A randomised controlled trial of EP395, a novel anti-inflammatory macrolide, in stable COPD patients Authors: Henrik Watz, Stephanie Korn, Oliver Kornmann, Dave Singh, Tom Wilkinson, Kate Hanrott, Karl J. Staples, Jodie Ackland, Virginia Norris Published: 2026 ABSTRACT BackgroundMacrolide antibiotics have immunomodulatory activity and when taken chronically reduce exacerbations of COPD. However, chronic use can cause bacterial resistance. EP395 (glasmacinal), a novel macrolide, is being developed as a treatment to reduce exacerbations of COPD without inducing antimicrobial resistance.MethodsIn this double-blind, placebo-controlled, phase 2a trial (NCT05572333), patients (≥45 years old, diagnosed with COPD for ≥2 years and stable on at least one maintenance inhaled therapy) were randomised (2:1) to EP395 or placebo daily for 12 weeks. The primary objective was safety, with key secondary objectives assessing pharmacodynamic effects of EP395.ResultsA total of 61 patients were randomised (42 EP395, 19 placebo). A 12-week course of EP395 was well tolerated: no serious adverse events were considered related to EP395, and adverse events occurred in similar proportions in both groups (64.3% EP395, 63.2% placebo). Four patients were withdrawn due to adverse events (three EP395, one placebo). Sputum neutrophil elastase and myeloperoxidase, mediators of neutrophil activation, were reduced with EP395 (treatment difference (log scale): neutrophil elastase −0.415 ng·mL−1, 95% CI −0.787 to −0.043 ng·mL−1, p=0.030; myeloperoxidase −0.282 ng·mL−1, 95% CI −0.640 to 0.076 ng·mL−1, p=0.119). Relative changes in neutrophil elastase and myeloperoxidase from baseline with EP395 were 66% and 75%, respectively, of those observed with placebo. Exploratory 16S rRNA sequencing of sputum showed EP395 had no detectable effect on the lung microbiome, including the proportion of pathogenic Proteobacteria species.ConclusionIn patients with stable COPD, EP395 for 12 weeks was well tolerated, demonstrated selective anti-inflammatory activity and had no detectable effect on the lung microbiome. Background Chronic use of macrolide antibiotics, in particular azithromycin, reduces exacerbations in patients with chronic obstructive pulmonary disease (COPD) [1–4]. In placebo-controlled trials, long-term treatment with azithromycin significantly reduces the frequency and rate of COPD exacerbations by up to 40% [4, 5]. Off-label azithromycin is recommended in the Global Initiative for Chronic Obstructive Lung Disease (GOLD) treatment guidelines as an option for patients who continue to have exacerbations despite the use of triple inhaled therapy or dual bronchodilators when inhaled corticosteroids (ICS) are not indicated [6]. However, long-term use is associated with the development of bacterial resistance to macrolide antibiotics [4, 7–9]. Additionally, gastrointestinal disturbances are commonly reported side-effects of azithromycin (and other antimicrobials), and azithromycin has been associated with hearing disturbances (including hearing loss, deafness and/or tinnitus) and QT prolongation [10]. The efficacy of macrolides in the prevention of COPD exacerbations is thought to be principally the result of their immunomodulatory effects. Macrolides have anti-inflammatory activity and promote the resolution of inflammation [1, 11, 12]. EP395 (glasmacinal) is a novel oral macrolide with preclinical anti-inflammatory effects and negligible antimicrobial activity [13, 14] that is being developed as a potential treatment to reduce COPD exacerbations without the risk of developing antimicrobial resistance or disrupting the microbiome. This was the first trial of EP395 in patients with COPD. The primary objective was to assess safety and tolerability over 3 months, with secondary objectives evaluating the pharmacodynamic effects of EP395, including the impact on lung and systemic inflammatory markers. An exploratory assessment of the impact of EP395 on the lung microbiome was conducted. Methods Trial design This randomised, double-blind, placebo-controlled, parallel group, phase 2a trial ran from November 2022 to November 2023 at six sites in Germany and the UK. Eligible patients were randomised 2:1 to 375 mg EP395 or placebo once daily orally, after an overnight fast, for 12 weeks. Once daily 375 mg EP395 was the maximum dose supported by available toxicology data and was shown to be well tolerated in the first-in-human trial of EP395. EP395 and placebo capsules were identical in appearance. Patients took the first dose at the clinical site and further doses at home (except on clinic visit days). Patients attended fortnightly site visits, with final safety checks approximately 16 days after their last dose. The trial design was reviewed by an independent group of respiratory experts, and the trial was approved by relevant regulatory authorities and ethics committees and conducted in accordance with International Council for Harmonisation of Technical Requirements for Pharmaceuticals for Human Use and Good Clinical Practice. All patients provided full written informed consent (ClinicalTrials.gov: NCT05572333; registration date 6 October 2022; EudraCT number: 2021-005787-22). Population Eligible patients had a COPD diagnosis for ≥2 years, a post-bronchodilator forced expiratory volume in 1 s (FEV1)/forced vital capacity ratio <0.70, FEV1 <70% predicted and a history of sputum production (bronchitic phenotype). They had to be stable on at least one maintenance inhaled therapy without change, or COPD exacerbation, within the 3 months prior to screening. Current and recent (≤3 months) smokers were limited to approximately 25% of the trial population. The full list of inclusion/exclusion criteria is detailed in supplement 1. Assessments Safety The primary end-points were safety assessments, including standard laboratory safety tests, 12-lead ECGs and vital signs (heart rate, blood pressure, respiratory rate, temperature) measured at each visit. 24-h urine cortisol (baseline and end of treatment) was conducted in ICS users to measure cortisol. Adverse events (AEs) were monitored throughout. Induced sputum Sputum induction was conducted at baseline (Screening and Day 1, pre-dose), Week 6 and end of treatment (Weeks 10 and 12), using 3%, 4% and 5% saline inhaled sequentially up to 15 min via an ultrasonic nebuliser. Sputum cells and supernatant were processed as detailed in supplement 1. A differential cell count (400 cells) included neutrophils, eosinophils, macrophages, lymphocytes and epithelial cells. Sputum supernatant was analysed for tumour necrosis factor (TNF)-α, interleukin (IL)-6, IL-8, IL-1β, macrophage inflammatory protein (MIP)-1α, MIP-1β, monocyte chemotactic protein (MCP)-1 and granulocyte–macrophage colony-stimulating factor (GM-CSF) (Bio-Rad Bio-Plex Pro Human Cytokine 27-Plex assay kit); IL-23, IL-25, IL-33 and CXC motif chemokine ligand (CXCL)-1 (Human U-plex Custom Biomarker Group 1, MSD); total neutrophil elastase (NE) (ELISA, Abcam); matrix metalloproteinase (MMP)-9 (Human U-Plex Custom Immuno-Oncology Group 1 Singleplex assay, MSD); myeloperoxidase (MPO) (R-Plex Human MPO Singleplex assay, MSD); and surfactant protein (SP)-D (ELISA, R&D Systems). Blood biomarkers Blood samples were collected at baseline (Day 1, pre-dose), Week 6 and Week 12 for biomarker analysis. Serum was analysed for high-sensitivity C-reactive protein (CRP) (Tina-quant, Roche Diagnostics); IL-6 and TNF-α (Human Pro-inflammatory Panel 1 V-Plex, MSD); and α2 macroglobulin (DiAgam). Citrate plasma was analysed for fibrinogen (STA, Roche Diagnostics). Health-related quality of life and symptoms The St George's Respiratory Questionnaire (SGRQ) was administered monthly during site visits, and the Exacerbations of COPD tool (EXACT)-respiratory symptoms (E-RS) was completed daily via eDiary. For SGRQ total score, the change from baseline at each post-treatment time-point was calculated (0 to 100 scale; lower scores=better quality of life; minimum clinically important difference (MCID), 4 points [15]), as well as the proportion of patients with a ≥4-point improvement. For E-RS, summary statistics were calculated for total scores (0 to 40 scale; lower scores=milder respiratory symptoms) and domain scores (breathlessness, chest symptoms, cough/sputum). EP395 plasma concentrations Sparse plasma sampling for EP395 was analysed using a validated high-performance liquid chromatography/mass spectrometry method (validated range 0.5–500 ng·mL−1). Genotypic analysis of the lung microbiome The effect of EP395 on the lung microbiome was assessed using DNA extracted from sputum (supplement 1). After sputum biomarker analysis, 17 patients (11 EP395, 6 placebo) had sufficient sputum sample remaining for analysis of 16S rRNA V4–V5 regions. Details of DNA extraction and analysis are provided in supplement 1. Statistics Because the trial's primary objective was safety, no formal sample size calculation was performed; a 2:1 randomisation was used to maximise EP395 safety data. Patients were stratified by ICS use and smoking status. The randomised allocation sequence was generated by an independent statistician not involved in the conduct of the trial. A central web-based randomisation system was used. The Safety Analysis Set included all randomised patients who received at least one dose; the Full Analysis Set included those who also provided post-Day 1 sputum or blood biomarker data. Primary end-points were analysed descriptively using the Safety Analysis Set; secondary and exploratory end-points were analysed descriptively using the Full Analysis Set. Descriptive statistics were used to explore the secondary and exploratory data. For sputum and blood biomarker data, because the distribution was anticipated to be skewed, absolute values were log-transformed using the natural logarithm. All further analyses were done using log-scale values. Group differences in absolute change from baseline were assessed using Welch's t-test (two-sided, 5% significance). No adjustment for multiple testing was made owing to the trial's exploratory nature. No imputation of missing values was used. Genotypic lung microbiome analyses are detailed in supplement 1. Results Trial population Of 123 patients screened, 61 patients were randomised (42 EP395, 19 placebo). Of the 42 EP395 patients, 39 (92.9%) completed the trial and three (7.1%) discontinued due to AEs. Of the 19 placebo patients, 18 (94.7%) completed the trial and one (5.3%) discontinued due to an AE (see below). Demographic and baseline characteristics of the randomised patients (table 1) were comparable between groups, and consistent with the expected COPD population, including spirometry and concomitant medication use. TABLE 1Demographic and baseline characteristics of randomised patientsEP395 375 mgPlacebo Sex Female15 (35.7)9 (47.4) Male27 (64.3)10 (52.6) Age, years (median (min–max)) 66 (46–82)66 (50–76) Body mass index, kg·m−2 27.2±4.5726.6±4.27 Race White41 (97.6)19 (100.0) Black or African American1 (2.4)0 Ethnicity, not Hispanic or Latino 42 (100.0)19 (100.0) Current smoker 12 (28.6)4 (21.1) Tobacco consumption, pack years# 42±2238±15 Maintenance therapy includes ICS 30 (71.4)14 (73.7) Baseline¶ pre-bronchodilator spirometry FEV1, L1.44±0.521.44±0.55 FEV1 % predicted48.83±11.6050.29±10.96 FEV1/FVC ratio0.48±0.100.49±0.10Baseline¶ SGRQ Total score47.79±15.5641.66±17.34 Symptom score65.41±19.3057.07±20.99 Baseline blood eosinophils <0.1×109·L−16 (15.4)3 (16.7) 0.1 to <0.3×109·L−126 (66.7)12 (66.7) >0.3×109·L−17 (17.9)3 (16.7)Data are presented as mean±sd or n (%), unless otherwise stated. FEV1: forced expiratory volume in 1 s; FVC: forced vital capacity; ICS: inhaled corticosteroid; SGRQ: St George's Respiratory Questionnaire. #: pack years=(number of cigarettes·day−1/20)×years of smoking; ¶: baseline=Day 1, pre-dose (screening used if Day 1, pre-dose was missing). Demographic and baseline characteristics of randomised patients Data are presented as mean±sd or n (%), unless otherwise stated. FEV1: forced expiratory volume in 1 s; FVC: forced vital capacity; ICS: inhaled corticosteroid; SGRQ: St George's Respiratory Questionnaire. #: pack years=(number of cigarettes·day−1/20)×years of smoking; ¶: baseline=Day 1, pre-dose (screening used if Day 1, pre-dose was missing). Safety and tolerability On treatment adverse events AE incidence was similar between groups: EP395 64.3% (27 AEs) and placebo 63.2% (12 AEs). The incidence of AEs assessed as treatment related was also similar across groups: EP395 21.4% (nine AEs) and placebo 26.3% (five AEs) (table 2). Most AEs were mild or moderate. Three patients had severe AEs (two EP395, one placebo). Two serious AEs (pneumonia and radius fracture) occurred in the EP395 group, both assessed as unrelated to treatment. TABLE 2Adverse events (AEs) considered related to treatment (Safety Analysis Set)EP395 375 mg#Placebo¶AEs, nPatients with AE,n (%)AEs, nPatients with AE,n (%) Preferred term Vision blurred00 (0)11 (5.3) Abdominal pain upper00 (0)11 (5.3) Diarrhoea11 (2.4)00 (0) Duodenitis11 (2.4)00 (0) Dyspepsia11 (2.4)00 (0) Gastro-oesophageal reflux disease11 (2.4)11 (5.3) Nausea33 (7.1)11 (5.3) Fatigue00 (0)11 (5.3) Urinary tract infection00 (0)11 (5.3) Hepatic enzyme increased22 (4.8)00 (0) Back pain00 (0)11 (5.3) Headache22 (4.8)00 (0) Rhinitis allergic11 (2.4)00 (0) Hyperhidrosis11 (2.4)00 (0) Pruritus22 (4.8)00 (0) Rash11 (2.4)00 (0) Total 16 9 (21.4) 7 5 (26.3) Related AEs were defined by the investigator as probably or possibly related to treatment or with a missing assessment. #: n=42; ¶: n=19. Adverse events (AEs) considered related to treatment (Safety Analysis Set) Related AEs were defined by the investigator as probably or possibly related to treatment or with a missing assessment. #: n=42; ¶: n=19. Four patients discontinued trial treatment due to AEs: three in the EP395 group (duodenitis/dyspepsia (possibly related), rash (probably related), pneumonia (unrelated)) and one in the placebo group (reflux/abdominal pain). The EP395 patient with duodenitis/dyspepsia had Barrett's oesophagus on endoscopy, suggesting pre-existing gastric issues. Notably, AE incidence was similar between groups for gastrointestinal disorders (EP395 19.0%, placebo 21.1%); infections and infestations (EP395 21.4%, placebo 31.6%); and respiratory, thoracic and mediastinal disorders (EP395 26.2%, placebo 21.1%). No auditory related AEs (considered AEs of special interest) were reported. Skin-related AEs considered possibly or probably related to treatment occurred only in the EP395 group (one rash, two pruritus, one hyperhidrosis). Other safety assessments No changes were observed in mean vital signs or laboratory parameters, including 24-h urinary cortisol. Two EP395 patients had mild liver enzyme elevations (reported as AEs, table 2), but concentrations of alanine aminotransferase and aspartate aminotransferase did not exceed three times the upper limit of normal, and both patients completed the trial. There were no changes from baseline in mean ECG parameters, except a potential increase in mean change from baseline QTcF compared with placebo (approximately 5 ms), but the finding was not consistent across time-points and was not correlated with EP395 plasma concentration. No patients had a QTcF >500 ms or QTcF increase ≥60 ms from baseline, and no ECGs were abnormal and clinically significant. Inflammatory biomarkers Lung inflammation (sputum) Baseline sputum differential cell counts demonstrated neutrophilic inflammation (neutrophil count was approximately 80% in both groups). EP395 had no effect on sputum total or differential cell counts, including neutrophils (supplement 1: table S1). Biomarkers of lung inflammation were analysed in sputum (total NE, MPO, TNF-α, IL-6, IL-8, IL-1β, IL-23, IL-25, IL-33, MIP-1α, MIP-1β, MCP-1, GM-CSF, SP-D, MMP-9 and CXCL-1; figure 1a). The sputum biomarkers affected by EP395 were total NE and MPO, and they were reduced following treatment. The reduction in NE was statistically significant (p=0.030), and there was a tendency for a reduction in MPO (p=0.119) (tables 3 and 4). Relative changes in NE and MPO from baseline with EP395 were 66% and 75%, respectively, of those observed with placebo. Repeated measures correlation revealed a strong intra-individual association between log NE and log MPO (r=0.76, p<0.001; figure 1b). Subgroup analyses by smoking status (current/former) showed no differences; only NE and MPO were affected by EP395, consistent with the Full Analysis Set. Sputum inflammatory markers. a) Sputum inflammatory markers: difference in mean change from baseline (log scale) between EP395 and placebo. Baseline is the mean of Screening and Day 1, and end of treatment is the mean of Week 10 and 12. 95% confidence intervals shown. NE, MMP-9, MPO and SP-D are in ng·mL−1; all other markers are in pg·mL−1. b) Repeated measures correlation (rrm) analysis of the intra-individual association for paired measurements across time-points for sputum log NE and sputum log MPO. CXCL-1: CXC motif chemokine ligand-1; GM-CSF: granulocyte–macrophage colony-stimulating factor; IL: interleukin; MCP-1: monocyte chemotactic protein-1; MIP: macrophage inflammatory protein; MMP-9: matrix metalloproteinase-9; MPO: myeloperoxidase; NE: neutrophil elastase; SP-D: surfactant protein-D; TNF-α: tumour necrosis factor-α. TABLE 3Full summary statistics of sputum inflammatory markers affected by EP395 (Full Analysis Set)NE (ng·mL−1)MPO (ng·mL−1)EP395#Placebo¶EP395#Placebo¶ Baseline + Geometric mean19451446824787 Log scale7.506±1.2997.205±0.7586.653±1.3126.590±0.982 Week 6 Geometric mean17001557687795 Log scale7.438±1.3497.351±0.5626.532±1.4076.678±0.844 Log scale abs. CFB−0.229±0.751−0.061±0.650−0.266±0.779−0.073±0.707 End of trial § Geometric mean15701685644873 Log scale7.292±1.3487.362±0.6696.402±1.3476.697±0.981 Log scale abs. CFB−0.243±0.6560.172±0.531−0.215±0.6860.067±0.491Treatment difference baseline to end of trial Log scale (95% CI)−0.415 (−0.787 to −0.043)−0.282 (−0.640 to 0.076) p-valueƒ0.0300.119Of note, not all patients were able to provide sufficient sputum for all time-points or all analytes. Data presented as mean±sd, unless otherwise indicated. abs.: absolute; CFB: change from baseline; MPO: myeloperoxidase; NE: neutrophil elastase. #: n=39; ¶: n=18; +: mean of Screening and Day 1 pre-dose, or single result if sputum sample only available from one of the two visits; §: mean of Week 10 and 12, or single result if sputum sample only available from one of the two visits; ƒ: two-sided Welch's test with significance level of 5%. Full summary statistics of sputum inflammatory markers affected by EP395 (Full Analysis Set) Of note, not all patients were able to provide sufficient sputum for all time-points or all analytes. Data presented as mean±sd, unless otherwise indicated. abs.: absolute; CFB: change from baseline; MPO: myeloperoxidase; NE: neutrophil elastase. #: n=39; ¶: n=18; +: mean of Screening and Day 1 pre-dose, or single result if sputum sample only available from one of the two visits; §: mean of Week 10 and 12, or single result if sputum sample only available from one of the two visits; ƒ: two-sided Welch's test with significance level of 5%. TABLE 4Summary statistics of sputum inflammatory markers not affected by EP395 (Full Analysis Set)EP395#, geometric meanPlacebo¶, geometric meanp-valueƒ (baseline to end of trial)Baseline+Week 6End of trial§Baseline+Week 6End of trial§ IL-6 (pg·mL−1) 96.471.568.010677.898.10.511 IL-8 (pg·mL−1) 2150279023652821226724300.116 IL-1β (pg·mL−1) 19.420.214.214.311.014.60.501 IL-25 (pg·mL−1) 3.703.653.914.125.104.360.643 IL-33 (pg·mL−1) 4.814.994.955.175.465.120.240 MIP-1α (pg·mL−1) 9.209.6310.515.013.611.80.068 MIP-1β (pg·mL−1) 85.684.580.310583.81040.879 MCP-1 (pg·mL−1) 77.555.455.594.272.375.20.420 SP-D (ng·mL−1) 2.631.832.627.456.865.050.478 MMP-9 (ng·mL−1) 7407847146596167290.636Of note, not all patients were able to provide sufficient sputum for all time-points or all analytes. Data for IL-23, tumour necrosis factor-α, granulocyte–macrophage colony-stimulating factor and CXC motif chemokine ligand-1 are not presented because >50% of values were below or above the lower/upper limit of quantification for the assay. IL: interleukin; MCP-1: monocyte chemotactic protein-1; MIP: macrophage inflammatory protein; MMP-9: matrix metalloproteinase-9; SP-D: surfactant protein-D. #: n=39; ¶: n=18; +: mean of Screening and Day 1 pre-dose, or single result if sputum sample only available from one of the two visits; §: mean of Week 10 and 12, or single result if sputum sample only available from one of the two visits; ƒ: two-sided Welch's test with significance level of 5%. Summary statistics of sputum inflammatory markers not affected by EP395 (Full Analysis Set) Of note, not all patients were able to provide sufficient sputum for all time-points or all analytes. Data for IL-23, tumour necrosis factor-α, granulocyte–macrophage colony-stimulating factor and CXC motif chemokine ligand-1 are not presented because >50% of values were below or above the lower/upper limit of quantification for the assay. IL: interleukin; MCP-1: monocyte chemotactic protein-1; MIP: macrophage inflammatory protein; MMP-9: matrix metalloproteinase-9; SP-D: surfactant protein-D. #: n=39; ¶: n=18; +: mean of Screening and Day 1 pre-dose, or single result if sputum sample only available from one of the two visits; §: mean of Week 10 and 12, or single result if sputum sample only available from one of the two visits; ƒ: two-sided Welch's test with significance level of 5%. Systemic inflammation Blood biomarkers measured were IL-6, CRP, TNF-α, fibrinogen and α2 macroglobulin (tables 5 and 6). There was a tendency for a reduction in blood IL-6 with EP395 (p=0.105). TABLE 5Full summary statistics of blood interleukin-6 (pg·mL−1) (Full Analysis Set)EP395#Placebo¶ Baseline + Geometric mean1.371.53 Log scale0.318±0.7370.423±1.121 Week 6 Geometric mean1.351.52 Log scale0.297±0.9460.417±1.175 Log scale abs. CFB−0.039±0.562−0.006±0.672 Week 12 Geometric mean1.151.71 Log scale0.141±0.8060.537±1.080 Log scale abs. CFB−0.179±0.4630.113±0.666Treatment difference, baseline to Week 12 Log scale (95% CI)−0.292 (−0.650 to 0.066) p-value§0.105Data presented as mean±sd, unless otherwise indicated. abs: absolute; CFB: change from baseline. #: n=39; ¶: n=18; +: baseline is Day 1, pre-dose; §: two-sided Welch's test with significance level of 5%. Full summary statistics of blood interleukin-6 (pg·mL−1) (Full Analysis Set) Data presented as mean±sd, unless otherwise indicated. abs: absolute; CFB: change from baseline. #: n=39; ¶: n=18; +: baseline is Day 1, pre-dose; §: two-sided Welch's test with significance level of 5%. TABLE 6Summary statistics of blood inflammatory markers not affected by EP395 (Full Analysis Set)EP395#, geometric meanPlacebo¶, geometric meanp-value§ (Baseline to Week 12)Baseline+Week 6Week 12Baseline+Week 6Week 12 Fibrinogen (g·L−1) 3.703.593.523.863.973.960.091 CRP (mg·L−1) 2.782.292.252.692.222.440.725 TNF-α (pg·mL−1) 2.052.052.101.831.901.890.769 α2 macroglobulin (g·L−1) 1.821.841.881.991.971.960.112CRP: C-reactive protein; TNF-α: tumour necrosis factor-α. #: n=39; ¶: n=18; +: baseline is Day 1, pre-dose; §: two-sided Welch's test with significance level of 5%. Summary statistics of blood inflammatory markers not affected by EP395 (Full Analysis Set) CRP: C-reactive protein; TNF-α: tumour necrosis factor-α. #: n=39; ¶: n=18; +: baseline is Day 1, pre-dose; §: two-sided Welch's test with significance level of 5%. Lung function Spirometry was monitored at intervals before, during and at the end of treatment. There was no evidence of an effect of EP395 on FEV1: mean FEV1 at baseline was 1.44 L in both groups, and at Week 12 was 1.42 L in the EP395 group and 1.52 L in the placebo group (supplement 1: table S2). Health-related quality of life and symptoms There was no evidence of an effect of EP395 treatment on E-RS daily scores (supplement 1: table S3); however, for the SGRQ, at Week 12, 44.7% of EP395 patients had a ≥4-point improvement from baseline in total score (the MCID), compared with 16.7% on placebo (table 7). No statistical analysis was performed due to the exploratory nature of the assessment. Data from the EP395 patients were reviewed to assess a potential relationship between a ≥4-point improvement from baseline in SGRQ total score with a reduction in sputum NE; however, in patients with a reduction in NE, a similar number of patients had an improvement in SGRQ versus no improvement in SGRQ (supplement 1: table S4). TABLE 7St George's Respiratory Questionnaire total score (Full Analysis Set)Absolute change from baseline#Improvement from baseline# of ≥4 pointsEP395 375 mg¶Placebo+EP395 375 mg¶Placebo+ Week 4 −0.70±7.981.44±8.77No26 (66.7)14 (82.4)Yes13 (33.3)3 (17.6) Week 8 −2.67±10.972.44±10.50No22 (56.4)15 (83.3)Yes17 (43.6)3 (16.7) Week 12 −4.12±11.13−0.08±9.00No21 (55.3)15 (83.3)Yes17 (44.7)3 (16.7)Data are presented as mean±sd or n (%). #: baseline is the last non-missing assessment before first dose; ¶: n=39; +: n=18. St George's Respiratory Questionnaire total score (Full Analysis Set) Data are presented as mean±sd or n (%). #: baseline is the last non-missing assessment before first dose; ¶: n=39; +: n=18. EP395 plasma concentration Most patients reported 100% compliance via eDiary (range 91.8–100%), which was supported by EP395 plasma concentrations. Mean trough plasma concentrations showed that EP395 steady state was achieved by Day 14, consistent with the estimated terminal t1/2 of EP395 of approximately 75 h. Estimates for EP395 maximum concentration and time to maximum concentration were comparable at Week 6 and 12. Phylogenetic analysis of the lung microbiome An exploratory 16S rRNA analysis of sputum samples at baseline and post-treatment assessed the impact of EP395 on the lung microbiome. No significant changes were observed in ɑ and β diversity between time-points (baseline versus post-treatment) for both treatment groups (figure S1). The microbial phylogenetic composition of the sputum samples was assessed (figure 2). There were no significant differences between the baseline and post-treatment samples at either the phylum or genus level within treatment groups (table 8 and figure S2). TABLE 8Statistical analysis of major phyla present in sputum samples by time-point and treatmentEP395 375 mg#Placebo#Baseline¶Week 12p-valueBaseline¶Week 12p-value Actinobacteriota 10.06 (7.95)13.48 (22.87)0.3117.01 (14.28)18.78 (14.84)0.52 Bacteroidota 39.03 (39.88)9.94 (37.78)0.441.33 (12.83)5.89 (10)0.58 Campylobacterota 0.18 (0.51)0.09 (0.52)0.590.04 (0.09)0.29 (0.66)0.08 Desulfobacterota 0 (0.02)0 (0)0.370 (0.01)0 (0)0.36 Fibrobacterota 0 (0)0 (0)–0 (0)0 (0)– Firmicutes 11.03 (11.08)22.58 (22.14)0.4440.53 (26.74)50.38 (24.28)0.46 Fusobacteriota 1.81 (2.41)2.82 (4.45)1.000.22 (2.62)0.32 (4.8)0.70 Patescibacteria 0.04 (0.04)0.14 (0.21)0.560.02 (0.2)0.1 (0.75)0.12 Proteobacteria 21.25 (16.91)20.49 (19.26)0.8415.45 (23.89)16.05 (31.58)0.46 Spirochaetota 0.01 (0.08)0.12 (0.3)0.860 (0)0 (0.1)0.67 Synergistota 0.01 (0.11)0.03 (0.09)1.000 (0)0 (0)0.59 Verrucomicrobiota 0 (0)0 (0)–0 (0)0 (0)–Data presented as aggregated percentage (median (interquartile range)) of each of the major phylum present in each sample. Values were analysed by Wilcoxon signed rank test to assess the difference between Baseline and Week 12 for each treatment group. There were no significant differences in phylum-level proportions observed between Baseline and Week 12. #: n=11 patients in the EP395 group and n=6 in placebo with sufficient sputum available for microbial analysis (after sputum plug for biomarker analysis had been removed) were included in the analysis; ¶: mean of Screening and Day 1 pre-dose, or single result if sputum sample only available from one of the two visits. Statistical analysis of major phyla present in sputum samples by time-point and treatment Data presented as aggregated percentage (median (interquartile range)) of each of the major phylum present in each sample. Values were analysed by Wilcoxon signed rank test to assess the difference between Baseline and Week 12 for each treatment group. There were no significant differences in phylum-level proportions observed between Baseline and Week 12. #: n=11 patients in the EP395 group and n=6 in placebo with sufficient sputum available for microbial analysis (after sputum plug for biomarker analysis had been removed) were included in the analysis; ¶: mean of Screening and Day 1 pre-dose, or single result if sputum sample only available from one of the two visits. Microbial phylogenetic abundance of sputum samples by time-point and treatment. Data shown are median percentages. Taxonomic summaries at the phylum level were generated by aggregating amplicon sequence variant data using the taxglom function in the phyloseq package (v1.46.0, https://www.bioconductor.org/packages/release/bioc/html/phyloseq.html). Proportional abundances of each phylum were calculated for each group and time-point (Baseline and Week 12) and visualised as stacked bar plots using ggplot2 (v3.5.1, https://ggplot2.tidyverse.org/). A subset of patients with sputum available for microbial analysis were included in the analysis. a) Placebo; b) EP395. Discussion This randomised, double-blind, placebo-controlled trial assessed the safety and pharmacodynamics of 3 months of treatment with the novel oral macrolide EP395 (glasmacinal) in patients with stable COPD. EP395 (375 mg) was well tolerated and showed selective anti-inflammatory effects, significantly reducing sputum total NE, and reducing sputum MPO and blood IL-6, although not significantly. An exploratory assessment indicated that EP395 had no impact on the sputum microbiome. AE incidence was similar between groups; most AEs were mild or moderate, and no treatment-related serious AEs were reported. AEs that occurred only on EP395 were skin-related (rash, pruritus and hyperhidrosis), and two cases of hepatic enzyme elevation (neither was associated with bilirubin change or reached the protocol withdrawal criteria). Larger trials are needed to further evaluate these potential side-effects. The incidence of gastrointestinal AEs was similar between groups, which is important in the context of other macrolides, for which gastrointestinal disturbances are commonly reported side-effects. Similarly, no auditory related AEs were reported, and hearing disturbances (including hearing loss, deafness and/or tinnitus) have been associated with azithromycin [10]. Emphasis was placed on the analysis of QT data, because QT prolongation has been reported following azithromycin treatment [10]. There was a potential increase of approximately 5 ms in mean change from baseline QTcF compared with placebo; however, it was inconsistent across time-points and showed no correlation to EP395 plasma concentration, so is of unclear clinical significance. ECG monitoring will continue in future EP395 trials. The impact of EP395 on lung function (FEV1) and health-related quality of life and symptom measures (monthly SGRQ and daily E-RS) was assessed. EP395 had no impact on FEV1, which was expected based on other trials of macrolides [16]. In a trial of this size and duration, detecting an impact of EP395 on health-related quality of life or symptoms was not anticipated. However, while there was no impact on E-RS scores, 44.7% of patients on EP395 had an improvement from baseline in SGRQ total score of ≥4 points compared with 16.7% of patients on placebo. Because this was a preliminary, under-powered assessment of the effect of EP395, no statistical analysis was conducted; however, this is considered an interesting preliminary finding, which warrants further investigation in fully powered, longer-term trials. EP395 showed selective anti-inflammatory effects but did not affect sputum total or differential cell counts, including neutrophils. It reduced sputum total NE, sputum MPO and blood IL-6 at Weeks 6 and 12, with only the NE reduction reaching statistical significance. There was a strong intra-individual relationship between the changes in sputum NE and MPO (repeated measures correlation analysis). Overall, these findings align with preclinical data showing that pretreatment with EP395 inhibits neutrophilic lung inflammation in mice challenged with respiratory syncytial virus, lipopolysaccharide (LPS) and tobacco smoke [13], and are consistent with the modest anti-inflammatory effects reported for other macrolides in patients with COPD and asthma [17–20]. In a systematic review of clinical immunomodulatory properties of macrolides, a decrease in NE and IL-6 were among the most frequently reported outcomes [21]. The mechanisms underlying the non-antimicrobial immunomodulatory properties of macrolides, and how they translate to reducing COPD exacerbations, are not fully elucidated. Azithromycin, the most studied macrolide, exhibits biphasic and time-dependent effects, initially enhancing innate immune defence responses, followed within hours (or days) by inhibition and then inflammation resolution, dependent on cellular activation and phenotype status [11, 22, 23]. In this trial, EP395 showed selective anti-inflammatory effects, indicating inhibition of neutrophil activation, in patients with stable COPD with chronic inflammation. In another recent EP395 trial [24], 3 weeks of EP395 pretreatment of healthy participants enhanced the innate immune defence response to inhaled LPS, including the epithelial response (demonstrated by an increase in blood SP-D and bronchoalveolar lavage fluid IL-33) and neutrophil activation (increase in NE, MPO and IL-8). Similarly, in the AZIMUNE trial, 12 weeks of treatment with azithromycin in patients with severe asthma increased antiviral defence markers (interferon (IFN)-β and IFN-λ) after ex vivo rhinovirus challenge of bronchial cells [25]. Collectively, these findings suggest that EP395 elicits a biphasic response, like that of azithromycin, with acute enhancing effects on immune defence and selective anti-inflammatory effects in patients with non-exacerbating COPD with chronic inflammation. Both EP395 and antibiotic macrolides appear to have a broader immunomodulatory effect upon challenge (such as viral or LPS challenge), compared with the effect on chronic stable inflammation, which may be related to changes in cellular sensitivity during chronic inflammation. The relative importance of these two phases of macrolide activity in the reduction of exacerbations is unknown. EP395 is a novel macrolide being developed to reduce exacerbations of COPD without the risk of developing antimicrobial resistance or adversely affecting the microbiome. In vitro, minimal inhibitory concentration testing shows that EP395 has negligible antimicrobial activity when tested against 117 isolates from 17 bacterial species [14]. Consistently, 16S rRNA sequencing in this trial found no significant changes in sputum microbiome composition after 12 weeks of EP395 in patients with COPD, including key phyla targeted by azithromycin (Proteobacteria, Firmicutes, Actinobacteria). Overall, the results support the in vitro findings that EP395 has negligible antimicrobial activity, but microbial testing will be continued in further trials with EP395. Given that EP395 is being developed as a treatment for reducing COPD exacerbations, the target population is patients with frequent exacerbations. A limitation of this trial is that it was conducted in patients with stable COPD, without a recent history of exacerbation and stable on maintenance therapy. That patient population was selected because exacerbations and changes in maintenance therapy affect inflammatory markers in both blood and sputum, which would have confounded the interpretation of biomarker data in this small trial. The magnitude of anti-inflammatory effects observed in this stable population was modest, but it is predicted that EP395 will have a greater effect on sputum neutrophil activity in a frequent exacerbator population. Increased sputum MPO is reported to be associated with increased risk of exacerbation [26]. However, the primary objective of this trial was safety, and the selected population was representative in terms of age and concomitant medication use, and appropriate for an initial safety evaluation in a COPD population. Conclusions In patients with stable COPD, 12 weeks of EP395 (glasmacinal) was well tolerated and demonstrated selective anti-inflammatory effects, notably a reduction in markers of neutrophilic inflammation. Exploratory analysis showed that EP395 had no detectable effect on the sputum microbiome of patients with COPD, supporting the in vitro data that EP395 has minimal antimicrobial activity. Overall, this trial indicates that EP395 retains azithromycin's selective anti-inflammatory effects, but without the potential for antibiotic resistance development. These data support the continued development of EP395 as a potential treatment for reducing exacerbations of COPD.