Sodium Chloride versus Lactose as a Carrier for House Dust Mite Allergen in Allergen Chamber Studies: A Clinical Study to Assess Noninferiority

Introduction: In the Fraunhofer allergen challenge chamber (ACC), a standardized, universal, good manufacturing practice-conforming technology using a spray dried solution of lactose (L) and allergen extract has been established. In this study, we investigated the noninferiority of hypertonic sodium chloride (S) versus L as a carrier for house dust mite (HDM) allergen to simplify manufacturing, reduce costs, and allow for wider use. Methods: Using a participant-blinded, sham exposure-controlled, single-arm, sequential intervention study, we challenged adults with HDM allergic rhinitis five times in the ACC. Participants were first exposed to S, L, and clean air (block 1), followed by S + HDM and L + HDM (block 2). Primary endpoints were mean total nasal symptom score (TNSS) and mean nasal secretion weight. Results: 19 participants were enrolled in the study (10 females; mean age 32 years [22–49], 4 with mild allergic asthma). The safety profile of S + HDM and L + HDM was similar; eight participants experienced mild procedure-related adverse events including tiredness, cough, and dyspnea. Due to dropouts, 13 participants completed the study and were evaluated. Mean TNSS and nasal secretion were as follows: S 0.98, 0.28 g; L 1.1, 0.20 g; clean air 1.1, 0.23 g; S + HDM 5.7, 4.8 g; L + HDM 5.1, 5.1 g. Separate block 1/block 2 MANOVAs with TNSS and nasal secretion as dependent variables revealed no significant differences between the carriers, neither alone and compared with clean air (p = 0.2059, Wilk’s λ = 0.78) nor combined with HDM (p = 0.3474, Wilk’s λ = 0.89). Noninferiority of S was established using a meta-analysis-based minimal clinical important difference of −0.55: mean TNSS difference between S + HDM and L + HDM was 0.62 (90% CI: −0.51 to 1.74). Conclusion: S as an HDM carrier was safe and well tolerated. It was noninferior to L which makes it an adequate and easy-to-use carrier substitute.


Introduction
Standardized exposures of patients with allergic rhinitis to qualified allergen atmospheres in chamber settings are often used for efficacy testing of allergen immunotherapy.Aerosolization of certain aeroallergens such as house dust mite (HDM) requires standardization of raw material since natural sources are variable.Worldwide, 8 out of 11 allergen exposure chambers have validated HDM challenges [1].Allergen material and particle size distributions differ, e.g., the Vienna Challenge Chamber (Austria) reported use of large particles >25 μm from dried raw material [2].Kingston Environmental Exposure Unit (Canada), GA 2- LEN chamber (Germany), Biogenics Research Chamber (USA), and ALL-MED Allergen Exposure Chamber (Poland) also used dried raw material, though not selected by particle size [3][4][5][6].In contrast, AARHUS (Denmark) exposure chamber described to have collected dust from office buildings or homes that was subsequently homogenized, with particle sizes <5 μm, dominating during exposure [7,8].ALYATEC (France) used lyophilized Dermatophagoides extracts diluted in saline [9].Allergen chambers work with different dispersion techniques such as computer-controlled feeders, powder dispensers, or nebulizers [1].At Fraunhofer Institute for Toxicology and Experimental Medicine (ITEM), an allergen chamber challenge (ACC) with HDM was validated in 2016 [10].For the challenge atmosphere, HDM particles were produced by spray drying an aqueous solution of commercially available allergen extracts and good manufacturing practice-grade lactose (L) as a carrier molecule.This patented technique (DE 10 2008 060 614) allows to produce aerosols with adjustable particle size, depending on the solute concentration (mass [m]/volume [v]) to reach the respective target organ (nose: >10 μm, lung: <5-10 μm) [10].L is also a well-established carrier for drug particles of dry powder inhalers due to its safe toxicity profile and stability [11,12].For use in the Fraunhofer ACC, good manufacturing practice-grade L has been aliquoted, dissolved in normal saline solution, vialed, sterilized, and labeled with respective manufacturing authorization.After release by a qualified person, it was used in clinical studies as 10% (m/v) solute for allergen extracts with reconstitution under the responsibility of the investigator.But since this L-based spray drying technique is expensive and complexcompared to effort and cost of pollen challengesalternatives were evaluated.Apart from widespread use of authorized normal saline (0.9% sodium chloride [S]), hypertonic saline (e.g., 3-7% S) is licensed for inhaled use to treat mucus retention in patients with airways disease [13][14][15].Using hypertonic S as a carrier instead of L for nasal allergen deposition in challenge chamber studies simplifies manufacturing and reduces costs.Our first study using S as a HDM carrier was done in patients with atopic dermatitis and HDM sensitization (NCT05019209) to study skin responses and T-cell immunology (unpublished data).However, until now, the noninferiority of S to L as a carrier has not been in-vestigated.To document potential differences of S and L, we challenged a group of allergic patients sequentially with S and L carriers alone, clean air, and carriers combined with HDM.

Study Design
This was a participant-blinded, sham exposure-controlled, single-arm, sequential intervention study in adult patients with HDM-induced allergic rhinitis (NCT05245175).The study was performed at Fraunhofer Institute for Toxicology and Experimental Medicine ITEM in Hannover, Germany.The objectives were to document the effect of L and S carriers on patient-reported nasal symptoms and production of nasal secretion.

Ethics
The study was approved by the local Ethics Committee of Hannover Medical School and was conducted in compliance with the Declaration of Helsinki (7th revision, Fortaleza, Brazil), International Council for Harmonisation Guideline for Good Clinical Practice (E6, Step 4, November 2016), and all national laws and regulations.All participants gave written informed consent before being included in the study.

Participants
Adults 18-65 years of age were eligible if they had a history of HDM-induced allergic rhinitis for 1 year or longer, a positive skin prick test (diameter ≥3 mm) to Dermatophagoides pteronyssinus (D. pteronyssinus), and a D. pteronyssinus-specific immunoglobulin IgE level ≥0.7 kU/L.Participants were excluded if they were taking anti-allergic or immunosuppressive drugs including corticosteroids; had asthma requiring step 2 treatment or higher according to the Global Initiative for Asthma guidelines (daily use of inhaled steroids with or without long-acting beta agonists) [16]; had a FEV 1 <80% of the predicted normal value; had concomitant allergies to seasonal aeroallergens which were anticipated to be active during study participation or to animal dander with regular exposure; or had received specific immunotherapy for dust mite allergy within 2 years before participation.Women could not be pregnant or lactating.

Study Conduct
Participants were challenged in the ACC five times sequentially with the two HDM allergen exposures 8 days apart.The five challenges were parted into two blocks; within each block, the challenge sequence was randomized with the following resulting order: • Block 1: S, L, clean air • Block 2: S + HDM, L + HDM

Study Endpoints
Nasal symptoms were evaluated as a patient-reported outcome prior to and every 20 min during allergen challenge using the Total Nasal Symptom Score (TNSS).The TNSS is an ordinal scale with 4 items: nasal congestion, rhinorrhea, nasal itching, and sneezing.Each item is scored according to symptom severity with 0 = none, 1 = mild, 2 = moderate, and 3 = severe.The item scores are summed up to calculate the TNSS of each assessment time (every 20 min).The TNSS ranges between 0 and 12 points.Eventually, this resulted in a prechallenge score and 12 within-challenge scores for each participant.The 12 within-challenge scores were averaged for each participant (individual TNSS mean) and for the whole group (group TNSS mean).
Nasal secretion was assessed using pre-weighed handkerchief sachets.Each participant had to use one sachet per hour to collect nasal discharge.After the challenge, the handkerchief sachets were weighed again.The weight difference resembled the amount of nasal secretion produced during the challenge.

Technical Methods
The spray drying technique was described by Lueer et al. [10].The exposure atmosphere was generated in the validated Fraunhofer ACC according to site-specific standard operation procedures.Airflow through the chamber was 600 m³/h.The chamber was operated with HEPA-filtered purified outdoor air, temperature, and humidity control, and then supplemented with a qualitatively and quantitatively determined particle/ allergen load [1].For the HDM exposures, lyophilized allergen extract of D. pteronyssinus 100 HEP/mL (LETI Pharma, Ismaning, Germany) was dissolved in either 10% L solution (Fraunhofer ITEM, Hannover, Germany) made from Pharmatose ® (DMV-Fonterra Excipients GmbH & Co. KG, Goch, Germany) and 0.9% saline (B.Braun Melsungen AG, Melsungen, Germany), or 5% saline (diluted from 10% saline, B. Braun Meslungen AG, Melsungen Germany).The spray solution contained either 5% saline, 10% L, 5% saline + D. pteronyssinus 100 HEP/mL (S + HDM), or 10% L + D. pteronyssinus 100 HEP/mL (L + HDM).For each exposure session, about 50 mL solution was needed.The spray dried aerosol containing either S, L, S + HDM, or L + HDM was added to the chamber air.The solute concentration determines the size of the particles after spray drying.The resulting challenge aerosol was characterized by a mass distribution with less than 30% of the aerosol mass in the size range <10 μm.This assured that primarily the nasal region was challenged.

Sample Size
The sample size was based on feasibility due to the maximum ACC capacity of 18 individuals.The required standard deviation of the group mean difference was calculated from unpublished data of two HDM challenge studies with paired samples exposed to the same allergen concentration.With a fixed sample size of 18, the pooled standard deviation 1.26 of mean group difference, a type 1 error probability of 2.5%, a true mean difference of 0.7 (derived from the above-mentioned unpublished data), and a noninferiority margin of −0.55, a power of 80% could be achieved [18].The power was calculated with SampSize [19].

Statistics
Statistical analysis was performed using R core version 4.2.1 and GraphPad Prism 9.3.1.Missing and incomplete data were not replaced, and no imputation was performed.Safety was assessed in all exposed participants.The confirmative analysis was based on the per protocol population, who completed all five exposures.One-way repeated measures univariate and multivariate analyses of variance (ANOVA/MANOVA) was performed to assess the effect of the challenge agent carrier on TNSS and nasal secretion.MANOVA was performed additionally because it has more power to detect group differences than separate ANOVAs.The ANOVA and MANOVA were performed separately for block 1 (S, L, clean air) and block 2 (S + HDM, L + HDM).In block 2, univariate analysis was performed with a paired t test since it compared just two groups.Exposures to S, L, clean air, S + HDM and L + HDM were evaluated by testing the following hypotheses at a two-sided 5% significance level: • H0: There is no difference between challenge agent carriers.
• H1: There is a difference between challenge agent carriers.
Assumptions for ANOVA/MANOVA were tested, including sphericity, univariate/multivariate outliers, normality, linearity, independence of observations, and equality of variance-covariance matrices.Post hoc analysis was done with Tukey's multiple comparisons test.Noninferiority of S as a carrier compared with L was determined by testing if the lower limit of the 90% confidence interval (CI) for the mean TNSS difference between S + HDM and L + HDM was above the noninferiority margin δ = −0.55[20,21].Descriptive statistics are provided for TNSS and nasal secretion (n, arithmetic mean, 95% CIs).

Participants and Demographics
19 participants were enrolled in the study between February 22 and April 13, 2022 (Fig. 1).Of those, approximately 10 were female, average age was 32 years (Table 1).Eventually, 13 participants completed the study per protocol.

Safety
Adverse events (AEs) were experienced by 17 participants, eight of them had mild procedure-related symptoms such as tiredness, cough, and dyspnea, including two participants who had a late phase asthmatic reaction after L + HDM exposure with peak flow reduction of >20% that was resolved immediately after 200 μg salbutamol inhalation.One of them also experienced an early phase reaction, 1 h into the S + HDM exposure that responded well to 200 μg salbutamol inhalation.One participant with procedurerelated AEs had a dry cough during peak flow measurements regardless of exposure atmosphere.The other procedure-related AEs were expected allergic side effects of HDM exposure.All other adverse events Sodium Chloride as a Carrier for House Dust Mite Allergen Extract were not procedure-related and mostly mild, including headache, stomachache, and symptoms of common cold.Four participants were withdrawn due to adverse events that were not procedure-related (Fig. 1).Overall, both S and L exposures with and without HDM had similar, favorable safety profiles.

Sample Size
The sample size decreased to 13 due to dropouts during the study (Fig. 1), including six withdrawals because of infection and one at participant's own request.
Block 1: S, L, and Clean Air Technical results are depicted in Figure 2. In block 1, group means of TNSS (Fig. 3a) and nasal secretion weight (Fig. 3b) lay close together (Table 2): mean TNSS ranged around 1 and mean nasal secretion weight around 0.24 g.Mean peak TNSS was also similar in all three groups, ranging around 1.8.These findings suggest a mild nocebo effect and that the carriers did not have a greater effect on nasal symptoms than clean air.
Univariate analysis with one-way repeated measures ANOVA indicated that the carriers alone likely had no Similarly, carriers alone likely had no influence on nasal secretion either, F(1.3, 15.6) = 1.1, p = 0.3381.Individual comparisons did also not reveal significant differences between interventions.
Combing TNSS and nasal secretion in a one-way repeated measures MANOVA showed that the slightly different effects of S, L, and clean air likely indicated no true differences between the carriers and clean air, F(4, 46) = 0.15, p = 0.2059, Wilk's λ = 0.78.Wilk's λ signified that 78% of the variance in the dependent variables was unaccounted for by the independent variable, i.e., the carrier agents.This suggests that the carrier agents did not influence the difference in TNSS and nasal secretion means.
Block 2: S + HDM and L + HDM.In block 2, mean TNSS and nasal secretion increased significantly compared to block 1 exposures (Table 2).Between S + HDM and L + HDM, mean difference in TNSS was 0.6 and in nasal secretion weight 0.3 g.In the S + HDM group, symptoms plateaued at 100-120 min and achieved a mean peak TNSS of 7.2.In the L + HDM group, symptoms plateaued at 100 min and achieved a mean peak TNSS of 6.9.
Univariate analysis with a paired t test indicated that the difference in carriers combined with HDM likely had no influence on TNSS, p = 0.3474.Similarly, it likely had no influence on nasal secretion either, p = 0.7829.Also, combing TNSS and nasal secretion in a one-way repeated measures MANOVA showed that the differences between S + HDM and L + HDM likely indicated no true influence of the carriers, F(2, 11) = 0.71, p = 0.5149, Wilk's λ = 0.89.Wilk's λ signified that 89% of the variance in the dependent variables was unaccounted for by the independent variable, i.e., the carrier agents.This suggests that the carrier agents did also not influence the difference in TNSS and nasal secretion means when combined with HDM allergen.
Analysis of TNSS subscales revealed that nasal congestion and rhinorrhea were the highest rated TNSS items in the S + HDM and L + HDM group (Fig. 5a, red and yellow bars).However, the largest mean difference between S + HDM and L + HDM was in the item sneezing (Fig. 5b, green bar).This difference was driven by the same participants who experienced a much higher TNSS with S + HDM as described in the previous paragraph.Both were the only individuals who had a mean TNSS difference >1 unit on the sneezing item.

Noninferiority of S
In univariate and multivariate analyses, the null hypothesis (no difference between the carriers) could not be rejected.However, this does not mean equivalence or proving the null hypothesis as described elsewhere [22,23].To investigate anti-allergic treatments, allergen chamber challenges usually need to induce an adequate level of allergic symptoms.Thus, the noninferiority of S to L as a carrier for HDM was tested.It was irrelevant whether S would be equivalent or better than L but rather that it would not perform worse than L.
The noninferiority margin δ was set to −0.55 based on the following considerations: The minimal clinically important difference (MCID) in TNSS has been derived by Barnes et al. [24] in a meta-analysis to be approximately 0.55 units.Noninferiority is likely, if the lower limit of the 90% CI of the mean group difference is above the noninferiority margin −δ [20].The lower limit of the 90% CI of the mean difference between S + HDM and L + HDM was −0.51, i.e., 0.04 units above −δ.Therefore, noninferiority of S could be established (Fig. 6).

Discussion
Our study indicated that S and L as carriers for HDM likely have no different influence on TNSS and nasal secretion.We tested this by exposing the same individuals to either carriers alone, clean air, and HDM + carrier in the ACC.As expected, no significant differences were found between S, L, and clean air.A mild nocebo effect was documented in the clean air exposure with a mean TNSS of 0.98 and nasal secretion of 0.23 g.In the S and L group, mean TNSS and nasal secretion were in similar ranges as in the clean air exposure, indicating that carriers alone did not have an effect on nasal symptoms besides a nocebo effect.When the carriers were combined with HDM, mean TNSS and nasal secretion increased significantly in both S + HDM and L + HDM groups compared to carriers alone.S was noninferior to L as a carrier; in fact, the observed mean TNSS increase was numerically higher in the S + HDM than in the L + HDM group.This descriptive mean difference slightly exceeded the MCID in TNSS of approximately 0.55 units that has been derived by Barnes et al. [24].This MCID approach might suggest a clinically meaningful difference between the carriers; however, it was not statistically significant.It should also be noted that the 90% CI calculated for noninferiority assessment was wide, ranging from −0.51 to 1.74, due to the large group variance.Thus, S might perform slightly worse than L in a repetition of the experiment.On closer examination, the lower mean TNSS in the L + HDM group can be attributed to two participants who reported fewer nasal symptoms during this exposure.They seemed to be poor L + HDM responders compared to the other participants.Removing these cases from the analysis set results in the L + HDM TNSS group mean to be slightly higher than the S + HDM TNSS group mean.Both individuals had not participated in HDM challenges at our site before, so we could not collate the findings with earlier results.We observed that both participants were the only ones who reported a mean difference >1 in the sneezing item of the TNSS with S + HDM.An irritant effect of S particles was unlikely since both participants scored a mean TNSS of <1 in the exposure with S alone.So far, we assumed that both L and S are immunologically inert and have no influence on IgE-mediated type 1 hypersensitivity reaction.Particle size distributions during S + HDM and L + HDM exposures were comparable, making varying upper airway deposition of allergen particles unlikely.Although this phenomenon of markedly different responses cannot be sufficiently explained, the total mean nasal secretion weights in both S + HDM and L + HDM groups were similar and therefore support the equivalence of S and L as carriers.
A direct comparison of our results to those of other groups is hardly possible since allergen materials, allergen concentrations, and participant selection differs.Zieglmayer et al. [2] reported a TNSS plateau >6 using 0.3 g mite allergen material/h.Hossenbaccus et al. [3] tested two different allergen concentrations and observed a mean peak TNSS of 5.74 at a modest concentration and 8.17 at a high concentration.Similarly, Zemelka-Wiacek et al. [4] documented mean peak TNSS values >6 at higher allergen concentrations and 4-6 at the lowest tested concentration.Khayath et al. [9] who also used a spray drying technique with allergen extract and normal saline, reported a positive dose-response relationship: mean TNSS >6 was found in 58% of participants at 15 ng/m 3 , in 63% at 25 ng/m 3 , and in 75% at 46 ng/m 3 Der p1 concentration.In our own HDM validation study, we reported a mean TNSS of 3.3 at 500 SQE/m 3 that increased only slightly when the allergen concentration was doubled [10].Basically, the TNSS results of this study are in the middle to lower range compared to the findings of other groups.The safety profile in all exposures was favorable, with only mild allergen-related adverse events.Considering other medical settings, changing from a sugar-based diluent to a saline-based diluent was found to be safe for parenteral use in intensive care settings, but more episodes of hypernatremia were reported [25].This potential side effect, however, is not of concern for inhaled use.Some limitations are worth noting.The ultimately small sample size reduced the power to detect differences between S and L. Also, the unexpected results of two apparently poor L + HDM responders could not be explained.This would have required resource-intensive repetition and probably extensions of the experiment.However, since our study showed that S + HDM was not inferior to L + HDM, this peculiarity may not be of future relevance to reliably induce moderate allergic nasal symptoms on average to test anti-allergic drugs.
In conclusion, S as a solute for spray drying of allergen extracts was safe and well tolerated when used as a carrier for HDM allergen exposure in the ACC.There was no difference to L which makes S an adequate and easy-touse carrier substitute.

Fig. 4 .
Fig. 4. Mean TNSS (a) and nasal secretion weight (b) with S + HDM and L + HDM for all participants.Colored shapes indicate cases with extreme differences between exposures of TNSS alone (square), nasal secretion alone (diamond, hexagon), or both (triangle).

Table 2 .
Descriptive results of TNSS and nasal secretion of per protocol population (n = 13)