P-values, 95% CIs, and Mean Differences were calculated using an analysis of covariance model (ANCOVA) model adjusted for pre-treatment baseline total nasal symptom score (TNSS, scale 0 [best] to 12 [worst]) area under the curve (AUC) at the 0.05 level of significance. Mean differences were calculated as follows: Sublingual immunotherapy– Placebo; Subcutaneous immunotherapy – Placebo; Sublingual immunotherapy – Subcutaneous immunotherapy. The weighted 10 hour AUC was calculated as the (early phase response (0–1 hour)/1) + (late phase response (1–10 hours)/9). Participants in the ITT population with missing primary endpoint data had their data imputed. Imputation was performed within treatment group using participants who had available TNSS AUC values. Specifically, a linear regression line and 95% confidence bands were fit where Year 3 values were regressed on TNSS AUC values at time t (t = Baseline, Year 1 or Year 2). Within each treatment arm, a missing Year 3 TNSS AUC value was imputed as the value predicted from the linear regression line. The primary endpoint was also calculated in the modified ITT population.