Radiological score, asthma and NSAID-exacerbated respiratory disease predict relapsing chronic rhinosinusitis

Abstract

Objectives: The aim was to evaluate the predictive potential of Sinonasal Radiological (SR) and the Lund-Mackay (LM) score of sinus computed tomography (CT) scans on postoperative relapses of chronic rhinosinusitis (CRS).

Materials and methods: CRS patients (n = 483, 12-80 years) underwent routine sinus CT scans. The SR score was defined by obstructed frontal recess (0 = no, 1 = yes) and visualization of middle and inferior turbinate (0 = anatomy can be easily visualized, 1 = anatomy cannot be easily visualized) on each side (a total of 0-6 points). Associations were analyzed by nonparametric, survival and Cox's proportional hazard models.

Results: Revision endoscopic sinus surgery (ESS) was performed in 133 (28.0%) patients on average (min-max) of 3.2 (0-12) years after performing the sinus CT scans. Of the 408 patients who underwent the baseline ESS, high preoperative SR or LM scores significantly predicted revision ESS (p < 0.001) and peroral corticosteroid courses purchased during the follow-up (p = 0.009 and p < 0.001, respectively for SR- and LM-scores). In multivariable analysis, both SR score and asthma and/or NSAID exacerbated respiratory disease (N-ERD) were significantly associated with revision ESS risk (p = 0.035, p = 0.007, respectively).

Conclusion: LM and SR and a history of asthma or N-ERD predict CRS relapses, which may help in decision-making.

Keywords: ESS; antrochoanal polyp; aspirin exacerbated respiratory disease; aspirin intolerance; inflammation; nasal polyp; recurrence; revision surgery; sinus surgery; sinusitis.

FIGURE 1

Predicting probability of the two radiological scores of sinus computer tomography (CT) scans of chronic rhinosinusitis (CRS) patients. The scores were Sinonasal Radiological (SR) score (a total of 0–6 points) and Lund‐Mackay (LM) score (a total of 0–24 points). (A) All 483 subjects. Both scores poorly identify the patient group undergoing endoscopic sinus surgery (ESS) at the baseline. (B) Only 408 subjects underwent baseline ESS within one year after baseline CT. Both scores significantly identify revision ESS. With the threshold value LM score ≥ 13/24, the sensitivity was 65% and specificity 61% for detecting those who needed revision ESS during follow‐up after the baseline ESS. With the threshold value SR score ≥ 2/6, the sensitivity was 72% and specificity 60% for detecting those who needed revision ESS during follow‐up after the baseline ESS. The threshold values were counted by the coordinates of the ROC curve (data not shown). AUC, area under curve;AUROC, area under the receiver operating characteristic curve; CI, confidence interval.

FIGURE 2

Kaplan–Meier survival curves showing how two scores of computed tomography scans taken before the baseline surgery are able to predict the risk of revision endoscopic sinus surgery (ESS). The two radiological scores were Sinonasal Radiological (SR) score (a total of 0–6 points) and Lund‐Mackay (LM) score (a total of 0–24 points). The patient group (408 patients) who had undergone baseline ESS within one year after the baseline CT was observed. Baseline = ESS performed within one year from the date of CT scan; Revision = ESS performed over a year after the date of the CT scan; p‐values by log rank test.