Pseudomonas aeruginosa isolates recovered from clinical and environmental samples in Buea, Cameroon: current status on biotyping and antibiogram
R. N. Ndip (corresponding author), H. M. Dilonga, L. M. Ndip, J. F. K. Akoachere and T. Nkuo Akenji, Department of Life Sciences, Faculty of Science, University of Buea, Box 63, Buea, Cameroon. Tel.: +237 7526884; Fax: +237 3322272; E-mail: email@example.com
Objective The study was aimed at determining the prevalent biotypes of Pseudomonas aeruginosa in the environment of Buea and the susceptibility of isolates to antibiotics.
Methods One hundred and fifty clinical specimens (urine, wound and sputum) collected from patients attending various health institutions in Buea, and 50 environmental swabs from furniture, appliances and surroundings of these institutions were screened for P. aeruginosa using standard microbiological and biochemical methods. Antimicrobial susceptibility of the isolates was determined by the disc diffusion assay.
Results Fifty-one (25.5%) of the 200 specimens were positive for the organism with urine (30%) being the most common source of isolation. Biochemical characterization grouped the isolates into eight biotypes with biotypes B III (25.59%), B II (23.53%) and B I (21.57%) being the most prevalent. Antimicrobial susceptibility results of isolates revealed 13 antibiotype patterns based on resistance to the antimicrobial agents investigated. The resistance pattern, cefotaxime, gentamicin and tetracycline (CTXR GENR TETR) was the most common (21.6%) amongst the isolates. Over 40% of the isolates exhibited multi-drug resistance to five or more antibiotics, especially environmental isolates. However, there was a significant difference (P < 0.05) in the susceptibility of isolates to ciprofloxacin (98%), amikacin (90.2%) and netilmicin (80.4%) compared with other drugs used in the study.
Conclusion A combination of biotyping and antibiogram, which are relatively cheap and routinely available methods in our environment, could be useful for clinical and epidemiological studies particularly in laboratories in the developing world with limited resources.
Pseudomonas aeruginosa belongs to a vast genus of obligate aerobic, non-fermenting, saprophytic, Gram-negative bacilli widespread in nature, particularly in moist environments (Volk et al. 1991; DuBois et al. 2001). Naturally, this organism is endowed with weak pathogenic potentials. However, its profound ability to survive on inert materials, minimal nutritional requirement, tolerance to a wide variety of physical conditions and its relative resistance to several unrelated antimicrobial agents and antiseptics, contributes enormously to its ecological success and its role as an effective opportunistic pathogen (Gales et al. 2001). The organism is pathogenic when introduced into areas devoid of normal defences (Jawetz et al. 1991) and infections are both invasive and toxigenic (Todar 2002).
The organism has been incriminated in cases of meningitis, septicaemia, pneumonia, ocular and burn infections, hot tubs and whirlpool-associated folliculitis, osteomyelitis, cystic fibrosis-related lung infection, malignant external otitis and urinary tract infections with colonized patients being an important reservoir (Hernández et al. 1997). Cross-transmission from patient to patient may occur via the hands of the health care staff or through contaminated materials and reagents (DuBois et al. 2001). However, it is believed that P. aeruginosa is generally environmentally acquired and that person-to-person spread occurs only rarely (Harbour et al. 2002). As such, contaminated respiratory care equipment, irrigating solutions, catheters, infusions, cosmetics, dilute antiseptics, cleaning liquids, and even soaps have been reported as vehicles of transmission (Joklik et al. 1992; Berrouane et al. 2000; DuBois et al. 2001).
As a result of their clinical significance, many methods exist for the epidemiological investigation of infections caused by P. aeruginosa. In different parts of the world, biotyping, serotyping, antibiogram, phage typing, bacteriocin typing, plasmid profile, and more recent techniques like pulsed-field gel electrophoresis and random amplified polymorphic DNA analyses have been used in typing the organism (Hernández et al. 1997). However, most of these methods are not widely available in routine diagnostic laboratories in developing countries because of the technical difficulties and expenses they involve. It would therefore be an important advance if a cheap, reliable and simplified scheme could be developed, especially for the developing countries.
Interestingly, previous studies (Higaki et al. 1990; Podschum & Ullman 1998) have reported that biotyping, which is based on an extended panel of biochemical tests is certainly the most practical method of typing organisms for smaller laboratories that are epidemiologically not optimally equipped. Against this background, we resolved to use the API 20E (Biomerieux, SA, France) to biotype our isolates.
In the present study, we also used antibiogram as an epidemiological marker for our isolates because there is a high frequency of misuse of antibiotics in the environment of Buea, Cameroon (Ndip et al. 2001). The tetracyclines, oxacillin and to a lesser extent the aminoglycosides, are relatively cheap and commonly available without prescription to the population who tend to misuse them, thus contributing to the emergence of resistance (Ndip et al. 2003). However, it has been widely acclaimed that the susceptibility of pathogens to antibiotics varies with time and geographical location (Abimbola et al. 1993). P. aeruginosa accounts for a significant proportion of nosocomial infections and the tendency of nosocomial pathogens to develop or acquire new antibiotic resistance traits poses a great problem in their treatment and control (Livrelli et al. 1996). We deemed it necessary therefore, to study the susceptibility patterns of P. aeruginosa isolates to some commonly used and relatively reserved antibiotics in Cameroon in order to update our knowledge on the use of these antimicrobial agents in the management of P. aeruginosa infections.
In Buea, Cameroon, there appears to be no report on the biotype distribution and antibiogram of P. aeruginosa. We determined the biotypes and antibiogram of P. aeruginosa isolates in an attempt to provide baseline data for clinical management and epidemiological surveys.
A total of 200 specimens were examined, of which 150 were of clinical origin constituting 50 each of urine (from patients with urinary tract infection), sputum (from patients with respiratory tract infection), and wound swabs (from patients with wounds/burns or surgical patients). These samples were consecutive samples from the various tissue sites sent to the laboratory. They were collected from hospitalized (81) and non-hospitalized (69) patients attending various health institutions in the Buea municipality. Fifty randomly selected environmental swabs were also collected from furniture, medical appliances and the surroundings of these institutions. These included patients’ bed, tables, ward sinks as well as surgical/dressing equipment. Demographic data such as age and sex of patients were recorded. The aim of the study was explained to the patients and their consent to participate in the study solicited. Ethical approval to undertake the study was obtained from the management boards of local hospitals and the Provincial Delegation of Public Health for the South West Province.
Specimens were collected and transported to the laboratory following standard methods (Koneman et al. 1992; Forbes et al. 1998; Cheesbrough 2000). They were plated primarily onto blood agar and incubated at 37 °C for 24–48 h. Suspect isolates were presumptively identified by colony morphology, pigment formation, mucoidy, haemolysis on blood agar, positive oxidase test, grape-like odour, growth at 42 °C on nutrient agar, positive motility, and Gram reaction (Cheesbrough 2000). Isolates were confirmed and grouped into biotypes using the API 20E (BioMerieux SA) kit according to the manufacturer's instruction.
Antibiotic susceptibility testing
The disc diffusion (Kirby–Bauer) technique, which conforms to the recommended standard of the National Committee for Clinical Laboratory Standards (2000) was used as previously described (Bauer et al. 1966). Briefly, a small inoculum of each bacterial isolate was emulsified in 3 ml sterile normal saline in Bijou bottles and the turbidity compared with barium chloride standard (0.5 McFarland). A sterile cotton swab was dipped into the standardized solution of bacterial cultures and used to evenly inoculate Mueller-Hinton plates (Biotec, UK). The plates were allowed to dry. Thereafter, antibiotic discs with the following drug contents aztreonam (30 μg), netilmicin (30 μg), tobramycin (10 μg), gentamicin (10 μg), and amikacin (30 μg) (Becton Dickinson and Company Sparks, USA), ciprofloxacin (5 μg), cefotaxime (30 μg), ceftazidime (30 μg) and tetracycline (10 μg) (Le point de Claix, France) were placed on the plates. Discs were placed at least 15 mm apart and from the edges of the plates to prevent overlapping of inhibition zones. The plates were incubated at 37 °C for 16–18 hours after which zones of inhibition were compared with recorded diameters of the control organism, Escherichia coli ATCC 25922 to determine susceptibility or resistance (Pitout et al. 1998; Ndip et al. 2001).
The chi-square test was employed at 5% significance level.
Prevalence of P. aeruginosa
Of the 200 specimens examined, 51 (25.5%) were positive for P. aeruginosa. There was no significant difference (P > 0.05) in prevalence between clinical 24.7% (37/150), and environmental 28% (14/50) specimens. The highest isolation rate (30%) was observed in urine specimens and the least (16%) in sputum (P > 0.05) (Table 1). Table 2 represents the distribution of environmental isolates by site of isolation.
Table 1. Prevalence of P. aeruginosa in the samples
|No. positive (%)||15 (30)||14 (28)||8 (16)||14 (28)||51 (25.5)|
|No. negative (%)||35 (70)||36 (72)||42 (84)||36 (72)||149 (74.5)|
|Total (%)||50 (100)||50 (100)||50 (100)||50 (100)||200|
Table 2. Distribution of environmental isolates by site of isolation
|Patients’ bed||13||6 (42.9)|
|Ward sink||12||3 (21.4)|
|Surgical/dressing equipment||13||2 (14.3)|
Prevalence of P. aeruginosa with respect to age and sex
There were no significant differences (P > 0.05) in the isolation rate between age groups and gender (27.6% females and 24.5% males); as expected, positive urine samples were more common in females. In addition, there was a trend for a higher positive rate in those aged ≥60 years.
Biochemical characterization of isolates
The biotyping scheme, comprising sugar fermentation and other biochemical parameters is presented in Table 3. Results indicated that 11 (21.5%), 12 (23.5%), and 13 (25.5%) isolates belonged to biotypes B I, B II and B III respectively. Four (7.8%) of the isolates were grouped under biotypes B IV and B V; three (5.8%) as biotype B VIII while two (3.9%) were classified as biotypes B VI and B VII.
Table 3. Biotypes of P. aeruginosa strains isolated from clinical and environmental samples
As shown in Table 4, the more prevalent (66.7%) biotype B VIII occurred in wound isolates. Most of the isolates from urine samples (50%) belonged to biotype B II while biotype B III was more prevalent (38.5%) in sputum isolates. Although biotypes B I and B III constituted just 36.4% and 30.8% of environmental isolates, respectively, they each had four isolates from the 14 obtained in this study. However, these were not statistically significant (P > 0.05). Interestingly, there was a possible link between biotype B I from wound and environmental isolates because the same biotype was seen in some wounds and patients’ bed.
Table 4. Specimen sources and occurrence of biotypes of P. aeruginosa isolates
|Urine||3 (27.3)||6 (50)||3 (23.1)||1 (25)||1 (25)||1 (50)||0 (0.0)||0 (0.0)||15|
|Wound||4 (36.4)||4 (33.3)||1 (7.7)||1 (25)||1 (25)||0 (0.0)||1 (50)||2 (66.7)||14|
|Sputum||0 (0.0)||1 (8.3)||5 (38.5)||0 (0.0)||1 (25)||0 (0.0)||0 (0.0)||1 (33.3)||8|
|Environmental||4 (36.4)||1 (8.3)||4 (30.8)||2 (50)||1 (25)||1 (50)||1 (50)||0 (0.0)||14|
A composite biotype–antibiotype profile is shown in Table 5. We observed a maximum of six isolates belonging to the same biotype (B III) and antibiotype (A11). Of the six isolates, five (83.3%) were from sputum and one (16.6%) from a wound specimen. Interestingly, most of these patients were hospitalized.
Table 5. Composite biochemical–antimicrobial profile of isolates
|B I|| || || ||4|| || || ||2|| || ||2||2||1||11|
|B II||1|| ||1||5||1|| || ||2|| || || ||1||1||12|
|B III||1||1|| ||1|| || || || || ||1||6||1||2||13|
|B IV|| || || || || || || ||1||1||1|| || ||1||4|
|B V||1|| || ||1||1|| || || || || ||1|| || ||4|
|B VI||1|| || || || ||1|| || || || || || || ||2|
|B VII|| || || || || || ||1|| || || || ||1|| ||2|
|B VIII||2|| || || || ||1|| || || || || || || ||3|
Antimicrobial susceptibility patterns
The antimicrobial patterns exhibited by the isolates are shown in Table 6. The predominant resistance pattern, cefotaxime, gentamicin and tetracycline (CTXR GENR TETR) was observed in 21.6% of isolates. Nine (17.7%) showed resistance to ceftazidime, cefotaxime, tobramicine, gentamicin, aztreonam and tetracycline (CAZR CTXR NNR GENR ATMR TETR), six (11.7%) to cefotaxime and tetracycline (CTXR TETR). The least resistance patterns were exhibited respectively by one (1.9%) isolate showing resistance corresponding to patterns A2, A3, A7 and A9. Ninety-eight per cent of the isolates were susceptible to ciprofloxacin. Amikacin and netilmicin showed good activity with 90.2% and 80.4% susceptibility rates, respectively (Table 7).
Table 6. Antimicrobial resistance patterns of P. aeruginosa strains isolated from clinical and environmental samples
|A1||CTXR TETR||6 (11.7)|
|A2||CTXR GENR||1 (1.9)|
|A3||GENR TETR||1 (1.9)|
|A4||CTXR GENR TETR||11 (21.6)|
|A5||CTXR ATMR TETR||2 (3.9)|
|A6||CTXR NNR TETR||2 (3.9)|
|A7||CTXR NNR ATMR TETR||1 (1.9)|
|A8||CTXR NNR GENR TETR||5 (9.8)|
|A9||CIPR NNR ATMR TETR||1 (1.9)|
|A10||CTXR NNR GENR ATMR TETR||2 (3.9)|
|A11||CAZR CTXR NNR GENR ATMR TETR||9 (17.7)|
|A12||CAZR CTXR NNR NETR GENR ATMR TETR||5 (9.8)|
|A13||CAZR CTXR ANR NNR NETR ATMR TETR||5 (9.8)|
|Total|| ||51 (100)|
Table 7. Antibiotic sensitivity results of P. aeruginosa strains isolated from clinical and environmental samples
|Ciprofloxacin||1 (2)||0 (0)||1 (2)||50 (98)|
|Ceftazidime||12 (23.6)||7 (13.7)||19 (37.3)||32 (62.7)|
|Cefotaxime||13 (25.5)||36 (70.6)||49 (96.1)||3 (3.9)|
|Amikacin||2 (3.9)||3 (5.9)||5 (9.8)||46 (90.2)|
|Netilmicin||7 (13.7)||3 (5.9)||10 (19.6)||41 (80.4)|
|Gentamicin||10 (19.6)||24 (47.1)||34 (66.7)||17 (33.3)|
|Tobramicin||14 (27.4)||16 (31.4)||30 (58.8)||21 (41.2)|
|Aztreonam||13 (25.5)||12 (23.5)||25 (49)||26 (51)|
|Tetracycline||14 (27.4)||36 (70.6)||50 (98)||1 (2)|
Pseudomonas aeruginosa is a leading cause of hospital-acquired infections with a high propensity to develop, acquire or transfer antimicrobial resistance genes (Gales et al. 2001). This phenomenon is associated with increased rates of morbidity, mortality and high cost of treatment (Köhler et al. 2001). In this study, we report on the prevalence and antimicrobial susceptibility pattern of this organism in the locality of Buea, Cameroon.
The organism was isolated from all four specimen types (urine, wound, sputum and environmental swabs) screened with an overall prevalence of 25.5% (51/200), which reflects the ubiquitous nature of the organism as previously reported (Baron 1986; Worgall et al. 2001; Doern et al. 2002). The prevalence of the organism in clinical specimens (24.7%) was not significantly different (P > 0.05) from that of environmental specimens (28%). The highest and least occurrences were in urine (30%) and sputum (16%) samples, respectively.
Inspite of the clinical importance of P. aeruginosa, there is no definite scheme developed locally for biotyping P. aeruginosa isolates in our environment. The biotyping scheme in this study made use of sugar fermentation and other miniaturized tests in the API 20E kit. Biotyping with this kit identifies the organism with up to 99% accuracy (Koneman et al. 1992). Based on these tests, eight biotypes were identified (Table 3). Our results indicated that 11, 12, and 13 of the isolates belonged to biotypes B I, B II and B III, respectively. Four, three and two of the isolates were grouped under biotypes B IV and B V, B VIII and B VI, and B VII in that order. The differences in the patterns were mainly at the level of adenine dihydrolase and glucose, for which 66.7% and 54.9% of the isolates were positive, respectively.
Biotypes B II, B III and B V occurred in all specimens. While biotype B V was equally distributed in all the samples, B II and B III were more prevalent in urine (50%) and sputum (38.5%) samples respectively (Table 4). All except biotype B VIII were isolated from both clinical and environmental samples suggesting a relationship between environmental and clinical acquisition of the organism as previously reported (Harbour et al. 2002). Our finding linking biotype B I from wound and environmental isolates corroborates this as the environmental isolates belonging to this biotype was mostly from patients’ bed. We are therefore constrained to speculate that patients’ beds might have served as a vehicle for contaminating wounds. In a similar study in Brazil, Freitas and Barth (2004) typed P. aeruginosa using a combination of biotyping, antibiotyping and DNA typing to characterize P. aeruginosa isolates from hospitalized patients. They detected 10 biotypes among their isolates and concluded that biotyping has a low discriminatory power thus corroborating our results. We therefore think that elaborate studies need to be carried out with more discriminatory typing methods such as genotyping to be able to draw definitive conclusions because other studies have also shown that some conventional methods like biochemical profiles for strain identification can only detect unstable differences between clonally related strains or may be applied to just few isolates (Gallego et al. 2000). This could be linked to variation in certain phenotypic characteristics of genetically related strains.
Antimicrobial resistance patterns revealed a total of thirteen patterns of which the most prevalent displayed resistance to cefotaxime, gentamicin and tetracycline (CTXR GENR TETR) and they accounted for 11 (21.6%) of P. aeruginosa isolates. This was closely followed by the pattern showing resistance (17.7%) to ceftazidime, cefotaxime, tobramicine, gentamicin, aztreonam and tetracycline (CAZR CTXR NNR GENR ATMR TETR). The least resistance patterns were noted for cefotaxime and gentamicin (CTXR GENR), gentamicin and tetracycline (GENR TETR) and ciprofloxacin, tobramicine, aztreonam and tetracycline (CIPR NNR ATMR TETR) exhibited by 1.9% of the isolates respectively. Approximately 84% of isolates were resistant to three or more antibiotics. Of these, 41.2% were resistant to five or more drugs, the majority coming from environmental samples. Multi-drug resistance in environmental isolates might be linked to the uncontrolled disposing of antibiotics and chemicals into the environment creating a selective pressure on these drugs. The use of antibiotics in hospital and the community at large serves as a major selective pressure for antibiotic resistant bacteria (Moreira et al. 2002). Multi-drug-resistant nosocomial infections by this organism are increasing worldwide (Luh et al. 1998). The existence of metallo-β-lactamases and extended-spectrum β-lactamase-producing strains exhibiting resistance to most β-lactams antimicrobial agents greatly complicate the clinical management of patients infected with such multi-drug-resistant strains (Moreira et al. 2002; Pagani et al. 2002).
It was interesting to note that except for the pattern (CTXR GENR), tetracycline showed resistance to all the isolates in the other 12 patterns. The remarkable multiple resistances to tetracycline could be attributed to the fact that this antibiotic is highly misused because of constant and indiscrimate usage in our environment (Akoachere et al. 2002), and an intrinsic and acquired resistance mechanism caused mainly by an active efflux system, which efficiently expels the compound from the cell (Köhler et al. 2001). However, susceptibility of P. aeruginosa to tetracycline has been reported in patients with burns and wounds (Estahbanati et al. 2002).
Pseudomonas aeruginosa resistance to antibiotics is a serious problem in hospitals in Africa. A number of similar studies conducted in South Africa, Cote d'Ivoire, Tunisia and Nigeria (Aka et al. 1987; Kamoun et al. 1992; Rotimi et al. 1994; Poirel et al. 2001, 2002) documented the existence of multi- resistant strains of P. aeruginosa responsible for nosocomial infections. This high level resistance is thought to be plasmid-mediated.
However, in the present study, most of our isolates (98%) were susceptible to ciprofloxacin. This corroborates a previous report, which documented ciprofloxacin as the most potent agent available in oral form for the treatment of P. aeruginosa infections (Gales et al. 2001). The high cost of this drug in our environment, which limits its use, may account for this observation. However, resistance to this drug has been reported in other regions of the world (Bouza et al. 1999), and is thought to be due to improper dosage prescription. It is likely that the dose normally prescribed is too low to kill the micro-organism. Similarly, high susceptibility was observed for amikacin and netilmicin with 90.2% and 80.4% of isolates demonstrating susceptibility to the drugs, respectively. This is in agreement with a previous finding (DuBois et al. 2001) that documented good activity of some β-lactams, fluoroquinolones and aminoglycosides against P. aeruginosa.
Despite the limitations of both biotyping and antibiogram methods, it is occasionally useful to employ these two simple and routinely available methods in combination to produce a composite antimicrobial–biochemical profile (Balows et al. 1991). Putting them together to establish their relatedness in strain identification, we observed a low concordance with a maximum of six isolates belonging to the same biotype (B III) and antibiotype (A11). This reflects the poor discriminatory power of traditional phenotypic typing methods, which are based on unstably expressed characteristics for strain identification (Gallego et al. 2000), as phenotypic variation could exist in genetically related strains.
Our results indicate a high level of resistance of P. aeruginosa to commonly prescribed antibiotics, a situation that merits immediate attention. We are of the opinion that health care practitioners and policy makers could address this problem by implementing a more rational and appropriate use of antibiotics. This may include amongst others restricting sale of antibiotics by pharmacies only to patients having prescription from qualified medical practitioners.
In conclusion, however, we recommend a combination of these methods, which are cheap and routinely available in most laboratories in the developing world with limited resources as useful tools for clinical and epidemiological studies.
We are grateful to the University of Buea for the equipment put at our disposal. We also remain indebted to the management of the hospitals used for sample collection.