Lactobacillus Sepsis Associated With Probiotic Therapy
http://www.100md.com
《小儿科》
Departments of Pediatrics Anesthesiology, Wake Forest University School of Medicine, Winston-Salem, North Carolina
ABSTRACT
Probiotic strains of lactobacilli are increasingly being used in clinical practice because of their many health benefits. Infections associated with probiotic strains of lactobacilli are extremely rare. We describe 2 patients who received probiotic lactobacilli and subsequently developed bacteremia and sepsis attributable to Lactobacillus species. Molecular DNA fingerprinting analysis showed that the Lactobacillus strain isolated from blood samples was indistinguishable from the probiotic strain ingested by the patients. This report indicates, for the first time, that invasive disease can be associated with probiotic lactobacilli. This report should not discourage the appropriate use of Lactobacillus or other probiotic agents but should serve as a reminder that these agents can cause invasive disease in certain populations.
Key Words: Lactobacillus probiotic bacteremia sepsis
Abbreviations: CVC, central venous catheter MIC, minimal inhibitory concentration
In recent years, several articles have reviewed the efficacy, mechanism of action, and safety of probiotics in the treatment of infectious disease.1–5 The term probiotic refers to "live microorganisms which when administered in adequate amounts confer a health benefit on the host."6 The microorganisms most frequently used as probiotic agents are lactic acid bacteria (species of Lactobacillus, Enterococcus, and Bifidobacterium) and nonpathogenic, antibiotic-resistant, ascosporic yeasts, especially Saccharomyces boulardii.2 Lactobacillus rhamnosus strain GG (ATCC53103), which was originally isolated from human intestinal flora, is the most widely studied probiotic agent for adults and children.7 L rhamnosus strain GG can prevent diarrhea and atopy among children.8,9 A recent meta-analysis suggested that Lactobacillus is a safe effective treatment for children with acute infectious diarrhea.10
Serious infections attributable to probiotic lactobacilli are extremely rare. Only 2 well-documented adult cases have been reported in which the L rhamnosus strain isolated from the clinical specimen was indistinguishable from the probiotic strain consumed by the patient.11,12 We report the first pediatric case of invasive disease attributable to a Lactobacillus strain indistinguishable from the L rhamnosus strain GG administered to the infant.
CASE REPORTS
Case 1
A 6-week-old, white, male patient was admitted for scheduled repair of a double-outlet right ventricle and pulmonic stenosis. The birth weight was 3.2 kg, and the patient was born at term. The postoperative course was complicated by pacemaker placement, pulmonary artery banding, seizures, acute renal insufficiency, and prolonged respiratory support. The infant received several 7- to 10-day courses of broad-spectrum antibiotics, including vancomycin and ceftriaxone, for treatment of suspected sepsis; however, blood and urine cultures remained negative. On day 57 of hospitalization, the patient developed copious nonbloody diarrhea. He had completed his last antibiotic course 3 days before the onset of diarrhea. The patient was afebrile, and the diarrhea persisted despite changes in formula. It was presumed that the diarrhea was a consequence of the courses of broad-spectrum antibiotics. Results of routine stool cultures, an enzyme immunoassay for Rotavirus, and an assay for Clostridium difficile toxin were negative. Beginning on day 79 of hospitalization, Lactobacillus GG (Culturelle, 10 x 109 cells per capsule; ConAgra Foods, Omaha, NE) was administered through the gastrostomy tube (1 capsule mixed with formula, once daily) for its probiotic effect. The patient's diarrhea improved with probiotic therapy but, on day 99 of hospitalization, while continuing to receive probiotic therapy, the patient developed new-onset fever and marked leukocytosis and appeared ill.
A physical examination revealed a toxic pale infant with a temperature of 39.7°C, pulse of 171 beats per minute, respiratory rate of 80 breaths per minute, and blood pressure of 60/30 mm Hg. Ventilatory support was maintained with a bilevel airway pressure of 18 cm H2O and a continuous positive airway pressure of 8 cm H2O. Respiratory distress was evidenced by sternal and intercostal retractions. Bilateral coarse breath sounds were noted. A grade 3/6 pansystolic murmur was audible over the precordium. The extremities were cool and mottled, with poor peripheral perfusion. The Broviac central venous catheter (CVC) exit site and subcutaneous tract were without redness, warmth, discharge, tenderness, or induration. The patient showed no peripheral signs of endocarditis. The rest of the examination results were unremarkable.
The results of laboratory tests were as follows: white blood cell count of 38 900 cells per mm3, with 87% segmented neutrophils and 10% bands. Urinalysis results were normal, and electrolyte levels were consistent with metabolic acidosis. A blood culture obtained from the CVC yielded Gram-positive rods. Chest radiographic findings were normal. Transesophageal echocardiography revealed a 10- x 12-mm echogenic mass at the tip of the CVC (superior vena cava-right atrium junction), firmly adherent to the posterior wall of the right atrium. No vegetations were noted on the heart valves. Intravenous vancomycin (40 mg/kg per day [10 mg/kg every 6 hours]) and gentamicin (5 mg/kg per day [2.5 mg/kg every 12 hours]) therapy was initiated.
In the subsequent 72 hours, 3 additional blood cultures (2 central and 1 peripheral) obtained at different times yielded >100 colony-forming units of Gram-positive rods per mL. All 4 isolates were identified as Lactobacillus species. The isolate was susceptible to penicillin G, erythromycin, clindamycin, gatifloxacin, tetracycline, and trimethoprim-sulfamethoxazole, with minimal inhibitory concentrations (MICs) of 1, <0.25, <0.25, <0.25, <1, and <10 mg/L, respectively, and was resistant to vancomycin, meropenem, ceftriaxone, and cefuroxime, with MICs of >4, >0.5, >2, and >2 mg/L, respectively. When antimicrobial susceptibilities became available, antibiotic therapy was changed to high-dose penicillin G (400 000 U/kg per day, every 4 hours); gentamicin was continued for its synergistic effect. When blood cultures were positive for Lactobacillus species on 3 consecutive days, the CVC was promptly removed and oral Lactobacillus GG therapy was discontinued.
In the subsequent 48 to 72 hours, the patient exhibited clinical improvement, with resolution of respiratory distress and improvement in perfusion. Low-grade fevers persisted but were <38°C. The white blood cell count normalized at 7700 cells per mm3 (56% segmented neutrophils, 0% bands). Repeat blood cultures obtained 48 hours after the initiation of high-dose penicillin G therapy yielded negative results. In the subsequent 5 weeks, serial echocardiograms revealed resolution of the thrombus at the superior vena cava-right atrium junction. The patient completed a 6-week course of penicillin G and gentamicin treatment for his presumed endocarditis-equivalent infection.
Case 2
A 6-year-old, white, female patient with cerebral palsy, microcephaly, mental retardation, and a seizure disorder, who required feeding through a gastrojejunostomy tube, was admitted for treatment of a urinary tract infection. One week before hospitalization, she had undergone revision of a spinal rod to improve her scoliosis. She also had developed regurgitation of feedings and abdominal pain associated with feedings in the week before admission.
The patient's urine culture yielded Escherichia coli, for which she received a 10-day course of ceftriaxone. The fever resolved but the abdominal pain with feedings persisted, with development of diarrhea. Therefore, enteral feedings were stopped and parenteral nutrition was initiated through a previously implanted CVC. On day 5 of hospitalization, the patient began treatment with vancomycin, after developing a fluid collection over the posterior portion of her left spinal rod. Vancomycin treatment was discontinued when culture of the fluid yielded oxacillin-susceptible Staphylococcus aureus, and the patient was treated with intravenously administered oxacillin for 2 weeks.
On hospital day 18, the patient developed catheter-related enterococcal sepsis, which was treated with vancomycin for 2 days followed by ampicillin plus gentamicin for 14 days. Follow-up blood cultures on the fourth and sixth days of this treatment course yielded negative results.
Throughout the first 2 months, the patient continued to experience intermittent diarrhea, which was sometimes explosive and was not always contained within her diapers. Multiple stool assays for C difficile toxin and assays for Rotavirus yielded negative results. Because antibiotic-associated diarrhea was a possibility, Lactobacillus GG (Culturelle, 10 x 109 cells per capsule; ConAgra Foods) was administered beginning on day 25 of hospitalization, for its probiotic effect; each day, the contents of 1 capsule were mixed with water and the mixture was injected into the gastrojejunostomy tube with a syringe. The syringe was prepared in a medication room remote from the patient's room. The patient remained hospitalized for ongoing evaluation and management of her gastrointestinal feeding intolerance.
The patient became febrile to 40.3°C on day 69 of hospitalization. Her pulse rate, blood pressure, and respiratory rate remained stable within the ranges of the previous few weeks. She was more irritable and less active than usual but otherwise demonstrated no new findings in her physical examination. The results of laboratory tests were as follows: white blood cell count of 9800 cells per mm3, with 87% segmented neutrophils, 7% bands, and 1% metamyelocytes. Chest radiographic findings, urinalysis results, and serum electrolyte levels were normal. Intravenous vancomycin (40 mg/kg per day, every 8 hours) and ceftazidime (150 mg/kg per day, every 8 hours) therapy was initiated.
A peripheral blood culture obtained on day 69 of hospitalization yielded >100 colony-forming units of Lactobacillus species per mL, with MICs identical to those in case 1 except that the penicillin MIC was 0.5 mg/L (vancomycin was not tested). After these culture results became known on the third day of illness (hospital day 71), Lactobacillus GG treatment was discontinued and intravenous antibiotic therapy was changed to ampicillin (100 mg/kg per day, every 6 hours). A blood culture obtained from the implanted port on day 71 yielded 1 colony-forming unit of Lactobacillus species per mL. The patient became afebrile on the fourth day of therapy. Four additional blood cultures obtained from the port during the subsequent week yielded negative results. The patient received a total of 10 days of ampicillin therapy.
The patient was discharged from the hospital after 86 days of hospitalization, at which time she was tolerating partial enteral feedings, demonstrated resolution of her seroma, and otherwise had returned to her preadmission status. She had no signs of illness 1 month after discharge.
DNA Fingerprinting Analysis
To determine the involvement of probiotic Lactobacillus GG as a possible source of bacteremia, samples of the probiotic capsules that had been administered to the 2 patients were cultured. Lactobacillus isolates from the available blood cultures and the probiotic capsules were analyzed with repetitive element sequence-based polymerase chain reaction DNA fingerprinting. Extracted DNA was amplified with a Lactobacillus DiversiLab kit (Bacterial Barcodes, Houston, TX), according to previously described methods.13 The resulting DNA fingerprint patterns were viewed as electropherograms, and analysis was performed with DiversiLab software; Pearson correlation coefficients were used to determine similarity, and the unweighted pair-group method with arithmetic means was used to create a dendrogram. The 4 isolates appeared indistinguishable from one another, with similarity coefficients of >99% for each of the 2 patient pairs (Fig 1).
DISCUSSION
Lactobacillus species are ubiquitous anaerobic or facultative anaerobic Gram-positive rods that colonize the mucosal surfaces of the mouth, gastrointestinal tract, and genitourinary tract.14 Their presence in clinical specimens is usually considered commensal or contaminating. Although Lactobacillus species are increasingly identified as opportunistic pathogens among both immunocompetent and immunocompromised adult patients,15–17 invasive disease attributable to Lactobacillus species is rare among children and has been reported primarily for immunocompromised hosts.17–23 Among adults, endocarditis is the most frequently reported clinical presentation, typically associated with a structural heart defect.15 Rarely reported clinical syndromes caused by Lactobacillus species among children include bacteremia,17–19 neonatal sepsis and meningitis,20,21 pneumonia,22 and local suppurative infections.23 Lactobacillus bacteremia among children is very unusual, and few cases have been reported, primarily among immunocompromised hosts; the underlying conditions include acquired immunodeficiency syndrome,18 immunosuppression after bone marrow transplantation for treatment of aplastic anemia,19 and dental abscesses.17 Lactobacillus infections have been noted among adult patients with serious underlying diseases, including cancer, organ transplants, diabetes mellitus, and recent surgery.15 To our knowledge, our cases represent the first pediatric cases of probiotic agent-related invasive disease attributable to L rhamnosus GG.
Since the beginning of the probiotic era in the early 1900s,24 many studies have been conducted to elucidate the efficacy and safety of probiotics, especially Lactobacillus species.1–5 A national surveillance program in Finland found no increase in the number of cases of Lactobacillus bacteremia between 1990 and 2000, despite a substantial increase in the probiotic use of L rhamnosus during that period.25 A recent review concluded that the risk of infection with probiotic lactobacilli is similar to that of infection with commensal strains and that probiotic consumption presents a negligible risk to consumers, including immunocompromised individuals.4,26 Cases of Lactobacillus bacteremia have primarily occurred among patients with severe underlying diseases, which suggests a low virulence potential of Lactobacillus species.27
Because of occasional reports of infections associated with lactobacilli, including 2 well-documented adult cases suggested to be associated with probiotic strains,11,12 the safety of Lactobacillus strains used as probiotic agents has been the subject of some debate.4,28–30 The first reported case of probiotic agent-associated invasive disease involved a 74-year-old female patient with non–insulin-dependent diabetes mellitus who developed a liver abscess associated with a right-sided pleural empyema.11 She had consumed 1.5 L per day of dairy drinks containing L rhamnosus GG in the 4 months before presentation.11 The liver abscess yielded a L rhamnosus strain that was indistinguishable from the probiotic L rhamnosus GG strain in pulsed-field gel electrophoretic analyses.11 The second reported case of probiotic agent-related Lactobacillus infection involved a 67-year-old male patient with mild mitral valve regurgitation who developed endocarditis a few days after a dental extraction, despite antimicrobial prophylaxis. Blood cultures persistently yielded positive results for L rhamnosus. The Lactobacillus isolate was indistinguishable from the probiotic strains reportedly consumed by the patient, on the basis of identical appearances in culture, API 50CHL test results, antimicrobial susceptibility patterns, and pyrolysis mass spectrometry findings.12
In the adult literature, there have been occasional reports of invasive disease associated with other probiotic agents. S boulardii, a probiotic used for treatment of recurrent C difficile-associated diarrhea, was associated with fungemia among immunocompromised patients; molecular typing confirmed the clonality of strains isolated from patients and the S boulardii probiotic capsules.31,32 Other case reports suggested a role for probiotic lactobacilli in the pathogenesis of bacteremia and endocarditis,33–35 but those reports lacked sufficient details for definitive conclusions.
The pathogenesis of infection attributable to Lactobacillus species is poorly understood. Adhesion of probiotic strains to the intestinal mucosa and subsequent colonization are considered important prerequisites for probiotic action, because they prolong persistence in the intestine.36 One study found that blood culture isolates of Lactobacillus adhered to intestinal mucosa better than did dairy product strains, which suggests that the high degree of adhesion exhibited by some strains may play some role in the translocation of lactobacilli into the circulation.37 Impaired integrity of the gut, such as that caused by mucosal ulcerations or mucositis (eg, among patients with hematopoietic malignancies, who may receive cytotoxic chemotherapy or may undergo invasive procedures involving the gastrointestinal tract), may allow translocation of bacteria across oral or intestinal mucosal barriers and may lead to bacteremia and invasive disease.5
For our patients, recent enteral administration of Lactobacillus GG seemed to be the only plausible portal of entry. The bacteremia among our patients most likely originated from the probiotic stain, through bacterial translocation. Gut ischemia attributable to poor cardiac function might have compromised the gastrointestinal mucosa and facilitated the passage of the probiotic strain into the bloodstream. Alternatively, CVC contamination, either during opening of the probiotic capsule or through hand-related transmission, is possible.
Although the beneficial effects of probiotic agents for normal hosts, including children with diarrheal illnesses, are well documented, probiotic therapy may be associated occasionally with adverse effects, such as bacteremia, sepsis, or endocarditis, for a select subset of patients, such as immunocompromised or severely debilitated hosts. This report should not discourage the appropriate use of Lactobacillus or other probiotic agents but should serve as a reminder that these agents can cause invasive disease in certain populations. When Lactobacillus species are isolated from patients with invasive disease who are receiving probiotic preparations, molecular characterization of the patient and probiotic isolates may be helpful for elucidation of the origin of the patient strain.
ACKNOWLEDGMENTS
We thank Julie Bassett and James R. Lupski, MD, PhD, from Bacterial BarCodes, for their assistance with molecular typing of the Lactobacillus isolates.
FOOTNOTES
Accepted Sep 30, 2004.
No conflict of interest declared.
REFERENCES
Vanderhoof JA, Young RJ. Probiotics in pediatrics. Pediatrics. 2002;109 :956 –958
Alvarez-Olmos MI, Oberhelman RA. Probiotic agents and infectious diseases: a modern perspective on a traditional therapy. Clin Infect Dis. 2001;32 :1567 –1576
Salminen S, Arvilommi H. Probiotics demonstrated efficacy in clinical settings. Clin Infect Dis. 2001;32 :1577 –1578
Borriello SP, Hammes WP, Holzapfel W, et al. Safety of probiotics that contain lactobacilli or bifidobacteria. Clin Infect Dis. 2003;36 :775 –780
Ishibashi N, Shoji Y. Probiotics and safety. Am J Clin Nutr. 2001;73(suppl) :465S –470S
Gorbach SL. Probiotics and gastrointestinal health. Am J Gastroenterol. 2000;95(suppl) :S2 –S4
Szajewska H, Mrukowicz J. Probiotics in the treatment and prevention of acute infectious diarrhea in infants and children: a systematic review of published randomized, double-blind, placebo-controlled trials. J Pediatr Gastroenterol Nutr. 2001;33(suppl) :S17 –S25
Kalliomaki M, Salminen S, Arvilommi H, Kero P, Koskinen P, Isolauri E. Probiotics in primary prevention of atopic disease: a randomized placebo-controlled trial. Lancet. 2001;357 :1076 –1079
Van Neil CW, Feudtner C, Garrison MM. Lactobacillus therapy for acute infectious diarrhea in children: a meta-analysis. Pediatrics. 2002;109 :678 –684
Rautio M, Jousimies-Somer H, Kauma H, et al. Liver abscess due to a Lactobacillus rhamnosus strain indistinguishable from L rhamnosus strain GG. Clin Infect Dis. 1999;28 :1159 –1160
Mackay AD, Taylor MB, Kibbler CC, Hamilton-Miller JM. Lactobacillus endocarditis caused by a probiotic organism. Clin Microbiol Infect. 1999;5 :290 –292
Antonia MA, Hillier SL. DNA fingerprinting of Lactobacillus crispatus strain CTV-05 by repetitive element sequence-based PCR analysis in a pilot study of vaginal colonization. J Clin Microbiol. 2003;41 :1881 –1887
Isenberg HD, D'Amato RF. Indigenous and pathogenic micro-organisms of humans. In: Murray PR, Baron EJ, Pfaller MA, Tenover FC, Yolken RH, eds. Manual of Clinical Microbiology. 6th ed. Washington, DC: ASM Press; 1995:5–18
Husni RN, Gordon SM, Washington JA, Longworth DL. Lactobacillus bacteremia and endocarditis: review of 45 cases. Clin Infect Dis. 1997;25 :1048 –1055
Horwitch CA, Furseth HA, Larson AM, Jones TL, Ollife JF, Spach DH. Lactobacillemia in three patients with AIDS. Clin Infect Dis. 1995;21 :1460 –1462
Bayer AS, Chow AW, Betts D, Guze LB. Lactobacillemia: report of 9 cases: important clinical and therapeutic considerations. Am J Med. 1978;64 :808 –813
Schlegel L, Lemerle S, Geslin P. Lactobacillus species as opportunistic pathogens in immunocompromised patients. Eur J Clin Microbiol Infect Dis. 1998;17 :887 –888
Kalima P, Masterton RG, Roddie PH, Thomas AE. Lactobacillus rhamnosus infection in a child following bone marrow transplant. J Infect. 1996;32 :165 –167
Thompson C, McCarter YS, Krause PJ, Herson VC. Lactobacillus acidophilus sepsis in a neonate. J Perinatol. 2001;21 :258 –260
Brughton RA, Gruber WC, Haffar AA, Baker CJ. Neonatal meningitis due to Lactobacillus. Pediatr Infect Dis J. 1983;2 :382 –384
Sriskandan S, Lacey S, Fischer L. Isolation of vancomycin-resistant lactobacilli from three neutropenic patients with pneumonia. Eur J Clin Microbiol Infect Dis. 1993;12 :649 –650
Brook I. Isolation of non-sporing anaerobic rods from infections in children. J Med Microbiol. 1996;45 :21 –26
Fuller R. Probiotics in man and animals. J Appl Bacteriol. 1989;66 :365 –378
Salminen MK, Tynkkynen S, Rautelin H, et al. Lactobacillus bacteremia during a rapid increase in probiotic use of Lactobacillus rhamnosus GG in Finland. Clin Infect Dis. 2002;35 :1155 –1160
Naidu AS, Biblack WR, Clemens RA. Probiotic spectra of lactic acid bacteria (LAB). Crit Rev Food Sci Nutr. 1999;39 :13 –126
Saxelin M, Chuang NH, Chassy N, et al. Lactobacilli and bacteremia in southern Finland, 1989–1992. Clin Infect Dis. 1996;22 :564 –566
Sipsas NV, Zonious DI, Kordossis T. Safety of Lactobacillus strains used as probiotic agents. Clin Infect Dis. 2002;34 :1283 –1284
Salminen S, Arvilommi H. Safety of Lactobacillus strains used as probiotic agents [reply]. Clin Infect Dis. 2002;34 :1284 –1285
Reid G. Safety of Lactobacillus strains used as probiotic agents. Clin Infect Dis. 2002;35 :349 –350
Riquelme AJ, Calvo MA, Guzman AM, et al. Saccharomyces cervisiae fungemia after Saccharomyces boulardii treatment in immunocompromised patients. J Clin Gastroenterol. 2003;36 :41 –43
Bassetti S, Frei R, Zimmerli W. Fungemia with Saccharomyces cervisiae after treatment with Saccharomyces boulardii. Am J Med. 1998;105 :71 –72
Barton LL, Rider ED, Coen RW. Bacteremic infection with Pediococcus: vancomycin-resistant opportunist. Pediatrics. 2001;107 :775 –776
Avlami A, Kordossis T, Vrizidis N, Sipsas NV. Lactobacillus rhamnosus endocarditis complicating colonoscopy. J Infect. 2001;42 :283 –285
Presterl E, Kneifel W, Mayer HK, Zehetgruber M, Makristathis A, Graninger W. Endocarditis by Lactobacillus rhamnosus due to yogurt ingestion Scand J Infect Dis. 2001;33 :710 –714
Saarela M, Mogensen G, Fonden R, Matto J, Matilla-Sandholm T. Probiotic bacteria: safety, functional and technological properties. J Biotechnol. 2000;84 :197 –215
Apostolou, E, Kirjavainen PV, Saxelin M, et al. Good adhesion properties of probiotics: a potential risk for bacteremia FEMS Immunol Med Microbiol. 2001;31 :35 –39(Michael H. Land, MD, Kell)
ABSTRACT
Probiotic strains of lactobacilli are increasingly being used in clinical practice because of their many health benefits. Infections associated with probiotic strains of lactobacilli are extremely rare. We describe 2 patients who received probiotic lactobacilli and subsequently developed bacteremia and sepsis attributable to Lactobacillus species. Molecular DNA fingerprinting analysis showed that the Lactobacillus strain isolated from blood samples was indistinguishable from the probiotic strain ingested by the patients. This report indicates, for the first time, that invasive disease can be associated with probiotic lactobacilli. This report should not discourage the appropriate use of Lactobacillus or other probiotic agents but should serve as a reminder that these agents can cause invasive disease in certain populations.
Key Words: Lactobacillus probiotic bacteremia sepsis
Abbreviations: CVC, central venous catheter MIC, minimal inhibitory concentration
In recent years, several articles have reviewed the efficacy, mechanism of action, and safety of probiotics in the treatment of infectious disease.1–5 The term probiotic refers to "live microorganisms which when administered in adequate amounts confer a health benefit on the host."6 The microorganisms most frequently used as probiotic agents are lactic acid bacteria (species of Lactobacillus, Enterococcus, and Bifidobacterium) and nonpathogenic, antibiotic-resistant, ascosporic yeasts, especially Saccharomyces boulardii.2 Lactobacillus rhamnosus strain GG (ATCC53103), which was originally isolated from human intestinal flora, is the most widely studied probiotic agent for adults and children.7 L rhamnosus strain GG can prevent diarrhea and atopy among children.8,9 A recent meta-analysis suggested that Lactobacillus is a safe effective treatment for children with acute infectious diarrhea.10
Serious infections attributable to probiotic lactobacilli are extremely rare. Only 2 well-documented adult cases have been reported in which the L rhamnosus strain isolated from the clinical specimen was indistinguishable from the probiotic strain consumed by the patient.11,12 We report the first pediatric case of invasive disease attributable to a Lactobacillus strain indistinguishable from the L rhamnosus strain GG administered to the infant.
CASE REPORTS
Case 1
A 6-week-old, white, male patient was admitted for scheduled repair of a double-outlet right ventricle and pulmonic stenosis. The birth weight was 3.2 kg, and the patient was born at term. The postoperative course was complicated by pacemaker placement, pulmonary artery banding, seizures, acute renal insufficiency, and prolonged respiratory support. The infant received several 7- to 10-day courses of broad-spectrum antibiotics, including vancomycin and ceftriaxone, for treatment of suspected sepsis; however, blood and urine cultures remained negative. On day 57 of hospitalization, the patient developed copious nonbloody diarrhea. He had completed his last antibiotic course 3 days before the onset of diarrhea. The patient was afebrile, and the diarrhea persisted despite changes in formula. It was presumed that the diarrhea was a consequence of the courses of broad-spectrum antibiotics. Results of routine stool cultures, an enzyme immunoassay for Rotavirus, and an assay for Clostridium difficile toxin were negative. Beginning on day 79 of hospitalization, Lactobacillus GG (Culturelle, 10 x 109 cells per capsule; ConAgra Foods, Omaha, NE) was administered through the gastrostomy tube (1 capsule mixed with formula, once daily) for its probiotic effect. The patient's diarrhea improved with probiotic therapy but, on day 99 of hospitalization, while continuing to receive probiotic therapy, the patient developed new-onset fever and marked leukocytosis and appeared ill.
A physical examination revealed a toxic pale infant with a temperature of 39.7°C, pulse of 171 beats per minute, respiratory rate of 80 breaths per minute, and blood pressure of 60/30 mm Hg. Ventilatory support was maintained with a bilevel airway pressure of 18 cm H2O and a continuous positive airway pressure of 8 cm H2O. Respiratory distress was evidenced by sternal and intercostal retractions. Bilateral coarse breath sounds were noted. A grade 3/6 pansystolic murmur was audible over the precordium. The extremities were cool and mottled, with poor peripheral perfusion. The Broviac central venous catheter (CVC) exit site and subcutaneous tract were without redness, warmth, discharge, tenderness, or induration. The patient showed no peripheral signs of endocarditis. The rest of the examination results were unremarkable.
The results of laboratory tests were as follows: white blood cell count of 38 900 cells per mm3, with 87% segmented neutrophils and 10% bands. Urinalysis results were normal, and electrolyte levels were consistent with metabolic acidosis. A blood culture obtained from the CVC yielded Gram-positive rods. Chest radiographic findings were normal. Transesophageal echocardiography revealed a 10- x 12-mm echogenic mass at the tip of the CVC (superior vena cava-right atrium junction), firmly adherent to the posterior wall of the right atrium. No vegetations were noted on the heart valves. Intravenous vancomycin (40 mg/kg per day [10 mg/kg every 6 hours]) and gentamicin (5 mg/kg per day [2.5 mg/kg every 12 hours]) therapy was initiated.
In the subsequent 72 hours, 3 additional blood cultures (2 central and 1 peripheral) obtained at different times yielded >100 colony-forming units of Gram-positive rods per mL. All 4 isolates were identified as Lactobacillus species. The isolate was susceptible to penicillin G, erythromycin, clindamycin, gatifloxacin, tetracycline, and trimethoprim-sulfamethoxazole, with minimal inhibitory concentrations (MICs) of 1, <0.25, <0.25, <0.25, <1, and <10 mg/L, respectively, and was resistant to vancomycin, meropenem, ceftriaxone, and cefuroxime, with MICs of >4, >0.5, >2, and >2 mg/L, respectively. When antimicrobial susceptibilities became available, antibiotic therapy was changed to high-dose penicillin G (400 000 U/kg per day, every 4 hours); gentamicin was continued for its synergistic effect. When blood cultures were positive for Lactobacillus species on 3 consecutive days, the CVC was promptly removed and oral Lactobacillus GG therapy was discontinued.
In the subsequent 48 to 72 hours, the patient exhibited clinical improvement, with resolution of respiratory distress and improvement in perfusion. Low-grade fevers persisted but were <38°C. The white blood cell count normalized at 7700 cells per mm3 (56% segmented neutrophils, 0% bands). Repeat blood cultures obtained 48 hours after the initiation of high-dose penicillin G therapy yielded negative results. In the subsequent 5 weeks, serial echocardiograms revealed resolution of the thrombus at the superior vena cava-right atrium junction. The patient completed a 6-week course of penicillin G and gentamicin treatment for his presumed endocarditis-equivalent infection.
Case 2
A 6-year-old, white, female patient with cerebral palsy, microcephaly, mental retardation, and a seizure disorder, who required feeding through a gastrojejunostomy tube, was admitted for treatment of a urinary tract infection. One week before hospitalization, she had undergone revision of a spinal rod to improve her scoliosis. She also had developed regurgitation of feedings and abdominal pain associated with feedings in the week before admission.
The patient's urine culture yielded Escherichia coli, for which she received a 10-day course of ceftriaxone. The fever resolved but the abdominal pain with feedings persisted, with development of diarrhea. Therefore, enteral feedings were stopped and parenteral nutrition was initiated through a previously implanted CVC. On day 5 of hospitalization, the patient began treatment with vancomycin, after developing a fluid collection over the posterior portion of her left spinal rod. Vancomycin treatment was discontinued when culture of the fluid yielded oxacillin-susceptible Staphylococcus aureus, and the patient was treated with intravenously administered oxacillin for 2 weeks.
On hospital day 18, the patient developed catheter-related enterococcal sepsis, which was treated with vancomycin for 2 days followed by ampicillin plus gentamicin for 14 days. Follow-up blood cultures on the fourth and sixth days of this treatment course yielded negative results.
Throughout the first 2 months, the patient continued to experience intermittent diarrhea, which was sometimes explosive and was not always contained within her diapers. Multiple stool assays for C difficile toxin and assays for Rotavirus yielded negative results. Because antibiotic-associated diarrhea was a possibility, Lactobacillus GG (Culturelle, 10 x 109 cells per capsule; ConAgra Foods) was administered beginning on day 25 of hospitalization, for its probiotic effect; each day, the contents of 1 capsule were mixed with water and the mixture was injected into the gastrojejunostomy tube with a syringe. The syringe was prepared in a medication room remote from the patient's room. The patient remained hospitalized for ongoing evaluation and management of her gastrointestinal feeding intolerance.
The patient became febrile to 40.3°C on day 69 of hospitalization. Her pulse rate, blood pressure, and respiratory rate remained stable within the ranges of the previous few weeks. She was more irritable and less active than usual but otherwise demonstrated no new findings in her physical examination. The results of laboratory tests were as follows: white blood cell count of 9800 cells per mm3, with 87% segmented neutrophils, 7% bands, and 1% metamyelocytes. Chest radiographic findings, urinalysis results, and serum electrolyte levels were normal. Intravenous vancomycin (40 mg/kg per day, every 8 hours) and ceftazidime (150 mg/kg per day, every 8 hours) therapy was initiated.
A peripheral blood culture obtained on day 69 of hospitalization yielded >100 colony-forming units of Lactobacillus species per mL, with MICs identical to those in case 1 except that the penicillin MIC was 0.5 mg/L (vancomycin was not tested). After these culture results became known on the third day of illness (hospital day 71), Lactobacillus GG treatment was discontinued and intravenous antibiotic therapy was changed to ampicillin (100 mg/kg per day, every 6 hours). A blood culture obtained from the implanted port on day 71 yielded 1 colony-forming unit of Lactobacillus species per mL. The patient became afebrile on the fourth day of therapy. Four additional blood cultures obtained from the port during the subsequent week yielded negative results. The patient received a total of 10 days of ampicillin therapy.
The patient was discharged from the hospital after 86 days of hospitalization, at which time she was tolerating partial enteral feedings, demonstrated resolution of her seroma, and otherwise had returned to her preadmission status. She had no signs of illness 1 month after discharge.
DNA Fingerprinting Analysis
To determine the involvement of probiotic Lactobacillus GG as a possible source of bacteremia, samples of the probiotic capsules that had been administered to the 2 patients were cultured. Lactobacillus isolates from the available blood cultures and the probiotic capsules were analyzed with repetitive element sequence-based polymerase chain reaction DNA fingerprinting. Extracted DNA was amplified with a Lactobacillus DiversiLab kit (Bacterial Barcodes, Houston, TX), according to previously described methods.13 The resulting DNA fingerprint patterns were viewed as electropherograms, and analysis was performed with DiversiLab software; Pearson correlation coefficients were used to determine similarity, and the unweighted pair-group method with arithmetic means was used to create a dendrogram. The 4 isolates appeared indistinguishable from one another, with similarity coefficients of >99% for each of the 2 patient pairs (Fig 1).
DISCUSSION
Lactobacillus species are ubiquitous anaerobic or facultative anaerobic Gram-positive rods that colonize the mucosal surfaces of the mouth, gastrointestinal tract, and genitourinary tract.14 Their presence in clinical specimens is usually considered commensal or contaminating. Although Lactobacillus species are increasingly identified as opportunistic pathogens among both immunocompetent and immunocompromised adult patients,15–17 invasive disease attributable to Lactobacillus species is rare among children and has been reported primarily for immunocompromised hosts.17–23 Among adults, endocarditis is the most frequently reported clinical presentation, typically associated with a structural heart defect.15 Rarely reported clinical syndromes caused by Lactobacillus species among children include bacteremia,17–19 neonatal sepsis and meningitis,20,21 pneumonia,22 and local suppurative infections.23 Lactobacillus bacteremia among children is very unusual, and few cases have been reported, primarily among immunocompromised hosts; the underlying conditions include acquired immunodeficiency syndrome,18 immunosuppression after bone marrow transplantation for treatment of aplastic anemia,19 and dental abscesses.17 Lactobacillus infections have been noted among adult patients with serious underlying diseases, including cancer, organ transplants, diabetes mellitus, and recent surgery.15 To our knowledge, our cases represent the first pediatric cases of probiotic agent-related invasive disease attributable to L rhamnosus GG.
Since the beginning of the probiotic era in the early 1900s,24 many studies have been conducted to elucidate the efficacy and safety of probiotics, especially Lactobacillus species.1–5 A national surveillance program in Finland found no increase in the number of cases of Lactobacillus bacteremia between 1990 and 2000, despite a substantial increase in the probiotic use of L rhamnosus during that period.25 A recent review concluded that the risk of infection with probiotic lactobacilli is similar to that of infection with commensal strains and that probiotic consumption presents a negligible risk to consumers, including immunocompromised individuals.4,26 Cases of Lactobacillus bacteremia have primarily occurred among patients with severe underlying diseases, which suggests a low virulence potential of Lactobacillus species.27
Because of occasional reports of infections associated with lactobacilli, including 2 well-documented adult cases suggested to be associated with probiotic strains,11,12 the safety of Lactobacillus strains used as probiotic agents has been the subject of some debate.4,28–30 The first reported case of probiotic agent-associated invasive disease involved a 74-year-old female patient with non–insulin-dependent diabetes mellitus who developed a liver abscess associated with a right-sided pleural empyema.11 She had consumed 1.5 L per day of dairy drinks containing L rhamnosus GG in the 4 months before presentation.11 The liver abscess yielded a L rhamnosus strain that was indistinguishable from the probiotic L rhamnosus GG strain in pulsed-field gel electrophoretic analyses.11 The second reported case of probiotic agent-related Lactobacillus infection involved a 67-year-old male patient with mild mitral valve regurgitation who developed endocarditis a few days after a dental extraction, despite antimicrobial prophylaxis. Blood cultures persistently yielded positive results for L rhamnosus. The Lactobacillus isolate was indistinguishable from the probiotic strains reportedly consumed by the patient, on the basis of identical appearances in culture, API 50CHL test results, antimicrobial susceptibility patterns, and pyrolysis mass spectrometry findings.12
In the adult literature, there have been occasional reports of invasive disease associated with other probiotic agents. S boulardii, a probiotic used for treatment of recurrent C difficile-associated diarrhea, was associated with fungemia among immunocompromised patients; molecular typing confirmed the clonality of strains isolated from patients and the S boulardii probiotic capsules.31,32 Other case reports suggested a role for probiotic lactobacilli in the pathogenesis of bacteremia and endocarditis,33–35 but those reports lacked sufficient details for definitive conclusions.
The pathogenesis of infection attributable to Lactobacillus species is poorly understood. Adhesion of probiotic strains to the intestinal mucosa and subsequent colonization are considered important prerequisites for probiotic action, because they prolong persistence in the intestine.36 One study found that blood culture isolates of Lactobacillus adhered to intestinal mucosa better than did dairy product strains, which suggests that the high degree of adhesion exhibited by some strains may play some role in the translocation of lactobacilli into the circulation.37 Impaired integrity of the gut, such as that caused by mucosal ulcerations or mucositis (eg, among patients with hematopoietic malignancies, who may receive cytotoxic chemotherapy or may undergo invasive procedures involving the gastrointestinal tract), may allow translocation of bacteria across oral or intestinal mucosal barriers and may lead to bacteremia and invasive disease.5
For our patients, recent enteral administration of Lactobacillus GG seemed to be the only plausible portal of entry. The bacteremia among our patients most likely originated from the probiotic stain, through bacterial translocation. Gut ischemia attributable to poor cardiac function might have compromised the gastrointestinal mucosa and facilitated the passage of the probiotic strain into the bloodstream. Alternatively, CVC contamination, either during opening of the probiotic capsule or through hand-related transmission, is possible.
Although the beneficial effects of probiotic agents for normal hosts, including children with diarrheal illnesses, are well documented, probiotic therapy may be associated occasionally with adverse effects, such as bacteremia, sepsis, or endocarditis, for a select subset of patients, such as immunocompromised or severely debilitated hosts. This report should not discourage the appropriate use of Lactobacillus or other probiotic agents but should serve as a reminder that these agents can cause invasive disease in certain populations. When Lactobacillus species are isolated from patients with invasive disease who are receiving probiotic preparations, molecular characterization of the patient and probiotic isolates may be helpful for elucidation of the origin of the patient strain.
ACKNOWLEDGMENTS
We thank Julie Bassett and James R. Lupski, MD, PhD, from Bacterial BarCodes, for their assistance with molecular typing of the Lactobacillus isolates.
FOOTNOTES
Accepted Sep 30, 2004.
No conflict of interest declared.
REFERENCES
Vanderhoof JA, Young RJ. Probiotics in pediatrics. Pediatrics. 2002;109 :956 –958
Alvarez-Olmos MI, Oberhelman RA. Probiotic agents and infectious diseases: a modern perspective on a traditional therapy. Clin Infect Dis. 2001;32 :1567 –1576
Salminen S, Arvilommi H. Probiotics demonstrated efficacy in clinical settings. Clin Infect Dis. 2001;32 :1577 –1578
Borriello SP, Hammes WP, Holzapfel W, et al. Safety of probiotics that contain lactobacilli or bifidobacteria. Clin Infect Dis. 2003;36 :775 –780
Ishibashi N, Shoji Y. Probiotics and safety. Am J Clin Nutr. 2001;73(suppl) :465S –470S
Gorbach SL. Probiotics and gastrointestinal health. Am J Gastroenterol. 2000;95(suppl) :S2 –S4
Szajewska H, Mrukowicz J. Probiotics in the treatment and prevention of acute infectious diarrhea in infants and children: a systematic review of published randomized, double-blind, placebo-controlled trials. J Pediatr Gastroenterol Nutr. 2001;33(suppl) :S17 –S25
Kalliomaki M, Salminen S, Arvilommi H, Kero P, Koskinen P, Isolauri E. Probiotics in primary prevention of atopic disease: a randomized placebo-controlled trial. Lancet. 2001;357 :1076 –1079
Van Neil CW, Feudtner C, Garrison MM. Lactobacillus therapy for acute infectious diarrhea in children: a meta-analysis. Pediatrics. 2002;109 :678 –684
Rautio M, Jousimies-Somer H, Kauma H, et al. Liver abscess due to a Lactobacillus rhamnosus strain indistinguishable from L rhamnosus strain GG. Clin Infect Dis. 1999;28 :1159 –1160
Mackay AD, Taylor MB, Kibbler CC, Hamilton-Miller JM. Lactobacillus endocarditis caused by a probiotic organism. Clin Microbiol Infect. 1999;5 :290 –292
Antonia MA, Hillier SL. DNA fingerprinting of Lactobacillus crispatus strain CTV-05 by repetitive element sequence-based PCR analysis in a pilot study of vaginal colonization. J Clin Microbiol. 2003;41 :1881 –1887
Isenberg HD, D'Amato RF. Indigenous and pathogenic micro-organisms of humans. In: Murray PR, Baron EJ, Pfaller MA, Tenover FC, Yolken RH, eds. Manual of Clinical Microbiology. 6th ed. Washington, DC: ASM Press; 1995:5–18
Husni RN, Gordon SM, Washington JA, Longworth DL. Lactobacillus bacteremia and endocarditis: review of 45 cases. Clin Infect Dis. 1997;25 :1048 –1055
Horwitch CA, Furseth HA, Larson AM, Jones TL, Ollife JF, Spach DH. Lactobacillemia in three patients with AIDS. Clin Infect Dis. 1995;21 :1460 –1462
Bayer AS, Chow AW, Betts D, Guze LB. Lactobacillemia: report of 9 cases: important clinical and therapeutic considerations. Am J Med. 1978;64 :808 –813
Schlegel L, Lemerle S, Geslin P. Lactobacillus species as opportunistic pathogens in immunocompromised patients. Eur J Clin Microbiol Infect Dis. 1998;17 :887 –888
Kalima P, Masterton RG, Roddie PH, Thomas AE. Lactobacillus rhamnosus infection in a child following bone marrow transplant. J Infect. 1996;32 :165 –167
Thompson C, McCarter YS, Krause PJ, Herson VC. Lactobacillus acidophilus sepsis in a neonate. J Perinatol. 2001;21 :258 –260
Brughton RA, Gruber WC, Haffar AA, Baker CJ. Neonatal meningitis due to Lactobacillus. Pediatr Infect Dis J. 1983;2 :382 –384
Sriskandan S, Lacey S, Fischer L. Isolation of vancomycin-resistant lactobacilli from three neutropenic patients with pneumonia. Eur J Clin Microbiol Infect Dis. 1993;12 :649 –650
Brook I. Isolation of non-sporing anaerobic rods from infections in children. J Med Microbiol. 1996;45 :21 –26
Fuller R. Probiotics in man and animals. J Appl Bacteriol. 1989;66 :365 –378
Salminen MK, Tynkkynen S, Rautelin H, et al. Lactobacillus bacteremia during a rapid increase in probiotic use of Lactobacillus rhamnosus GG in Finland. Clin Infect Dis. 2002;35 :1155 –1160
Naidu AS, Biblack WR, Clemens RA. Probiotic spectra of lactic acid bacteria (LAB). Crit Rev Food Sci Nutr. 1999;39 :13 –126
Saxelin M, Chuang NH, Chassy N, et al. Lactobacilli and bacteremia in southern Finland, 1989–1992. Clin Infect Dis. 1996;22 :564 –566
Sipsas NV, Zonious DI, Kordossis T. Safety of Lactobacillus strains used as probiotic agents. Clin Infect Dis. 2002;34 :1283 –1284
Salminen S, Arvilommi H. Safety of Lactobacillus strains used as probiotic agents [reply]. Clin Infect Dis. 2002;34 :1284 –1285
Reid G. Safety of Lactobacillus strains used as probiotic agents. Clin Infect Dis. 2002;35 :349 –350
Riquelme AJ, Calvo MA, Guzman AM, et al. Saccharomyces cervisiae fungemia after Saccharomyces boulardii treatment in immunocompromised patients. J Clin Gastroenterol. 2003;36 :41 –43
Bassetti S, Frei R, Zimmerli W. Fungemia with Saccharomyces cervisiae after treatment with Saccharomyces boulardii. Am J Med. 1998;105 :71 –72
Barton LL, Rider ED, Coen RW. Bacteremic infection with Pediococcus: vancomycin-resistant opportunist. Pediatrics. 2001;107 :775 –776
Avlami A, Kordossis T, Vrizidis N, Sipsas NV. Lactobacillus rhamnosus endocarditis complicating colonoscopy. J Infect. 2001;42 :283 –285
Presterl E, Kneifel W, Mayer HK, Zehetgruber M, Makristathis A, Graninger W. Endocarditis by Lactobacillus rhamnosus due to yogurt ingestion Scand J Infect Dis. 2001;33 :710 –714
Saarela M, Mogensen G, Fonden R, Matto J, Matilla-Sandholm T. Probiotic bacteria: safety, functional and technological properties. J Biotechnol. 2000;84 :197 –215
Apostolou, E, Kirjavainen PV, Saxelin M, et al. Good adhesion properties of probiotics: a potential risk for bacteremia FEMS Immunol Med Microbiol. 2001;31 :35 –39(Michael H. Land, MD, Kell)