Bifidobacterium breve Bif195 Protects Against Small-intesti…

Atlantia is a world-class provider of human clinical studies. For over a decade we have delivered clinical results to our many clients around the world, on time and on budget. We strive to offer the highest quality science while maintaining a cost-conscious approach to maximize the value proposition for each of our clients.

Accepted Manuscript

Bifidobacterium breve Bif195 Protects Against Small-intestinal Damage Caused by Acetylsalicylic Acid in Healthy Volunteers Brynjulf Mortensen, PhD, Clodagh Murphy, MD, John O'Grady, MD, Mary Lucey, MsC, Gafer Elsafi, MD, Lillian Barry, MsC, Vibeke Westphal, PhD, Anja Wellejus, PhD, Oksana Lukjancenko, PhD, Aron C. Eklund, PhD, Henrik Bjørn Nielsen, PhD, Adam Baker, PhD, Anders Damholt, PhD, Johan E.T. van Hylckama Vlieg, PhD, Fergus Shanahan, Prof., Martin Buckley, MD.

PII: DOI:

S0016-5085(19)40880-9

https://doi.org/10.1053/j.gastro.2019.05.008

Reference:

YGAST 62645

To appear in:

Gastroenterology

Accepted Date: 7 May 2019

Please cite this article as: Mortensen B, Murphy C, O'Grady J, Lucey M, Elsafi G, Barry L, Westphal V, Wellejus A, Lukjancenko O, Eklund AC, Nielsen HB, Baker A, Damholt A, van Hylckama Vlieg JET, Shanahan F, Buckley M, Bifidobacterium breve Bif195 Protects Against Small-intestinal Damage Caused by Acetylsalicylic Acid in Healthy Volunteers, Gastroenterology (2019), doi: https:// doi.org/10.1053/j.gastro.2019.05.008. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

A CCEPTED MANUSCRIPT

Bifidobacterium breve Bif195 Protects Against Small-intestinal Damage Caused by Acetylsalicylic Acid in Healthy Volunteers A randomized, placebo-controlled, double-blind clinical trial Aspirin + Bif195 Aspirin + Placebo Bif195 Placebo

Bif195 / Placebo

Aspirin

=35

= 31

A CCEPTED MANUSCRIPT

Bifidobacterium breve Bif195 Protects Against Small-intestinal Damage Caused by

Acetylsalicylic Acid in Healthy Volunteers

Brynjulf Mortensen 1 PhD, Clodagh Murphy 2,3 MD, John O'Grady 2,3 MD, Mary Lucey 4 MsC, Gafer Elsafi 3 MD, Lillian Barry 4 MsC, Vibeke Westphal 1 PhD, Anja Wellejus 1 PhD, Oksana Lukjancenko 5 PhD, Aron C. Eklund 5 PhD, Henrik Bjørn Nielsen 5 PhD, Adam Baker 1 PhD, Anders Damholt 1 PhD, Johan E. T. van Hylckama Vlieg 1 PhD, Fergus Shanahan 2,3 Prof. and Martin Buckley 2,3,4 MD.

1 Chr. Hansen A/S, Human Health Innovation, Hoersholm, Denmark

2 APC Microbiome Ireland, Cork, Ireland

3 Department of Medicine, University College Cork, National University of Ireland

MANUSCRIPT

4 Centre for Gastroenterology, Mercy University Hospital, Cork, Ireland

5 Clinical Microbiomics A/S, Copenhagen, Denmark

Corresponding Author: Professor Fergus Shanahan, APC Microbiome Ireland, Cork, Ireland.

Email: f.shanhan@ucc.ie Phone: +353 (0)86 2804881

1

A CCEPTED MANUSCRIPT

Declaration of interest: FS is supported, in part, by Science Foundation Ireland in the form of a

research centre grant (SFI/12/RC/2273). BM, VW, AW, AB, AD and JETvHV are all employees of

the Sponsor Chr. Hansen A/S.

Contributors: BM contributed to the trial conception and trial design, wrote the trial protocol and

prepared and managed the trial on behalf of the sponsor, drafted the article and contributed to the

final version. CM reviewed the VCE data and contributed to the final version. JOG reviewed the

VCE data and contributed to the final version. ML reviewed the VCE data and contributed to the

final version. GE reviewed the VCE data and contributed to the final version. LB oversaw the VCE

procedure, reviewed the VCE data and contributed to the final version. VW contributed to the trial

conception and trial design and contributed to the final version . AW contributed to the trial

conception and trial design and contributed to the final version. OL performed analysis and

MANUSCRIPT

statistics on microbiome data obtained from fecal samples and contributed to the final version. ACE

performed analysis and statistics on microbiome data obtained from fecal samples and contributed

to the final version. HBN performed analysis and statistics on microbiome data obtained from fecal

samples, drafted parts of the article and contributed to the final version. AB contributed to the trial

conception and trial design and contributed to the final version. AD contributed to the trial

conception and trial design and contributed to the final version. JETVHV contributed to the trial

conception and trial design and contributed to the final version. FS contributed to the trial

conception and trial design, was Co-PI during the trial, oversaw the VCE procedure, reviewed the

VCE data, wrote in part the manuscript and contributed to the final version . MB contributed to the

trial conception and trial design, was PI during the trial, oversaw the VCE procedure, reviewed the

VCE data and contributed to the final version .

2

A CCEPTED MANUSCRIPT

Role of the funding source: This clinical trial was funded entirely by the sponsor Chr. Hansen

A/S. The authors of the trial (some of whom are Chr. Hansen employees) jointly designed the trial.

The sponsor was not directly involved in trial conduct nor the statistical analysis on the dataset

obtained. All authors had access to the data after unblinding, contributed to writing the manuscript,

approved the manuscript before submission and vouch for its integrity. The corresponding author

had the final responsibility for the decision to submit for publication.

Abbreviations:

AUC

Area-under-the-curve

ASA

Acetylsalicylic Acid

NSAID

Non-steroidal anti-inflammatory drugs

PPIs

Proton pump inhibitors

MANUSCRIPT Video Capsule Endoscopy

VCE

Acknowledgements : We would like to acknowledge all of the involved staff at the CRO Atlantia

Food Clinical Trials, Cork Ireland for an efficient and professional conduct of the trial on behalf of

the sponsor. We would like to acknowledge all of the involved staff at the CRO Signifikans,

Vedbæk Denmark for their professional handling of all data management and statistical analysis

performed. We would also like to thank Medtronic, Ireland for technical assistance and dialogue in

planning of the VCE procedures conducted during the trial.

3

A CCEPTED MANUSCRIPT

Abstract

Background & Aims: Enteropathy and small-intestinal ulcers are common side effects of non-

steroidal anti-inflammatory drugs such as acetylsalicylic acid (ASA). Safe, cytoprotective strategies

are needed to reduce this risk. Specific Bifidobacteria might have cytoprotective activities, but little

is known about these effects in humans. We used serial video capsule endoscopy (VCE) to assess

the efficacy of a specific Bifidobacterium strain in healthy volunteers exposed to ASA.

Methods: We performed a single-site, double-blind, parallel-group, proof of concept analysis of 75

heathy volunteers given ASA (300 mg) daily for 6 weeks, from July 31 through October 24, 2017.

The participants were randomly assigned (1:1) to groups given oral capsules of Bifidobacterium

10 colony forming units) or placebo, daily for 8 weeks. Small-intestinal

breve (Bif195; ≥ 5*10

damage was analyzed by serial VCE at 6 visits. The area under the curve (AUC) for intestinal

damage (Lewis score) and the AUC value for ulcers were the primary and first-ranked secondary

MANUSCRIPT

endpoint of the trial, respectively.

Results: Efficacy data were obtained from 35 participants given Bif195 and 31 given placebo. The

AUC for Lewis score was significantly lower in the Bif195 group (3040 ± 1340 arbitrary units) than

the placebo (4351 ± 3195 arbitrary units) ( P =.0376). The AUC for ulcer number was significantly

lower in the Bif195 group (50.4 ± 53.1 arbitrary units) than in the placebo group (75.2 ± 85.3

arbitrary units) ( P =.0258). Twelve adverse events were reported from the Bif195 group and 20 from

the placebo group. None of the events were determined to be related to Bif195 intake.

ClinicalTrials.gov no: NCT03228589

Conclusions: In a randomized double-blind trial of healthy volunteers, we found oral Bif195 to

safely reduce the risk of small-intestinal enteropathy caused by ASA.

KEY WORDS: aspirin, bacteria, microbiota, bleeding

4

A CCEPTED MANUSCRIPT

Introduction

Non-Steroidal Anti-Inflammatory Drugs ( NSAIDs) are used worldwide both as prescription and

over-the-counter products for their analgesic, anti-inflammatory and cardiovascular disease (CVD)

risk-reduction properties, and are among the most used pharmaceuticals in the world today 1 .

Chronic, low-dose use (commonly defined as 75-325 mg daily) of the NSAID Acetylsalicylic Acid

(ASA) is widely recommended for both primary and secondary prevention of CVD. More than 30%

of the US population aged above 40 are estimated to be on chronic, daily, low-dose ASA for that

reason alone 2 . However, chronic use of ASA is also associated with adverse side effects including

small-intestinal mucosal lesions and ulcers, perforations, major hemorrhage and in rare instances

death 3,4,5 . A recent review and meta-analysis addressing both the efficacy of ASA in prevention of

CVD and also bleeding-related side-effects concluded that a balanced, cautious approach should be

taken in the case of primary CVD prevention due to these side-effects 6 , highlighting the unmet need

MANUSCRIPT

to reduce the risk of side-effects of chronic ASA use.

For decades endoscopists have acknowledged the vulnerability of the gastroduodenal mucosa to

NSAID-induced enteropathy. Complications include ulceration, blood loss, protein loss, perforation

and occasional strictures. The pathogenesis of tissue injury at the gastric and small-intestinal sites

appears to differ 7,8 , and therefore distinct and separate preventative strategies are probably required

to combat enteropathy and gastropathy. For example, the risk of gastropathy can be offset by acid

suppression, usually with proton pump inhibitors (PPIs). However, the pathogenesis of NSAID-

induced damage in the small bowel seems to be much more complex and has been shown to involve

microbiota composition, bile and enterohepatic circulation of the certain NSAIDs 8 . Moreover, there

is evidence to suggest that PPIs may actually increase the risk of NSAID-associated small bowel

injury 9 , possibly by disturbing the composition of the small bowel microbiota 10 . The importance of

the microbiota is emphasised by the fact that administration of NSAIDs to germ-free animals is

5

A CCEPTED MANUSCRIPT

associated with minimal damage to the small intestinal mucosa and co-administration of antibiotics

reduces NSAID-induced injury 8,7 . Besides the well-established inhibitory effect of cyclooxygenase (COX), ASA specifically has been recognised to compromise the phospholipid layer in mucus 11

increasing access to luminal aggressors like lipopolysaccharide and bile as well as disrupt intestinal

permeability and cause inflammation 12 . Given that deleterious compositional changes to the

microbiota, in addition to direct effects on mucus and epithelial tissue, may increase the risk of

NSAID-enteropathy, we hypothesised that an intervention targeting microbiome-host interactions

may offer an attractive, preventative strategy. Our strain selection was based on the anti-

inflammatory properties of certain bifidobacteria 13,14 and experimental pre-clinical evidence for a role of bifidobacteria in NSAID-associated ulceration 15,16,17 as well as unpublished pre-clinical

screening data suggesting a particular potential of efficacy for the specific strain belonging to this

genus. In addition, another Bifidobacterium breve has been shown to express a pilus-associated

MANUSCRIPT

protein (Tad E) in vivo, but not in vitro , which promotes colonic epithelial proliferation 18 .

Here, we describe the development of a clinical model to assess the quantitative and time-resolved

induction of small intestinal injury upon ASA administration. Using this model, we addressed

whether oral co-administration of a single bacterial strain of Bifidobacterium breve (Bif195) can

reduce the risk of low-dose ASA-induced intestinal ulceration in humans in a randomized, placebo-

controlled, parallel-group, double-blind trial using serial video capsule endoscopy (VCE) as a

rigorous demonstration of efficacy.

6

A CCEPTED MANUSCRIPT

Methods

Study design

This clinical trial was a single-site, randomized, double-blind, placebo-controlled, parallel-group,

proof-of-concept trial. The trial was conducted at the CRO Atlantia Food Clinical trials (Cork,

Ireland). The trial was conducted in accordance with the ethical principles set forth in the current

version of the Declaration of Helsinki, the International Conference on Harmonisation E6 Good

Clinical Practice (ICH-GCP). The trial was approved by The Clinical Research Ethics Committee of

the Cork Teaching Hospitals (Cork, Ireland) prior to trial conduct. The trial conduct period was July

– December 2017. The trial was registered at ClinicalTrials.gov under the ID number

NCT03228589.

Participants

MANUSCRIPT

All subjects were carefully informed about the trial before they signed the informed consent form

and were screened for participation criteria. Main inclusion criteria for participation were: Age

between 18 and 40 years, healthy and without GI symptoms, sedentary lifestyle and willingness to

refrain from other bacterial products and medication known to alter GI function throughout trial

participation.

Main exclusion criteria were: History of abdominal surgery (except appendectomy and

cholecystectomy), history of peptic ulcers, known bleeding disorders, known allergy to ASA,

history of diseases related to H. pylori infection, diastolic blood pressure ≥ 90 mmHg, systolic

blood pressure ≥ 140 mmHg, BMI > 27, smoking or use of other nicotine products, lactose

intolerance, pregnancy, lactation and regular use of probiotics, systemic antibiotics, steroids (except

contraceptives), NSAIDs, laxatives, anti-diarrheals, PPIs, and/or immunosuppressant drugs prior to

7

A CCEPTED MANUSCRIPT

screening. After inclusion, subjects went through a two-week run-in period before baseline data

were obtained at Visit 2 with randomization being performed at the very end of visit 2.

Randomization and masking

Prior to trial conduct, the allocation of subjects in a 1:1 ratio to Bif195 or placebo intervention was

planned according to randomization lists. The randomization procedure was stratified by gender and

the lists were drawn up to n=50 for each strata using the proc plan procedure in SAS.

Randomization blocks of n=6 was used throughout and trial site and sponsor were kept blinded to

the use of randomization blocks. The randomization list and unblinding list were produced by a

third party not otherwise involved in the trial.

At screening, subjects were assigned a 4-digit screening number according to their chronological

entry into the trial. If a subject was found eligible and enrolled for trial participation, they received

MANUSCRIPT

their randomization number by blinded trial staff after all baseline assessments performed at Visit 2.

Randomization numbers included the stratification number and was allocated sequentially by trial

staff in the order in which the subjects completed Visit 2.

Test and placebo product were produced by the sponsor to be similar in smell, taste and appearance.

All trial product was packaged in identical packs with identical labelling, except for the

randomization number. All trial subjects, the clinical team, statisticians and the sponsor were all

blinded during the entire trial until database lock and signature of the request for unblinding

document.

An emergency unblinding procedure using emergency code break opaque sealed envelopes was

established to allow the investigator the option of disclosing the product assignment for any

individual subject if clinical circumstances required such unblinding. This option was not used in

the conduct of this trial. The randomization list and production of emergency code break envelopes

8

A CCEPTED MANUSCRIPT

were performed by a third party not otherwise involved in the trial. The labeling of product vials,

based on the randomization list, was also performed by a third-party not otherwise involved in the

trial.

Procedures

Bif195 or placebo were administered in a 1:1 ratio daily to 75 randomized subjects for 8 weeks. To

induce damage to the small intestine, all subjects were co-treated daily with 300 mg of ASA for the

first 6 weeks of the 8 week Bif195/placebo intervention period.

In order to document small-intestinal damage, we performed VCE at 6 visits during the 8 week

intervention period (Suppl. Figure 1 and Suppl. Figure 4). The time course kinetics of ASA-induced

damage, as well as a potentially protective effect by Bif195, were expressed as area-under-the-curve

(AUC) for the 8 week intervention period for all datasets obtained.

MANUSCRIPT

All subjects were given 2 hypromellose capsules daily with or without Bifidobacterium breve

Bif195 starting the day after visit 2 with a duration of 8 weeks. The product stability was monitored

in parallel to trial conduct and showed at least 5*10 10 Colony forming units (CFU) of Bif195 per

daily dose during the period of trial conduct. Detailed trial product and placebo description is

provided in Supplementary Table 1.

All randomized subjects were also given 300 mg of ASA (Alliance Pharmaceuticals, Ireland) to

induce small-intestinal damage. This dose was taken daily from the day after visit 2 with a duration

of 6 weeks.

VCE is the widely accepted reference standard for assessment of occult gastrointestinal bleeding.

Current use include exploration and surveillance of bowel pathology such as in Crohn’s disease,

polyps, small bowel malignancy and drug-induced mucosal injury 19 . To standardise the findings

from VCE, we used a reproducible, clinical scoring system to categorise small intestinal mucosal

damage, the Lewis score. The Lewis score is a validated tool that evaluates villous edema, ulcers

9

A CCEPTED MANUSCRIPT

and stenosis in order to quantify small bowel inflammatory change in one score 20 . This scoring

system uses specific definitions for each of the recorded parameters to reduce inter-reviewer

variability. In addition, we also counted red spots as observed during VCE.

For all VCE analyses (visit 2-7), data were recorded using the SB3 TM Pillcam video recording

capsule (Medtronic, Ireland). For all visits, subjects met fasting in the morning and the Pillcam

capsule was swallowed with water. Video images were recorded for a total of 8 hours during each

visit, after which the capsule was verified in the video to have passed the small intestine.

Four experienced gastroenterologists, blinded to intervention and not allowed to communicate

internally regarding obtained VCE data, reviewed the video material retrieved from the capsules

using the Pillcam TM Reader Software Version 9.0 from Medtronic. The VCE video material from all

6 VCE visits for each of the subjects were evaluated by two randomized reviewers, and mean

values for each subject visit were calculated. In cases where the data from a specific visit differed

MANUSCRIPT

with 4 or more number of ulcers, a third reviewer would review the VCE dataset. A mean value of

all 3 datasets was then calculated and used as the final data point for that specific visit. All VCE

reviews were performed prior to database lock and unmasking of the randomization key.

Representative pictures of the VCE material obtained are shown in Supplementary Figure 4.

Fecal samples and blood samples were obtained during all visits from visit 2 to visit 7 for secondary

and exploratory analyses.

At all visits, subjects completed the GI symptoms rating score (GSRS) questionnaire to assess GI

symptoms 21 .

Intestinal fatty acid binding protein (I-FABP) was measured by Nordic Biosite, Finland, in triplicate

heparin plasma samples using the HK406 human I-FABP ELISA kit from Hycult Biotech under

GLP conditions.

10

A CCEPTED MANUSCRIPT

Serum calprotectin was measured in duplicate serum samples under GLP conditions by Nordic

Biosite, Finland, using the HK379 Human Calprotectin ELISA kit from Hycult Biotech.

Fecal calprotectin was measured in duplicates under GLP conditions by Synlab, Switzerland, using

an ELISA kit from Immundiagnostik AG, Germany.

Outcomes

The primary outcome of this trial was the effect of the Bif195 intervention on the AUC Lewis score

obtained by VCE from visit 2 (randomization) to visit 7 (end of treatment). As the first secondary

endpoint, the effect of the Bif195 intervention on the AUC number of ulcers obtained by VCE from

visit 2 to visit 7 was tested. Other secondary endpoints were, in hierarchical order: AUC of the pain

module from the GSRS questionnaire, AUC of the total score from the GSRS questionnaire, AUC

of blood I-FABP, AUC of red spots from the VCE procedure, AUC of fecal calprotectin and AUC

MANUSCRIPT

of blood calprotectin.

As exploratory endpoints, data stratified into tertiles (small intestine divided into thirds) on effects

of the Bif195 intervention on ulcerations observed by VCE was analysed and further post-hoc

analyses on intervention effects on prostaglandin E2 (PGE2) and thromboxane B2 (TBX2) in serum

samples downstream of COX were studied.

Safety was assessed by means of adverse events. A complete list of adverse events is provided in

Table 2.

Statistical analysis

For all data obtained, area-under-the-curve (AUC) was calculated in order to evaluate the

intervention effects by comparing the AUC in the Bif195 arm versus the placebo arm. For this

purpose, the kinetics of Lewis score for each subject over the 6 VCE visits, was fitted to a third-

11

A CCEPTED MANUSCRIPT

degree polynomial and the total AUC was calculated by computing the integral. This approach was

taken for all VCE-obtained data.

Statistical tests were pre-defined and agreed in the statistical analyses plan finalised and signed

prior to unblinding of the randomization key. The randomization list was made, and the labelling of

trial product was performed by third parties not otherwise involved in the trial. No imputation of

data was carried out in cases of missing data, but all available data were used.

Subject characteristics and all efficacy data presented are based on the Full Analysis Set (FAS)

population. Criteria for inclusion in FAS was defined as maximum one missing visit in between the

randomization visit (visit 2) and end of trial (Visit 7). The safety reporting by listing of adverse

events included all subjects that were randomized (n=75).”

A sample size calculation was performed prior to trial initiation based on the primary endpoint of

the trial. The curve shapes were assumed to fit with a third-degree polynomial. We considered a

MANUSCRIPT

30% lower AUC following treatment of Bif195 compared to placebo to be clinically relevant and

aimed at a trial design that would have 80% power in detecting an intervention effects of this size as

statistical significant. No previously knowledge exists on AUC values and SD. Sample size

calculation was therefore performed on percent difference of AUC between two normalised curves

(Active vs. placebo) as an approximation. We assumed similar standard deviation in each arm and

planned for two-sided testing with a significance level of 5%. Given the above assumptions the

number of subjects needed in each arm was 30. To account for potential drop-out subjects, we

aimed to randomize a total of 75 subjects. Subjects who withdrew within one week of

randomization were replaced by standby-subjects.

In general, datasets were modelled as the dependent variable in a linear mixed model. The model

included the baseline value as covariate and gender and Bif195/placebo intervention as factors.

Model check was always assessed for all datasets using QQ residual plots together with

12

A CCEPTED MANUSCRIPT

Kolmogorov-Smirnov test for normality. In cases where datasets did not meet a normal distribution,

a log transformation was performed and check for normality performed again. In cases where a

normal distribution was still not obtained, the dataset was tested for intervention effects using a

non-parametric Mann-Whitney test. Curves in Figure 2,3 and 5 are shown as mean values or

medians, depending on normality. Bars in Figure 2-5 are always shown as mean ± SEM.

All authors had access to the study data and reviewed and approved the final manuscript.

Results Between July 31st, 2017 to October 24 th 2017, 109 subjects were screened for eligibility, of whom

75 were enrolled and randomized. Among the 75 randomized subjects, 9 subjects discontinued

during the intervention (n=3 active and n=6 placebo) and therefore efficacy data was obtained in a

total of 66 subjects, the analysis population (n=35 active arm and n=31 placebo, Figure 1).

MANUSCRIPT

The arms were in general similar in their baseline parameters as shown in Table 1, including gender

distribution, age, BMI and blood pressure. Accountability of both ASA and trial product were in

general very high in both two arms (Table 1).

This clinical trial met its primary endpoint with a statistically significantly (p=0.0376) lower AUC

Lewis Score, as captured by VCE, during the 8 weeks intervention in the Bif195 arm versus the

placebo arm (3040 ± 1340 arbitrary units (au) in the Bif195 arm vs 4351 ± 3195 au in the placebo

arm, Figure 2A and B). In addition, the trial met its secondary endpoint with a significantly

(p=0.0258) lower AUC ulcer number as captured using VCE during the intervention in Bif195

subjects versus the placebo group (50.4 ± 53.1 au in the Bif195 arm vs 75.2 ± 85.3 au in the placebo

arm, Figure 2C and D).

13

A CCEPTED MANUSCRIPT

An exploratory tertile stratification of VCE data showed that the damage induced by ASA occurs

primarily in the first tertile (Figure 3) where a significant Bif195 protective effect (p=0.03) was also

observed (31.0 ± 16.8 au in the Bif195 arm vs 41.6 ± 25.2 au in the placebo arm, Figure 3A and B).

The other secondary endpoints GSRS pain AUC, GSRA total score AUC, plasma I-FABP AUC,

red spots from VCE AUC and serum calprotectin AUC did not meet statistical significance (Figure

4) while fecal calprotectin AUC was significantly lower (p=0.0347) in the Bif195 arm compared to

the placebo arm (Figure 4E.)

ASA and trial product were both generally well-tolerated by the subjects. In total, 32 adverse events

were registered from 22 different subjects included in the n=75 safety analysis set. Twelve of these

adverse events were reported from the Bif195 arm and 20 from the placebo arm (Table 2). None of

the adverse events were related to Bif195 intake, while in total 10 of them were assumed related to

ASA intake, as assessed by the principal investigator. The number of adverse events related to ASA

MANUSCRIPT

did not differ between the two intervention arms (4 and 6 in the Bif195 and placebo arm,

respectively).

DNA sequencing of all fecal samples obtained showed an increase after randomization in

abundance of Bifidobacterium breve in fecal samples obtained from subjects in the Bif195 arm

compared to the placebo arm, confirming trial product compliance (Supplementary Figure 2). The

Bif195 intervention was not associated with significant changes in abundance of specific microbial

taxa nor in the changes of the overall microbiome composition (as revealed by Bray-Curtis

dissimilarity index, Supplementary Figure 3).

Serum PGE2 and TXB2 concentrations showed a robust decline during ASA intake and a reversal

to baseline levels during the final 2 weeks recovery period. The Bif195 intervention did not have

significant effects on these datasets (Figure 5).

14

A CCEPTED MANUSCRIPT

Discussion

The trial results indicate that Bifidobacterium breve Bif195 confers significant and objectively

verifiable protection against small-intestinal damage caused by a 6 week ASA challenge in healthy

volunteers. The primary and first secondary efficacy criteria for the trial were met, thereby

highlighting the potential of Bif195 co-treatment in future prevention strategies for a growing

population experiencing silent or overt small-intestinal enteropathy from chronic ASA use.

Although prior studies have described gastric damage from NSAIDs, this is, to the best of our

knowledge, the first trial to record the detailed time-resolved kinetics of ASA-induced, and reversal

of, small-intestinal damage. This dataset shows a gradual increase in damage observed by VCE

during the 6 weeks of daily ASA intake and a partial reversal towards baseline levels over a 2-week

recovery period. Furthermore, the small-intestinal tertile stratification clearly shows that ASA-

induced enteropathy is mainly a duodenal phenomenon. This site coincides with localisation of the

MANUSCRIPT

main effect of the Bif195 intervention on ulceration, further highlighting the potential of protective

intervention with this strain. The strategy to perform serial capsule endoscopies in this trial, enabled

us to obtain the sensitivity needed to observe a significant effect in a dynamic environment where

damage formation and healing co-exists. Thus, it represents a superior and more sensitive form of

assessment than the more usually adopted before/after intervention trial design.

The efficacy of Bif195 in NSAID-associated small intestinal injury may be partly explained by the

difference in pathogenesis between NSAID-associated small intestinal injury and NSAID-

associated gastropathy. Whereas acid and pepsin are the principal luminal aggressors in NSAID-

gastropathy, bile and indeed bacteria are the luminal factors in NSAID-enteropathy 22 . Although pre-

clinical studies in experimental animals have been encouraging, previous trials in humans of

putative probiotics in NSAID-enteropathy have been inconsistent. Certain strains of Bifidobacteria,

are known to strengthen the intestinal epithelium layer, to modulate the local immunoinflammatory

15

A CCEPTED MANUSCRIPT

response as well as compete with potential bacterial aggressors. The molecular details of

bifidobacterial-mediated protection against small-intestinal epithelial injury are currently under

investigation, but one candidate includes the pilus-associated protein Tad E which exerts a

proliferative effect on host colonic epithelium following oral consumption of B. breve 18 . This

appears to be a characteristic of all B. breve and supports our choice of the stain used in this trial.

Interestingly, fecal microbiome analysis revealed changes were limited to a marked increase in the

total B. breve population in the Bif195 arm. These data provide further evidence that microbial

intervention strategies targeting the microbiome can be clinically efficacious without inducing

major alterations in the overall microbial population structure.

Our 6-week ASA challenge model yielded minor responses in the GSRS questionnaire and in the

biomarkers of damage, I-FABP in blood and calprotectin in blood and feces. Although trends were

observed for I-FABP, only the fecal calprotectin endpoint reached statistical significance indicating

MANUSCRIPT

a modest Bif195 protective effect. Our data suggest that VCE is the method of choice when

conducting human challenges with mild induction of small-intestinal damage by NSAIDs over a

limited time period.

Although encouraging, the present clinical trial has limitations in terms of translation to a real-life

clinical setting. The relatively short-term challenge in healthy volunteers, for proof-of-concept, used

a higher dose of ASA than is most commonly prescribed for primary CVD prevention. However, it

is a dose that is readily available for over-the-counter usage. It is also noteworthy, that a recent

report suggested that the current cardioprotective dosage of ASA may be insufficient

and recommended doses based on a mg/kg basis 23 .

Due to our AUC approach based on a polynomic curve fitted to data-points obtained from 6

different visits, data imputation is not feasible for drop-out subjects where only baseline data are

available. Therefore, we acknowledge that long-term intervention clinical trials will be needed to

16

A CCEPTED MANUSCRIPT

confirm if Bif195 has long-term clinical efficacy in a larger intention-to-treat population of chronic

users of ASA taking lower doses for CVD prevention.

In addition, we acknowledge that the division of the small intestine into tertiles by VCE is based on

assumptions and that tertile-specific data are an approximation.

As expected, the ASA intake was associated with robust inhibition downstream of the COX enzyme

on serum PGE2 and TXB2 concentrations. It is noteworthy that the Bif195 intervention did not alter

these well-described ASA-induced changes in metabolites downstream of COX 24,25 . This suggests

that the small-intestinal protective actions of Bif195 is unlikely to interfere with the specific

cardiovascular-protective properties of ASA. Close monitoring of adverse events during this trial

suggests that daily, oral intake of Bif195 is safe and without side-effects. Further clinical trials are

required to test whether the strain has clinical efficacy also in other settings and populations, i.e. in

chronic users of ASA.

MANUSCRIPT

17

A CCEPTED MANUSCRIPT

Figure legends:

Figure 1.

Enrollment and randomization of subjects according to the CONSORT Flow Diagram.

Figure 2.

Primary and Secondary endpoint. Mean Lewis Score per visit (A) and the primary

endpoint mean Lewis Score AUC ± SEM (B) per treatment arm. Median number of ulcers per visit

(C) and the secondary endpoint ulcer number AUC ± SEM (D) per treatment arm. * indicates p <

0.05. Effects sizes were 30% lower AUC in Bif195 arm (B) and 33% lower AUC in Bif195 arm

(D).

Figure 3.

Tertile-stratification of ulceration. Median ulcer numbers, both per visit (A, C and E)

MANUSCRIPT

and mean ulcer number AUC ± SEM (B, D and F) from Video Capsule Endoscopy stratified on

small-intestinal tertiles (thirds of small intestine). * indicates p < 0.05.

Figure 4.

Other secondary endpoints measured in the trial. AUC ± SEM of the Pain module (A)

and total score (B) from in the GSRS questionnaire. AUC ± SEM of blood I-FABP (C), AUC ±

SEM of red spots from VCE (D), AUC ± SEM of fecal (E) and blood (F) calprotectin. * indicates p

< 0.05.

Figure 5.

Mean serum concentrations of Prostaglandin E2 per visit (A) and AUC ± SEM (B).

Mean serum concentrations of Thromboxane B2 per visit (C) and AUC ± SEM (D).

18

A CCEPTED MANUSCRIPT

References

1. Zhou, Y., Boudreau, D. M. & Freedman, A. N. Trends in the use of aspirin and nonsteroidal

anti-inflammatory drugs in the general U.S. population. Pharmacoepidemiol. Drug Saf. 23,

43–50 (2014).

2. Stuntz, M. & Bernstein, B. Recent trends in the prevalence of low-dose aspirin use for

primary and secondary prevention of cardiovascular disease in the United States, 2012–2015.

Prev. Med. Reports 5, 183–186 (2017).

3. Goldstein, J. L. & Cryer, B. Gastrointestinal injury associated with NSAID use: a case study

and review of risk factors and preventative strategies. Drug. Healthc. Patient Saf. 7, 31–41

(2015).

4. Sostres, C. & Gargallo, C. J. Gastrointestinal lesions and complications of low-dose aspirin

in the gastrointestinal tract. Best Pract. Res. Clin. Gastroenterol. 26, 141–151 (2012).

MANUSCRIPT

5. McNeil, J. J. et al. Effect of Aspirin on Cardiovascular Events and Bleeding in the Healthy

Elderly. N. Engl. J. Med. 379, 1509–1518 (2018).

6. Zheng, S. L. & Roddick, A. J. Association of Aspirin Use for Primary Prevention With

Cardiovascular Events and Bleeding Events: A Systematic Review and Meta-analysis. JAMA

321, 277–287 (2019).

7. Bjarnason, I. et al. Mechanisms of Damage to the Gastrointestinal Tract From Nonsteroidal

Anti-Inflammatory Drugs. Gastroenterology 154, 500–514 (2018).

8. Wallace, J. L. NSAID gastropathy and enteropathy: distinct pathogenesis likely necessitates

distinct prevention strategies. Br. J. Pharmacol. 165, 67–74 (2012).

9. Washio, E. et al. Proton Pump Inhibitors Increase Incidence of Nonsteroidal Anti-

Inflammatory Drug–Induced Small Bowel Injury: A Randomized, Placebo-Controlled Trial.

Clin. Gastroenterol. Hepatol. 14, 809–815.e1 (2016).

19

A CCEPTED MANUSCRIPT

10. Wallace, J. L. et al. Proton Pump Inhibitors Exacerbate NSAID-Induced Small Intestinal

Injury by Inducing Dysbiosis. Gastroenterology 141, 1314–1322.e5 (2011).

11. Anand, B. S., Romero, J. J., Sanduja, S. K. & Lichtenberger, L. M. Phospholipid association

reduces the gastric mucosal toxicity of aspirin in human subjects. Am. J. Gastroenterol. 94,

1818–1822 (1999).

12. Lai, Y. et al. Rebamipide Promotes the Regeneration of Aspirin-Induced Small-Intestine

Mucosal Injury through Accumulation of β -Catenin. PLoS One 10, e0132031 (2015).

13. Schiavi, E. et al. The Surface-Associated Exopolysaccharide of Bifidobacterium longum

35624 Plays an Essential Role in Dampening Host Proinflammatory Responses and

Repressing Local TH17 Responses. Appl. Environ. Microbiol. 82, 7185–7196 (2016).

14. Konieczna, P., Akdis, C. A., Quigley, E. M. M., Shanahan, F. & O’Mahony, L. Portrait of an

immunoregulatory Bifidobacterium. Gut Microbes 3, 261–6 (2012).

MANUSCRIPT

15. Syer, S. D. et al. Su1724 Bifidobacteria Exert a Protective Effect Against NSAID-Induced

Enteropathy That is Dependent on Lactate Production. Gastroenterology 142, S-489 (2012).

16. Kinouchi, T. et al. Culture supernatants of Lactobacillus acidophilus and Bifidobacterium

adolescentis repress ileal ulcer formation in rats treated with a nonsteroidal antiinflammatory

drug by suppressing unbalanced growth of aerobic bacteria and lipid peroxidation. Microbiol.

Immunol. 42, 347–55 (1998).

17. Uejima, M., Kinouchi, T., Kataoka, K., Hiraoka, I. & Ohnishi, Y. Role of intestinal bacteria

in ileal ulcer formation in rats treated with a nonsteroidal antiinflammatory drug. Microbiol

Immunol 40, 553–560 (1996).

18. O’Connell Motherway, M. et al. A Bifidobacterial pilus-associated protein promotes colonic

epithelial proliferation. Mol. Microbiol. (2018). doi:10.1111/mmi.14155

19. Mustafa, B. F., Samaan, M., Langmead, L. & Khasraw, M. Small bowel video capsule

20

A CCEPTED MANUSCRIPT

endoscopy: an overview. Expert Rev. Gastroenterol. Hepatol. 7, 323–329 (2013).

20. Gralnek, I. M. et al. Development of a capsule endoscopy scoring index for small bowel

mucosal inflammatory change. Aliment. Pharmacol. Ther. 27, 146–154 (2008).

21. Svedlund, J., Sjödin, I. & Dotevall, G. GSRS--a clinical rating scale for gastrointestinal

symptoms in patients with irritable bowel syndrome and peptic ulcer disease. Dig. Dis. Sci.

33, 129–34 (1988).

22. Bjarnason, I. et al. Mechanisms of Damage to the Gastrointestinal Tract From Nonsteroidal

Anti-Inflammatory Drugs. Gastroenterology 154, 500–514 (2018).

23. Rothwell, P. M. et al. Effects of aspirin on risks of vascular events and cancer according to

bodyweight and dose: analysis of individual patient data from randomised trials. Lancet 392,

387–399 (2018).

24. Ritter, J. M., Cockcroft, J. R., Doktor, H. S., Beacham, J. & Barrow, S. E. Differential effect

MANUSCRIPT

of aspirin on thromboxane and prostaglandin biosynthesis in man. Br. J. Clin. Pharmacol.

28, 573–9 (1989).

25. Pulcinelli, F. M. et al. Inhibition of platelet aggregation by aspirin progressively decreases in

long-term treated patients. J. Am. Coll. Cardiol. 43, 979–984 (2004).

21

A CCEPTED MANUSCRIPT

Table 1. Subject baseline characteristics and trial compliance of the analysis population.

Bif195

Placebo

N

35

31

Age (years)

30.5 ± 6.8

31.2 ± 6.4

Gender (m/f)

16/19

14/17

Ethnicity (non-caucasians)

2

0

Height (cm)

172.2 ± 12.1 173.4 ± 10.2

Weight (kg)

73.5 ± 12.5

72.0 ± 11.4

BMI (kg/m^2)

24.6 ± 2.1

23.8 ± 2.2

Blood pressure, Systolic (mm hg)

124.1 ± 7.8

121.6 ± 10.2

Blood pressure, Diastolic (mm hg)

78.7 ± 6.9

77.1 ± 7.6

alcohol consumption (“drinks” per week)

5.1 ± 3.2

5.5 ± 3.7

MANUSCRIPT 98.7 ± 2.4 99.1 ± 1.9 98.6 ± 2.4 99.0 ± 1.9

Compliance of ASA intake (%, 100% = product subj. should have taken during trial)

Compliance of trial product (%, 100% = product subj. should have taken during trial)

Body-mass index, BMI, is the weight in kilograms divided by the square of height in meters. Numbers are given as Mean ± SD.

A CCEPTED MANUSCRIPT

Table 2. Trial adverse events overview.

Treatment

Bif195

Placebo

Total

N (%) E N (%) E N (%) E

Number of subjects

38

37

75

14 (37.8) 20

All adverse events

8 (21.1) 12

22 (29.3) 32

Back pain

1 ( 2.6) 1

0 (0) 0

1 (1.3) 1

Blocked sinuses

0 (0) 0

1 (2.7) 1

1 (1.3) 1

Chest infection

1 (2.6) 1

0 (0) 0

1 (1.3) 1

Cold and flu

0 (0) 0

1 (2.7) 1

1 (1.3) 1

Cold flu symptoms

1 (2.6) 1

0 (0) 0

1 (1.3) 1

Cold/flu symptoms including a nose bleed.

0 (0) 0

1 (2.7) 1

1 (1.3) 1

Cough, nasal congestion

0 (0) 0

1 (2.7) 1

1 (1.3) 1

Cramping in the stomach

1 (2.6) 1

0 (0) 0

1 (1.3) 1

MANUSCRIPT 1 (2.6) 1 0 (0) 0 1 (1.3) 1 1 (2.6) 1 2 (5.4) 2 3 (4.0) 3 0 (0) 0 1 (2.7) 1 1 (1.3) 1

Cystitis

Headache

Headache, sore throat, rhinorrhea.

Heartburn

0 (0) 0

2 (5.4) 2

2 (2.7) 2

Inflammation in kidneys due to kidney stones

1( 2.6) 1 0(0) 0

1 (1.3) 1

Lower abdominal pain

0 (0) 0

1 (2.7) 1

1 (1.3) 1

Nasal congestion

1 (2.6) 1

0 (0) 0

1 (1.3) 1

Nausea

1 (2.6) 1

0 (0) 0

1 (1.3) 1

Nausea, vomiting, headache, fatigue

0 (0) 0

1 (2.7) 1

1 (1.3) 1

Pain and discomfort in the stomach and gut region.

0 (0) 0

1 (2.7) 1

1 (1.3) 1

Pain and discomfort in the stomach/gut region.

1 (2.6) 1

0 (0) 0

1 (1.3) 1

Pain/discomfort in the stomach and gut region.

0 (0) 0

1 (2.7) 1

1 (1.3) 1

Painful headache which caused vomiting, thigh and back pain.

0 (0) 0

1 (2.7) 1

1 (1.3) 1

A CCEPTED MANUSCRIPT

Sore throat and flu symptoms

0 (0) 0

1 (2.7) 1

1 (1.3) 1

Stomach cramps

0 (0) 0

1 (2.7) 1

1 (1.3) 1

Stomach cramps and loose stools

1 (2.6) 1

0 (0) 0

1 (1.3) 1

Stomach discomfort

1 (2.6) 1

0 (0) 0

1 (1.3) 1

Subject became pregnant

0 (0) 0

1 (2.7) 1

1 (1.3) 1

Subject was physically assaulted and suffered facial injuries

0 (0) 0

1 (2.7) 1

1 (1.3) 1

Vomiting bug

0 (0) 0

2 (5.4) 2

2 (2.7) 2

N = Number of subjects in the group having the event. E = Number of events in total in the group.

() = Percentage of subjects in the group having the event.

MANUSCRIPT

A CCEPTED MANUSCRIPT

MANUSCRIPT

A CCEPTED MANUSCRIPT

MANUSCRIPT

A CCEPTED MANUSCRIPT

MANUSCRIPT

A CCEPTED MANUSCRIPT

MANUSCRIPT

A CCEPTED MANUSCRIPT

MANUSCRIPT

A CCEPTED MANUSCRIPT

Supplementary table 1. Trial product composition

Placebo capsules

Probiotic capsules Bifidobacterium breve

Manufacturing

Chr. Hansen A/S, Denmark Chr. Hansen A/S, Denmark

Brief description

Capsules with excipients only

Capsules containing Bifidobacterium breve and excipients

Capsules

Size 1 HPMC capsules

Size 1 HPMC capsules

Capsules shell

73.6 mg Hypmellose 1.4 mg Titanium dioxide

73.6 mg Hypmellose 1.4 mg Titanium dioxide

Active Ingredients None

Bifidobacterium breve Bif195

Excipients

Microcrystalline Cellulose 6 mg per capsule Magnesium Stearate 1.5 mg per capsule Maltodextrin 277.8 mg per capsule Sodium Ascorbate 14.7 mg per capsule

Supplied as

CSP Activ-Vials containing 24 capsules in each vial

Storage conditions Store at +2-8 ºC

MANUSCRIPT

1

A CCEPTED MANUSCRIPT

Supplementary table 2. Discontinued subjects overview.

Subject ID

Trial product taken

Aspirin taken

Baseline Lewis score

Reason for trial discontinuation

Bif195/placebo

1013

Subject became pregnant. Contraindicates continuation.

Placebo

yes

yes

0

1026

Withdrawal for personal reasons

Placebo

yes

yes

184

1029

Baseline VCE capsule did not reach caecum. Contraindicates continuation.

Bif195

No

No

-

1042

Baseline VCE capsule retained in stomach. Contraindicates continuation.

Placebo

No

No

-

1048

Withdrawal for personal reasons

Bif195

yes

yes

0

1077

Baseline VCE capsule retained in stomach. Contraindicates continuation.

Placebo

No

No

-

1083

Baseline VCE capsule did not reach caecum. Contraindicates continuation.

Placebo

No

No

-

1089

SAE due to prolonged hospitalisation (back pain). Event unrelated to intake of trial product.

Placebo

yes

yes

67.5

1108

Withdrawal for personal reasons

Bif195

yes

yes

67.5

Supplementary Figure 1: Trial design.

MANUSCRIPT

2

A CCEPTED MANUSCRIPT

Supplementary Figure 2:

MANUSCRIPT

3

A CCEPTED MANUSCRIPT

Supplementary Figure 3:

MANUSCRIPT

4

A CCEPTED MANUSCRIPT

Supplementary figure 4. Representative Video Capsule Endoscopy data.

MANUSCRIPT

5

A CCEPTED MANUSCRIPT

Figure legends:

Supplementary Figure 1. The trial consisted of a 2 week run-in phase after the screening visit.

Subjects were then randomised to 8 weeks of Bif195 or placebo intervention for 8 weeks. The first

6 of these 8 weeks, all subjects took 300 mg ASA daily. In total 6 visits with video capsule

endoscopies were performed during the 8 week intervention period.

Supplementary Figure 2. Box-plot showing the relative abundances of Bifidobacterium breve in

stool at visits 2-7. The boxed extends from the first quartile (Q1) to the third quartile (Q3) and the

line within the box shows the median value. The lower whisker extends to the smallest value within

Q1 - 1.5 x inter-quartile range (IQR) and the upper whisker extends to the largest value within Q3 +

1.5 x IQR. Values outside the whiskers are shown as circles. After unblinding, a post-hoc lab and

bioinformatic analysis was performed on DNA extracted from all obtained fecal samples using a

MANUSCRIPT

NucleoSpin 96 Soil kit (Macherey-Nagel) and randomly sheared into 350 bp fragments. Libraries

were constructed using NEBNext Ultra Library Prep Kit for Illumina (New England Biolabs) and

sequenced to at least 30 million read pairs per sample (2 x 150 bp paired-end Illumina sequencing).

Sequencing reads were filtered to remove human and low-quality reads, mapped to the Clinical

Microbiomics Human Gut 22M gene catalog, and summarised as a taxonomic relative abundance

table as described previously 21 . The involved parties were kept blinded for intervention during

analyses. Changes in relative abundances of taxa between visit 2 and the integral of later time-

points was tested using Wilcoxon rank sum test and corrected for multiple comparison using a

Bonferroni correction. Similarly, the Bray-Curtis distance between visit 2 and later time-points were

compared between the two arms (t-test).

6

Page 1 Page 2-3 Page 4 Page 5 Page 6 Page 7 Page 8 Page 9 Page 10 Page 11 Page 12 Page 13 Page 14 Page 15 Page 16 Page 17 Page 18 Page 19 Page 20 Page 21 Page 22 Page 23 Page 24 Page 25 Page 26 Page 27 Page 28 Page 29 Page 30 Page 31 Page 32 Page 33 Page 34 Page 35 Page 36 Page 37 Page 38 Page 39

atlantiaclinicaltrials.com

Powered by