Discussion

The present study shows an increased likelihood of the need for medical polypectomy or revision surgery in younger patients with a history of previous FESS, higher peripheral blood eosinophil counts, a higher Lund–Mackay score and the presence of aspirin-exacerbated respiratory disease. For example, the odds ratios in patients with an average peripheral blood eosinophil count of 300 cells/μl or higher, indicate a 73 per cent significantly increased likelihood of requiring medical polypectomy and 73 per cent likelihood for combined polypectomy. Radiological Lund–Mackay scores of more than 17 were associated with an increased likelihood of medical polypectomy (58 per cent), revision surgery (89 per cent) and combined polypectomy (62 per cent).

We also observed significant differences in peripheral blood eosinophil count, total IgE values, nasal polyp scores and Lund–Mackay scores between chronic rhinosinusitis with nasal polyps patients depending on the presence or absence of asthma or aspirin-exacerbated respiratory disease. These markers of chronic rhinosinusitis with nasal polyps increased in severity, from being the lowest in chronic rhinosinusitis with nasal polyps only patients to the highest in chronic rhinosinusitis with nasal polyps plus aspirin-exacerbated respiratory disease patients.

There is conflicting evidence regarding an association of patient age with revision surgery during follow-up.^{Reference Koskinen, Salo, Huhtala, Myller, Rautiainen and Kaariainen10} We hypothesise that patients younger than 55 years of age have higher odds for revision surgery as they may be more fit for general anaesthetic compared to patients aged over 55 years. Following surgical and anaesthetic assessments, older patients with concerns about undergoing an operation requiring general anaesthetic may have opted against having surgery. Nevertheless, as we did not evaluate co-morbidities in each patient, this is merely speculative.

The mean age of patients in our cohort was 52.9 years. The incidence of nasal polyps increases with age, with a peak in the sixth decade; it is more common in men and less common before the third decade of life.^{Reference Hopkins5,Reference Stevens, Schleimer and Kern11,Reference Lal, Rounds and Divekar17} In our cohort, only 4.5 per cent of patients were younger than 30 years of age, while there were more males than females. There might be other factors that explain why men were being operated on three times more. Physician bias has been reported as having an effect on decision-making in situations where the indications for surgery are less clear.^{Reference Lal, Rounds and Divekar17}

Up to 50 per cent of patients with chronic rhinosinusitis with nasal polyps have co-existent asthma, which has been shown to be associated with more severe sinonasal disease.^{Reference Fokkens, Lund, Hopkins, Hellings, Kern and Reitsma4,Reference Stevens, Schleimer and Kern11,Reference Promsopa, Kansara, Citardi, Fakhri, Porter and Luong15} In our cohort, 51.1 per cent of patients had asthma or aspirin-exacerbated respiratory disease, 38.9 per cent had asthma, and 12.2 per cent had aspirin-exacerbated respiratory disease; these rates are similar to previous reports.^{Reference Koskinen, Salo, Huhtala, Myller, Rautiainen and Kaariainen10,Reference Stevens, Schleimer and Kern11,Reference Kim and Kountakis16} Although not statistically significant, a higher proportion of female chronic rhinosinusitis with nasal polyps patients had aspirin-exacerbated respiratory disease compared to males, which has been reported before.^{Reference Hopkins5,Reference Kim and Kountakis16,Reference Stevens, Peters, Suh, Norton, Kern and Conley18}

The finding of higher Lund–Mackay scores when comparing chronic rhinosinusitis with nasal polyps only, chronic rhinosinusitis with nasal polyps plus asthma and chronic rhinosinusitis with nasal polyps plus aspirin-exacerbated respiratory disease patients supports the notion of more severe sinonasal inflammation. We should emphasise that chronic rhinosinusitis with nasal polyps plus asthma and chronic rhinosinusitis with nasal polyps plus aspirin-exacerbated respiratory disease are two distinct groups. As there are no in vitro tests available to diagnose aspirin-exacerbated respiratory disease, patients should be specifically asked whether nasal or respiratory symptoms are exacerbated by the intake of salicylates, non-steroidal anti-inflammatory agents, and dietary sources such as fresh berries and nuts.^{Reference Hopkins5,Reference Stevens, Schleimer and Kern11,Reference Kim and Kountakis16}

Chronic rhinosinusitis with nasal polyps plus aspirin-exacerbated respiratory disease patients were more likely to have had previous FESS, in keeping with previous literature.^{Reference Stevens, Schleimer and Kern11,Reference Tosun, Arslan, Karslioglu, Deveci and Durmaz12,Reference Kim and Kountakis16} We suggest that higher levels of peripheral blood eosinophils might alert clinicians to the presence of possible aspirin-exacerbated respiratory disease, as chronic rhinosinusitis with nasal polyps plus aspirin-exacerbated respiratory disease patients had a higher peripheral blood eosinophil count. The concept of the unified airway suggests that upper airway inflammation may influence lower airway inflammation and vice versa.^{Reference Tosun, Arslan, Karslioglu, Deveci and Durmaz12,Reference Promsopa, Kansara, Citardi, Fakhri, Porter and Luong15} Medical and surgical treatment of chronic rhinosinusitis in patients with asthma has been shown to decrease asthmatic and sinonasal symptoms.^{Reference Promsopa, Kansara, Citardi, Fakhri, Porter and Luong15}

Nineteen patients (9 per cent) in our cohort underwent revision surgery; more specifically, 4.9, 6 and 37.5 per cent underwent revision surgery for chronic rhinosinusitis with nasal polyps only, chronic rhinosinusitis with nasal polyps plus asthma and chronic rhinosinusitis with nasal polyps plus aspirin-exacerbated respiratory disease, respectively. Unsurprisingly, a history of previous FESS is associated with more severe sinus disease and has been reported as a factor associated with revision surgery, although Tosun et al. reported that previous FESS has no effect on post-operative recurrences.^{Reference Tosun, Arslan, Karslioglu, Deveci and Durmaz12} Somewhat similar results to our findings have been reported in the literature in aspirin-exacerbated respiratory disease patients, but the incidence of revision surgery in chronic rhinosinusitis with nasal polyps only and chronic rhinosinusitis with nasal polyps plus asthma patients differs.^{Reference Koskinen, Salo, Huhtala, Myller, Rautiainen and Kaariainen10,Reference Tosun, Arslan, Karslioglu, Deveci and Durmaz12} In our cohort, patients who underwent previous FESS had a higher peripheral blood eosinophil count.

Computed tomography scans are usually performed as part of surgical planning, but are also used for the assessment of chronic rhinosinusitis severity.^{Reference Hopkins5} No difference in Lund–Mackay score was present between genders in our cohort, which differs from other reports.^{Reference Stevens, Schleimer and Kern11} Patients with a Lund–Mackay score of more than 17 had higher odds for requiring revision surgery, which corresponds with other reports of patients with higher CT scores and the failure of FESS at two years.^{Reference Hopkins, Slack, Lund, Brown, Copley and Browne3}

Patients with aspirin-exacerbated respiratory disease were more likely to undergo revision surgery. Peripheral blood eosinophil count and total IgE did not differ among patients who did or did not undergo revision surgery.

There were fewer patients who received medical polypectomy in the chronic rhinosinusitis with nasal polyps only group compared to the rest of the patients who also had asthma and aspirin-exacerbated respiratory disease (Table 3 in the supplementary material). When analysed in more detail, however, the findings revealed no significant difference between the chronic rhinosinusitis with nasal polyps only patients and the chronic rhinosinusitis with nasal polyps plus asthma patients, but a significant difference between the chronic rhinosinusitis with nasal polyps plus asthma patients and the aspirin-exacerbated respiratory disease patients. We can therefore say that chronic rhinosinusitis with nasal polyps patients with asthma do not have higher risk for further medical polypectomy compared to chronic rhinosinusitis with nasal polyps only patients; it is aspirin-exacerbated respiratory disease that increases the risk for medical polypectomy.

More patients younger than 55 years of age received medical polypectomy compared to those older than 55 years. This could be partially explained by older patients having higher risk factors for potential side effects of oral corticosteroids, such as diabetes, osteoporosis and glaucoma, and this might have been the reason why they were prescribed medical polypectomy less frequently.

Patients with a higher peripheral blood eosinophil count were more likely to receive medical polypectomy, reflecting a higher type 2 inflammatory burden. This might have been influenced by surgeons being more likely to prescribe medical polypectomy to patients with a higher peripheral blood eosinophil count, as it is known that corticosteroids have better therapeutic effect in patients with blood eosinophilia.^{Reference Fokkens, Lund, Hopkins, Hellings, Kern and Reitsma4,Reference Tosun, Arslan, Karslioglu, Deveci and Durmaz12} This is somewhat confirmed, as peripheral blood eosinophil count had no association with patients undergoing revision surgery.

Type 2 inflammation has been characterised by increased levels of the cytokines interleukin (IL)-4, IL-5 and IL-13 resulting in peripheral blood eosinophilia, elevated fractional exhaled breath nitric oxide and total IgE.^{Reference Fokkens, Lund, Hopkins, Hellings, Kern and Reitsma4,Reference Promsopa, Kansara, Citardi, Fakhri, Porter and Luong15} Currently, tests of these cytokines are not commonly performed in rhinology or asthma clinics, but we believe peripheral blood eosinophil count and total IgE should be determined as part of standard clinical practice in patients with chronic rhinosinusitis with nasal polyps, as the results may be useful in guiding treatment and predicting prognosis.^{Reference Hopkins5} The question arises whether fractional exhaled breath nitric oxide, a specific marker of IL-13 expression, should be performed routinely in chronic rhinosinusitis with nasal polyps patients without known asthma in order to help stratify the type 2 inflammatory burden, although in our hospital it would not be feasible because of the current workload.^{Reference Chan, Stewart, Misirovs and Lipworth19,Reference Khatri, Iaccarino, Barochia, Soghier, Akuthota and Brady20}

We have chosen medical polypectomy and revision surgery as pragmatic endpoints, and assessed which patients ended up receiving them. We elected to combine medical polypectomy and revision surgery as a composite endpoint. In our rhinology clinic, medical polypectomy consists of a two-week course of oral prednisolone 25 mg tablet once a day with or without antibiotics. A Cochrane review of short-course oral corticosteroids for chronic rhinosinusitis has shown improvement in quality of life, less severe nasal symptoms and smaller nasal polyps that lasted up to three to six months, which is similar to a randomised clinical trial by Vaidyanathan et al. using a two-week course of 25 mg of prednisolone.^{Reference Vaidyanathan, Barnes, Williamson, Hopkinson, Donnan and Lipworth1,Reference Head, Chong, Hopkins, Philpott, Burton and Schilder21} Short-term therapy with oral corticosteroids should not be given more than twice per year because of the cumulative systemic burden.^{Reference Van Zele, Gevaert, Holtappels, Beule, Wormald and Mayr6,Reference De Corso, Pipolo, Cantone, Ottaviano, Gallo and Canevari22} The most common short-term side effects of oral corticosteroids are gastrointestinal disturbances and insomnia, and the long-term side effects are osteoporosis, type II diabetes, weight gain, skin thinning, proximal myopathy, hypertension, cataracts, glaucoma and psychiatric disturbance.^{Reference Rice, White, Scarpati, Wan and Nelson23}

About 10–20 per cent of patients undergo a revision surgical procedure within five years.^{Reference Hopkins, Slack, Lund, Brown, Copley and Browne3,Reference Koskinen, Salo, Huhtala, Myller, Rautiainen and Kaariainen10} As per our results, patients with aspirin-exacerbated respiratory disease have higher odds for undergoing revision surgery; therefore, we suggest having a high level of suspicion for unreported aspirin-exacerbated respiratory disease in chronic rhinosinusitis with nasal polyps patients who require revision surgery, especially in patients with high eosinophil counts and a high Lund–Mackay score.^{Reference Tosun, Arslan, Karslioglu, Deveci and Durmaz12} Some studies report that having asthma or aspirin-exacerbated respiratory disease is not associated with revision surgery during follow-up.^{Reference Koskinen, Salo, Huhtala, Myller, Rautiainen and Kaariainen10} In our cohort, chronic rhinosinusitis with nasal polyps plus asthma patients did not have higher revision surgery rates, which is different from other reports.^{Reference Stevens, Schleimer and Kern11} This is possibly because, in this study, chronic rhinosinusitis with nasal polyps plus asthma patients were carefully separated from nasal polyps plus aspirin-exacerbated respiratory disease patients. When patients have a diagnosis of aspirin-exacerbated respiratory disease, additional effort should be made to educate patients about non-surgical management options, perhaps including anti-leukotriene medication, aspirin desensitisation and/or dupilumab, and the overall long-term post-operative prognosis.^{Reference Hopkins5,Reference Oykhman, Paramo, Bousquet, Kennedy, Brignardello-Petersen and Chu7,Reference Kim and Kountakis16} Additionally, surgeons should consider performing more extensive FESS in patients at high risk for recurrence of nasal polyposis, as Hopkins et al. reported a significantly lower requirement for revision surgery when additional sinus surgery was performed during surgical polypectomy.^{Reference Hopkins, Slack, Lund, Brown, Copley and Browne3}

Studies suggest that patients with prolonged and chronic disease might benefit more from surgery than continued medical therapy.^{Reference Koskinen, Salo, Huhtala, Myller, Rautiainen and Kaariainen10} This might be overturned by the availability of biologics based on phase 3 trials, with the use of dupilumab in patients with chronic rhinosinusitis with nasal polyps.^{Reference Bachert, Han, Desrosiers, Hellings, Amin and Lee24} Revision surgery is associated with more surgical risks than primary FESS because of major surgical complications, such as injury to the orbit or a cerebrospinal fluid leak.^{Reference Koskinen, Salo, Huhtala, Myller, Rautiainen and Kaariainen10} Of note, no major surgical complications following either primary FESS or revision surgery were reported in our study.

This study has limitations. It is retrospective in nature and relies on data recorded in medical notes. Not all patients had their peripheral blood eosinophil count, and total IgE and specific IgE responses tested; hence, there might be selection or sampling bias for the reported results. Information on patients having asthma or aspirin-exacerbated respiratory disease might have been missing, and some patients might therefore have been marked as having chronic rhinosinusitis with nasal polyps only or chronic rhinosinusitis with nasal polyps plus asthma. Data on medical polypectomy prescribed in the community by general practitioners were not available, and so there might be a higher percentage of patients who had medical polypectomy following FESS.

When chronic rhinosinusitis with nasal polyps plus asthma patients were given oral corticosteroids for asthma in the follow-up period, this was not counted as medical polypectomy, as the indication was for asthma, not chronic rhinosinusitis with nasal polyps. Some authors would count this as medical polypectomy, but in this study it was not as the primary indication for oral corticosteroids was not chronic rhinosinusitis with nasal polyps.

The study concentrated on patients who underwent FESS; therefore, some patients, who were candidates for the surgery but for unknown reasons did not undergo it, were missed from the overall analysis of the phenotype of patients with severe chronic rhinosinusitis with nasal polyps requiring FESS. Patients who underwent FESS in the private sector were also missed, and so there might be a higher number of patients with chronic rhinosinusitis with nasal polyps who underwent FESS and/or required revision surgery than is reported by us.

It was not possible precisely to assess patients having allergic rhinitis from clinical notes as symptoms and diagnosis were not mentioned accurately. Although information was gathered on the use of antihistamines as a surrogate, it is likely there is a proportion of patients who were taking over-the-counter antihistamines and the use of these was not recorded anywhere in the medical notes.

The reported numbers of chronic rhinosinusitis with nasal polyps patients who had medical polypectomy or revision surgery following FESS is believed to be potentially lower than they actually were. There is no information on the patients who were not referred back to the ENT clinic despite having troublesome chronic rhinosinusitis with nasal polyps symptoms and/or who chose to see ENT doctors in the private sector. As part of this retrospective study, it was not possible to analyse the exact role of the extent of FESS and the need for medical polypectomy and/or revision surgery.