Indirect Treatment Comparison of Tezepelumab Versus Other Biologics in Severe Chronic Rhinosinusitis with Nasal Polyps: A Systematic Literature Review and Network Meta-analysis Authors: Joseph K. Han, Joaquim Mullol, Claire Hopkins, Oliver Pfaar, Martin Desrosiers, Jason Steenkamp, Danny Gibson, Santiago Zuluaga Sanchez, Andrew Foster, Christopher S. Ambrose Published: 2026 ABSTRACT IntroductionThis study compared the efficacy of tezepelumab with other approved biologics and endoscopic sinus surgery (ESS) using indirect treatment comparisons (ITCs) in adults with severe, uncontrolled chronic rhinosinusitis with nasal polyps (CRSwNP).MethodsRandomized controlled trials (RCTs) identified via a systematic literature review (SLR) were synthesized in a network meta-analysis (NMA). Outcomes included change from baseline in nasal polyp score (NPS), nasal congestion/obstruction score (NCS/NOS), difficulty with sense of smell (DSS), Sino-Nasal Outcome Test-22 (SNOT-22) score, Lund Mackay score (LMS), time to nasal polyposis (NP) surgery decision, proportions requiring NP surgery or systemic corticosteroids (SCS), time to SCS use, and the University of Pennsylvania Smell Identification Test (UPSIT). Mean differences, hazard ratios, and odds ratios with 95% credible intervals were estimated for available treatments at 52 weeks. Not all endpoints were reported across trials; comparisons were conducted only where both treatments reported the outcome.ResultsAt 52 weeks, tezepelumab showed comparable efficacy to dupilumab across all evaluated endpoints, apart from the proportion of patients requiring surgery, where tezepelumab was associated with a lower surgery rate than dupilumab. Tezepelumab consistently outperformed mepolizumab, depemokimab, and placebo. In analyses of < 30-week endpoints, tezepelumab outperformed omalizumab for all outcomes except SNOT-22, where no difference was observed. Sensitivity analyses including ESS showed benefits for tezepelumab across endpoints except for NPS, for which ESS ranked highest.ConclusionTezepelumab achieved similar efficacy to dupilumab across all evaluated endpoints with the exception of the proportion of patients requiring surgery, where tezepelumab demonstrated a numerically greater reduction compared with dupilumab. Results consistently favoured tezepelumab and dupilumab over depemokimab, mepolizumab, and omalizumab.Supplementary InformationThe online version contains supplementary material available at 10.1007/s12325-026-03655-8. Key Summary Points Why carry out this study?With multiple biologics approved for chronic rhinosinusitis with nasal polyps (CRSwNP), but no head-to-head trials, there is an unmet need to understand their comparative efficacy.This study used network meta-analyses to evaluate the relative efficacy of tezepelumab versus other biologics and endoscopic sinus surgery (ESS) for adults with severe, uncontrolled CRSwNP.What was learned from the study?At 52 weeks, tezepelumab showed comparable efficacy to dupilumab across most outcomes and demonstrated a lower proportion of patients requiring nasal polyp (NP) surgery, while also providing improvements over mepolizumab, depemokimab, omalizumab, and placebo across multiple endpoints.These findings suggest that tezepelumab may offer a favourable efficacy profile, including reduced need for surgery and systemic corticosteroids (SCS), for patients with severe CRSwNP and may help guide future treatment selection in the absence of direct comparative trials. Why carry out this study?With multiple biologics approved for chronic rhinosinusitis with nasal polyps (CRSwNP), but no head-to-head trials, there is an unmet need to understand their comparative efficacy.This study used network meta-analyses to evaluate the relative efficacy of tezepelumab versus other biologics and endoscopic sinus surgery (ESS) for adults with severe, uncontrolled CRSwNP.What was learned from the study?At 52 weeks, tezepelumab showed comparable efficacy to dupilumab across most outcomes and demonstrated a lower proportion of patients requiring nasal polyp (NP) surgery, while also providing improvements over mepolizumab, depemokimab, omalizumab, and placebo across multiple endpoints.These findings suggest that tezepelumab may offer a favourable efficacy profile, including reduced need for surgery and systemic corticosteroids (SCS), for patients with severe CRSwNP and may help guide future treatment selection in the absence of direct comparative trials. Introduction Chronic rhinosinusitis with nasal polyps (CRSwNP) is an inflammatory disease of the sinonasal tissue that affects approximately 1–4% of the global population. CRSwNP is characterized by bilateral nasal polyps accompanied by symptoms such as nasal obstruction or congestion, anterior or posterior nasal drainage, hyposmia, and facial pressure or pain [1]. Persistent inflammation of the airway and sinonasal tissue results in chronic symptoms that substantially impair quality of life by disrupting olfaction, sleep, emotional well-being, concentration, and productivity [2–5]. Although its precise etiology remains unclear, most patients exhibit a type 2 inflammatory endotype characterized by elevated interleukin (IL)-4, IL-5, IL-13, increased IgE, eosinophil counts, and other inflammatory markers in nasal polyp tissue and sinonasal mucosa [3, 6, 7]. Type 2 inflammation is present in approximately 70–90% of CRSwNP cases and is associated with more severe symptoms, reduced health-related quality of life, diminished responsiveness to conventional therapies, and increased risk of recurrence compared with non-type 2 disease [3, 5, 8]. CRSwNP is often comorbid with other type 2 inflammatory conditions, including asthma, allergic rhinitis, and aspirin-exacerbated respiratory disease (AERD) [8–11]. Given these features, biologic therapies targeting inflammatory pathways represent a major advancement in the management of CRSwNP. The current treatment paradigm for CRSwNP includes nasal saline irrigation, intranasal corticosteroids, and the conservative use of oral corticosteroids (i.e. one to two courses per year). Patients whose disease remains uncontrolled with medical management are candidates for surgical intervention, which aims to remove the diseased tissue to improve anatomical function and access for irrigation and intranasal corticosteroids [8]. Patients with type 2 inflammation-mediated CRSwNP, however, are often refractory to intranasal corticosteroid therapy and can experience recurrence following surgical intervention [3, 12]. A population-based study of patients with CRSwNP who had previously undergone endoscopic sinus surgery (ESS) found that 15.9% of patients required at least one revision surgery and 4.4% underwent at least two revisions within 1–3 years of the initial surgery [12]. Additionally, a retrospective analysis using long-term population-based data with a mean follow-up of 9.7 years found that 29.9% of patients with CRSwNP required at least one revision. Surgery also entails risks of postoperative complications such as infection, bleeding, haemorrhage, orbital complications, and cerebrospinal fluid leaks [8, 13]. Long-term systemic corticosteroid use is associated with substantial adverse effects including hip fractures, vertebral fractures, cataract formation, and psychiatric episodes requiring inpatient admission [14]. One analysis found that the risks of oral corticosteroid use exceeded the risks of ESS in patients with CRSwNP that require more than one course of oral corticosteroids in 2 years [14]. Overall, one third of patients are not adequately controlled with the current treatment options, necessitating the use of disease-modifying biologics to control inflammation in refractory CRSwNP [8, 13, 15, 16]. Biologics for the treatment of CRSwNP are monoclonal antibodies targeting the type 2 inflammatory pathway [16]. Briefly, dupilumab inhibits IL-4 and IL-13 activity, mepolizumab (as well as the recently approved longer-acting depemokimab) inhibit IL-5 activity, and omalizumab binds to IgE [17, 18]. Tezepelumab is a recently approved monoclonal antibody that inhibits thymic stromal lymphopoietin (TSLP) to reduce inflammation in patients with CRSwNP. TSLP is an epithelial cytokine that initiates and amplifies type 2 and other inflammation pathways and acts upstream of the type 2 mediators targeted by other biologics [19]. By targeting TSLP, tezepelumab reduces downstream inflammatory mediators and biomarkers, including IL-5, IL-13, eosinophils, IgE, and fractional exhaled nitric oxide (FeNO). The efficacy of TSLP blockade by tezepelumab in adults with CRSwNP was established in the randomized, double-blind, placebo-controlled, phase 3 WAYPOINT trial (NCT04851964) [20]. WAYPOINT participants received either tezepelumab 210 mg or placebo every 4 weeks for a total duration of 52 weeks and demonstrated statistically significant improvements in nasal congestion scores (NCS), nasal polyp scores (NPS), Lund Mackay scores (LMS), the Sino-Nasal Outcome Test (SNOT-22), loss of smell as measured by the nasal polyposis symptom diary (NPSD) and University of Pennsylvania Smell Identification Tests (UPSIT), and a significantly reduced need for systemic corticosteroids (SCS) compared with placebo. The need for surgical intervention was nearly eliminated in patients treated with tezepelumab, with a 98% reduction [20]. The comparative efficacy of tezepelumab relative to other biologics has not been evaluated in head-to-head trials. In the absence of direct evidence, indirect comparison methods provide a means to assess relative efficacy. A network meta-analysis (NMA) is a widely used approach to estimate treatment effects between interventions that lack direct head-to-head comparisons in clinical trials, or for which both direct and indirect evidence is available for synthesis. This methodology uses summary (or aggregate) data from published studies and has been broadly adopted in recent years by health technology assessment agencies worldwide, including Canada’s Drug Agency (CDA), the National Institute for Health and Care Excellence (NICE), and the European Network for Health Technology Assessment (EUnetHTA) in the context of evaluating therapies for indications with multiple active comparators. The objective of this study was to conduct an indirect treatment comparison (ITC) evaluating the comparative efficacy of tezepelumab versus other biologic therapies and/or ESS for CRSwNP using a NMA approach. Methods This article is based solely on previously published studies and does not include any new research involving human participants or animals conducted by the authors. Systematic Literature Review A systematic literature review (SLR) was conducted in October 2023 to identify all relevant randomized controlled trials (RCTs) and real-world evidence studies for inclusion in the ITCs. A systematic search of Embase, MEDLINE®, and Evidence-Based Medicine Reviews (EBMR) databases was conducted by an information specialist and peer-reviewed by another senior information specialist in accordance with NICE recommended Peer Review of Electronic Search Strategies (PRESS) checklist [21]. The titles and abstracts of records identified in the literature search were screened against the pre-defined eligibility criteria by two independent reviewers. Briefly, the outlined eligibility criteria captured studies reporting on tezepelumab, benralizumab, dupilumab, mepolizumab, omalizumab, or ESS for the treatment of CRSwNP in patients aged 18 and older (Table 1). Full text review and data extraction were performed by one reviewer and independently validated by a second reviewer; disagreements were resolved by consensus. Extracted data included study design, patient characteristics, and key outcomes of interest. Risk of bias assessments were conducted according to the Cochrane Risk of Bias Assessment Tool for Randomized Controlled Trials [22]. The SLR was performed according to the Cochrane Handbook for Systematic Reviews of Interventions and reported following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines [23, 24]. The WAYPOINT (NCT04851964) and ANCHOR trials were incorporated as additional evidence beyond the SLR, as results from all three studies were published after completion of the SLR. Efficacy data for tezepelumab were sourced from the phase III WAYPOINT trial, which reported results in 2025 and provides the foundational evidence for tezepelumab in CRSwNP [20]. The ANCHOR-1 and ANCHOR-2 depemokimab trials were similarly added because their findings became available only after completion of the SLR and represent the first phase III evaluations of depemokimab as a comparator in CRSwNP [20]. Table 1Eligibility criteria for the identification of relevant studiesCriteriaInclusion criteriaExclusion criteriaPopulationPatients ≥ 18 years old with CRSwNPKey subgroups of interest included patients with eosinophilic disease and patients with comorbid asthmaOnly patients < 18 years oldInterventionAny treatment for CRSwNPTreatments not for CRSwNPComparatorsAny treatment for CRSwNPTreatments not for CRSwNPOutcomesNPSNCS/NOSSNOT-22LMSNeed for NP surgeryaUPSITaTotal symptom score, VASAEsOutcomes not related to the efficacy or safety of CRSwNP treatmentsStudy designRCTsNon-randomized studies:Observational/RWE studies (e.g. case-control, cross-sectional, longitudinal, and cohort studies)Single-arm studiesPilot studiesPublished articles (January 1, 2008 to October 23, 2023)Conference abstracts (January 1, 2021 to October 23, 2023)Animal studies, in vitro studies, case reports, expert opinion articles, commentaries, lettersArticles published prior to 2008Conference abstracts published prior to January 1, 2021LanguageArticles in EnglishbAll non-English articlesAE adverse event, CRSwNP chronic rhinosinusitis with nasal polyps, NCS/NOS nasal congestion/obstruction score, NP nasal polyposis, NPS nasal polyp score, RCTs randomized controlled trials, RWE real-world evidence, SNOT-22 Sino-Nasal Outcome Test, UPSIT University of Pennsylvania Smell Identification Test, VAS visual analogue scaleaOnly applicable to studies investigating biologicsbCitation retrieval was not limited by language. Records were categorized based on language during title and abstract screening stage, and non-English abstracts were excluded. English abstracts with non-English articles were excluded at the full-text screening stage Table 1Eligibility criteria for the identification of relevant studiesCriteriaInclusion criteriaExclusion criteriaPopulationPatients ≥ 18 years old with CRSwNPKey subgroups of interest included patients with eosinophilic disease and patients with comorbid asthmaOnly patients < 18 years oldInterventionAny treatment for CRSwNPTreatments not for CRSwNPComparatorsAny treatment for CRSwNPTreatments not for CRSwNPOutcomesNPSNCS/NOSSNOT-22LMSNeed for NP surgeryaUPSITaTotal symptom score, VASAEsOutcomes not related to the efficacy or safety of CRSwNP treatmentsStudy designRCTsNon-randomized studies:Observational/RWE studies (e.g. case-control, cross-sectional, longitudinal, and cohort studies)Single-arm studiesPilot studiesPublished articles (January 1, 2008 to October 23, 2023)Conference abstracts (January 1, 2021 to October 23, 2023)Animal studies, in vitro studies, case reports, expert opinion articles, commentaries, lettersArticles published prior to 2008Conference abstracts published prior to January 1, 2021LanguageArticles in EnglishbAll non-English articlesAE adverse event, CRSwNP chronic rhinosinusitis with nasal polyps, NCS/NOS nasal congestion/obstruction score, NP nasal polyposis, NPS nasal polyp score, RCTs randomized controlled trials, RWE real-world evidence, SNOT-22 Sino-Nasal Outcome Test, UPSIT University of Pennsylvania Smell Identification Test, VAS visual analogue scaleaOnly applicable to studies investigating biologicsbCitation retrieval was not limited by language. Records were categorized based on language during title and abstract screening stage, and non-English abstracts were excluded. English abstracts with non-English articles were excluded at the full-text screening stage Eligibility criteria for the identification of relevant studies Patients ≥ 18 years old with CRSwNP Key subgroups of interest included patients with eosinophilic disease and patients with comorbid asthma NPS NCS/NOS SNOT-22 LMS Need for NP surgerya UPSITa Total symptom score, VAS AEs RCTs Non-randomized studies: Observational/RWE studies (e.g. case-control, cross-sectional, longitudinal, and cohort studies) Single-arm studies Pilot studies Published articles (January 1, 2008 to October 23, 2023) Conference abstracts (January 1, 2021 to October 23, 2023) Animal studies, in vitro studies, case reports, expert opinion articles, commentaries, letters Articles published prior to 2008 Conference abstracts published prior to January 1, 2021 AE adverse event, CRSwNP chronic rhinosinusitis with nasal polyps, NCS/NOS nasal congestion/obstruction score, NP nasal polyposis, NPS nasal polyp score, RCTs randomized controlled trials, RWE real-world evidence, SNOT-22 Sino-Nasal Outcome Test, UPSIT University of Pennsylvania Smell Identification Test, VAS visual analogue scale aOnly applicable to studies investigating biologics bCitation retrieval was not limited by language. Records were categorized based on language during title and abstract screening stage, and non-English abstracts were excluded. English abstracts with non-English articles were excluded at the full-text screening stage Feasibility Assessment The validity of results generated by NMA based on summary-level published data is dependent on the evidence meeting the exchangeability assumption [25]. Under this assumption, all interventions being studied could have been included as comparators in a single clinical trial; thus, all treatments under study are truly competing interventions. Failure to meet this assumption can result in biased estimates of comparative effect. As such, a rigorous qualitative assessment of between-trial heterogeneity was conducted based on trial design, patient eligibility criteria, baseline patient characteristics, placebo response, and trial-specific outcome definitions, which builds upon existing recommendations regarding NMAs [25–28]. To facilitate a connected network via a common comparator, this analysis was restricted to RCTs. Treatments considered included dupilumab, depemokimab, omalizumab, mepolizumab, tezepelumab, and ESS. Control arms varied across the included studies and typically consisted of a placebo plus intranasal corticosteroids, medical therapy, or combinations thereof (e.g. placebo plus intranasal corticosteroids and medical therapy such as nasal rinses). Three RCTs (Gevaert et al. 2011; NCT01066104; Gevaert et al. 2013) reported placebo control arms but did not provide explicit information on concomitant intranasal corticosteroids or additional medical therapy. While benralizumab was initially identified in the literature search, it was excluded during the feasibility assessment as it was not considered a relevant comparator for this analysis. Although primary trials (e.g. OSTRO) reported statistical significance, the magnitude of clinical benefit and the drug’s current trajectory in the CRSwNP treatment landscape led to its exclusion in favour of therapies more closely aligned with current and emerging clinical practice. NMA Methods NMAs were performed to derive simultaneous comparisons of tezepelumab with relevant biologics and ESS for the treatment of CRSwNP for the outcomes presented in Table 2. To distinguish the analytical approach from previously published NMAs in CRSwNP, several methodological considerations were incorporated into the present analysis [29, 30]. A Bayesian hierarchical modelling framework was specified to support probabilistic inference and stabilize treatment effect estimates in a sparse network. The evidence base was limited to CRSwNP-specific RCTs, thereby reducing heterogeneity attributable to inclusion of non-CRSwNP populations (i.e. asthma populations). In addition, time-to-event endpoints, including time to surgery decision and time to SCS use, were incorporated alongside symptom and polyp-based outcomes. Table 2Summary of outcomesOutcomeOutcome typeDescriptionChange from baseline in NPSContinuousEndoscopic scoring of nasal polyp size and severity (rated 0–4 per nostril; maximum sum of 8 bilaterally) [31]Change from baseline in NCS/NOSContinuousAssesses the severity of nasal congestion/obstruction (rated 0–3)Change from baseline in DSS scoreContinuousAssesses worst difficulty in sense of smell (rated 0–3) [32]Change from baseline in SNOT-22 scoreContinuousAssesses the impact of CRS symptoms on QoL (22 questions rated 0–5; total score 0–110) [2, 33]Change from baseline in LMSContinuousAssesses the degree of sinus opacification on CT (scored 0–2 per sinus; maximum sum of 24 bilaterally) [34]Time to NP surgery decisionSurvivalNumber of days from baseline to first decision to perform, or scheduling of, NP surgeryProportion of patients requiring surgery for NPBinaryEstimate of percentage of patients who required NP surgery (actual or planned) by week 52 using Kaplan–Meier methodTime to SCS useSurvivalNumber of days from baseline to first course of SCS at any timepoint up to week 52Proportion of patients requiring SCSBinaryPercentage of patients who received at least one course of SCS during follow-upChange from baseline in UPSIT scoreContinuousObjective measure of olfactory function (40 item identification test; scored 0–40) [32]CRS chronic rhinosinusitis, CT computed tomography, DSS difficulty with sense of smell, LMS Lund Mackay score, NCS/NOS nasal congestion/obstruction score, NP nasal polyposis, NPS nasal polyp score, SCS systemic corticosteroids, SNOT-22 Sino-Nasal Outcomes Test, UPSIT University of Pennsylvania Smell Identification Test Table 2Summary of outcomesOutcomeOutcome typeDescriptionChange from baseline in NPSContinuousEndoscopic scoring of nasal polyp size and severity (rated 0–4 per nostril; maximum sum of 8 bilaterally) [31]Change from baseline in NCS/NOSContinuousAssesses the severity of nasal congestion/obstruction (rated 0–3)Change from baseline in DSS scoreContinuousAssesses worst difficulty in sense of smell (rated 0–3) [32]Change from baseline in SNOT-22 scoreContinuousAssesses the impact of CRS symptoms on QoL (22 questions rated 0–5; total score 0–110) [2, 33]Change from baseline in LMSContinuousAssesses the degree of sinus opacification on CT (scored 0–2 per sinus; maximum sum of 24 bilaterally) [34]Time to NP surgery decisionSurvivalNumber of days from baseline to first decision to perform, or scheduling of, NP surgeryProportion of patients requiring surgery for NPBinaryEstimate of percentage of patients who required NP surgery (actual or planned) by week 52 using Kaplan–Meier methodTime to SCS useSurvivalNumber of days from baseline to first course of SCS at any timepoint up to week 52Proportion of patients requiring SCSBinaryPercentage of patients who received at least one course of SCS during follow-upChange from baseline in UPSIT scoreContinuousObjective measure of olfactory function (40 item identification test; scored 0–40) [32]CRS chronic rhinosinusitis, CT computed tomography, DSS difficulty with sense of smell, LMS Lund Mackay score, NCS/NOS nasal congestion/obstruction score, NP nasal polyposis, NPS nasal polyp score, SCS systemic corticosteroids, SNOT-22 Sino-Nasal Outcomes Test, UPSIT University of Pennsylvania Smell Identification Test Summary of outcomes CRS chronic rhinosinusitis, CT computed tomography, DSS difficulty with sense of smell, LMS Lund Mackay score, NCS/NOS nasal congestion/obstruction score, NP nasal polyposis, NPS nasal polyp score, SCS systemic corticosteroids, SNOT-22 Sino-Nasal Outcomes Test, UPSIT University of Pennsylvania Smell Identification Test Network Connectivity As follow-up time was considered to be an important source of heterogeneity, the feasibility of analyses considered studies reporting at < 30 weeks follow-up and ≥ 30 weeks follow-up separately. Network connectivity was considered for studies not reporting the use of systemic or oral corticosteroids as background therapy, while the feasibility of constructing sensitivity analysis networks was considered including such studies. Considering the outcomes of interest, 15 trials reported SNOT-22, 11 reported LMS, 14 reported NCS or nasal obstruction score (NOS), 18 reported NPS, 14 reported difficulty with sense of smell (DSS), 8 reported UPSIT, and 10 reported treatment response (5 NPS response, 6 SNOT-22 response, 7 other response measures). Five trials reported the proportion of patients requiring SCS, time to first course of SCS, and time to nasal polyposis (NP) surgery decision. Three trials reported on the proportion of patients requiring NP surgery. Statistical Analysis Bayesian NMAs were conducted in accordance with standard ITC methodology detailed in NICE Decision Support Unit (DSU) Technical Support Document (TSD) 2 [31]. Analyses were performed in R (version 3.5.3) using the rjags package (version 4-8) to interface with JAGS (version 4.3.0) [31]. For binary outcomes, NMAs employed a binomial likelihood with a logit link, modelling treatment effects on the log odds ratio scale. Continuous outcomes were analyzed using a normal likelihood with an identity link and treatment effects expressed as mean differences. Time-to-event outcomes were modelled using a normal likelihood with an identity link and treatment effects expressed as log hazard ratios. Vague prior distributions that assume no pre-existing information were assigned for the treatment effects, trial baselines, and between-study variance in all analyses. Trial results were utilized as published, with no adjustment or modification to the reported aggregate data. Posterior distributions were estimated using Markov chain Monte Carlo sampling with four chains, each run for 20,000 iterations following a 20,000-iteration burn-in. Convergence was assessed using the Gelman–Rubin diagnostic (\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\hat{R}$$\end{document}R^ ≤ 1.05), with additional iterations applied if convergence criteria were not met. Placebo served as the reference treatment for all analyses. Primary analyses were based on intention-to-treat (ITT) populations, while subgroup analyses focused on populations using oral corticosteroids. Both fixed-effect and random-effects models were evaluated. Random-effects models assumed a common estimate of between-trial heterogeneity, expressed as the standard deviation of the random effect on the linear predictor scale. Model fit was compared using the deviance information criterion (DIC), with a difference of ≥ 5 points considered meaningful. In the absence of meaningful differences and given limited data for random-effects models (e.g. single-study connections), fixed-effect models were preferred and reported as the primary analyses. Results SLR Results Out of 153 RCTs identified in the clinical SLR, 10 investigated biologics of interest against placebo: three trials of dupilumab, three trials of mepolizumab, and four trials of omalizumab. A further three studies investigated ESS against medical therapy. In trials investigating corticosteroid treatments, improvements were generally observed across outcomes, regardless of steroid type or route of administration (intranasal or oral). Similarly, trials reporting on surgical treatments most commonly reported on ESS and found that surgery typically resulted in improved outcomes. Two trials compared ESS to medical therapy, and both found that ESS improved SNOT-22 scores compared to medical therapy alone; however, one study had a small sample size (n = 31) and the other noted that the difference between groups did not meet the threshold for a minimal important difference. Improvements in the key outcomes of interest (SNOT-22, LMS, NCS/NOS, and NPS) were observed with most treatments in RCTs. Data gaps include a lack of head-to-head trials directly comparing biologic therapies, corticosteroid administration, and medical therapy with surgery. Additionally, few studies provided subgroup analyses for patients with eosinophilic disease, a key subgroup of interest. Feasibility Assessment Results The feasibility assessment encompassed the 16 studies outlined in Table 3, focusing on trials investigating biologic therapies or surgery for the treatment of CRSwNP. These therapies included depemokimab, dupilumab, mepolizumab, omalizumab, tezepelumab, and ESS. Table 3Overview of RCTs included in NMAsTrial; NCTPhaseTreatment armsRandomized NStudy duration (weeks)Tezepelumab WAYPOINT; NCT048519643Tezepelumab + intranasal corticosteroid40052Placebo + intranasal corticosteroidDepemokimab ANCHOR-1; NCT05274750a [18]3Depemokimab 100 mg SC + medical therapyb27652Placebo + medical therapyb ANCHOR-2; NCT05281523a [18]3Depemokimab 100 mg SC + medical therapyb26452Placebo + medical therapybDupilumab SINUS-52; NCT02898454 [35]3Dupilumab + mometasone furoate44852Placebo + mometasone furoate SINUS-24; NCT02912468 [35]3Dupilumab + mometasone furoate27624Placebo + mometasone furoate NCT01920893 [36]2Dupilumab + mometasone furoate6016Placebo + mometasone furoateMepolizumab SYNAPSE; NCT03085797 [37]3Mepolizumab + mometasone furoate40752Placebo + mometasone furoate NCT01362244 [38]2Mepolizumab + fluticasone propionate10725Placebo + fluticasone propionate Gevaert et al. [39]NRMepolizumab3048PlaceboOmalizumab POLYP 1; NCT03280550 [40]3Omalizumab + mometasone13824Placebo + mometasone POLYP 2; NCT03280537 [40]3Omalizumab + mometasone12724Placebo + mometasone NCT01066104 [41]2Omalizumab2726Placebo Gevaert et al. [42]NROmalizumab2416PlaceboESS Dhiman et al. [43]NRESS + fluticasone furoate nasal spray3112Fluticasone furoate nasal spray Sahin-Onder et al. [44]NRESS + fluticasone propionate nasal spray + clarithromycin + nasal douche + oral prednisone4812Fluticasone propionate nasal spray + clarithromycin + nasal douche + oral prednisone Lourjisen et al. [45]NRESS + medical therapyb23852Medical therapybESS endoscopic sinus surgery, NMA network meta-analysis, NR not reported, RCT randomized controlled trial, SC subcutaneous injection, SOC standard of careaANCHOR-1 and ANCHOR-2 were parallel replicate trialsbMedical therapy included, but was not limited to, nasal corticosteroids (spray or drops, or both), nasal rinsing with saline solution, systemic corticosteroids (short courses or long-term tapered treatment), or systemic antibiotics. Biologic therapies were excluded Table 3Overview of RCTs included in NMAsTrial; NCTPhaseTreatment armsRandomized NStudy duration (weeks)Tezepelumab WAYPOINT; NCT048519643Tezepelumab + intranasal corticosteroid40052Placebo + intranasal corticosteroidDepemokimab ANCHOR-1; NCT05274750a [18]3Depemokimab 100 mg SC + medical therapyb27652Placebo + medical therapyb ANCHOR-2; NCT05281523a [18]3Depemokimab 100 mg SC + medical therapyb26452Placebo + medical therapybDupilumab SINUS-52; NCT02898454 [35]3Dupilumab + mometasone furoate44852Placebo + mometasone furoate SINUS-24; NCT02912468 [35]3Dupilumab + mometasone furoate27624Placebo + mometasone furoate NCT01920893 [36]2Dupilumab + mometasone furoate6016Placebo + mometasone furoateMepolizumab SYNAPSE; NCT03085797 [37]3Mepolizumab + mometasone furoate40752Placebo + mometasone furoate NCT01362244 [38]2Mepolizumab + fluticasone propionate10725Placebo + fluticasone propionate Gevaert et al. [39]NRMepolizumab3048PlaceboOmalizumab POLYP 1; NCT03280550 [40]3Omalizumab + mometasone13824Placebo + mometasone POLYP 2; NCT03280537 [40]3Omalizumab + mometasone12724Placebo + mometasone NCT01066104 [41]2Omalizumab2726Placebo Gevaert et al. [42]NROmalizumab2416PlaceboESS Dhiman et al. [43]NRESS + fluticasone furoate nasal spray3112Fluticasone furoate nasal spray Sahin-Onder et al. [44]NRESS + fluticasone propionate nasal spray + clarithromycin + nasal douche + oral prednisone4812Fluticasone propionate nasal spray + clarithromycin + nasal douche + oral prednisone Lourjisen et al. [45]NRESS + medical therapyb23852Medical therapybESS endoscopic sinus surgery, NMA network meta-analysis, NR not reported, RCT randomized controlled trial, SC subcutaneous injection, SOC standard of careaANCHOR-1 and ANCHOR-2 were parallel replicate trialsbMedical therapy included, but was not limited to, nasal corticosteroids (spray or drops, or both), nasal rinsing with saline solution, systemic corticosteroids (short courses or long-term tapered treatment), or systemic antibiotics. Biologic therapies were excluded Overview of RCTs included in NMAs ESS endoscopic sinus surgery, NMA network meta-analysis, NR not reported, RCT randomized controlled trial, SC subcutaneous injection, SOC standard of care aANCHOR-1 and ANCHOR-2 were parallel replicate trials bMedical therapy included, but was not limited to, nasal corticosteroids (spray or drops, or both), nasal rinsing with saline solution, systemic corticosteroids (short courses or long-term tapered treatment), or systemic antibiotics. Biologic therapies were excluded Five trials were identified as suitable for comparison with WAYPOINT at ≥ 30 weeks, based on assessments of heterogeneity in trial design, eligibility criteria, baseline characteristics, and outcome definitions. Four of these trials—SINUS-52, SYNAPSE, ANCHOR-1, and ANCHOR-2—were deemed sufficiently similar to support reliable estimates of comparative efficacy via NMA. Although the network was initially specified to include outcomes at ≥ 30 weeks, completion of the SLR and feasibility assessment demonstrated that all four eligible trials reported endpoints at 52 weeks; accordingly, the network was finalized and is herein referred to as the 52-week network. The fifth trial, Lourjisen 2022, differed significantly in permitted background therapy and was deemed appropriate for inclusion in sensitivity analyses only. NMA Results The primary analyses presented herein are based on endpoints reported at 52 weeks of follow-up. An exception was made for omalizumab, for which no 52-week data were identified in the SLR. NMA results for endpoints assessed at < 30 weeks for all other comparators (i.e. depemokimab, dupilumab, mepolizumab, and ESS) are provided in the Supplementary Material and were broadly consistent with those observed at 52 weeks (Supplementary Tables 1–7). NPS Five relevant studies reporting NPS at week 52 were identified, resulting in a network of 1746 patients (WAYPOINT, SINUS-52, SYNAPSE, ANCHOR-1, and ANCHOR-2). Significant reductions in NPS were observed with tezepelumab relative to mepolizumab, depemokimab, and placebo (Fig. 1, Supplementary Table 8). NPS reductions were similar between tezepelumab and dupilumab.Fig. 1NMA forest plots for change from baseline outcomes, 52-week network. Tools used to evaluate loss of smell were not consistent across trials; however, all scales employed a 0–3 Likert scale which allowed for a comparison of outcomes between treatments. For dupilumab and depemokimab, loss of smell was assessed using a 0–3 scale where 0 indicated no symptoms and 3 indicated severe symptoms. Dupilumab trials utilized a categorical scale while depemokimab trials utilized a verbal response scale. For tezepelumab, loss of smell was assessed using the DSS scale (within the nasal polyposis symptom diary), where 0 indicated no difficulty and 3 indicated severe difficulty. Note the UPSIT results for MEPO were only presented within a subset of patients, though which subset was unclear. CrI credible interval, DEPE depemokimab, DSS difficulty with sense of smell, DUPI dupilumab, LMS Lund Mackay score, MD mean difference, MEPO mepolizumab, NCS/NOS nasal congestion/obstruction score, NMA network meta-analysis, NPS nasal polyp score, SNOT-22 Sino-Nasal Outcome Test 22, UPSIT University of Pennsylvania Smell Identification Test NMA forest plots for change from baseline outcomes, 52-week network. Tools used to evaluate loss of smell were not consistent across trials; however, all scales employed a 0–3 Likert scale which allowed for a comparison of outcomes between treatments. For dupilumab and depemokimab, loss of smell was assessed using a 0–3 scale where 0 indicated no symptoms and 3 indicated severe symptoms. Dupilumab trials utilized a categorical scale while depemokimab trials utilized a verbal response scale. For tezepelumab, loss of smell was assessed using the DSS scale (within the nasal polyposis symptom diary), where 0 indicated no difficulty and 3 indicated severe difficulty. Note the UPSIT results for MEPO were only presented within a subset of patients, though which subset was unclear. CrI credible interval, DEPE depemokimab, DSS difficulty with sense of smell, DUPI dupilumab, LMS Lund Mackay score, MD mean difference, MEPO mepolizumab, NCS/NOS nasal congestion/obstruction score, NMA network meta-analysis, NPS nasal polyp score, SNOT-22 Sino-Nasal Outcome Test 22, UPSIT University of Pennsylvania Smell Identification Test NCS/NOS Two relevant studies reporting NCS/NOS at week 52 were identified, resulting in a network of 856 patients (WAYPOINT, SINUS-52). Significant reductions in NCS/NOS were observed with tezepelumab relative to placebo (Fig. 1, Supplementary Table 9). NCS/NOS reductions were similar between tezepelumab and dupilumab. DSS Four relevant studies reporting DSS at week 52 were identified, resulting in a network of 1335 patients (WAYPOINT, SINUS-52, ANCHOR-1, and ANCHOR-2). Significant reductions in DSS were observed with tezepelumab relative to depemokimab and placebo (Fig. 1, Supplementary Table 10). DSS reductions were similar between tezepelumab and dupilumab. SNOT-22 Five relevant studies reporting the SNOT-22 at week 52 were identified, resulting in a network of 1736 patients (WAYPOINT, SINUS-52, SYNAPSE, ANCHOR-1, and ANCHOR-2). Significantly lower SNOT-22 scores were observed with tezepelumab relative to mepolizumab, depemokimab, and placebo (Fig. 1, Supplementary Table 11). SNOT-22 score reductions were similar between tezepelumab and dupilumab. LMS Four relevant studies reporting LMS at week 52 were identified, resulting in a network of 1334 patients (WAYPOINT, SINUS-52, ANCHOR-1, and ANCHOR-2). Significantly lower LMS were observed with tezepelumab relative to depemokimab and placebo (Fig. 1, Supplementary Table 12). LMS reductions were similar between tezepelumab and dupilumab. UPSIT Three relevant studies reporting UPSIT at week 52 were identified, resulting in a network of 1004 patients (WAYPOINT, SINUS52, SYNAPSE). Significantly higher UPSIT scores were observed with tezepelumab relative to mepolizumab and placebo (Fig. 1, Supplementary Table 13). UPSIT scores were similar between tezepelumab and dupilumab. Time to NP Surgery Decision Five relevant studies reporting the time to NP surgery decision up to 52 weeks of follow-up were identified, resulting in a network of 1347 patients (WAYPOINT, Pooled SINUS [SINUS-52 and SINUS-24], SYNAPSE, ANCHOR-1, and ANCHOR-2). Treatment with tezepelumab prolonged the time to NP surgery decision compared to mepolizumab and placebo (Fig. 2, Supplementary Table 14). The time to surgery decision was similar between tezepelumab and dupilumab.Fig. 2NMA forest plots for surgical and SCS outcomes, 52-week network. CrI credible interval, DEPE depemokimab, DUPI dupilumab, HR hazard ratio, MEPO mepolizumab, NMA network meta-analysis, NP nasal polyp, OR odds ratio, SCS systemic corticosteroids NMA forest plots for surgical and SCS outcomes, 52-week network. CrI credible interval, DEPE depemokimab, DUPI dupilumab, HR hazard ratio, MEPO mepolizumab, NMA network meta-analysis, NP nasal polyp, OR odds ratio, SCS systemic corticosteroids Proportion of Patients Requiring NP Surgery Three relevant studies reporting the proportion of patients requiring NP surgery (actual or planned), as Kaplan–Meier probability estimates, up to 52 weeks of follow-up were identified, resulting in a network of 1263 patients (WAYPOINT, SYNAPSE, SINUS-52). Patients treated with tezepelumab were less likely to require NP surgery than patients treated with dupilumab, mepolizumab, or placebo (Fig. 2, Supplementary Table 15). Tezepelumab was ranked as the most effective treatment with respect to the need for surgical intervention in CRSwNP. Time to SCS Use Three relevant studies reporting the time to SCS use up to 52 weeks of follow-up were identified, resulting in a network of 1266 patients (WAYPOINT, SYNAPSE, pooled SINUS [SINUS-52 and SINUS-24]). Treatment with tezepelumab significantly prolonged the time to SCS use compared to mepolizumab and placebo (Fig. 2, Supplementary Table 16). The time to SCS use was similar between tezepelumab and dupilumab. Proportion of Patients Requiring SCS Use Five relevant studies reporting the proportion of patients requiring SCS up to 52 weeks of follow-up were identified, resulting in a network of 1791 patients (WAYPOINT, SINUS-52, SYNAPSE, ANCHOR-1, and ANCHOR 2). Patients treated with tezepelumab were significantly less likely to require SCS than patients treated with mepolizumab, depemokimab, or placebo (Fig. 2, Supplementary Table 17). The proportion of patients requiring SCS use was similar between tezepelumab and dupilumab. NMAs Including Omalizumab (< 30 Weeks Follow-up) Endpoint reporting in the omalizumab trials (POLYP 1 and POLYP 2) was limited to follow-up durations < 30 weeks. To facilitate the inclusion of omalizumab in the ITC, networks for endpoints reported at < 30 weeks were constructed. Omalizumab trials did not include all endpoints considered for this ITC; only data for NPS, NCS/NOS, DSS, SNOT-22, and UPSIT was available. Patients treated with tezepelumab were shown to have significant reductions in NPS, NCS/NOS, and DSS, and significant increases in UPSIT compared to patients treated with omalizumab. SNOT-22 scores were similar between tezepelumab and omalizumab. The full analyses, including the comparative efficacy of omalizumab versus other biologics, are presented for all < 30 weeks follow-up networks in the Supplementary Material (Supplementary Tables 18 and 19). Discussion In this NMA, tezepelumab demonstrated statistically significant improvements across all endpoints assessed at 52 weeks compared to depemokimab, mepolizumab, and placebo. These outcomes reflect improvements in the symptom burden of CRSwNP and reductions in nasal polyp size, which directly impact nasal airflow, olfaction, sense of smell, and quality of life. Furthermore, reducing reliance on SCS and surgery helps minimize treatment-related risks and may benefit patients who have not responded adequately to these interventions, potentially reducing the need for therapies that offer limited effectiveness. Tezepelumab showed comparable efficacy to dupilumab across all outcomes, with the exception of the proportion of patients requiring surgery, where tezepelumab achieved a lower proportion than dupilumab (95% CrI 0.003–0.546). Overall, tezepelumab ranked highest in the 52-week network for change from baseline in SNOT-22 scores, proportion requiring SCS use, proportion requiring NP surgery, and time to NP surgery decision, suggesting a favourable profile for patients with severe disease or those at high risk of surgical intervention. Analyses incorporating omalizumab, which were necessarily limited to < 30-week follow-up, showed that tezepelumab was associated with significant improvements in NCS/NOS and NPS compared with omalizumab. A recent meta-analysis by Xu et al. ranked dupilumab as the most effective treatment for CRSwNP based on improvements in NPS, SNOT-22, and NCS/NOS, followed by tezepelumab [29]. The discrepancy between the present ITC and Xu et al. likely reflects differences in population, intervention, comparison, and outcomes (PICOS) criteria and evidence networks. Notably, Xu et al. included subgroup data from the LIBERTY QUEST [32] and LIBERTY VENTURE [33] phase III trials, which evaluated dupilumab in patients with moderate-to-severe asthma, a subset of whom had comorbid CRSwNP of any severity. Subgroup results from asthma trials were excluded from the present analysis, as the asthma trial patient population was not recruited based on CRSwNP specific attributes, was not required to have severe uncontrolled CRSwNP, and CRSwNP baseline characteristics were not adequately reported. Inclusion of CRSwNP subgroups from asthma trials may introduce bias by disproportionately increasing the representation of patients with comorbid asthma in the analysis population. The robustness of these findings is supported by key methodological strengths, including the conduct of analyses according to best practices for conducting and reporting NMAs as described by NICE TSDs to ensure transparency and reproducibility [31]. Additionally, the NMAs were informed by a comprehensive SLR designed to capture all relevant evidence. All trials identified through the SLR underwent a rigorous feasibility assessment to evaluate similarity in design, population, and outcomes, minimizing heterogeneity and enhancing the validity of the network. The sensitivity analyses, performed for both SNOT-22 (Supplementary Table 20) and NPS (Supplementary Table 21) at 52 weeks follow-up and < 30 weeks follow-up, were consistent with the results of the primary analysis. Finally, in contrast to methodologies that require adjustment or reweighting of patient-level data, the NMAs synthesized evidence directly from aggregate trial-level estimates without applying any additional statistical modification. Several limitations should be considered when interpreting these findings, which are exploratory in nature. Residual heterogeneity between trials could not be fully eliminated, even after feasibility screening. NMAs do not incorporate population adjustment; therefore, prognostic factors and effect modifiers, including pre-treatment eosinophil count and Asian ancestry, were not adjusted for. In addition, some trials lacked sufficient detail to fully assess comparability; these were assumed similar and included in the network, introducing potential uncertainty. However, discrepancies in study methodology may confound comparisons; specifically, differences in NPS scoring methods between the dupilumab and tezepelumab trials could have contributed to greater apparent NPS improvement with dupilumab [34, 46]. Most treatment comparisons were also informed by a single trial, necessitating the use of fixed-effect models. Fixed-effect models were the most appropriate given the sparse networks of evidence for some outcomes; however, they may limit the assessment of variability across studies. Challenges were also encountered for the time to NP surgery decision endpoint. Low event rates resulted in extreme effect estimates with wide credible intervals, reducing precision in these findings. No specific statistical approach or correction method was employed for studies where the number of events was zero in one or more treatment arms within our NMA. Explicit corrections, such as adding a continuity constant, were avoided to prevent introducing bias. Instead, robust Bayesian hierarchical models operating on the log scale were employed, as these methods inherently and reliably borrow strength from the overall network to inform and stabilize estimates involving sparse data. Conclusions Overall, this NMA suggests tezepelumab offers a compelling efficacy profile in the treatment of CRSwNP. Tezepelumab demonstrated comparable efficacy to dupilumab across all evaluated outcomes with the exception of the proportion of patients requiring NP surgery, where tezepelumab achieved a lower surgery rate than dupilumab. In addition, tezepelumab showed significant benefit over mepolizumab for surgical outcomes and significantly prolonged the time to NP surgery decision relative to depemokimab. The marked reductions in the proportion of patients requiring NP surgery relative to dupilumab and mepolizumab were driven in part by the observation of a single surgical event in the tezepelumab arm of the WAYPOINT trial. These findings support tezepelumab as a valuable therapeutic option for CRSwNP, with the potential to reduce long-term surgical and SCS burden. Ultimately, there remains a need for future head-to-head comparisons to confirm these findings. Supplementary Information Below is the link to the electronic supplementary material.Supplementary file1 (PDF 431 kb) Supplementary file1 (PDF 431 kb)