Introduction
Allergic rhinitis, an immune response mediated by immunoglobulin E (IgE), affects an estimated 10 to ~40 per cent of the global population, demonstrating a sustained upward trend.Reference Chong and Chew1–Reference Leth-Møller, Skaaby and Linneberg3 Common allergic rhinitis symptoms include nasal itching, sneezing, runny nose, and nasal congestion. Additionally, some patients may encounter symptoms of allergic nasal conjunctivitis such as tearing, eye itching and/or eye redness. Severe allergic rhinitis can significantly impair physical and mental function, leading to diminished quality of life and economic repercussions.Reference Brożek, Bousquet, Agache, Agarwal, Bachert and Bosnic-Anticevich4,Reference Yoo, Choi and Mo5 The treatment of allergic rhinitis varies depending on the severity of the symptoms, including avoiding contact with allergens, medication, immunotherapy, surgical treatment, and combination therapy.Reference Bousquet, Schünemann, Togias, Bachert, Erhola and Hellings6,Reference Wise, Damask, Roland, Ebert, Levy and Lin7 Pharmacological intervention is widely employed as the primary approach for managing allergic rhinitis, with commonly used medications including H1 antihistamines, leukotriene receptor antagonists, glucocorticoids, anticholinergics, decongestants, and specific immunotherapy.
Despite the good results shown by drug therapy, approximately 10 to ~22 per cent of allergic rhinitis patients still do not respond effectively to such treatment measures.Reference Klimek, Sperl, Becker, Mösges and Tomazic8,Reference Balai, Gupta, Jolly and Darr9 Therefore, surgical treatment has emerged as a viable alternative for situations where there is no response to medication or low adherence, drowsiness, and decreased efficacy due to long-term continuous use of medication.Reference Lin, Lin, Su and Chang10 Various surgical treatment options for allergic rhinitis exist, including inferior turbinate ablation or partial turbinate resection for the inferior turbinate, alar nerve resection or posterior nerve resection for the posterior nasal nerve, and inferior turbinate surgery combined with posterior nasal nerve resection.Reference Balai, Gupta, Jolly and Darr9 Among these, radiofrequency ablation of the inferior turbinate is a frequently employed surgical technique in clinical practice.
Radiofrequency ablation is a well-established technology that enables precise targeting of narrow areas while preserving the integrity of surrounding healthy mucosal tissue. It has been widely employed in various medical disciplines, including otolaryngology. This innovative approach involves localised heating of submucosal tissue without causing any damage to its epidermal and mucosal layers. As part of the healing process, it induces fibrosis and promotes tissue volume reduction. Moreover, radiofrequency ablation offers the advantages of simplicity in operation and minimally invasive techniques, making it suitable for outpatient settings under local anesthesia.Reference Irawati, Bashiruddin, Rengganis, Lisnawati, Dharma and Harahap11,Reference Neiderman NN, Caspi, Eisenberg, Halevy, Wengier and Shpigel12
Numerous researchers have reported the favourable therapeutic effect of radiofrequency ablation on allergic rhinitis, with no significant increase in adverse reactions. However, there is a lack of a comprehensive systematic review and meta-analysis that assesses the safety and efficacy of radiofrequency ablation for allergic rhinitis, leaving a void in solid evidence supporting its use as a treatment alternative. Therefore, this study aims to fill this gap by conducting an extensive systematic review and meta-analysis that includes original research listed in major databases. Our goal is to synthesise size effects and expand the sample size to draw more dependable conclusions about the safety and effectiveness of radiofrequency ablation in the treatment of allergic rhinitis. These results will provide valuable insights for clinical practice.
Material and methods
Registration
The systematic review and meta-analysis protocol has been officially registered on the international prospective register of systematic reviews (PROSPERO) platform with registration number CRD42023486427. This study strictly adhered to the guidelines set forth by the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (‘PRISMA’) protocols statement.Reference Page, McKenzie, Bossuyt, Boutron, Hoffmann and Mulrow13
Search strategy
To ensure a comprehensive search for relevant studies, we conducted an extensive search across multiple databases, spanning from inception until October 2023. These included PubMed, the Cochrane Library, Embase, Web of Science, China National Knowledge Infrastructure Database (CNKI), WanFang Database (WF), Chinese Scientific Journal Database (VIP) and Chinese Biomedical Literature Database (CBM). In addition to these databases, we also performed supplementary searches on Baidu Scholar, Google Scholar and the Chinese Clinical Trials Registry. The complete literature retrieval strategy, exemplified by our approach with PubMed, is presented in Table 1.
Searching strategy.
Eligibility criteria
Inclusion criteria
Following the population, intervention, comparator, outcome, and study (PICOS) design format, we established the following inclusion criteria for our investigation. (1) Patient: individuals ≥ 18 years of age with allergic rhinitis. (2) Intervention: the intervention group underwent radiofrequency ablation or a combination of radiofrequency ablation and medication treatment. (3) Comparison: the control group received conventional drug treatment or other non-surgical treatments such as immunotherapy. (4) Outcome: we assessed both primary and secondary outcomes. The primary outcome measure was the total effective rate, defined as the combined cure rate, significant improvement rate, and overall response rate. Secondary outcomes included adverse events and recurrence rate. (5) Study design: randomised, controlled trial. Additionally, eligible studies needed to be published in either English or Chinese.
Exclusion criteria
The following conditions led to the exclusion of studies: (1) patients with allergic rhinitis who had complications related to asthma or other severe atopic diseases; (2) publications that contained duplicated experimental data; (3) publications with incomplete data that could not be fully retrieved through various sources; (4) review articles and literature based on animal experiments; (5) individuals undergoing concurrent nasal surgeries, such as septoplasty, polypectomy, sinusotomy, etc.; (6) non-randomised, controlled experiments; and (7) publications for which full-text materials were unattainable through any means.
Study selection and data extraction
The process of literature screening and data extraction was independently carried out by two researchers, adhering strictly to the pre-established inclusion and exclusion criteria. In case of any disagreements, a third researcher was consulted for resolution. Data extraction was facilitated using Excel 2019 software. The extracted information included various details such as title of the literature, first author, year of publication, average age, follow-up duration, sample size, intervention measures employed in both experimental and control groups, number of observed effective cases, reported adverse events, and recurrence size in both groups. Whenever significant gaps or incompleteness were identified in the reviewed literature sources, proactive communication via phone or email with either the first author or corresponding author was initiated to obtain complete information. The specific approach used for literature selection was in adherence to the Preferred Reporting Items for Systematic Review and Meta-Analyses (‘PRISMA’) extension for scoping reviews guidelines.Reference Tricco, Lillie, Zarin, O'Brien, Colquhoun and Levac14
Assessment of the methodological quality
The Cochrane Risk of Bias Assessment ToolReference Higgins, Altman, Gøtzsche, Jüni, Moher and Oxman15 was utilised in this study to conduct quality evaluation of the included literature. This tool scrutinises six crucial aspects to determine the quality of the literature: (1) methods of randomisation used, (2) concealment of allocation, (3) implementation of blinding, (4) completeness of the outcome data, (5) selective reporting of outcomes, and (6) any other potential sources that could introduce bias. Each of these criteria is classified as having a low, high, or unclear risk of bias. Furthermore, the quality of evidence related to primary outcomes was graded using the grading of recommendations assessment, development and evaluation (‘GRADE’) system and specifically analysed with GRADEprofiler version 3.6 software (Informer Technologies, Los Angeles).
Data synthesis
For data analysis, we utilised the Revman 5.4.1 software, freely available on the official Cochrane website. We employed the Cochrane Q test statistics (chi squared) and I2 to assess heterogeneity among the original literature when integrating research data.Reference Yang, Chen, Yi, Huang, Han and Zhang16 Heterogeneity levels are categorised according to Cochrane's classificationReference Higgins, Thomas, Chandler, Cumpston, Li and Page17 as follows: 0–40 per cent (mild heterogeneity), 30–60 per cent (moderate heterogeneity), 50–90 per cent (significant heterogeneity), and 75–100 per cent (extreme heterogeneity). If the probability value (p) of the statistic is > 0.1 and I2 is < 50 per cent, it indicates no significant heterogeneity among the included studies, leading us to use a fixed-effect model. Conversely, if there significant heterogeneity was present, we applied a random-effects model.
To assess result stability, we conducted a sensitivity analysis, which involves sequentially excluding one study at a time and conducting statistical analyses on the remaining original literature through multiple iterations. In our analysis of binary variables such as overall effectiveness rate, incidence rate of adverse events, and recurrence rate, we employed odds ratio as the effect measure. The 95 per cent confidence interval (CI) represents the potential range of this effect measure with a 95 per cent level of confidence. The combined statistical results are determined by the 95 per cent CI of the combined statistics displayed in a forest plot. We utilised Stata 15.1 (StataCorp, College Station, Texas) software to generate Begg's funnel plots and conduct Egger's tests, assessing publication bias by examining symmetry within the funnel-plot distribution. In cases where asymmetry indicated potential publication bias, we employed the trim-and-fill method to adjust for size effect.
Results
Study selection
Our initial search yielded 483 literature sources. These were distributed as follows: 9 from PubMed, 9 from the Cochrane Library, 9 from Embase, 12 from Web of Science, 95 from China National Knowledge Infrastructure Database (CNKI), 138 from WanFang, 101 from Chinese Scientific Journal Database (VIP), and 110 from Chinese Biomedical Literature Database (CBM). We utilised NoteExpress 3.8 (Aegean Software, Beijing) software for the classification and screening process, discarding reviews that did not meet our inclusion standards. Ultimately, we incorporated 15 randomised, controlled trialsReference Liu18–Reference Li, Hu, Xing, Yue, Zhang and Liang32 into the analysis (Figure 1).
Flowchart of literature selection; CNKI = China National Knowledge Infrastructure Database; WF = WanFang Database; VIP = Chinese Scientific Journal Database; CBM = Chinese Biomedical Literature Database.
Study characteristics
All characteristics of the included trials were meticulously documented. The study encompassed a total sample size of 1430 participants. Of these, 726 individuals were assigned to the intervention group and received radiofrequency ablation, while 704 individuals were allocated to the control group and treated with medication. Table 2 summarises an in-depth overview of the included studies.
Basic characteristics of eligible randomised, controlled trials; I = intervention; C = comparison.
Risk of bias assessment
We employed the Cochrane bias risk assessment tool to evaluate the quality of the literature included in this study. Among the 15 studies, the completeness of outcome data and selective outcome reporting were satisfactorily implemented. However, the random allocation methods, allocation concealment, and blinding were inadequately addressed. Detailed evaluation results can be found in Table 3. A bias-risk graph was generated using Revman 5.4.1 software, as displayed in Figure 2.
Risk of bias in the included randomised, controlled trials.
Risk-of-bias graph.
Effects of interventions
Total effective rate
Fourteen studies reported the total effectiveness rate as an outcome indicator. We performed a meta-analysis on these 14 papers using Revman 5.4.1 software. The total effectiveness rate, which can be categorised as effective or ineffective, represents a binary variable. We utilised the odds ratio as the combined statistic and selected the fixed-effect model for calculation using the Mantel–Haenszel method. Heterogeneity testing results demonstrated homogeneity among the 14 studies containing the outcome indicator of total effectiveness rate (chi squared = 10.50, p = 0.65, I2 = 0 per cent), thus validating the choice of employing a fixed-effect model for combining statistics. Our analysis revealed a statistically significant difference between the experimental group and control group (odds ratio = 3.27, 95 per cent CI = 2.37~4.51, p 0.00001). This indicates that radiofrequency ablation exhibited an overall higher effectiveness rate in treating allergic rhinitis compared to drug-treatment methods. The forest plot generated by Revman 5.4.1 software is presented in Figure 3.
Forest plots of total effective rate; CI = confidence interval; df = degrees of freedom; M–H = Mantel–Haenszel.
Adverse events
Adverse events were documented as outcome indicators in five studies. These studies exhibited homogeneity (chi squared = 2.11, p = 0.72, I2 = 0 per cent), and the odds ratio was used as the combined statistic. Using a fixed-effect model, we synthesised effect sizes. The meta-analysis results revealed no statistically significant difference in adverse reaction incidence between the experimental and control groups (odds ratio = 1.18; 95 per cent CI = 0.67–2.08; p = 0.57), indicating that the observed difference lacked statistical significance (Figure 4).
Forest plots of adverse events; CI = confidence interval; df = degrees of freedom; M–H = Mantel–Haenszel.
Rate of recurrence
Three studies investigated recurrence rates post-treatment, and the heterogeneity test results showed no significant variation (chi squared = 0.37, p= 0.83, I2 = 0 per cent). Consequently, a fixed-effect model was chosen for meta-analysis. The findings demonstrated that the experimental group exhibited markedly reduced recurrence rates following treatment compared to the control group (odds ratio = 0.27; 95 per cent CI = 0.12~0.62; p = 0.002), as depicted in Figure 5.
Forest plots of rate of recurrence; CI = confidence interval; df = degrees of freedom; M–H = Mantel–Haenszel.
Data synthesis
Publication bias
Data on the overall effectiveness rate was contributed by 14 studies, which served as an indicator for constructing a funnel plot to analyse publication bias. We employed Begg's funnel plot and conducted Egger's test to assess bias in the literature.Reference Chen, Wu, Wu, Zheng, Mei and Mei33 The regression diagram of Egger's test is depicted in Figure 6. The results indicated T = 2.54 and p = 0.026, suggesting a low likelihood of symmetry in the funnel plot and a high probability of publication bias. To adjust for potential size-effect distortion, we applied the trim-and-fill method. After adjustment, the odds ratio remained stable at 2.60 with a 95 per cent CI of 1.89–3.57. Notably, there was no significant change compared to before adjustment, implying that publication bias had minimal effect on our meta-analysis results and ensuring their stability.Reference Zhang and Zhong34 The trim-and-fill funnel plot is displayed in Figure 7.
Regression diagram of Egger's test.
Funnel plot after trimming and filling; SE = standard error. These three data points with squares around them are the dummy studies added by the trim-and-fill method, and the funnel plot is able to achieve symmetry if the three dummy studies are added.
Sensitivity analysis
We utilised sensitivity analysis to evaluate the stability of our meta-analysis findings. This analysis was performed using Stata 15.1 software, with a particular emphasis on the overall effective rate. The outcomes consistently indicated a robust total effective rate. Figure 8 provides a visual representation of the sensitivity analysis plot.
Sensitivity analysis plot; CI = confidence interval.
Level of evidence
We utilised GRADEprofiler 3.6 software to assess evidence quality associated with the total effectiveness rate, which served as an outcome indicator. This tool evaluates quality by examining various factors, including risk of bias, inconsistency in research, indirectness of evidence, precision of results, and publication bias. When multiple influencing factors affect the total effectiveness rate in this study, a downgrade in the evidence quality is justified.Reference Zhao, Hu, Guan, Yi, Zhang and Chen35 Figure 9 outlines detailed evaluation specifics. The evidence grade for the total effectiveness rate, when used as an outcome indicator, is categorised as low.
Level of evidence.
Discussion
By integrating the effect sizes derived from the 15 studies included, we obtained the following results: total effectiveness rate (odds ratio = 3.27, 95 per cent CI 2.37–4.51); recurrence rate (odds ratio = 0.27, 95 per cent CI 0.12–0.62); and incidence of adverse events (odds ratio = 1.18, 95 per cent CI 0.67–2.08). For binary variable data, the odds ratio for values less than one suggests that intervention measures in the intervention group can decrease event occurrence; odds ratio values equal to one suggest no effect on event occurrence; and values greater than one suggest an increased event occurrence due to intervention measures in the intervention group.
Our findings show that both the upper and lower limits of the odds ratio for the total effectiveness rate exceed one, while both the upper and lower limits of the odds ratio for the recurrence rate are less than one. The upper and lower limits of the odds ratio for the incidence of adverse events encompass one. In conclusion, radiofrequency ablation may exhibit superior treatment effectiveness rates and reduced recurrence rates compared to drug treatment without significantly increasing adverse reaction incidence levels when compared clinically.
The grading of recommendations assessment, development and evaluation (‘GRADE’) evaluation system classifies evidence levels and clearly presents evaluation items, aiding clinicians in understanding the effectiveness and feasibility of intervention measures for clinical decision-making.Reference Zhao, Hu, Guan, Yi, Zhang and Chen35 For the outcome indicator of total effectiveness rate, we used GRADEprofiler 3.6 software to evaluate the evidence quality. The grading result was of moderate quality, not high quality, due to the lack of sufficiently high-quality papers containing this outcome indicator. Among these 14 papers, 10 did not explicitly state their randomisation methods, while 3 employed an open random list method that failed to implement allocation concealment. Furthermore, blinding was not mentioned in 10 papers and was unachievable in 4 others because patients signed informed consent forms or group agreements.
Egger's test indicated a high probability of publication bias (T = 2.54, p = 0.026). However, after conducting initial screening and post-retrieval re-screening, as per our appeal process, we obtained literature with trim-and-fill analysis results consistent with those prior to the analysis. We thus concluded that, despite potential publication bias concerns, radiofrequency ablation treatment is more effective for managing allergic rhinitis. If future research produces additional high-quality original papers on this topic, we will perform further analyses incorporating them into our meta-analysis to boost credibility and potentially elevate the grade level assigned based on evidence quality.
Nasal patency, intact mucosa, and active mucociliary clearance are vital for optimal nasal airflow. The turbinates, which maximise the respiratory surface area, play a critical role in nasal functioning, including air warming, moistening, and filtration. However, enlarged turbinates can diminish nasal airflow, impair function, and cause nasal obstruction.Reference Hsu and Suh36
Allergic rhinitis triggers a series of events characterised by type 2 inflammation. These events involve the recruitment of effector cells, release of mediators, and production of cytokines, leading to vasodilation, vascular congestion, inflammatory swelling, and eventual tissue remodelling. These factors collectively contribute to the enlargement of the turbinates.Reference Ciprandi, Cirillo, Klersy, Marseglia, Caimmi and Vizzaccaro37,Reference Ciprandi and Tosca38
Radiofrequency ablation is a preferred surgical method for reducing inferior turbinates. By decreasing their size, radiofrequency ablation enhances the nasal airway and improves nasal symptom perception.Reference Türk, Korkut, Kaya, Salepçi, Ünsal and Coşkun39 Radiofrequency ablation has been reported to alleviate sneezing symptoms in allergic rhinitis by damaging the posterior branches of the nasopalatine nerve.Reference Lin, Lin, Friedman, Chang, Su and Chen40 Moreover, an imbalance in the regulation of sympathetic, parasympathetic and nociceptive nerves that innervate the nasal mucosa is implicated in controlling vascularity and glandular secretion within the mucosa. When radiofrequency energy is applied to submucosal regions, it obliterates small vessels and destroys mucosal glands, leading to circumferential scar formation, which is a significant contributor to the observed benefits of radiofrequency ablation.
While some studies attribute sneezing alleviation after radiofrequency ablation to the destruction of post-nasal nerve branches, it remains intriguing that turbinate reduction alone yields such remarkable results, given that the posterior nasal nerve innervates the entire nasal mucosa. Furthermore, the involvement of sympathetic, parasympathetic and nociceptive nerves in nasal mucosa regulation has been proposed. By applying radiofrequency energy to the submucosal layer of inferior turbinates, radiofrequency ablation achieves scar formation, contributing to its positive effects.Reference Lin, Lin, Friedman, Chang, Su and Chen40–Reference Banov and Lieberman42 However, the precise mechanisms by which radiofrequency improves itching perception, rhinorrhoea symptoms, and sneezing in allergic rhinitis patients remain unclear.
While the procedure of radiofrequency ablation of the nasal concha is relatively simple and can be conducted in outpatient settings, it remains an invasive surgical intervention for patients. Factors such as intra-operative bleeding, post-operative pain, and other immediate or delayed reactions may lead to a less-than-optimal patient experience. Moreover, radiofrequency ablation carries a certain risk of recurrence, which could present challenges for patients who may be hesitant to undergo repeat radiofrequency ablation procedures. As a result, current treatment strategies for allergic rhinitis primarily favour pharmacological interventions.
Compared to radiofrequency ablation, drug therapy provides a non-invasive approach that enables self-medication. Radiofrequency ablation of the nasal concha should only be considered as an alternative option when drug treatment is ineffective or when patient compliance is poor.Reference Yang, Lee, Kim, Ryu, Kang and Kim43,Reference Park, Lee, Kim, Kim, Yang and Kim44 Some studies have suggested that combined drug therapy following radiofrequency ablation yields superior efficacy compared to radiofrequency ablation alone.Reference Huang, Zhang, Fu and Zhong45 This meta-analysis also shows that radiofrequency ablation outperforms standalone drug treatment in terms of effectiveness. Patients with good compliance can consider combining drug therapy with radiofrequency ablation of the nasal concha to further enhance therapeutic outcomes in allergic rhinitis.
• Based on previous reports, efficacy and safety of radiofrequency ablation are compared to nonsurgical medical therapy for treating allergic rhinitis
• Radiofrequency ablation is a safe and effective method for managing allergic rhinitis
• Utilisation of radiofrequency ablation presents a viable alternative for patients suffering from allergic rhinitis who exhibit inadequate response to pharmacotherapy or demonstrate suboptimal adherence
Conclusion
This systematic review and meta-analysis provide the first comprehensive evaluation of radiofrequency ablation for treating allergic rhinitis. Although radiofrequency ablation has proven to be effective and safe in managing allergic rhinitis, careful consideration is required due to methodological limitations and the quality of evidence. As a result, future research should focus on large-scale, multicentre, high-quality randomised, controlled trials to validate these conclusions.
