UBC Faculty Research and Publications

Azithromycin and risk of COPD exacerbations in patients with and without Helicobacter pylori Ra, Seung W; Sze, Marc A; Lee, Eun C; Tam, Sheena; Oh, Yeni; Fishbane, Nick; Criner, Gerard J; Woodruff, Prescott G; Lazarus, Stephen C; Albert, Richard; Connett, John E; Han, Meilan K; Martinez, Fernando J; Aaron, Shawn D; Reed, Robert M; Man, S. F P; Sin, Don D May 30, 2017

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata


52383-12931_2017_Article_594.pdf [ 837.6kB ]
JSON: 52383-1.0361804.json
JSON-LD: 52383-1.0361804-ld.json
RDF/XML (Pretty): 52383-1.0361804-rdf.xml
RDF/JSON: 52383-1.0361804-rdf.json
Turtle: 52383-1.0361804-turtle.txt
N-Triples: 52383-1.0361804-rdf-ntriples.txt
Original Record: 52383-1.0361804-source.json
Full Text

Full Text

RESEARCH Open AccessAzithromycin and risk of COPDexacerbations in patients with and withoutHelicobacter pyloriSeung Won Ra1,2, Marc A. Sze3, Eun Chong Lee1, Sheena Tam1, Yeni Oh1, Nick Fishbane1, Gerard J. Criner4,Prescott G. Woodruff5, Stephen C. Lazarus5, Richard Albert6, John E. Connett7, Meilan K. Han8,Fernando J. Martinez9, Shawn D. Aaron10, Robert M. Reed11, S. F. Paul Man1, Don D. Sin1* and on behalf of theCanadian Respiratory Research NetworkAbstractBackground: Helicobacter pylori (HP) infection is associated with reduced lung function and systemic inflammationin chronic obstructive pulmonary disease (COPD) patients. Azithromycin (AZ) is active against HP and reduces therisk of COPD exacerbation. We determined whether HP infection status modifies the effects of AZ in COPD patients.Methods: Plasma samples from 1018 subjects with COPD who participated in the Macrolide Azithromycin (MACRO)in COPD Study were used to determine the HP infection status at baseline and 12 months of follow-up using aserologic assay. Based on HP infection status and randomization to either AZ or placebo (PL), the subjects weredivided into 4 groups: HP+/AZ, HP-/AZ, HP+/PL, and HP-/PL. Time to first exacerbation was compared across the 4groups using Kaplan-Meier survival analysis and a Cox proportional hazards model. The rates of exacerbation werecompared using both the Kruskal-Wallis test and negative binomial analysis. Blood biomarkers at enrolment and atfollow-up visits 3, 12, and 13 (1 month after treatment was stopped) months were measured.Results: One hundred eighty one (17.8%) patients were seropositive to HP. Non-Caucasian participants were nearlythree times more likely to be HP seropositive than Caucasian participants (37.4% vs 13.6%; p < 0.001). The mediantime to first exacerbation was significantly different across the four groups (p = 0.001) with the longest time in theHP+/AZ group (11.2 months, 95% CI; 8.4–12.5+) followed by the HP-/AZ group (8.0 months, 95% CI; 6.7–9.7). Hazardratio (HR) for exacerbations was lowest in the HP+/AZ group after adjustment for age, sex, smoking status, ethnicity,history of peptic ulcer, dyspnea, previous hospital admission, GOLD grade of severity, and forced vital capacity(HR, 0.612; 95% CI, 0.442–0.846 vs HR, 0.789; 95% CI, 0.663–0.938 in the HP-/AZ group). Circulating levels of solubletumor necrosis factor receptor-75 were reduced only in the HP+/AZ group after 3 months of AZ treatment(−0.87 ± 0.31 μg/L; p = 0.002); levels returned to baseline after discontinuing AZ.Conclusions: AZ is effective in preventing COPD exacerbations in patients with HP seropositivity, possibly bymodulating TNF pathways related to HP infection.Keywords: Helicobacter pylori, COPD, Exacerbation, Azithromycin* Correspondence: Don.Sin@hli.ubc.ca1Centre for Heart Lung Innovation, St. Paul’s Hospital, & Department ofMedicine (Respiratory Division), University of British Columbia, Don D Sin,Room 8446-1081 Burrard Street, Vancouver, BC V6Z 1Y6, CanadaFull list of author information is available at the end of the article© The Author(s). 2017 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, andreproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link tothe Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver(http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.Ra et al. Respiratory Research  (2017) 18:109 DOI 10.1186/s12931-017-0594-xBackgroundPersistent systemic inflammation affects ~15 to 20% ofpatients with chronic obstructive pulmonary disease(COPD), which in turn is associated with an increasedrisk of exacerbations and mortality [1]. The etiology forthis persistent inflammation in COPD, however, islargely a mystery [2]. There is growing evidence that thegastrointestinal tract is a major modulator and driver ofinflammation and that the gut-lung axis may be per-turbed in COPD [3, 4]. Previous studies have reportedan increased prevalence of Helicobacter pylori (HP) in-fection in patients with COPD [5]. Using serum samplesfrom the Lung Health Study (LHS), we recently showedthat ~18% of patients in that cohort had serologic evi-dence for HP infection, which was associated with sys-temic inflammation and reduced lung function [6]. HPinfection may promote persistent low-grade inflamma-tion by up-regulating antigenic stimulation in mucosalsurfaces and by skewing the lymphocyte response to-wards a T helper (Th) lymphocyte 1 bias [7–9].Interestingly, azithromycin (AZ), which is being in-creasingly used to prevent exacerbations in COPD hasbactericidal activity against HP both in vitro and in vivo[10, 11]. In general, the immunomodulatory effects ofAZ preferentially attenuate Th1 (rather than Th2) re-sponses [12, 13] and decrease tumor necrosis factor(TNF)-α production by human monocytes [14]. How-ever, whether HP modifies the effectiveness of AZ inpatients with COPD is unknown. Using data from theMACRO (MACROlide azithromycin to prevent COPDexacerbations) Study [15], we determined the impact ofHP infection status on the beneficial effects of AZ inpreventing exacerbations in patients with COPD.MethodsDetails of the MACRO Study design and results have beenpublished previously [15]. The study cohort consisted of1142 subjects with COPD who were randomized either toazithromycin (AZ, 250 mg) or a placebo tablet (PL) takendaily for 12 months in addition to usual care. The primaryoutcome of interest was the time to first exacerbation ofCOPD, defined as a complex of respiratory symptoms (in-creased or new onset) consisting of two or more of the fol-lowing: cough, sputum, wheezing, dyspnea, or chesttightness with a duration of at least three days and requir-ing treatment with antibiotics or systemic steroids or bothin combination [15]. At the time of study entry, all sub-jects had to have been free of an acute exacerbation ofCOPD (AECOPD) for at least 4 weeks prior torandomization. Subjects were monitored for AECOPDs atclinic visits, which occurred at 3-month intervals, and bymonthly telephone contact between each clinic visit.For the current study, we used blood samples that wereavailable from 1018 subjects to determine the prevalenceof HP seropositivity (HP+ or HP-). Plasma samples, col-lected at enrollment and at the 12 month visit, were usedto measure concentrations of immunoglobulin G (IgG) toHP cytotoxin-associated gene A (CagA) antigen using acommercially available ELISA kit (DRG Diagnostics,GmbH, Marburg, Germany) according to the manufac-turer’s protocol. Samples with values greater than 18 DU/mL were regarded as positive for HP infection (HP+).Values lower than 18 DU/mL were considered HP nega-tive (HP-) [6]. Using this cutoff value, seroconverters weredefined as patients who were seronegative at enrollmentbut became seropositive at the 12 month visit. Serorever-ters were defined as subjects who were seropositive at en-rolment but became seronegative at the 12 month visit. Asignificant reduction in HP antibody was defined as a≥50% decrease in IgG antibody levels over 12 months [16,17]. Fifteen patients failed to come in for follow-up visitsand were excluded from the remaining analyses. Thus,1003 subjects were used to evaluate the time to first ex-acerbation and rate of exacerbation during the 1 year.These 1003 patients were further divided into 4 groupsbased on their initial HP status and randomization to AZor PL treatment (HP+/AZ, HP-/AZ, HP+/PL, HP-/PL).To determine whether AZ and HP status modifiedcommon biomarkers of systemic inflammation, we relatedthese parameters to plasma levels of C-reactive protein(CRP) and soluble tumor necrosis factor receptor-75(sTNFR75), which had been previously measured at base-line and at the 3, 12 and 13 (1 month after treatment dis-continuation) month visits in the MACRO participants.The details of these measurements have been previouslyreported [18]. The MACRO study received approval fromeach participating institution’s local research ethics boardand the present study received approval from the Univer-sity of British Columbia/Providence Health Care ResearchEthics Committee (Approval No. H11-00786).Statistical analysesExacerbation-free survival was assessed using Kaplan-Meier curves stratified by HP serostatus and AZ use, andthe groups were compared by a log-rank test. Bootstrapmethods were used to generate the confidence intervals(CIs) for median time to the first exacerbation [19]. Pair-wise comparisons were performed to determine the differ-ences in survival rate between each pair of groups using ageneralized Wilcoxon test (Breslow test). A Cox propor-tional hazards regression was performed to adjust forpotential confounders including age, sex, smoking status,ethnicity, history of peptic ulcer, dyspnea, previoushospital admission for COPD during past year, GlobalInitiative for Chronic Obstructive Lung Disease (GOLD)grade of severity of airflow limitation, and forced vitalcapacity (FVC) on these relationships using HP-/PL as thereference group. The rates of exacerbation were determinedRa et al. Respiratory Research  (2017) 18:109 Page 2 of 9by dividing the number of acute exacerbations by theperson-years of follow-up, and were compared using boththe Kruskal-Wallis test and negative binomial analysis.We used paired t-tests to compare HP antibody levels atbaseline vs. the 12 month follow-up in each of the 4groups. Plasma concentrations of CRP and sTNFR75levels were log-transformed and paired t-tests were usedto determine the differences in concentrations betweensamples collected at enrollment and at the 3, 12, and13 month visits within the 4 groups. All statistical analyseswere performed using SPSS version 21.0 for Windows(IBM Corp., Armonk, NY, USA) and R version 3.2.2(available from https://www.r-project.org).ResultsCharacteristics of the study subjects based on HP statusPlasma samples from 1018 subjects at enrollment wereanalyzed to determine the prevalence of HP seropositiv-ity; 181 patients (17.8%) had positive titres. HP sero-prevalence was substantially lower among Caucasianthan non-Caucasian subjects (13.6% vs 37.4%; p < 0.001),even though the history of peptic ulcer was similarbetween the two groups (13.8% vs 13.7%; p = 0.985 byChi-square test). After excluding 15 patients withoutfollow-up visits, the baseline characteristics of theremaining 1003 patients according to HP status aresummarized in Table 1. In the HP seropositive group,the proportion of a peptic ulcer history was higher andFVC (%) was lower as compared to HP seronegativegroup. The patient characteristics stratified by HP statusand treatment (HP+/AZ, HP-/AZ, HP+/PL, or HP-/PL)can be found in Additional file 1: Table S1. There wereno significant differences in age, sex, smoking status,pack years of smoking, dyspnea grade, hospital admis-sion for COPD or the use of systemic steroids and/orantibiotics past year, lung function, GOLD grade ofseverity of airflow limitation, or baseline levels ofbiomarkers (CRP and sTNFR75) between the 4 groups.The proportion of a peptic ulcer history was higher inHP seropositive groups (HP+/AZ and HP+/PL) than HPseronegative groups (HP-/AZ and HP-/PL).Time to first exacerbation in relation to HP status andazithromycin useThe median time to first exacerbation was the longest inthe HP+/AZ group (11.2 months, 95% CI; 8.4–12.5+),followed by the HP-/AZ group (8.0 months, 95% CI; 6.7–9.7), the HP+/PL group (7.5 months, 95% CI; 4.9–8.8) andTable 1 Characteristics of Helicobacter pylori positive and negative patientsH. pylori positive, n = 179 H. pylori negative, n = 824 p-value*Age, years 66.2 ± 8.5 65.3 ± 8.7 0.18Male sex 118 (65.9) 485 (58.9) 0.08Current smoker 36 (20.1) 175 (21.3) 0.73Smoking history - Pack years 58.5 ± 32.5 58.3 ± 31.8 0.94Ethnicity (Caucasian) 112 (62.6) 716 (86.9) <0.001Peptic ulcer history 36 (20.1) 102 (12.4) 0.006Dyspnea (MRC grade) 1.55 ± 0.89 1.53 ± 0.91 0.78Hospitalization for COPD past year 80 (44.7) 423 (51.3) 0.11FEV1, Liter 1.1 ± 0.5 1.1 ± 0.5 0.25FEV1, % predicted 38.5 ± 15.1 39.9 ± 15.7 0.27FVC, Liter 2.5 ± 0.8 2.7 ± 0.9 0.07FVC, % predicted 67.8 ± 16.4 70.7 ± 18.3 0.049FEV1/FVC % 42.5 ± 12.2 42.6 ± 12.8 0.97GOLD grade, n (%) 0.65II 42 (23.5) 216 (26.3)III 75 (41.9) 344 (42.0)IV 62 (34.6) 260 (31.7)Baseline biomarker dataCRP (mg/L) 4.83 ± 3.89 4.91 ± 3.87 0.80sTNFR75 (μg/L) 8.79 ± 4.41 8.71 ± 4.87 0.85Data are presented as means ± standard deviation or absolute number (%)MRC medical research council, FEV1 forced expiratory volume in one second, FVC forced vital capacity, GOLD Global Initiative for Chronic Obstructive Lung Disease,CRP C-reactive protein, sTNFR75 soluble tumor necrosis factor receptor-75*Unpaired t-test or Chi-square testThe results of FEV1, FVC, and FEV1/FVC are post-bronchodilator valuesRa et al. Respiratory Research  (2017) 18:109 Page 3 of 9the HP-/PL group (5.7 months, 95% CI; 4.5–7.2) with asignificant difference across these 4 groups by a log-ranktest (p = 0.001; Fig. 1). Pair-wise comparisons were per-formed across the 4 groups using a generalized Wilcoxontest (Breslow test), which showed significant differences insurvival function for the time to first exacerbation betweenthe HP+/AZ group and the HP-/AZ group (p = 0.04); theHP+/PL group (p = 0.006) and the HP-/PL group (p =0.001); and between the HP-/AZ and the HP-/PL groups(p = 0.02). A Cox proportional hazards regression was per-formed to investigate the relationship between the 4 groupsin regards to the time to first exacerbation. Hazard ratio(HR) for the time to first exacerbation was the lowest in theHP+/AZ group after adjusting for age, sex, smoking status,ethnicity, history of peptic ulcer, dyspnea, previous hospitaladmission for COPD during past year, GOLD grade of se-verity of airflow limitation, and FVC (HR, 0.612; 95% CI,0.442–0.846; p = 0.003 vs HR, 0.789; 95% CI, 0.663–0.938;p = 0.007 in the HP-/AZ group vs HR, 1.096; 95% CI,0.832–1.442; p = 0.52 in the HP+/PL group vs HR, 1 in thereference group [HP-/PL]). The rates of exacerbationper patient-year were 1.21, 1.54, 1.73, and 1.85, respect-ively (p = 0.007 by Kruskal-Wallis test; Additional file 1:Table S2). The HP+/AZ group had a 31% lower exacer-bation rate compared to the HP-/PL group (p = 0.02),whereas the HP-/AZ group had a 14% lower exacerba-tion compared to the HP-/PL group (p = 0.08).Change in HP status and antibody levels over 1 yearWe also measured HP antibodies in plasma collected at the12 month visit in 643 subjects. The HP seroprevalencechanged from 17.8% (181/1018) to 16.6% (107/643) after1 year.Seroconversion rates and changes in antibodyconcentrationsOf the 530 COPD patients at risk of acquiring HP infec-tion (i.e. seronegative at the baseline plasma measure-ment), 21 (8 from the HP-/AZ group and 13 from theHP-/PL group) became seropositive, resulting in an an-nual seroconversion rate of 4.0% (Table 2). Overall, therewas a significant increase in HP antibody titres over1 year of follow-up among participants in the HP-/PLgroup (3.21 ± 1.56 DU/ml; mean ± standard error ofmean; p = 0.041; Fig. 2). In contrast, the HP antibodytitres did not change significantly during follow-up inthe HP-/AZ group (1.38 ± 1.07 DU/ml ; p = 0.198).Seroreversion rates and changes in antibody concentrationsThe seroreversion rate was not statistically different be-tween AZ and PL use group (20.4% vs 27.1%; p = 0.401by Chi-square test; Table 2). Of the 113 COPD patientswho were IgG seropositive at study entry, 27 (11 fromthe HP+/AZ group, 16 from the HP+/PL group) becameseronegative, resulting in an annual seroreversion rate ofFig. 1 Proportion of participants free from acute exacerbations of chronic obstructive pulmonary disease (COPD). AZ, Azithromycin; PL, Placebo;HP, Helicobacter pylori. Pair-wise comparisons between each group showed significant differences in the time to first exacerbation between: TheHP+/AZ group versus HP-/AZ group (11.2 months, 95% CI; 8.4–12.5+ vs 8.0 months, 95% CI; 6.7–9.7; p = 0.040); The HP+/AZ group versus the HP+/PL group (7.5 months, 95% CI; 4.9–8.8; p = 0.006); and The HP+/AZ group versus the HP-/PL group (5.7 months, 95% CI; 4.5–7.2; p = 0.001). Therewas a significant difference between the HP-/AZ and the HP-/PL groups (p = 0.020). The remaining pair-wise comparisons were notstatistically significantRa et al. Respiratory Research  (2017) 18:109 Page 4 of 923.8% (Table 2); 25 patients (22.1%) demonstrated atleast a 50% reduction in their baseline IgG antibodytitres at 12 months (Additional file 1: Table S3). In theHP+/AZ group, HP antibody titres decreased signifi-cantly during follow-up (−35.42 ± 9.41 DU/ml; p < 0.001;Fig. 2). Patients in the HP+/PL group also experienced asignificant decline in HP titres during this time (−45.91± 11.83 DU/ml; p < 0.001).Serial change in CRP and sTNFR75 in relation to baselineHP status and treatmentPlasma CRP and sTNFR75 levels were available for 1001of the 1003 subjects (99.8%) at baseline and 859 (85.6%),719 (71.7%), and 697 (69.5%) subjects at 3, 12, and13 months of follow-up, respectively. We analyzed themean (± SEM) changes in the log-transformed CRP orsTNFR75 levels within subjects from baseline to 3, 12,and 13 months using paired t-tests (Fig. 3). CRP de-creased significantly after 3 months in the HP-/AZgroup (−0.33 ± 0.18 mg/L; p = 0.047) and in the AZ usegroup (−0.34 ± 0.16 mg/L; p = 0.02; Fig. 3 and Additionalfile 1: Table S4). After discontinuing AZ at 12 months,CRP returned to baseline levels at 13 months in thesegroups. In terms of sTNFR75, only the HP+/AZ groupshowed a significant decrease at 3 months (−0.87 ±0.31 μg/L; p = 0.002; Fig. 3, Table 3) and the levelsreturned to baseline levels after discontinuing AZ. Thedifferences in CRP and sTNFR75 concentrations be-tween the other time points and baseline levels were notsignificant.DiscussionThe present study is the first to evaluate the effect of azi-thromycin on COPD exacerbations according to serosta-tus for Helicobacter pylori (HP). The most importantfindings were that approximately 18% of patients withCOPD in the MACRO Study were seropositive for HPand that these individuals experienced the largest reduc-tion in the risk for exacerbation from chronic prophylacticAZ therapy. AZ therapy in these patients was associatedwith a significant decline in plasma sTNFR75 levels, butnot with seroreversion rate as compared to PL use group.Together, these data suggest that AZ therapy alone prob-ably does not eradicate HP infection but may have asignificant immunomodulatory role that mitigates the riskof exacerbations in HP seropositive individuals.It is now well-established and accepted that chronicprophylactic therapy with macrolides such as AZ re-duces the risk of exacerbations in patients with COPD.However, many national and international guidelinesand strategic documents avoid strong endorsement fortheir use because of the concerns about long-term sideeffects, including decreased hearing acuity, arrhythmo-genesis, and the possibility of promoting widespreadantimicrobial resistance [20, 21]. Given these concerns,it would be highly desirable to develop biomarkers thatcan identify patients who would benefit the most fromthese therapies. In a post hoc analysis of the MACROdata, Han and colleagues showed that the benefits of AZwere greatest in ex-smokers, older patients, and subjectswith mild COPD. However, in their study, reductions inexacerbations were still seen in younger patients andthose with higher GOLD grades of airflow limitation,particularly for exacerbations that required treatmentwith antibiotics and steroids [22]. We extend these find-ings by demonstrating that the risk of exacerbation is re-duced by 39% with AZ in patients who were seropositiveto HP (versus only 21% in HP- individuals; p = 0.04).Table 2 HP status at 12 months according to baseline HP statusand treatment in 643 COPD patientsBaseline HP statusand treatmentHP status after 12 months TotalNegative n = 536 Positive n = 107HP-/AZ 269 (97.1%) 8 (2.9%) 277 (100%)HP-/PL 240 (94.9%) 13 (5.1%) 253 (100%)Total (HP-) 509 (96.0%) 21 (4.0%) 530 (100%)HP+/AZ 11 (20.4%) 43 (79.6%) 54 (100%)HP+/PL 16 (27.1%) 43 (72.9%) 59 (100%)Total (HP+) 27 (23.8%) 86 (76.2%) 113 (100%)The percentages in the table were calculated row-wiseHP Helicobacter pylori, AZ azithromycin, PL placeboFig. 2 Individual IgG antibody titres to H. pylori (HP) CagA antigen inplasma of COPD patients at 0 (baseline) and 12 months. AZ,Azithromycin; PL, Placebo. The horizontal lines denote arithmeticmeans for individual groups. The differences in antibodyconcentrations within an individual patient between two time pointswere compared using a paired t-test. During follow-up, there was asignificant decrease in antibody level from baseline (mean ± SEM) inboth the HP+/AZ group (−35.42 ± 9.41; p < 0.001) and the HP+/PLgroup (−45.91 ± 11.83; p < 0.001) over the year. There was nosignificant increase in antibody level from baseline in the HP-/AZ group(1.38 ± 1.07; p = 0.198). In contrast, there was a significant increase inthe HP-/PL group (3.21 ± 1.56; p = 0.041)Ra et al. Respiratory Research  (2017) 18:109 Page 5 of 9Interestingly, using the same cohort, Woodruff et al.found that a decline in sTNFR75 concentrations over3 months identified COPD patients who benefited fromAZ therapy [18]. We extend these findings by demonstrat-ing that the largest (and most significant) reductions incirculating sTNFR75 levels after 3 months were observedin HP+ patients who were treated with AZ (N = 76; −0.87± 0.31 μg/L; p = 0.002), followed by ex-smokers (N = 360;−0.17 ± 0.24 μg/L; p = 0.03; Table 3). However, AZ therapywas also beneficial in HP seronegative individuals(although the impact was less striking than those whowere HP+), suggesting that the mechanisms and pathwaysby which AZ mitigates exacerbations in COPD are diverseand complicated and extend beyond those with HPseropositivity.In contrast to sTNFR75, CRP concentration at 3 monthswas significantly reduced in the AZ group (N = 432; −0.34± 0.16 mg/L; p = 0.02; Fig. 3, Additional file 1: Table S4)and not reduced in the HP+/AZ group (N = 76; −0.40 ±0.38 mg/L; p = 0.17), suggesting that potential mechanismsfor the reduction in CRP levels include an anti-inflammatory effect independent of the presence of HP in-fection. Previously it was shown that there was an increasein CRP level in mild to moderate COPD patients whowere stable and HP positive [6], whereas the present studyshowed no significant difference in CRP or sTNFR75levels between HP negative and positive COPD patients.In addition, there was no significant difference in the timeto first exacerbation or rates of exacerbations between theHP+/PL and HP-/PL groups (Fig. 1). We postulate thatthis may reflect a selection bias towards inclusion of indi-viduals prone to high systemic inflammation and exacer-bation in the MACRO Study. It is likely that theseindividuals had multiple drivers of exacerbations (not justHP infection), which may have obscured the effects of HPinfection on systemic inflammation and the overall risk ofexacerbations in patients who received placebo duringfollow-up. Consistent with this notion, the average CRPFig. 3 Serial change in blood biomarkers according to baseline H. pylori (HP) status and treatment. AZ, Azithromycin; PL, Placebo. The points anderror bars indicate the means and standard errors of change in C-reactive protein (CRP) and soluble tumor necrosis factor-75 (sTNFR75) levelsfrom baseline to 3, 12, and 13 months for each of the 4 groups. Paired t-tests were used to determine differences in biomarker concentrationsbetween the three time points and baseline levels. *p < 0.05, **p < 0.01. a CRP decreased significantly at 3 months in the HP-/AZ and AZ usegroup. After stopping AZ at 12 months, CRP returned to baseline levels at 13 months. b sTNFR75 levels decreased significantly at 3 months onlyin the HP+/AZ group and returned to baseline levels after stopping AZTable 3 Effect of azithromycin on changes in sTNFR75 levels at3 months according to subgroupsSubgroup sTNFR75 level (μg/L) p-value*N Baseline 3 months DifferenceHP + 76 8.65 ± 0.50 7.78 ± 0.45 −0.87 ± 0.31 0.002HP - 356 8.52 ± 0.24 8.63 ± 0.32 0.11 ± 0.25 0.64Total 432 8.54 ± 0.22 8.48 ± 0.27 −0.06 ± 0.21 0.10Ex-smoker 360 8.73 ± 0.24 8.56 ± 0.31 −0.17 ± 0.24 0.03Smoker 89 8.10 ± 0.41 8.11 ± 0.43 0.01 ± 0.30 0.89Age > 65 220 8.52 ± 0.29 8.46 ± 0.42 −0.06 ± 0.36 0.25Age≤ 65 229 8.68 ± 0.31 8.48 ± 0.32 −0.20 ± 0.20 0.08GOLD 2 120 9.07 ± 0.39 9.08 ± 0.67 0.01 ± 0.60 0.15GOLD 3 183 8.68 ± 0.32 8.49 ± 0.34 −0.19 ± 0.22 0.29GOLD 4 145 8.12 ± 0.40 7.86 ± 0.40 −0.26 ± 0.27 0.19Data are presented as means ± standard errors of the meansTNFR75 soluble tumor necrosis factor-75, GOLD Global Initiative for ChronicObstructive Lung Disease, HP Helicobacter pylori*Paired t-tests on log-transformed dataRa et al. Respiratory Research  (2017) 18:109 Page 6 of 9values in the MACRO Study participants were signifi-cantly higher compared to those of previous studies,which had enrolled stable COPD patients [1, 6].Seropositivity to HP, and annual rates of seroconver-sion and reversal, is not well-known in patients withCOPD. Our previous study using serum samples fromthe lung health study (LHS) and ELISA plates coatedwith CagA protein demonstrated a cross-sectionalprevalence of HP seropositivity of 17.6% [6]. However,this study was limited in that serum samples were col-lected in the 1990’s before the widespread availability oftriple therapy for HP eradication. Here, using a morecontemporaneous cohort (MACRO), we report a similarHP seropositivity rate despite advances in diagnosis andavailability of effective therapies for HP. It is also notablethat in the MACRO Study we found a substantiallyhigher rate of HP seropositivity among non-Caucasiansubjects than Caucasian subjects (37.4% vs 13.6%). Con-sidering that approximately 60% of Western HP strainsdemonstrate CagA antigen expression, these data areconsistent with results from the National Health andNutrition Examination Survey (1999–2000) conductedin US adults, which used HP antigen coated ELISAplates (Wampole Laboratories, Princeton, New Jersey)and reported a higher age-standardized seroprevalenceof HP in Mexican Americans (64.0%) and non-Hispanicblacks (52.0%) compared with Caucasians (21.2%) [23].Despite the difference of HP seropositivity betweenCaucasian and non-Caucasian subjects, the proportionwith a peptic ulcer history was similar between twogroups in our study, suggesting that additional studieswould be required to find the mechanism to explain thedisparities among ethnicities who may have differentdietary cultures and diverse intestinal microbiota com-munities. AZ may have differential effect on the intes-tinal microbiota including helicobacter genus. Theoverall annual seroconversion rate in our study subjectswas 4% with rates numerically lower in the AZ groupcompared to the PL group. A 4% annual seroconversionrate in our study subjects is higher than previously re-ported in high-income countries (~0.01–1% per year inthe general adult population) [24–27]. A previous studysuggests that older age and smoking are associated withan increased risk of HP [28], possibly due to immunedysregulation and senescence. The overall annual serore-version rate in our study subjects was 23.8% (27/113)and the rate between the PL and AZ groups was not sta-tistically different. In line with the high rate of serocon-version in our study subjects, seroreversion rate waseven higher than other studies performed in the generaladult population [24, 27, 29]. Kuipers et al. documentedthat HP positivity reverted in only 3 of 56 subjects in theabsence of specific antimicrobial therapy over 11 years[25]. Since the many of our study subjects were usinghome oxygen therapy and some had advanced COPD,they experienced frequent exacerbations leading to re-peated exposures to antibiotics and steroids during thestudy period.Importantly, in the HP-/AZ group, there was no sig-nificant increase in HP antibody titres over the 12 monthperiod; by contrast, there was a significant increase inHP antibody titres in the HP-/PL group, suggesting thatAZ use prevented significant increases in HP antibodylevels in HP- patients. Moreover, although AZ has somein vitro and in vivo activity against HP, we did not find asignificant difference in the rate of seroreversionbetween the PL and AZ groups. We also used ≥50% de-cline in titres at 12 months of follow-up compared tobaseline levels as indirect evidence for HP eradication[16, 17] since absolute antibody titres remain in thepositive range for 1 to 4 years after successful eradica-tion [30, 31]. Interestingly, HP+ patients who experi-enced at least one exacerbation during the 12 monthfollow-up were more likely to demonstrate a ≥50% re-duction in HP titres during follow-up compared to thosewho did not experience any exacerbations (30.1% eradica-tion rate vs 7.5%; p = 0.006; Additional file 1: Table S3).Again, this raises the possibility that antibiotic exposureduring exacerbations may have played a significant role indecreasing antibody levels. Consistent with this hypoth-esis, a previous study showed that children who had beentreated with antibiotics demonstrated lower risk of HPinfection compared with those who had never beentreated with antibiotics (12.5% vs 30%) [32]. Together,these data suggest that although AZ monotherapy prob-ably does not prevent new HP infection or eradicate exist-ing infection, it likely downregulates the host’s immuneresponses [10, 11] and dampens macrophage responsesand Th1 immunity related to TNF pathways [12–14].Limitations of the studyThere were several limitations that should be noted in thecurrent study. First, we do not know how many patientswith an HP infection developed resistance to macrolidesafter 1 year of AZ exposure. HP readily becomes resistantto macrolides when given alone, and short-term AZmonotherapy is not considered sufficient for HP eradica-tion [33–35]. Moreover, we had no available data on thepresence of gastric diseases or symptoms, or signs thatcould indicate active HP infection, though we did deter-mine HP serostatus by using the CagA antigen, which is awell-known virulent strain causing more severe gastrodu-odenal disease [36]. However, to date, AZ monotherapyfor COPD patients with HP seropositivity cannot be advo-cated without further study. For those who are seroposi-tive, HP infection should be further explored and treatedwith the appropriate triple therapy regimen. Second, wedid not have detailed information regarding the use ofRa et al. Respiratory Research  (2017) 18:109 Page 7 of 9prescription or over-the-counter antisecretory drugs,which may be responsible for prevention (or even treat-ment) of COPD exacerbations in patients with gastricsymptoms and could have affected HP antibody levels inour patient population. Third, the use of antibiotics or ste-roids other than AZ could also have affected HP titres inour study subjects. Fourth, the present study used a retro-spective design and did not focus on the potential patho-genic mechanisms underlying the association between HPinfection and COPD exacerbation. It is well known thatHP can stimulate the release of a variety of pro-inflammatory cytokines, including interleukin-1 (IL-1),IL-8 and TNF-α [37, 38] and these cytokines are alsothought to be involved in the pathogenesis of COPD [39,40]. Therefore, HP infection might play a pro-inflammatory role and be a cofactor in the pathogenesis ofCOPD exacerbation. Finally, although we demonstrateddifferential effect of AZ therapy on the risk of exacerba-tion in HP seropositive COPD patients, the mechanism bywhich this occurs is unknown. We posit that AZ sup-presses (but not fully eradicates) HP infection, which inturn, may downregulate the TNF pathway. Experimentalstudies will be required in the future to tease out the fullmechanisms to explain the beneficial effects of AZ on HPinfection in COPD patients.ConclusionsHP infection status was associated with differentialeffects of AZ in COPD patients with frequent exacerba-tions, highlighting the importance of evaluating HP sta-tus in these group. Chronic AZ monotherapy suppressedHP antibody levels after 1 year of treatment, but did notsignificantly modify the rate of seroreversion in the HP+group. The HP+/AZ group exhibited a decline insTNFR75 levels at 3 months. Patients who were HPseropositive experienced a significantly greater protec-tion against exacerbation from chronic AZ therapy thanthose who were HP seronegative. A larger prospectivestudy is needed to definitively prove the benefit of eradi-cation of HP with triple therapy in seropositive HPpatients with COPD, who frequently exacerbate.Additional fileAdditional file 1: Table S1. Patient characteristics according toHelicobacter pylori status and treatment. Table S2. Effect of Helicobacterpylori status at enrollment and treatment on the rates of exacerbationper person-year. Table S3. Comparison of the proportion of subjects thatshowed significant reduction in HP antibody level at 12 months in 113COPD patients. Table S4. Effect of treatment on changes in CRP levelfrom baseline to 3 months according to HP status. (DOCX 27 kb)AbbreviationsAECOPD: Acute exacerbation of COPD; AZ: Azithromycin; BD: Bronchodilator;CagA: Cytotoxin-associated gene A; CI: Confidence interval; COPD: Chronicobstructive pulmonary disease; CRP: C-reactive protein; FEV1: Forced expiratoryvolume in one second; FVC: Forced vital capacity; GOLD: Global Initiative forChronic Obstructive Lung Disease; HP: Helicobacter pylori; HR: Hazard ratio;LHS: Lung Health Study; MACRO: MACROlide azithromycin to prevent COPDexacerbations; PL: Placebo; SEM: Standard error of the mean; sTNFR75: Solubletumor necrosis factor receptor-75; Th: T helper; TNF: Tumor necrosis factorAcknowledgementsWe thank Kim Schmidt for comments on the manuscript.FundingThis study was funded by the Canadian Respiratory Research Network (CRRN)and the Canadian Institutes of Health Research (CIHR). The MACRO Study wasfunded by the US National Heart Blood and Lung Institute (NHLBI). DDS is a Tier1 Canada Research Chair in COPD. SWR was funded by Ulsan UniversityHospital (Biomedical Research Center Promotion Fund, 10–03).Availability of data and materialsPlease contact author for data requests.Authors’ contributionsSWR designed the biomarker experiment, wrote all drafts, and performedstatistical analysis; MAZ contributed to the design of the experiments; ST, YO,and ECL processed the blood samples and performed the assays; NFcontributed significantly to statistical analyses; GC, PW, RA, JC, SCL, MH, FJM,RMR were involved in the MACRO biomarker studies and contributed to thedesign of the experiments; SDA and SFPM were intimately involved in thedesign of the experiments and interpretation of results; DDS is the principalinvestigator of the present study and is the guarantor of the results. Allauthors provided critical intellectual contributions to the present work andparticipated in the revision and finalization of the manuscript. All authorsread and approved the final manuscript.Competing interestsThe authors declare that they have no competing interests.Consent for publicationNot applicable.Ethics approval and consent to participateThe MACRO study received approval from each participating institution’slocal research ethics board and the present study received approval from theUniversity of British Columbia/Providence Health Care Research EthicsCommittee (Approval No. H11-00786). Written informed consent was ob-tained from all patients.Publisher’s NoteSpringer Nature remains neutral with regard to jurisdictional claims inpublished maps and institutional affiliations.Author details1Centre for Heart Lung Innovation, St. Paul’s Hospital, & Department ofMedicine (Respiratory Division), University of British Columbia, Don D Sin,Room 8446-1081 Burrard Street, Vancouver, BC V6Z 1Y6, Canada. 2UlsanUniversity Hospital, University of Ulsan College of Medicine, Ulsan, SouthKorea. 3Department of Microbiology & Immunology, University of Michigan,Ann Arbor, MI, USA. 4Department of Thoracic Medicine and Surgery, TempleUniversity, Philadelphia, PA, USA. 5Department of Medicine, University ofCalifornia San Francisco, San Francisco, CA, USA. 6Pulmonary Sciences andCritical Care Medicine, University of Colorado, Denver, CO, USA. 7School ofPublic Health, University of Minnesota, Minneapolis, MN, USA. 8Departmentof Internal Medicine, University of Michigan, Ann Arbor, MI, USA. 9Joan andSanford I. Weill Department of Medicine, Weill Cornell Medical College,Cornell University, New York, NY, USA. 10Department of Medicine, Universityof Ottawa, Ottawa, ON, Canada. 11Division of Pulmonary and Critical CareMedicine, University of Maryland School of Medicine, Baltimore, MD, USA.Ra et al. Respiratory Research  (2017) 18:109 Page 8 of 9Received: 17 January 2017 Accepted: 19 May 2017References1. Agusti A, Edwards LD, Rennard SI, MacNee W, Tal-Singer R, Miller BE, et al.Persistent systemic inflammation is associated with poor clinical outcomesin COPD: a novel phenotype. PLoS One. 2012;7:e37483.2. Rennard SI, Locantore N, Delafont B, Tal-Singer R, Silverman EK, Vestbo J, etal. Identification of five chronic obstructive pulmonary disease subgroupswith different prognoses in the ECLIPSE cohort using cluster analysis. AnnAm Thorac Soc. 2015;12:303–12.3. Sze MA, Tsuruta M, Yang SW, Oh Y, Man SF, Hogg JC, et al. Changes in thebacterial microbiota in gut, blood, and lungs following acute LPS instillationinto mice lungs. PLoS One. 2014;9:e111228.4. Young RP, Hopkins RJ, Marsland B. The gut-liver-lung axis. Modulation ofthe innate immune response and its possible role in chronic obstructivepulmonary disease. Am J Respir Cell Mol Biol. 2016;54:161–9.5. Wang F, Liu J, Zhang Y, Lei P. Association of Helicobacter pylori infectionwith chronic obstructive pulmonary disease and chronic bronchitis: a meta-analysis of 16 studies. Infect Dis (Lond). 2015;47:597–603.6. Sze MA, Chen YW, Tam S, Tashkin D, Wise RA, Connett JE, et al. Therelationship between Helicobacter pylori seropositivity and COPD. Thorax.2015;70:923–9.7. Lehmann FS, Terracciano L, Carena I, Baeriswyl C, Drewe J, Tornillo L, et al.In situ correlation of cytokine secretion and apoptosis in Helicobacterpylori-associated gastritis. Am J Physiol Gastrointest Liver Physiol.2002;283:G481–8.8. Blaser MJ, Atherton JC. Helicobacter pylori persistence: biology and disease.J Clin Invest. 2004;113:321–33.9. D’Elios MM, Amedei A, Benagiano M, Azzurri A, Del Prete G. Helicobacterpylori, T cells and cytokines: the “dangerous liaisons”. FEMS Immunol MedMicrobiol. 2005;44:113–9.10. Loo VG, Sherman P, Matlow AG. Helicobacter pylori infection in a pediatricpopulation: in vitro susceptibilities to omeprazole and eight antimicrobialagents. Antimicrob Agents Chemother. 1992;36:1133–5.11. Dong J, Yu XF, Zou J. Azithromycin-containing versus standard tripletherapy for Helicobacter pylori eradication: a meta-analysis. World JGastroenterol. 2009;15:6102–10.12. Murphy BS, Sundareshan V, Cory TJ, Hayes Jr D, Anstead MI, Feola DJ.Azithromycin alters macrophage phenotype. J Antimicrob Chemother.2008;61:554–60.13. Yamauchi K, Shibata Y, Kimura T, Abe S, Inoue S, Osaka D, et al.Azithromycin suppresses interleukin-12p40 expression in lipopolysaccharideand interferon-gamma stimulated macrophages. Int J Biol Sci.2009;5:667–78.14. Khan AA, Slifer TR, Araujo FG, Remington JS. Effect of clarithromycin andazithromycin on production of cytokines by human monocytes. Int JAntimicrob Agents. 1999;11:121–32.15. Albert RK, Connett J, Bailey WC, Casaburi R, Cooper Jr JA, Criner GJ, et al.Azithromycin for prevention of exacerbations of COPD. N Engl J Med.2011;365:689–98.16. Kosunen TU, Seppala K, Sarna S, Sipponen P. Diagnostic value of decreasingIgG, IgA, and IgM antibody titres after eradication of Helicobacter pylori.Lancet. 1992;339:893–5.17. Fallone CA, Loo VG, Barkun AN. Utility of serology in determiningHelicobacter pylori eradication after therapy. Can J Gastroenterol.1998;12:117–24.18. Woodruff PG, Chatila W, Connett JE, Criner GJ, Curtis JL, Dransfield MT, et al.Tumour necrosis factor receptor-75 and risk of COPD exacerbation in theazithromycin trial. Eur Respir J. 2014;43:295–8.19. Efron B, Tibshirani RJ. An introduction to the bootstrap. New York: Chapman& Hall; 1993.20. Qaseem A, Wilt TJ, Weinberger SE, Hanania NA, Criner G, van der Molen T,et al. Diagnosis and management of stable chronic obstructive pulmonarydisease: a clinical practice guideline update from the American College ofPhysicians, American College of Chest Physicians, American ThoracicSociety, and European Respiratory Society. Ann Intern Med.2011;155:179–91.21. Global Initiative for Chronic Obstructive Lung Disease. Global Strategy forthe diagnosis, management, and prevention of Chronic Obstructive LungDisease. updated 2016. http://www.goldcopd.org. Accessed 1 May 2017.22. Han MK, Tayob N, Murray S, Dransfield MT, Washko G, Scanlon PD, et al.Predictors of chronic obstructive pulmonary disease exacerbation reductionin response to daily azithromycin therapy. Am J Respir Crit Care Med.2014;189:1503–8.23. Grad YH, Lipsitch M, Aiello AE. Secular trends in Helicobacter pyloriseroprevalence in adults in the United States: evidence for sustained race/ethnic disparities. Am J Epidemiol. 2012;175:54–9.24. Rosenstock S, Jorgensen T, Andersen L, Bonnevie O. Seroconversion andseroreversion in IgG antibodies to Helicobacter pylori: a serology basedprospective cohort study. J Epidemiol Community Health. 2000;54:444–50.25. Kuipers EJ, Pena AS, van Kamp G, Uyterlinde AM, Pals G, Pels NF, et al.Seroconversion for Helicobacter pylori. Lancet. 1993;342:328–31.26. Parsonnet J, Blaser MJ, Perez-Perez GI, Hargrett-Bean N, Tauxe RV.Symptoms and risk factors of Helicobacter pylori infection in a cohort ofepidemiologists. Gastroenterology. 1992;102:41–6.27. Kumagai T, Malaty HM, Graham DY, Hosogaya S, Misawa K, Furihata K, et al.Acquisition versus loss of Helicobacter pylori infection in Japan: results froman 8-year birth cohort study. J Infect Dis. 1998;178:717–21.28. Krueger WS, Hilborn ED, Converse RR, Wade TJ. Environmental risk factorsassociated with Helicobacter pylori seroprevalence in the United States: across-sectional analysis of NHANES data. Epidemiol Infect. 2015;143:2520–31.29. Jung JH, Choi KD, Han S, Jung HY, Do MY, Chang HS, et al. Seroconversionrates of Helicobacter pylori infection in Korean adults. Helicobacter.2013;18:299–308.30. Wang WM, Chen CY, Jan CM, Chen LT, Perng DS, Lin SR, et al. Long-termfollow-up and serological study after triple therapy of Helicobacter pylori-associated duodenal ulcer. Am J Gastroenterol. 1994;89:1793–6.31. Cutler AF, Prasad VM. Long-term follow-up of Helicobacter pylori serologyafter successful eradication. Am J Gastroenterol. 1996;91:85–8.32. Daugule I, Rumba I, Ejderhamn J. Previous antibacterial treatment due toconcomitant infections in preschool children is associated with a lowerHelicobacter pylori positivity. Scand J Infect Dis. 2005;37:326–9.33. Dohmen W, Seelis RE. The role of azithromycin in the treatment ofHelicobacter pylori infection - a retrospective report. Infection.1998;26:256–62.34. Gerrits MM, van Vliet AH, Kuipers EJ, Kusters JG. Helicobacter pylori andantimicrobial resistance: molecular mechanisms and clinical implications.Lancet Infect Dis. 2006;6:699–709.35. Malfertheiner P, Megraud F, O’Morain CA, Atherton J, Axon AT, Bazzoli F, etal. Management of Helicobacter pylori infection–the Maastricht IV/ FlorenceConsensus Report. Gut. 2012;61:646–64.36. Peek Jr RM, Miller GG, Tham KT, Perez-Perez GI, Zhao X, Atherton JC, et al.Heightened inflammatory response and cytokine expression in vivo to cagA+ Helicobacter pylori strains. Lab Invest. 1995;73:760–70.37. Perri F, Clemente R, Festa V, De Ambrosio CC, Quitadamo M, Fusillo M, et al.Serum tumour necrosis factor-alpha is increased in patients withHelicobacter pylori infection and CagA antibodies. Ital J GastroenterolHepatol. 1999;31:290–4.38. Russo F, Jirillo E, Clemente C, Messa C, Chiloiro M, Riezzo G, et al. Circulatingcytokines and gastrin levels in asymptomatic subjects infected byHelicobacter pylori (H. pylori). Immunopharmacol Immunotoxicol.2001;23:13–24.39. Keatings VM, Collins PD, Scott DM, Barnes PJ. Differences in interleukin-8and tumor necrosis factor-alpha in induced sputum from patients withchronic obstructive pulmonary disease or asthma. Am J Respir Crit CareMed. 1996;153:530–4.40. Chung KF. Cytokines in chronic obstructive pulmonary disease. Eur Respir JSuppl. 2001;34:50s–9.Ra et al. Respiratory Research  (2017) 18:109 Page 9 of 9


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items