Twelve-year follow-up study after endoscopic sinus surgery in patients with chronic rhinosinusitis with nasal polyposis

The patient population of this 12-year prospective cohort study covers 47 patients, 18 years or older, with CRSwNP diagnosed based on history, clinical examination, nasal endoscopy and CT scan, following the current EPOS guidelines [2]. All patients underwent extended ESS for nasal polyposis at the department of Otorhinolaryngology of the Ghent University, Belgium, between 1998 and 2000. The description of the surgical technique can be found in this article’s Additional file 1.

Before surgery, all patients underwent skin prick testing for common aero-allergens to diagnose allergic sensitization [17]. The presence of asthma was investigated by a lung physician [18] and the diagnosis of aspirin intolerance was primarily based on the clinical picture, namely the presence of asthma, nausea, erythema or other complaints shortly after ingestion of aspirin or NSAIDs. The exclusion criteria were pregnancy, cystic fibrosis, primary ciliary dysfunction, Kartagener syndrome, parasitic infection, and fungal infections.

In the scope of this study, three visits at the ENT-department were organized: pre-operatively, approximately 6 and 12 years after initial surgery.

The ethics committee of Ghent University Hospital approved the study and all patients provided written informed consent before participation in the study.

As primary outcome NP recurrence and the need of revision surgery were examined. The NP recurrence over 12 years’ time was defined as a nasal polyp score greater than 0 at one of the two follow-up visits or a history of revision surgery for recurrent NP during the 12-year follow-up period. The revision surgery rate is based on the exact surgical dates collected at each visit and by revision of medical records.

At each time point the following data were collected: demographics, comorbidities, symptom score (nasal obstruction, rhinorrhea, smell disturbance, sneezing, headache and eye symptoms), total nasal endoscopic polyp score (Davos score), history of previous surgery, medication use, general therapeutic response and EPOS control test. More detailed information can be found in this article’s Additional file 1.

Nasal polyp tissue, nasal secretion and serum were collected and investigated for different inflammatory markers [IL-5, IL-5Rα, TGF-β1, MPO, IL-18, ECP, total IgE and specific IgE antibodies against S. aureus enterotoxins (SAE-IgE)]. More detailed information about the used techniques can be found in this article’s Additional file 1.

Data are expressed as absolute numbers and percentages for categorical and as median and interquartile range (IQR) for continuous variables. Further details about statistics can be found in this article’s Additional file 1.

Forty-seven CRSwNP patients were included prior to ESS. Table 1 shows the clinical baseline characteristics of these 47 patients. A follow-up was performed over 12 years and in 2006 and 2012 patients were invited for an extensive follow up visit. After 12 years the response rate was 79% (38 out of 47), of which 19 patients underwent in 2000 primary surgery and 19 underwent revision surgery. We show in Table 2 the clinical baseline characteristics of both groups and we observe at baseline a significant higher amount of IL-5, IL-5Ra and ECP in patients undergoing revision surgery compared to primary surgery. Over the 12-year period 3 patients deceased, 4 patients could not be traced, and 2 patients were not prepared to participate.

Smell disturbance and nasal obstruction were the most predominant symptoms pre-operatively, as shown in Fig. 1. These symptoms are scored bothersome, i.e. moderate to severe, in 91.9% and 89.2% patients respectively for smell disturbance and nasal obstruction. Twelve years after surgery the smell disturbances (P < 0.01) and the nasal obstruction (P < 0.001) were still significantly improved compared to baseline (see Fig. 1). Rhinorrhoea and headache, two other symptoms mentioned in the EPOS definition, were also pre-operatively reported as bothersome in 51.3% and 40.5% of the patients respectively. After 12 years those complaints decreased to respectively 18.4% and 13.1%. Compared to the aforementioned symptoms, patients were less troubled by sneezing and eye symptoms pre-operatively and during follow-up.

In 2000, prior to ESS, all patients had considerable nasal polyps (Fig. 2). Compared to the NP score prior to ESS, the NP score was significantly decreased in 2006 (P < 0.05) and in 2012 (P < 0.001). Twelve years after ESS, endoscopic examination of the nasal cavity showed in 40% of the patients no nasal polyps. However, this percentage also includes patients that underwent revision surgery during the 12-year follow-up study. This percentage thus underestimates the presence on nasal polyps over the 12-year period.

Thirty out of 38 CRSwNP patients or 78.9% developed NP recurrence at a certain time point during 12 years of follow-up. Figure 3a shows the development of recurrence over time.

Comparing the ‘NP recurrence’ (N = 30) and ‘no recurrence’ group (N = 8; see Additional file 1: Table S1), we identified that 60% of the NP recurrence group had comorbid allergic sensitization however not significant different with 25% in the ‘no recurrence’ group (P = 0.117). Figure 4a shows the contribution of asthma and AERD. In the NP recurrence group 80.0% expressed asthma and 90.0% AERD compared to 78.3% asthma and 75.0% AERD in the non-recurrence group (Fig. 4a). The logistic regression showed a trend towards increased risk of NP recurrence in allergic patients (OR 4.5, 95% CI 0.78 to 26.1, P = 0.094) however not significant. For asthmatic patients (OR 1.1, 95% CI 0.22 to 5.5, P = 0.898) and patients with AERD (OR 3.0, 95% CI 0.32 to 28, P = 0.336) the risk is less obvious. This is confirmed by the Kaplan–Meier graphs (allergy Fig. 3b P = 0.093; asthma Fig. 3c P = 0.738; AERD Fig. 3d P = 0.136).

Looking at the tissue markers there was no statistical difference between tissue IL-5, ECP and IgE (see Table S1 in Additional file 1).

The non-significant influence of tissue IL-5 (Fig. 3f; P = 0.233), IgE (Fig. 3g; P = 0.446) and SAE-IgE (Fig. 3h; P = 0.459) on the development of NP recurrence over time is shown in the Kaplan–Meier figures.

Fourteen out of 38 or 36.8% of the CRSwNP patients had a need of revision surgery in the 12-year follow-up period. Consequently over 12-year follow-up the patient group can be divided in three groups: no recurrence (21.05%), NP recurrence but no revision ESS (42.11%) and NP recurrence with revision ESS (36.84%).

Figure 5a illustrates Kaplan–Meier survival analysis for revision surgery within the 12-year follow-up based on the date of revision surgery, with an overall ‘surgery-free rate’ of 84.2% (32/38) at 6 years and 63.2% (24/38) at 12-year follow-up. The time to revision ESS ranged from 18 to 153 months, with a median of 91 months.

The possible contribution of allergic sensitization, asthma and AERD to the risk of revision surgery was investigated and only allergic sensitization could be identified as a significant predictor for the need of revision surgery during follow-up (OR 6.1, 95% CI 1.3 to 28, P < 0.05 in allergic sensitization). The percentage patients who underwent revision surgery in asthmatics (53.3%) and patients with AERD (60.0%) was higher compared to patients without asthma (30.4%) or AERD (28.6%), however there was no statistical difference of revision surgery in the presence or absence of asthma and AERD (Fig. 4b).

Moreover, the survival curve showed that the ‘surgery-free survival rate’ was significantly longer in non-allergic patients compared to allergic patients (Fig. 5b; P < 0.05). The ‘surgery-free survival rate’ for revision ESS did not significantly differ in patients with primary or revision surgery (Fig. 5e; P = 0.220), in patients with or without asthma (Fig. 5c; P = 0.350) and in patients with and without AERD (Fig. 5d; P = 0.098).

Comparing the baseline characteristics, tissue IL-5 is significantly higher in patients undergoing revision surgery [360.84 pg/ml (103.92–521.84)] compared to patients who did not need revision surgery (with and without NP recurrence; 112.94 pg/ml (43.00–228.88); P < 0.05; see Additional file 1: Table S2). Importantly, we identified tissue IL-5 as a significant predictor for revision surgery over the follow-up period (OR 1.004, 95% CI 1.001 to 1.008, P < 0.05). This implicates that the higher the IL-5 levels pre-operatively the higher the risk for a revision surgery.

Additionally, tissue IgE (Fig. 5g; P = 0.166) or SAE-IgE (Fig. 5h; P = 0.294) was not a predictive factor for revision surgery.

The use of medication did not significantly differ between the 3 moments of contact, although asthma medication use tended to be higher during follow-up (Additional file 1: Fig. S1). In 2006 64.0% of patients and in 2012 52.6% of patients used intranasal corticosteroids as an ongoing treatment, compared to 57.9% pre-operatively. Additionally, 5 patients received during the 12-year follow-up period monoclonal antibodies in double blind randomized controlled trials (3 patients received anti-IL-5 [22], 1 patient anti-IgE [25] and 1 patient both). Four of these patients demonstrated a clinical response to the treatment, and four patients required surgery after termination of treatment. The Kaplan–Meier curves were generated for revision surgery and revision surgery plus biologics but did not show any difference.

When inquiring general therapeutic relief after the ESS in 2000 over the 12-year period, 8 out of the 38 (21.1%) patients reported a complete, 36.8% a marked, 26.3% a moderate and 13.2% a slight relief over time.

The EPOS control test was performed retrospectively and showed that initially all patients (97.4%) were uncontrolled before surgery. Six years after surgery 40% was uncontrolled and 44% was partially controlled based on symptoms and medication needed. Only 16% was controlled 6 years after surgery. At the endpoint, 12 years after inclusion, 47.4% was uncontrolled, 26.30% was partially controlled and 26.30% was controlled.

We asked patients if they would do the ESS again 12 years ago with the knowledge they have today. Thirty-six (94.7%) out of the 38 patients answered ‘Yes’. Two patients (5.3%) answered ‘No’, of which one was due to postoperative bleeding.