Report of Two Cases of Recurrent Clostridium Difficile Infections in Children and Literature Review

Wilson Daza Carreño MD (1), Silvana Dadán (2), Michelle Higuera MD (3)

(1) Pediatric Gastroenterologist, Director of the Gastroenterology, Hepatology and Nutrition Unit of Gastronutriped, Associate Professor and Director of Graduate Pediatric Gastroenterology program at the Universidad El Bosque in Bogotá DC, Colombia. Mail: wilson.daza@gastronutriped.com.

(2) Nutritionist and Dietitian. Master of Clinical Nutrition, Gastroenterology, Hepatology and Nutrition Unit of Gastronutriped, Associate Professor of Pediatrics and Pediatric Gastroenterology at the Universidad El Bosque in Bogotá DC, Colombia. Mail: silvana.dadan@gastronutriped.com

(3) Pediatrician, Professor and Researcher in the Pediatric Gastroenterology, Hepatology and Nutrition Unit in Bogotá DC, Colombia. Mail: michellehiguera@yahoo.com

Received:    01-05-15   Accepted:    26-01-16

Abstract

Background: Indiscriminate use of antibiotics is not only related to resistance, but also to increased incidence of some bacterial infections such as Clostridium difficile (CD). Despite appropriate treatment, these infections have a high recurrence rate which requires appropriate management and monitoring guidelines. Methodology: This article presents the cases of two pediatric patients who experienced CD reinfection on more than one episode. Patients were treated in Gastroenterology, Hepatology and Nutrition Unit of Gastronutriped in Bogotá, Colombia. In addition, the article provides an update on identification of risk factors, diagnosis and treatment of recurrent CD infections. Conclusions: Cases of CD infections in the pediatric population present a challenge because of their similarity to other infectious processes. When there are bloody stools in a patient with a history of prior use of antibiotics, the infection should be suspected. Reinfection can occur up to 3 months after the initial infection. Management of the first reinfection in mild cases does not require antibiotics, but moderate or severe may be treated with metronidazole if it was used for initial treatment. In cases of a second reinfection, vancomycin should be used. The most appropriate treatment for cases of three or more episodes is still a matter of debate. In recent years, the use of probiotics and fecal microbiota transplantation have shown great benefits in cases of reinfection, though the available evidence is still inconclusive.

Keywords

Clostridium difficile, pediatrics, antibacterial, diarrhea.

INTRODUCTION

Clostridium difficile (CD) are spore-forming anaerobic bacteria. (1, 2) Rising incidence has led to increased mortality secondary to this infection. (3, 4) About two-thirds of all cases occur in the community while the rest occur in hospitals. (2) Even though many children are colonized with CD and are asymptomatic carriers. (5) Few develop the disease, and in most of these cases the symptoms resolve with proper management. (6, 7) Nevertheless, up to 19.6% of pediatric cases develop into recurrent clostridium difficile (RCD) which requires treatment. (8)

The intestinal microbiota plays multiple roles. (7) It can be altered by several factors, including route of delivery (Caesarian section), absence or short duration of breastfeeding, early initiation of complementary feeding, prematurity, the use of agents to suppress gastric acid and the use of antimicrobials. (7, 9, 10, 11)  Individually or together, these factors favor colonization by several pathogens including CD. (12) Increased incidence of Ruminococcus and Klebsiella has been in infants colonized by Clostridium difficile. (13) CD infections have also been observed in patients with inflammatory bowel disease (IBD) and patients with cystic fibrosis (CF). (14, 15)

We present the cases of two pediatric with recurrent CD infections who were cared for at Gastronutriped a  Gastroenterology, Hepatology and Nutrition Unit. We also present a  supplementary literature review with emphasis on the management of recurrent CD infections.

CLINICAL CASES

Case 1

The patient was a three year and nine month old girl who had lived alternately in the United States of America and Bogotá during the year prior to coming to the clinic. In the United States she developed respiratory symptoms which were diagnosed as bronchitis and treated with amoxicillin and clavulanic acid for three days. Six days after completing treatment, she developed hematochezia associated with very intense cramping abdominal pain. She was taken to an emergency department in the USA and found to be positive for CD toxin . She was treated with 10 mg/kg of metronidazole treatment plus saccharomyces boulardii for 10 days and put on a dairy-free diet.

Six days after the start of treatment, hematochezia associated with cramping abdominal pain recurred, and the child also developed diarrhea. A pediatric gastroenterologist reconfirmed CD with three cultures. Metronidazole at 10 mg/kg was reinstated for 21 days. The patient’s symptoms improved, but within four days of the end of treatment hematochezia recurred although without abdominal pain. By this time, the patient was in Colombia. She was taken to a pediatric infectious disease specialist who suspected a relapse and prescribed 40 mg/kg/day of oral vancomycin administered every eight hours for 14 plus continuation of saccharomyces boulardii plus administration of zinc sulfate.

The patient’s first consultation at Gastronutriped happened eight weeks after onset of initial respiratory symptoms and the patient had been seen by the pediatric infectious disease specialist.  Failure to thrive was evident on physical examination. The child and members of her family had medical histories of atopic diseases. Studies to rule out food allergies, Celiac disease, IBD and cystic fibrosis were conducted. A colonoscopy was performed to check recurrent lower gastrointestinal bleeding. Multivitamins were added to the girl’s treatment and the dosage of vancomycin was adjusted to 30 mg/kg/day administered every six hours for 14 days. At subsequent consultations, immune system pathologies were discarded.

At the end of antibiotic treatment, the child suffered another relapse with Bristol type 6 and 7 stool. Probiotics and zinc were prescribed, but no additional antibiotics were prescribed. Laboratory tests were negative for CD toxins A and B but were positive for  Immunoglobulin G for Saccharomyces cerevisiae. An upper endoscopy was conducted because colonoscopy was not authorized by the child’s parents. Other studies are presented in Table 1. At her last follow-up examination 12 weeks after her first assessment at Gastronutriped, the patient’s evolution was adequate, but results of tests to rule out CF (pancreatic fecal elastase and electrolytes in perspiration) were still pending. Treatment with saccharomyces boulardii continued at a dosage of 200 mg every 12 hours to complete 3 months.

Case 2

This patient was a seven-year-old boy from Bogotá who, two months after taking 30 mg/kg/day of cefuroxime for 10 days for rhinitis and sinusitis, suffered a two-day bout of non-explosive diarrhea with mucus and blood. Subsequently his stools became Bristol types 5 and 6. The child was brought to the emergency room of a local institution which began outpatient management with oral rehydration solution and bifidobacteria and freeze-dried live lactobacilli. The patient did not improvement and was reexamined seven days later. The patient was diagnosed with amebiasis after stool samples showed trophozoites of Entamoeba histolyitica and Entamoeba dispar. Treatment was begun with 30 mg/kg/ day of metronidazole and 5 mg/kg/day of nifuroxazide for 7 days.

Symptoms persisted, so the boy was hospitalized and treated with 50 mg/kg/day of nalidixic acid for 7 days together with intravenous hydration and oral Bacillus clausii. Three days after this treatment began, white membranes began to appear in the patient’s stool. Lab studies for CD were positive for toxins A and B. A colonoscopy was performed. Within 24 hours, the patient became feverish. He was treated with 30 mg/kg/day of metronidazole and 100 mg/kg/day of ampicillin for 7 days. After completing a total of 10 days of treatment in the hospital, he was discharged with a prescription for symbiotic treatment with bifidobacteria, lactobacilli, inulin and fructooligosaccharides together with metronidazole. Five days later he suffered a relapse, but stool samples showed no microorganisms. It was decided to treat the patient with 10 mg/kg/ day of pyrantel pamoate for three days.

The patient came to Gastronutriped two months after the onset of symptoms. Two days prior to consultation, his had sparse Bristol 6 stool with tenesmus and blood but without other symptoms. He was treated with 50 mg of Saccharomyces boulardii 2 every 12 hours. Acute malnutrition was evident. He was diagnosed with colitis due to CD and post enteritis syndrome due to persistent diarrhea. He was treated with sugar-free zinc gluconate, L-glutamine, Lactobacillus reuteri, protein supplement, and Saccharomyces boulardii was continued. Testing for IBD and CF began. Eight days after the first consultation at Gastronutriped,  he suffered a four-day recurrence of diarrhea with mucus and blood without fever or general commitment. The patient was taken to the emergency room, where outpatient management with oral rehydration solution, diosmectite, and continued Saccharomyces boulardii was prescribed. A follow-up examination 24 days after the first consultation in Gastronutriped showed a resolution of symptoms, evident weight gain, and symptoms suggestive of constipation. Polyethylene glycol without electrolytes was prescribed and nutritional recommendations were made. Laboratory tests were positive for fecal calprotectin (> 300mcg/g) indicating the possibility of IBD. Other test results are presented in Table 2.

Tests for fecal calprotectin continued to be positive (> 300 mcg/g) at the patient’s six month follow-up examination in Gastronutriped, so we decided to perform an upper endoscopy and colonoscopy (Figures 1 and 2). Biopsies from the upper endoscopy (Figure 3) indicated grade I esophagitis, mild chronic foveolar gastritis, grade I duodenitis, but no Helicobacter pylori. Treatment with 20 mg of Omeprazole every 12 hours for 3 months was begun. The patient’s evolution was favorable and without gastrointestinal symptom. At a follow-up examination a month and a half later, the patient tested negative for fecal calprotectin. Diagnoses of CF and IBD were ruled out.

DISCUSSION

A recurrent clostridium difficile (RCD) infection is defined as laboratory confirmation of a new episode of diarrhea within 90 days of resolution of the initial episode and at least ten days after discontinuation or completion of therapy for CD. (16) Infections recur in up to 20% of patients despite successful treatment of initial infection. The risk increases to 65% for those who have a previous history. (17) Recurrent infections most often occur within a week treatment interruption. (18, 19)

There are many factors that predisposing patients to development, persistence and recurrence of CD infections, but the most important independent risk factor for infection is treatment with antibiotics. This is true at the community level and at the level of the health care provider. (20-22) Risk varies according to the type of antibiotic administered (Table 3). The use of agents that suppress gastric acid is another risk factor, (23, 24) but this topic is still under review. Studies that evaluated the effect of these drugs in children infected with CD have found finding that recurrence of infection in 10.4% of those treated and that relapses occurred most frequently among patients who had taken medications to suppress gastric acid. (25)

Other factors that have been mentioned as posing risks for CD infections include previous hospitalization, previous surgical procedures, immunosuppression and use of oral or nasal gastric feeding tubes. (24) Factors associated with RCD include severity of disease, extreme age, and parallel antibiotic therapy for another infection. (1) The two patients discussed in this article did not present these risk factors although CD did reoccur.

The clinical spectrum of CD infections ranges from asymptomatic colonization through mild illnesses to severely complicated presentations (Table 4). Severity of the infection and mortality are associated with extreme ages, however instances of the entire spectrum occur at all ages. (1, 2, 26, 27) It is important to note, that there are reports of onset of symptoms two to three months after exposure to antibiotics or other risk factors. (1, 2, 18) The patients studies here had mild clinical manifestations that developed a week after antibiotic treatment had ended. This correlates with findings from the literature. (1, 2, 21)

Egressy et al. described infected adults with both CD and CF who had atypical symptoms and increased risk of pancolitis. Therefore, when a patient has other comorbidities such as CF, and if several risk factors are combined, a CD infection should be suspected. (28)

Diagnosis is based on a combination of clinical and laboratory findings: clinically significant diarrhea, a positive lab test for CD, or colonoscopic evidence of pseudomembranous colitis. (24)

The key diagnostic reference studies used over the past 30 years have been the CD cytotoxin neutralization assay and toxigenic culturing of stool samples. The latter, despite frequent prescription, is not currently recommended because it fails to detect toxin producing strains which is important considering the high frequency of colonization in children. (29, 30) The enzyme immunoassay test for stool toxin A and/or B is widely used for diagnosis of CD infections. Other studies, such as specific antigen tests for glutamate dehydrogenase, an enzyme produced by all strains of CD, have been implemented and studied in recent years. Nevertheless, it cannot distinguish between toxigenic and non-toxigenic strains which makes it is useful only as a screening test in a diagnostic algorithm in which samples that test positive must be subjected to additional tests. (31) This may be an alternative in the future. (2) The patients studied in this article both tested positive for Toxins A and B which confirmed the diagnosis. Subsequent to treatment, they tested negative which helped the monitoring of these patients.

Some studies have report that leukocytosis, leukopenia, neutrophilia, impaired renal function, and increased acute phase reactants can be used as indicators of CD infections, but these parameters can be within normal ranges, as happened in our patients, but a CD diagnosis cannot be excluded. These factors are more useful for assessing complications. (1, 2, 29). Feghaly et al. have shown that neutrophil count, C-reactive protein and proinflammatory cytokines were all higher in patients with severe cases of the disease than among asymptomatic controls. (32)

Gastrointestinal endoscopy is rarely used to diagnose CD infections because it can lead to complications. Nevertheless, it may be justified in emergency situations when infection is suspected and stool tests are negative. (27, 33) For children who suffer from chronic diarrhea and impaired nutritional status, it is important to assess intestinal malabsorption and rule out other differential diagnoses as was done in the patients studied here. (34)

Studies of other factors such as fecal calprotectin have also been evaluated. Fecal calprotectin is one of the most stable biomarkers and provides information on intestinal inflammation. It is useful for tracking and predicting relapses of some diseases of inflammatory origins. (35) Its use was initially described in cases of IBD, but it is currently beginning to be used for other pathologies as well. (36) In case 2, the fecal calprotectin level was high which led to follow-up. This patient’s positive test result indicated colitis that was probably due to CD rather than IBD. However, there are insufficient studies of the sensitivity, specificity and usefulness of this study for patients infected CD. Table 5 shows a comparison of key findings in our clinical cases and reported in the literature.

The American Academy of Pediatrics recommends that the first step in treatment be suspension of the antibiotic that led to the episode. (37) This action by itself leads to improvement in up to a third of CD cases. (38) When patients are dehydrated, appear toxic and/or have severe illness, they should be hospitalized. (39) When antibiotic treatment is used for a first-time CD infection, metronidazole is the drug of choice. First recurrences in children with mild symptoms can be managed conservatively without antibiotic treatment. (40) Those with moderate to severe symptoms usually respond to a second course of metronidazole or to vancomycin. (40 ) In case of a second or third recurrence, or when prolonged therapy is decided upon, the use of vancomycin is always recommended. (40) In cases of recurrence after a regimen of vancomycin, fecal microbiota transplants have been shown to produce benefits with cure rates of 91-98%. (41-44) Nevertheless, the optimal route of administration remains to be determined, and there is insufficient evidence regarding the short and long term consequences and/or complications of fecal microbiota transplants. Some studies have reported inflammatory colitis following transplants. (45) In case-studies, patients have relapsed following this treatment. In fact, in Case 1, the patient required oral vancomycin. This article proposes an algorithm for diagnosis and treatment of CD infections (Figure 4).

Although other approaches such as the use of immunoglobulin, probiotics and nitazoxanide rifaximin have been described, more studies are required before they can be recommended. (37, 46) Two recent meta-analyses support the use of probiotics for management of CD infections. A study by Mcfarland found that patients managed with S. boulardii had RR = 0.50 (95% CI: 0.29 - 0.85), those managed with L. casei DN114001 had RR = 0.07 (95% CI: 0.01 to 0.55), a mixture of L. acidophilus and Bifido. bifidum had RR = 0.41 (95% CI: 0.21 to 0.80), and the mixture of L. acidophilus, L. casei and L. rhamnosus had RR = 0.21 (95% CI: 0.11 to 0.40). (47) The meta-analysis by Szajewska et al. showed that the risk of diarrhea decreased among pediatric patients with CD treated with S. boulardii  with an RR =  0.25 (95% CI: 0.08 to 0.73). (48)

Intravenous administration of monoclonal immunoglobulin  could be an option for treating patients whose initial therapy has failed or for seriously ill patients. (27, 33) However, more studies are needed in order to evaluate usefulness and cost effectiveness. (27) Fidaxomicin may be a treatment alternative, especially because it has few negative effects on the Bacteroid count and could be beneficial for intestinal microbiota. (49) A synthetic mixture of intestinal microbiota has also been evaluated as a way to reduce transmission of pathogens. However, more studies are needed. (50, 51). Currently, a vaccine against CD toxins is being experimented with. It may prove to be useful for preventing recurrences. (27)

CONCLUSIONS

CD infections in the pediatric population present challenges for medical staff because of their similarities to infectious diseases and because of the high risk of re-infection. When a patient with a history of administration of antibiotics has bloody stools, an infection should be suspected.

Treatment of first reinfections when symptoms are mild does not require antibiotics. Many of these patients improve with discontinuation of antibiotic therapy. In moderate and severe cases, vancomycin is an option for initial treatment. Vancomycin can also be used to treat a second recurrence, but there is still no consensus about the most appropriate treatment for the third recurrence and for cases of more than three recurrences.

In recent years, the use of probiotics has shown great benefits for treating reinfections, but the available evidence is still contradictory.

It is important to avoid the indiscriminate use of antibiotics for pediatric patients especially because such a high percentage of respiratory and gastrointestinal infections have viral etiologies. We believe that the use of probiotics could have favorable implications for management of patients with recurrent CD infections. The benefits are greatest in mild cases in which antibiotics are not required, but probiotics can be used as supplement therapy for moderate and severe cases.

Financial support

This study was conducted with the financial support of the Gastroenterology, Hepatology and Nutrition Unit of Gastronutriped.

Conflicts of interest

The authors of this research state that they have no conflicts of interest related to this study’s objective.

REFERENCES