1. Introduction
Despite advances in health care, infectious diseases remain a major cause of morbidity and mortality [1] . Control of infectious diseases requires multiple approaches that rely on improved methods of diagnosis and treatment [2] . Gastroenteritis is one of the most infectious diseases with high morbidity, mortality, and serious public health significance particularly in low and middle-income countries [3] [4] . Gastroenteritis can be acute or chronic, caused by viral, bacterial, and more rarely parasitic pathogens [5] . Bacterial pathogens are responsible for 20% - 40% of gastroenteritis with diarrhoeal episodes [6] [7] , and the increase of their antimicrobial resistance has become another health challenge for therapy, leading to treatment failures [8] [9] [10] .
Generally, one of the management dilemmas in the evaluation of patients with gastroenteritis is deciding when to look for etiological agents and when to initiate antimicrobial therapy [8] . Locally, parasites are researched, and testing for pathogenic bacteria is usually limited to Salmonella sp and Shigella sp. Escherichia coli (E. coli) is systematically tested and considered only in children’s cases, with non-differentiation of pathogenic or non-pathogenic strains. As commonly known, E. coli is a bacterium found in the commensal flora of the gut of humans and warm-blooded animals. However, although most E. coli are harmless, some are pathogenic and these species can cause significant gastrointestinal diseases [11] [12] . Indeed, worldwide, the epidemiology is changing with an increasing burden of gastroenteritis associated with diarrheic Escherichia coli (DEC) [13] . Pathogenicity is acquired through the capture of genetic elements containing genes coding for virulence factors necessary to cause infection [14] . Pathogenic DEC can be categorized as enteropathogenic E. coli (EPEC), enterotoxigenic E. coli (ETEC), enteroinvasive E. coli (EIEC), enteroaggregative E. coli (EAEC), and Shiga toxin-producing E. coli (STEC) [12] . In addition, Escherichia coli can harbor resistance genes using the same capture system [15] . Accordingly, any pathogenic or antimicrobial-resistant Escherichia coli can be harmful to its host [16] .
Previous studies in Cameroon showed the low implication of Salmonella sp, Shigella sp, and Yersinia sp in cases of gastroenteritis [17] [18] [19] . However, there are very little data in the literature regarding the involvement of DEC in cases of gastroenteritis due to the limitation of diagnosis methods. Indeed, in the laboratory, phenotypic detection methods by E. coli culture use conventional selective media EMB or MacConkey, where the characteristic colonies will be green colonies with a metallic shine and pink lactose fermenting colonies respectively. Although these media are selective for E. coli, they showed limits to screening the pathogenic potential species in gastroenteritis cases. Recent studies have revealed that the new chromogenic media, such as CHROMagar™STEC, is a good screening media for DEC detection [20] [21] [22] .
Our study was focused on the Littoral region, in the city of Douala, one of the most populated and cosmopolite in Cameroon, and added to that, the absence of existing data concerning the frequency of DEC in cases of gastroenteritis. This study investigated the frequency of common causes of enteric infection with an emphasis on pathogenic DEC in cases of gastroenteritis, and their antibiotic resistance profiles.
2. Methods
2.1. Study Design
Between October 2021 and June 2022, we carried out a cross-sectional study targeting human patients in Douala. With more than 3 million inhabitants, Douala is the economic city of Cameroon; located in Central Africa. This city concentrates almost 20% of the urban population of the country. Laquintinie Hospital has been chosen as a sampling site because it constitutes the regional hospital of the city, where sample stool could be handled in the context of the COVID-19 pandemic. All patients complaining of gastroenteritis were referred there for sample analysis.
2.2. Ethical Considerations
According to the guidelines for human experimental models in clinical research, as stated by the Cameroon Ministry of Public Health, ethical approval was obtained from the Institutional Ethics Committee for Research Human Health of the University of Douala (Reg 2880 IEC-UD/07/2021/T). This was followed by the administrative agreement of the Regional Delegate of the Ministry of Public Health in the Littoral Region and the research authorization of the administrative authority of Laquintinie Hospital.
2.3. Sample Collection
All patients with gastroenteritis complaints consulted by the physician at the hospital which required analysis of stool samples were approached and included consecutively after obtaining their consent. Stools samples were collected using sterile universal containers and then labelled with the gender, age of the patient, and postcode. Samples were transported to the laboratory within 2 hours of collection and analyzed according to the recommendations of the reference in medical microbiology (REMIC) [23] .
2.4. Parasite Identification
Stools samples obtained were observed under a microscope (Olympus XSZ-107BN; x40) in fresh state immediately arrive in the laboratory. Parasites were identified based on their morphological characteristics (size, motility, shape) and life stage.
2.5. Bacterial Pathogens Isolation and Identification
We screened different bacterial pathogens: Campylobacter species, diarrheic Escherichia coli, Shigella species, Salmonella species, and Yersinia species using selective media. Before use, each culture media in this study was subjected to internal quality control using different microorganisms according to the manufacturer’s instructions (Table 1).
Screening of Campylobacter sp
REMIC protocol was used with minor modifications [23] . We used combination methods of filtration and culture with media Campylobacter Selective Agar (Merck, UK) to optimize the isolation of Campylobacter species. Briefly, a Pasteur pipette was used to place eight to ten drops of the sample diluted onto a cellulose triacetate membrane with 0.45 µm pores placed on the surface of the selective agar plate. The membrane was left on the agar surface until all the fluid had passed through; this took approximately 20 to 30 minutes. Plates were incubated microaerobically using GENbag anaer (Biomérieux, France), at 37˚C for two days. Suspect colonies (grey, translucent colonies, sometimes with a silver sheen) were identified to the genus level by a positive oxidase reaction and a typical Gram stain appearance (slender, curved, “seagull wing-shaped”, Gram-negative rods).
Screening of DEC
Samples were enriched in Trypticase Soya Broth (Oxoid, UK) at 37˚C for 24 h to optimize the recovery of pathogenic E. coli [20] . Subsequently, a loop of 10 µl of the enrichment was plated on CHROMagar™STEC and incubated for 24 h at 37˚C. Next, one to three mauve colonies were purified in PCA and incubated overnight at 37˚C for 18 hours. The isolates were confirmed as E. coli using morphological characteristics such as motility, Gram staining, and biochemical characteristics from media such as Kliger’s iron agar, Simmon’s citrate agar, and urea-indole media according to the manufacturer’s instructions.
Screening of Salmonella sp and Shigella sp
Traditional detection methods involve enrichment in a selective liquid culture medium followed by isolation using selective and differential agar [23] . Briefly, stool samples were inoculated into 9 ml of Selenite broth (Biolab, Hungary) and incubated at 37˚C for 24 hours. Subsequently, 10 µL of the culture was plated onto Salmonella-Shigella agar and incubated for 24 hours at 37˚C. Based on the morphology and appearance of the colonies, presumptive Salmonella colonies (colorless with a black center) and Shigella colonies (colorless), a subculture was made by plating onto Plate Count agar (PCA) (Oxoid, UK) and incubated at
Table 1. Internal control quality of media used in the study.
a. species were identified using biochemical characters in API20E. b. species identification was performed using API20E and serotyping.
37˚C for 24 hours. Colony species were biochemically identified using API 20E (Biomérieux, France) according to the manufacturer's instructions.
Screening of Yersinia sp
Stool samples were diluted at 10−1 in distilled sterile water, and subsequently, a loop of 10 µL of the suspension was plated onto MacConkey agar and incubated for 24 - 48 hours at 37˚C. Lactose-negative small colonies (1 - 2 mm diameter) colorless or pale pink colonies, and flat were selected [24] . A subculture of these colonies was made by plating onto PCA and incubating at 37˚C for 24 hours. Finally, colony species were biochemically identified using API 20E (Biomérieux, France) according to the manufacturer’s instructions.
2.6. In-Vitro Antimicrobial Susceptibility Testing of Bacterial Pathogens
All isolated strains were subjected to susceptibility testing and evaluated to commonly used antibiotics using the Kirby–Bauer disc diffusion method according to the European Committee of Antimicrobial Susceptibility Testing criteria (EUCAST) [25] . Table 2 presents the different antibiotics used with their concentrations and their breakdown used to categorize results as Sensible or Resistant. With a bacterial cell culture of 24 h on PCA, a loop of each isolate was emulsified in a sterile physiological water solution in a test tube and the density was measured with a McFarland densitometer to obtain 0.5 McFarland
Table 2. Antibiotics tested (and the respective antibiotic classes) and interpretation of zone of inhibition (mm), from EUCAST 2021.
∅ = diameter of inhibition zone in mm.
standards. Using a sterile cotton swab, the suspension was emulsified onto a Mueller Hinton agar plate and incubated at 37˚C for 18 h.
The zone of inhibition was measured, and the results were interpreted. Escherichia coli ATCC 25922 was used as quality control. Isolates observed resistant to at least three classes of antimicrobials were considered Multidrug Resistant (MDR). Pansusceptible was defined as isolates susceptible to all antibiotics tested.
2.7. Data Analysis
All data were recorded into an Excel spreadsheet and used the sheet for descriptive statistical analysis (frequencies, proportions, and Chi-square test) with SPSS 23.0. A McNemar of Chi-square test was performed to compare the frequency of enteric pathogenic species found in positive samples. We considered an association statistically significant if P-values < 0.05.
3. Results
3.1. Study Population
We collected samples from 296 patients included in the study with ages ranging between 5 months to 90 years old (Median = 35.5; SD = 20.8). The sex ratio was 0.73 with more females (57.8%) than males.
3.2. Parasitic Pathogens
Parasites were identified in 1.7% (n = 5/296) patients as the etiological agent of gastroenteritis. Parasites were identified in monoparasitism and included Entamoeba histolytica (60%; n = 3/5), Trichomonas intestinalis (20%; n = 1/5), and Giardia intestinalis (20%; n = 1/5).
3.3. Bacterial Pathogens
Out of the 296 samples, 27% (n = 80/296) of the samples were positive for bacterial pathogens. The pathogens identified in the samples were DEC (90%; n = 72/80) followed by, Salmonella sp (6.3%; n = 5/80) and Shigella sp (3.7%; n = 3/80) (Figure 1). No Yersinia sp and Campylobacter sp were found in stool samples in this study. Also, no co-infections were identified.
The diarrheic E. coli were frequently isolated in population studies with an age range between 5 to 50 years old (Table 3) and was not detected in patients with an age range between 0 to 5 years old. There was no statistical difference between females and males across bacterial pathogens.
3.4. Antibiotic Resistance Profiles of Bacterial Pathogens Isolates
Strains of diarrheic E. coli were most resistant to AMO (93.1%; n = 67/72), followed by CIP (75%, n = 54/72), SXT (73.6%, n = 53/72), DOX (68.1%, n = 49/72), AMC (52.8%, n = 38/72), FOX (47.2%, n = 34/72), CTX (45.8%, n = 33/72), AZM (38.9%, n = 28/72), AMC (38.9%, n = 28/72), AKN (34.7%, n = 24/72) and CHL (34.7%, n = 24/72) (Figure 2). In contrast, strains showed low resistance to NIT and IMI, with respective rates of 6.9% (n = 5/72) and 2.8% (n = 2/72).
Fifty-six distinctive antimicrobial resistance profiles (ARPs) were recorded with resistance levels ranging from one to ten antibiotics from the twelve antibiotics tested. The most common resistance levels were recorded in five classes of antibiotics, and the common phenotype of resistance was AMO-AMC-CIP-SXT-DOX. Among these ARPs, we found 88.9% MDR strains (n = 64/72) and no pan-susceptible isolates as presented in Table 4.
Salmonella sp strains were resistant to only three drugs: AMO (100%; n = 5/5), SXT (80%, n = 4/5), and DOX (60%, n = 3/5). Shigella sp isolates were resistant to SXT (66.7%, n = 2/3), AMO (33.3%, n = 1/3), and CIP (33.3%, n = 1/3). MDR strains were found in Salmonella sp (80%, n = 4/5) (Table 5).
Globally, resistance to AMO, CIP, and SXT was common in Salmonella sp, Shigella sp, and, diarrheic Escherichia coli (Figure 2).
4. Discussion
The main objective of this work was to investigate the contribution of pathogenic Escherichia coli and their antibiotic resistance profiles in gastroenteritis. While the research of pathogens is usually limited to parasites and two bacterial pathogens namely Salmonella sp and Shigella sp in the context of Cameroon, the
Figure 1. Distribution of enteropathogens identified in samples analyzed. Asterisk denotes a significant difference between the proportion of this pathogen across bacterial pathogens identified, p-value < 0.001 was obtained using McNemar of Chi-square test.
Figure 2. Resistance profiles of different antibiotics tested on enteric bacteria pathogen isolates. Abbreviations: AMO, amoxicillin; AMC, amoxicillin + clavulanic acid; FOX, cefoxitin; CTX, cefotaxime; IPM, imipenem; CIP, ciprofloxacin; AKN, amikacin; SXT, trimethoprim-sulfamethoxazole; NIT, nitrofurantoin; AZM, azitromycin; CHL, chloramphenicol; DOX, doxycycline; MDR, Multidrug resistant.
Table 3. Distribution of diarrheic E. coli, Salmonella sp, and Shigella sp per sex and age of the population study.
a. represents the number of samples analyzed. b. represents the number of positive samples population. c. P-value was calculated using the Pearson Chi-square test and the significance level was considered with p < 0.05.
Table 4. Antibiotic resistance profiles (ARPs) of diarrheic E. coli isolates on CHRO Magar™ STEC.
Table 5. Antibiotic resistance profiles of Salmonella sp and Shigella sp isolates from cases of gastroenteritis.
current worldwide trend shows an important involvement of diarrheic E. coli in cases of diarrhea in countries with more advanced methods of pathogen identification [13] .
Although our results are online with the conclusion of Riddle et al. [5] that parasites are rarely found in case of diarrhea, our low values found are not similar to the results of Belay et al. [26] which found a high prevalence of intestinal parasites in human samples. In this study, parasite pathogens were found at a low frequency, representing 1.7% of cases of gastroenteritis. Regarding the situation in Cameroon, the frequency of intestinal parasites in this study are contrary to previous studies in other regions, which found 8.4%, 15.4%, and 21.9% respectively [27] [28] [29] . Similarly, regarding trends in other countries of the world, our frequency values obtained are much lower concerning the carriage of intestinal parasites [30] [31] [32] [33] . Various sizes of samples analyzed in these previous studies ranging from at least 500 to 50,000 cases, could be strongly associated with this difference in our value obtained.
If this study reports the first results of a screening of parasites and bacterial pathogens in cases of gastroenteritis in Douala, the frequency of detection of parasites (1.7%) was very low than the detection of bacterial pathogens (27%) among cases. Our results are online with the conclusion of Moro et al. [3] which showed in a minireview presenting the causes of infectious gastroenteritis, that parasites are less commonly implicated in gastroenteritis than bacterial pathogens. However, Belay et al. [26] in Ethiopia, found a high prevalence of intestinal parasites (20.7%) than bacterial pathogens (6.6%) in human samples in Ethiopia. The low rate of detection of parasites as observed in the present study might be due to the increasing awareness of the people about personal and environmental hygiene and, as well Douala is an urban zone. It should be noted that all parasites found were protozoans with Entamoeba histolytica mainly detected in positive cases. This may be justified by the fact that this parasite remains one of the top three parasitic implicates in cases of gastroenteritis and causes of mortality worldwide [34] .
Among the 27% positives samples to bacterial pathogens, Salmonella sp and Shigella sp were detected at very low frequencies (6.3% and 1.7% respectively) than diarrheic Escherichia coli. These findings are in line with the previous study which found relatively low frequency in the city of Douala, Littoral region (10.3% of Salmonella sp and 3.99 of Shigella sp) [18] , and in the city of Buea, North West Region of Cameroon (8.7% of Salmonella sp) [17] . Salmonella sp and Shigella sp are bacterial pathogens that represent major public health problems in terms of mortality and morbidity for both developed and undeveloped countries. The possible explanation for the low prevalence of these pathogens could be due to the fact that the health center has initiated treatment of patient referral to this hospital, or associated with an increase in self-medication which has been described previously [35] .
Diarrheic E. coli was isolated with a significantly higher proportion than other enteric pathogens. If here we reported a first analysis of common enteric pathogens in other to show the real contribution of each pathogen in cases of gastroenteritis in Douala, especially pathogenic Escherichia coli, a recent study highlighted the important involvement of diarrheic E. coli associated with gastroenteritis in the city of Mbouda, West region in Cameroon (19.7% of cases) [36] . If Escherichia coli is a commensal bacterium representing 80% of digestive flora, it is easily found in high proportions using classical media as in some studies, which unfortunately did not provide specific data about the proportion of pathogenic strains. In addition to the framework of this current study, no co-infections were found in the analysis of samples. So, our findings suggest that diarrheic strains of E. coli could be one of the main causes of consultation for gastroenteritis in hospitals, and should be taken into account when suspecting enteric pathogens.
In the study population, patients’ age ranging from 5 to 50 years old showed a high frequency of isolation of diarrheic E. coli. Our result is similar to that obtained for patients of ages ranging between 20 to 50 years old (20 to 30 years old, 26.89%; 30 to 40 years old, 18.49%; 40 to 50 years old, 30.25%) by Marbou et al. [36] . However, in this study, no isolates were obtained from children aged between 0 to 5 years old. This is contrary to the results of a previous study in the Littoral region, where diarrheic E. coli were identified in children at a low rate [37] [38] . This contrast could be explained by the fact that identification methods were different from this study. In addition, worldwide the major cause of childhood diarrhea is Rotavirus as an infectious agent. Our result highlighted that the use of antibiotics in children aged between 0 - 5 years old should be better controlled, as children routinely receive antibiotics when Escherichia coli has been isolated from their stool samples in cases of gastroenteritis in Douala.
Regarding the antibiotic resistances profiles, if good activity was observed among isolates of diarrheic E. coli against imipenem and nitrofurans in this study, high levels of resistance were observed against amoxicillin (91%), ciprofloxacin (75%) and trimetoprim + sulfamethazole (73.5%). Analysis of the correlation between diarrheic E. coli antimicrobial resistance and virulence profiles can help physicians avoid treatment failure. Indeed, the choice of antimicrobial therapies depends on the type of diarrheic E. coli as well as its virulence and resistance profiles [39] . Similar high rates of resistance have been described previously on diarrheic E. coli in the case of Mbouda, West region which was found with amoxicillin and trimetoprim + sulfamethazole [36] . In a recent review of human health in Cameroon, these same resistances have been described in E. coli from extra digestive infections [40] .
Of the diarrheic E. coli isolates, 47.2% showed resistance to azithromycin. While studies in Cameroon on E. coli have not described resistance to this antibiotic in humans, our results are contrary to a recent study in Congo, which found a low level of resistance in E. coli strains of fecal origin [41] . It would be important to note that this lack of data regarding azithromycin resistance in our context, could be related to the reference used in the laboratory which recommends systematically testing azithromycin in particular for Salmonella sp and Shigella sp [25] . However, azithromycin is a promising alternative with excellent activity against the most common enteric pathogens including diarrheic E. coli [42] [43] . Our results could be related to the use of azithromycin for treatment during the COVID-19 pandemic, where in our area self-medication was a common phenomenon and has already been described. The expression of this resistance to antibiotics represents a serious problem worldwide. Indeed, E. coli may harbor resistance genes that may be transferred to pathogenic or opportunistic bacteria. For these reasons, E. coli has been classified by the World Health Organization as a priority pathogen due to its widespread resistance to antibiotics [44] .
Among Salmonella sp, Shigella sp, and diarrheic Escherichia coli isolates, resistance to amoxicillin, ciprofloxacin, and trimetoprim + sulfamethazole was commonly observed. These three medicines are on the national list of essential medicines in Cameroon [45] . While high resistance to amoxicillin is commonly described, the WHO has recently reported high resistance levels for Escherichia coli and Salmonella sp to ciprofloxacin [46] . Ciprofloxacin is an antibiotic substance usually prescribed for the treatment of salmonellosis. Trimetoprim + sulfamethazole (known as Metronidazole) is an antibiotic and antiparasitic substance widely used in the treatment of several infections caused by bacteria and some types of protozoa. The resistance especially in this case of them could be associated with the poor quality of these two drugs in the market of Cameroon [47] , which promotes antibiotic resistance, and finally can lead to the reduction or absence of effectiveness of first-line therapies [48] .
Based on antibiotic resistance profiles, diarrheic E. coli showed 56 distinctive resistance profiles with resistance levels ranging from one to ten antibiotic classes, which allowed us to find that 88.9% (n = 64/72) of isolates were MDR. Similar results have been described in diarrheic E. coli in Egypt, which found 90% MDR among isolates [39] . These results could be linked to a carriage of genetic elements such as integrons, genetic structures that will allow the bacteria to capture many antibiotic resistance genes.
This study was limited by the lack of identification of viral enteropathogens among stool samples, which could allow us to give complete profiles of etiological agents responsible for gastroenteritis cases. In further studies, a molecular analysis of diarrheic Escherichia coli obtained could be necessary to identify the different pathotypes and virulence genes among the isolates obtained, and genes of resistance associated with the resistance observed.
5. Conclusion
These results emphasize the need to consider diarrheic Escherichia coli as the main cause of consultation in cases of gastroenteritis in our hospitals in Douala and reiterate the urgent need to rationalize antibiotic use.
Acknowledgements
This research did not receive external funding. We thank G.T. Tchoupe Alix for the English revision and Medi Sike Christiane for administrative assistance at Laquintinie Hospital.