Introduction

Inflammatory bowel diseases (IBD) have an ongoing, unmet medical need for new medicinal products with a different mode of action and a favourable safety profile. MH002 is a live biotherapeutic product (LBP) consisting of 6 commensal bacteria and is entering phase 3 clinical development as a therapeutic treatment developed to restore dysbiosis and depletion of butyric acid producing-bacteria that are associated with IBD. The non-pathogenic nature of the MH002 strains and absence of systemic bioavailability are expected to result in a particularly favourable safety profile.

Aims & Methods

To assess the safety of MH002 in ulcerative colitis (UC) and pouchitis (PC).
Safety data from 45 UC patients randomized (2:1) to double-blind treatment with 400mg MH002 or placebo daily for up to 16 weeks (MH002-UC-201), and from 14 PC patients treated open label with 400mg MH002 daily for 8 weeks (MH002-PC-201) were collected prospectively, the primary endpoint being occurrence of treatment emergent adverse events (AEs). UC patients were on stable 5-ASA treatment with or without low-dose corticosteroid throughout the study. Concomitant and recent antibiotic use was prohibited in all, as was treatment with advanced therapies. Placebo-treated UC patients who did not have a clinical response after 8 weeks were switched to MH002 treatment for 8 weeks.

Results

A total of 52 patients were exposed to MH002 with a total of about 136 patient-months of exposure. At baseline, the mean (range) modified Mayo score in UC patients was 5.76 (4.0-8.0) and the modified mean (range) Pouchitis Disease Activity Index score in PC patients was 6.91 (5.3-9.3). In UC patients, 20% were on stable, low dose corticosteroid, none was in the PC study. In both studies, no deaths, no severe AEs, no related serious AEs were reported in MH002-treated subjects, and no discontinuation occurred for an AE in MH002-treated participants. In particular, no bacteremia or sepsis was reported. Most AEs were mild and unrelated. Numerically less patients experienced an AE with MH002 than with placebo in UC.

	MH002-UC-201		MH002-PC-201	Total
Nr (%) of subjects	MH002 (n=31)	Placebo (n=14)	MH002 (n=14)	MH002 (n=45)​
with a severe Adverse Event (AE)	0	0	0	0
discontinuing for an AE	0	1 (7%)	0	0
with at least 1 AE	11 (35%)	8 (57%)	11 (79%)	22 (49%)
with a Serious AE (not related)	1 (3%)	1 (7%)	1 (7%)	2 (4%)
with a possibly related AE	1 (3%)	1 (7%)	3 (21%)	4 (9%)
with an AE of gastro-intestinal disorders	2 (6%)	4 (29%)	4 (29%)	6 (13%)

Possibly related AEs reported with MH002 included rash (UC) and abdominal pain, diarrhoea, flatulence and nausea (PC), while abdominal distension was reported with placebo (UC). Type 2 diabetes, arterial hypertension (MH002) and Clostridioides difficile infection (placebo) were the unrelated serious AEs that were reported. No clinically significant changes were observed during physical examination, vital signs assessment and in routine laboratory safety analyses (biochemistry, hematology and urinalysis) with MH002 or placebo. Efficacy data showed positive trends for clinical and endoscopic improvement and have been reported previously.

Conclusion

MH002 was well tolerated and safe in the first-in-disease phase 2 clinical trials conducted in UC and PC. MH002 may exhibit a particularly favourable safety profile if these results are confirmed in larger, late-stage clinical studies.

Disclosure

SV: financial support for research from AbbVie, J&J, Pfizer, Takeda and Galapagos; speakers’ and/or consultancy fees from AbbVie, Abivax, AbolerIS Pharma, AgomAb, Alimentiv, Arena Pharmaceuticals, AstraZeneca, BMS, Boehringer Ingelheim, Celgene, Cytoki Pharma, Dr Falk Pharma, Ferring, Galapagos, Genentech-Roche, Gilead, GSK, Hospira, Imidomics, Janssen, J&J, Lilly, Materia Prima, Mestag Therapeutics, MiroBio, Morphic, MRM Health, Mundipharma, MSD, Pfizer, Prodigest, Progenity, Prometheus, Robarts Clinical Trials, Surrozen, Takeda, Theravance, Tillots Pharma AG, VectivBio, Ventyx, Zealand Pharma.
SD: consultancy fees from AbbVie, Alimentiv, Allergan, Amgen, Applied Molecular Transport, AstraZeneca, Athos Therapeutics, Biogen, Boehringer Ingelheim, Bristol Myers Squibb, Celgene, Celltrion, Dr Falk Pharma, Eli Lilly, Enthera, Ferring Pharmaceuticals Inc., Gilead, Hospira, Inotrem, Janssen, Johnson & Johnson, Morphic, MSD, Mundipharma, Mylan, Pfizer, Roche, Sandoz, Sublimity Therapeutics, Takeda, Teladoc Health, TiGenix, UCB Inc., Vial, Vifor; reports lecture fees from Abbvie, Amgen, Ferring Pharmaceuticals Inc., Gilead, Janssen, Mylan, Pfizer, Takeda.
FM: served as speaker and received honoraria from: Abbvie, Arena, Biogen, Bristol-Myers Squibb, Falk, Ferring, Hospira, Janssen, Laboratórios Vitoria, Pfizer, Lilly, Merck Sharp & Dohme, Sandoz, Takeda, UCB, Vifor.
SP, SB & LH: employee, paid consultant and/or shareholder of MRM Health.

Introduction

Microplastic particles (MPs), defined as plastic fragments smaller than 5 mm, are pervasive pollutants that accumulate in ecosystems and the human food chain (1). Literature points at various health risks, including the induction of carcinogenesis (2). Emerging evidence from animal models (3) suggests that MPs may influence the gut microbiome, but the impact on the human gut microbiome remains poorly understood.

Aims & Methods

This study aimed to evaluate how various MP types influence gut microbial composition and metabolism using an ex vivo bioreactor model, with a focus on exploring potential carcinogenic effects arising from MP–microbiome interactions.
Stool samples from healthy donors were used to inoculate bioreactor cultures, which were maintained under anaerobic conditions for five days with daily nutrient feeding. These cultures were exposed to five common MP types—polystyrene (PS), polypropylene (PP), low-density polyethylene (LDPE), poly(methyl methacrylate) (PMMA), and polyethylene terephthalate (PET)—at concentrations mimicking both estimated human exposure (4) and higher doses to assess dose-dependence. Cell viability and total bacterial counts, as well as culture pH, were measured throughout the experiment. To further explore potential microbiome and metabolic alterations, 16S rRNA gene sequencing and targeted metabolomics were performed.

Results

MP exposure did not lead to significant changes in total or viable bacterial cell counts. However, MP-treated cultures exhibited a consistent and significant decrease in pH compared to controls, suggesting changes in microbial metabolic activity. Microbiome sequencing revealed plastic-type-dependent alterations in microbial composition, with certain bacterial taxa increasing or decreasing in abundance depending on the MP type. These shifts occurred across multiple genera within diverse families, including Lachnospiraceae, Oscillospiraceae, Enterobacteriaceae, and Ruminococcaceae, with the majority of changes occurring withing the phylum Bacillota. These compositional changes were accompanied by shifts in the metabolomic profiles, some of which correlated with the observed pH changes. Several plastic types induced changes in valeric acid levels, while individual MP types were associated with alterations in distinct metabolites such as uracil, lactic acid, and acetic acid.

Conclusion

This study demonstrates that MPs can affect gut microbial activity and metabolism without significantly altering overall bacterial abundance in the short term. The variation in microbiome response across plastic types highlights the complexity of MP-microbiome interactions. Changes in faecal pH are known to be associated with various gastrointestinal diseases (5). Interestingly, some of the MP-induced changes in microbial composition resembled patterns linked to diseases such as depression and colorectal cancer. In contrast, other changes showed patterns that differed from those typically associated with conditions like Parkinson’s disease and irritable bowel syndrome, highlighting the complexity of these associations and the dependency on the context. The bioreactor model provided a controlled environment for identifying direct MP-microbiome interactions; however, in humans, additional factors like diet, immune response, and individual microbiome variations play a role in determining long-term effects. Exposure duration may also play a critical role in shaping the extent and nature of microbiome alterations (6), underscoring the need for further research into the effects of chronic MP exposure and associated health risks.

References

(1) Mamun, A. A., Prasetya, T. A. E., Dewi, I. R., & Ahmad, M. (2023). Microplastics in human food chains: Food becoming a threat to health safety. The Science of the total environment, 858(Pt 1), 159834. https://doi.org/10.1016/j.scitotenv.2022.159834
(2) Kumar, R., Manna, C., Padha, S., Verma, A., Sharma, P., Dhar, A., Ghosh, A., & Bhattacharya, P. (2022). Micro(nano)plastics pollution and human health: How plastics can induce carcinogenesis to humans?. Chemosphere, 298, 134267. https://doi.org/10.1016/j.chemosphere.2022.134267
(3) Hirt, N., & Body-Malapel, M. (2020). Immunotoxicity and intestinal effects of nano- and microplastics: a review of the literature. Particle and fibre toxicology, 17(1), 57. https://doi.org/10.1186/s12989-020-00387-7
(4) Senathirajah, K., Attwood, S., Bhagwat, G., Carbery, M., Wilson, S., & Palanisami, T. (2021). Estimation of the mass of microplastics ingested - A pivotal first step towards human health risk assessment. Journal of hazardous materials, 404(Pt B), 124004. https://doi.org/10.1016/j.jhazmat.2020.124004
(5) Yamamura, R., Inoue, K. Y., Nishino, K., & Yamasaki, S. (2023). Intestinal and fecal pH in human health. Frontiers in Microbiomes, 2, 1192316.
(6) Yin, L., Yang, M., Teng, A., Ni, C., Wang, P., & Tang, S. (2025). Unraveling Microplastic Effects on Gut Microbiota across Various Animals Using Machine Learning. ACS nano, 19(1), 369–380. https://doi.org/10.1021/acsnano.4c07885

Introduction

Side effects, patient preference, and high costs may prompt cessation of anti-TNF therapy in ulcerative colitis (UC) patients in sustained remission. Given the risk of relapse, careful patient selection is essential. This study aimed to assess relapse rates and identify predictors of relapse following anti-TNF cessation in UC patients in remission.

Aims & Methods

A systematic search on cohort studies reporting relapse rates after cessation of anti-TNF therapy in ≥30 UC patients in clinical remission was performed 6 September 2024. Individual patient data was requested. We excluded patients with steroid use, ≤6 months anti-TNF therapy, ≤16 years, previous colectomy (n=4), golimumab as anti-TNF therapy (n=3) or without follow-up. Continuation of treatment with immunomodulators or mesalamine at cessation of anti-TNF was allowed. Relapse was defined as the need to (re)start treatment with steroids, anti-TNF or other advanced therapies, or the need for surgical treatment, as applied by the original author. Pooled relapse rates were computed using a random-effects model. A prediction model was built with a Cox proportional hazards model, performance assessed by iteratively leaving one study out (internal-external cross-validation, IECV). Missing covariate data were multiply imputed 100 times using the MICE algorithm. Variable selection was guided by three criteria: ≤20% missing data, a univariate p-value ≤ 0.2, and input from clinical experts.

Results

In total, 654 patients from 8 studies were included in the individual participant data meta-analysis. At anti-TNF cessation, median age was 40 years (IQR 31-51) with a median anti-TNF treatment of 18 months (IQR 18-36). The majority of patients were treated with infliximab (n = 610, 93%). Only 42 patients (9%) received a second-line anti-TNF therapy. The pooled relapse rate were 24% [95% CI 19-29%, I² = 44%] and 34% [95% CI 29-40%, I² = 39] at 1 and 2 years after cessation, respectively. Upon cessation, 211 patients (33%) continued mesalamine, 239 (39%) thiopurines, and 136 (22%) a combination of both. In our model, thiopurine therapy and older age at cessation were associated with a reduced risk of relapse, whereas immunomodulator refractoriness as an indication for starting anti-TNF therapy was associated with an increased risk. Other factors included in the prediction model were disease extent, type of anti-TNF therapy, history of anti-TNF therapy, fecal calprotectin level and endoscopic activity, leading to a c statistic of 0.55 [95% CI 0.51-0.59] (Table 1).

Table 1: Predictors of relapse in a pooled Cox regression analysis

Variable	Hazard Rate	95% Confidence interval
Age at discontinuation, per 10 years	0.80	0.72 – 0.88
Ulcerative colitis disease extent, Proctitis vs. extensive Left-sided vs. extensive	1.4 1.1	0.92 – 2.1 0.87 – 1.5
Type of anti-TNF therapy, adalimumab vs. infliximab	0.91	0.55 – 1.5
≥1 previous anti-TNF, yes vs. no	1.5	0.95 – 2.3
Indication anti-TNF therapy Immunomodulator refractoriness, yes vs no/unknown	1.5	1.1 – 2.2
Concomitant treatment after anti-TNF therapy, None vs thiopurine Mesalamine vs. thiopurine Combination of mesalamine and thiopurines vs thiopurine	1.8 1.2 1.2	1.1 – 3.0 0.86 – 1.6 0.81 – 1.7
Fecal Calprotectin (µg/g), per 100	1.1	0.95 – 1.3
Endoscopic activity, yes (mild, moderate or severe) vs. no	1.3	0.91 – 2.0

Conclusion

Approximately 1 in 4 patients experience a relapse of UC within 1 year after anti-TNF cessation, increasing to 1 in 3 within 2 years. Continuation of thiopurines is associated with a reduced risk. A prediction model for relapse was developed in this large IPD meta-analysis cohort and may aid in shared decision-making on anti-TNF cessation. However, its limited discriminative ability underscores the need for further refinement, incorporating data on microscopic remission and molecular and immunological markers to better capture deeper disease control.