A Multidrug-Resistant Staphylococcus epidermidis Clone (ST2) Is an Ongoing Cause of Hospital-Acquired Infection in a Western Australian Hospital

Abstract

We report the molecular epidemiology of 27 clinical multidrug-resistant Staphylococcus epidermidis (MDRSE) isolates collected between 2003 and 2007 in an Australian teaching hospital. The dominant genotype (sequence type 2 [ST2]) accounted for 85% of the isolates tested and was indistinguishable from an MDRSE genotype identified in European hospitals, which may indicate that highly adaptable health care-associated genotypes of S. epidermidis have emerged and disseminated worldwide in the health care setting.

TEXT

Staphylococcus epidermidis is an important pathogen involved in health care-associated bloodstream infections and infections related to vascular catheters and prosthetic devices (17). Several investigations have demonstrated that certain multidrug-resistant S. epidermidis (MDRSE) genotypes become established as opportunistic pathogens in the health care setting as a novel ecological niche (8, 10, 12). In addition, recent studies identified several worldwide epidemic clonal lineages (9, 12, 13, 18, 21). Currently, little information is available on the molecular epidemiology of S. epidermidis in the health care setting in Australia (16, 18).

We have previously documented the occurrence and potential dissemination of two genotypes of MDRSE in 11 hospitals in northern Europe between 2001 and 2008. The aims of this study were to examine the molecular epidemiology of clinical isolates of MDRSE collected in a teaching hospital in Australia and determine the possible presence of previously described health care-associated MDRSE clones.

Twenty-seven MDRSE isolates were collected between 2003 and 2007 from patients admitted to Royal Perth Hospital, Western Australia. One isolate per patient was included in the study.

The S. epidermidis isolates were identified using a phenotypic disc method (2) or bioMérieux Vitek2 Compact GP identification card (1). In addition, one isolate from each pulsed-field gel electrophoresis (PFGE) type was identified as S. epidermidis using rpoB sequencing (11). Antimicrobial susceptibility testing was performed by disc diffusion on Mueller-Hinton agar (BBL; Becton Dickinson, Cockeysville, MD) using Oxoid discs according to the guidelines of the Clinical and Laboratory Standards Institute (CLSI) (4). The antimicrobials tested included cefoxitin (10 μg), erythromycin (15 μg), ciprofloxacin (5 μg), fusidic acid (10 μg), gentamicin (10 μg), rifampin (5 μg), tetracycline (30 μg), and mupirocin (5 μg). CLSI susceptibility interpretive criteria (5) were utilized for all antimicrobials with the exception of fusidic acid (3) and mupirocin (7).

Retrospective medical chart review of the 27 patients whose isolates were included in the study showed that 23 were treated for hospital-acquired MDRSE infections (Table 1). The median age among these 9 women and 14 men was 53 years (range, 20 to 88 years). Eight patients with hematological malignancy, all on hematology wards, had positive blood cultures and presumed line-related sepsis between February 2005 and June 2007. Six of these were treated with either vancomycin or linezolid, one with meropenem and ciprofloxacin, and in one case the treatment details are not available. Four patients with intracranial hemorrhage and shunt insertion were diagnosed with ventriculitis or ventriculoperitoneal shunt infection between January 2004 and February 2006. These patients were treated with IV vancomycin with or without intraventricular vancomycin. Four patients with prosthetic-joint infections were all treated with vancomycin with or without other antibiotics between April 2006 and April 2007. In addition, six patients were diagnosed with other health care-associated MDRSE infections and treated with vancomycin: one pacemaker infection, one postoperative discitis following laminectomy, one postoperative endophthalmitis, one postoperative infection following surgery for trauma, one postoperative thigh collection, and one meliodosis complicated by line-related sepsis. Lastly, one patient with line-related sepsis was treated by line removal and other antibiotics (isolate 11). The time span for these positive cultures was June 2003 to June 2007. The median length of stay (LOS) for patients with significant infection (available data on 21 patients) was 36 days (range, 10 to 282 days).

Table 1.

Epidemiological and clinical data for the 27 methicillin-resistant Staphylococcus epidermidis isolates included in the study in order of appearance in Fig. 1^c

No.	Isolate					Patient
	PFGE type	ST typeb	SCCmec type	Antibiogram result	Source	Gender and age (yr)	Underlying condition	Diagnosis	Admission LOS (no. of days/yr)	Antibiotic treatment
										Preculture	Postculture
1a	I	5	IV (2B)	CIP ERY GEN MUP	Blood	NA	NA	Presumed line-related sepsis	NA/2006	NA	NA
2	I		III (3A&5)	CIP ERY GEN MUP	Knee tissue	F 58	Osteoarthritis	Septic total knee replacement	27/2006	CEF	VAN
3	D		IV (2B)	CIP ERY FUS GEN	Blood	F 63	Burkitt's lymphoma	Line-related sepsis	39/2006	TZP AZM	MEM CIP
4	D		IV (2B)	CIP ERY FUS GEN	Blood	M 48	AML	Line-related sepsis	23/2006	TZP	VAN
5	D		IV (2B)	CIP ERY FUS GEN MUP	Blood	M 29	Heart transplant, idiopathic myopathy	Septic thrombophlebitis second to pacemaker insertion	16/2006	SXT TIM	VAN
6	D		IV (2B)	CIP ERY FUS GEN MUP	Blood	F 72	AML	Line-related sepsis	24/2006	None	LZD
7	D		IV (2B)	CIP ERY FUS GEN	Blood	M 56	Multiple myeloma, allogeneic stem cell transplant, GVHD	Line-related sepsis	37/2007	NA	FEP VAN
8	D		IV (2B)	CIP ERY FUS GEN MUP RIF	Blood	M 19	ALL, bone marrow transplant	Line-related sepsis	47/2007	NA	VAN
9	D		IV (2B)	CIP ERY FUS GEN	Blood	M 67	AML	Line-related sepsis	36/2007	TZP	FEP LZD VAN
10	D		IV (2B)	CIP ERY FUS GEN MUP	Hip tissue	F 88	Total hip replacement	Septic total hip replacement	63/2007	CFZ CIP TIM TMP	LZD TZP VAN
11	D	2	IV (2B)	CIP ERY FUS GEN	Blood	M 37	Metastatic carcinoma, Staphylococcus aureus sepsis	Continuing fever, IV line in situ	18/2007	FCX	GEN MEM TIM line removed
12	D		IV (2B)	CIP ERY FUS GEN MUP	Hip liquid	M 63	Metastatic renal cell carcinoma	Thigh collection postsurgery to femur for metastic lesion	12/2006	None	VAN
13a	D		IV (2B)	CIP FUS MUP	Bone	M 85	Diabetes mellitus foot infection	Necrotic bone from 3rd metatarsal head	O/P/2005	NA	CIP DOX
14	D		IV (2B)	CIP ERY FUS GEN MUP	Hip tissue	F 54	Hemiarthroplasty	Septic hemiarthroplasty	70/2006	AMC CRO TIM VAN	MEM TIM VAN
15	D		IV (2B)	CIP ERY FUS GEN MUP	Disc tissue	M 42	Laminectomy	Discitis postsurgery	24/2006	CFZ LEX	VAN
16a	D		IV (2B)	CIP ERY FUS GEN MUP	Bone	F 51	ALL, bone marrow transplant	GVHD, manibular mucormycosis, surgical debridement on date of positive culture; no other obvious source	106/2007	AMC CIP CRO FEP LZD MEM TZP VRC	AMB, no anti-staphylococcal therapy added
17	D		IV (2B)	CIP ERY FUS MUP TET	Blood	M 20	AML, stem cell transplant, severe cutaneous GVHD	Febrile illness	NA/2005	NA	NA
18	D		IV (2B)	CIP ERY FUS GEN MUP	Blood	M 22	ALL, blood stem cell transplant	Line-related sepsis	282/2007	CIP MEM TZP	LZD
19	H	2	III (3A&5)	CIP ERY FUS GEN RIF	Blood	F 39	Meliodosis	Line-related sepsis	25/2007	MEM	VAN
20	G		III (3A) Variant SCCmercury absent	ERY FUS GEN MUP	Vitreous fluid	F 77	NA	Postoperative endophthalmitis	10/2003	NA	OFX CAZ Intravitreal VAN
21a	G	2	III (3A&5)	CIP ERY GEN TET	Sclera	M 73	Lung transplant scedosporium eye infection	Enucleation of eye	1/2005	TRB VRC	TRB VRC
22	F		III (3A&5)	CIP ERY FUS GEN MUP RIF	Hip tissue	F 55	Rheumatoid arthritis	Infected total hip replacement	108/2007	FUS RIF	VAN
23	F	2	III (3A&5&4) Extra ccrA4B4	CIP ERY FUS GEN RIF	CSF	M 83	Subarachnoid hemorrhage	Ventriculitis	NA/2004	LZD MEM VAN	FEP VAN
24	F		III (3A&5&4) Extra ccrA4B4	CIP ERY FUS GEN RIF	CSF	M 49	Subarachnoid hemorrhage	VP shunt infection	32/2004	MEM VAN	VAN
25	F		Isolate not available	CIP ERY FUS GEN RIF	Leg wound	M 46	Trauma, splenic rupture	Extensive soft tissue infection requiring amputation of leg	61/2007	CIP CLI	CIP MEM VAN
26	J	23	IV (2B)	CIP ERY MUP RIF TET	CSF	M 67	Cerebral hemorrhage	Ventriculitis	105/2006	None	VAN (IV and intraventricular), MEM LZD
27	J		IV (2B)	CIP GEN MUP RIF	CSF	F 58	Subarachnoid hemorrhage	Ventriculitis	164/2004	MEM TZP VAN	VAN (IV and intraventricular)

^aIsolate of uncertain significance.

^bPerformed according to Thomas et al. Allelic profile of the seven housekeeping loci (arcC, aroE, gtr, mutS, pyrR, tpi, yqiL) (18a).

^cAMC, amoxicillin-clavulanic acid; AZM, azithromycin; CFZ, cefazolin; FCX, flucloxacillin; FEP, cefepime; CAZ, ceftazidime; CRO, ceftriaxone; LEX, cephalexin; CEF, cephalothin; CIP, ciprofloxacin; CLI, clindamycin; DOX, doxycycline; ERY, erythromycin; GEN, gentamicin: LZD, linezolid; MEM, meropenem; MUP, mupirocin; OFX, ofloxacin; TZP, piperacillin-tazobactam; RIF, rifampin; TET, tetracycline; TIM, ticarcillin-clavulanic acid; TMP, trimethoprim; SXT, trimethoprim-sulfamethoxazole; VAN, vancomycin; AMB, Amphotericin B; TRB, terbinafine; VRC, voriconazole; CSF, cerebrospinal fluid; GVHD, graft versus host disease; NA, none available; ALL, acute lymphoblastic leukemia; AML, acute myeloid leukemia; VP, ventriculoperitoneal; LOS, length of stay; F, female; M. male.

After initial identification and antimicrobial susceptibility testing, the isolates were stored at −70°C until further analysis. Isolates resistant to cefoxitin and at least three of the other antibiotics tested were defined as multidrug resistant (MDR). PFGE, staphylococcal cassette chromosome mec (SCCmec) typing, and multilocus sequence typing (MLST) were performed on the 27 isolates as previously described (6, 19). One isolate from each PFGE type was analyzed using MLST. One isolate was not available for SCCmec typing. In addition, the PFGE patterns of the 27 strains were compared with a previously described compilation of 277 MDRSE isolates collected between 2001 and 2008 at 11 hospitals in northern Europe (14, 19, 20).

DNA macrorestriction analysis of the 27 MDRSE isolates revealed six PFGE types (Fig. 1). The predominant PFGE type detected, PFGE type D, consisted of 16 isolates (57%). Four PFGE types (G, H, I, J) contained one or two isolates, and PFGE type F consisted of four isolates (Fig. 1). PFGE type D isolates were from patients in six different wards between 2005 and 2007 and accounted for 11 of the 12 blood culture isolates (92%) (Fig. 1). Multilocus sequence type 2 (ST2) included four different PFGE types (D, F, G, and H), representing 85% of isolates (Fig. 1). PFGE type D was indistinguishable from an MDRSE genotype previously described in two Swedish hospitals (19). SCCmec characterization, performed on 26 isolates, identified two SCCmec types, type III (27%) and type IV (73%). ST2 MDRSE isolates harbored both SCCmec types. Type IV SCCmec was found in all 16 PFGE type D isolates.

Fig 1.

Cluster analysis of the genetic similarity of 27 isolates of multidrug-resistant Staphylococcus epidermidis using pulsed-field gel electrophoresis (PFGE). The horizontal upper bar represents genetic similarity (percent). The dotted lines in the center represent digitalized transformation of the PFGE-DNA pattern. Source of culture, PFGE type, sequence type (ST), ward, and year of isolation are described in the columns to the right. ^a BMTU, bone marrow transplant unit.

The results document the occurrence and possible endemicity of one PFGE type which represented the majority of the examined S. epidermidis isolates. The prevailing genotype in the current study, PFGE type D, was identified among patients treated in six different wards over a 3-year period. Interestingly, all PFGE type D isolates harbored type IV SCCmec, which presently is found in the majority of methicillin-resistant Staphylococcus aureus (MRSA) strains of community origin (15). Furthermore, MLST results showed that four of the six identified genotypes (PFGE types D, F, G, and H) belonged to ST2. This genotype has been detected in strains isolated in as many as 25 different countries across the world (18) (http://sepidermidis.mlst.net). A limitation of this study is that the isolates were a convenience sample; hence, it is not possible to determine the burden of disease. A prospective study characterizing all S.epidermidis isolates causing health care-associated infections would be necessary to determine the prevalence of this clone.

In summary, this report demonstrates the occurrence, persistence, and potential spread of an MDR genotype of S. epidermidis causing health care-associated infections in an Australian teaching hospital. This genotype (ST2) accounted for 85% of isolates tested, was indistinguishable from an MDRSE genotype identified in European hospitals, and has been reported as the most widely disseminated health care-associated ST type. More studies are needed to increase our understanding of the mechanisms that contribute to the evolutionary success of this extremely versatile microorganism.