Evaluation of the Amplidiag CarbaR+MCR Kit for Accurate Detection of Carbapenemase-Producing and Colistin-Resistant Bacteria

LETTER

As carbapenemase-producing Gram-negative bacilli (CP-GNB) coproducing MCR-1 are now emerging and, thus, compromising the use of last-resort antibiotics (carbapenems and colistin), there is an urgent need for accurate and fast diagnostic tests (1–4). We have previously shown that the Amplidiag CarbaR+VRE assay displayed a high sensitivity and specificity on colonies of CP-GNB but failed to detect GES carbapenemase-producing organisms (GES-CPOs), which represent 11% of the carbapenemase-producing Pseudomonas aeruginosa (CP-Pa) in France (5). Moreover, this assay was not validated on clinical samples, such as rectal swabs.

Here, we have evaluated an updated assay, the Amplidiag Carba-R+MCR assay, which is a multiplex nucleic acid-based in vitro diagnostic test intended for the detection of CP-GNB, including GES carbapenemase-producing organisms (GES-CPOs) and colistin resistance gene mcr-1/2 from rectal swabs and from cultured colonies. The assay was evaluated during a retrospective evaluation on colonies of 215 GNB with well-characterized resistance mechanisms, during a prospective study on 51 consecutive enterobacterial isolates with reduced susceptibility to carbapenems referred to the French National Reference Center, and on extracted DNA from 100 rectal swabs (including 40 positives) sent to the hygiene unit for CP-GNB screening. This assay was used as recommended by the manufacturer (Mobidiag, Paris, France).

The Amplidiag CarbaR+MCR efficiently detected the big 5 carbapenemase families (KPC, including novel variants tested, such as KPC-5, KPC-6, KPC-12, KPC-14, and KPC-28; NDM, including NDM-4, NDM-5, NDM-6, NDM-7, and NDM-9; VIM, including VIM-4, VIM-9, and VIM-19; IMP, including IMP-8, IMP-11, and IMP-13; OXA-48-like, including OXA-162, OXA-181, OXA-232, and OXA-244) and the main acquired carbapenem-hydrolyzing oxacillinases from Acinetobacter baumannii (OXA-23, OXA-24/OXA-40, OXA-58, and the overexpressed OXA-51-like β-lactamase) in a similar manner to the Amplidiag CarbaR+VRE assay (Table 1). The sensitivity for all carbapenemases in cultured colonies was 99.2%; only one GES-CPO (GES-2-producing P. aeruginosa) was not detected. Overall, the specificity of detection of GES carbapenemases was not optimal since GES noncarbapenemase variants were also detected but with lower threshold cycle (C_T) values. The three GES-7 producers have been falsely detected as carbapenemase producers. Indeed, the β-lactamase GES-7 does not host the mutation (Gly170Ser) conferring carbapenemase activity; its spectrum of hydrolysis is restricted to expended-spectrum cephalosporins (Table 1). However, C_T values at 23 to 27 were obtained for the extended-spectrum β-lactamase (ESBL) GES-7, unlike GES-CPOs (GES-5 and GES-6) that displayed lower C_T values of 12 to 17 (Table 1). Among the GES non-CPOs, the specificity was 38.5%, lowering the specificity of carbapenemase detection to 86.2% (Table 1). Plasmid-encoded colistin resistance genes mcr-1 and mcr-2 were perfectly detected (100% sensitivity; Table 1). As claimed by the manufacturer, MCR variants other than MCR-1 and -2 (here, MCR-3, -3.2, -4, and -5 [6] were not detected; 100% specificity; Table 1).

TABLE 1.

Global performances of the Amplidiag Carba+MCR kit on Enterobacteriaceae, P. aeruginosa, and Acinetobacter spp. colonies grown on MH agar isolates with decreased susceptibility to carbapenems or to colistin^a

Organism(s) (no. of isolates)	β-lactam- or colistin-resistance mechanism	Amplidiag Carba+MCR CT valuesb
		KPC, NDM, VIM, IMP, OXA-48-like, AcOXA	GES	MCR
Decreased susceptibility to carbapenems (n = 187)
Non-GES/Non-carbapenemase/Nontargeted carbapenemase producers (45)
Escherichia coli (2), E. cloacae (4), Enterobacter asburiae (1), Citrobacter freundii (1), Morganella morganii (1), Hafnia alvei (2), Serratia marcescens (2), Klebsiella oxytoca (1), Klebsiella pneumoniae (3), P. aeruginosa (17), A. baumannii (11)	↗↗↗ Casec , ↗↗↗ Pase, CTX-M-15, TEM, SHV, OXA-163, OXA-405, PER-1, VEB-1, OXA-32, ↗↗↗ Case, OprD deficiency, Efflux, IMI, NmcA, GIM, AIM, SPM, DIM, OXA-198, None, PER-1, VEB-1, SCO-1, RTG-4, OXA-10, OXA-21, OXA-69, SIM, OXA-143, OXA-253	-d	-	-
Non-GES other carbapenemase producers (113)
Enterobacteriaceae (53), P. aeruginosa (30), A. baumannii (30)	KPC, NDM, VIM, IMP, OXA-48-like, ISAba1-OXA-51, OXA23, OXA-40, OXA-58	10–20	-	-
GES non-CPO (13)
K. pneumoniae (1)	GES-1	-	-	-
A. baumannii (2)	GES-11, GES-12 + OXA-51+ISAba1	14	-	-
P. aeruginosa (2)	GES-1, GES-9	-	-	-
Citrobacter amalonaticus (1), E. cloacae (1), K. oxytoca (1)	GES-7	-	23–27	-
A. baumannii (5)	GES-11/GES-12, + OXA-23, + ISAba1-OXA-51	11–16	26–32	-
GES CPO (16)
P. aeruginosa (1)	GES-2	-	-	-
E. cloacae (5), K. pneumoniae (2), Citrobacter braakii (1), Citrobacter youngae (1)	GES-5, GES-6	-	12–17	-
P. aeruginosa (2)	GES-5	-	14	-
A. baumannii (4)	GES-14	-	14–17	-
Decreased susceptibility to colistin (n = 28)
Non-/Non-targeted-MCR producers (15)
E. coli (4), K. pneumoniae (1), Salmonella spp. (4)	ΔpmrBe , ΔmgrB, ND	-	-	-
E. coli (4), Salmonella spp. (2)	MCR-3, MCR-3.2, MCR-4, MCR-5	-	-	-
Targeted MCR producers (13)
E. coli (4), Salmonella spp. (3), K. pneumoniae (3)	MCR-1, MCR-2	-	-	12–19
E. coli (3)	MCR-1 + OXA-48, + NDM-1	14–16	-	12–14

^aSensitivity for carbapenemases, 99.2 % (95% confidence interval [CI], 95.1%–100%); sensitivity for GES-CPO (15/16), 93.7%; for mcr-1/-2, 100% (95% CI, 71.6%–100%); specificity for carbapenemases, 86.2% (95% CI, 74.1%–93.4%); specificity among GES non-CPO (5/13), 38.5 %; for mcr-1/-2, 100% (95% CI, 74.6%–100%).

^bNumber of isolates tested. Dark gray boxes represent discrepant results. C_T values were rounded up.

^c↗↗↗ Case, abbreviation for overexpressed cephalosporinase.

^dNo amplification.

^eΔ, mutation or deletion; ND, not determined.

DNA was extracted from rectal swabs, and the sensitivity was 92.5%; two NDM producers and one OXA-48 producer were not detected by the Amplidiag CarbaR+MCR. These 3 samples contained low concentrations of bacteria, requiring an overnight enrichment step in brain heart infusion medium supplemented with 0.5 µg/ml ertapenem for detection of two of them (Table 2). Conversely, one positive result was obtained with an AcOXA gene in a sample containing a cultivable OXA-48 producing Enterobacter cloacae isolate. In-house PCR confirmed the presence of the bla_OXA-23 gene in the DNA extract, despite the absence of Acinetobacter spp. in the culture (7, 8). Accordingly, the global performances of the Amplidiag Carba-R+MCR assay were high (92 to 100% sensitivity and 86 to 100% specificity). Moreover, it can provide a result not only from colonies growing on Mueller-Hinton agar (MHA) or on selective screening medium but also from DNA extracted from clinical rectal swabs in less than three hours (1 h of DNA extraction and less than 2 h for PCR). The Amplidiag Carba-R+MCR assay is well adapted to the French epidemiology of CP-GNB, as it is able to detect 99.57% of CPEs (missing 13 IMI producers and one FRI-1 producer), 100% of CP A. baumannii (CP-Abs), and now 100% of CP-Pas (including the 11% of GES variants that were missed by the previous Carba-R+VRE assay) isolated over the period from 2012 to 2016 in France (9).

TABLE 2.

Global performances of the Amplidiag Carba+MCR kit on DNA extracted from clinical rectal swabs^a

Culture on ChromID Carba SMART medium (no. of isolates tested)		Carbapenemasea	Amplidiag Carba+MCR CT valuesb
Direct inoculation from the rectal swab	After enrichment in 0.5 μg/ml ertapenem containing BHI		KPC	NDM	VIM	IMP	OXA-48-like	AcOXA	MCR	GES
Negative (60)	Negative (60)	-	-	-	-	-	-	-	-	-
E. coli + K. pneumoniae (2)	Not performed	KPC-2	23–24	-	-	-	-	-	-	-
Negative	K. pneumoniae	KPC-3	25	-	-	-	-	-	-	-
Negative (GeneXpert result :VIM)	Negativec	VIM-type	-	-	33	-	-	-	-	-
E. coli	Not performed	NDM-5	-	31	-	-	-	-	-	-
K. pneumoniae (1), A. baumannii (1)	Not performed	NDM-1	-	32	-	-	-	-	-	-
K. pneumoniae + A. baumannii (3)	Not performed	NDM- 1 + OXA-23	-	23–29	-	-	-	25–27	-	-
Negative	A. baumannii	NDM-1	-	33	-	-	-	- (38)d	-	-
Negative	A. baumannii	NDM-1	-	- (39)	-	-	-	-	-	-
Negative	A. baumannii	NDM-1	-	-	-	-	-	-	-	-
A. baumannii (2)	Not performed	OXA-23	-	-	-	-	-	15–22	-	-
C. freundii (1), E. cloacae + K. varicola (2), E. cloacae (4)	Not performed	OXA-48	-	-	-	-	18–25	-	-	-
E. aerogenes	Not performed	OXA-48	-	-	-	-	34	-	-	-
E. coli (6), E. coli + C. freundii (1)	Not performed	OXA-48	-	-	-	-	19–29	-	-	-
K. pneumoniae (3), E. coli + K. pneumoniae (5)	Not performed	OXA-48	-	-	-	-	19–30	-	-	-
E. cloacaee	Not performed	OXA-48 + OXA-23	-	-	-	-	23	27	-	-
K. oxytocaf	Not performed	OXA-48	-	-	-	-	- (37)	-	-	-

^an = 100. Sensitivity, 92.5% (95% CI, 78.5%–98.0%); specificity, 100% (95% CI, 92.5%–100%). Positive predictive value, 100% (95% CI, 88.3%–100%). Negative predictive value, 95.2% (95% CI, 85.8%–98.8%). Dark gray boxes represent discrepant results. -, negative result.

^b-, no amplification.

^cNo growth was detected on ChromID Carba SMART after plating 100 μl of an overnight culture of the ESwab in brain heart infusion supplemented with 0.5 μg/ml ertapenem.

^dThe negativity cutoff value being fixed at C_T ≥ 35 by the manufacturer, this sample is counted as negative.

^eAlthough no culture of A. baumannii was obtained with this sample, an in-house PCR confirmed the presence of the bla_OXA-23 gene.

^fOnly 2 CFU were obtained on ChromID Carba Smart.