An 11-Year Review of Lactobacillus Bacteremia at a Pediatric Tertiary Care Center

This was a retrospective, descriptive study conducted in a 323-bed tertiary-care pediatric hospital. The hospital includes a 66-bed NICU, 47-bed pediatric ICU, 28-bed bone marrow transplant and hematology and oncology unit and 12-bed intestinal rehabilitation unit. Cases of Lactobacillus bacteremia were identified through a search of all positive inpatient blood culture reports from January 1, 2010 through December 31, 2020. All patients admitted to the hospital who were diagnosed with Lactobacillus bacteremia, regardless of age, were included. The clinical chart for each case was independently reviewed for the following risk factors: documented probiotic use, presence of a central venous catheter, immunocompromising condition (defined as primary immune deficiency or use of immunosuppressive agents), neutropenia (defined as an absolute neutrophil count less than 1.0 × 10⁹/L), impaired intestinal function because of an underlying disorder (e.g., inflammatory bowel disease or short-bowel syndrome), and age below 3 months. Additionally, the clinical presentation and management were reviewed for each case, including the presence of fever, need for vasoactive medications, gastrointestinal symptoms, and treatment course. Results were summarized descriptively. Total inpatient probiotic administration was assessed over the same time period and was reported based on the number of patients who received at least 1 dose of probiotics during hospital admission. Distribution of probiotic use by clinical service was collected for a random 3-month period between May and July 2020. Patient data were collected using Microsoft Excel (Microsoft Corporation). The study was acknowledged as a quality improvement project and did not constitute human subjects research, and thus did not require oversight by the hospital’s institutional review board.

Eight cases of Lactobacillus bacteremia were identified in inpatients over the study period (Table 1) during which 127 845 inpatient admissions occurred. Two cases received probiotics in the hospital, initiated 5 and 10 days before bacteremia, and a third case had documented outpatient administration of probiotics before bacteremia. Of the 3 cases with prior documented probiotic use, 2 were administered via a nasogastric or gastrostomy tube. The remaining 5 cases had no documented probiotic use. During the study period, the probiotic on formulary contained Lactobacillus rhamnosus GG. The patient who received outpatient probiotics also reported using a Lactobacillus rhamnosus GG product. For one case in our review – a 13 year old female with acute myeloid leukemia who received probiotics orally during inpatient admission – the mass spectrometric proteomic tracing of the strain of L. rhamnosus that grew in blood culture was compared with that of the strain in the probiotic administered. The tracings were essentially identical, suggesting that the 2 strains may have been identical organisms.

Regarding underlying risk factors for Lactobacillus bacteremia, 5 cases had underlying impaired intestinal function, including 4 cases in individuals with short-bowel syndrome and 1 case in an individual with Crohn’s disease. One case was in an oncology patient and there were no cases in solid organ or bone marrow transplant patients. Two cases, including the oncology patient and 1 patient with impaired intestinal function, were neutropenic at the time of bacteremia. All but 1 case had a central venous catheter in place.

All 8 cases presented clinically with fever at the time of bacteremia. No cases required vasoactive medications. Five cases had gastrointestinal symptoms, which included at least 1 of the following: abdominal pain, diarrhea, bloody stools or vomiting. In 3 cases, all in individuals with impaired intestinal function, multiple organisms grew on blood culture. All cases received antibiotics and had resolution of bacteremia and associated symptoms. However, in many cases the antibiotics prescribed were determined by a confirmed or suspected concurrent infection. Choice of antibiotics, as well as duration of treatment, differed significantly from case to case.

Probiotics were administered to 4478 inpatients during the study period, which represents approximately 3.5% of all admitted patients. Over a random three-month period, 41.5% of all probiotics were prescribed by the hospitalist service. The oncology service prescribed 7.7% of probiotics. The gastroenterology service and intestinal rehabilitation and liver service prescribed 4.6% and 1.5%, respectively. No probiotics were prescribed by the pediatric or neonatal intensive care units or the bone marrow transplant service. The greatest number of patients received probiotics in the years 2016 and 2017, and the cases of Lactobacillus bacteremia peaked in 2019 and 2020 (Fig 1).

Overall, in this review of 127 845 pediatric hospital admissions over an 11-year period, we found Lactobacillus bacteremia to be uncommon with the peak number of cases and peak utilization of probiotics occurring at different times. This finding is consistent with a previous study in Finland that did not show an increase in Lactobacillus bacteremia after a rapid increase in probiotic use nationally.¹²  Additionally, in 5 out of the 8 cases we identified, there was no documented prior use of probiotics. This suggests that the rates of hospital-wide probiotic use may not be of as significant concern as individual risk factors. Nevertheless, for individual patients with risk factors for Lactobacillus bacteremia, probiotics may increase their risk for bacteremia as was the case for at least 1 patient in our review – an oncology patient for whom the strain of Lactobacillus causing bacteremia matched the probiotic strain on mass spectroscopy. Using whole-genome sequencing, Yelin et al demonstrated that strains causing Lactobacillus bacteremia could be traced to probiotic strains in the pediatric ICU.¹³ 

In our review, we found that the majority of cases of Lactobacillus bacteremia occurred in individuals with underlying impaired intestinal function, which is consistent with previous reports in this population.^9,10  Importantly, 4 of 5 cases in this population, and 7 out of 8 cases overall had central venous catheters in place, which has also been identified as a risk factor for Lactobacillus bacteremia following probiotic use.¹⁴  We did not find any cases in neonates or a significant number of cases in the immunocompromised individuals, which have previously been described as having the highest risk for Lactobacillus bacteremia.⁴  Although the former may be because of little to no prescribing of probiotics in the neonatal intensive care, the latter is true despite administration of probiotics to oncology patients.

Although Lactobacillus is often considered to have low pathogenicity, all of the cases in our review presented with clinical symptoms, including fever, and most also had other complaints, most commonly gastrointestinal in nature. Nonetheless, no cases required vasoactive medications and the bacteremia resolved in all cases.

This review has several limitations. We did not account for probiotic use by hospital unit or by patient diagnosis. Prescribing differences by unit or patient diagnosis may be a confounding factor in development of Lactobacillus bacteremia in certain populations. However, we note that probiotics were not restricted in any unit or patient population during the study period. History of outpatient probiotic use was limited by information documented in the chart. As outpatient medications are typically collected by caretaker history, this data may be incomplete. Additionally, the study was performed at a single site, which may prevent the findings from being generalizable to other settings. Multiple sites would have provided greater variability in the patient population and prescribing practices. We were also unable to extend these results to non-Lactobacillus containing probiotic formulations.

Although Lactobacillus bacteremia is uncommon, certain pediatric populations, such as individuals with impaired intestinal function, underlying immune-compromise, and those with central venous catheters, may be at higher risk. Decisions about probiotic use in these and other high-risk populations should be informed by a risk-benefit decision with careful attention to the available efficacy data for the proposed use in the context of known, albeit small, risk.