Bacterial lysate add‐on therapy to reduce exacerbations in severe asthma: A double‐blind placebo‐controlled trial

Abstract Background Asthma exacerbations are frequently induced by respiratory tract infections (RTIs). Bacterial lysates have been described to possess immune‐modulatory effects and reduce RTIs as well as asthma symptoms in children. However, whether bacterial lysates have similar effects in adult asthma patients is unknown. Aims To reduce asthma exacerbations by add‐on bacterial lysate therapy in adults with severe asthma and to characterize the clinical and immune‐modulatory effects of this treatment. Methods Asthma patients (GINA 4) with ≥2 annual exacerbations in the previous year were included. The intervention regimen consisted of OM‐85/placebo for 10 consecutive days per month for 6 months during two winter seasons. Primary end‐point was the number of severe asthma exacerbations within 18 months. The study was approved by the national and local ethical review board and registered in the Dutch Trial Registry (NL5752). All participants provided written informed consent. Results Seventy‐five participants were included (38 OM‐85; 37 placebo). Exacerbation frequencies were not different between the groups after 18 months (incidence rate ratio 1.07, 95%CI [0.68–1.69], p = 0.77). With the use of OM‐85, FEV1% increased by 3.81% (p = 0.04) compared with placebo. Nasopharyngeal swabs taken during RTIs detected a virus less frequently in patients using OM‐85 compared to placebo (30.5% vs. 48.0%, p = 0.02). In subjects with type 2 inflammation adherent to the protocol (22 OM‐85; 20 placebo), a non‐statistically significant decrease in exacerbations in the OM‐85 group was observed (IRR = 0.71, 95%CI [0.39–1.26], p = 0.25). Immune‐modulatory effects included an increase in several plasma cytokines in the OM‐85 group, especially IL‐10 and interferons. Peripheral blood T‐ and B cell subtyping, including regulatory T cells, did not show differences between the groups. Conclusion Although OM‐85 may have immune‐modulatory effects, it did not reduce asthma exacerbations in this heterogeneous severe adult asthma group. Post hoc analysis showed a potential clinical benefit in patients with type 2 inflammation.


| INTRODUC TI ON
Asthma patients with recurrent exacerbations suffer the highest disease burden and account for over half of the asthma-associated healthcare expenditure, with 80% of the total direct costs of asthma attributed to the treatment of exacerbations. 1 About 50% of exacerbations are elicited by viral respiratory tract infections (RTIs). 2,3 Therefore, the prevention of RTIs is critical for reducing exacerbation frequencies. 4 In several European and Asian countries, bacterial lysates have been used for the prevention of recurrent RTIs since the early 1950s. 3,[5][6][7] Bacterial lysates are non-viable bacterial extracts obtained by either chemical or mechanical cellular lysis of bacterial cultures and lyophilization. 5,[8][9][10] Although the most well-known bacterial lysate OM-85 has shown clinical efficacy in preventing RTIs, the immune-modulatory effects of OM-85 in humans are still not fully elucidated. 5,7,[11][12][13] Murine studies suggest that dendritic cell (DC) activation in gut-associated lymphoid tissue results in local and systemic stimulation of antiviral cytokine production, including Th1 cytokines such as interferons, as well as local immunoglobulin secretion. 14,15 Asthma is a phenotypically heterogeneous disease, mainly driven by type 2 inflammation. In murine asthma studies, a decrease in pulmonary Th2 cytokines was observed after OM-85 treatment. 10,[14][15][16] Additionally, OM-85 induced pulmonary recruitment of regulatory T cells along with a decrease in bronchial hyperreactivity in animal models. 15 Some studies in children suggest shifts in the Th1/Th2 cytokine balance. 17,18 Several recent reports and meta-analyses concluded bacterial lysate therapy has a favourable safety profile and reduced RTIs, COPD exacerbations and preschool wheezing episodes. 5,7,[11][12][13]18 Also, a recent study showed bacterial lysates Conclusion: Although OM-85 may have immune-modulatory effects, it did not reduce asthma exacerbations in this heterogeneous severe adult asthma group. Post hoc analysis showed a potential clinical benefit in patients with type 2 inflammation.

K E Y W O R D S
asthma, bacterial lysates, exacerbations, immune modulation, type 2 inflammation

G R A P H I C A L A B S T R A C T
Seventy-five severe asthma patients were included (38 OM-85; 37 placebo). While intention to treat (ITT) analysis revealed no difference in exacerbation frequencies between the groups after 18 months, per protocol analysis in patients with type 2 inflammation (PPT2) showed a non-significant decrease. Nasopharyngeal swabs detected a virus less frequently in the OM-85 group during respiratory tract infections. Immune-modulatory effects were seen in several plasma cytokines. To conclude, OM-85 might be an effective add-on therapy in severe asthma patients with type 2 inflammation.

Key Messages
• OM-85 did not reduce exacerbation frequency in the studied group of heterogeneous adult severe asthma patients; • OM-85 may be an effective add-on therapy to reduce exacerbations in asthma patients with type 2 inflammation; • Immune-modulatory effects were observed with the use of OM-85.
to be effective in the improvement of the clinical course of allergic rhinitis. 19 There is some evidence for a decrease in asthma symptoms in school children. 17,18,20 In contrast, studies of adult patients with asthma are scarce. Therefore, our aim was to investigate the potential of add-on bacterial lysate therapy to reduce asthma exacerbations in adults with severe asthma and to characterize the clinical, microbiological and immunemodulatory effects of this treatment.

| Study design and setting
We conducted Remaining drugs were destroyed.

| Study protocol
The protocol included 3 monthly visits at baseline (T = 0) and thereafter at 3, 6, 9, 12, 15 and 18 months (study end) and exacerbation visits ( Figure 1). During these visits, we assessed clinical condition and asthma control and took a nasopharyngeal swab. Every 6 months, blood was drawn. During asthma exacerbations, additional nasopharyngeal swabs and blood were collected when the patient presented at the hospital. During common colds, patients were asked to collect a nasopharyngeal swab at home. Participants received a weekly digital questionnaire addressing ACQ, number of RTIs and exacerbations. Participants that stopped prematurely with the study, were included in the analysis of the primary end-point after consent.

| Study outcomes
The primary study outcome was the number of asthma exacerbations within the first 18 months after start of intervention. An asthma

| Sample size calculation
Literature on the use of OM-85 for exacerbation prevention in adults with asthma was unavailable. Power analysis was targeted at a 20% reduction in asthma exacerbations, based on previous literature describing a 20%-40% reduction of RTIs in children and adults. 24,25 Two x 36 patients were required to have 80% chance of detecting the 20% decrease in asthma exacerbations measured at a 5% significance level (two-sided, assumed participants having on average 2 ± 0.6 exacerbations per year based on pilot data).

| Randomization and treatment allocation
Block randomization of all patients was stratified for atopy status (de-   Table S1 lists antibodies used for PBMC phenotyping; Figure S3 shows the gating strategies for CD4 + and CD8 + T cells and intracellular cytokines, regulatory T cells and B cells. 29

| Post hoc analysis
Post hoc subgroup analyses, with identical outcomes and statistical plan as described above, were performed for patients being adherent to the study protocol (PP) and PPT2 (N = 42). Rationale of this PPT2 analysis is that bacterial lysate therapy decreases childhood asthma exacerbations, in asthma that often is allergic in origin. 17,18,20 Patients in the PP analysis were either found to be non-adherent or
Adherence, as monitored by questionnaires and returned study medication packages, was not different between the two groups ( Figure 2).

| Primary outcome; intention to treat
The  (Table S2). A separate analysis of patients <40 years of age and >40 years of age did not show differences between OM-85 and placebo groups.   (Figure 3c; Table 3).

| Microbiological and immunological analysis; intention to treat
We next compared relevant microbiological and immunological parameters between the two treatment groups. At baseline, 3 and 6 months, the proportion of samples in which viruses could be detected did not differ ( Figure S2). During asthma exacerbations, an additional nasopharyngeal swab was taken when participants (N = 63) were visiting the research facility for evaluation. A respiratory virus was detected in 54% of these swabs, with no differences between treatment groups ( Figure S2; virus specification can be found in Table S3).
No significant differences between treatment groups were observed for absolute numbers of blood eosinophils and neutrophils, serum IgG, IgA, IgM and IgE at baseline and during the 18-month follow-up (Table 3). IFNλ was higher at baseline within   Table 3).
This was reflected in a higher fold-change ratio for plasma IL-10 (p = 0.001) in the OM-85 group as compared to placebo group ( Figure 4).     Table 4). Fold-change ratios between T0 and T12 were increased for IL-10 (p = 0.0001) and

TA B L E 3 Secondary end-points for patients included in the intention to treat analysis (N = 75)
IFNγ (p = 0.0005) in the OM-85 group compared to placebo group ( Figure 4).
Finally, flow cytometric analysis of intracellular cytokine levels in circulating CD4 + and CD8 + T cells in the PPT2 cohort did not show differences between the treatment groups over time ( Table 4). The proportions of naive and memory CD4 + T cell subsets and B cell subsets did not differ between OM-85 and placebo-treated patients (Table S4). While proportions of CD86 + regulatory T cells of CD4 + T cells showed a trend towards a significant increase over time, regulatory T cells did not show differences over time. Regulatory T cell activation, defined as proportions of naïve Tregs and active Tregs, also did not differ between the treatment groups over time (Table S4).
Gating strategy can be found in Figure S3. Previous studies in children showed that bacterial lysates were possibly effective in reducing asthma exacerbations. 18,20 Unfortunately, these studies were not all designed or powered to monitor exacerbation frequency. Studies in adults with COPD reported a decrease in exacerbation frequency with the use of bacterial lysates. 7,32 In our study, age was inserted as a covariate in the binomial regression model and was not statistically significant. We did not detect differences in the primary outcome in the younger age group <40 years of age. Childhood asthma is known to be mostly driven by allergic airway inflammation. 33 Asthma in adulthood is more heterogeneous, with several endotypes identified according to the presence of allergies, eosinophilia, neutrophilia and obesity.

| DISCUSS ION
Although ~60% of patients included in our study showed allergydriven asthma, the complete study group was small and heterogeneous with regard to the other asthma endotypes. As our post hoc analysis in patients with type 2 inflammation revealed a trend towards exacerbation reduction, we hypothesize that a longer treatment duration and/or different treatment regimen in a larger and more homogenous group of patients with type 2 inflammation may reveal a beneficial effect for OM-85.
Several studies and meta-analyses describe a reduction of recurrent RTIs with the use of OM-85 in adults. 15 In the self-collected nasopharyngeal swabs, we did less frequently detect a virus in patients using OM-85 during a clinical episode of RTI (OM-85 30.5% vs. placebo 48.0%, p = 0.02), which is in accordance with previous literature.
However, self-reported RTIs are prone to subjective self-reporting and could include episodes of allergic rhinitis. Nevertheless, we did not observe a difference in the numbers of viral RTI-associated asthma exacerbations between patients using OM-85 and placebo.
Possibly, when having acquired a viral infection, immunomodulation by bacterial lysates in severe asthma patients does not prevent further progression to an asthma exacerbation.
Despite the absence of effects on exacerbation frequency in the complete study group, we did find evidence for immune-modulatory effects induced by bacterial lysate therapy. These effects were F I G U R E 4 Plasma cytokine fold-change ratios between baseline and 12 months for both the OM-85 and placebo group (shown as mean fold-changes). Significant differences (p < 0.005) after Mann-Whitney U test shown as *for intention to treat (ITT) and #for patients with type 2 inflammation being adherent to the study protocol (PPT2)  Although carefully designed, our study has limitations. First, due to the unknown effect of OM-85 in adults with asthma, we powered our cohort based on paediatric studies. Next, the number of asthma exacerbations during the study was lower than anticipated in the complete study group (mean number of 1.2 ± 1.4, as compared to 2.7 ± 1.2 before study start). Reasons for this could be a trial participation effect, as a result of the more frequent hospital visits or possibly increased therapy compliance and regression to the mean.
This was observed in several other asthma studies and referred to as the Hawthorne effect. Nevertheless, our results suggest a potential clinically relevant beneficial effect for patients with type 2 inflammation, and therefore, this might be a good selection indicator as a starting point for future studies into the effect of bacterial lysate therapy. A different treatment regimen of OM-85, which is currently being studied in the PrecISE network study (NCT04129931), might be more effective in adults.
In conclusion, our study did not demonstrate a benefit of bacterial lysate therapy on the incidence of asthma exacerbations in a heterogeneous group of adult patients with severe asthma and recurrent exacerbations. However, we did find an increase in FEV1% and non-specific immune-modulatory effects. Interestingly, post hoc subgroup analysis did show a trend towards a positive effect on exacerbation frequency in asthma with type 2 inflammation.

ACK N OWLED G EM ENTS
We thank all patients for their endless efforts to participate in this study. Also, we thank M. Mondeel for her assistance in contacting pharmacies and general practitioners, and M. Dorleijn for analysing baseline ELISA data.

E TH I C A L A PPROVA L
The study was approved by the national and local ethical review board and registered in the Dutch Trial Registry (NL5752). All participants provided written informed consent.

DATA AVA I L A B I L I T Y S TAT E M E N T
Data are available upon reasonable request via the corresponding author.