This study was the first to quantify adequacy of prior surgical completeness among symptomatic AERD patients presenting to a tertiary rhinology clinic. Most referrals included patients who had their surgery at non-tertiary healthcare centres prior to referral. This does not necessarily mean however, that their surgeon was not rhinology fellowship trained or experienced in sinus surgery. Not unexpectedly, uncinectomies and maxillary antrostomies were completed in most patients. However, four independent fellowship trained rhinologists who were blinded to the patient histories concluded that the remaining sinuses were adequately opened in less than half of included patients. Frontal sinus openings were the least adequately opened with approximately one-quarter of included patients having their frontal sinuses adequately opened.

In general, the sinuses that were graded as most often adequately complete by the surgeons represent sinuses that are part of the initial and foundational steps in classic functional endoscopic sinus surgery [13]. On the other hand, endoscopic surgery of the sphenoid and frontal sinuses can be regarded as more complex with a higher surgical risk [14]. This can be particularly true in AERD patients who have severe inflammatory disease which may make the surgical fied more difficult to navigate when performing ESS. While middle turbinate resection is not standard of normal ESS, it is a consideration in revision surgery, especially with AERD patients. While it is impossible to know why most patients had incomplete ESS prior to referral to our centre, it is possible that surgical difficulty may have been a contributing factor. Another possible cause of lack of completeness may be related to more intraoperative bleeding in AERD patients potentially leading to early termination of surgery. However, it is also possible that some of these unopened sinuses did not have significant sinus disease at initial presentation and previous surgeons may not have had the same surgical philosophy or intent of performing “complete” sinus surgery during the initial ESS.

In general, the agreement scores across all the sinuses were similar, ranging from fair agreement to moderate agreement [12]. The agreement when all sinus agreement scores were averaged was moderate. Interestingly, the total score mean agreement was the only agreement metric that received “slight agreement” categorization. This is likely due to the nature of such analyses as for all individual measurements the agreement scores were calculated using binary values (0 or 1) hence the variation between graders would only be either 0 or 1. Whereas for the total scores, the variation could be between 0 and 14 leaving greater opportunity for worse agreement scores. Despite this, the average total scores across all patients for each surgeon grader were 6.6, 7.1, 6.4 and 6.4. Variations between the surgeons may represent individual training, personal preference, and institutional differences.

The case for complete sinus surgery in patients with CRS relies on the fact that the mainstay of medical treatment in this disease is ongoing topical therapy (i.e. saline and nasal steroid irrigations). From a delivery standpoint, prior studies have demonstrated an improved penetration of topical delivery when all sinuses have been completely opened, with little to no penetration of the frontal and sphenoid sinuses in the unoperated state [15]. Furthermore, in this disease of epithelial and immune barrier dysfunction, it has been postulated that complete sinus surgery lessens the inflammatory load from the sinuses which may be driving the ongoing cycle of mucosal inflammation [16]. From an outcome perspective, full-house revision ESS can improve patient symptoms, mucosal swelling and LM scores [17]. Hwang et al. recently evaluated the correlation between extent of sinus surgery and postoperative outcomes [18].They demonstrated that patients who had appropriate extent of surgery in concordance with the severity of their disease had the best long term outcomes (greatest reduction in SNOT-22 scores) [18]. DeConde et al. [19] found that complete sinus surgery improved postoperative SNOT-22 scores in CRS patients compared with more targeted, limited surgery, although this difference did not reach the minimal clinically important difference threshold. In other words, patients with more severe disease may benefit from more complete sinus surgery. Given that AERD patients tend to have higher LM scores—on average 20, as demonstrated by Grose et al. [20], it can be hypothesized that these patients could benefit from more comprehensive surgery to match their disease burden. Some authors have advocated for even more extended procedures such as the endoscopic modified Lothrop procedure (EMLP) in AERD patients. Outcome studies in AERD patients who have undergone the EMLP have demonstrated a 22.5% revision surgery rate, which is slightly less than the published rate of 26.2% for complete ESS in this same population [20, 21]. Despite this, additional studies need to be undertaken to determine if additional extended procedures truly result in improved and more durable outcomes.

Biologic therapies such as dupilumab, mepolizumab [21] as well as AD are beneficial for AERD patients [22, 23]. Nonetheless, they do not come without potential costs and risks. For example, for biologic treatment, risks have been described such as: atopic dermatitis, injection site infections, and most commonly conjunctivitis which can occur in 38% of patients [24]. Adverse effects of AD risks can include: major bleeding, gastritis, and rashes with a relative risk of 4.39 [25]. Additionally, reducing the patient’s inflammatory load increases the effectiveness of AD. Hence, given the aforementioned possible additional benefit of complete surgery for AERD patients, assessment of surgical completeness prior to initiating adjuvant treatments may be prudent. With these assessments, surgeons may begin to have conversations with their patients who have had prior surgery regarding surgical completeness. This may allow for collaborative decision making with the patient’s perception of their disease and surgeon regarding the need for revision surgery versus biologic therapy.

This study has certain limitations. Primarily, it is retrospective and was limited by unavailability of certain patient data leading to some patients being excluded from the study. Additionally, the agreement between the four rhinology surgeons was ‘moderate’ as opposed to ‘substantial’ or ‘almost perfect’, hence there is evidence of variability between the surgeons. Nonetheless, the trends between the different sinuses and completeness remained consistent between the surgeons. More so, surgical completeness was only being measured by one factor—postoperative imaging. Furthermore, the scores used were binary and perhaps more granular data could more accurately reflect extent of prior surgery. Finally, we were not able to calculate intra-rater agreement as well as compare the scoring system used with newer scoring systems, such as the ACCESS score [26]. The LM score is very similar to the ACCESS score as it addresses each sinus cavity (maxillary, anterior ethmoid, posterior ethmoid, sphenoid and frontal sinus). The LM score has the additional granularity of scoring adequacy of uncinectomy and whether a middle turbinate resection was completed. As noted in our methodology, the rhinologists would only score a LM 1 if minimal or no additional bony dissection would be required for the opening of a particular sinus cavity. This is in essence equivalent to a score of 0 on the ACCESS score (i.e. no additional dissection is required). While we do not have the additional granularity of scoring "partial surgery" (i.e. ACCESS score of 1) for a particular sinus cavity, we believe that the LM score provides a similar overall assessment of the degree and extent of surgery that has been previously performed. In fact, one could argue that there is no utility in only partial opening of a particular sinus (i.e. partial opening of the posterior ethmoid) if the goal of “adequate” or full-house ESS in severe inflammatory CRS (i.e. AERD in our study) is to provide maximal ventilation and access for topical steroid delivery.

Future research should continue to study objective and prospective ways to define what constitutes adequate surgical therapy for patients with AERD. Clearly, incorporating newer scoring systems such as the ACCESS score will be important to enhance cross-study comparisons. Most importantly, studying long-term outcomes of patients who may have had different degrees of sinus surgery will help define what constitutes the optimal surgical environment for this difficult endotype of CRS with nasal polyps. The heterogeneity of sinus surgery is clearly an area which needs further investigation and may be an important variable in the management of patients with AERD.