Prevalence and spread of extended-spectrum β-lactamase-producing Enterobacteriaceae in Ngaoundere, Cameroon

Authors


Abstract

During April 2010 and June 2010, 334 Enterobacteriaceae isolates from 590 participants (outpatients, inpatients, inpatient carers, hospital workers and members of their households) were collected from faecal samples. Based on β-lactamase pattern, origin of strains and the relationship between participants, 44 isolates of extended-spectrum β-lactamase (ESBL)-producing Escherichia coli and Klebsiella pneumoniae were selected from 44 participants (in Ngaoundere Protestant Hospital and Ngaoundere Regional Hospital, Cameroon). To determine the relatedness of bacterial strains, these isolates were fingerprinted using the automated, repetitive-sequenced-based PCR-based DiversiLab system. Subsequently, E. coli isolates that had undergone DiversiLab analysis were examined with respect to their phylogenetic group and detection of the ST131 clone to shed light on the epidemiology of these isolates in the Ngaoundere hospitals. The prevalence of faecal carriage of ESBL-producing Enterobacteriaceae among the study participants was 54.06%. According to participant groups, the prevalence of faecal carriage was also high (outpatients 45%; inpatients 67%; inpatient carers 57%; hospital workers 44%; and members of their households 46%). Analysis of the molecular epidemiology of ESBL-producing E. coli and K. pneumoniae showed a close relationship of the isolates between related and non-related individuals. In addition, DiversiLab results of E. coli identified four related isolates (4/22) from cluster III belonging to the epidemiologically important clone ST131. Our results highlight the importance of outpatients, inpatients, their carers, hospital workers and their families as reservoirs of ESBL-producing Enterobacteriaceae

Studies have shown the presence of extended-spectrum β-lactamase-producing Enterobacteriaceae (E-ESBLs) in Cameroon, not only in hospitals but also in the community [1-4]. However, although E-ESBLs are present and increasingly prevalent in the country, little is known about their dissemination. To find out the prevalence of faecal carriage of E-ESBLs and how they are spreading, an epidemiological study of E-ESBLs involving outpatients, inpatients, inpatient carers, hospital workers and members of their households was performed in hospitals in Ngaoundere, Cameroon. Written informed consent was obtained from all participants. Participants' characteristics are shown in Table 1.

Table 1. Characteristics of all participants included in the study
CharacteristicOutpatients (= 232)Inpatients (= 208)Inpatient carers (= 63)Hospital workers (= 48)Household members (= 39)
E-ESBL+ = 104 (45%)E-ESBL = 128 (55%)E-ESBL+ = 140 (67%)E-ESBL = 68 (33%)E-ESBL+ = 36 (57%)E-ESBL = 27(43%)E-ESBL+ = 21 (44%)E-ESBL = 27 (56%)E-ESBL+ = 18 (46%)E-ESBL = 21 (54%)
  1. E-ESBL, extended-spectrum β-lactamase-producing Enterobacteriaceae; E-ESBL, E-ESBL-negative; E-ESBL+, E-ESBL-positive; SD, standard deviation.

Median age, years (±SD)35 ± 1536 ± 15.537 ± 17.441 ± 1842 ± 14.839 ± 13.535 ± 6.7237 ± 10.5217 ± 1314 ± 9.24
Male gender (n)4060713968916814
Previous hospital admission
Yes13234614533314
No78838853312417191413
Unknown132261001534
Previous antimicrobial treatment
Yes485673287611653
No4454523227198161114
Unknown1218158222524

A total of 334 Enterobacteriaceae were isolated on two selective media, Drigalski and MacConkey agars, supplemented, respectively with cefotaxime and ceftazidime. Detection of ESBL producers was carried out by the double-disc synergy test [5]. Susceptibility of the isolates to antibiotics was determined using the Vitek system (bioMérieux, Marcy l'Etoile, France). All the presumptive ESBL producers were further analysed by PCR aimed at detecting ESBL genes (one isolate of each morphotype from each patient). PCR amplification of bla genes (blaTEM, blaSHV, blaOXA and blaCTX-M) was performed using primers and methods described previously [6, 7] (Table 2). All PCR products were sequenced using a 3100 ABI Prism Genetic Analyser (Applied Biosystems, Foster City, CA, USA). Sequence alignment and analyses were performed online using the BLAST program available at the National Center for Biotechnology Information (http://www.ncbi.nlm.nih.gov).

Table 2. Primers used for PCR amplification
PCR nameβ-Lactamase(s) targetedPrimer nameSequence (5′-3′)Amplicon size (bp)Primer concentration (pmol/μL)Reference
Multiplex TEM,TEM variants including, TEM-1 and TEM-2MultiTSO-T_forCATTTCCGTGTCGCCCTTATTC8000.4 [ [6] ]
  MultiTSO-T_revCGTTCATCCATAGTTGCCTGAC 0.4 
SHVSHV variants, including SHV-1MultiTSO-S_forAGCCGCTTGAGCAAATTAAAC7130.4 
  MultiTSO-S_revATCCCGCAGATAAATCACCAC 0.4 
and OXA-1- likeOXA-1, OXA-4 and OXA-30MultiTSO-O_forGGCACCAGATTCAACTTTCAAG5640.4 
  MultiTSO-O_revGACCCCAAGTTTCCTGTAAGTG 0.4 
CTX-M group 1Variants of CTX-M group 1, including CTX-M-1, CTX-M-3 and CTX-M-15CTX-M3G_forGTTACAATGTGTGAGAAGCAG10500.5 [ [7] ]
  CTX-M3G_revCCGTTTCCGCTATTACAAAC 0.5 

To determine the relatedness of bacterial strains, 44 isolates (22 Escherichia coli and 22 Klebsiella pneumoniae) were recovered from 44 participants and fingerprinted using the automated repetitive-sequence-based PCR DiversiLab system (bioMérieux). The strains were selected based on their β-lactamase pattern, their origin and the relationship between participants. The relationships between repetitive-sequence-based PCR profiles were designated as recommended by the manufacturer: different, 3+ band differences (similarity <95%); similar, 1–2 band differences (similarity 95–97%); and indistinguishable, no band differences (similarity >97%).

Subsequently, the 22 E. coli isolates that had undergone DiversiLab analysis were investigated by a series of PCRs [8-10] to determine the phylogenetic groups (A, B1, B2 and D), the presence of ISEcp1 elements and the aac(6)-Ib-cr variant (the gene that can induce resistance to aminoglycoside and fluoroquinolone simultaneously).

All 22 E. coli ESBL-producing isolates were screened for sequence type 131 (ST131) using a PCR for the pabB allele, described by Clermont et al. [11]. Multilocus sequence typing was performed on positive pabB isolates using the Achtman typing scheme (http://www.mlst.ucc.ie/mlst/dbs/Ecoli).

Statistical analysis was performed with Epi Info version 3.5.3. Fisher's exact test, when appropriate, was used for the univariate comparison of all variables. A p value of p <0.05 was considered to be statistically significant.

The overall prevalence of faecal carriage of E-ESBLs in this study was 54.06%. Regarding participant groups and their relatives, the prevalence was also high (see Table 2). However, the prevalence of faecal carriage among inpatients was not significantly different from that of their carers (p >0.05). Similarly, the prevalence of faecal carriage among hospital workers was not significantly different from that of their household members (p >0.05).

Of the 334 bacteria isolated, 216 (64.67%) were E. coli, 74 (22.15%) were Klebsiella spp., 23 (6.88%) were Enterobacter spp. and 21 (6.28%) were Citrobacter freundii. Regardless of group, participants presented similar ESBLs and CTX-M-15 was the most widespread ESBL found in all strains (Table 3).

Table 3. Overview of extended-spectrum β-lactamase (ESBL) types produced by Enterobacteriaceae isolated in all groups of participants
ESBL types (number of isolates)
Strain (= 334)OutpatientsInpatientsInpatient carersHospital workersHousehold members
Citrobacter freundii (= 21)CTX-M-15, OXA-1 (8)CTX-M-15 (2)  CTX-M-15, TEM-1 (1)
CTX-M-15, TEM-1 (2)CTX-M-15, OXA-1, TEM-1 (2)  CTX-M-15, OXA-1, TEM-1 (1)
SHV-12, TEM-1 (2)CTX-M-15, TEM-1 (1)   
 SHV-12, TEM-1 (1)   
 SHV-12 (1)   
Enterobacter spp. (= 23)CTX-M-15, OXA-1, TEM-1 (14)CTX-M-15, OXA-1, TEM-1 (5)  CTX-M-15, OXA-1, TEM-1 (2)
SHV-12, TEM-1, OXA-1 (1)SHV-12, TEM-1 (1)   
Escherichia coli (= 216)CTX-M-15, OXA-1 (26)CTX-M-15, OXA-1, TEM-1(42)CTX-M-15, OXA-1, TEM-1 (11)CTX-M-15, OXA-1 (8)CTX-M-15, TEM-1 (8)
CTX-M-15, TEM-1 (15)CTX-M-15, OXA-1 (30)CTX-M-15, OXA-1 (9)CTX-M-15, OXA-1, TEM-1 (4)CTX-M-15, OXA-1, TEM-1 (5)
CTX-M-15, OXA-1, TEM-1 (14)CTX-M-15, TEM-1 (24)CTX-M-15, TEM-1 (8)CTX-M-15, TEM-1 (2) 
CTX-M-15 (3)CTX-M-15 (3) CTX-M-15, SHV-12, OXA-1, TEM-1(1) 
 SHV-12, TEM-1 (1) CTX-M-15 (1) 
 SHV-12 (1)   
Klebsiella spp. (n = 74)CTX-M-15, SHV-1, TEM-1 (8)CTX-M-15, SHV-1, TEM-1 (13)CTX-M-15, SHV-1, TEM-1 (7)CTX-M-15, SHV-1, TEM-1 (2)CTX-M-15, SHV-1, TEM-1 (1)
CTX-M-15, TEM-1 (8)CTX-M-15, TEM-1 (6)CTX-M-15, SHV-1, OXA-1, TEM-1(1)CTX-M-15, OXA-1, TEM-1 (2)CTX-M-15 (1)
CTX-M-15, OXA-1, TEM-1 (5)CTX-M-15, SHV-1, OXA-1, TEM-1(5)CTX-M-15, TEM-1 (1)CTX-M-15, OXA-1 (1) 
CTX-M-15, OXA-1 (2)CTX-M-15, OXA-1, TEM-1 (4)CTX-M-15, OXA-1 (1)CTX-M-15, TEM-1 (1) 
CTX-M-15, SHV-1, OXA-1, TEM-1 (1)CTX-M-15, SHV-12, TEM-1 (1)SHV-12 (1)  
SHV-12 (1)CTX-M-15 (1)   

As demonstrated by the DiversiLab results (see Supplementary material, Fig. S1 and Fig. S2), multiple clones were seen among ESBL-producing E. coli and using a cut-off of 95% similarity (vertical dashed red line, Fig. S1) it was possible to establish six distinct clusters (types I, III, V, VI, VII and X), indicating the overall close relationship of the isolates, and five singleton patterns (types II, IV, VIII, IX and XI) (Table 4). These six outbreaks occurred between related or non-related individuals. This finding confirms that in some cases the acquisition of E-ESBLs has occurred via a common source or a common environmental reservoir (hands of hospital workers, through contaminated surfaces or via food) as described previously [12-17].

Table 4. Epidemiological data of the 22 selected strains of extended-spectrum β-lactamase (ESBL) -producing Escherichia coli in Ngaoundere Protestant Hospital
Alternative identificationType of participants (department in which the participant was at the time of sampling)Type of relationship/gender (age)Phylogenetic groupCluster typeST131 PCRPMQRISEcp1 elementESBL typesResistance to antibiotics (other than β-lactam)
  1. M, male; F, female; PMQR, plasmid-mediated quinolone-resistance determinants; GEN, gentamicin; SXT, trimethoprim/sulfamethoxazole; CIP, ciprofloxacin; Nit, nitrofurantoin.

f32 Household member of p3Daughter (3)DICTX-M-15, TEM-1GEN, SXT
20 Maintenance staff (medicine)M (42)DI aac(6′)Ib-cr CTX-M-15, OXA-1, TEM-1GEN, SXT
hn50 Inpatient (medicine)F (29)DII aac(6′)Ib-cr +CTX-M-15, OXA-1, TEM-1CIP, GEN, SXT
hn89 Inpatient (medicine)M (31)B2III+ aac(6′)Ib-cr +CTX-M-15, OXA-1CIP, SXT
p13 Nurse (medicine)F (28)B2IIIa+ aac(6′)Ib-cr +CTX-M-15, OXA-1CIP, SXT
p3 Nurse (surgery)F (41)B2IIIb++CTX-M-15, TEM-1CIP, SXT
p25 Nurse (intensive care unit)F (35)B2III+CTX-M-15, TEM-1CIP, SXT
hn69 Inpatient (surgery)M (40)B2IV aac(6′)Ib-cr +CTX-M-15, OXA-1, TEM-1CIP, GEN, SXT
p5 Nurse (medicine)F (30)AV aac(6′)Ib-cr CTX-M-15, OXA-1CIP, GEN, NIT, SXT
16 Maintenance staff (medicine)M (38)AV aac(6′)Ib-cr CTX-M-15, OXA-1CIP, GEN, NIT, SXT
hn82 Inpatient (medicine)M (39)AVa aac(6′)Ib-cr CTX-M-15, OXA-1CIP, GEN, NIT, SXT
12 Maintenance staff (operating room)F (25)AVb aac(6′)Ib-cr CTX-M-15, OXA-1CIP, GEN, NIT, SXT
p4 Nurse (medicine)M (38)AVc aac(6′)Ib-cr CTX-M-15, OXA-1CIP, NIT, SXT
p17 Nurse (surgery)F (27)AVI aac(6′)Ib-cr CTX-M-15, OXA-1CIP, SXT
3 Maintenance staff (operating room)M (41)AVIa aac(6′)Ib-cr CTX-M-15, OXA-1CIP, SXT
f2 Household member of 11Sister (18)AVII aac(6′)Ib-cr +CTX-M-15, OXA-1, TEM-1GEN, SXT
11 Maintenance staff (maternity)F (35)AVII aac(6′)Ib-cr +CTX-M-15, OXA-1, TEM-1GEN, SXT
ng33 Carer of hn57Husband (53)AVIIICTX-M-15, OXA-1, TEM-1CIP, SXT
ng41 Carer of hn69Wife (35)B1IXCTX-M-15, OXA-1, TEM-1CIP, SXT
ng32 Carer of hn57Sister (55)B1X aac(6′)Ib-cr +CTX-M-15, OXA-1, TEM-1CIP, GEN, SXT
p16 Nurse (medicine)F (33)B1X aac(6′)Ib-cr +CTX-M-15, OXA-1, TEM-1GEN, SXT
hn57 Inpatient (medicine)F (43)DXICTX-M-15, OXA-1, TEM-1CIP, GEN, SXT

Molecular typing of K. pneumoniae isolates showed them to be more genotypically diverse than the E. coli isolates. The repetitive-sequence-based PCR analysis allowed the identification of two clusters and 17 unique profiles (see Supplementary material, Fig. S3 and Fig. S4). Our results emphasize the high endemicity of CTX-M-15 producers in the study setting (demonstrated by the very high prevalence in all participant categories). In addition, these findings could also explain the dramatic increase in the prevalence of faecal carriage of E-ESBLs previously observed in the same area during two non-outbreak periods separated by 1 year (in 2009 and 2010) [3, 4] (and data of this study). Moreover, examination of the different clusters (III, V and VI) showed that different strains of E. coli produced the same type of β-lactamase, indicating that the spread does not occur by strain but by another mode of dissemination (e.g. dissemination by plasmids: further studies are ongoing to clarify this).

The PCR for the pabB allele of ST131 status identified cluster III with four related isolates as belonging to ST131. The ST131 status was confirmed by multilocus sequence typing. In addition, isolates from this cluster could be assigned to phylogenetic group B2. The remaining ESBL-producing E. coli isolates (negative by PCR for pabB) belonged to phylogenetic groups A, B2, D and B1 (ten, one, four and three isolates, respectively). Interestingly, strains that belonged to the same phylogenetic group and clustered together showed a similar resistance profile to the antibiotics tested (Table 4). Many of the E. coli carrying blaCTX-M-15 from different countries in Europe and North America are homogeneously grouped as E. coli O25:H4-ST131 [11, 18-20]. In addition, in all blaCTX-M-15 tested, the aac(6′)-Ib-cr gene was carried by 16 strains (16/22) and the ISEcp1 element was found in nine of the strains analysed (9/22). One study has reported that CTX-M ESBL genes are often associated with an ISEcp1 element, which may provide a higher level of expression of the plasmid-located blaCTX-M genes and facilitate the spread of resistance [21]. Our study provides the first evidence for the presence of an E. coli ST131 clone in Cameroon. Strict hygiene measures, staff training to improve hand-washing procedures and changes to the antibiotic policy are essential to limit the spread of these E-ESBLs in hospitals and the community.

Acknowledgements

We would like to acknowledge the Chief Doctor, Simon Z. Aroga, and technicians at the microbiology laboratory in Ngaoundere Protestant Hospital who contributed to the study.

Transparency Declaration

All the authors declare that they have no conflicts of interest.

Ancillary