Failure of mecA/mecC PCR Testing to Accurately Predict Oxacillin Resistance in a Patient with Staphylococcus aureus Infective Endocarditis

ABSTRACT

Genotypic testing for mecA/mecC is heavily relied upon for rapid optimization of antimicrobial therapy in infections due to Staphylococcus aureus. Little is known regarding optimal reporting and/or therapy for patients demonstrating lack of genotypic evidence of mecA or mecC but phenotypic oxacillin resistance. We report a case of a 77-year-old patient with S. aureus bloodstream infection and infective endocarditis with discordance between mecA/mecC genotypic results and phenotypic susceptibility testing.

CASE PRESENTATION

A 77-year-old male presented to the outpatient setting with a 1-week history of fever, chills, nausea, and 1 month of generalized weakness. Past medical history was notable for bioprosthetic aortic valve replacement (19 months prior) complicated by sternal wound abscess with methicillin-susceptible Staphylococcus aureus (MSSA) bacteremia occurring 8 weeks after valve replacement. A transesophageal echocardiogram (TEE), at the time of the postoperative MSSA bacteremia, did not demonstrate infective endocarditis (IE); however, he received 6 weeks of cefazolin followed by suppression with cefadroxil.

Upon outpatient presentation, he noted onset of the aforementioned symptoms coincided with a decrease in suppressive cefadroxil dose from 1,000 mg to 500 mg orally twice daily due to gastrointestinal intolerance. Outpatient blood cultures (hospital day −2 [HD −2]) were collected and were positive for Gram-positive cocci at 16 h in two sets. PCR testing, using the BioFire blood culture identification 2 panel (BCID2; bioMérieux, Durham, NC), detected S. aureus without mecA/mecC. The patient was referred to the emergency department (ED) at a separate outside institution, admitted, and initiated on empirical, renally adjusted (creatinine clearance, 32 mL/min) vancomycin (15 mg/kg intravenous, every 24 h [i.v. q24h]). On HD 0, initial inpatient blood cultures were drawn and were positive for S. aureus at 18 h in 2/2 sets without mecA/mecC detection by BioFire BCID2, and vancomycin was deescalated to cefazolin (2 g i.v. q8h) (Fig. 1). On HD3, TEE revealed IE of the prosthetic aortic valve and native mitral valve. Blood cultures from HD3 grew S. aureus in 2/2 sets after 48 h. Given TEE findings, cefazolin was changed to oxacillin (2 g i.v. q4h) and synergistic gentamicin (1 mg/kg i.v. q12h), with plans to add rifampin once blood cultures cleared and source control was achieved.

FIG 1.

Timeline of microbiological, treatment, and therapeutic events.

On HD4, antimicrobial susceptibility test (AST) results from blood cultures on HD −2 collected at the outpatient encounter (laboratory 1) demonstrated an oxacillin MIC of ≥4 μg/mL with cefoxitin susceptibility determined by Vitek2 AST-GP67 (bioMérieux; performed according to the manufacturer’s instructions for use [IFU]; breakpoints were based on Clinical and Laboratory Standards Institute [CLSI] M100) (Table 1) (1). A lateral flow assay (Abbott Clearview PBP2a SA; performed according to IFU) for the detection of penicillin binding protein 2a (PBP2a) was negative. Due to the oxacillin-resistant phenotypic result, the isolate was reported as MRSA despite the negative genotypic results; however, the MIC for oxacillin was suppressed, with only the interpretive category (i.e., resistant) reported. Given the different institutional affiliations of laboratory 1 and laboratory 2, it is likely that this update to AST at laboratory 1 was not communicated to the patient’s current care team, as evidenced by the team’s continuation of the oxacillin. Blood cultures were repeated on HD4 and were negative. Repeat blood cultures were obtained daily through HD8 and all remained negative.

TABLE 1.

Blood culture testing and reporting^a

Testing	Outside microbiology (laboratory 1)	Initial inpatient microbiology (laboratory 2)	Tertiary-care reference microbiology (laboratory 3)
Blood culture collection	HD −2	HD 0	HD 3
Time to positivity	16 h	18 h	48 h
mecA/mecC result	Negative (BCID2)	Negative (BCID2)	Negative (LDT)
PBP2a assay	Negative	Negative	NA
Cefoxitin	S (VITEK2)	S (VITEK2)	R (disk diffusion)
AST platform or methodb	VITEK2	VITEK2	Agar dilution
Antimicrobial susceptibility testing result (MIC)
Clindamycin	S (≤0.25 μg/mL)	S (≤0.12 μg/mL)	S (≤0.5 μg/mL)
Daptomycin		S (0.5 μg/mL)	S (0.5 μg/mL)c
Erythromycin	S (≤0.25 mcg/mL)	S (≤0.25 μg/mL)
Gentamicin		S (≤0.5 μg/mL)
Oxacillin	R	S	R (>2 μg/mL)
Rifampin	S (≤0.5 μg/mL)	S (≤0.5 μg/mL)	S (≤0.5 μg/mL)
Tetracycline	S (≤1 μg/mL)	S (≤1 μg/mL)
TMP/SMX		S (≤10 μg/mL)	S (≤0.5/9.5 μg/mL)
Vancomycin	S (2 μg/mL)	S (1 μg/mL)	S (1 μg/mL)
Ceftaroline			S (0.5 μg/mL)
Linezolid	S (≤0.5 μg/mL)		S (≤2 μg/mL)
Microbiology laboratory interpretation and reporting
Result	MRSA	MSSA	MRSA

^aAbbreviations: S, susceptible; R, resistant; LDT, laboratory-developed test (developed for mecA only); TMP/SMX, trimethoprim-sulfamethoxazole; NA, not applicable.

^bLaboratory 1 used VITEK2 with method AST-GP67; laboratory 2 used VITEK2 with method AST-GP75; laboratory 3 used the CLSI agar dilution reference method (2).

^cDaptomycin was tested by Etest (bioMérieux) at laboratory 3.

The patient transferred to our tertiary care center on HD5 in anticipation of cardiac surgery. The same day, AST results from the first set of inpatient blood cultures (HD0) were found to be oxacillin resistant (MIC, ≥4 μg/mL; Vitek2 AST-GP75; performed according to IFU; breakpoints applied using CLSI M100), cefoxitin susceptible, and negative for PBP2a (Abbot Clearview PGP2a SA) (1). The initial inpatient hospital laboratory (laboratory 2) reported the organism as MSSA and released the “susceptible” interpretive category for oxacillin with the MIC suppressed. As such, the patient was continued on therapy with oxacillin. On HD7, the tertiary care team noted discrepancies between both the phenotypic AST and mecA/mecC results and subsequent reporting by two microbiology laboratories. Thus, HD3 blood isolates (one from each set) were referred to our laboratory (laboratory 3) for repeat AST and further testing. On HD9 the patient underwent aortic root reconstruction and bioprosthetic replacement of the aortic and mitral valves. Intraoperative cultures from the aortic valve, mediastinum, and sternum were negative. Due to rising serum creatinine on HD9, the patient was transitioned from oxacillin to cefazolin.

On HD11, our institution’s AST results (agar dilution performed according to CLSI M07) on both isolates referred from HD3 confirmed an oxacillin MIC of >2 μg/mL (breakpoints applied using CLSI M100), and a laboratory-developed PCR assay verified the absence of mecA (1–3). β-Lactamase activity was assessed using a nitrocefin disk (Cefinase, BD BBL, Franklin Lakes, NJ) and penicillin zone edge test, and both isolates were positive for both assays. Cefoxitin was resistant by disk diffusion at exactly the clinical breakpoint (21 mm; CLSI resistant breakpoint, ≤21 mm) (1). Given these findings, our microbiology laboratory reported the organism as MRSA.

CHALLENGE QUESTION

Given the mismatch between genotypic mecA/mecC and phenotypic AST results, as well as clearance of blood cultures on MSSA-directed therapy, which definitive therapy is most appropriate for this patient?

• A.Cefazolin, given mecA/mecC negativity and blood culture clearance on MSSA-directed therapy
• B.Oxacillin, given mecA/mecC negativity and blood culture clearance on MSSA-directed therapy
• C.Vancomycin, given phenotypic AST result of oxacillin MIC of >2 mcg/mL
• D.Daptomycin, given phenotypic AST result of oxacillin MIC of >2 mcg/mL

TREATMENT AND OUTCOME

Following in-house confirmation of non-mecA/mecC-mediated MRSA phenotype, and to avoid nephrotoxic antimicrobials, the patient was transitioned to daptomycin (8 mg/kg i.v. q24h) and rifampin (300 mg orally [p.o.] q8h) on HD11. This regimen was continued for 6 weeks following valve replacement. He was then transitioned to lifelong suppressive doxycycline (100 mg p.o. q12h). He was discharged to a skilled nursing facility on HD88. He remains on suppressive doxycycline without further infectious complications.

The predominant methicillin resistance mechanism in S. aureus is expression of PBP2a encoded by mecA or mecC genes (4, 5). The rapidity and reliability of testing for mecA and mecC genes lends to early optimization of therapy. Implementation of deescalation based on mecA/mecC testing has been shown to reduce durations of empirical vancomycin and hospital length of stay (6). Currently, CLSI recommends that isolates that test positive for mecA or PBP2a or are resistant by any of the recommended phenotypic methods should be reported as methicillin (oxacillin) resistant (1). In this case, laboratory 1 reported the isolate as MRSA, while laboratory 2 reported it as MSSA (despite a resistant oxacillin MIC of ≥4 μg/mL); both labs suppressed oxacillin MIC values from Vitek2 and only reported interpretations. Laboratory 3 reported the isolate as MRSA, based on the oxacillin MIC and cefoxitin resistance. These inconsistencies demonstrated a lack of standardization in reporting isolates such as these, emphasizing a need for clearer guidelines. It should be emphasized that cefoxitin and oxacillin susceptibility testing are equally important, as one method may detect resistant isolates that the other cannot, and either a phenotypic or genotypic resistant result would require reporting isolates as MRSA. Discussions within the literature have attempted to address such issues (7).

Borderline-oxacillin resistant S. aureus (BORSA) phenotype may be one explanation for this discrepancy. BORSA isolates are characterized by elevated oxacillin MICs (i.e., 1 to 8 μg/mL) without detectable mecA, mecC, or PBP2a. The precise mechanism behind BORSA is unclear; however, these isolates often contain blaZ, resulting in hyperproduction of a β-lactamase (8). Such an organism has also previously been implicated in a neonatal intensive care unit outbreak (9). Alternatively, this organism could represent a modified S. aureus (MODSA) phenotype, in which mutations in non mec-type genes (e.g., pbp, gdpP, and yibH) result in increased oxacillin MICs (10). Selective pressure from β-lactams is a possible mechanism for MODSA mutation selection (10). As such, it is plausible that postoperative suppression with cefadroxil played a role in the selection of the mec-independent oxacillin-resistant phenotype observed in this case. It seems unusual that blood cultures were cleared on β-lactam therapy despite the organism’s oxacillin resistance. However, some reports of MODSA indicate mutation acquisition may come at a high fitness cost (11). It is possible that reduced bacterial fitness, in conjunction with high-dose, intravenous β-lactam therapy and limited vancomycin exposure, impacted blood culture clearance in our patient.

Limited data are available describing optimal treatment of S. aureus bacteremia in the context of mec-independent oxacillin resistance. One retrospective cohort study showed that there was no difference in acute treatment outcomes between patients with an oxacillin MIC of <1 μg/mL and an oxacillin MIC of 1 to 2 μg/mL when treated with either cefazolin or antistaphylococcal penicillins (12). As such, those authors concluded that the oxacillin MIC should not dictate β-lactam selection in S. aureus isolates that fall within the qualitative oxacillin susceptibility breakpoint as defined by CLSI. However, organisms with MICs of >2 μg/mL, as seen in this case, were notably not included in this analysis. Another case report described a patient with S. aureus endocarditis with an oxacillin MIC of 12 μg/mL and which was found to be mecA and PBP2a negative (13). Therapeutic success was reported in this case after transition from cloxacillin to vancomycin. However, use of vancomycin in less resistant isolates may be undesirable given previously observed increases in risk of mortality in patients with MSSA bacteremia treated with vancomycin compared to that for patients treated with cefazolin or antistaphylococcal penicillins (14). Daptomycin has been demonstrated to be an effective alternative to vancomycin in the management of MRSA bacteremia with or without endocarditis (15, 16). In this case, given previous acute kidney injury, our patient was transitioned to therapy with daptomycin (Answer D) once an oxacillin MIC of >2 μg/mL was confirmed.

There is no routine diagnostic method to detect BORSA or MODSA strains in clinical laboratories. One approach for detecting BORSA isolates is phenotypic oxacillin resistance and cefoxitin susceptibility, leading to the assumption that these isolates are β-lactamase producers (17). However, as seen in this case, cefoxitin may be resistant in the absence of mecA or mecC, leading to further confusion regarding discrepant results. Standardization of reporting of discrepant results between genotypic mecA/mecC testing and phenotypic cefoxitin and oxacillin testing is needed.

Given reliance on genotypic identification for directed antistaphylococcal therapy, an increased prevalence of non-mecA/mecC-mediated MRSA would have dramatic implications. In this case, utilization of mecA/mecC PCR for deescalation may have resulted in up to 10 days of potentially suboptimal therapy, including inappropriate perioperative antimicrobials. Nonetheless, blood cultures remained clear after HD4 and intraoperative cultures were negative despite treatment with cefazolin and oxacillin.

This case further highlights the need for consensus on reporting S. aureus isolates that are genotypically mecA/mecC negative but phenotypically oxacillin resistant and cefoxitin susceptible, as well as further investigation into infection prevention implications and optimal antimicrobial therapy for these isolates.