Safety of Lactobacillus reuteri DSM 17938 in Healthy Children 2 to 5 Years of Age

Abstract

Probiotics are increasingly used for diarrhea, but studies under FDA IND are few. We conducted a phase-one placebo-controlled study of Lactobacillus reuteri DSM 17938 under IND in 60 children 2–5 years of age (41 L. reuteri, 19 placebo) in a resource constrained community in Peru. No differences in objective data on adverse events were noted, although some differences based on subjective parental reports for fever and diarrhea were seen.

Introduction.

Probiotics are live bacteria which are increasingly used for treating conditions such as diarrhea, infantile colic, and inflammatory bowel disease.¹ Research on probiotics has increased, resulting in a regulatory framework to assure proper oversight and systematic safety data on products classified as dietary supplements. In the United States, the Food and Drug Administration (FDA) regulates clinical trials for products evaluated for clinical indications, including probiotics, under the Investigational New Drug (IND) program, beginning with Phase One safety studies. Worldwide and historically the majority of probiotic clinical trials have not been done under IND, so when results are conflicting and it is not clear whether product characteristics such as viability, purity, and strain content were as stated on the label. Clinical trials have shown highly divergent findings regarding the induction of carriage,² as well as clinical outcomes, ³ indicating the need for more stringency with independently scrutinized clinical documentation, protocol design, study oversight, and reporting.

Lactobacillus reuteri ATCC 55730 has shown therapeutic benefit for infantile colic, ⁴ and the closely related strain DSM 17938 used here was cured of an antibiotic resistance plasmid.⁵ L. reuteri DSM 17938 has recently been shown to be safe for infants with colic in studies under IND.⁶ It was also effective for treatment of diarrhea in children and adults,^{7, 8} but studies for this indication under IND are lacking. We previously conducted a safety study in adults ⁹ and progressed to do the same in children 2–5 years of age in order to support licensing of the product for clinical indications in pediatric populations.

Methods.

This was a Phase I double-blinded randomized controlled trial of safety and tolerability of Lactobacillus reuteri DSM 17938 given daily for five days (the recommended duration of therapy for infectious diarrhea). Sixty subjects were randomized by blocks in a ratio of treatment to placebo of 2:1 to the following groups:

1. Lactobacillus reuteri (LR): 10⁸ (100 million) organisms of Lactobacillus reuteri DSM 17938 per dose, given once daily for five days (dose based on previous reports¹⁰ and manufacturer guidelines [BioGaia AB, Stockholm, Sweden]). Whole genome sequencing was performed to confirm strain identity and viable counts confirmed on arrival and monthly that confirmed minimal viable active product counts.

2. Placebo (PL): 5 drops of the oil vehicle used for the LR suspension, given once daily for five days.

Eligible participants were children 2–5 years of age from Santa Clara, a peri-urban community in Peru with high rates of diarrheal diseases. Exclusion criteria included: 1) exclusionary household members (another study participant, pregnant or breastfeeding women, infant younger than 6 months, immunosuppressed individuals); 2) allergy to penicillin, cephalosporins, or gentamicin; 3) antibiotic use within 30 days; 4) use of probiotics within 90 days; 5) diarrhea within 30 days; 6) fever or a pre-existing adverse event; 7) positive HIV antibody test; 8) severe anemia; 9) out of range laboratory values on Day 0; and 10) positive PCR for L. reuteri in the pre-enrollment stool. Systematic quality control assays of study products were performed monthly and all results fell within appropriate ranges.

Participants had active surveillance for 28 days, and parents used a diary card to record symptoms based on the clinical scoring system of the NIH Division of AIDS Table for Grading the Severity of Adult and Pediatric Adverse Events.¹¹ Field workers administering study product and conducting home visits were well versed in symptom classification and assisted parents in selecting the appropriate symptom grade. During study visits (daily during 5-day product administration and at 3-day intervals following that until day 28) temperature was monitored and presence of diarrhea (defined as 3 or more liquid stools daily) was noted, along with the daily number of unformed stools. After study day 28, subjects continued with passive surveillance up to 6 months post-enrollment. Data was analyzed in Access using Stata, using chi square, Fisher’s exact, and Wilcoxon tests to compare frequencies between groups. Written informed consent from both parents was obtained and the study was approved by the IRBs of Tulane Medical Center and Asociacion Benefica Prisma. The study was conducted under FDA IND and with oversight by the Peruvian National Health Institute (INS).

Results.

Eighty-seven subjects were screened and 60 were enrolled and randomized. Of the 27 not enrolled, 18 had an exclusionary pre-enrollment laboratory value, 8 had a positive PCR assay for L. reuteri in stool, and one was excluded for medical reasons. Forty-one children received L. reuteri (mean age 4.04 years [SD= 1.27]; 21 females) and 19 children (mean age 3.31 years [SD= 1.1]; 13 females) received placebo. No differences by sex were observed between groups, but there was a significant difference of 0.73 years in the mean age (p-value = .035), so subsequent comparisons were adjusted for this age difference.

When body temperature was analyzed based on values from all observations (n = 823) and compared based on frequency of temperatures above Grade 0 (i.e. values ≥ 38.0 °C), no differences between study groups were observed (LR 2/555 observations ≥ 38.0 °C; PL 2/268 observations ≥ 38.0°C; p= .604). No subject met the criteria for obtaining a blood culture (T >38.5°C). No differences in diarrhea prevalence at home visits were noted (LR diarrhea present 10/555 observations =1.80%; PL diarrhea present 7/268 observations =2.61%; age adjusted p=0.87). At least one diarrhea episode was observed in 6/19 subjects in PL group and 8/41 subjects in LR group, and 1 child in the PL group and 2 children in the LR group experienced 2 episodes of diarrhea.

Laboratory assessment on Day 0 (baseline), Day 5 (end of treatment), and Day 28 included total leucocyte count, serum hemoglobin; hepatic enzymes; total bilirubin, blood urea nitrogen, and creatinine. No significant differences between groups were noted.

Symptoms histories.

The prevalence of symptoms monitored by parents during 28 days of active surveillance is reported in Table 1. Symptoms were graded based on defined criteria,¹² and subjects with grade 2 symptoms or new symptoms of concern were evaluated by a physician. In all 532 daily observations were recorded in 19 subjects in the PL group and 1148 observations in 41 subjects in the LR group. Subjects in the LR group reported a higher period prevalence of subjective fever (5.4% of days observed LR vs 3.01% of days observed PL) and more days with reported wheezing (1.48% of days observed LR vs 0% of days observed PL). Subjects in the PL group reported more days with diarrhea (4.89% PL vs 1.48% LR), more days with rash (2.44% PL vs 0.09% LR), and more days with pruritis (2.26% PL vs 0.87% LR). Number and duration of symptom episodes were similar between groups in all categories, including subjective fever (2.41 days PL, 2.56 days LR) and diarrhea (2.5 days PL, 1.3 days LR). When analyzed based on number of subjects reporting each symptom by study group, no significant differences were found.

Table 1.

Clinical Symptoms Based on Parental Home Record Reporting

	Placebo Group ( n = 532 symptom reports from 19 subjects, recorded during 28 days of active surveillance)			LR Group ( n = 1148 symptom reports from 41 subjects, recorded during 28 days of active surveillance)			Chi-Square Test
	Frequency* (Percentage)	Grade 1 (# Patients)	Grade 2 (# Patients)	Frequency* (Percentage)	Grade 1 (# Patients)	Grade 2 (# Patients)	p-value (Age-adjusted p-value)
FEVER	16 (3.01%)	16 (n=6)	0 (0)	62 (5.40%)	59 (n=18)	3 (n=3+)	0.0301 (0.0089)
VOMITING	5 (0.94%)	4 (n=3)	1 (n=1)	11 (0.96%)	10 (n=9)	1 (n=1)	0.9713 (0.8159)
DIARRHEA	26 (4.89%)	25 (n=5)	1 (n=1)	17 (1.48%)	15 (n=10)	2 (n=2)	< 0.0001 (0.0015)
IRRITABILITY	12 (2.26%)	12 (n=3)	0 (0)	32 (2.79%)	30 (n=7)	2 (n=1+)	0.5255 (0.1513)
RASH	13 (2.44%)	13 (n=3)	0 (0)	1 (0.09%)	1 (n=1)	0 (0)	< 0.0001f (0.053)
PURITIS/ITCHING	12 (2.26%)	12 (n=3)	0 (0)	10 (0.87%)	9 (n=4)	1 (n=1+)	0.0202 (0.1692)
WHEEZING	0 (0.00%)	0 (0)	0 (0)	17 (1.48%)	9 (n=2)	8 (n=4++)	0.0048 -----
OTHERS	99 (18.61%)			172 (14.98%)			0.0601 (0.9409)

^*Frequency = the number of individual reports of the symptom during the 28 day active surveillance period. Each day that a symptom is reported is reported separately for subjects reporting the same symptom on consecutive days.

^f:Fisher’s Exact Test

⁺Each also had Grade 1

⁺⁺one also had Grade 1 and one also had 2 Grade 1

Adverse event reports

are based on physician evaluations of subjects referred to the study clinic. The study physician reviewed 31 adverse events for the PL group (1.63 AEs per subject) and 45 adverse events in the LR group (1.10 AEs per subject). The most common diagnoses were respiratory infections (32 of 45 AEs in the LR group [71.1%], 15 of 31 in PL [48.4%]), and urinary tract infections (3 of 45 AEs in the LR group [6.7%], 2 of 31 in PL [6.5%]). No statistically significant differences were observed, and no adverse events were determined to be related to study products.

Discussion:

This study conducted under IND did not suggest safety concerns with the administration of LR in Peruvian children 2–5 years of age. Based on the large numbers of home diary observations, differences in home-reported symptom rates appear to be statistically significant but not biologically meaningful. No severe adverse events were observed, and adverse event rates did not differ between groups. Episodes of fever and respiratory symptoms were associated with routine conditions such as respiratory and urinary tract infections. While not designed to evaluate efficacy, one interesting observation was the difference in subjective reports of diarrhea. Subjects in the PL group had more than twice the period prevalence of diarrhea days recorded compared with the LR group (4.89% of days observed PL vs 1.48% LR; p=0.0015), and while diarrhea episode duration did not differ statistically, the mean duration for PL was about double the mean for LR, a 1 day difference in duration which compares favorably to the effect size of treatment with antibiotics for travelers with diarrhea. Phase Two studies are needed to evaluate the efficacy of LR on diarrhea, linear growth, intestinal and systemic markers of inflammation, and intestinal microbiota.