization

Three to four weeks after ESS, patients underwent a second aspirin desensitization according to the same protocol.¹¹ Besides administering montelukast prior to this desensitization, patients who were unable to complete aspirin desensitization the first time due to the persistent gastro-intestinal (GI) symptoms (epigastric and abdominal pain and nausea/vomiting) were also premedicated with famotidine (40 mg) or ranitidine (150 mg), omeprazole (40 mg), and ondansetron (4 to 8 mg) 1 hour before the procedure start.

Patients who tolerated the second desensitization attempt were started on 650 mg of aspirin twice daily. After 4 weeks of aspirin treatment, patients were re-evaluated and advised to continue aspirin treatment with 650 mg in the morning and 325 mg in the evening for a total of 975mg of aspirin daily for additional 8 weeks (until a total of 12 weeks of aspirin therapy), then 325 mg twice daily. Patients were prospectively followed for a total of 24 weeks of aspirin treatment.

Biomarkers measurement

All patients underwent measurement of exhaled Fractional excretion of Nitric Oxide (FeNO) [NIOX VERO, Aerocrine AB, Morrisville, NC], spirometry (Puritan Bennett Renaissance® II spirometer, Pleasanton, CA), and determination of nasal inspiratory peak flow (NPF) (In-Check™ Nasal inspiratory flow device, Clement International, Ltd., Essex, UK). The best of three efforts was recorded for spirometry and NPF. Blood was collected at baseline for measurement of the eosinophil count and total serum immunoglobulin E levels.

Statistical Analysis

All summary statistics were expressed as means ± SEM or as medians and interquartile ranges (IQRs). Paired data were analyzed by using paired Student’s t-test or Wilcoxon signed-rank test as appropriate. Categorical data were analyzed by chi-square test or Fisher exact test, as appropriate. All statistical analyses were performed with STATA 15.1 software (StataCorp, College Station, Texas). Significant differences were identified at a conventional 0.05 alpha level.

RESULTS

Study Cohort and Response to Aspirin Treatment

We identified 40 patients with aspirin-challenge confirmed AERD who were offered aspirin desensitization. Eight patients (8/40, 20%) were unable to complete the desensitization protocol due to persistent reactions (bronchospasm, gastrointestinal symptoms, and/or skin rash) during the procedure. The remainder of the patients were started on high-dose aspirin treatment which was continued for four weeks. Treatment outcomes following aspirin treatment in responders and non-responders is summarized in Figure 1. 19 patients classified as “responders” had a significant increase in the ACT score after 4 weeks of treatment compared with baseline (23 [IQR 21–25] vs 20.5 [IQR 17–22], p<0.01) and a significant decline in RQLQ score compared to baseline (2.5 [IQR 1.5–3.0] vs. 1.2 [IQR 0.5–2.0], p<0.05). Twenty-one patients classified as “non-responders” had a decline in ACT score at 4 weeks compared with baseline (15 [IQR 12–18] vs. 19 [IQR 15–22], p<0.05) but no significant change in RQLQ score at 4 weeks compared to baseline (2.9 [IQR 2.2–3.7] vs. 2.6 [IQR 2.0–3.8], p=0.6).

Figure 1:

ACT (A) and RQLQ (B) outcomes in responders and non-responders on primary aspirin desensitization and therapy before ESS; vertical bars represent Interquartile Ranges

The baseline characteristics for responders (N=19) and non-responders (N=21) are presented in Table 1. Compared with responders, non-responders were more likely to be African American than Caucasian (71% vs. 26%, p<0.01) with a lower baseline NPF (94 ± 8.7 L/min vs. 125 ± 12.2 L/min, p<0.05), fewer sinus surgeries prior to aspirin desensitization (1 [IQR 0–1] vs. 1 [IQR 1–2], p<0.01), and higher total serum immunoglobulin E levels (252 [IQR 96–479] kU/L vs. 87 [IQR 53–193] kU/L, p<0.01). Additionally, the non-responders had higher FeNO levels than responders (48 [IQR 29–86] vs 32 [IQR 18–43] ppb, p=0.05), although this outcome was of borderline significance. After 4 weeks, all non-responders discontinued aspirin treatment and most of the responders continued aspirin treatment for at least 12 weeks (Figure 1).

Table 1.

Baseline comparison of responders and non-responders

	Responders (N=19)	Non-responders (N=21)	p value
Age (y), mean ±SE	43.6 ± 2.5	42.2 ± 2.5	0.7
Female patients, N (%)	9 (47)	15 (71)	0.1
Race, N (%)			<0.01
African American	5 (26)	15 (71)
Latino	4 (22)	5 (24)
White	9 (47)	1 (5)
Asian	1 (5)	0
Baseline FEV1, % predicted (±SEM)	81.3 ± 4.7	84.6 ± 4.8	0.6
ACT scores at baseline, median (IQR)	20.5 (17–22)	19 (15–22)	0.3
Baseline nasal peak flow, L/min (±SEM)	125 ± 12.2	94 ± 8.7	<0.05
FeNO*, ppb median (IQR)	32 (18–43)	48 (29–86)	0.05
Number of surgeries at baseline, median (IQR)	1 (1–2)	1 (0–1)	<0.01
SNOT-22 scores at baseline, median (IQR)	55 (43–58)	51 (38–66)	0.9
RQLQ scores at baseline, median (IQR)	2.5 (1.5–3.0)	2.6 (2 – 3.8)	0.2
High dose ICS use, %	52	48	0.5
Systemic prednisone use, %	40	60	0.9
Time between preceding sinus surgery and desensitization to aspirin, months median (IQR)	48 (24–72)	48 (20–96)	0.8
Lund-Mackay score prior to surgery, median (IQR)	19 (14–21)	19 (14–22)	0.8
Baseline peripheral blood eosinophil counts, K/uL, median (IQR)	0.6 (0.4–0.8)	0.7 (0.5–0.9)	0.2
Baseline total serum immunoglobulin E level, kU/L, median (IQR)	87 (53–193)	252 (96–479)	<0.01*

FEV1 – Forced expiratory volume in 1 second, ACT – Asthma Control Test, FeNO – Fractional excretion of Nitric Oxide, SNOT-22 – Sino-nasal Outcome Test, RQLQ – Rhinoconjunctivitis Quality of Life Questionnaire, ICS – Inhaled corticosteroids

Effect of Subsequent Endoscopic Sinus Surgery (ESS) on Aspirin Treatment Outcomes

Twenty-four of the 40 patients (9 from the responder group and 15 from the non-responder group) required subsequent ESS. Of these 24 patients, one (4.2%) had no frontal sinus involvement. Thirteen patients (54.2%) had Draf 2A frontal sinusotomies, 8 patients (33.3%) had Draf 2B frontal sinusotomies, and 2 patients (8.3%) had a modified Lothrop (Draf 3) procedure. All 24 patients underwent total ethmoidectomy. Wide maxillary antrostomies were created in all patients and, of these, 8 (33.3%) had maxillary mega-antrostomies.

At 4 weeks after the ESS, all 24 patients had a significant improvement in RQLQ score compared to baseline prior to ESS (0.7 [IQR 0.3–1.9] vs. 2.5 [IQR 1.9–3.6], p<0.001) (Figure 2) as well as an improvement in NPF compared to baseline (140 [IQR 100–175] L/min vs. 100 [IQR 80–140] L/min, p=0.01) (Figure 3). However, 4 weeks after the ESS, there was no significant change in ACT score compared with baseline (22 [IQR 19–24] vs. 19.5 [IQR 17–22], p=0.1) (Figure 2), or in FEV1% (78% [IQR 65%−102%] vs. 81% [IQR 70%−99%], p=0.8) (Figure 3).

Figure 2:

ACT and RQLQ changes in patients after ESS; vertical bars represent Interquartile Ranges

Figure 3:

FEV1 % and NPF changes in patients after ESS; vertical bars represent Interquartile Ranges

Re-initiation of aspirin therapy after ESS

Three to four weeks after surgery, all 24 patients were able to complete the aspirin desensitization protocol and start aspirin therapy. Of note, three out of five patients who could not complete desensitization before ESS due to persistent GI symptoms had no such GI symptoms during the desensitization after ESS. The other two patients had mild GI symptoms. These symptoms were treated and did not preclude continuation of aspirin desensitization.

After 4 weeks of aspirin therapy post-ESS, there was continuous improvement in NPF (140 [IQR 100–175] L/min vs 155 [IQR 100–205] L/min, p<0.05) (Figure 4). ACT and RQLQ scores as well as FEV1% remained unchanged at 4 weeks after ESS compared with the values immediately post-ESS (Figure 4). Long-term follow-up consistently showed stability in ACT score (Figure 4a), RQLQ score (Figure 4b), NPF values (Figure 4c) and FEV1% (Figure 4d) at the postoperative visit and at 12 and 24 weeks from the initiation of maintenance therapy.

Figure 4:

Post-ESS Response of Aspirin Therapy as measured by ACT score (A), RQLQ score (B), NPF (C), and FEV1% (D); vertical bars represent Interquartile Ranges

Changes in biomarkers after aspirin treatment prior to ESS

Peripheral blood eosinophil counts after the initial aspirin treatment increased from baseline to 4 weeks after treatment in both responders (0.6 [IQR 0.4–0.8] K/μL vs. 0.7 [IQR 0.4–1.0] K/μL, p<0.01) and non-responders (0.7 [IQR 0.5–0.9] K/μL vs. 1.7 [IQR 0.8–1.9] K/μL, p<0.001) (Figure 5a). It remained stable in responders over the course of 12 weeks of aspirin treatment (Figure 5a).

Figure 5:

Serum Eosinophil Count (A), IgE levels (B) and FeNO levels (C) changes in responders and non-responders to aspirin desensitization and therapy before ESS; vertical bars represent Interquartile Ranges

Total serum IgE levels did not significantly increase in responders from baseline to the level measured after 4 weeks of aspirin treatment (87 [IQR 53–193] kU/L vs. 91 [IQR 32–122] kU/L, p=0.5) (Figure 5b) and remained stable after 12 weeks of aspirin treatment (113 [IQR 29–125] kU/L, p=0.6 vs. IgE levels at 4 weeks of aspirin treatment). However, in non-responders, total serum IgE levels increased significantly from baseline to after 4 weeks of aspirin therapy (252 [IQR 96–479] kU/L vs. 474 [IQR 197–911] kU/L, p=0.02) (Figure 5b).

FeNO levels did not significantly change in responders from baseline to after 4 or 12 weeks of aspirin treatment (Figure 5c). Similarly, in non-responders, FeNO levels did not significantly increase from baseline to after 4 weeks of aspirin therapy (Figure 5c).

Changes in biomarkers after aspirin treatment following ESS

In the 24 patients who either failed initial desensitization or did not sustain an adequate response to aspirin therapy and had to undergo subsequent ESS, peripheral blood eosinophil counts dropped significantly from baseline to post-ESS (0.6 [IQR 0.4–0.8] K/μL vs. 0.3 [IQR 0.1–0.5] K/μL, p<0.01) (Figure 6a). Eosinophil counts then increased significantly from post-ESS to after 4 weeks of aspirin treatment (0.3 [IQR 0.1–0.5] K/μL vs. 0.5 [IQR 0.2–0.8] K/μL, p<0.05) (Figure 6a) and remained stable thereafter (Figure 6a).

Figure 6:

Post-ESS Changes to Aspirin Therapy as measured by Serum Eosinophil Count (A), IgE levels (B) and FeNO levels (C); vertical bars represent Interquartile Ranges

Serum IgE levels remained unchanged from a pre-ESS baseline to any point post-ESS or after aspirin treatment (Figure 6b).

FeNO levels significantly decreased from pre-ESS to post-ESS (45 [IQR 26–69] ppb vs. 28 [15–41] ppb, p<0.01) (Figure 6c). However, FeNO levels significantly increased from a point post-ESS to after 4 weeks of aspirin therapy (28 [IQR 15–41] ppb vs. 38 [IQR 24–77] ppb, p<0.05) (Figure 6c) but then remained stable at 12 and 24 weeks after aspirin therapy (Figure 6c).

DISCUSSION

Aspirin desensitization and daily maintenance therapy benefits many patients with AERD.^3,7 Aspirin therapy has been shown to impart a long-term benefit of polyp regrowth reduction, improved sense of smell, and need for revision sinus surgery in prospective clinical trials^6,18–20. In addition to aspirin therapy, ESS has been shown to improve the disease process in both physician reported (Nasal Endoscopy scores, Lund-MacKay CT scores) and patient reported (SNOT-22 score) measures²¹. While aspirin therapy may be beneficial, its impact does not appear to be universal. In previously published studies, the response rate of aspirin desensitization and therapy varies from 77%−93^1,19,22. Cahill et al. reported that 17% of AERD patients were unable to complete aspirin desensitization with failures being due to gastrointestinal intolerance, persistent skin rash, or severe respiratory reactions.²³ Other patients have worsening of respiratory symptoms during daily aspirin treatment.²⁴ There is currently a lack of data on the factors that can predict which patients respond well to desensitization. A confounder in the treatment protocol may be the length of time between preceding ESS, if one exists, and the initiation of aspirin therapy. While some studies that show response and benefit to aspirin therapy have tested patients who have undergone ESS 3–6 weeks prior to aspirin therapy^22,25, other studies have not shown any correlation between preceding ESS time and initiation of aspirin therapy^6,18,24.

In our cohort, 20% of patients failed the initial desensitization protocol due to GI intolerance, similar to the 17% reported by Cahill et al²³. A third of the patients were able to complete desensitization but reported no change or worsening of respiratory symptoms on aspirin maintenance therapy. Ultimately, this group of “non-responders” were not able to benefit from aspirin therapy for their AERD. They had lower NPF values and higher total serum IgE levels compared to the patients that were able to tolerate aspirin desensitization and therapy with symptomatic improvement. Interestingly, there was no difference in timing of the preceding ESS, LM-scores, lung function, or medication use between the two groups. However, when non-responders underwent ESS and then underwent aspirin desensitization, they were all able to tolerate the same desensitization protocol and aspirin maintenance therapy with symptomatic improvement.

In the present study we report that aspirin responders tend to have lower total serum IgE levels than non-responders. While the role of serum IgE is not well established in the natural progression of AERD, previous studies have reported high levels of IgE in patients with AERD in the absence of atopy²⁶ and also as a response to staphylococcal superantigens.²⁷ Interestingly, aspirin-responders in this study—as well as in another study on aspirin treatment responders where IgE levels were reported—appear to have lower IgE levels than aspirin non-responders¹⁰. Non-responders were more likely to be African American. Several studies indicated that African Americans tend to have higher IgE levels than Caucasians.^28,29. In a presumably majority Caucasian AERD population from Poland, higher IgE levels were associated with more ED visits and hospitalizations for asthma exacerbations and poorer asthma control as measured through the ACT, further supporting the concept of IgE being a significant biomarker of worse disease in AERD³⁰.

While serum IgE levels did not change over the course of treatment in aspirin responders in this study as well as in another recent report²⁰, IgE levels significantly increased in non-responders during aspirin treatment. This is consistent with the recent observation that high-dose aspirin therapy increases markers of type 2 inflammation in AERD patients²⁰ and may make patients with higher baseline levels of IgE more prone to treatment failure. Interestingly, there was no change in IgE levels in our cohort after the second, post-ESS, aspirin desensitization. This may indicate that the absence of nasal polyp burden, due to the recent ESS, prior to aspirin desensitization, suppressed the increase in type 2 inflammation in AERD patients treated with aspirin. We previously reported that aspirin challenges are more likely to be asymptomatic after recent ESS³¹. In this study we also report that recent ESS is likely to improve aspirin desensitization and treatment outcomes, especially for patients with high baseline serum IgE levels. Of note, peripheral blood eosinophil counts were observed to significantly decrease after ESS. It is possible that this decrease was due to the preceding corticosteroid taper that was administered to patients in the immediate postoperative period.

It is postulated that aspirin therapy can prevent the regrowth of polyps that have been recently surgically cleared, but has no effect on polyps already existing in the nasal cavity^32,33. However, there was no difference in Lund-MacKay scores between the responders and non-responders in our cohort (score of 19 for both groups, p=0.8), suggesting that anatomical burden of nasal polyps alone did not contribute to the outcome of response. Of note, non-responders had a lower baseline nasal peak flow rate.

While ESS seems to benefit outcomes of aspirin desensitization and therapy, some patients are reluctant to undergo surgery for various reasons. Our results may help to stratify patients, via serum biomarkers and other clinical characteristics, into groups that may gain clinical benefit from aspirin desensitization and therapy without preceding ESS and those who require ESS before desensitization to decrease disease burden and increase chances of success. Our study supports previously published reports that some groups of patients may benefit from aspirin desensitization regardless of preceding ESS timing^6,10 and shows that others, possibly with serum biomarkers characteristics of higher T2 inflammation, may require ESS to be able to tolerate desensitization and fully benefit from long-term aspirin therapy.

This study has several limitations. The patients in our study were not systematically tested for atopy, which could potentially explain some of the peripheral biomarker levels and subsequent difference in aspirin challenge response. Additionally, while we analyzed prospectively collected data, not all confounders and systematic biases can be excluded without blinding and randomization. All patients who underwent ESS had mometasone-eluting implants placed at the time of surgery and received postoperative oral corticosteroid taper. Consequently, the potential impact of these treatments on the subsequent response to aspirin desensitization and therapy could not be evaluated. Furthermore, though we noted that aspirin therapy non-responders were more likely to be African American, there was insufficient power to evaluate aspirin treatment outcomes by race or ethnic background due to the relatively modest sample size.

CONCLUSION

Antecedent ESS enhances aspirin treatment response in AERD patients and may convert patients who failed aspirin treatment prior to surgery to a more responsive phenotype after ESS. Patients with higher baseline serum IgE levels may especially benefit from ESS performed shortly before aspirin desensitization and long-term treatment initiation.