Subjects aged 4-13 years were recruited from outpatients that visited the Hainan General Hospital from August 2019 to April 2021. Eligible participants were randomly assigned to the morning dosing (MD) group and the evening dosing (ED) group by a computer-generated randomization method and stratified by sensitization status (monosensitized vs. polysensitized). All participants received daily SLIT between 7:00 a. m. and 9:00 a.m. (for the MD group) or between 8:00 p. m. and 10:00 p.m. (for the ED group) with glycerinated Dermatophagoides farinae drops. The standardized drops (Chanllergen; Zhejiang Wolwo Bio-Pharmaceutical Co., Ltd., Zhejiang, China) were labeled from 1 to 4 with the total protein concentration of 1, 10, 100, and 333 μg/mL, respectively. The drops were self-administered sublingually for 1-3 min before swallowing. Nothing was allowed by mouth for 15 min after the administration of the drops.

All patients were required to take daily doses in strict accordance with the manufacturer’s instructions. The young children were administered the SLIT extracts under the supervision of their guardians. In the first 3 weeks, participants were instructed to take drops No. 1, drops No. 2 and drops No. 3, respectively, as increment phase, in a gradually increasing order of 0.05, 0.10, 0.15, 0.20, 0.30, 0.40, 0.50 mL/day. From week 4, patients were treated with 0.15 mL of drops No. 4/day until the treatment completed (15, 16).

Investigators collected participants’ demographic characteristics and clinical data before initiation of treatment. Telephone follow-ups were provided to patients monthly to supervise medication and assess adverse events (AEs). Patients returned to hospital at 0.5 and 1 year for evaluation, including clinical efficacy, lung function and the FeNO level. The efficacy and safety analysis only involved data of patients who completed the study, while the data of patients who did not complete the study were excluded. The present study was approved by the Ethics Committee of Hainan General Hospital and conducted in compliance with the Ethical Guidelines for Clinical Studies and Good Clinical Practice. All patients and their guardians were informed of the relevant information prior to their participation in the study (17).

Patients with AS in remission due to house dust mite (HDM) were enrolled into the study. Inclusion criteria were as follows: all patients were diagnosed with mild-to-moderate AS according to Global Initiative for Asthma (GINA) (18); patients were sensitized to D. farinae and/or D. pteronyssinus. Mild asthma is defined as asthma that is well controlled with as-needed ICS-formoterol alone. or with low-intensity maintenance controller treatment such as low dose ICS, leukotriene receptor antagonists or chromones. Moderate asthma is defined as asthma that is well controlled with low or medium dose ICS-LABA. Patients only sensitized to D. farinae and/or D. pteronyssinus were allocated to the monosensitized group, while patients who were sensitive to D. farinae and/or D. pteronyssinus and other coexisting inhaled allergens were allocated to the polysensitized group. Sensitization to D. farinae, D. pteronyssinus and other aeroallergens were further confirmed by the presence of specific immunoglobulin E (sIgE) ≥ 0.7 KU/L, using UniCAP system (Phadia, Uppsala, Sweden). Exclusion criteria were as follows: forced expiratory volume in one second (FEV₁) < 70% of predictive value; severe systemic diseases such as poorly controlled cardiovascular diseases, immune diseases, or malignant tumors; receiving β-blockers or angiotensin-converting enzyme inhibitors; serious psychological barriers or failed to understand the risks and limitations of treatment.

Participants or their guardians were required to record symptom and medication consumption 10 min before the SLIT administration daily throughout the study period. The investigators calculated the weekly average scores at every visit. The total asthma symptom score (TASS) was the sum of daytime asthma symptom scores and nighttime asthma symptom scores (1). The daytime asthma symptoms were scored based on a scale of 0-5 points, in accordance with the general severity of asthma symptoms (wheeze, shortness of breath, dyspnea, and cough) and its impact on daily life. 0, no symptoms; 1, symptoms are rare and short lasting; 2, two or more short lasting symptoms; 3, mild symptoms for more of the day, but had little impact on life and work; 4, severe symptoms for more of the day and affect life and work; 5, the symptom is so serious that the subject cannot work and live normally. The nighttime asthma symptoms were scored based on a scale of 0-4 points, in accordance with the frequency of nocturnal and early morning awakening induced by asthma. 0, no symptoms; 1, wake up once or wake up early; 2, wake up twice, including wake up early; 3, wake up many times (≥ 3 times); 4, can’ t fall asleep at night. Patients were prescribed pharmacologic therapy for symptom remission and required to use according to physicians’ instruction. Rescue medication score was calculated as follows (per day): 0 = no use of relief medication; 1 = use of oral antihistamines, anti-leukotrienes, or β2 receptor agonists; 2 = use of inhaled corticosteroids; 3 = use of combination therapy (corticosteroids with β2 receptor agonists). The total asthma medicine score (TAMS) was the sum of all the recorded medicine scores (19). The Asthma Control Questionnaire (ACQ) has strong evaluative and discriminative properties and can be used with confidence to measure asthma control. The ACQ includes five questions about the most important symptoms for asthma control assessment, one question about β2 agonist use and another about FEV₁. The items are equally weighted and the ACQ score is the mean of the 7 items and therefore between 0 (well controlled) and 6 (extremely poorly controlled) (20).

The pulmonary function of the children was determined by the Master Screen lung function instrument (Jaeger GmbH, Cologne, Germany) at every visit. The measurement parameters included FEV₁ and FEV₁/forced volume vital capacity (FVC) (21).

The fractional exhaled nitric oxide (FeNO) was evaluated at every visit by using the exhaled nitric oxide tester (SUNVOU, Wuxi, China). FeNO level was detected in accordance with the FeNO standardized monitoring guidelines recommended by the American Thoracic Society (ATS) and the European Respiratory Society (ERS) (22). The children seated comfortably and breathed quietly for approximately 5 min, then inhaled the gas contain low NO concentration (< 5 ppb) to near total lung capacity (TLC) and immediately exhaled at a constant flow rate of 50 ml/second, until an NO plateau of at least 2 s could be identified during an exhalation of at least 4 seconds.

The occurrence rate, duration, and severity (23) were recorded to assess AEs. All AEs were addressed under the instruction of the physicians.

Statistical analysis was performed in demographic analysis and efficacy assessment with IBM SPSS Statistics 21.0. The differences between the MD group and the ED group at baseline demographic characteristics were analyzed by χ² test. The intragroup comparisons of clinical characteristics were performed by the Friedman M test. The intergroup comparisons of clinical characteristics were performed by the Mann-Whitney U test. The 2-tailed level of statistical significance was set at p = 0.05.

A total of 163 patients [mean age 8.39 ± 2.69 years, 24.54% female (n = 40), 75.46% male (n = 123)] fulfilling the eligibility requirements were randomized and divided into two groups (Figure 1). The MD group and the ED group included 82 patients and 81 patients, respectively. After 1 year, 62 patients in the MD group (75.61%) and 63 patients in the ED group (77.77%) completed the entire study. Reasons for discontinuation were as follows: lost to follow-up (n = 16), protocol non-compliance (n = 6), improvement of symptoms (n = 11) and other reasons (n = 5). Both groups were well balanced in respect of age, gender and sensitization status (Table 1, all p > 0.05). Similarly, there was no statistical difference in all items of clinical data between two groups (all p > 0.05).

The TASS of patients in both groups decreased significantly during the treatment in comparison of baseline (Figure 2A; MD group, 3.03 ± 0.79 at baseline vs. 0.31 ± 0.53 in 0.5 year vs. 0.19 ± 0.40 in 1 year; ED group, 3.19 ± 0.91 at baseline vs. 0.38 ± 0.55 in 0.5 year vs. 0.27 ± 0.51 in 1 year; all p < 0.001compared to baseline). The significant reduction of TAMS was consistent with the TASS results (Figure 2B; MD group, 5.03 ± 0.18 at baseline vs. 2.79 ± 0.41 in 0.5 year vs. 0.00 ± 0.00 in 1 year; ED group, 5.00 ± 0.00 at baseline vs. 2.84 ± 0.37 in 0.5 year vs. 0.00 ± 0.00 in 1 year; all p < 0.001 compared to baseline). No significant difference between MD and ED group was shown in those clinical scores (p > 0.05).

The score of ACQ declined from baseline to the end of treatment in both groups (Figure 2C; MD group, 1.66 ± 0.33 at baseline vs. 0.61 ± 0.24 in 0.5 year vs. 0.29 ± 0.34 in 1 year; ED group, 1.53 ± 0.44 at baseline vs. 0.33 ± 0.42 in 0.5 year vs. 0.28 ± 0.44 in 1 year; all p < 0.001 compared to baseline). Notably, significantly greater improvement in ACQ was manifested in the ED group compared with the MD group at 0.5 year (p < 0.001). This difference disappeared by the end of the study.

There were significant increases in FEV₁ (Figure 3A; MD group, 86.03 ± 10.05 at baseline vs. 100.89 ± 8.27 in 0.5 year vs. 98.92 ± 9.66 in 1 year; ED group, 83.63 ± 7.99 at baseline vs. 101.17 ± 10.29 in 0.5 year vs. 100.70 ± 9.57 in 1 year; all p < 0.001 compared to baseline) and FEV₁/EVC (Figure 3B; MD group, 91.16 ± 9.07 at baseline vs. 105.60 ± 6.73 in 0.5 year vs. 106.19 ± 6.99 in 1 year; ED group, 91.11 ± 7.29 at baseline vs. 106.16 ± 6.19 in 0.5 year vs. 105.68 ± 9.44 in 1 year; all p < 0.001 compared to baseline) during the whole treatment in both groups. Comparison in respect of pulmonary function revealed no significant difference between two groups throughout the treatment (p > 0.05).

In the MD group, there was significant decrease of FeNO level during the whole therapy (Figure 4; 29.47 ± 14.01 at baseline vs. 12.84 ± 8.48 in 0.5 year vs. 11.42 ± 7.12 in 1 year, all p < 0.001 compared to baseline). The level of FeNO also significantly ameliorated in the ED group (26.08 ± 14.04 at baseline vs. 13.10 ± 8.98 in 0.5 year vs. 9.75 ± 6.78 in 1 year, all p < 0.001 compared to baseline), whereas a significant better improvement than the MD group occurred at 1 year (p < 0.05).

No deaths, anaphylactic shocks or life-threatening events were reported during the entire study (Table 2). Seven AEs that occurred in the MD group were 3 oral numbness or itching, 3 local rash, and 1 asthma attack. Three AEs that occurred in the ED group were 1 oral numbness or itching, and 2 fatigue. The AE rate was 11.29 and 4.76% in the MD group and the ED group, respectively, with no statistically significant difference between the two groups (p > 0.05). All AEs occurred in the first month of SLIT, and mitigated with or without medication treatment within a week.