Abstract
Antibiotic-resistant Enterobacteriaceae and Acinetobacter baumannii are problematic pathogens, with few treatment options for multidrug-resistant (MDR)-A. baumannii and few oral options for extended spectrum β-lactamase (ESBL)-producing and MDR-Enterobacteriaceae. Omadacycline, a newer tetracycline derivative, has activity against some of these pathogens. We tested the in vitro activity of omadacycline against a contemporary collection of over 2,600 consecutive unique clinical isolates of Enterobacteriaceae and A. baumannii, a previous collection of carbapenem-resistant Klebsiella pneumoniae and A. baumannii from a surveillance study in 2013–2014, and a group of K. pneumoniae and A. baumannii isolates with previously defined resistance mechanisms. For the contemporary collection, over 96% of Escherichia coli and 70% of K. pneumoniae isolates were inhibited by omadacycline at ≤4 μg/mL including 95% of E. coli and 49% of K. pneumoniae with presumptive ESBLs. Nearly 90% of A. baumannii were inhibited by omadacycline at ≤4 μg/mL. The omadacycline MIC50/90 was 1/4 μg/mL, 4/>8 μg/mL, and 0.5/8 for E. coli, K. pneumoniae, and A. baumannii, respectively. For the carbapenem-resistant collection of isolates, 56% of A. baumannii were inhibited by omadacycline at ≤4 μg/mL, but only 30% of Klebsiella pneumoniae carbapenemase (KPC)-possessing K. pneumoniae were susceptible. Expression of the efflux gene adeB appeared to affect the activity of omadacycline against A. baumannii, but could not fully explain resistance to this agent. Omadacycline may prove to be a parenteral or oral option for some infections due to ESBL-producing Enterobacteriaceae and carbapenem-resistant A. baumannii, and clinical studies are warranted.
Introduction
Antibiotic resistance is an ongoing problem affecting medical centers around the world. The WHO has identified multidrug-resistant (MDR) Klebsiella pneumoniae and Acinetobacter baumannii as Gram-negative pathogens posing serious health care threats. 1 Because of limited therapeutic options for many MDR pathogens, there has been a renewed interest in tetracyclines and their derivatives. A new formulation of injectable minocycline was approved by the Food and Drug Administration (FDA) in 2015 because of its activity against some strains of MDR A. baumannii. Over 70% of MDR A. baumannii isolates from North America were reportedly susceptible to minocycline, 2 but evidence of clinical efficacy in humans is limited. Tigecycline has greater in vitro activity than older tetracyclines against some strains of carbapenem-resistant K. pneumoniae and A. baumannii. The use of tigecycline has been limited by pharmacokinetic issues, possible higher in-hospital mortality rates, and lower microbiologic cure rates. 3 Eravacycline, a newer fluorocycline antibiotic, also has enhanced activity against carbapenem-resistant K. pneumoniae and A. baumannii and has proven to be effective in the treatment of intraabdominal infection. 4 However, it was less effective than comparators in urinary tract infections; its efficacy for pneumonia and other infections remains to be determined.
Omadacycline is a novel derivative of minocycline and the first aminomethylcycline antibiotic. 5 It was approved by the FDA in late 2018 for the treatment of community-acquired bacterial pneumonia and acute bacterial skin and skin structure infections. As pointed out in a recent editorial, there are several available options for treating those infections. 6 Omadacycline has demonstrated in vitro activity against some strains of carbapenem-resistant Enterobacteriaceae and carbapenem-resistant A. baumannii. 7 A potential role for omadacycline in the management of MDR Gram-negative infections would be of great interest. In this report, we determined the activity of omadacycline against a recent collection of Gram-negative clinical isolates from several New York City hospitals and assessed its activity against a collection of strains with well-characterized resistance mechanisms.
Materials and Methods
Bacterial isolates
The study included three groups of isolates. The first group (Group 1) consisted of clinical isolates collected in 2017 at seven hospitals in Brooklyn, NY. For a 3 months period, all isolates of Escherichia coli, K. pneumoniae, Enterobacter spp., and A. baumannii were collected. Only one isolate of each species was collected per patient. A total of 1,878 E. coli, 518 K. pneumoniae, 172 Enterobacter spp., and 47 A. baumannii were tested. The second group (Group 2) consisted of an older collection of 79 carbapenem-resistant isolates of A. baumannii and 111 klebsiella pneumoniae carbapenemase (KPC)-possessing isolates of K. pneumoniae. These resistant isolates were a subset of a similar surveillance study conducted at 11 New York City hospitals in 2013–2014.8,9 Finally, Group 3 consisted of 34 isolates of K. pneumoniae and 39 A. baumannii selected from surveillance studies conducted at New York City hospitals between 2001 and 2006. These isolates represented some of the dominant clones prevalent at regional hospitals and had previous characterization of resistance mechanisms including the genetic expression of β-lactamase, porin, and efflux genes.10,11
Susceptibility testing
Omadacycline powder was provided by Paratek Pharmaceuticals (King of Prussia, PA). Minimum inhibitory concentrations (MICs) of omadacycline and most antibiotics were determined by the Clinical and Laboratory Standards Institute (CLSI) agar dilution method, which facilitated testing of large numbers of isolates. 12 The CLSI broth microdilution method was used for tigecycline. 12 E. coli ATCC 25922 was included as a control. The FDA recommended breakpoint for K. pneumoniae and E. cloacae (4 μg/mL) 13 was used to determine susceptibility to omadacycline for all Enterobacteriaceae. The FDA recommended breakpoint (2 μg/mL) was used to determine susceptibility of Enterobacteriaceae to tigecycline. For E. coli, the European Committee on Antimicrobial Susceptibility Testing (EUCAST) recommended tigecycline breakpoint (0.5 μg/mL) was also used.13,14 For all other antibiotics, CLSI susceptibility breakpoints were used.
Statistical analysis
Student's t-test, Fisher's exact test, and multiple linear regression analysis were used to compare the susceptibilities of isolates with various resistance mechanisms. A two-tailed p value of ≤0.05 was considered significant.
Results
Group 1, contemporary clinical isolates
Overall, 2,615 Gram-negative clinical isolates were collected during the 3 months study period in 2017 (Tables 1 and 2). For E. coli, minocycline had slightly greater activity than tetracycline and doxycycline with 84% susceptible. Tigecycline had the lowest MICs, with 96% inhibited by 0.5 μg/mL or less and 99.5% by 2 μg/mL or less. Omadacycline had slightly better activity than minocycline, with nearly 97% inhibited by 4 μg/mL or less. Four isolates possessed blaKPC; two of four were susceptible to tetracycline and doxycycline and three of four were susceptible to minocycline and omadacycline. There were 224 E. coli isolates (11.9%) presumed to have extended spectrum β-lactamase (ESBLs) (ceftazidime MIC >1 μg/mL). For these isolates, susceptibility rates were 45%, 46%, and 83% to tetracycline, doxycycline, and minocycline, respectively. Omadacycline inhibited 95% at 4 μg/mL and tigecycline inhibited 92% and 99.1% of presumed ESBL isolates at 0.5 and 2 μg/mL, respectively.
Susceptibility Results for Omadacycline and Comparators
CLSI breakpoints used except as indicated.
Using FDA breakpoint.
Using EUCAST breakpoint.
% MIC ≤4 μg/mL.
CLSI, Clinical and Laboratory Standards Institute; EUCAST, European Committee on Antimicrobial Susceptibility Testing; FDA, Food and Drug Administration; KPC, Klebsiella pneumoniae carbapenemase; MIC, minimum inhibitory concentration.
For K. pneumoniae, the activity of omadacycline was fairly similar to that of tetracycline, doxycycline, and minocycline with 70% susceptible (Tables 1 and 2). Tigecycline MICs were ∼8-fold lower than omadacycline versus K. pneumoniae. Twenty isolates (3.9%) possessed blaKPC. Of the KPC-possessing isolates, only 10% were susceptible to minocycline and 15% to omadacycline (MIC50/90 8/>8 μg/mL for both). Tigecycline MICs were 4- to 8-fold lower than omadacycline for the KPC-possessing isolates with 100% inhibited by 2 μg/mL (MIC50/90 1/2 μg/mL). Seventy-one isolates lacking blaKPC had ceftazidime MICs >1 μg/mL and were presumed to possess ESBLs. For these isolates, susceptibility rates were 37.5%, 37.5%, 44%, and 49% to tetracycline, doxycycline, minocycline, and omadacycline, respectively. Tigecycline inhibited 89% of presumed ESBL isolates at 2 μg/mL.
MIC Distributions for Omadacycline and Other Agents
There were 172 isolates of Enterobacter spp. (104 E. cloacae, 59 E. aerogenes, and 9 other species). As with K. pneumoniae, the activity of omadacycline was similar to the older tetracyclines and the MICs of tigecycline were ∼8-fold lower (Tables 1 and 2).
There were 47 isolates of A. baumannii (Tables 1 and 2), of which 49% were carbapenem-nonsusceptible. The activity of omadacycline was greater than the older tetracyclines, with 89% inhibited by 4 μg/mL or less. Tigecycline MICs were 2-fold lower than omadacycline, but only 71% were ≤2 μg/mL. Eleven isolates possessed carbapenemase; including 8 with blaOXA23-like, 2 with blaOXA24-like, and one with blaKPC. All but three of these isolates were inhibited by minocycline and omadacycline at 4 μg/mL.
Group 2, KPC-producing K. pneumoniae and carbapenem-resistant A. baumannii collection
From a prior surveillance study conducted in 2013–2014, 111 isolates of KPC-possessing K. pneumoniae were tested (Tables 1 and 2). The activity of omadacycline was similar to doxycycline and minocycline and somewhat less than tetracycline, with only 30% susceptible to omadacycline. Similar to the Group 1 K. pneumoniae isolates, tigecycline MICs were ∼8-fold lower than omadacycline MICs.
There were 79 isolates of carbapenem-resistant A. baumannii, including 47 isolates (59%) possessing blaOXA23-like (Tables 1 and 2). Overall, 56% of the carbapenem-resistant A. baumannii isolates were inhibited by 4 μg/mL or less of omadacycline. Tigecycline MICs were ≤2 μg/mL in only 38% of isolates. The activity of omadacycline against the subset of OXA23-possessing isolates was identical to the entire group.
Group 3, collection of isolates with well-characterized resistance mechanisms
Susceptibility testing was performed for 34 isolates of K. pneumoniae that had characterization of the presence of β-lactamases and the genetic expression of blaKPC, acrB, ompK35, and ompK36. 10 Of the 34 isolates, 14 possessed blaKPC and 20 possessed SHV-type ESBLs. Omadacycline MICs were higher in ESBL-possessing isolates than those without SHV ESBLs (14.9 ± 2.3 vs. 8.2 ± 2.2 μg/mL, p = 0.04). No relationship was found between omadacycline MICs and the expression of acrB, ompK35, or ompK36, nor with the expression of blaKPC for the KPC-positive isolates.
Testing was also performed on 39 isolates of A. baumannii that had characterization of the presence of β-lactamases and the genetic expression of ompA, adeB, and abeM. 11 Of the 39 isolates, 19 possessed SHV-type ESBLs. Omadacycline MICs were higher in ESBL-possessing isolates than those without SHV ESBLs (11.6 ± 2.2 vs. 4.8 ± 0.9 μg/mL, p = 0.009). No correlation was found between omadacycline MICs and the expression of ompA or abeM. However, MICs tended to be higher in isolates with expression of adeB more than 9 times that of the control than in isolates with lower expression of adeB (15.4 ± 5.7 vs. 6.7 ± 1.0 μg/mL, p = 0.17). Insertional inactivation of the adeB gene was performed in two isolates as previously described, 15 including one isolate with normal adeB expression (1.4 times control) and one isolate with increased adeB expression (42 times control). For the isolate with normal adeB expression, the omadacycline MIC went from 8 to 4 μg/mL after adeB inactivation. For the isolate with increased adeB expression, the omadacycline MIC went from 32 to 4 μg/mL after adeB inactivation.
Discussion
Our data demonstrate that omadacycline has greater in vitro activity against E. coli than the older tetracyclines, including the vast majority of ESBL-producing strains. While tigecycline MICs are lower, the pharmacokinetics of tigecycline and the lack of an oral formulation significantly limit its use. The activity of omadacycline was similar to the older tetracyclines against K. pneumoniae, and nearly half of ESBL-producing strains were susceptible. However, the activity against KPC-producing K. pneumoniae was substantially less. Finally, omadacycline had better activity than the older tetracyclines against A. baumannii, with nearly 90% of isolates inhibited by ≤4 μg/mL. When compared to a study of a large collection of isolates from hospitals in the United States and Europe conducted by the SENTRY program in 2017, omadacycline MICs were similar overall. 16 In both studies, omadacycline had excellent activity against ESBL-producing E. coli and substantial activity against A. baumannii.
The testing of the previously characterized (Group 3) isolates does not provide an explanation for the mechanisms of omadacycline resistance in K. pneumoniae or a full explanation of the mechanisms of resistance in A. baumannii. It does appear that the AdeB efflux system affects the activity of omadacycline against A. baumannii, as evidenced by the sharp drop in omadacycline MIC after adeB gene inactivation in a strain over-expressing this gene. However, some isolates with normal or low expression of adeB had high omadacycline MICs, suggesting the presence of other factors. Additional study is needed to further clarify the mechanisms of omadacycline resistance in these species.
In recent years, several newer agents have become available with activity against KPC-producing Enterobacteriaceae including ceftazidime-avibactam, meropenem-vaborbactam, plazomicin, and most recently imipenem-relebactam. However, options for managing infections due to MDR A. baumannii remain very limited. Colistin and polymyxin B have been the mainstay of therapy for MDR A. baumannii, but questions regarding efficacy, nephrotoxicity and emergence of resistance make them suboptimal choices. 17 Tigecycline has also been used for MDR A. baumannii infections, but pharmacokinetic limitations, possible higher mortality rates, and emergence of resistance limit its use as well.17,18 Eravacycline, a fluorocycline agent recently available in the United States, has 4-fold greater in vitro activity than tigecycline against A. baumannii. 15 Pharmacokinetic studies suggest that eravacycline should achieve adequate concentrations in the lungs, the site of many A. baumannii infections. 19 Clinical data on the use of eravacycline for human respiratory infections are pending, and an oral formulation is not available. In addition to A. baumannii, there is a need for oral agents to treat infections caused by ESBL-producing E. coli and K. pneumoniae. Oral options for these infections are limited due to the frequent coexistence of resistance to fluoroquinolones and trimethoprim-sulfamethoxazole. Antibiotics with activity against these pathogens, particularly orally bioavailable agents, are sorely needed.
This study suggests a possible role for omadacycline in the management of some infections caused by A. baumannii and ESBL-producing Enterobacteriaceae, particularly as a possible oral option for therapy. Additional studies are warranted to evaluate the clinical utility of omadacycline for these pathogens.
Footnotes
Disclosure Statement
No competing financial interests exist.
Funding Information
Funding for the study was provided by Paratek Pharmaceuticals, Inc.
