Meta-analyses of the efficacy of pharmacotherapies and sublingual allergy immunotherapy tablets for allergic rhinitis in adults and children

Background: Treatment options for seasonal and perennial allergic rhinitis (SAR/PAR) include pharmacotherapies and allergy immunotherapy. These meta-analyses evaluated the efficacy of pharmacotherapies and sublingual immunotherapy tablets (SLIT-tablets) versus placebo on nasal symptoms associated with SAR and PAR. Methods: Randomized, double-blind, placebo-controlled trials were identified from systematic PubMED/EMBASE searches through 7/18/2019 (PROSPERO protocol CRD42018105632). The primary outcome was mean numerical difference in total nasal symptom score (TNSS; 0-12) between active treatment and placebo at the end of the assessment period. Random-effects meta-analyses estimated the mean difference for each medication group weighted by the inverse of the trial variance. Publication bias assessments and sensitivity analyses were conducted. Results: Rescue symptom-relieving pharmacotherapy was prohibited in most pharmacotherapy trials but was allowed in all SLIT-tablet trials. For adult/adolescent SAR, the mean numerical difference (95% CI) in TNSS versus placebo was: intranasal corticosteroids (INCS)=1.38 (1.18, 1.58; 39 trials); combination intranasal antihistamine/INCS=1.34 (1.15, 1.54; 4 trials); intranasal antihistamines=0.72 (0.56, 0.89; 13 trials); oral antihistamine=0.62 (0.35, 0.90; 18 trials); SLIT-tablets=0.57 (0.41, 0.73; 4 trials); and montelukast=0.48 (0.36, 0.60; 10 trials). For adult/adolescent PAR, mean difference in TNSS versus placebo (95% CI) was: INCS=0.82 (0.66, 0.97; 14 trials); SLIT-tablets=0.65 (0.42, 0.88; 3 trials); and oral antihistamine=0.27 (0.11, 0.42; 3 trials). The number of eligible trials limited meta-analyses for pediatric SAR/PAR. Conclusions: All treatments significantly improved nasal symptoms versus placebo. SLIT-tablets provided improvement in TNSS despite access to rescue symptom-relieving pharmacotherapy. Extensive trial heterogeneity and strong indications of publication bias preclude the comparison of treatment effects among treatment classes.