In Vitro Potency and Spectrum of the Novel Polymyxin MRX-8 Tested against Clinical Isolates of Gram-Negative Bacteria

ABSTRACT

The polymyxins display excellent in vitro antimicrobial activity against most Enterobacterales, Pseudomonas aeruginosa, and Acinetobacter baumannii isolates, but their clinical utility has been limited because of class-specific toxicity problems. Therefore, new polymyxin analogs with improved safety properties are needed to combat serious infections caused by resistant Gram-negative pathogens. MRX-8 is a novel polymyxin B analog that displays reduced toxicity in in vitro and animal assays and is currently being evaluated in a phase 1 clinical trial. In this nonclinical study, the in vitro potency and spectrum of MRX-8 and comparators were evaluated against a large set of Gram-negative clinical isolates collected in the United States in 2017 to 2020. MRX-8, colistin, and polymyxin B exhibited nearly identical antimicrobial activities against the Enterobacterales, Pseudomonas aeruginosa, and Acinetobacter baumannii isolate sets. MRX-8 MIC₅₀ and MIC₉₀ values were 0.12 and 0.25 mg/L, respectively, for the set of Enterobacterales isolates not intrinsically resistant to colistin and 0.5 and 1 mg/L, respectively, against both the A. baumannii and P. aeruginosa isolate sets. All three polymyxin-class compounds retained activity against meropenem-resistant and multidrug-resistant isolate subsets but were inactive against isolates displaying acquired or intrinsic resistance to polymyxins. These results support the continued development of MRX-8 to treat serious Gram-negative infections.

INTRODUCTION

Many recent reports and reviews have emphasized the need for the development of novel or improved antimicrobial agents to treat serious infections caused by Gram-negative species or groups like the Enterobacterales, Pseudomonas aeruginosa, and Acinetobacter baumannii, which often display resistance to frontline drugs (1–4). Of particular concern are Gram-negative isolates that are multidrug resistant (MDR) or carbapenem resistant due to multifactorial mechanisms or the production of class B metallo-β-lactamases (3, 5).

The clinically relevant polymyxin antimicrobial class is composed of colistin (polymyxin E) and polymyxin B, which are structurally related polycationic molecules with a cyclic lipodecapeptide structure composed of a heptapeptide core, a tripeptide linear linker, and an N-terminal fatty acyl group. The structures of colistin and polymyxin B differ only at the R6 peptide position, which is d-phenylalanine in polymyxin B and d-leucine in colistin (6). Commercial preparations of polymyxin B and colistin also typically contain mixtures of components with related but distinct fatty acid groups (7). Three free amino groups are present within the heptapeptide core, and two free amino groups are present within the linker region. Both colistin and polymyxin B exhibit rapid in vitro bactericidal activity against many important Gram-negative pathogens. The mode of action is principally mediated by disruption of the bacterial outer membrane through binding of the lipopeptide cationic groups to lipopolysaccharide, followed by disruption of the inner membrane, which likely involves interaction with the fatty acid polymyxin tail (6, 8).

Gram-positive species and several Enterobacterales species (e.g., Providencia spp. and Proteus spp.) are intrinsically resistant to the polymyxins (9, 10). In contrast, Escherichia coli, Klebsiella spp., Citrobacter spp., Enterobacter spp., P. aeruginosa, and A. baumannii are intrinsically susceptible to the polymyxins but resistance can be acquired by various methods, including mutations in chromosomal genes or the acquisition of mobile genetic elements that contain mcr-1 or related genes encoding enzymes that modify the bacterial lipopolysaccharide structure (9, 11, 12).

Although the discovery of polymyxins was reported in 1947 (for a review, see reference 7), their clinical use was short-lived due to concerns about nephrotoxicity and neurotoxicity and the discovery of safer alternative antimicrobials (7, 13). Because of the recent rise in highly drug-resistant Gram-negative pathogens, however, there has been a resurgence in the clinical use of the polymyxins as drugs of last resort (14).

Several groups have launched research efforts to improve the clinical utility of the polymyxins with the goal of maintaining their excellent antimicrobial features while abrogating their toxicity (6, 14). Other groups have instead focused on designing truncated polymyxins, with decreased antimicrobial activity and decreased toxicity, that still maintain the ability to potentiate the entry of other antimicrobials through the Gram-negative outer membrane (8).

MRX-8 is a polymyxin-class antimicrobial under development that was designed using soft drug principles, in which the goal is to produce molecules that are “deactivated in a predictable and controllable way after achieving their therapeutic goal” (15). To that end, MRX-8 is an analog of polymyxin B in which the fatty acid tail is linked to the rest of the molecule by a polar ester group. MRX-8 maintains in vitro antimicrobial activity (reference 16 and see below). After exerting its antimicrobial effect in vivo, however, MRX-8 is intended to be converted by endogenous esterases into a nontoxic metabolite that displays minimized cell culture toxicity and animal nephrotoxicity properties, compared to polymyxin B (16).

Lepak et al. (17) recently investigated the pharmacodynamic (PD) properties of MRX-8, compared to polymyxin B, in mouse thigh and lung infection models involving Enterobacterales, P. aeruginosa, and A. baumannii strains. They found that MRX-8 exhibited efficacy in those animal models and that AUC/MIC and maximum drug concentration (C_max)/MIC were the pharmacokinetic (PK)/PD indices that best correlated with the observed antimicrobial therapeutic effects.

The safety, tolerability, and human PK features of MRX-8 following intravenous dosing are currently being evaluated in a phase 1 clinical trial (ClinicalTrials registration no. NCT04649541), with results expected in 2022. In this study, we investigated the in vitro antimicrobial potency and spectrum of MRX-8, compared to colistin and polymyxin B, when tested against a large set of Gram-negative pathogens collected in the United States from 2017 to 2020.

RESULTS

Cumulative distributions of MIC values for each isolate set tested against MRX-8, colistin, and polymyxin B are shown in Table 1, Table S1, and Table S2, respectively. Summary tables that display MIC₅₀ and MIC₉₀ values, MIC ranges, and percentages of susceptible, intermediate, or resistant isolates for various species are displayed in Table 2 and in Tables S3 to S6 in the supplemental material.

TABLE 1.

Cumulative MIC distributions for MRX-8 tested against the main species and organism groups

Species/organism group (no. of isolates)	No. (cumulative %) of isolates inhibited with MIC of:													MIC50 (mg/L)	MIC90 (mg/L)
	≤0.015 mg/L	0.03 mg/L	0.06 mg/L	0.12 mg/L	0.25 mg/L	0.5 mg/L	1 mg/L	2 mg/L	4 mg/L	8 mg/L	16 mg/L	32 mg/L	>a
Enterobacterales (787)
Enterobacterales isolates not intrinsically resistant to colistin (685)b	0 (0.0)	2 (0.3)	86 (12.8)	450 (78.5)	109 (94.5)	20 (97.4)	0 (97.4)	1 (97.5)	0 (97.5)	3 (98.0)	2 (98.2)	1 (98.4)	11 (100.0)	0.12	0.25
Citrobacter spp. (47)		0 (0.0)	10 (21.3)	30 (85.1)	6 (97.9)	1 (100.0)								0.12	0.25
Enterobacter cloacae species complex (52)		0 (0.0)	8 (15.4)	27 (67.3)	6 (78.8)	1 (80.8)	0 (80.8)	0 (80.8)	0 (80.8)	1 (82.7)	0 (82.7)	0 (82.7)	9 (100.0)	0.12	>32
Escherichia coli (261)	0 (0.0)	2 (0.8)	46 (18.4)	164 (81.2)	40 (96.6)	7 (99.2)	0 (99.2)	0 (99.2)	0 (99.2)	1 (99.6)	1 (100.0)			0.12	0.25
Klebsiella aerogenes (23)		0 (0.0)	1 (4.3)	15 (69.6)	5 (91.3)	2 (100.0)								0.12	0.25
Klebsiella oxytoca (37)		0 (0.0)	6 (16.2)	30 (97.3)	1 (100.0)									0.12	0.12
Klebsiella pneumoniae (265)		0 (0.0)	15 (5.7)	184 (75.1)	51 (94.3)	9 (97.7)	0 (97.7)	1 (98.1)	0 (98.1)	1 (98.5)	1 (98.9)	1 (99.2)	2 (100.0)	0.12	0.25
Colistin-resistant (17)									0 (0.0)	3 (17.6)	2 (29.4)	1 (35.3)	11 (100.0)	>32	>32
Meropenem-resistant (10)		0 (0.0)	1 (10.0)	3 (40.0)	3 (70.0)	0 (70.0)	0 (70.0)	0 (70.0)	0 (70.0)	1 (80.0)	0 (80.0)	0 (80.0)	2 (100.0)	0.25	>32
MDR (54)		0 (0.0)	3 (5.6)	34 (68.5)	11 (88.9)	3 (94.4)	0 (94.4)	0 (94.4)	0 (94.4)	1 (96.3)	0 (96.3)	0 (96.3)	2 (100.0)	0.12	0.5
Enterobacterales isolates intrinsically resistant to colistin (102)c											0 (0.0)	2 (2.0)	100 (100.0)	>32	>32
Acinetobacter baumannii (264)	0 (0.0)	1 (0.4)	3 (1.5)	33 (14.0)	58 (36.0)	116 (79.9)	35 (93.2)	12 (97.7)	1 (98.1)	0 (98.1)	2 (98.9)	1 (99.2)	2 (100.0)	0.5	1
Meropenem-resistant (74)			0 (0.0)	9 (12.2)	20 (39.2)	32 (82.4)	9 (94.6)	1 (95.9)	0 (95.9)	0 (95.9)	1 (97.3)	1 (98.6)	1 (100.0)	0.5	1
Colistin-nonresistant (258)	0 (0.0)	1 (0.4)	3 (1.6)	33 (14.3)	58 (36.8)	116 (81.8)	35 (95.3)	12 (100.0)						0.5	1
Colistin-resistant (6)								0 (0.0)	1 (16.7)	0 (16.7)	2 (50.0)	1 (66.7)	2 (100.0)	16
MDR (104)			0 (0.0)	15 (14.4)	23 (36.5)	44 (78.8)	15 (93.3)	4 (97.1)	0 (97.1)	0 (97.1)	1 (98.1)	1 (99.0)	1 (100.0)	0.5	1
Pseudomonas aeruginosa (263)		0 (0.0)	3 (1.1)	9 (4.6)	19 (11.8)	166 (74.9)	64 (99.2)	2 (100.0)						0.5	1
Meropenem-resistant (31)			0 (0.0)	3 (9.7)	3 (19.4)	20 (83.9)	5 (100.0)							0.5	1
MDR (46)		0 (0.0)	1 (2.2)	4 (10.9)	6 (23.9)	22 (71.7)	12 (97.8)	1 (100.0)						0.5	1

^aGreater than the highest concentration tested.

^bSpecies included Citrobacter amalonaticus/farmeri (1 isolate), Citrobacter freundii (1 isolate), C. freundii species complex (26 isolates), Citrobacter koseri (19 isolates), Enterobacter asburiae (1 isolate), Enterobacter cloacae (14 isolates), E. cloacae species complex (36 isolates), Enterobacter hormaechei (1 isolate), Escherichia coli (261 isolates), Klebsiella aerogenes (23 isolates), Klebsiella oxytoca (37 isolates), and Klebsiella pneumoniae (265 isolates).

^cSpecies included Morganella morganii (10 isolates), Proteus mirabilis (46 isolates), Proteus penneri (1 isolate), Proteus vulgaris group (5 isolates), Providencia rettgeri (7 isolates), Providencia stuartii (8 isolates), Serratia liquefaciens complex (3 isolates), and Serratia marcescens (22 isolates).

TABLE 2.

Antimicrobial activity of MRX-8 and comparator agents tested against Escherichia coli, Klebsiella pneumoniae, Acinetobacter baumannii, and Pseudomonas aeruginosa isolates

Species and antimicrobial agent	No. of isolates	MIC findings (mg/L)			Susceptibility results (%) using:
					CLSI criteriaa			EUCAST criteriaa
		MIC50	MIC90	MIC range	S	I	R	S	I	R
Escherichia coli
MRX-8	261	0.12	0.25	0.03 to 16
Colistin	261	0.25	0.25	0.06 to 8		99.2	0.8	99.2		0.8
Polymyxin B	261	0.25	0.5	0.12 to 8		99.6	0.4
Amikacin	261	2	8	0.5 to 16	100.0	0.0	0.0	98.9b		1.1
Ceftazidime	261	0.25	8	0.03 to >32	88.1	2.7	9.2	82.8	5.4	11.9
Ceftazidime-avibactam (fixed at 4 mg/L)	77	0.06	0.25	≤0.015 to 0.25	100.0		0.0	100.0		0.0
Ceftriaxone	261	≤0.06	>8	≤0.06 to >8	83.5	0.0	16.5	83.5,c 83.5d	0.0d	16.5,c 16.5d
Gentamicin	261	1	>16	≤0.12 to >16	87.0	0.4	12.6	86.6b		13.4
Levofloxacin	260	≤0.03	16	≤0.03 to >16	68.8	0.8	30.4	68.8	0.8	30.4
Meropenem	261	0.015	0.03	0.008 to 8	99.6	0.0	0.4	99.6,c 99.6d	0.4d	0.4,c 0.0d
Piperacillin-tazobactam (fixed at 4 mg/L)	261	2	8	1 to >128	96.6	1.5	1.9	94.3		5.7
Tigecycline	261	0.12	0.25	≤0.06 to 1	100.0e	0.0	0.0	98.9		1.1
Klebsiella pneumoniae
MRX-8	265	0.12	0.25	0.06 to >32
Colistin	265	0.12	0.25	0.12 to >32		98.1	1.9	98.1		1.9
Polymyxin B	265	0.25	0.5	0.12 to >32		98.1	1.9
Amikacin	265	1	2	0.5 to >32	98.9	0.0	1.1	98.5b		1.5
Ceftazidime	265	0.25	16	0.03 to >32	88.3	1.1	10.6	87.5	0.8	11.7
Ceftazidime-avibactam (fixed at 4 mg/L)	102	0.12	0.25	0.03 to 1	100.0		0.0	100.0		0.0
Ceftriaxone	265	≤0.06	>8	≤0.06 to >8	87.9	0.0	12.1	87.9,c 87.9d	0.0d	12.1,c 12.1d
Gentamicin	265	0.25	0.5	≤0.12 to >16	94.0	0.4	5.7	94.0b		6.0
Levofloxacin	265	0.06	1	≤0.03 to >16	86.0	4.2	9.8	86.0	4.2	9.8
Meropenem	265	0.03	0.03	0.015 to >16	97.4	0.4	2.3	97.7,c 97.7d	0.4d	2.3,c 1.9d
Piperacillin-tazobactam (fixed at 4 mg/L)	265	4	16	0.12 to >128	93.6	2.3	4.2	89.4		10.6
Tigecycline	265	0.5	1	≤0.06 to 8	98.1e	1.5	0.4
Acinetobacter baumannii
MRX-8	264	0.5	1	0.03 to >32
Colistin	264	0.25	1	0.06 to >32		97.7	2.3	97.7		2.3
Polymyxin B	264	0.25	0.5	0.12 to 16		98.5	1.5
Amikacin	264	4	>32	0.5 to >32	83.3	1.9	14.8	79.2b		20.8
Ceftazidime	264	8	>32	1 to >32	65.2	7.2	27.7
Ceftazidime-avibactam (fixed at 4 mg/L)	91	8	32	1 to >32
Ceftriaxone	224	>8	>8	4 to >8	21.4	0.0	0.0
Gentamicin	264	1	>16	≤0.12 to >16	75.4	5.7	18.9	75.4b		24.6
Levofloxacin	264	0.25	>16	≤0.015 to >16	65.9	1.1	33.0	62.9	2.3	34.8
Meropenem	264	0.5	>16	0.06 to >16	71.2	0.8	28.0	71.2,c 71.2d	1.1d	28.8,c 27.7d
Piperacillin-tazobactam (fixed at 4 mg/L)	258	8	>128	≤0.06 to >128	58.5	7.0	34.5
Tigecycline	264	0.5	4	≤0.06 to 8
Pseudomonas aeruginosa
MRX-8	263	0.5	1	0.06 to 2
Colistin	263	0.5	1	0.12 to 2		100.0	0.0	100.0		0.0
Polymyxin B	263	0.5	1	0.12 to 2		100.0	0.0
Amikacin	263	4	8	≤0.25 to >32	96.2	1.1	2.7	96.2b		3.8
Ceftazidime	263	2	32	0.25 to >32	82.1	4.2	13.7	f	82.1	17.9
Ceftazidime-avibactam (fixed at 4 mg/L)	92	2	4	0.25 to >32	98.9		1.1	98.9		1.1
Ceftriaxone	211	>8	>8	0.5 to >8
Gentamicin	263	2	8	≤0.12 to >16	87.8	6.5	5.7
Levofloxacin	263	0.5	8	≤0.03 to >16	68.4	11.8	19.8	f	68.4	31.6
Meropenem	263	0.5	8	0.015 to >16	82.5	5.7	11.8	82.5,c 82.5d	8.7d	17.5,c 8.7d
Piperacillin-tazobactam (fixed at 4 mg/L)	263	4	128	≤0.06 to >128	79.8	7.6	12.5	f	79.8	20.2
Tigecycline	263	8	>8	0.25 to >8

^aCriteria as published by CLSI (10) and EUCAST (23). I, intermediate; R, resistant; S, susceptible.

^bFor infections originating from the urinary tract. For systemic infections, aminoglycosides must be used in combination with other active therapy.

^cUsing meningitis breakpoints.

^dUsing nonmeningitis breakpoints.

^eUsing FDA breakpoints.

^fAn arbitrary susceptible breakpoint of ≤0.001 mg/L and/or >50 mm has been published by EUCAST, indicating that susceptible should not be reported for this organism-agent combination and intermediate should be interpreted as susceptible-increased exposure.

Activity of MRX-8 and comparators against Enterobacterales species.

The activity of MRX-8 and comparator lipopeptides was first measured against Enterobacterales species that are not intrinsically resistant to colistin, including E. coli (Table 2), Klebsiella pneumoniae (Table 2), Citrobacter spp. (see Table S3), Enterobacter cloacae species complex (see Table S4), Klebsiella aerogenes (see Table S5), and Klebsiella oxytoca (see Table S6). Within each species or group of isolates, the MIC₅₀ and MIC₉₀ values for MRX-8, colistin, and polymyxin B agreed within 2-fold. For example, against the E. coli subset (Table 2), the MIC₅₀ and MIC₉₀ values for MRX-8, colistin, and polymyxin B were 0.12 and 0.25 mg/L, 0.25 and 0.25 mg/L, and 0.25 and 0.5 mg/L, respectively. There was also little variation in the different polymyxin MIC₅₀ and MIC₉₀ values among the various Enterobacterales species and groups. Except for the E. cloacae species complex isolate subset, the MIC₅₀ and MIC₉₀ values were 0.12 and 0.12 to 0.25 mg/L, respectively, for MRX-8 (Table 1), 0.12 to 0.25 and 0.25 mg/L, respectively, for colistin (see Table S1), and 0.25 and 0.25 to 0.5 mg/L, respectively, for polymyxin B (see Table S2). In contrast, the E. cloacae species complex isolate subset MIC₉₀ values for MRX-8 (Table 1), colistin (see Table S1), and polymyxin B (see Table S2) were much higher (>32 mg/L for each lipopeptide) than those for the other species groups of Enterobacterales that are not intrinsically resistant to colistin. This difference is due to the larger percentage of Enterobacter species isolates that displayed acquired colistin resistance (19.2%) (see Table S4), compared to the other species.

MRX-8 (Table 1) and colistin (see Table S1) were 2-fold more potent (MIC₅₀ and MIC₉₀ values for both antimicrobials of 0.12 and 0.25 mg/L, respectively) than polymyxin B (MIC₅₀ and MIC₉₀ of 0.25 and 0.5 mg/L, respectively) (see Table S2) against the combined set of Enterobacterales isolates not intrinsically resistant to colistin. This potency difference can be observed in the scatterplot of polymyxin B MIC values versus MRX-8 MIC values shown in Fig. 1A.

FIG 1.

Scatterplot of MRX-8 and polymyxin B MIC values when tested against Enterobacterales and Acinetobacter baumannii isolates. The figure displays a scatterplot of MIC values for MRX-8 and polymyxin B tested against Enterobacterales isolates (787 isolates) (A) and Acinetobacter baumannii isolates (264 isolates) (B). The gray cells represent identical MIC values for the two antimicrobials.

Against the meropenem-resistant subset of 10 Enterobacterales isolates from species that are not intrinsically resistant to colistin, the MIC₅₀ and MIC₉₀ values for MRX-8, colistin, and polymyxin B were 0.25 and >32 mg/L (Table 1), 0.12 and >32 mg/L (see Table S1), and 0.25 and 16 mg/L (see Table S2), respectively. Against the MDR subset of 54 Enterobacterales isolates from species that are not intrinsically resistant to colistin, the MIC₅₀ and MIC₉₀ values for MRX-8, colistin, and polymyxin B were 0.12 and 0.5 mg/L (Table 1), 0.12 and 0.5 mg/L (see Table S1), and 0.25 and 0.5 mg/L (see Table S2), respectively. MRX-8, colistin, and polymyxin B were all inactive (MIC₅₀ values of ≥32 and MIC₉₀ values of >32 mg/L) against the subset of 17 Enterobacterales isolates from species that are not intrinsically resistant to colistin but displayed acquired colistin resistance (Table 1; also see Tables S1 and S2 in the supplemental material).

Finally, we investigated the activity of MRX-8 and comparator lipopeptides against a subset of Enterobacterales species that are intrinsically resistant to colistin, such as Morganella morganii (10) (Table 1; also see Tables S1 and S2). As expected, MRX-8, colistin, and polymyxin B were all inactive against this isolate set (MIC₉₀ values of >32 mg/L for all three antimicrobials).

Activity of MRX-8 and comparators against Acinetobacter baumannii.

The in vitro antimicrobial activities of MRX-8 and comparators against the A. baumannii isolate subset are displayed in Table 2. The MIC₅₀ and MIC₉₀ values for the three lipopeptides agreed within 2-fold (MIC₅₀ and MIC₉₀ ranges of 0.25 to 0.5 and 0.5 to 1 mg/L, respectively). Colistin and polymyxin B were 2-fold more potent than MRX-8 according to MIC₅₀ values. This effect can be observed in the scatterplot of polymyxin B MIC values and MRX-8 MIC values shown in Fig. 1B. The three lipopeptides maintained their nearly equivalent potencies against the meropenem-resistant A. baumannii isolate subset. In vitro activity was lost for each antimicrobial, however, against the subset of colistin-resistant A. baumannii isolates, although 2 of the 6 colistin-resistant isolates remained intermediate to polymyxin B (Table 1; also see Tables S1 and S2). MDR status did not significantly affect the MIC₅₀ and MIC₉₀ values for any of the three polymyxins (Table 1; also see Tables S1 and S2).

Activity of MRX-8 and comparators against Pseudomonas aeruginosa.

The in vitro antimicrobial activities of MRX-8 and comparators against the P. aeruginosa isolate subset are displayed in Table 2. The MIC₅₀ and MIC₉₀ values for the three polymyxins were identical (MIC₅₀ and MIC₉₀ values of 0.5 and 1 mg/L, respectively), and the antimicrobial activities were unchanged against the subsets of meropenem-resistant and MDR P. aeruginosa isolates (Table 1; also see Tables S1 and S2).

No colistin-resistant P. aeruginosa isolates were present within the randomly selected set tested in this study (Table 2). However, we measured MIC values for the three polymyxins against a JMI Laboratories colistin-resistant P. aeruginosa isolate (collection no. 991784). All three polymyxins were inactive against this isolate, which displayed modal MIC values for MRX-8, colistin, and polymyxin B of >32 mg/L, 32 mg/L, and 8 mg/L, respectively (data not shown).

DISCUSSION

In this study, the in vitro antimicrobial activity of MRX-8 and polymyxin comparators was measured against a large set of clinically relevant Gram-negative pathogens, including various Enterobacterales species, A. baumannii, and P. aeruginosa. MRX-8, colistin, and polymyxin B exhibited identical or nearly identical in vitro antimicrobial activities against almost all of the Gram-negative species and groups tested, including MDR and meropenem-resistant subsets. Against the Enterobacterales isolates not intrinsically resistant to colistin, MRX-8 was 2-fold more potent than polymyxin B according to MIC₅₀ and MIC₉₀ values. In contrast, MRX-8 was about 2-fold less potent than polymyxin B against the A. baumannii set and equipotent against the P. aeruginosa isolate set according to MIC₅₀ and MIC₉₀ values. In general, all three polymyxins exhibited potent in vitro activity against the pathogen groups, e.g., the MRX-8 MIC₅₀ and MIC₉₀ values were 0.12 and 0.25 mg/L, respectively, against the set of Enterobacterales isolates not intrinsically resistant to colistin and 0.5 and 1 mg/L, respectively, against the A. baumannii and P. aeruginosa isolate sets. MRX-8, colistin, and polymyxin B were all inactive against isolate subsets that displayed acquired or intrinsic resistance to colistin.

MRX-8 merits additional study for the potential treatment of serious infections caused by Gram-negative pathogens. This conclusion is based on the following considerations. First, the in vitro potency and spectrum of MRX-8 are similar to those of colistin and polymyxin B. Second, the unique soft drug structure of MRX-8, featuring an ester bond that is cleaved by enzymes in vivo, is designed to decrease the potential for nephrotoxicity, which is a significant treatment-limiting side effect of colistin and polymyxin B use. Although further work is needed, in vitro and in vivo toxicology studies have generated favorable data in support of this hypothesis (16). Third, recent animal data suggest that MRX-8 exhibits PK/PD properties equivalent or superior to those of colistin and polymyxin B (17).

Finally, when evaluating the potential clinical utility of MRX-8, it is important to consider the existing and potential future colistin resistance rates within target pathogen groups, because MRX-8, like other polymyxins, is inactive against such resistant isolates. Global longitudinal rates of colistin resistance have recently been reviewed for the Enterobacterales (18), A. baumannii (19), and P. aeruginosa (20). In general, European Committee on Antimicrobial Susceptibility Testing (EUCAST) colistin susceptibility rates remain high for these target pathogens, but significant global variation exists. Against the Enterobacterales, colistin susceptibility rates were generally >90%, but increased resistance has been noted for Klebsiella spp. and Enterobacter spp. (9, 18). Against A. baumannii isolates, colistin susceptibility was >93% overall in each global region examined from 2005 to 2016, although resistance rates did increase over this period, particularly in Europe (9, 19). Importantly, colistin resistance rates are also typically higher against carbapenem-resistant A. baumannii isolate subsets than against randomly selected isolate sets. In contrast to the Enterobacterales and A. baumannii findings, P. aeruginosa susceptibility rates for colistin have remained >99% from 2005 to the present (9, 20).

In summary, we have shown that MRX-8, which is a lipopeptide designed to abrogate the cytotoxic properties common to other polymyxins, displayed in vitro antimicrobial activity against Enterobacterales, A. baumannii, and P. aeruginosa clinical isolates that was identical or nearly identical to the activities of colistin and polymyxin B.

MATERIALS AND METHODS

Bacterial isolates.

A total of 1,314 nonduplicate, Gram-negative clinical isolates from the SENTRY Antimicrobial Surveillance Program (21) were randomly selected from 77 medical centers located in all 9 U.S. Census Bureau divisions in 2017 to 2020 (see Table S7 in the supplemental material). The isolate sets were composed of various Enterobacterales species (including some species intrinsically resistant to polymyxins), Acinetobacter baumannii-calcoaceticus species complex (referred to as A. baumannii), and Pseudomonas aeruginosa. All organisms were isolated from various documented infection types (see Table S8), and only 1 isolate per patient infection episode was included in the surveillance collection. The number of isolates tested per species was arbitrary and should not be interpreted as representing the clinical prevalence of the species for these infection types. Species identification was performed at the participating medical centers and confirmed at the monitoring laboratory (JMI Laboratories, North Liberty, IA, USA) using standard microbiological methods and matrix-assisted laser desorption ionization–time of flight mass spectrometry (Bruker, Billerica, MA, USA).

Isolates were categorized as MDR if they were nonsusceptible to at least one antimicrobial from ≥3 drug classes using Clinical and Laboratory Standards Institute (CLSI) breakpoints (10). For the A. baumannii isolate set, the antimicrobial classes were extended-spectrum cephalosporins (ceftriaxone, ceftazidime, or cefepime), carbapenems (meropenem, doripenem, or imipenem), piperacillin-tazobactam (fixed at 4 mg/L), quinolones (ciprofloxacin or levofloxacin), aminoglycosides (gentamicin, tobramycin, or amikacin), polymyxins (colistin or polymyxin B), tetracyclines (tetracycline, doxycycline, or minocycline), and ampicillin-sulbactam (2:1). For the P. aeruginosa isolate set, the antimicrobial classes were cephalosporins (ceftazidime or cefepime), carbapenems (meropenem, doripenem, or imipenem), piperacillin-tazobactam (fixed at 4 mg/L), quinolones (ciprofloxacin or levofloxacin), aminoglycosides (gentamicin, tobramycin, or amikacin), and polymyxins (colistin or polymyxin B). For the set of Enterobacterales isolates that are not intrinsically resistant to colistin (E. coli, Enterobacter spp., Citrobacter spp., and Klebsiella spp.), the antimicrobial classes were extended-spectrum cephalosporins (cefotaxime, ceftriaxone, ceftazidime, or cefepime), carbapenems (meropenem, doripenem, or imipenem), piperacillin-tazobactam (fixed at 4 mg/L), quinolones (ciprofloxacin or levofloxacin), aminoglycosides (gentamicin, tobramycin, or amikacin), polymyxins (colistin or polymyxin B), and tigecycline. The MIC data for MDR categorizations were obtained from the SENTRY Antimicrobial Surveillance Program (21). A subset of MIC data for comparator antimicrobials is displayed in Table 2 and Tables S3 to S6. We also tested the activity of MRX-8 and comparators against a rare colistin-resistant P. aeruginosa isolate (JMI 991784) collected in 2017.

Susceptibility testing methods.

MRX-8 powder was supplied by MicuRx Pharmaceuticals. Colistin (catalog no. C4461) and polymyxin B (catalog no. 1547007) were obtained from Sigma-Aldrich and United States Pharmacopeia, respectively. MRX-8, colistin, and polymyxin stocks were made in water.

Susceptibility to MRX-8 and comparator agents was measured using current CLSI methods (10, 22). The test medium was cation-adjusted Mueller-Hinton broth. Non-tissue-culture-treated polystyrene plates were used. CLSI and EUCAST interpretive criteria were applied according to current guidelines (10, 23). U.S. FDA product package insert interpretive criteria were used for tigecycline (24).

The current EUCAST interpretive criteria for colistin against Enterobacterales, P. aeruginosa, and A. baumannii isolates are ≤2 mg/L for susceptible and ≥4 mg/L for resistant (23). In contrast, CLSI categorizes an Enterobacterales, P. aeruginosa, or A. baumannii isolate as resistant if it exhibits a colistin or polymyxin MIC value of ≥4 mg/L (10). CLSI does not currently recognize a susceptible category for colistin or polymyxin B against any of these organism groups or species. Rather, isolates from these groups and species that display a colistin or polymyxin B MIC value of ≤2 mg/L are categorized as intermediate (10).

JMI Laboratories followed current CLSI quality assurance practices when performing the susceptibility tests. MIC values were validated by concurrently testing CLSI- and EUCAST-recommended (10, 25) ATCC or National Collection of Type Cultures (NCTC) quality control (QC) reference strains. The QC strains included E. coli ATCC 25922, P. aeruginosa ATCC 27853, K. pneumoniae ATCC BAA 1705, K. pneumoniae ATCC 700603, and E. coli NCTC 13846. QC ranges for the tested reference strains were the criteria published by CLSI or EUCAST (10, 25). The inoculum density during susceptibility testing was monitored by bacterial colony counts.

Data availability.

Data will be made available upon reasonable request.