Clostridioides difficile, a Gram-positive, spore-forming, anaerobic bacterium, is the foremost cause of healthcare-associated diarrhea worldwide. While it remains the primary contributor to healthcare-related diarrhea among adults in the United States, it is also increasingly acknowledged as a significant pathogen impacting children.1

The carriage rate of C. difficile is notably high among newborns, implying a commensal status for this bacterium within this specific population segment.1,2 For reasons still unknown, these babies colonized with toxigenic C. difficile rarely develop clinical symptoms, unlike older children and adults who are susceptible to severe diarrhea and colitis. Colonization rates tend to decrease with increasing age, with children older than two years exhibiting rates similar to those observed in adults.2 Nevertheless, toxigenic strains may also colonize asymptomatic infants during hospitalization and within the community.2

In recent decades, a significant increase in the incidence of CDI has been reported, even in populations previously considered low-risk, such as children. Despite the rising number of CDI cases in children, along with an increase in morbidity related to frequent recurrences,3 and evidence that infection in hospitalized children has a similar economic impact to that observed in adults,4 research on pediatric CDI remains considerably limited when compared to study in adults. Moreover, there has been a noteworthy rise in the prevalence rates of both Healthcare-Associated (HA-CDI) and Community-Associated (CA-CDI) CDI among pediatric populations.5–7

Given the limited information and diagnosis of C. difficile infection in Brazilian hospitals, especially in the Northeast region, and the global spread of this infection, our study aimed to conduct a local epidemiological investigation. Specifically, this study sought to identify comorbidities and risk factors associated with CDI in pediatric patients and investigate molecular epidemiology and antimicrobial resistance.

A total of 56 specimens were evaluated. Diarrheic stool samples were collected from 49 children; five patients contributed with more than one sample, collected at different points during their hospitalization. Detailed information regarding patients' demographic, socio-economic, and health characteristics is provided in Table 1.

In this study, the frequency of C. difficile positivity (toxins and/or isolation) was 30.4 % (17/56). Initially, C. difficile toxins were detected in 13 of 17 samples through toxin A/B rapid immunoassay (76.4 %), among which two isolates were obtained. Four isolates were obtained from the toxin-negative samples, of which two were subsequently tested as toxigenic through PCR. Consequently, the percentage of C. difficile isolated in culture was 35 % (6/17). None of the children presented more than one positive sample (Fig. 1).

Fig. 1.

Flow Chart indicating the recruitment and isolation of Clostridioides difficile in children with diarrhea at a tertiary hospital in Fortaleza, Brazil, between 2015 and 2017. Toxigenic ribotypes are highlighted in orange, while non-toxigenic ribotypes are highlighted in green.

(0.4MB).

According to the Laboratory of Biology of Anaerobes at UFRJ, the isolates of C. difficile belonged to six different ribotypes: 046, 106, 002, 143, 133, and 012 (Table 2).

Toxin and tpi gene fragments (definitive identification) were performed on all the isolated strains. PCR targeting tpi, tcdA, and tcdB genes were found in Medical Record Number (MRN)-1 (RT046), MRN-15 (RT106), MRN-54 (RT002), and MRN-58 (RT012) samples, but no deletion of the tcdC gene was found in these samples. The samples MRN-55 (RT143) and MRN-56 (RT133) exhibited non-toxigenic characteristics, and therefore genotyping by PFGE was not performed. The binding gene of the binary toxin was not detected in any of the samples (Table 2).

The analysis of the PFGE profile by digestion with SmaI revealed new pulsotypes in three isolates (RT046, RT002, and RT012). In addition, RT106 was identified as NAP11-0499. The analysis of the identified number was standardized according to the database of the University of Costa Rica and NML-Canada (Table 2).

In terms of antimicrobial susceptibility, all strains were susceptible to metronidazole and vancomycin, commonly used for treating CDI. The children with CDI, including those six patients whose stool samples tested positive for C. difficile, were treated with metronidazole and evolved with diarrhea remission during the follow-up period. All strains were susceptible to the tested antimicrobials except two that showed levofloxacin resistance (MRN-1 and MRN-15), as shown in Table 2.

Regarding clinical manifestations, shown in Table 1, most affected children experienced acute diarrhea, typically lasting up to 14 days. However, it's worth noting that three patients experienced persistent diarrhea lasting between 14 and 30 days, while two children had chronic diarrhea lasting more than 30 days. Alongside diarrhea, additional symptoms such as vomiting, fever, abdominal pain, and abdominal distension were noted. However, only nausea exhibited a significant association (p = 0.035) with patients who tested positive for CDI in comparison to those with a negative C. difficile infection test (Table 1).

When evaluating risk factors for CDI, which include recent hospitalization (within the last 30 days), antibiotic or proton pump inhibitor usage, and the presence of comorbidities or chronic diseases, it was observed that 88.2 % (15/17) of patients with positive samples had underlying conditions. Importantly, each of these 15 children presented one or more risk factors. Only two children with positivity to C. difficile, constituting 11.7 % (2/17), exhibited no risk factors (Table 3). Hospitalization within the last 30 days was noted in 41.1 % (7/17) of patients with positivity to C. difficile, with no significant difference compared to those with negative C. difficile test (p = 0.206). Antibiotics had been previously prescribed to 47 % (8/17) of children with CDI, and proton pump inhibitors had also been once used by 47 % (8/17) of these children (Table 3).

While comorbidities emerged as the most prevalent risk factor, affecting 76.5 % (13/17) of CDI patients, no statistical differences were found between patients with and without the infection (p = 0.758). These comorbidities included inflammatory bowel disease (35.3 %; 6/17), chronic kidney disease (23.5 %; 4/17), neurologic disease (11.8 %; 2/17), and hepatopathies (11.8 %; 1/17), as shown in Table 3. However, it is crucial to emphasize that all six patients infected with C. difficile, from whom fecal isolates were obtained, had previously been diagnosed with a chronic disease.

Considering the changing patterns of C. difficile infection in childhood,7,13–19 the present study investigated the molecular epidemiology and antimicrobial resistance in C. difficile strains isolated from children and adolescents in a tertiary referral pediatric hospital in Fortaleza, Brazil, and the associated risk factors. As far as we know, no previous research has examined C. difficile infection among children in northeastern Brazil. Studies on C. difficile-associated diarrhea in Brazil have been limited. In the last two decades, only two pediatric research groups reported the presence of C. difficile infection in Brazilian children: a study carried out in Rio de Janeiro with hospitalized and non-hospitalized patients20 and a study conducted in São Paulo between June 2000 and July 2001.21 In these studies, C. difficile was isolated from 6.7 % and 5.5 % of diarrheal stool samples, respectively. In this current study, the frequency of C. difficile positivity stood notably much higher at 30.4 % (17/56), with the rate of toxigenic C. difficile reaching 88.2 %, detected by toxin A/B rapid immunoassay and PCR, accounting for 15 out of 17 cases.

Further epidemiological studies are needed to confirm whether pediatric CDI has increased in Brazil, but our data, compared with studies from Brazilian hospitals in the 2000s, point in that direction. This hypothesis aligns with studies by other research groups in different countries.22,23

This difference between the Brazilian studies conducted in the 2000s and our study can be attributed to a combination of factors. During this time gap, changes in health conditions, hygiene practices, antibiotic use, emergence of new strains, and other factors that affect the prevalence of C. difficile and other intestinal pathogens may have occurred. Furthermore, in Brazil's case, it's crucial to consider the vast continental proportions of the country, which are geopolitically divided into five regions. For instance, Fortaleza is the capital of the state of Ceará, situated in the northeastern region, while Rio de Janeiro and São Paulo are in southeastern Brazil's wealthiest region. Tailored epidemiological studies in each region are essential to gather relevant, region-specific data on Clostridioides difficile-associated diarrhea. In this context, our study stands out as a significant contribution, as it provides insights into the local epidemiological landscape of CDI among children in northeastern Brazil for the first time.

We should also consider the age group differences among these Brazilian studies as a variable that may affect the results: Rio de Janeiro (3 months ‒ 7 years), São Paulo (0‒5 years), and Fortaleza (18 months ‒ 18 years). In this study, we excluded children under 18 months old due to elevated asymptomatic carriage rates among infants and young children. This significantly complicates diagnosing C. difficile-associated disease in this group, leading to challenges in defining infection and determining the necessity of treatment.23 Two exceptions were made for infants under 18 months. One had Hirschsprung's disease, a potential risk factor for severe CDI. The other shared a ward with two children already diagnosed with CDI, heightening the risk of infection transmission due to the persistent presence of C. difficile spores in healthcare environments. Indeed, both infants were confirmed to be infected with non-toxigenic C. difficile strains (ribotypes 106 and 002).

We also examined commonly reported CDI risk factors, which include hospitalization, previous antibiotics exposure, previous use of Proton Pump Inhibitors (PPIs), and comorbidities. Overall, 88.2 % (15/17) of the children infected with C. difficile were identified as having at least one of these risk factors. Comorbidity was the most prevalent risk factor 76.5 % (13/17), followed by previous use of antibiotics at 47 % (8/17) and PPIs at PIs 47 % (8/17). However, it's important to note that none of these risk factors were found to predispose to CDI in the present study. Although the above-listed factors have been commonly described for CDI in adults, it has been argued that not all CDI risk factors act equally in pediatric and adult patients.24,25

A case-control study to identify potential risk factors for CDI at a pediatric quaternary care hospital in the USA reported that recent antibiotic exposure and certain comorbid conditions (solid organ transplant, presence of a gastrostomy, or jejunostomy tube) were associated with CDI.26 One of the toxigenic strains identified in the present study (ribotype 046 with a new pulsotype 1174) came from a child with a history of hospitalization, antibiotic administration, hepatic transplantation, and immunosuppressive medication therapy. In addition to ribotype 046, we identified five more C. difficile ribotypes (106, 002, 143, 133, 012). Among these, three new pulsotypes were detected (1174, NML-1234, NML-1235) and NAP 11–0499.

The heterogeneity found in the PCR-ribotyping of C. difficile strains performed in the present study does not suggest an outbreak. In a study conducted in eight US states to examine the association between strain type and disease outcomes in adults, NAP11 was previously identified as the third most common strain type (9.1 %).26 A study conducted in Chicago, USA, with pediatric patients diagnosed with CDI reported the evolution and transmission of a multi-drug resistant DH/NAP11/106 C. difficile clone.27 In accordance, NAP11/RT106, isolated in our study, was resistant to levofloxacin. Fluoroquinolone resistance in a BI/NAP1/027 strain has been implicated as a significant contributor to its pathogenesis and global spread.28,29 All C. difficile strains isolated in the present study were sensitive to metronidazole and vancomycin, the most used antibiotics to treat CDI in Brazil. Therefore, all children responded well to metronidazole treatment.

Our study obtained four isolates from samples that initially tested negative for toxins. Based on a careful evaluation of the clinical features and associated risk factors, we decided to treat these four patients with metronidazole, as there were strong indications that they might be infected with C. difficile. Subsequently, two toxigenic strains were isolated from this group, emphasizing the need to revise existing clinical protocols for diagnosing and treating CDI, especially in the pediatric population.

Our study provides insights into the current epidemiological status of Clostridioides Difficile-associated Diarrhea (CDI) among children in northeastern Brazil. It reveals a substantial CDI frequency of 30.4 %, with toxigenic strains detected in 76.4 % of cases, highlighting a higher prevalence compared to earlier Brazilian studies. While our study provides valuable insights, it is essential to acknowledge its limitations, notably the small sample size and restriction to a single pediatric hospital. These constraints underline the necessity for more extensive, multicenter studies across diverse healthcare settings in Brazil.

This study received approval from the Research Ethics Committee (REC) of Albert Sabin Children's Hospital (protocol number 861363). Written informed consent was obtained from all patients. The REC's primary responsibility is safeguarding the dignity, rights, safety, and well-being of individuals participating in biomedical research. It also upholds public accountability by publishing its decisions. All institutions involved in this research provided their consent to participate.

Not applicable.

The data generated and analyzed during the current study are available from the corresponding author upon reasonable request.

All authors contributed to the study conception, design, and experiment. GACB and RFCL contributed equally to this work. HBRB, CLC, EOF, and DMP prepared material and collected data. HBRB, GACB, RFCL, and CQG analyzed the data; HBRB, CLC, RFCL, EOF, and GACB wrote the first draft of the manuscript. CLC, RFCL, and GACB did the final review. All authors read and approved the final manuscript.

This work was supported by PRONEX/FUNCAP/CNPq of Brazil through grant PR2-0101-00060.01.00/15.