A synbiotic preparation (SIM01) for post-acute COVID-19 syndrome in Hong Kong (RECOVERY): a randomised, double-blind, placebo-controlled trial

Background

Post-acute COVID-19 syndrome (PACS) affects over 65 million individuals worldwide but treatment options are scarce. We aimed to assess a synbiotic preparation (SIM01) for the alleviation of PACS symptoms.

Methods

In this randomised, double-blind, placebo-controlled trial at a tertiary referral centre in Hong Kong, patients with PACS according to the US Centers for Disease Control and Prevention criteria were randomly assigned (1:1) by random permuted blocks to receive SIM01 (10 billion colony-forming units in sachets twice daily) or placebo orally for 6 months. Inclusion criterion was the presence of at least one of 14 PACS symptoms for 4 weeks or more after confirmed SARS-CoV-2 infection, including fatigue, memory loss, difficulty in concentration, insomnia, mood disturbance, hair loss, shortness of breath, coughing, inability to exercise, chest pain, muscle pain, joint pain, gastrointestinal upset, or general unwellness. Individuals were excluded if they were immunocompromised, were pregnant or breastfeeding, were unable to receive oral fluids, or if they had received gastrointestinal surgery in the 30 days before randomisation. Participants, care providers, and investigators were masked to group assignment. The primary outcome was alleviation of PACS symptoms by 6 months, assessed by an interviewer-administered 14-item questionnaire in the intention-to-treat population. Forward stepwise multivariable logistical regression was performed to identify predictors of symptom alleviation. The trial is registered with ClinicalTrials.gov, NCT04950803.

Findings

Between June 25, 2021, and Aug 12, 2022, 463 patients were randomly assigned to receive SIM01 (n=232) or placebo (n=231). At 6 months, significantly higher proportions of the SIM01 group had alleviation of fatigue (OR 2·273, 95% CI 1·520–3·397, p=0·0001), memory loss (1·967, 1·271–3·044, p=0·0024), difficulty in concentration (2·644, 1·687–4·143, p<0·0001), gastrointestinal upset (1·995, 1·304–3·051, p=0·0014), and general unwellness (2·360, 1·428–3·900, p=0·0008) compared with the placebo group. Adverse event rates were similar between groups during treatment (SIM01 22 [10%] of 232 vs placebo 25 [11%] of 231; p=0·63). Treatment with SIM01, infection with omicron variants, vaccination before COVID-19, and mild acute COVID-19, were predictors of symptom alleviation (p<0·0036).

Interpretation

Treatment with SIM01 alleviates multiple symptoms of PACS. Our findings have implications on the management of PACS through gut microbiome modulation. Further studies are warranted to explore the beneficial effects of SIM01 in other chronic or post-infection conditions.

Funding

Health and Medical Research Fund of Hong Kong, Hui Hoy and Chow Sin Lan Charity Fund, and InnoHK of the HKSAR Government.

Translation

For the Chinese translation of the abstract see Supplementary Materials section.

Introduction

Post-acute COVID-19 syndrome (PACS) is a multisystemic condition comprising debilitating symptoms following SARS-CoV-2 infection.

1

According to the US Centers for Disease Control and Prevention (CDC), these symptoms persist beyond 4 weeks after the acute infection and could last for up to several years.

2

The incidence of PACS is approximately 10–30% and at least 65 million individuals worldwide are estimated to have PACS, with cases increasing daily.

1

3

PACS commonly affects the respiratory, gastrointestinal, and neuropsychiatric systems and is accompanied by fatigue.

1

3

Although several hypotheses have been proposed for the pathogenesis of PACS,

4

there is not yet an effective treatment for this complex condition.

Studies have shown that patients with PACS had significant alterations in the gut microbiota beyond 1 year after acute infection.

5

6

The gut microbiomes of patients with PACS were characterised by decreased microbial diversity and richness and reduced abundance of short-chain fatty-acid producing bacteria after SARS-CoV-2 clearance.

5

Furthermore, metagenomic sequencing of faecal samples showed depletion of several beneficial bacteria, such as Bifidobacterium adolescentis, in association with specific PACS symptoms.

6

Given the potentially debilitating symptoms of PACS and the emerging evidence of gut dysbiosis in association with various PACS symptoms, the role of microbiome modulation with probiotics or prebiotics in the management of PACS warrants investigation.

Research in context

Evidence before this study

We did a literature review on PubMed for randomised controlled trials published from database inception with the keywords “Post-acute COVID-19 Syndrome” or “PACS” or “Post-COVID” or “Long COVID” in combination with the terms “Gut Microbiome” or “Gut Microbiota” or “Probiotics” or “Synbiotics” or “Prebiotics” to identify relevant studies on June 1, 2023. To date, no randomised controlled trials have assessed the effect of gut microbiome modulation on post-acute COVID-19 syndrome (PACS). SIM01 is a synbiotic preparation of three lyophilised Bifidobacteria strains and three prebiotic compounds derived from our metagenomics dataset. In a pilot study, we showed that, in COVID-19 patients staying in hospital, SIM01 improved gut dysbiosis, hastened antibody formation, and decreased nasopharyngeal viral load and plasma pro-inflammatory immune markers.

Added value of this study

This randomised, double-blind, placebo-controlled trial showed that treatment with SIM01, which targets gut dysbiosis and potentially modifies the immune response, was effective in alleviating multiple symptoms of PACS. We identified favourable changes in the gut microbiome, including increased bacterial diversity and short-chain acid-producing bacteria and decreased resistome in the SIM01 group but not placebo group after 6 months of treatment, as plausible mechanisms to account for the clinical improvement.

Implications of all the available evidence

SIM01 is a safe and promising treatment for PACS, which is worth further confirmation in a pragmatic, multi-centre trial. Our findings on gut microbiota provided plausible mechanisms to account for the observed clinical benefits.

We did a randomised controlled trial on the efficacy in alleviating PACS of adult recovered patients using a synbiotic preparation SIM01 (RECOVERY trial). The synbiotic preparation (SIM01) is a micro-encapsulated lyophilised powder containing 20 billion colony-forming units of three bacterial strains, B adolescentis, Bifidobacterium bifidum, and Bifidobacterium longum with three prebiotic compounds including galacto-oligosaccharides, xylo-oligosaccharides, and resistant dextrin, which has been shown to promote the growth of these bacterial strains but also other probiotic strains.

7

8

The specific ratio of the three probiotic bacteria was decided based on the relative abundance of these species naturally present in the healthy Chinese population. We previously found that the relative abundance of several species, including B adolescentis and B longum, and the bacterial pathway of short-chain fatty acid production, were all significantly lower in the gut of COVID-19 patients compared with healthy individuals.

9

10

11

Decreased abundance of short-chain fatty acid-producing bacteria in the gut of patients with COVID-19 might represent one of the crucial mechanisms contributing to the gut–lung interaction and thereby disease severity in COVID-19. We previously reported prolonged loss of short-chain fatty acids after clearance of SARS-CoV2.

10

Using our metagenomics dataset, we found that B adolescentis, B bifidum, and B longum were the species that had the greatest positive correlations with the relative abundance of short-chain fatty acid-producing bacterial species, which could potentially boost immunity and prevent and treat respiratory infections (unpublished). In an open-label pilot study of patients with COVID-19 receiving treatment in hospital, SIM01 was associated with restoration of gut dysbiosis, hastened antibody formation, and decreased nasopharyngeal viral load and plasma pro-inflammatory immune markers.

7

A randomised, double-blind, placebo-controlled trial also reported that SIM01 was associated with reduced adverse clinical outcomes and improvement in sleep quality and mood among patients with type 2 diabetes and individuals older than 65 years.

8

Given that gut dysbiosis persisted in the post-acute phase of COVID-19, there is a potential to extend the use of microbiome modulation to patients with PACS.

In this study, we hypothesised that treatment with SIM01 would alleviate symptoms of PACS. We also evaluated changes in the gut microbiome and blood cytokines after treatment.

Methods

Study design

In this single-centre, randomised, double-blind, placebo-controlled trial, individuals were recruited from the Prince of Wales Hospital, a tertiary referral centre in Hong Kong, between June 25, 2021, and Aug 12, 2022.

This investigator-initiated study was approved by the local research ethics committee (The Joint Chinese University of Hong Kong–Hospital Authority New Territories East Cluster Clincal Research Ethics Committee, CRE-2021.259). This study was conducted following the Declaration of Helsinki and Guideline for Good Clinical Practice of the International Council for Harmonisation of Technical Requirements for Pharmaceuticals for Human Use. All participants provided written informed consent.

Participants

Eligible individuals were aged 18 years and older and had been previously diagnosed with COVID-19. SARS-CoV-2 infection must have been confirmed by either RT-PCR test or rapid antigen test with documentary proof from the local health authority. According to the epidemiological and genome sequencing records of our central registry,

12

omicron cases were defined as infections occurring in or after January 2022, and infections occurring in or before December 2021 were defined as cases of earlier variants. Individuals must have also had PACS symptoms according to the CDC criteria.

2

Eligibility also required patients to have at least one of 14 symptoms included in the post-acute COVID-19 syndrome 14-item improvement questionnaire (PACSQ-14) for 4 weeks or more after SARS-CoV-2 infection. Symptoms included in PACSQ-14 were fatigue, memory loss, difficulty in concentration, insomnia, mood disturbance, hair loss, shortness of breath, coughing, inability to exercise, chest pain, muscle pain, joint pain, gastrointestinal upset (defined as having at least one of the following symptoms: diarrhoea, constipation, abdominal pain, epigastric pain, bloating, nausea, vomiting, or acid reflux), and general unwellness. PACSQ-14 comprised a standardised framework assessing PACS symptoms by trained interviewers (appendix 2 pp 1–5). The questionnaire was developed by a group of physicians, biostatisticians, and public health professionals, and validated by an independent expert panel for face and content validity. PACS symptoms were selected and included in the questionnaire based on global data available during the early pandemic combined with clinical observations in local settings.

13

Individuals were excluded if they were immunocompromised, pregnant or breastfeeding, unable to receive oral fluids, or if they had received gastrointestinal surgery in the 30 days before randomisation.

Randomisation and masking

Participants were randomly assigned to treatment groups centrally by random permuted blocks using a web-based service to receive either an oral synbiotic preparation (SIM01) or placebo (vitamin C) in a 1:1 ratio for 6 months. Allocation concealment was done using sequentially numbered, opaque, sealed envelopes prepared before the trial by an independent staff member of the study. SIM01 consisted of 20 billion colony-forming units of three bacterial strains (ie, B adolescentis, B bifidum, and B longum) with three prebiotic compounds (ie, galacto-oligosaccharides, xylo-oligosaccharides, and resistant dextrin), administered as 10 billion colony-forming units in sachets twice daily. Placebo consisted of a very low-dose vitamin C (1 mg in sachets twice daily) and an inert substance made of starch filler, flavour, and colouring. The contents of the sachet containing synbiotics or placebo, which had been maintained at 0–25°C, were sprinkled into room temperature drinks or poured into the mouth directly. Sachets containing synbiotics or placebo were identical in appearance, smell, and weight and dispensed to participants at the baseline visit by an independent researcher who was masked to treatment allocations. Study products were manufactured in Good Manufacturing Practice-accredited facilities and were stored and dispensed in the research pharmacy of the Prince of Wales Hospital.

Procedures

At the baseline visit, attending investigators assessed the eligibility of individuals based on inclusion and exclusion criteria, and performed the clinical assessments. To minimise inter-observer variation, the three masked interviewers (senior research staff with biomedical backgrounds) who delivered the PACSQ-14 questionnaire during the study received standardised training before study commencement. The trained interviewers conducted structured interviews using a set of standardised questions to assess whether the symptoms were present at baseline (defined as symptoms affecting an individual's ability to carry out daily activities). Targeted follow-up questions were asked to elicit supplementary information whenever appropriate. Demographic information, including age and sex, and history of COVID-19 including severity based on WHO classification and vaccination status, were obtained from a territory-wide electronic clinical management system by a clinician. At baseline, quality of life and physical activity were assessed by the trained interviewers using a visual analogue scale, with scores ranging from 0–100, and International Physical Activity Questionnaire.

14

Blood tests for complete blood count, liver function, and renal function were done, and plasma samples were used for inflammatory marker profiling using human inflammation panel 1 (known as 13-plex) of LEGENDplex (BioLegend, San Diego, CA, USA; appendix 2 p 6). Faecal samples were also collected for microbiome profiling, which was done using metagenomic sequencing (appendix 2 p 6).

Participants were followed up at 3 months and 6 months by a clinician in a research clinic, to assess compliance and adverse events. Compliance was assessed by counting the number of unused study products returned, searching the territory-wide electronic prescription database for other drugs prescribed to the participants during the study period, and direct questioning of usage of unprescribed medication, non-study probiotics, or off-label therapies (eg, herbal medicine or nutritional supplements). Faecal and blood samples were collected at 6 months for metagenomic gut microbiome analysis, complete blood count, liver and renal function tests, and inflammatory cytokine profiling. Use of other probiotics, prebiotics, or antibiotics during the study period were considered as drug violations. The trained interviewers repeated the PACSQ-14 questionnaire at 6 months to assess symptom alleviation. Quality of life and physical activity were assessed with the visual analogue scale and International Physical Activity Questionnaire

14

by the trained interviewers at 6 months.

An independent Data and Safety Monitoring Committee (DSMC) consisting of two physicians and a biostatistician regularly reviewed data on adverse events and monitored patient safety. If serious adverse events associated with study intervention occurred, the trial could be terminated early by the Trial Steering Committee upon recommendation by the DSMC. An independent Adjudication Committee consisting of three physicians masked to treatment allocation was also set up to assess endpoints and attribute causality of any adverse event during the trial.

Outcomes

The primary outcome was the alleviation of PACS symptoms by 6 months. Alleviation of PACS symptoms (alleviated vs not alleviated) was defined as reduction in the severity of symptoms leading to improvement in activities of daily living. Evaluation of the primary outcome was done via the PACSQ-14 questionnaire. Symptom alleviation for a specific domain was achieved if no more than one response to the set of questions in that domain was negative. Symptom alleviation was considered clinically meaningful if the study participants perceived the change could positively affect the activities of their daily living (appendix 2 pp 1–5). The intention-to-treat analysis included all randomly assigned patients who had taken at least one dose of the study product. Prespecified secondary outcomes included quality of life assessment at 6 months, changes in the faecal microbiota composition, changes in microbiome resistome and microbiome functions (assessed using shotgun metagenomic sequencing), changes in inflammatory profiles (assessed by multiplex ELISA against 13 different human cytokines), and rates of adverse events.

Statistical analysis

Assuming that 70% of the patients in the SIM01 group and 50% in the placebo group achieved the primary outcome and that the drop-out rate was 25%, a total of 224 patients per group was required to achieve a power of 80% at a two-sided significance level of 0·0036 after Bonferroni correction. Bonferroni correction was for multiple comparisons across 14 symptoms involved in the primary outcome. All analyses were specified a priori. The intention-to-treat analysis included data from all randomly assigned patients who had taken at least one dose of the study product. The last observation carried forward method was used to impute missing data of the primary outcome.

For both the SIM01 and placebo groups, symptom outcomes assessed at 3 months were carried forward if data were missing at 6 months. We assumed no change in symptoms from baseline to 6 months for any individual who withdrew or was lost to follow-up before the follow-up visit at 3 months. Relative benefit increase for each symptom after SIM01 supplementation was calculated by the difference in the proportion of individuals with symptom alleviation between groups divided by the proportion of placebo group, expressed in percentage. Bonferroni correction was applied to adjust for multiple comparisons. A multivariable logistical regression analysis using the forward stepwise method was done to identify possible factors for symptom alleviation. We further correlated microbial changes and response to specific symptom by use of the MaAsLin 2 function in R. All statistical analyses were done in SPSS (version 26.0) or R (version 4.0.3; appendix 2 p 7).

This trial is registered with ClinicalTrials.gov, NCT04950803.

Role of the funding source

The funders of the study had no role in the study design, data collection, data analysis, data interpretation, or writing of the report.

Results

Between June 25, 2021, and Aug 12, 2022, 700 individuals were screened for eligibility. Among them, 237 individuals were excluded (figure 1), leaving a final total of 463 participants. 232 individuals were allocated to SIM01 and 231 were allocated to the placebo. Participants were randomly assigned to treatment groups at a median duration of 4 months (IQR 3–11) after COVID-19 infection. 142 (31%) of 463 individuals stayed in hospital and 59 (13%) took antibiotics during acute COVID-19. The proportion of individuals who completed follow-up were 93% (215 of 232) in the SIM01 group and 91% (210 of 231) in the placebo group at 3 months and 88% (204 of 232) in the SIM01 group and 86% (199 of 231) in the placebo group at 6 months. Baseline characteristics of patients were well-matched between SIM01 and placebo groups (table 1). Participants presented with a mean of 8·6 (SD 3·8) of 14 PACS symptoms included in PACSQ-14.

Figure 1Trial profileShow full caption • View Large Image • Figure Viewer • Download Hi-res image • Download (PPT)

Table 1Baseline characteristics

Data are n (%) or mean (SD) unless otherwise specified.

• Cardiovascular diseases: coronary heart disease, stroke, and pulmonary embolism; mood disturbances: depression, bipolar disorder, anxiety, panic disorder, adjustment disorder, and sleep disturbance; thyroid dysfunctions: hypothyroidism, hyperthyroidism, thyroid nodules, and thyroid tumour; musculoskeletal conditions: gout, osteoarthritis, osteoporosis, chronic back pain, and chronic neck pain; renal disorders: chronic kidney disease, kidney stones, overactive bladder, urinary incontinence, and acontractile bladder; and functional gastrointestinal disorders: irritable bowel syndrome, and gastroesophageal reflux disease.

† Gastrointestinal upset: at least one of the following symptoms: diarrhoea, constipation, abdominal pain, epigastric pain, bloating, nausea, vomiting, and acid reflux.

‡ Off-label therapies consisted of herbal medicine and nutritional supplements.

• Open table in a new tab

At 6 months, a significantly higher proportion of individuals who received SIM01 had alleviations in fatigue (OR 2·273, 95% CI 1·520–3·397, p=0·0001), memory loss (1·967, 1·271–3·044, p=0·0024), difficulty in concentration (2·644, 1·687–4·143, p<0·0001), gastrointestinal upset (1·995, 1·304–3·051, p=0·0014), and general unwellness (2·360, 1·428–3·900, p=0·0008) compared with the placebo group, after adjusting for multiple comparisons (figure 2). The relative benefit increase after SIM01 were 47% for fatigue, 56% for memory loss, 62% for difficulty in concentration, 30% for gastrointestinal upset, and 31% for general unwellness. More patients in the SIM01 group compared with the placebo group had alleviation in joint pain, ability to exercise, shortness of breath, insomnia, muscle pain, coughing, hair loss, chest pain, and mood disturbance, although these results were not significant after Bonferroni correction (ie, p>0·0036). On the multivariable logistical regression analysis, treatment with SIM01, infection with omicron variants, vaccination before COVID-19, and mild or acute COVID-19 were independent predictors of alleviation of PACS symptoms (table 2). Because the last observation carried forward method expected the data to be missing completely at random, we repeated the analyses using multiple imputation and observed consistent results for the primary outcome (appendix 2 p 12). Visual analogue scale scores on quality of life at 6 months had a mean of 76·0 (SD 12·0) for the SIM01 group and 74·5 (12·3) for the placebo group (p=0·17). For the post-hoc analysis of physical activity at 6 months, there was no significant difference in the total metabolic equivalent of task-minutes per week between the two groups (SIM01 median 1646·3, IQR 815·6–2899·5; placebo 1902·0, 956·0–3290·0; p=0·37).

Figure 2Proportion of PACS symptoms alleviation by 6 monthsShow full caption • View Large Image • Figure Viewer • Download Hi-res image • Download (PPT)

Table 2Independent predictors of PACS alleviation identified by forward stepwise multivariate analysis

OR=odds ratio.

• Gastrointestinal upset: at least one of the following symptoms: diarrhoea, constipation, abdominal pain, epigastric pain, bloating, nausea, vomiting, and acid reflux.
• Open table in a new tab

22 (10%) of 232 individuals in the SIM01 group and 25 (11%) of 231 individuals in the placebo group had adverse events during the treatment period (p=0·63; table 3). All adverse events were deemed unlikely to be associated with the study products as assessed by an independent safety adjudication committee. The most common gastrointestinal adverse events included diarrhoea (SIM01 2 [1%] of 232 vs placebo 5 [2%] of 231, p=0·25), bloating (1 [0%] vs 2 [1%], p=0·56), epigastric pain (1 [0%] vs 1 [0%], p=1), abdominal pain(1 [0%] vs 0, p=0·32), and flatulence (1 [0%] vs 0, p=0·32), which were all mild and self-limiting.

Table 3Adverse events, study compliance, vaccination, and re-infection during study period

Data are n (%).

• Off-label therapies consisted of herbal medicine and nutritional supplements.
• Open table in a new tab

Metagenomic analysis of faecal samples showed increased bacteria diversity (p=0·0019) and bacteria richness (observed number of species; p=0·051) in the SIM01 group but no difference in diversity (p=0·61) and richness (p=0·44) in the placebo group (appendix 2 p 8) at 6 months compared with baseline. The relative abundance of several bacteria from the Bifidobacterium genus (B bifidum, B adolescentis, B longum, Bifidobacterium pseudocatenulatum) as well as Roseburia intestinalis, Roseburia hominis, Faecalibacterium prausnitzii, and Akkermansia muciniphila were significantly increased whereas Ruminococcus gnavus and some from the Klebsiella genus were significantly decreased in the faecal samples at 6 months after SIM01 treatment (false discovery rate <0·05). These changes were not found in the placebo group (appendix 2 p 8). At 6 months, the faecal microbiota showed a higher richness and distinct composition in the SIM01 group compared with the placebo group. At the species level, the faecal microbiota of the SIM01 group was characterised by enrichment of B pseudocatenulatum and B longum, with depletion of Klebsiella pneumoniae, Klebsiella variicola, and Parabacteroides merdae, compared with the placebo group (FDR<0·05, appendix 2 p 8).

There was a significant decrease in the total number (antimicrobial resistant gene type p=0·028; antimicrobial resistant gene subtype p=0·0014) and relative abundance (p=0·0003) of the antimicrobial resistant genes in the gut microbiome of the SIM01 group, but not the placebo group (appendix 2 p 9). At 6 months, the total number and relative abundance of antimicrobial resistant genes in the gut microbiome of the SIM01 group were all significantly lower than that of the placebo group (appendix 2 p 9). The SIM01 group also exhibited significant changes in microbiome functions at 6 months compared with baseline (p<0·001, appendix 2 p 10). The relative abundance of three microbial metabolic pathways involved in short-chain fatty acid production, including pyruvate fermentation to acetone (PWY-6588), pyruvate fermentation to butanoate (CENTFERM-PWY), and pyruvate fermentation to acetate and lactate (PWY-5100), significantly increased in the SIM01 group, but not placebo group, after 6 months of treatment (FDR<0·05, appendix 2 p 10). At 6 months, principal coordinates analysis could significantly separate the microbial pathways between the SIM01 group and placebo group (p=0·017; appendix 2 p 10). The faecal microbiome function profiles of the SIM01 group were characterised as enrichment of several pathways involved in short-chain fatty acid production with depletion of the superpathway of fatty acid biosynthesis (PWY0-881) and the urea cycle, when compared with placebo group at 6 months (FDR<0·05; appendix 2 p 10). Correlation of microbial changes and response to a specific symptom showed that alleviation in different symptoms was associated with distinct changes in the microbiome at both the compositional and functional levels (FDR<0·05; appendix 2 p 11). For instance, Bifidobacterium adolescentis showed positive correlations with alleviation in fatigue, gastrointestinal upset, and memory loss. We also found that alleviation in fatigue and general unwellness correlated with an elevated relative abundance of Bifidobacterium bifidum, whereas alleviation in difficulty in concentration correlated with a positive shift in Bifidobacterium longum.

Discussion

We set out to test the hypothesis that treatment with SIM01 would alleviate symptoms of PACS. We also evaluated changes in the gut microbiome after treatment. This double-blind, randomised controlled trial showed that treatment with SIM01 was effective in alleviating multiple symptoms of PACS. We found that SIM01 led to improved gut microbiota composition by promoting the abundance of beneficial bacteria and suppressing that of pathogenic bacteria associated with PACS, such as those of the Klebsiella genus. Microbiome pathway analysis showed that SIM01 was associated with a decrease in the relative abundance of the microbial species involved in the enhancement of urea cycle pathway, which is also known to be associated with PACS. Besides, gut microbiota composition has been linked to host immune response and blood cytokine profiling. Overall, our findings on gut microbiota provided plausible mechanisms to account for the observed clinical benefits of SIM01 in patients with PACS.

We found that SIM01 led to alleviation in gastrointestinal upset. Gastrointestinal symptoms following SARS-CoV-2 infection resembled a condition known as post-infectious irritable bowel syndrome (PI-IBS), characterised by chronic GI symptoms, such as abdominal pain and diarrhoea, following infectious gastroenteritis.

15

In a study comparing COVID-19 patients and non-infected controls, it was found that COVID-19 patients had significantly higher rates of irritable bowel syndrome (IBS) according to Rome IV criteria than controls (3·2% vs 0·5%), at 12-month follow-up.

16

In COVID-19 patients, an earlier study showed that incident PI-IBS occurred in 3% of those with persisting GI symptoms.

17

Probiotics have been extensively used in the treatment of IBS.

18

Persistent fatigue after COVID-19 is similar to a chronic condition called myalgic encephalomyelitis/chronic fatigue syndrome (ME/CFS).

19

ME/CFS is an illness that causes prolonged tiredness with largely unknown pathogenesis.

19

Metagenomic analyses of ME/CFS faecal samples have shown significant gut dysbiosis with altered microbial butyrate biosynthesis and a decrease in plasma butyrate.

20

Similarly, we have reported depletion of butyrate-producing bacterial species in faecal samples of patients with PACS.

5

6

As SIM01 has been shown to be associated with an increase in the abundance of butyrate-producing species,

8

this might be one of the possible mechanisms by which fatigue was alleviated in this study.

The microbiota–gut–brain axis has been extensively studied in both animal and human studies.

21

The gut microbiota communicates with the brain bidirectionally via multiple routes, such as the vagus nerve and the enteric nervous system, immune system, tryptophan metabolism, and microbial short-chain fatty acids,

21

which might in part explain some of the neuropsychiatric symptoms after the acute infection. A randomised controlled trial has shown the effects of probiotics in improving memory and cognition in stressed adults.

22

In mouse models, probiotics have been shown to reverse the memory decline associated with gut dysbiosis

23

and improve memory deficits.

24

The exact mechanisms of SIM01 in alleviating the neuropsychiatric symptoms of PACS including memory loss and difficulty in concentration, requires further investigation.

Antibiotics are an important confounder and are known to have a profound effect on the gut microbiome. Most (87%) individuals in our cohort did not take antibiotics. We found that the gut microbiome of patients who had not had antibiotics also showed decreased bacteria diversity and reduced abundance of different Bifidobacterium strains, suggesting that SARS-CoV-2 infection might also have an effect on the gut microbiome in these cases. The exact cause for this finding is unknown but could be the effect of SARS-CoV-2 or immune dysfunction in COVID-19. Although discerning whether the gut microbiome is causal of immune dysfunction in COVID-19 is difficult in human studies, animal studies have shown that the virus could drive changes in the gut microbiome.

25

26

Importantly, gut microbiome changes are also detectable in asymptomatic infants with SARS-CoV-2 who had not been exposed to antibiotics including lowered Bifidobacterium bifidum.

27

Prebiotic compounds in SIM01, including galacto-oligosaccharides, xylo-oligosaccharides, and resistant dextrin, can contribute to beneficial compositional shifts in gut microbiome composition. Human studies have shown that galacto-oligosaccharides can increase the abundance of different Bifidobacteria species in patients with irritable bowel syndrome, whereas xylo-oligosaccharides and resistant dextrin were shown to promote the growth of Bifidobacterium and Lactobacillus strains in in-vivo and in-vitro studies.

28

29

30

Resistant dextrin was also able to reduce the abundance of Clostridium and Bacteroides strains.

29

Importantly, our previous metagenomic analysis suggested that Bifidobacteria and Lactobacilli species were negatively correlated whilst Clostridium and Bacteroides species were positively correlated with different symptoms of PACS.

5

6

Taken altogether, these data support the synergistic effect of probiotics and prebiotics in modulating the gut microbiome. Furthermore, adverse events rates were low for SIM01.

The present study did not identify a significant difference in quality of life and physical activity between the two groups at 6 months. Quality of life indices commonly cover a broad range of domains, including personal, social, political, cultural, economic, and environmental aspects. Time spent on physical activity is determined by multiple factors in addition to physical health. Our primary outcome focused mainly on physical symptoms, which only partly contribute to the assessment of quality of life and time spent on physical activity.

The strengths of this study included a double-blind trial design, a large sample size, the inclusion of a diverse study population that mimicked the real-world scenario, and a high completion rate. The PACS symptoms included were selected on the basis of common symptoms reported in previous studies and symptom assessment was done by trained interviewers. Furthermore, the clinical efficacy of SIM01 correlated with gut microbial changes based on serial stool microbiome analyses. Our trial has some limitations. First, assessment of treatment outcome for PACS is challenging because symptoms of PACS are common and can be multifactorial. To ensure that we dealt with a coherent post-COVID cohort, we adopted measures including an objective diagnosis of COVID-19, fulfilment of the CDC definition of PACS, and the use of a locally validated questionnaire to assess PACS symptoms with substantial effect on activities of daily living. Using these stringent criteria, we found that most of our patients presented with multiple PACS symptoms. As PACS is a new entity and includes many symptoms affecting different organs, robust and validated assessment tools for individual symptoms were not available. Future studies should focus on the development of globally validated assessment tools for PACS, and objective endpoints and surrogate markers to measure therapeutic interventions. Second, although this early phase trial showed promising clinical outcomes supported by plausible biological mechanisms, further confirmation in a multi-centre, ethnically diverse trial with globally validated assessment tools is warranted. As our analysis on cytokine profiles did not yield any significant result (appendix pp 15–16), alternate mechanisms underlying the improved symptoms should also be explored in future studies. Third, two-thirds of our patients recovered in the community. The extent to which SIM01 is beneficial to all patients with varying severity during acute COVID-19 could only be explored by post-hoc subgroup analysis. On post-hoc analysis of patients who stayed in hospital during acute COVID-19, we observed a similar numerical trend in favour of SIM01 treatment, although the differences did not reach statistical significance after Bonferroni correction, possibly due to an insufficient sample size. Fourth, we adopted the CDC definition of PACS though the WHO definition might be more widely applicable outside the USA. If we were to adopt the WHO definition, 80% (369 of 463) of our patients would have been eligible for inclusion. On post-hoc analysis, the SIM01 group still significantly outperformed the placebo group in seven symptoms (two more symptoms showed improvement compared with the current cohort, including shortness of breath and joint pain). Therefore, our results have further confirmed the clinical benefits of SIM01 irrespective of the definition of PACS. To date, studies using probiotic intervention to alleviate PACS are sparse, and the optimal dose, duration, or type of bacteria requires further interventional trials. Finally, although faecal metagenomic sequencing confirmed the modulatory effects of SIM01 on gut microbiome at both the compositional and functional levels, we did not take tissue biopsy via colonoscopy to assess the mucosa microbiome as this would entail extra risk for our patients and would have reduced study compliance.

In conclusion, we found that our synbiotic preparation, SIM01, alleviated multiple symptoms of PACS at 6 months in adult patients after acute COVID-19 infection. Our findings support the potential of gut microbiome-targeted therapeutics for PACS in the post-COVID era.

For the randomisation website used see https://www.sealedenvelope.com

Contributors

RIL and QS conceived the study, accessed and verified all the data, and prepared the manuscript. ISFL, MCSW, LHSL, and SWHC contributed to participant recruitment and clinical assessment. JYLC contributed to clinical data management and study monitoring. HMT and CKPM contributed to cytokines profiling. YKT provided important comments on statistical analysis. CPC contributed to sample collection and biobank management. MKTL contributed to sample processing and metagenomic sequencing. GTYY and PKC provided important assistance during the clinical trial. FKLC and SCN designed the study and contributed to data interpretation and manuscript writing. All authors read, revised, and had final responsibility for the decision to submit the manuscript for publication and had full access to all the data in the study.

Data sharing

The metagenomic sequencing data generated in this study have been deposited in the NCBI Sequence Read Archive database under accession code PRJNA995807. De-identified clinical metadata is available upon appropriate request to the corresponding author.

Declaration of interests

All authors have completed the Unified Competing Interest form (available on request to the corresponding author). MCSW is an advisory committee member of Pfizer, an external expert of GlaxoSmithKline, a member of the advisory board of AstraZeneca, and has been paid consultancy fees for providing advice on research. LHSL is supported by grants from the Health and Medical Research Fund, General Research Fund, and Direct Grant for Research; received honoraria as a speaker for Olympus, Boston Scientific, and GenieBiome; travel grants from Olympus, Pfizer, and AbbVie; and received research product support from GenieBiome for his other research. FKLC is Board Member of the Chinese University of Hong Kong (CUHK) Medical Centre; is a co-founder, non-executive Board Chairman, non-executive scientific adviser, and shareholder of GenieBiome; receives patent royalties from GenieBiome through his affiliated institution, CUHK; has received fees as an adviser and honoraria as a speaker for Eisai, AstraZeneca, Pfizer, Takeda Pharmaceutica, and Takeda (China) Holdings; and received consulting fees from the American Gastroenterological Association Institute. SCN has served as an advisory board member for Pfizer, Ferring, Janssen, and AbbVie and received honoraria as a speaker for Ferring, Tillotts, Menarini, Janssen, AbbVie, and Takeda; received research grants from Olympus, Ferring, and AbbVie through her affiliated institution, CUHK; is a scientific co-founder and shareholder of GenieBiome; and receives patent royalties from GenieBiome, through CUHK. QS, HMT, FKLC, and SCN are named inventors of patent applications held by the CUHK and Microbiota I-Center that cover the therapeutic and diagnostic use of microbiome. FKLC and SCN are named inventors of a patent application related to SIM01 for improving immunity held by the CUHK, which is licensed exclusively to GenieBiome. No patents have been filed or are intended to be filed for the findings presented in this manuscript. All other authors declare no competing interests.

Acknowledgments

GenieBiome did not provide any financial support for this study. We thank Tak Chiu Wu, Man Yee Chu, Wai Shing Leung, and Shuk Ying Chan for case referral. This study was funded by the Health and Medical Research Fund under the Food and Health Bureau of the Government of the Hong Kong Special Administrative Region, and Hui Hoy and Chow Sin Lan Charity Fund. RIL, QS, JYLC, HMT, CPC, MKTL, FKLC and SCN are supported by InnoHK, the Government of the Hong Kong Special Administrative Region. RIL is supported by the Hong Kong PhD Fellowship Scheme (HKPFS) of the Research Grants Council of Hong Kong. This research has been conducted using the CU-Med Biobank Resource under Request ID R20231025.

Supplementary Materials

• Download .pdf (.25 MB) Help with pdf files Chinese translation of the abstract
• Download .pdf (4.85 MB) Help with pdf files Supplementary appendix 2

References

1. 1.Long COVID: major findings, mechanisms and recommendations.Nat Rev Microbiol. 2023; 21: 133-146View in Article
• Davis HE
• McCorkell L
• Vogel JM
• Topol EJ
• Scopus (484)
• PubMed
• Crossref
• Google Scholar
2. 2.Post-COVID conditions: information for healthcare providers.https://www.cdc.gov/coronavirus/2019-ncov/hcp/clinical-care/post-covid-conditions.htmlDate: 2023Date accessed: May 31, 2023View in Article
• Centers for Disease Control and Prevention
• Google Scholar
3. 3.Post-acute COVID-19 syndrome.Nat Med. 2021; 27: 601-615View in Article
• Nalbandian A
• Sehgal K
• Gupta A
• et al.
• Scopus (2274)
• PubMed
• Crossref
• Google Scholar
4. 4.Gut microbiota in COVID-19: key microbial changes, potential mechanisms and clinical applications.Nat Rev Gastroenterol Hepatol. 2023; 20: 323-337View in Article
• Zhang F
• Lau RI
• Liu Q
• Su Q
• Chan FKL
• Ng SC
• Scopus (38)
• PubMed
• Crossref
• Google Scholar
5. 5.Gut microbiota dynamics in a prospective cohort of patients with post-acute COVID-19 syndrome.Gut. 2022; 71: 544-552View in Article
• Liu Q
• Mak JWY
• Su Q
• et al.
• Scopus (170)
• PubMed
• Crossref
• Google Scholar
6. 6.Post-acute COVID-19 syndrome and gut dysbiosis linger beyond 1 year after SARS-CoV-2 clearance.Gut. 2023; 72: 1230-1232View in Article
• Su Q
• Lau RI
• Liu Q
• Chan FKL
• Ng SC
• Scopus (11)
• PubMed
• Crossref
• Google Scholar
7. 7.Gut microbiota-derived synbiotic formula (SIM01) as a novel adjuvant therapy for COVID-19: an open-label pilot study.J Gastroenterol Hepatol. 2022; 37: 823-831View in Article
• Zhang L
• Xu Z
• Mak JWY
• et al.
• Scopus (26)
• PubMed
• Crossref
• Google Scholar
8. 8.Effects of gut microbiome modulation on reducing adverse health outcomes among elderly and diabetes patients during the COVID-19 pandemic: a randomised, double-blind, placebo-controlled trial (IMPACT Study).Nutrients. 2023; 151982View in Article
• Wong MCS
• Zhang L
• Ching JYL
• et al.
• Scopus (1)
• Crossref
• Google Scholar
9. 9.Gut microbiota composition reflects disease severity and dysfunctional immune responses in patients with COVID-19.Gut. 2021; 70: 698-706View in Article
• Yeoh YK
• Zuo T
• Lui GC
• et al.
• Scopus (661)
• PubMed
• Crossref
• Google Scholar
10. 10.Prolonged impairment of short-chain fatty acid and L-isoleucine biosynthesis in gut microbiome in patients with COVID-19.Gastroenterology. 2022; 162 (61.e4): 548View in Article
• Zhang F
• Wan Y
• Zuo T
• et al.
• Scopus (101)
• PubMed
• Summary
• Full Text
• Full Text PDF
• Google Scholar
11. 11.Alterations in gut microbiota of patients with COVID-19 during time of hospitalization.Gastroenterology. 2020; 159: 944-955.e8View in Article
• Zuo T
• Zhang F
• Lui GCY
• et al.
• Scopus (922)
• PubMed
• Summary
• Full Text
• Full Text PDF
• Google Scholar
12. 12.Resurgence of omicron BA.2 in SARS-CoV-2 infection-naive Hong Kong.Nat Commun. 2023; 142422View in Article
• Xie R
• Edwards KM
• Adam DC
• et al.
• Google Scholar
13. 13.Long covid-mechanisms, risk factors, and management.BMJ. 2021; 374n1648View in Article
• Crook H
• Raza S
• Nowell J
• Young M
• Edison P
• PubMed
• Google Scholar
14. 14.International physical activity questionnaire: 12-country reliability and validity.Med Sci Sports Exerc. 2003; 35: 1381-1395View in Article
• Craig CL
• Marshall AL
• Sjöström M
• et al.
• Scopus (13129)
• PubMed
• Crossref
• Google Scholar
15. 15.Post-infectious irritable bowel syndrome.Curr Gastroenterol Rep. 2017; 19: 56View in Article
• Lee YY
• Annamalai C
• Rao SSC
• Scopus (36)
• PubMed
• Crossref
• Google Scholar
16. 16.Post COVID-19 irritable bowel syndrome.Gut. 2022; (gutjnl-2022-328483.)View in Article
• Marasco G
• Cremon C
• Barbaro MR
• et al.
• PubMed
• Google Scholar
17. 17.Persisting gastrointestinal symptoms and post-infectious irritable bowel syndrome following SARS-CoV-2 infection: results from the Arizona CoVHORT.Epidemiol Infect. 2022; 150: e136View in Article
• Austhof E
• Bell ML
• Riddle MS
• et al.
• Scopus (10)
• PubMed
• Crossref
• Google Scholar
18. 18.Probiotics in irritable bowel syndrome: an up-to-date systematic review.Nutrients. 2019; 112048View in Article
• Dale HF
• Rasmussen SH
• Asiller OO
• Lied GA
• Scopus (82)
• Crossref
• Google Scholar
19. 19.Analysis of post COVID-19 condition and its overlap with myalgic encephalomyelitis/chronic fatigue syndrome.J Adv Res. 2022; 40: 179-196View in Article
• Sukocheva OA
• Maksoud R
• Beeraka NM
• et al.
• Scopus (47)
• PubMed
• Crossref
• Google Scholar
20. 20.Multi-'omics of gut microbiome-host interactions in short- and long-term myalgic encephalomyelitis/chronic fatigue syndrome patients.Cell Host Microbe. 2023; 31 (87.e5): 273View in Article
• Xiong R
• Gunter C
• Fleming E
• et al.
• Scopus (0)
• PubMed
• Summary
• Full Text
• Full Text PDF
• Google Scholar
21. 21.The microbiota–gut–brain axis.Physiol Rev. 2019; 99: 1877-2013View in Article
• Cryan JF
• O'Riordan KJ
• Cowan CSM
• et al.
• Scopus (1879)
• PubMed
• Crossref
• Google Scholar
22. 22.Probiotic Lactobacillus plantarum P8 alleviated stress and anxiety while enhancing memory and cognition in stressed adults: a randomised, double-blind, placebo-controlled study.Clin Nutr. 2019; 38: 2053-2064View in Article
• Lew LC
• Hor YY
• Yusoff NAA
• et al.
• Scopus (145)
• PubMed
• Summary
• Full Text
• Full Text PDF
• Google Scholar
23. 23.Probiotics protect against gut dysbiosis associated decline in learning and memory.J Neuroimmunol. 2020; 348577390View in Article
• Roy Sarkar S
• Mitra Mazumder P
• Banerjee S
• Scopus (22)
• PubMed
• Summary
• Full Text
• Full Text PDF
• Google Scholar
24. 24.Probiotics modulate the microbiota-gut-brain axis and improve memory deficits in aged SAMP8 mice.Acta Pharm Sin B. 2020; 10: 475-487View in Article
• Yang X
• Yu D
• Xue L
• Li H
• Du J
• Scopus (168)
• PubMed
• Crossref
• Google Scholar
25. 25.Alteration of the gut microbiota following SARS-CoV-2 infection correlates with disease severity in hamsters.Gut Microbes. 2022; 142018900View in Article
• Sencio V
• Machelart A
• Robil C
• et al.
• Google Scholar
26. 26.SARS-CoV-2 infection in nonhuman primates alters the composition and functional activity of the gut microbiota.Gut Microbes. 2021; 13: 1-19View in Article
• Sokol H
• Contreras V
• Maisonnasse P
• et al.
• Scopus (57)
• PubMed
• Crossref
• Google Scholar
27. 27.Gut microbiota changes are detected in asymptomatic very young children with SARS-CoV-2 infection.Gut. 2022; 71: 2371-2373View in Article
• Nashed L
• Mani J
• Hazrati S
• et al.
• Scopus (19)
• PubMed
• Crossref
• Google Scholar
28. 28.In vitro study of the prebiotic xylooligosaccharide (XOS) on the growth of Bifidobacterium spp and Lactobacillus spp.Int J Food Sci Nutr. 2015; 66: 919-922View in Article
• Li Z
• Summanen PH
• Komoriya T
• Finegold SM
• Scopus (71)
• PubMed
• Crossref
• Google Scholar
29. 29.Dietary resistant dextrins positively modulate fecal and cecal microbiota composition in young rats.Acta Biochim Pol. 2015; 62: 677-681View in Article
• Śliżewska K
• Libudzisz Z
• Barczyńska R
• Kapuśniak J
• Zduńczyk Z
• Juśkiewicz J
• Scopus (9)
• PubMed
• Crossref
• Google Scholar
30. 30.Prebiotics in irritable bowel syndrome and other functional bowel disorders in adults: a systematic review and meta-analysis of randomized controlled trials.Am J Clin Nutr. 2019; 109: 1098-1111View in Article
• Wilson B
• Rossi M
• Dimidi E
• Whelan K
• Scopus (0)
• PubMed
• Summary
• Full Text
• Full Text PDF
• Google Scholar

Article info

Publication history

Published: December 07, 2023

Identification

DOI: https://doi.org/10.1016/S1473-3099(23)00685-0

Copyright

© 2023 Elsevier Ltd. All rights reserved.

ScienceDirect

Access this article on ScienceDirect

Figures

• Figure 1Trial profile
• Figure 2Proportion of PACS symptoms alleviation by 6 months

Tables

• Table 1Baseline characteristics
• Table 2Independent predictors of PACS alleviation identified by forward stepwise multivariate analysis
• Table 3Adverse events, study compliance, vaccination, and re-infection during study period