Dose-dependent effect of Lactobacillus rhamnosus on quantitative reduction of faecal rotavirus shedding in children

Abstract

Beneficial effects of probiotics in acute infectious diarrhoea in children are mainly seen in watery diarrhoea and viral gastroenteritis. Lactobacillus rhamnosus, one the most extensively studied probiotic strains, is effective in shortening courses of acute diarrhoea in children. However, the dose-dependent effect of Lactobacillus upon quantification of faecal rotavirus shedding in humans remains little known. Thus, an open-label randomized trial in 23 children with acute rotaviral gastroenteritis was undertaken by randomly allocating patients to receive one of the three regimens for 3 days: daily Lactobacillus rhamnosus 35 (Lcr35) with 0 CFU/day to six patients in the control group, 2 × 10⁸ CFU/day to nine patients in the low-dose group, and 6 × 10⁸ CFU/day to eight patients in the high-dose group. Faecal samples were collected before and after the 3-day regimen for measurements of rotavirus concentrations by ELISA. There was no statistically significant change in faecal rotavirus concentrations in either the control group (119.2 × 10⁵ particles/ml vs. 23.7 × 10⁵ particles/ml, p = 0.075) or the low-dose group (36.1 × 10⁵ particles/ml vs. 73.5 × 10⁵ particles/ml, p = 0.859). However, the high-dose group had a significant reduction of faecal rotavirus concentration (64.2 × 10⁵ particles/ml vs. 9.0 × 10⁵ particles/ml, p = 0.012). Without any exception, the faecal rotavirus concentrations of all eight patients in the high-dose Lcr35 group declined by 86% after 3 days when compared with those before Lcr35 administration. In conclusion, this is the first report to provide quantitative evidence of the dose-dependent effect of Lactobacillus rhamnosus, a minimal effective dose of 6 × 10⁸ CFU for 3 days, upon the faecal rotavirus shedding in paediatric patients.

Dose-dependent, Lactobacillus rhamnosus, rotavirus, virus shedding, ELISA

Introduction

Beneficial effects of probiotics in acute infectious diarrhoea in children so far appear to be moderate, strain-dependent, dose-dependent and significant mainly in watery diarrhoea and viral gastroenteritis. The effects would be more evident when treatment with probiotics is initiated early in the course of disease [1]. Lactobacillus rhamnosus GG (LGG), a similar probiotic strain to Lactobacillus rhamnosus 35 (Lcr35) to colonize the human gut, is effective in promoting a more rapid recovery of acute, watery diarrhoea due to rotavirus in children [2]. A number of dose-dependent effects of probiotic bacteria upon different models have been documented. The probiotic bacterium LGG has an inhibitory effect on bacterial translocation in a neonatal rabbit model and inhibits bacterial translocation of E. coli C25 in a dose-dependent manner in an in vitro cell-culture model [3].

Previous in vitro and in vivo animal experiments have sketched out a blueprint of the dose-dependent effect of Lactobacillus. In an in vivo study for evaluating the ability of LGG to survive gastrointestinal transit in dogs and assessing whether oral administration of LGG is safe, faecal colonization of LGG significantly increased in receiving 5 × 10¹¹ CFU/day of LGG for 5 days [4]. A dose–response effect in calves was demonstrated by oral administration of LGG with their morning milk feeding on 3 consecutive days at a low (2 × 10¹⁰ CFU), medium (1 × 10¹¹ CFU) or high (2 × 10¹¹ CFU) dosage. Twenty-four hours after the first feeding, LGG was recovered from 20%, 80% and 100% respectively in the three groups [5]. By treating mice perorally with a high dose of 10¹² viable bacteria per kilogram of body weight for 7 days, LGG has specific dose- and duration-dependent immunomodulatory effects on the proliferative activity of B and T lymphocytes [6].

Probiotics have proven effective in the treatment of infectious diarrhoea and had shown a beneficial effect in reducing the duration of diarrhoea. The greatest effect was evident in rotaviral gastroenteritis, and LGG showed the most consistent effect [1]. A dose-dependent inverse relationship between a daily Lactobacillus dose, with the dose of >10¹¹ CFU/48 h being the most effective, and the reduction of diarrhoea duration has been reported [1]. However, the dose-dependent effect of Lactobacillus in other applications, including influences upon the faecal rotavirus shedding in human subjects, has rarely been discussed. Thus, we designed a small open-label randomized trial to assess whether there is a potential difference between high doses and low doses of Lactobacillus rhamnosus in quantitative reduction of faecal rotavirus shedding in children.

Materials and methods

This study was conducted at Taiwan Adventist Hospital during a 1-year period to prospectively enrol 23 paediatric patients with acute gastroenteritis aged 9–72 months with moderate to severe dehydration receiving Lcr35 (Antibiophilus®, lyophilized culture of Lactobacillus rhamnosus 35, Laboratoires Lyocentre, France). Confirmed positive stool rotavirus antigens by Enzyme-Linked ImmunoSorbent Assay (ELISA) were determined in children, then patients were immediately allocated by unrestricted randomization to receive one of the three regimens for 3 days. Daily Lcr35 were administered with 0 CFU/day in the control group, 2 × 10⁸ CFU/day in the low-dose group and 6 × 10⁸ CFU/day in the high-dose group. All patients simultaneously received placebo with antiflatulent Simethicone 80 mg/day. Faecal samples were collected before commencement of the assigned regimen and within 24 h after ending of the 3-day regimen for every recruit. Absorbance (OD_{450 nm} value) of stool samples to quantify faecal rotavirus concentrations was measured in triplicate using a commercial ELISA kit RIDASCREEN® (R-Biopharm AG, Darmstadt, Germany).

The exclusion criteria included use of antibiotics, probiotic agents, long-term use of Chinese herbal medicine or steroids, immunodeficiency or cancer, bloody diarrhoea or positive stool occult blood and pus cells, neutropaenia, parasite infections, bacteraemia, lactose-containing foods, antidiarrhoeal medications 2 weeks before and during hospitalization, failure in completing the informed consents and loss to follow-up. Eventually, 23 patients (six in the control group, nine in the low-dose group and eight in the high-dose group) completed the entire protocol and their data were analysed according to the different grouping.

The Wilcoxon Sign Ranks test and Kruskal–Wallis test were respectively used to compare the faecal rotavirus concentrations before and after administration of the 3-day regimen in each group, and between each two of the three groups. The data were analysed using SPSS 14.0 statistical software. Faecal rotavirus concentrations in particles per millilitre were expressed as medians and individual values are shown as scatter plots in Figs 1–3. A p-value of <0.05 was considered statistically significant.

Fig. 1.

Faecal rotavirus concentrations before and after the placebo for 3 days showing an insignificant declining trend in the control group (filled square and filled triangle represent faecal rotavirus concentrations in each patient; horizontal lines indicate medians).

Results

There were 13 males and 10 females in the 23 children enrolled in the study. The control group had a female predominance in comparison with the low-dose and high-dose groups, but no sex difference between the latter two groups. The median age was 33.2 months and the median body weight was 12.5 kg. Neither timing of cessation of diarrhoea nor duration of fever differed among the three groups. No statistical difference existed between groups in blood cell counts and biochemical parameters. There was no significant difference in the faecal rotavirus concentrations between each two of the three groups either before (p = 0.6) or after (p = 0.143) administration of the assigned regimen for 3 days. There was also no statistically significant change in faecal rotavirus concentrations in either the control group (119.2 × 10⁵ particles/ml vs. 23.7 × 10⁵ particles/ml, p = 0.075, Fig. 1) or the low-dose group (36.1 × 10⁵ particles/ml vs. 73.5 × 10⁵ particles/ml, p = 0.859, Fig. 2). Although five of the six patients in the control group diminished in faecal rotavirus concentrations after the 3-day regimen treatment, nine patients in the low-dose Lcr35 group developed various trends including four to rise and five to fall. However, the high-dose group had a significant reduction of faecal rotavirus concentration (64.2 × 10⁵ particles/ml vs. 9.0 × 10⁵ particles/ml, p = 0.012, Fig. 3). With a loading dose of Lcr35 6 × 10⁸ CFU for 3 days, the faecal rotavirus concentrations in all of the eight patients showed an obvious decline by 86%.

Fig. 2.

Faecal rotavirus concentrations before and after low-dose Lcr35 for 3 days showing variously increasing or decreasing tendencies in the low dose Lcr35 group (filled square and filled triangle represent faecal rotavirus concentrations in each patient; horizontal lines indicate medians).

Fig. 3.

Faecal rotavirus concentrations before and after high-dose Lcr35 for 3 days showing an universally falling tendency in the high dose Lcr35 group (filled square and filled triangle represent faecal rotavirus concentrations in each patient; horizontal lines indicate medians; *p < 0.05).

Discussion

Very few human studies have evaluated the impact of different daily doses of Lactobacillus on faecal recovery over the last decade. In a dose–response study in healthy infants, a 2-week oral administration of powdered form LGG at varying levels (10⁸, 10⁹ or 10¹⁰ CFU/day) was effective in transiently colonizing the intestine for 7–14 days after discontinuing LGG and all levels resulted in a similar incidence of intestinal colonization, in which a minimum level of 10⁸ CFU/day of LGG had positive implications for the therapeutic use of LGG [7]. In another study in healthy human adults, recovery of LGG in faeces was evaluated after oral administration of LGG to 20 volunteers for 7 days in gelatine capsules with daily doses of 1.6 × 10⁸ CFU and 1.2 × 10¹⁰ CFU. All subjects in the higher dose group had detectable numbers of LGG in their faeces during the test period. The strain was detected in faeces of all the volunteers after 3 days of administration [8]. The above two human studies facilitated our study to choose two different doses. The doses of Lactobacillus rhamnosus in our study were 2 × 10⁸ CFU/day and 6 × 10⁸ CFU/day, well between the threshold doses of 10⁸ CFU/day and 1.2 × 10¹⁰ CFU/day suggested as most effective [8, 9]. After considering both drug compliance in children and minimally detectable amounts of Lcr35 in faeces, we adopted a shorter duration of 3 days in children with rotavirus infections. Additionally, Lcr35 was started in the early stage during the course of rotaviral gastroenteritis to achieve its better efficacy [1].

Our 3-day regimens showed that Lcr35 administered in a high dose of 6 × 10⁸ CFU/day could have a potential effect on clearance of stool rotavirus because all of the patients receiving high-dose Lcr35 significantly reduced faecal rotavirus shedding, but such a reduction did not occur in those patients receiving low-dose Lcr35 or only placebo. In a randomized clinical trial in acute watery diarrhoea of Indian children with ∼75% acquiring rotavirus infection, using oral rehydration solution with a lower concentration (1.2 × 10⁸ CFU/day) of LGG for at least 7 days neither decreased frequency and duration of diarrhoea and vomiting nor hospital stay, which could be due to malnutrition and comparatively lower dosage [10]. Malnutrition, which may alter normal gut flora by repeated infections, is related to impairment of cell-mediated immunity, phagocyte function, immunoglobulin A secretion, and cytokine responses. Similarly, using a milk formula with a higher concentration (10¹¹ CFU/day) of LGG without restriction of lactose-containing foods showed no positive effect on the clinical course of acute watery diarrhoea [9]. A possible beneficial effect could have been masked by worsening diarrhoea due to transient carbohydrate malabsorption and by various effects of Lactobacillus on viral growth. In contrast, using oral rehydration solution containing a relatively high concentration (10¹⁰ CFU/day) of LGG in a multi-centre European trial resulted in shorter duration of diarrhoea, less chance of a protracted course, and faster discharge from the hospital [11]. Such controversial results from two similar high-dose LGG trials designate existence of other influential factors such as food intolerance and host status. However, no data related to quantification of viral shedding was addressed. Using milk or formula containing relatively high concentration (6 × 10⁹ CFU/day) of LGG for a maximum of 5 days reduced not only duration of diarrhoea in rotavirus-positive and -negative children, but also the duration of stool rotavirus excretion in six days after the onset of diarrhoea [12]. Despite no kinetic quantitative analysis of faecal rotavirus concentrations, such reduction in stool rotavirus excretion is comparable with the declining tendency in the high-dose Lcr35 group of this study.

In addition to durations of diarrhoea and other clinical manifestations, our study mainly focused on the quantification of faecal rotavirus shedding before and after administration of Lactobacillus rhamnosus in order to observe its direct impact within a short period. To lessen any possible interference of lactose malabsorption in our study, we used only single Lcr35 rather than a synbiotic and food restriction of lactose. Interestingly, the faecal rotavirus concentrations of the most children in the control group seemed to decrease, although not significantly after using placebo for 3 days. The decreased faecal rotavirus shedding could be attributed to decay during the course of rotavirus diarrhoea. Such spontaneous shedding could be further enhanced by high-dose Lcr35 that might derive from the colonization characteristics of Lactobacillus [13]. Individual patients have different pathogenic viral loads and immunological competencies which could affect viral shedding. The nutritional status of children is also important to influence recovery of viral or bacterial diarrhoea. Therefore, the timing of their enrolment for starting the Lactobacillus regimen, the impact of osmolarity of faecal contents in patients with severe watery diarrhoea, and the correlation of the faecal rotavirus shedding with the intraluminal viral concentrations in the intestines should be taken into consideration in the future studies.

To our knowledge, this is the first randomized trial providing quantitative evidence of the dose-dependent effect of Lactobacillus rhamnosus upon the faecal rotavirus shedding in paediatric patients, and our preliminary results demonstrated a minimal effective dose of 6 × 10⁸ CFU Lcr35 for 3 days in children with rotaviral gastroenteritis. Although we could not draw a definite conclusion due to limited case numbers, the minimal effective dose and short duration of Lcr35 in reducing the faecal rotavirus shedding have been obtained. From a standpoint of public health, the implication of our clinical results can lower the risks of faecal-oral transmission when cost-benefit effectiveness is concerned. Prospective clinical trials can be designed to investigate whether durations of Lactobacillus administration could affect the clinical efficacy, which may facilitate development of optimal therapeutic strategies. If successful, application of Lactobacillus with appropriate dosage in paediatric rotavirus infection or various enteric infections is foreseeable in the future.

Funding

Wonderful Biotech Co, Ltd; the Medical Research Committee of Taiwan Adventist Hospital (TAH 921001 to S.B.F.); National Taipei University of Technology (NTUT 97-140-3 to H.W.F.).

Acknowledgements

We are grateful to Professor Mao-Yuan Chen for his support in the laboratory, to Dr Pei-Ying Ling for her revision of the English in this paper and to Miss Shu-Fen Chou for her assistance in data collection.