Multilaboratory Comparison of Omadacycline MIC Test Strip to Broth Microdilution MIC against Gram-Negative, Gram-Positive, and Fastidious Bacteria

ABSTRACT

The performance of the Liofilchem omadacycline MIC Test Strip (MTS) was evaluated in a multisite study. Three testing sites collected/tested clinical isolates and one site tested challenge isolates that totaled 175 S. aureus, 70 S. lugdunensis, 121 E. faecalis, 100 E. faecium, 578 Enterobacterales, 142 Haemophilus spp., 181 S. pneumoniae, 45 S. anginosus group, 35 S. pyogenes,and 20 S. agalactiae. MIC testing was performed by CLSI broth microdilution (BMD) and MTS. Fastidious isolates testing included BMD and MTS testing with both CLSI and EUCAST Mueller-Hinton Fastidious (MH-F). In addition, each site performed reproducibility for nonfastidious and fastidious isolates and QC by MTS and BMD. All BMD and MTS results for the QC strains were within expected ranges, with exception of one MTS HTM result for H. influenzae ATCC 49247. Among reproducibility isolates, omadacycline MTS results were within one dilution of the modal MIC for 95.2% of nonfastidious Gram-positive, 100% of Gram-negative, 99.3% and 98.5% of fastidious isolates tested on CLSI and EUCAST media, respectively. MTS results for all study isolates were within one doubling dilution of the CLSI BMD MIC for 98.9% of S. aureus, 100% of S. lugdunensis, 98.3% of E. faecalis, 100% of E. faecium, and 99.6% of Enterobacterales. Essential agreement rates for CLSI and EUCAST MH-F agar compared to CLSI BMD were 98.2% and 98.2%, for H. influenzae, 91.1% and 73.6%, for S. pneumoniae and 100% and 85–91.7% for other streptococcus species, respectively. Based on CLSI media, all categorical errors were minor errors and categorical agreement rates were >90% with exception of C. freundii, S. lugdunensis, E. faecalis, S. anginosus and S. constellatus.

INTRODUCTION

Omadacycline is a semisynthetic broad spectrum oral and intravenous antibiotic, which is a tetracycline antibiotic in the aminomethylcycline subclass. It was approved by the US FDA in October 2018 for the treatment of community-acquired bacterial pneumonia (CABP) caused by Streptococcus pneumoniae, Staphylococcus aureus (methicillin-susceptible isolates), Haemophilus influenzae, Haemophilus parainfluenzae, Klebsiella pneumoniae, Legionella pneumophila, Mycoplasma pneumoniae, and Chlamydophila pneumoniae and for the treatment of acute skin and skin structure infections caused by S. aureus (methicillin-susceptible and-resistant isolates), Staphylococcus lugdunensis, Streptococcus pyogenes, Streptococcus anginosus group (includes S. anginosus, S. intermedius, and S. constellatus), Enterococcus faecalis, Enterobacter cloacae, and Klebsiella pneumoniae. Its wide spectrum of in vitro activity comprises Enterobacterales (including ESBL-producing strains), Gram-positive bacteria (including methicillin-resistant S. aureus, vancomycin-resistant enterococcus, multidrug resistant S. pneumoniae and other streptococci), Haemophilus spp, some anaerobic bacteria (Bacteroides spp., Prevotella spp., Clostridioides difficile, and Clostridium perfringens), and atypical pathogens (including Legionella pneumophila, Mycoplasma pneumoniae, and Chlamydia spp.) (1–4). Omadacycline has activity against bacterial isolates that express the two primary mechanisms of tetracycline resistance (efflux and ribosomal protection). Similar to other tetracyclines, omadacycline is not active in vitro against Morganella spp., Proteus spp., and Providencia spp.

Liofilchem (Roseto degli Abruzzi, Italy) provides MIC test strips for in vitro susceptibility testing for a variety of antimicrobial agents. The Liofilchem MIC Test Strips (MTS) is a quantitative agar-based diffusion assay for determining the MIC. The strips are made of a special high-quality paper impregnated with a predefined concentration gradient of antibiotic across 15 2-fold dilutions like those of conventional broth and agar dilution MIC methods. The primary objective of this multi-lab study was to compare the omadacycline MTS to CLSI reference broth microdilution (CLSI BMD) for submission to FDA for device clearance. A secondary objective was to compare MTS and BMD results generated with EUCAST media to those generated with CLSI media for fastidious organisms.

In general, for testing of fastidious bacteria (for the purpose of this study, includes streptococcus and Haemophilus spp.), the broth for MIC and the agar for disk diffusion testing differs between Clinical and Laboratory Standards Institute (CLSI) and European Committee of Antimicrobial Susceptibility Testing (EUCAST) methods. To address harmonization efforts between these two groups, both CLSI and EUCAST broth and agar medias were tested in this study. MTS agar medium is typically based on the recommended agar for disk diffusion. The data submitted to the U.S. FDA for MTS clearance was based on CLSI recommended broth and agar. CLSI recommends the use of cation adjusted Mueller-Hinton broth + 5% lysed horse blood (CAMHB-LHB) for streptococci and Haemophilus Test Media (HTM) broth for Haemophilus spp. testing (5). EUCAST recommends the use of CAMHB + 5% LHB + 20 mg/liter β-NAD hydrate (MH-F broth) for both streptococci and Haemophilus spp. testing (6). The media recommended for disk diffusion testing by CLSI is Mueller-Hinton agar (MHA) + 5% sheep blood (MHA-SB) for streptococci and HTM agar for Haemophilus testing (7); EUCAST recommended medium is MHA + 5% defibrinated horse blood + 20 mg/liter β-NAD hydrate (MH-F agar) for streptococci and Haemophilus spp. (8). CLSI has recently validated use of MH-F agar for disk diffusion testing of S. pneumoniae and studies are on-going for similar validation of Haemophilus spp. (9).

MATERIALS AND METHODS

The overall study design and analysis was in accordance with the FDA, Center for Device and Radiological Health (CDRH), antimicrobial susceptibility methods guidance document (10). Non-fastidious and fastidious clinical isolates were collected and tested at three sites: (i) Arcispedale Santa Maria Nuova (Reggio Emilia, Italy), (ii) Wake Forest Baptist Medical Center (Winston-Salem, NC), and (iii) University of Rochester Medical Center (Rochester, NY). Laboratory Specialists, Inc. (Westlake, OH) was a fourth site that performed testing of challenge isolates. Gram-negative, Gram-positive, and fastidious isolates were collected within 6 months of testing for 65.7%, 79.7% and 64.8% of all clinical isolates, respectively.

The reference MIC method was performed with frozen plates according to the CLSI broth microdilution (BMD) guidelines and additionally with EUCAST MH-F media for fastidious isolates (5, 6). For nonfastidious isolate testing, frozen BMD plates (ThermoFisher, Oakwood Village, OH) were prepared using Difco-brand cation adjusted Mueller-Hinton broth (CAMHB; BD, Sparks, MD). For fastidious isolate testing, frozen BMD panels (LSI, Westlake, OH) for streptococcus were prepared using CAMHB-LHB and MH-F broth and for Haemophilus spp. were prepared using HTM broth (Remel, Lenexa, KS) and MH-F broth. Although only the CLSI media results were used for the FDA submission, the testing of both CLSI and EUCAST recommended broth for the BMD method allows for comparison of the MTS result for both CLSI and EUCAST agar media to both of the BMD media. Colony counts were performed from the CLSI broth positive growth control well of the BMD panel for at least 10% of isolates tested.

The MIC test strip (MTS) was tested according to manufacturer’s instructions. For nonfastidious isolate testing, each of the three testing sites utilized BBL brand MHA (Site 1: BD, Heidelberg, Germany and Site 2, 3, and 4: BD, Sparks, MD) and plates were incubated at 35 ± 2°C for 16–20 h in ambient atmosphere. Fastidious isolates were tested with MHA-SB for streptococci and HTM for Haemophilus spp. (Site 1: Liofilchem, Roseto degli Abruzzi, Italy; Site 2: Remel, Lenexa, KS; Site 3: BD, Sparks, MD; Site 4: MHA/SB-BD, Sparks, MD and HTM – Remel, Lenexa, KS) and plates were incubated at 35 ± 2°C for 20–24 h in 5% CO₂. In addition, all fastidious isolates were tested by MTS with MH-F agar (Site 1: Liofilchem, Roseto degli Abruzzi, Italy, prepared plates; Sites 2–4: LSI prepared plates; both batches of prepared plates were made with the same lot of BD MHA powder [Sparks, MD]).

Another set of isolates tested by both BMD and MTS methods were defined as “challenge isolates” and included a number of omadacycline non-susceptible isolates or isolates with elevated omadacycline MIC results (66 Enterobacterales, 31 S. aureus, 9 E. faecalis, 6× H. influenzae, and 5 S. pneumoniae). The total number of clinical isolates and challenge isolates tested for Enterobacterales were 496 and 82, for Gram-positive isolates were 474 and 116, and for fastidious isolates were 351 and 72, respectively (numbers by species are shown in Table 3). For verification of reproducibility, Gram-positive, Gram-negative, and fastidious sets of a minimum of 10 isolates each (representing clinically indicated and prevalent species and range of omadacycline MIC results) were sent to each of the three sites, and each site tested them in triplicate on 3 days by omadacycline MTS. Each of the three sites also tested 20 replicates of 5 QC strains (E. coli ATCC 25922, S. aureus ATCC 29213, E. faecalis ATCC 29212, H. influenzae ATCC 49247, and S. pneumoniae ATCC 49619) by BMD and MTS methods and results were compared to the CLSI expected ranges (11).

TABLE 3.

Summary TABLE of omadacycline MTS MIC compared to CLSI BMD MIC by bacterial species for combined clinical and challenge isolates

Bacterial species	Total tested	No. of EA	%EA	Total EVAL	No. of EA of EVAL	%EA of EVAL	#CA	%CA	No. of R	#VME	#ME
C. freundii	17	17	100	17	17	100	14	82.4	0	NA	0
C. koseri	17	17	100	16	16	100	16	94.1	1	0	0
E. cloacae	170	169	99.4	153	152	99.3	162	95.3	24	0	0
E. coli	140	139	99.3	135	134	99.3	137	97.9	6	0	0
K. aerogenes	35	35	100	34	34	100	35	100	2	0	0
K. oxytoca	20	20	100	19	19	100	20	100	1	0	0
K. pneumoniae	179	179	100	155	155	100	165	92.2	31	0	0
S. aureus a	175	173	98.9	174	172	98.9	168	96.0	31	0	0
S. aureus b	175	173	98.9	174	172	98.9	169	96.6	34	0	0
S. epidermidis a	40	40	100	40	40	100	NA	NA	NA	NA	NA
S. haemolyticus	26	26	100	26	26	100	NA	NA	NA	NA	NA
S. lugdunensis	70	70	100	70	70	100	61	87.1	1	0	0
S. saprophyticus	24	24	100	24	24	100	NA	NA	NA	NA	NA
E. avium	13	13	100	13	13	100	NA	NA	NA	NA	NA
E. faecalis	121	119	98.3	121	119	98.3	102	84.3	11	0	0
E. faecium	100	100	100	100	100	100	NA	NA	NA	NA	NA
E. gallinarum	21	20	95.2	21	20	95.2	NA	NA	NA	NA	NA
CLSI media: HTM and MHA/SB
H. influenzae	112	110	98.2	112	110	98.2	109	97.3	2	0	0
H. parainfluenzae	30	30	100	30	30	100	28	93.3	0	0NA	0
S. agalactiae	20	20	100	20	20	100	NA	NA	NA	NA	NA
S. anginosus	27	27	100	27	27	100	21	77.8	1	0	0
S. constellatus	18	18	100	18	18	100	14	77.8	0	NA	0
S. pneumoniae	181	166	91.7	181	166	91.7	175	96.7	4	0	0
S. pyogenes	35	35	100	35	35	100	32	91.4	1	0	0
EUCAST media: MH-F
H. influenzae	112	110	98.2	108	110	98.2	109	97.3	2	0	0
H. parainfluenzae	30	30	100	30	30	100	30	100	0	NA	0
S. agalactiae	17	20	85.0	17	20	85.0	NA	NA	NA	NA	NA
S. anginosus	27	23	85.2	27	23	85.2	19	70.4	1	0	0
S. constellatus	12	11	91.7	12	11	91.7	10	83.3	0	NA	0
S. pneumoniae	178	131	73.6	178	131	73.6	168	94.4	4	0	0
S. pyogenes	32	35	91.4	32	35	91.4	28	80	1	0	0

EA, essential agreement; EVAL, evaluable; CA, category agreement, R, resistant; VME, very major error; ME, major error; NA, not applicable (no breakpoints); HTM, Haemophilus test medium; MHA/SB, Mueller-Hinton agar + 5% Sheep Blood; MH-F, Mueller-Hinton, fastidious medium.

^aBased on FDA ABSSSI breakpoints.

^bBased on FDA CABP breakpoints.

Essential agreement (EA) rates were calculated based on the percentage of MTS results that were within ± one dilution of the BMD results. Categorical agreements and error rates were based on the FDA omadacycline breakpoints (https://www.fda.gov/drugs/development-resources/omadacycline-injection-and-oral-products). Essential agreement based on evaluable results excluded those from any isolate where either the MTS or reference BMD MIC result was ≤0.002 or ≥32 μg/ml. Very major errors were defined as isolates resistant by BMD MIC and susceptible by MTS MIC. Major errors were defined as isolates susceptible by BMD MIC and resistant by MTS MIC. Minor errors were defined as isolates susceptible by BMD MIC and intermediate by MTS MIC, intermediate by BMD MIC and susceptible by MTS MIC, intermediate by BMD MIC and resistant by MTS MIC, and resistant by BMD MIC and intermediate by MTS MIC. Very major error rates were based on number of very major errors divided by the number of resistant isolates and major error rates were based on number of major errors divided by the number of susceptible isolates and additionally, results were compared for determination of any trending (i.e., lower or higher results) compared to the reference method. Isolates with off-scale MIC results by both methods were excluded from this trending analysis. The total number of results ≥1 dilution lower and ≥1 dilution higher than the reference MIC was determined and percentages for each were calculated based on the total number of isolates. The percentage of results ≥1 dilution lower was subtracted from the percentage of results ≥1 dilution higher for determination of the overall trending rate. Based on recent FDA CDRH 510(k) review guidance (not included in current guidance document), a ≥30% trend is considered noteworthy for the device label.

The authors are committed to data transparency therefore deidentified data can be made available when permitted, feasible, and appropriate. Requests may be submitted to medinfo@paratekpharma.com for review.

RESULTS

Summary of QC, reproducibility and clinical/challenge isolate results are shown in Table 1, Table 2, and Table 3, respectively. QC results were within established CLSI QC ranges for all isolates by BMD and MTS, with exception of one H. influenzae ATCC 49247 MTS/HTM result of 4 μg/ml.

TABLE 1.

Omadacycline BMD and MTS MIC results (n) for quality control strains^a

QC organism	MIC μg/mL	Reference BMD MIC distribution				MTS MIC distribution
		Site 1	Site 2	Site 3	All sites	Site 1	Site 2	Site 3	All sites
E. coli ATCC 25922	0.12
	0.25
	0.5	17	21	20	58	20	21	19	60
	1	3			3			1	1
	2
	4
S. aureus ATCC 29213	0.06
	0.12	5			5	12	1		13
	0.25	15	20	17	52	8	15	17	40
	0.5			3	3		4	3	7
	1
	2
E. faecalis ATCC 29212	0.03
	0.06	3	4		7	4			4
	0.12	16	12	13	41	16	4	12	32
	0.25	1	4	7	12		13	8	21
	0.5						3		3
	1
CLSI Media
		HTM and CAMHB-LHB				HTM and MHA-SB
H. influenzae ATCC 49247	0.25
	0.5	3			3			1	1
	1	16		2	18	1	5	2	8
	2	1	20	18	39	20	15	17	52
	4					1			1
S. pneumoniae ATCC 49619	0.008
	0.016	5	7	5	17
	0.03	17	11	14	42
	0.06	1	2	4	7	9	11	17	37
	0.12					14	12	6	32
	0.25
EUCAST Media
		MH-F Broth				MH-F Agar
H. influenzae ATCC 49247	0.25
	0.5	9	14	19	42
	1	11	6	1	18	22	5	6	33
	2						15	14	29
	4
S. pneumoniae ATCC 49619	0.008
	0.016	15	3	5	23
	0.03	5	17	14	36
	0.06			4	4	7	9	9	25
	0.12					13	11	11	35
	NG	3			3

NG, No Growth.

^aShaded areas indicate CLSI established quality control ranges (2).

TABLE 2.

Summary of omadacycline MTS reproducibility results by organism group and isolate for all testing sites

Organism group (all-site and -isolate essential agreement)	Reproducibility strain no., species
		No. of results at 2-fold dilution differencea,b
		–2	–1	0	1	2
Gram Negative (270 of 270 results[100%])	R1, E. coli			27			0.5
	R2, E. coli			26	1		1
	R3, E. coli			26	1		2
	R4, E. coli		11	16			8
	R5, K. pneumoniae			14	13		1
	R6, K. pneumoniae		12	14	1		4
	R7, K. pneumoniae		8	19			8
	R8, E. cloacae		11	16			2
	R9, E. cloacae		5	22			2
	R10, E. cloacae		8	9			2
Gram Positive (257 of 270 results [95.2%])	R1, MSSA		10	15	2		0.25
	R5, MSSA		3	21	3		1
	R8, MRSA			12	8	7	0.12
	R10, MRSA		5	22			2
	R11, S. lugdunensis		6	21			0.12
	R12, S. lugdunensis		4	21	2		0.12
	R14, E. faecalis (VSE)		2	24	1		0.12
	R15, E. faecalis (VSE)			18	7	2	4
	R16, E. faecalis (VRE)		1	17	9		0.12
	R17, E. faecalis (VRE)			12	11	4	1
Fastidious HTM and MHA-SB (295 of 297 results [99.3%])	R1, S. pneumoniae		1	18	8		0.5
	R2, S. pneumoniae		6	21			0.12
	R3, S. pneumoniae		9	18			1
	R6, H. influenzae	1		24	2		2
	R7, H. influenzae		13	14			1
	R11, S. pyogenes			16	11		0.12
	R12, S. anginosus			19	8		0.5
	R14, S. constellatus		1	19	6	1	0.12
	R15, H. parainfluenzae			24	3		2
	R16, S. pyogenes		1	22	4		0.25
	R17, S. anginosus		12	15			0.12
Fastidious MH-F (266 of 270 results [98.5%])	R1, S. pneumoniae			20	7		0.5
	R2, S. pneumoniae		1	26			0.12
	R3, S. pneumoniae		2	24	1		1
	R6, H. influenzae		1	25	1		2
	R7, H. influenzae			21	6		0.5
	R11, S. pyogenes		3	24			0.25
	R12, S. anginosus			18	9		0.5
	R13, S. intermedius	4	8	15			0.03
	R14, S. constellatus		1	15	11		0.12
	R15, H. parainfluenzae		2	25			2

^aEssential agreement pertains to the shaded results, which are within ±1 dilution of the modal MIC.

^bThe 2-fold dilution difference is the difference between each MIC and the modal MIC for all results.

Enterobacterales.

The majority of MTS results for E. coli ATCC 25922 were at the low-middle end of the 4-dilution expected range (98.4% were 0.5 μg/ml, Table 1). Inter- and intralaboratory precision was excellent for the 10 Enterobacterales reproducibility isolates; all results were within ± one dilution of the modal MIC (Table 2).

EA rates for consolidated clinical and challenge Enterobacterales was 99.7% and 100%, respectively (Table 3, Fig. 1). Among the individual species of Enterobacterales, EA rates for the consolidated clinical and challenge isolates were 100% for C. freundii, C. koseri, K. aerogenes, K. oxytoca, and K. pneumoniae, and were 99.4% for E. cloacae and 99.3% for E. coli. There was no significant trending of MTS compared to BMD results noted for Enterobacterales. The mean colony counts (CFU/ml) for the MTS and BMD consolidated clinical and challenge isolate testing were 7.7 × 10⁷ and 3.8 × 10^5, respectively.

FIG 1.

Comparison of omadacycline MTS MIC distribution to BMD MIC distribution (n) for Enterobacterales (consolidated clinical and challenge isolates). Red lines in each scatterplot illustrate omadacycline FDA MIC breakpoints.

Nonfastidious Gram positive.

The majority of MTS results for S. aureus ATCC 29213 were at the low end of the 4-dilution expected range (21.7% were 0.12 μg/ml and 66.7% were 0.25 μg/ml, Table 1). The majority of MTS results for E. faecalis ATCC 29212 were in the middle of the 4-dilution expected range (53.3% were 0.12 μg/ml and 35.0% were 0.25 μg/ml, Table 1). Inter- and intralaboratory precision was acceptable for the 10 Gram-positive reproducibility strains; 95.2% of all results were within ± one dilution of the modal MIC (Table 2).

EA rates for consolidated clinical and challenge isolates were 98.9% for S. aureus, 100% for S. lugdunensis, 98.3% for E. faecalis, and 100% for E. faecium (Table 3, Fig. 2). Trending was limited to S. lugdunensis, for which a 41.4% trend of higher MTS MIC results was considered significant. The mean colony counts (CFU/ml) for the MTS and BMD consolidated clinical and challenge isolate testing were 6.8 × 10⁷ and 3.4 × 10⁵ for staphylococcus, 7.2 × 10⁷ and 3.6 × 10⁵ for enterococcus, respectively.

FIG 2.

Comparison of omadacycline MTS MIC distribution to BMD MIC distribution (n) for staphylococci and enterococci (consolidated clinical and challenge isolates). Red lines in each scatterplot illustrate omadacycline FDA MIC breakpoints. Red dotted lines indicate FDA CABP breakpoints for S. aureus.

Fastidious organisms.

The majority of HTM MTS results for H. influenzae ATCC 49247 were at the high end of the 3-dilution expected range (83.9% were 2 μg/ml) and there was one out-of-range result at 4 μg/ml (Table 1). The MH-F MTS results for H. influenzae ATCC 49247 were distributed at the mid and high end of the CLSI-based range (53.2% were 1 μg/ml and 46.8% were 2 μg/ml, Table 1). All MHA-SB MTS results for S. pneumoniae ATCC 49619 were at the high end of the four-dilution range, which differs from the CLSI BMD results (89.4% at the low end of the range). Inter- and intralaboratory precision was acceptable for the 10 reproducibility strains; 99.3% and 98.5% of all MTS results were within ±1 dilution of the modal MIC for CLSI media (HTM and MHA-SB) and EUCAST media (MH-F agar), respectively (Table 2). S. intermedius was initially tested in the study, but testing was paused as a result of poor to no growth in CAMHB-LHB in the BMD panels. It is noteworthy, that growth on the agar plates after 24 h was sufficient and MTS results could be obtained. However, since a reference MIC was not available, S. intermedius MTS results were not included in the final data set.

MTS on CLSI recommended agar compared to CLSI BMD.

EA agreement for MTS tested on HTM compared to HTM BMD for H. influenzae and H. parainfluenzae were 98.2% and 100%, respectively, with no significant trending. Essential agreement rates for MTS tested on MHA-SB compared to CAMHB-LHB were 91.7% for S. pneumoniae and 100% for S. agalactiae, S. anginosus, S. constellatus, and S. pyogenes. The majority of MHA-SB MTS MIC results for streptococci were one dilution higher than CAMHB-LHB BMD MIC results and therefore, trending rates of higher MIC results were observed (65.7–83.1%) (Table 4).

TABLE 4.

Comparison of fastidious organism MIC results using BMD and MTS methods with different media

	Media				Dilution differencea
MTS comparisons to BMD	MTS	BMD	Species	Total	>3	−3	−2	−1	0	1	2	3	>3	%EA	Notesb
CLSI MTS - LSI BMD	HTM	HTM	H. influenzae	112			1	19	49	42	1			98%	Data included in the FDA submission
	HTM	HTM	H. parainfluenzae	30				3	21	6				100%
	CAMHB-SB	CAMHB-LHB	S. agalactiae	20					5	15				100%
	CAMHB-SB	CAMHB-LHB	S. anginosus	27				1	6	20				100%
	CAMHB-SB	CAMHB-LHB	S. constellatus	18					5	13				100%
	CAMHB-SB	CAMHB-LHB	S. pneumoniae	181				2	27	137	14	3		92%
	CAMHB-SB	CAMHB-LHB	S. pyogenes	35				1	10	24				100%
EUCAST MTS - CLSI BMD	MH-F	HTM	H. influenzae	112				32	66	12	2			98%
	MH-F	HTM	H. parainfluenzae	30				5	22	3				100%
	MH-F	CAMHB-LHB	S. agalactiae	20					5	12	3			85%
	MH-F	CAMHB-LHB	S. anginosus	27				1	5	17	4			85%
	MH-F	CAMHB-LHB	S. constellatus	12				1	3	7	1			92%
	MH-F	CAMHB-LHB	S. pneumoniae	178				1	7	123	45	2		74%	3 NT
	MH-F	CAMHB-LHB	S. pyogenes	35				1	8	23	3			91%
EUCAST MTS - EUCAST BMD	MH-F	MH-F	H. influenzae	112				2	19	72	17	2		83%
	MH-F	MH-F	H. parainfluenzae	30					5	19	6			80%
	MH-F	MH-F	S. agalactiae	20					1	14	5			75%
	MH-F	MH-F	S. anginosus	26				1	3	19	3			88%	1 BMD NG
	MH-F	MH-F	S. constellatus	8					1	6	1			88%	5 BMD NG, 5 NT
	MH-F	MH-F	S. pneumoniae	176					5	102	55	7	7	61%	2 BMD NG, 3 NT
	MH-F	MH-F	S. pyogenes	33					6	22	4		1	85%	2 BMD NG
BMD comparisons	EUCAST	CLSI
EUCAST BMD - CLSI BMD	MH-F	HTM	H. influenzae	112		1	28	69	13	1				74%
	MH-F	HTM	H. parainfluenzae	30			6	21	3					80%
	MH-F	CAMHB-LHB	S. agalactiae	20				6	14					100%
	MH-F	CAMHB-LHB	S. anginosus	26				3	22	1				100%	1 MH-F BMD NG
	MH-F	CAMHB-LHB	S. constellatus	8					8					100%	5 MH-F BMD NG, 5 NT
	MH-F	CAMHB-LHB	S. pneumoniae	176	4	4	4	20	135	9				93%	2 MH-F BMD NG, 3 NT
	MH-F	CAMHB-LHB	S. pyogenes	33	1			3	29					97%	2 MH-F BMD NG

^aThe shaded area represents results that are within ±1 dilution (essential agreement) in comparison to the reference method.

^bNG, no growth; NT, not tested.

MTS on EUCAST (MH-F) agar compared to CLSI BMD.

EA agreement for MTS tested on MH-F compared to HTM BMD for H. influenzae and H. parainfluenzae were 98.2% and 100%. Essential agreement rates for MTS tested on MH-F compared to CAMHB-LHB were lower for streptococci, ranging from 73.6% for S. pneumoniae to 91.7% for S. constellatus (Table 3, Fig. 3). A significant trend of higher MTS MIC results for all streptococci and especially S. pneumoniae was observed (Table 4). The majority of MH-F MTS MIC results were one dilution higher than CAMHB-LHB BMD MIC results. The trending rates for S. pneumoniae were 94.9% and for other streptococci were 58.3–75.0%).

FIG 3.

Comparison of omadacycline MTS MIC distribution to BMD MIC distribution (n) for fastidious species (consolidated clinical and challenge isolates). Red lines in each scatterplot illustrate omadacycline FDA MIC breakpoints.

MTS on EUCAST (MH-F) agar compared to EUCAST (MH-F) BMD.

There was a trend of higher MTS results for all species tested, with the majority of MTS MH-F results one dilution higher than MH-F broth results. The trending rates for H. influenzae and H. parainfluenzae were 79.1% and 83.3%, respectively, and for S. pneumoniae were 33.7% and for other streptococci were 81.3–95.0%.

EUCAST (MH-F) BMD compared to CLSI BMD.

As shown in Table 4, there is a trend of lower MH-F BMD MIC results for H. influenzae and H. parainfluenzae compared to CLSI BMD MIC results (87.5% and 90%, respectively). There were also 7 streptococci (1 S. anginosus, 4 S. constellatus, 2 S. pneumoniae and 2 S. pyogenes) that did not grow in MH-F BMD and 12 S. pneumoniae with MH-F BMD MIC results at least 2 dilutions lower than the CLSI BMD MIC results.

The mean colony counts (CFU/ml) for the MTS and BMD consolidated clinical and challenge isolate testing for CLSI media were 8.8 × 10⁷ and 4.4 × 10⁵ for Haemophilus spp., 8.9 × 10⁷ and 4.4 × 10⁵ for all streptococcus and 8.1 × 10⁷ and 4.0 × 10⁵ for S. pneumoniae.

DISCUSSION

During the development of antimicrobial susceptibility testing devices for new antimicrobial agents, there are often many challenges to overcome. In the case of a broad-spectrum agent, the ability to test all species accurately and reproducibly using the same gradient strip and/or disk is an additional hurdle. In the case of the omadacycline MTS, the variety of diverse species which required testing of different media for fastidious organisms to address both CLSI and EUCAST reference methods was yet another challenge. The evaluation of an AST device must sufficiently test the common variables that can affect an MIC, such as interlaboratory methodology and reading variations and should also include isolates that are not susceptible to ensure detection of resistance (11).

The MTS product performed well in this multisite and varied-isolate study, particularly with regard to CLSI-based media and essential agreement analysis. All categorical errors were minor errors and for categorical agreement rates that fell below 90%, essential agreement rates were 100% for C. freundii, S. lugdunensis, S. anginosus and S. constellatus and 98.3% for E. faecalis. Although two different breakpoints are to be used for ABSSSI and CABP S. aureus isolates, the categorical agreements in this study were similar (96.0 and 96.6%, respectively). For nonfastidious species and Haemophilus spp., with the exception of S. lugdunensis, there was no significant trending of MTS results compared to BMD results (i.e., at least 30% of results scattered around higher or lower values). Although trending of MTS results compared to BMD results was observed for all streptococcus species, essential agreement rates of >98% were achieved for all species with the exception of S. pneumoniae, which had an EA rate of 91.7% (within the >90% FDA guidance criteria). Since the majority of both BMD and MTS results for H. influenzae ATCC 49247 were 2 μg/ml (the high end of the CLSI quality control range), verification of the CLSI range may be warranted. Insufficient growth of S. intermedius was observed after following the CLSI BMD method, which recommended a 20–24-h incubation period. Subsequently, troubleshooting was performed to allow for longer incubation periods and some isolates that were incubated for an additional 24 h did in fact grow sufficiently for MIC determination. Further testing at longer incubation times is thus recommended and amendment to the CLSI BMD method specific to the incubation period for S. intermedius may be necessary. Based on the FDA review, the MTS met the criteria for clearance with the following performance notations necessary in the product label: (1) MTS™ Omadacycline MIC values tend to be in exact agreement or at least one double dilution higher when testing S. lugdunensis, S. anginosus, S. constellatus, S. pneumoniae, and S. pyogenes compared to the CLSI reference broth microdilution (2) The performance of Streptococcus intermedius has not been established during the clinical study (3) The ability of MTS™ to detect non-susceptible isolates for C. freundii, H. influenzae, H. parainfluenzae, S. anginosus, S. constellatus and S. pyogenes is unknown because resistant isolates were either not available or an insufficient number was encountered at the time of comparative testing.

When streptococci were tested on MH-F agar, the trend for higher omadacycline MTS results was slightly more pronounced then results on MHA-SB, resulting in essential agreement rates below 90%. Although the same lot of MHA was used in both sources of prepared MH-F agar plates, different lots of horse blood and NAD were used. However, there were no site-based MIC variations that indicate these differences in the MH-F agar affected MIC results. Additional testing of streptococci with different sources of MH-F agar is warranted for further validation.

With regard to the comparison of MH-F BMD and CLSI BMD fastidious isolate MIC results, this study showed evidence of better growth in the CLSI broth (i.e., growth issues and lower MIC results with some streptococci and lower MIC results for Haemophilus spp. for MH-F BMD testing). A limitation of the study is that colony counts were not performed from the MH-F BMD control wells. However, growth in the HTM growth control wells was notated as better in comparison to the MH-F control wells in some cases. The same NAD preparation and different horse blood products (lysed for the broth and defibrinated for the agar) were used for production of the MH-F agar and MH-F broth. Further investigation is warranted with other lots of media and components.

The results for some non-indicated species were included in this study for information purposes (i.e., S. epidermidis, S. haemolyticus, S. saprophyticus, E. avium, E. faecium, E. gallinarum, and S. agalactiae). The essential agreement rates were 100% for all of these species, with exception of E. gallinarum, which was 95%. These species are examples of those that may be tested by a clinical laboratory and although the device may show acceptable performance relative to the MIC, the device label cannot provide supporting data. S. agalactiae is included in the microbiology section of the omadacycline label based on its in vitro activity, however, the current guidance from FDA, CDRH, is that non-indicated species cannot be included in the “in vitro only” list of the device label if the species does not have a breakpoint. Performance data based on essential agreement results only or utilizing a similar species breakpoint for categorical agreement and error rate determination are two options worth further discussion with CDRH. This study does not provide comparative data for other species that may be requested for testing (e.g., Acinetobacter baumannii).

For optimal correlation to reference BMD, the MTS reading instruction for nonfastidious species is to interpret the MIC at 80% inhibition where trailing is seen. However, for fastidious species, the MIC should be read at 100% inhibition. The final product label should be reviewed for complete method instructions and reading guide pictures.

Overall, the omadacycline MTS performed well in comparison to reference BMD for the species tested. The limitations, as discussed, were related to some trending in comparison of the MTS result to the reference BMD result, lack of resistant isolates for certain species and need for further investigation with regard to use of EUCAST media. This FDA cleared MTS method was shown to be a reliable method for testing omadacycline MIC against relevant clinical isolates.