A meta‐analysis on allergen‐specific immunotherapy using MCT® (MicroCrystalline Tyrosine)‐adsorbed allergoids in pollen allergic patients suffering from allergic rhinoconjunctivitis

Abstract Background The World Allergy Organization and the European Academy of Allergy and Clinical Immunology recommend to perform product‐specific meta‐analyses for allergen‐specific immunotherapies because of the high degree of heterogeneity between individual products. This meta‐analysis evaluates the efficacy and safety of Glutaraldehyde‐modified and MCT® (MicroCrystalline Tyrosine)‐adsorbed allergoids (MATA). Methods The databases MEDLINE, LILACS, embase, LIVIVO, Web of Science and Google (Scholar) were searched for publications on MATA up to June 2019. Primary endpoint was the combined symptom and medication score (CSMS). Secondary endpoints were single scores, immunogenicity and improvement of allergic condition. Secondary safety endpoints were the occurrence of side effects. A random effects model was applied with (standardized) mean differences ([S]MDs) including confidence intervals (CI). Heterogeneity was analyzed using the I2 index and publication bias using Egger's test and Funnel plots. Subgroups were analyzed regarding age and asthma status. Results Eight randomized double‐blind placebo‐controlled trials were selected for efficacy and 43 publications for safety analysis. In total, 4531 patients were included in this analysis including eight studies containing data on children and adolescents. AIT with MATA significantly reduced allergic symptoms and medication use with a SMD for CSMS of −0.8 (CI: −1.24, −0.36) in comparison to placebo. Heterogeneity was moderate between the studies. The total symptom score (−1.2 [CI: −2.11, −0.29]) and the total medication score (−2.2 [CI: −3.65, −0.74]) were also significantly reduced after MATA treatment. Patient's condition improved significantly after treatment with MATA, with an odds ratio of 3.05 (CI: 1.90, 4.90) when compared to placebo. The proportion of patients, who developed side effects was 38% (CI: 19%, 57%). No serious side effects occurred. Safety in the subgroups of asthmatic patients, children and adolescents did not differ from the overall patient population. Conclusions This meta‐analysis reveals a large body of evidence from publications investigating MATA. MATA significantly improved allergic symptoms and reduced the use of anti‐allergic medication in comparison to placebo, with an excellent safety profile. Especially for children and asthmatic patients, the use of MATAs can be considered as safe, because the safety profiles in these groups did not differ from the total patient population.

(CI: 19%, 57%). No serious side effects occurred. Safety in the subgroups of asthmatic patients, children and adolescents did not differ from the overall patient population.
Conclusions: This meta-analysis reveals a large body of evidence from publications investigating MATA. MATA significantly improved allergic symptoms and reduced the use of anti-allergic medication in comparison to placebo, with an excellent safety profile. Especially for children and asthmatic patients, the use of MATAs can be considered as safe, because the safety profiles in these groups did not differ from the total patient population.

K E Y W O R D S
adjuvanz, allergenimmuntherapie, allergische rhinokonjunktivitis, allergoid, Mikrokristallines Tyroson (MCT ® ) 1 | BACKGROUND Meta-analyses provide a systematic and statistically validated overview and combine the results of earlier studies; therefore, broadening and rating the evidence for the investigated body of research. 1 For example, a meta-analysis on allergen immunotherapy for allergic rhinoconjunctivitis by Dhami and colleagues gives such an overview but lacks product-specific details. 2 This may be sufficient in many areas of research, but in some cases, like treating patients with allergic rhinoconjunctivitis, the published meta-analysis lacks specificity because of the heterogeneity of the analyzed studies and diversity of included products. To overcome this, the World Allergy Organization (WAO) published a statement 3 in which they propose to conduct productspecific meta-analyses to establish clinical efficacy for individual products. Also, EAACI (European Academy of Allergy and Clinical Immunology) recommended in their allergen immunotherapy (AIT) guideline 4 a product-specific evaluation as a reasonable approach.
Since the products differ considerably with regard to allergenconcentrations, modifications, adjuvants, and dosing schemes, the results obtained from one product cannot be extrapolated to other products or be generalized. Because of their different structure, adjuvants, and adsorbents, modified allergen extracts like MicroCrystalline Tyrosine (MCT ® )-adsorbed allergoids cannot be compared with, for example, aqueous native allergen extracts.
The allergen extracts analyzed here are glutaraldehyde-modified (so called allergoids), as well as MCT ® -adsorbed. These MATAs (glutaraldehyde-modified and MCT ® -adsorbed allergoids), have been marketed already since the early 1970s. The modification of allergen extracts with glutaraldehyde was first described by Patterson and coworkers. 5 The modification with glutaraldehyde cross-links the proteins, resulting in high molecular weight allergen polymers. These inhibit IgE binding, leading to a reduced allergenicity, but do not decrease immunogenicity. 6 Because of the reduced allergenicity, the updosing can be done much faster without compromising safety. 7,8 However, allergoids are rarely applied without adjuvants like aluminium or MCT ® . MCT ® acts as a depot-adjuvant facilitating a slow release of the allergen extract, 9 it is biodegradable with a half-life of 48 h at the injection-site and a strong Th1 polarizing adjuvant without known safety concerns. 9 This short half-life in the tissue (biodegradability) and its biocompatibility indicates a favorable role as adjuvant compared to aluminium. 10 When looking especially at clinical trials, 1575 patients in placebo arms received MCT ® alone, without reporting any serious adverse events with regards to the treatment. 11 Also, a recent head-to-head comparison confirmed the superior safety profile of MCT ® when compared with aluminium. 12 The various MATA preparations include, for example, either a grass mix, a tree mix, ragweed or other weeds like plantain and/or mugwort. Many clinical trials, including randomized controlled trials (RCT) and double-blind placebo-controlled trials (DBPCT) demonstrated that MATAs are effective, safe, and well tolerated. 11,13,14 MATAs are licensed as a pre-seasonal and perennial treatment.
Here, we performed a systematic review and meta-analysis of studies conducted with the product-line of MATAs to compare the efficacy and safety with the current generalized meta-analysis by Dhami et al. and other product-specific analyses.

| Search strategy
An extensive search was conducted in the following databases: MEDLINE, LILACS, LIVIVO, and Web of Science. In addition, literature was searched using Google and Google Scholar. Search terms were "desensitization," "MATA," "tyrosine allergoid," and "glutaraldehyde-modified tyrosine adsorbate," including various combinations of these search terms (Additional file 1). Bencard Allergie GmbH has also provided the authors with a list of publications referring to studies utilizing MATAs.
Literature was included in this systematic review up to June 12, 2019. There were no language restrictions. Where applicable, relevant literature was translated into English. If full publications were not available online, they were requested (if possible) from the first author.

| Inclusion and exclusion criteria
For this meta-analysis, all publications were considered in which patients with allergic rhinoconjunctivitis (with or without asthma) were treated with MATA, independent of applied dose and treatment/study duration. Age was not restricted and studies with children and adolescents were also included. Studies using the following MATA allergens were included: tree pollen with or without other allergens, grass pollen mix with or without other allergens as well as ragweed with or without other allergens (Table 1). Regarding study design, DBPCTs, randomized controlled trials, controlled trials and uncontrolled studies that were either conducted prospectively or retrospectively, were included in the analysis. For the analysis of the efficacy endpoints (primary and secondary), only DBPCTs were included, because the blinding ensures an unbiased conduct of the studies.
Reviews, case reports, animal studies and studies containing data on aluminium absorbed allergen extracts were excluded.

| Endpoints
The primary endpoint was the combined symptom and medication score (CSMS) in randomized DBPCTs using MATA. The analysis was conducted similar to the primary analysis in the original studies. [14][15][16] In the two-year studies, only the first year of treatment was considered.
Covariates with regard to the primary endpoint were responders and non-responders as defined in the respective studies. [14][15][16][17][18] Secondary endpoints (also analyzed in the DBPCTs) were the total symptom score (TSS) including the individual symptoms sneezing, nasal obstruction, eye symptoms and cough, [14][15][16]19 the total medication score (TMS) [14][15][16]20 as well as immunologic parameters (allergen-specific Immunoglobulin G, sIgG). 13,14,16 Safety was assessed by analyzing the occurrence of solicited local and systemic adverse events as well as other adverse reactions with regard to the number of patients and the number of injections in the included DBPCTs, randomized controlled trials, controlled trials and uncontrolled studies.

| Data extraction and analysis
The data analyzed here were extracted from the publications by two different reviewers in Excel files. In case of discrepancies, this issue was discussed with the co-authors until a consensus was reached. The means and standard deviations (SD) of the CSMS, the TMS, the TSS, as well as the single symptom scores were extracted from the DBPCTs.
For the analysis of the total symptom score (TSS) all studies were used that provided results on nasal symptoms (sneezing, rhinorrhoea/congestion/nasal obstruction), eye symptoms, and coughing. [14][15][16]19 For the total medication score, all studies with data on T A B L E 1 Overview of the MATA allergens included in this meta-analysis with current tradenames of the respective market areas symptomatic medication were included (use of antihistamines, beclomethasone, chlorpheniramine, etc.). [14][15][16]20 The single scores were then added up to the TSS or TMS, respectively. If the data was not expressed in a daily score but for the whole pollen season/ observation period, the scores were converted by dividing the TSS or TMS by the days of the pollen season/observation period defined in the publications. The combined symptom and medication score (CSMS) was then generated by adding the TSS and TMS of the respective study. [14][15][16] To evaluate the treatment effect (improvement of allergic con-

| Statistical analysis
Random effects modeling was applied for meta-analysis if it was clinically or statistically suitable. The inverse variance (IV) method was applied for study weights. As summary statistics, the standardized mean difference (SMD) was used for results of a continuous outcome or odds/risk ratio (OR/RR), for results of a binary outcome.
Where applicable, a confidence interval (CI) of 95% was included.

| Analysis of heterogeneity and publication bias
Heterogeneity was identified with the I 2 test. With regard to the I 2 test, an I 2 value <25% was considered as homogenous between the analyzed studies. A I 2 value of 26%-50% would identify a low heterogeneity and 51%-75% a moderate heterogeneity between studies. An I 2 value >75% was defined as high heterogeneity. 23 Comparison of means between studies was done with the Z-test.
To test for publication bias, funnel plots (with 95% CI, where applicable) were created for the primary and secondary efficacy outcomes 24 and tested with the Egger's regression test 24

| Subgroup analysis
In addition, subgroup analyses were performed with regard to the safety of MATAs in vulnerable groups like children and adolescents, as well as asthmatics.
The publications were checked for inclusion of data on patients <18 years. Studies with a mixed patient collective (children/adolescents and adults), or only adults were also analyzed for comparison.

| Systematic review
Two hundred sixty-nine publications were found after database and  The DBPCTs were also analyzed regarding methodological quality: six of the eight studies gained three score points or more, [13][14][15][18][19][20] of which three studies scored five points. 13,14,18 Two studies (both from 1979) gained two points. 16,17 Details on the distribution of the score points are displayed in Table 2.

| Study characteristics and patient population
All DBPCTs which were considered for analysis were published be-

| Secondary efficacy analysis: total (TSS, TMS) and single scores
The TSS was evaluable in 4 DBPCTs with 234 patients: [14][15][16]19 using the random effects model, the SMD was −1.20 (CI: −2.11; −0.29), when comparing MATA treatment with placebo. However, the heterogeneity was high between the studies with an I 2 of 85%. The test for overall effects (Z) was 2.59 when comparing the means of the studies (p = 0.01, Figure 3A).
Four studies [14][15][16]20 were analyzed with regard to the TMS (N = 241 patients): with application of the random effects model, the

| Asthmatic patients
Three studies 26 occurred in 22% of the asthmatic patients. In comparison, 32% of non-asthmatics and 40% of patients included in the asthmatics/nonasthmatics studies experienced side effects. Of the patients for whom the asthma status was not known, 58% experienced side effects ( Figure 7A, Additional File 8A). More detailed, 18% of asthmatic patients experienced local reactions and 4% experienced systemic reactions. In the studies with only non-asthmatics and both (asthmatics as well as non-asthmatics), 34% and 28% of the patients, respectively, experienced local reactions. Systemic reactions occurred in 9% of the non-asthmatic and in 10% of the mixed population (asthmatic/non-asthmatic patients). Of the patients with unknown asthma status, 59% experienced local reactions and 21% experienced systemic reactions ( Figure 7B, C, Additional file 8B, C).

| DISCUSSION
This meta-analysis demonstrates that AIT with MATA significantly  Since the studies analyzed here are from the 1970s to 1990s, the CSMS used in these DBPCTs cannot be identical as the CSMS published in the EAACI position paper from 2014. 64 However, we merged data in analogy to the CSMS of the position paper. As we Evaluation of the quality of the DBPC studies using the Jadad score-that was published in 1996-revealed that only two of the eight studies lacked sufficient quality. 16,17 The six remaining studies scored three points or more. [13][14][15][18][19][20]  In addition to the overall safety analysis, special emphasis was laid on asthmatics and children/adolescents in our meta-analysis.
Asthmatic patients are considered to be prone to side effects 67 With this early treatment option, the progression to allergic asthma can be considerably inhibited, if not prevented entirely, as was shown by Möller at al. 69 In this study, asthma symptoms of children who were treated with immunotherapy improved significantly in comparison to the placebo group, highlighting the importance of AIT already in childhood.
The analysis presented here is based on the first year in case of 2-years studies. This is due to the fact that in many studies included in the meta-analysis, the treatment and endpoints were changed after the first year.
It may be considered as a limitation, that this product-line specific meta-analysis was not split with regard to the single allergens or allergen-mixtures. 3 Although this is recommended in the WAO statement, we believe it is sufficient to test the product line with the different allergens and allergen-mixtures, as we mainly aimed to evaluate the efficacy and safety of MATAs, independent of the allergen. Furthermore, the allergens evaluated in this meta-analysis are solely derived from pollen (trees, grasses, weed), which provides a certain homogeneity. Therefore, despite the use of different allergens in the analyzed studies, we did not perform allergen-specific subgroup analyses.
A similar approach (product-line specific meta-analyses without allergen-specific subgroup analyses) was already done by Mösges