Extended Spectrum β-Lactamase Producing Lactose Fermenting Bacteria Colonizing Children with Human Immunodeficiency Virus, Sickle Cell Disease and Diabetes Mellitus in Mwanza City, Tanzania: A Cross-Sectional Study

Rectal carriage of extended spectrum β-lactamase-lactose fermenters (ESBL-LF) is the major risk factor for the development of subsequent endogenous infections. This study determined the patterns and factors associated with the rectal carriage of ESBL-LF among children with Human Immunodeficiency Virus (HIV), Diabetes Mellitus (DM), and Sickle Cell Disease (SCD) attending clinics at different health care facilities in the city of Mwanza, Tanzania. A cross-sectional study was conducted among children living with HIV (n = 236), DM (n = 42) and SCD (n = 126) between July and September 2021. Socio-demographic and clinical data were collected using a structured questionnaire. Rectal swabs/stool samples were collected and processed to detect the rectal carriage of ESBL-LF following laboratory standard operating procedures (SOPs). Descriptive statistical analysis was conducted using STATA 13.0. The overall prevalence of ESBL-LF carriage was 94/404 (23.3%). Significantly higher resistance was observed to ampicillin, trimethoprim-sulfamethoxazole, and tetracycline among Enterobacteriaceae isolated from HIV infected children than in non-HIV infected children (p < 0.05). The commonest ESBL allele 45/62 (72.6%) detected was blaCTX-M. Generally, a parent’s low education level was found to be associated with ESBL-LF colonization among children living with HIV; (OR 4.60 [95%CI] [1.04–20], p = 0.044). A higher proportion of ESBL-LF from DM 10/10 (100%) carried ESBL genes than ESBL-LF from HIV 37/56 (66.1%) and SCD 15/28 (53.6%), p = 0.02. There is a need to collect more data regarding trimethoprim-sulfamethoxazole (SXT) prophylaxis and antibiotic resistance to guide the decision of providing SXT prophylaxis in HIV-infected children especially at this time, when testing and treatment is carried out.


Introduction
Extended Spectrum Beta-Lactamase (ESBL) are enzymes produced by Gram-negative bacteria mainly among members of the family Enterobacteriaceae. Infections associated with ESBLs-producing bacteria are linked to increased costs, morbidity, and mortality [1,2]. Rectal carriage of ESBL-producing bacteria is a potential endogenous source of infections and an independent risk of developing subsequent infections due to ESBL-producing bacteria [3][4][5]. Furthermore, the colonization of ESBL-producing bacteria was found to be high in individuals with comorbidities such as Human Immunodeficiency Virus (HIV) [6], diabetes mellitus (DM), organ transplant, and sickle cell disease (SCD) due to the fact that these patients are often in long-term antibiotic prophylaxis [7,8].
The burden of ESBL colonization varies in different populations and geographical locations due to the presence of different drivers of the emergence and spreading of ESBL-LF [9]. Data shows that the burden of ESBL-producing bacteria intestinal carriage is low in high-income countries (HIC) compared to low-and middle-income countries (LMIC) [10][11][12][13][14][15][16][17]. In Tanzania, the prevalence of ESBL rectal colonization has been found to range from 16.5% among adults and children in Mwanza to 54.6% among neonates at Bugando medical Centre [18][19][20][21][22]. In addition, a study in Dar es Salaam reported the ESBL-producing bacteria rectal colonization among HIV children to be 89.7% compared to 16.9% among non-HIV infected children [15]. ESBL families CTX-M, TEM, and SHV are among the commonest enzymes conferring resistance towards β-lactams at our setting [23].
Given these considerations and the lack of information globally, this study determined the patterns and associated factors of ESBL-LF colonization in children with SCD, HIV, DM attending clinics at Bugando Medical Centre (BMC), Sekou Touré regional referral hospital (SRRH), and Baylor Children Hospital (BCH) in Mwanza, Tanzania.

Study Design, Duration, Population and Setting
This hospital-based cross-sectional study was conducted from July to September 2021. The study included children living with HIV infection, diabetes mellitus (DM), and sickle cell disease (SCD), aged from 0-13 years and attending the pediatric outpatient clinic at Bugando Medical Centre (BMC), Sekou Toure Regional Referral Hospital (SRRH) and Baylor Children hospital (BCH). BMC is a zonal consultant and teaching hospital serving the Lake zone regions with a population of about 16 million. BCH is a zonal hospital providing treatment and care for children living with HIV infection and tuberculosis (TB), and SRRH is serving a population of about 2.773 million from Mwanza region.

Data and Sample Collections
Data were collected using a questionnaire customized in Epicollect5 (https://five. epicollect.net/, accessed on 30 September 2021) on smartphone by trained research assistants. Social demographic (e.g., age, gender and place of residence), clinical information (e.g., history of previous hospital admissions and antibiotics exposure), and other relevant information related to risk for ESBL colonization were collected. Rectal swabs were obtained by inserting a single time sterile swab (Guangzhou, China PC: 510530) approximately 2.5 cm and rotated twice in the rectal canal. The swab was removed from the rectal canal after ten seconds, inserted into Stuart transport medium (Guangzhou, China PC:510530), and transported to the Catholic University of Health and Allied Sciences (CUHAS) microbiology Laboratory. In case of rectal swab inconvenience, the participant was provided with a stool container and asked to collect a scoop of feces as an alternative.

Molecular Confirmation of ESBL Enzymes Production
All phenotypic confirmed ESBL-PE were tested for the presence of bla CTX-M , bla SHV , and bla TEM . Bacterial DNA was extracted using the heat treatment method as previously described [29] with some minor modifications. Briefly, loop full colonies of overnight growth of bacteria were suspended into Eppendorf tube containing 500 µL DNase/RNase free water, mixed by vortexing and boiled at 100 • C for 10 min. Tubes were centrifuged at 1200 rpm for 10 min, and 100 µL of the supernatant (DNA rich) was placed in an Eppendorf tube and stored at −20 • C before being used for the detection of bla CTX-M , bla SHV , and bla TEM . Quantity of extracted DNA was determined based on the fluorescence intensity of fluorescent dye binding to double-stranded DNA (dsDNA) using Qubit (Thermo Scientific, 33 Marsiling Industrial Estate Road 3, Singapore ), and quality of extracted DNA was evaluated by electrophoresis in 1.5% agarose gel.

Multiplex PCR Amplifications
A total of 4 µL of DNA samples was used in multiplex polymerase chain reaction (PCR) to detect ESBL genes. This was performed using a specific set of primers targeting ESBL resistance genes (bla CTX-M , bla SHV and bla TEM ), as shown in Table 1 [30]. The bla CTX-M , bla TEM , and bla SHV were selected for amplification because these alleles are the commonest genes encoding for ESBL production in the study setting. Multiplex PCR was carried out in a final volume of 25 µL reaction containing HotStarTaq ® DNA polymerase mastermix (QI-AGEN, Qiagen Str. 1, 40724 Hilden, Germany) and each primer at a reaction concentration of 10 mM. PCR reactions were carried out in thermal cycler machine (BIO-RAD, T100™, WA, USA) and consisted of three steps: initial denaturation at 95 • C for 15 min; 30 cycles of denaturation at 94 • C for 30 s, annealing at 60 • C for 30 s, and extension at 72 • C for 2 min; and a final extension at 72 • C for 10 min.. The amplified products were visualized by electrophoresis in 1.5% agarose gel in TBE (Tris Borate-EDTA) and stained with SYBR green dye safe (Thermo Scientific, 33 Marsiling Industrial Estate Road 3, Singapore). A 100 bp DNA ladder (New England Biolabs; MA, USA) was used as a DNA marker.

Data Management and Data Analysis
Each participant was given a unique identification number. Laboratory data were recorded in the counter book and then transferred to an Excel sheet tallying with specific demographic and clinical data for cleaning and coding. Data analysis was done using STATA software version 15 (Texas, USA). The categorical variables; sex, residence, parent economic status, education level etc., were presented as proportions, whereas the continuous variables such as age, and number of relative living in the same house were summarized as median (IQR: interquartile range). A Pearson Chi square test was used to compare ESBL proportion in various groups (i.e., DM, HIV and SCD). Stepwise logistic regression model was used to determine factors associated with the colonization of ESBL-LF among children with chronic diseases. Factors with p-value less than 0.05 in univariable analysis were subjected to multivariable analysis and factors with p-value < 0.05 were considered statistically significant.  (Table 2).

Factors Associated with ESBL Carriage
Age, sex, residence, parent/guardian level of education and occupation, child history of hospital admission in the past 3 months, child on antibiotic during enrollment, child on antibiotic in the past 3 months, number of members in the household, child on SXT prophylaxis (HIV seropositive only), and child on folic acid/pen V prophylaxis (SCD child only) were included in the model to determine their association with positive ESBL-LF carriage.
The univariate logistic regression analysis results showed that children whose parent were unemployed (OR 6.0, 95%CI: 1.4-26.1 p = 0.004), and being on SXT prophylaxis (OR 2.6, 95%CI 1.1-6.3, p = 0.020) were significantly associated with ESBL-LF colonization among children with HIV infection ( The univariate logistic regression analysis results demonstrated that children with a relative history of admission (OR 2.7, 95% CI: 1.0-7.9, p= 0.037) and parents' education (4.5 [1.69-12.2], p = 0.002) were significantly associated with ESBL colonization among children with SCD. Furthermore, according to the results of the multivariate analysis, a lower primary education level of the parent was found to be associated with ESBL colonization (OR [95%CI]; p value; 5 [1.54-11.3], p = 0.006), (Table 5).
Parent occupation (p = 0.002) and history of hospitalization (p = 0.001) were statistically significant associated with ESBL carriage among children with DM (Table 6).

Factors Associated with ESBL Carriage
Age, sex, residence, parent/guardian level of education and occupation, child history of hospital admission in the past 3 months, child on antibiotic during enrollment, child on antibiotic in the past 3 months, number of members in the household, child on SXT prophylaxis (HIV seropositive only), and child on folic acid/pen V prophylaxis (SCD child only) were included in the model to determine their association with positive ESBL-LF carriage.
The univariate logistic regression analysis results showed that children whose parent were unemployed (OR 6.0, 95%CI: 1.4-26.1 p = 0.004), and being on SXT prophylaxis (OR 2.6, 95%CI 1.1-6.3, p = 0.020) were significantly associated with ESBL-LF colonization among children with HIV infection ( The univariate logistic regression analysis results demonstrated that children with a relative history of admission (OR 2.7, 95% CI: 1.0-7.9, p= 0.037) and parents' education (4.5 [1.69-12.2], p = 0.002) were significantly associated with ESBL colonization among children with SCD. Furthermore, according to the results of the multivariate analysis, a lower primary education level of the parent was found to be associated with ESBL colonization (OR [95%CI]; p value; 5 [1.54-11.3], p = 0.006), (Table 5). Parent occupation (p = 0.002) and history of hospitalization (p = 0.001) were statistically significant associated with ESBL carriage among children with DM (Table 6).

Discussion
This study, conducted for the first time in Mwanza, represents the prevalence of ESBL carriage among children with chronic disease as 23.8%, 23.7%, and 22.2% for DM, HIV, and SCD, respectively, with no statistically significant differences despite the fact that children with HIV and SCD are on daily use of trimethoprim/sulfamethoxazole and Pen V, respectively. However, the study observed that Enterobacteriaceae isolates from HIV-infected children were significantly more resistant to ampicillin, trimethoprim/sulfamethoxazole, and tetracycline. The observed prevalence was significantly high compared to that which was found in the study done on the general population in the same setting, which observed a prevalence of 16.3% [22] and significantly low compared to the study conducted in the same setting among street children, which found a prevalence of 31.8% [19], signifying the role of hygiene in the transmission of ESBL-LF and variations among different population. The high prevalence in children with co-morbidities could be explained by the fact that children with chronic diseases are at increased chances of using antibiotics and being admitted to hospital. Previous studies documented that hospitalization and prior antibiotic use were the risk factors for ESBL colonization [31].
Nearly one third of the phenotypic confirmed ESBL-LF were not typed by multiplex PCR compared to a study done in the same setting [23] that typed 93.3%; this could be explained by the fact that the primers used in the current study targeted the three ESBL genes bla SHV , bla TEM , and bla CTX-M , and untyped isolates may be harboring other ESBL genes like bla OXA , bla PER , bla VEB , and bla BEL , which have been documented in our setting [32]. The bla CTXM has been the most frequently detected ESBL allele among ESBL-LF in our setting and other settings [23,33]. The predominance of the bla CTXM gene may be due to the fact that these genes are encoded by conjugative epidemic plasmids, i.e., IncFII, which play an important role in the effective spreading of this allele [34]; these plasmids have been commonly detected in this setting [35].
The majority of bacterial isolates identified in this study were resistant to multiple classes of antibiotics, as previously reported [36]. Among the three groups, isolates from children with a HIV infection showed significantly high resistance to ampicillin, trimethoprim/sulfamethoxazole, and tetracycline. Similar findings in the same region reported that HIV-infected children had significantly higher multidrug resistance than non-HIV infected children [6]. The level of resistance reflects the high use of these antibiotics in the healthcare system and the community [37]. It should be noted that children living with a HIV infection usually receive long-term trimethoprim/sulfamethoxazole for prophylaxis, as recommended by WHO guidelines [38], and ampicillins are commonly used as the antibiotic of choice in children presenting with upper respiratory infections in primary health care in the community setting.
Furthermore, isolated bacteria generally exhibited a lower resistance to meropenem than other antibiotics, as previously observed in this setting and other settings in Africa [24,39]. The low resistance of bacteria towards meropenem can be explained by the unavail-ability of this antibiotic in low tiers of healthcare facilities, its high cost in community pharmacies, and generally, its prescription in regional and tertiary hospitals is guided by the presence of culture results.
In this study, we observed that children with a HIV infection and those with SCD whose parents have a low education level were significantly more colonized with ESBL-LF. Low parent educational attainment is associated with low socioeconomic status, which has been associated with sub-optimal hygiene, which favors the emergence and spreading of ESBL-PE, as previously reported [19,40]. Furthermore, participants' or relatives' history of hospital admission was associated with ESBL colonization in children with DM, as observed previously [41]. Usually, the majority of hospitalized patients are on antibiotics treatment, hence the antibiotic selection pressure that increases the chances of ESBL colonization [8,24].

Conclusions
A significant proportion of children with DM, HIV, and SCD was colonized with ESBL-PE in the city of Mwanza with HIV-infected children being more colonized by Enterobacteriaceae resistant to AMP, SXT, and TET than non-HIV infected children. Moreover, the bla CTX-M allele was uniformly present in ESBL isolates from children with DM, HIV, and SCD. Nevertheless, children whose parents had a low level of education were significantly more colonized by ESBL-PE. Clinicians should consider ESBL-PE sepsis in children with chronic diseases and institute appropriate management early to prevent associated morbidity and mortality, because in most cases, colonization precedes invasive infections. There is a need to collect more data regarding SXT prophylaxis and resistance to guide the decisions on the practices of providing SXT prophylaxis in HIV-infected children, especially at this time, when testing and treatment is carried out.  Institutional Review Board Statement: Ethical clearance to conduct this study was obtained from the joint BMC/CUHAS research ethics and review committee (CREC) with clearance numbers CREC/475/2021. Permission to conduct the study was obtained from each respective hospital administration.
Informed Consent Statement: Informed consent was obtained from all subjects involved in the study. Parents/guardians were requested to sign written informed consent on behalf of their children, and assent was obtained from older children. Confidentiality was maintained throughout the study.