Does endoscopic sinus surgery improve olfaction in nasal polyposis?

BACKGROUND

Olfactory dysfunction in chronic rhinosinusitis (CRS) is among the most prominent manifestations of CRS and contributes to CRS diagnostic criteria.1 This is particularly true in CRS with nasal polyps (CRSwNP). The detailed mechanism of olfactory dysfunction in CRSwNP is unclear. Leading theories include mechanical obstruction of the airflow to the olfactory neuroepithelium, chronic inflammatory injury to the neuroepithelium, or a combination of both.2 Impaired olfaction not only leads to major safety concerns but also significantly impacts quality of life (QOL). Thus, it is an important element in several CRS health-related QOL measures.1

Endoscopic sinus surgery (ESS) is an effective treatment for medically refractory CRSwNP. The goals of surgery include the removal of polyps obstructing the nasal cavity and sinuses, as well as the creation of patent sinus drainage pathways. These wide sinusotomies facilitate the application of topical postoperative medical treatments designed to minimize disease progression and treat infection.3 Thus, it is possible that surgery also will improve olfaction. However, evidence supporting olfactory improvement after ESS is conflicting, with estimates of improvements ranging from 25% to 100%.4

This review seeks to evaluate the best available evidence, focusing on the change in objective olfactory outcomes in CRSwNP after surgical intervention.

LITERATURE REVIEW

Multiple prospective studies have evaluated the effects of ESS on olfaction in CRSwNP patients. This review focuses on the use of quantifiable olfactory measures to evaluate the effect of ESS on olfaction and assesses the sustainability of the outcome. Specifically, we have chosen a recent high-quality systematic review and individual studies (not included in the systematic review) that provide the best evidence on which to base clinical practice patterns.

Based on the data presented in 31 studies, Kohli et al. showed in a systematic review that CRSwNP patients experienced significant olfactory improvement after ESS on objective and subjective scales. Specifically, they examined the studies that looked at olfaction outcomes in CRSwNP patients and compared them to a heterogenous group of CRS patients that included CRS without nasal polyps (CRSsNP) and CRSwNP patients.4 Overall, polyp patients tended to have greater improvement in olfaction after surgery.

Objective olfactory measures included both the 40-item Smell Identification Test (SIT-40) and Sniffin' Sticks (Burghart, Wedel, Germany). CRSwNP patients, as compared to a heterogenous CRS cohort, have a greater improvement (higher score indicates better smell) in both measures. The SIT-40 mean difference was 7.87 for the CRSwNP group versus 3.49 for the comparison group (P = .006). The Sniffin' Sticks (Burghart) threshold, discrimination, and identification mean difference was 12.81 ± 5.87 for the CRSwNP group versus 28.84 ± 7.98 for the comparison group (P < 0.0001).4 The olfaction improvements in CRSwNP patients were found to be above the minimal clinically significant threshold for both the SIT-40 and Sniffin' sticks (Burghart) tests ( ≥ 4 and ≥ 5.5, respectively). Similarly, in seven studies (437 patients) the subjective measures of olfaction using the visual analog score (VAS) showed greater reduction (improvement in olfaction) among CRSwNP patients (7.92 ± 1.89 to 3.38 ± 2.21; P < .0001). It is important to note that CRSwNP patients generally started with a lower baseline olfaction and postoperatively did not achieve the same level of olfactory function as the heterogenous group. Overall, heterogenous CRS patients with baseline dysosmia experienced greater improvements compared to normosmic patients. However, all CRS (including those with and without polyps) patients remained in the hyposmic range postoperatively based on SIT-40 scores.

Even with a systematic review based on a large population of patients, several limitations were described by Kohli et al, such as a nonunified CRS diagnostic criteria, inconsistent reporting of perioperative medical management, variable follow-up durations among the studies, and unreported extent of surgical intervention. It is possible that perioperative medical treatment and a more extensive surgical intervention may influence the olfactory outcome. Furthermore, the use of different objective measures among the studies, without comparable validity, limits confidence in the conclusions.

Andrews et al. conducted a prospective study on 113 CRS patients, 60 of whom had CRSwNP.3 Olfaction was evaluated subjectively with VAS and objectively using the University of Pennsylvania Smell Identification Test (UPSIT) (University of Pennsylvania, Philadelphia, PA). Patients were also assessed using Nasal Obstruction and Symptom Evaluation (NOSE) Scale, Sino-Nasal Outcome Test (SNOT-22), and Lund-Kennedy Scores. All scores were taken preoperatively and at 6 months postoperatively. The study described lower baseline UPSIT (University of Pennsylvania) scores in CRSwNP group (21.9 [ ± 10.4]) compared to CRSsNP (29.5 [ ± 7.8]) (P < 0.0001). Postoperatively, UPSIT (University of Pennsylvania) scores were significantly improved among the CRSwNP group only (P < 0.04). Statistically significant improvements also were reported for the SNOT-22, VAS for olfaction and overall symptoms, and NOSE Scale and Lund-Kennedy Scores for the CRSwNP group (P < 0.01). Moreover, the olfactory improvement, measured by UPSIT (University of Pennsylvania), correlated with the decrease in SNOT-22 scores (r2 = 0.379, P < 0.001). Although the results of this study seem promising, 6 months of follow-up postoperatively might not reflect long-term sustainable results.

Levy et al. experienced a relatively similar olfactory improvement in CRSwNP in their cohort.1 122 patients were studied; of those, 38 had CRSwNP. Brief Smell Identification Test (BSIT) (Sensonics Inc, Haddon Heights, NJ) was documented preoperatively and at 6, 12, and 18 months postoperatively (score range 0–12; ≥ 9 normal, ≤ 8 indicates olfactory dysfunction). Levy et al. reported that CRSwNP patients, who had olfactory dysfunction at baseline (average BSIT = 3.9), experienced significant olfactory improvement at 6 months postoperatively (average BSIT = 7.6; P = 0.001), which was sustained at 12 months. However, CRSwNP patients with normal olfaction (average BSIT = 10.8) had no significant difference at 6 months (average BSIT = 11), with worsening scores documented at 12 months (average BSIT = 10.5) and 18 months (average BSIT = 10.1).

BEST PRACTICE

CRSwNP might be the most surgically responsive group in CRS in that current evidence supports short-term olfactory improvement after ESS in the treatment of CRSwNP. However, there is a lack of consensus on the extent of surgical intervention required to achieve improvement in olfaction. Also, variability in perioperative medical therapy, patient compliance, and severity of preoperative disease are factors that meaningfully impact the treatment outcome and are poorly adjusted for in the literature. When considering ESS, patients may be counseled regarding the unpredictable olfaction outcome after surgery, especially for the long-term, due to confounding factors such as medical treatments and variable disease severity.

In order to reach solid conclusions on the role of ESS on olfaction, the literature needs high-level and well-powered studies comparing olfaction outcomes of medical and surgical interventions separately, with consistent olfaction measures and long-term follow-up. We also suggest having a clear classification system of the extent of surgical intervention that is correlated with preoperative condition to better identify the surgery with the best olfaction outcomes.

LEVEL OF EVIDENCE

The highest level of evidence in this article is level 2a in the meta-analysis, and level 2b in the individual articles.