Rapid Detection of Methicillin Resistance in Coagulase-Negative Staphylococci by a Penicillin-Binding Protein 2a-Specific Latex Agglutination Test

Abstract

The detection of PBP 2a by the MRSA-Screen latex agglutination test with 201 clinical coagulase-negative staphylococci had an initial sensitivity of 98% and a high degree of specificity for Staphylococcus epidermidis strains compared to PCR for mecA. Determination of oxacillin MICs evaluated according to the new breakpoint (0.5 μg/ml) of the National Committee for Clinical Laboratory Standards exhibited an extremely low specificity for this population.

Coagulase-negative staphylococci (CoNS) are major nosocomial pathogens (10, 12, 18, 22, 24), and treatment of infections caused by CoNS is increasingly problematic due to the frequent occurrence of isolates resistant to multiple antibiotics (2).

The majority of clinical CoNS harbor the mecA gene, which encodes an additional penicillin-binding protein (PBP), PBP 2a, essential for expression of methicillin resistance (3). Phenotypic detection of methicillin resistance in CoNS is difficult due to the heterogeneous expression of mecA (6, 16, 21, 27). mecA detection by PCR is very sensitive but is not feasible for the busy clinical microbiology laboratory.

Thus, rapid, sensitive, and specific procedures for the detection of methicillin resistance in CoNS are urgently needed. A latex agglutination (LA) test (MRSA-Screen) detects PBP 2a by using latex particles sensitized with monoclonal antibodies specific for PBP 2a of Staphylococcus aureus (17). This test was evaluated for S. aureus with overall favorable results for sensitive and specific detection of methicillin-resistant S. aureus (MRSA) (5, 11, 15, 17, 25, 26).

(Part of this work will appear in the doctoral thesis of M. A. Horstkotte, Universitätsklinikum Hamburg-Eppendorf, Hamburg, Germany.)

CoNS strains (n = 197) were consecutively collected at the University Hospital Hamburg-Eppendorf from November 1997 to January 1998, including blood culture (n = 77), infected catheter (n = 91), wound (n = 15), and urine (n = 14) isolates. A single isolate per patient was included. S. epidermidis 1457, 1057, 9225, and RP62A were included as reference strains (14). S. aureus ATCC 29213 and ATCC 25923 were used for quality control. The strains were kept at −80°C and were subcultured twice onto Columbia blood-agar (CBA) plates before testing. Species identification was performed by using Gram stain morphology, the catalase test, the clumping factor test, and the ID 32 Staph system (bioMérieux, Marcy l'Etoile, France).

Oxacillin MICs were determined by the broth microdilution method, as recommended by the National Committee for Clinical Laboratory Standards (NCCLS) (9).

PCR detection of mecA was performed by using the conditions described previously (13, 20) with the primers mecAsense (181–5′-GAA ATG ACT GAA CGT CCG AT-3′-200) and mecAantisense (330–5′-GCG ATC AAT GTT ACC GTA GT-3′-311) (19). S. epidermidis 1457 and RP62A were included as negative and positive controls, respectively.

For the MRSA-Screen (Denka Seiken Co., Niigata, Japan), isolates were grown overnight on CBA. The manufacturer's instructions were essentially followed; however, a larger inoculum was used by suspending a loopful of bacteria to at least a McFarland no. 6 standard in extraction buffer. Agglutination results were read after 3 min.

A total of 201 isolates of CoNS (142 S. epidermidis, 15 S. haemolyticus, 10 S. hominis, 9 S. saprophyticus, 6 S. capitis, 4 S. lugdunensis, 4 S. warneri, 4 S. xylosus, 2 S. schleiferi, 2 S. cohnii, 1 S. chromogenes, 1 S. simulans, and 1 S. kloosii isolates) were tested. mecA was detected by PCR in 126 (62.7%) of all strains, of which 102 were S. epidermidis strains and 24 were non-S. epidermidis strains (Table 1). By the MRSA-Screen, 119 of 126 mecA-positive strains were LA positive. On initial testing 2 mecA-positive S. epidermidis strains were LA negative and 5 strains displayed weakly positive agglutination (Tables 1 and 2). Initially, 67 of 75 mecA-negative strains were LA negative, but 8 non-S. epidermidis strains displayed a positive (n = 1) or weakly positive (n = 7) agglutination result (Tables 1 and 2). When the weakly positive reactions were also counted as positive, PBP 2a was detected with sensitivities of 98.4, 98.0, and 100% and specificities of 89.3, 100, and 77.1% in all strains of CoNS, S. epidermidis strains only, and non-S. epidermidis strains, respectively.

TABLE 1.

Detection of methicillin resistance by latex agglutination of PBP 2a compared to mecA PCR

Isolate and mecA status	No. (%) of isolates tested	No. of isolates with the following MRSA-Screen resulta:
		Positive	Weakly positive	Negative
All CoNS	201 (100)
mecA positive	126 (62.7)	119 (124)	5 (0)	2 (2)
mecA negative	75 (37.3)	1 (1)	7 (1)	67 (73)
S. epidermidis	142 (70.6)
mecA positive	102 (71.8)	95 (100)	5 (0)	2 (2)
mecA negative	40 (28.2)	0 (0)	0 (0)	40 (40)
Non-S. epidermidis	59 (29.4)
mecA positive	24 (40.7)	24 (24)	0 (0)	0 (0)
mecA negative	35 (59.3)	1 (1)	7 (1)	27 (33)

^aEach value in parentheses is the number of strains with the given MRSA-Screen result upon retesting of weakly positive strains. 

TABLE 2.

Strains with discrepant results between mecA PCR and the PBP 2a latex agglutination test

Strain	Identification	mecA PCR result	Agglutination of PBP 2a		Oxacillin MIC (μg/ml)
			Initially	On repeat testing
66	S. epidermidis	Positive	Weakly positive	Positive	4
73	S. epidermidis	Positive	Negative	Positive	4
76	S. epidermidis	Positive	Weakly positive	Positive	512
85	S. epidermidis	Positive	Weakly positive	Positive	32
101	S. epidermidis	Positive	Weakly positive	Positive	4
105	S. epidermidis	Positive	Weakly positive	Positive	32
193	S. epidermidis	Positive	Negative	Positive	2
14	S. saprophyticus	Negative	Weakly positive	Negative	2
45	S. warneri	Negative	Weakly positive	Positive	0.5
65	S. capitis	Negative	Weakly positive	Negative	2
67	S. xylosus	Negative	Weakly positive	Negative	2
68	S. warneri	Negative	Weakly positive	Weakly positive	1
91	S. haemolyticus	Negative	Weakly positive	Negative	0.5
93	S. capitis	Negative	Weakly positive	Negative	1
180	S. chromogenes	Negative	Positive	Negative	0.5

The 12 weakly positive strains and the 3 strains with LA results discordant with the mecA PCR result were retested. All seven mecA-positive strains were LA positive on retesting. On retesting, six of eight mecA-negative strains exhibited a negative result by LA; however, two S. warneri strains still displayed false-positive LA results (Tables 1 and 2). All mecA-positive strains retested were S. epidermidis (Table 2). Among the mecA-negative strains retested, only non-S. epidermidis species were observed (Table 2). Retesting of weakly positive strains resulted in PBP 2a detection with sensitivities of 98.4, 98.0, and 100% and specificities of 97.3, 100, and 94.3% in all strains of CoNS, S. epidermidis strains only, and non-S. epidermidis strains, respectively.

MIC testing revealed for 125 of 126 mecA-positive strains oxacillin MICs of 4 μg/ml and for 1 S. epidermidis isolate an oxacillin MIC of 2 μg/ml. For only 9 of 75 mecA-negative strains were oxacillin MICs ≤0.25 μg/ml, but for 66 strains oxacillin MICs were between 0.5 and 2 μg/ml. Compared to mecA PCR, all mecA-positive strains of CoNS were found to be oxacillin resistant by use of the actual oxacillin breakpoint (0.5 μg/ml) of NCCLS (9). However, 66 of 75 mecA-negative strains of CoNS, for which oxacillin MICs ranged from 0.5 to 2 μg/ml, would have been falsely identified as oxacillin resistant, indicating that the actual breakpoint has a low degree of specificity.

In the present study detection of PBP 2a by MRSA-Screen was highly sensitive and specific and at least equivalent to other phenotypic techniques for the detection of methicillin resistance in CoNS (6, 16, 21, 27). Compared to the results of PCR, MRSA-Screen was superior to the standard broth microdilution assay when the present oxacillin breakpoint (0.5 μg/ml) suggested by NCCLS (9) was used. The latter method misidentified 66 of 75 mecA-negative strains of CoNS as oxacillin resistant. In contrast, an excellent specificity of the new breakpoint was reported for S. epidermidis, S. hominis, and S. haemolyticus, but the breakpoint was less accurate when it was applied to other species of CoNS (8). A lack of specificity of the new breakpoint could jeopardize the efforts directed at curtailing the overuse of glycopeptide antibiotics. When we evaluated our MIC data with the old oxacillin breakpoint (4.0 μg/ml), one mecA-positive S. epidermidis isolate would have been misclassified as susceptible and one mecA-negative S. kloosii isolate would have been misclassified as resistant.

The principal difficulties of performance of the MRSA-Screen with CoNS were reported to be regarding sensitivity (7, 15) or specificity (1, 28), or both (4, 23). A low initial sensitivity was reported, with satisfactory results obtained only after induction of PBP 2a synthesis with an oxacillin disk during overnight subcultivation (7). In contrast, all 60 mecA-positive strains of CoNS were reported to be LA positive after 3 min without induction (28). Apparently, the sensitivity of the MRSA-Screen depends on the amount of bacteria used in the inoculum or additional enhancement of PBP 2a expression. We prefer using a rather heavy inoculum, as oxacillin induction of PBP 2a requires subcultivation and could delay the results by 24 h. Additionally, extended agglutination times for increased sensitivity (7) can lead to false-positive results (28).

When CoNS were evaluated, weakly positive LA results occurred (7, 28). In our study, retesting of weakly positive strains led to positive test results for all of the mecA-positive strains and to negative results for all but two mecA-negative strains. For use in the clinical laboratory, it seems reasonable to retest a weakly positive isolate and to interpret a concordant LA result (positive or weakly positive on retesting) as positive. If the second LA test gives a discordant (negative) result, either the species diagnosis could help in the decision, as all false-positive results occurred with non-S. epidermidis strains, or an independent reference method could be used in parallel.

In our study two S. warneri strains displayed false-positive LA test results. False-positive LA test results were reported with S. lugdunensis, S. warneri, S. simulans, and S. hominis (1, 7, 23, 28).

Apparently, the MRSA-Screen performs favorably with clinical isolates of CoNS, especially the most frequently encountered species, S. epidermidis. False-positive results occur primarily with non-S. epidermidis strains. With a turnaround time of about 30 min, this assay could replace other phenotypic methods for determination of methicillin resistance in CoNS in the clinical microbiology laboratory.