In Vitro Activity of AZD0914, a Novel Bacterial DNA Gyrase/Topoisomerase IV Inhibitor, against Clinically Relevant Gram-Positive and Fastidious Gram-Negative Pathogens

Following the introduction of nalidixic acid into clinical practice in the 1960s, fluoroquinolones have undergone profound structural modifications, leading to the development of numerous molecules in this class (1–3). Modifications to the quinolone structure, particularly the addition of fluorine(s) at key positions, have provided compounds with increased potency, broader spectrum of activity, and acceptable safety profiles. However, safety has been problematic in fluoroquinolone development (4–6). Several potent compounds have either been determined to be unsuitable for human use due to unacceptable toxicities, have required black box warnings on their labels, have been restricted to topical applications, or have resulted in removal from the market (7–9). Resistance development to fluoroquinolones is also becoming a significant concern among several Gram-positive and Gram-negative pathogens, including Staphylococcus aureus, Escherichia coli, and Neisseria gonorrhoeae.

AZD0914 is a new orally administered spiropyrimidinetrione bacterial DNA gyrase inhibitor that demonstrates a novel mode of inhibition distinct from that of fluoroquinolones (10, 11). AZD0914 is a selective and potent inhibitor of the supercoiling and decatenation activity of DNA gyrase and topoisomerase IV, with the ability to overcome fluoroquinolone resistance by the inhibition of DNA biosynthesis through the accumulation of double-stranded cleaved DNA bound to the tetramer topoisomerase II.

AZD0914 has a primary spectrum of activity that includes clinically relevant Gram-positive and fastidious Gram-negative bacterial species, including N. gonorrhoeae (12–15). Published in vitro data on Chlamydia trachomatis and Chlamydia pneumoniae have shown that AZD0914 also has activity against these species (16). Its activity is maintained against strains with common fluoroquinolone resistance mutations in gyrase and topoisomerase IV within the quinolone resistance-determining region (QRDR) (12). Resistance to other key antimicrobial classes, such as β-lactams, macrolides, and glycopeptides, also do not diminish the activity of AZD0914. Currently, AZD0914 is being investigated in phase 2 trials for the treatment of uncomplicated N. gonorrhoeae infections (14, 15).

In this study, the in vitro activity of AZD0914 against key bacterial groups isolated from intra-abdominal, urinary tract, skin and soft tissue, and respiratory tract infections collected in a 2013 global surveillance survey was analyzed and compared to that of levofloxacin, moxifloxacin, and other nonfluoroquinolone compounds.

(This study was presented, in part, at the 54th Interscience Conference on Antimicrobial Agents and Chemotherapy in 2014 [12].)

Clinical isolates (n = 21,152) from hospitalized patients were collected at 169 medical centers in 39 countries distributed across North America, Latin America, Europe, Asia-Pacific, and Middle East/Africa during 2013. Isolates were obtained from specimens collected from patients with documented intra-abdominal infections (IAI), urinary tract infections (UTI), skin and soft tissue infections (SSTI), or lower respiratory tract infections (LRTI). Only one isolate per patient infection episode was included in the surveillance program. Confirmation of isolate identification to the species level was done using matrix-assisted laser desorption ionization–time of flight (MALDI-TOF) mass spectrometry (Bruker Daltonics, Bremen, Germany); susceptibility testing by broth microdilution, according to current CLSI guidelines (17–19), was both performed at a central laboratory (International Health Management Associates [IHMA], Inc., Schaumburg, IL). Of the 21,152 isolates collected, the percentage of isolates per region was Europe, 48.5%; Asia-Pacific, 21.5%; North America, 11.1%; Latin America, 10.9%; and Middle East/Africa, 8.0%. The source of the organisms by infection type included IAI, 7.6%; LRTI, 31.3%; SSTI, 53.1%; UTI, 7.8%; and unknown source or not given, 0.2%.

MICs were determined using custom frozen broth microdilution panels prepared at IHMA, Inc. All broth microdilution testing aspects, including panel manufacture, inoculation, incubation, and interpretation, were conducted according to current CLSI guidelines (17–19). The in vitro activities of levofloxacin, moxifloxacin, and other compounds were tested concurrently used as comparator agents to AZD0914. The activity of AZD0914 was analyzed against susceptible and resistant populations for comparative purposes where appropriate. Quality control testing (QC) using appropriate ATCC strains was performed on each day of testing, according to CLSI guidelines (17). The QC ranges for AZD0914 have been approved by the CLSI and are documented in the June 2014 CLSI antimicrobial susceptibility testing meeting minutes (http://clsi.org/standards/micro/microbiology-files/).

The in vitro activity of AZD0914 and comparator antibiotics was determined against 11,680 isolates of S. aureus (Table 1). These included methicillin-susceptible S. aureus (MSSA), methicillin-resistant S. aureus (MRSA), levofloxacin-susceptible (MIC, ≤1 mg/liter), and levofloxacin-resistant (MIC, ≥4 mg/liter) isolates. The MIC₉₀ of AZD0914 against S. aureus was 0.25 mg/liter, with equal potency against MRSA and MSSA. In contrast, the MIC₉₀ of both levofloxacin and moxifloxacin was >2 mg/liter for S. aureus. Whereas 93.2% of the MSSA isolates were susceptible to levofloxacin (MIC₉₀, 0.5 mg/liter), only 29.9% of the MRSA isolates were levofloxacin susceptible. Levofloxacin susceptibility did not affect the activity of AZD0914, with MIC₉₀s of 0.25 mg/liter against both levofloxacin-susceptible and -resistant isolates. One levofloxacin-resistant MSSA isolate collected in Europe resulted in an AZD0914 MIC of 1 mg/liter. Regardless of specimen source or geographic source, no major differences were observed in the activity of AZD0914 against S. aureus (data not shown).

AZD0914 and comparator antibiotics were tested against 1,923 coagulase-negative staphylococci (CoNS) (Table 1), including methicillin-susceptible, methicillin-resistant (74% of the isolates), levofloxacin-susceptible, and levofloxacin-resistant isolates. The overall MIC₉₀ of AZD0914 against CoNS was 0.25 mg/liter, which was ≥16-fold lower than that for levofloxacin and moxifloxacin (both MIC₉₀, >2 mg/liter). There was no change in the MIC₉₀s of AZD0914 when the population was segregated according to their methicillin resistance status (data not shown), but the levofloxacin MIC₉₀s against the methicillin-susceptible and methicillin-resistant CoNS isolates were 0.5 and >2 mg/liter, respectively. Resistance to levofloxacin did not affect the activity of AZD0914 (Table 1). The AZD0914 MIC₉₀ against the CoNS collected from each of the five geographic regions was identical, at 0.25 mg/liter. The in vitro activities of AZD0914 and comparator antibiotics specifically against 120 isolates of Staphylococcus lugdunensis are also shown in Table 1. The MIC₉₀ of AZD0914 (0.5 mg/liter) against S. lugdunensis was 2-fold higher than that against S. aureus. All isolates were susceptible to levofloxacin and moxifloxacin, with MIC₉₀s of 0.5 mg/liter and 0.12 mg/liter, respectively.

The in vitro activities of AZD0914 and comparators against S. pneumoniae, including penicillin-susceptible and nonsusceptible isolates, are shown in Table 2. The overall MIC₉₀ of AZD0914 against S. pneumoniae was 0.25 mg/liter and, unlike levofloxacin and moxifloxacin, the activity of AZD0914 was not affected by the penicillin resistance status of the population tested. All of the isolates had AZD0914 MICs of ≤0.5 mg/liter, with 99.4% of them being ≤0.25 mg/liter. Among the isolate collection, 37.6% were erythromycin resistant, and AZD0914 MICs were not affected by resistance to macrolides (Table 2). Regardless of specimen or geographical source, no major differences were observed in the activity of AZD0914 against isolates of S. pneumoniae (data not shown). The in vitro activities of AZD0914 and comparator antibiotics against β-hemolytic streptococci (βHS) and viridans streptococci also are shown in Table 2. The MIC₉₀s obtained with AZD0914 and moxifloxacin were 0.25 mg/liter for all species. Against these same species, levofloxacin was 4- to 8-fold less active, with MIC₉₀s of 1 to 2 mg/liter. The potency of AZD0914 against the βHS did not differ between the species tested or across the five geographic regions. Of the 1,358 Streptococcus pyogenes isolates tested, 11% were erythromycin resistant, and AZD0914 was equally active against these subpopulations, with only 4 isolates (2 erythromycin susceptible and 2 erythromycin resistant) having an AZD0914 MIC of 0.5 mg/liter (data not shown). The activity of AZD0914 was also similar against erythromycin-susceptible and erythromycin-resistant (32.2%) Streptococcus agalactiae isolates, with MIC₉₀s of 0.25 mg/liter (data not shown). In contrast, whereas levofloxacin was effective against the erythromycin-susceptible isolates (MIC₉₀, 2 mg/liter; 94.2% susceptible), its activity was decreased against the erythromycin-resistant S. agalactiae isolates (MIC₉₀, >8 mg/liter; 87.3% susceptible). Isolates of erythromycin-resistant β-hemolytic streptococci (610 isolates) showed a lower MIC₉₀ with AZD0914 and moxifloxacin (0.25 mg/liter) than that for levofloxacin (MIC₉₀, 2 mg/liter). Against β-hemolytic streptococci and viridans streptococci, of which 97% were susceptible to levofloxacin, the MIC₉₀ of both AZD0914 and moxifloxacin was 0.25 mg/liter, regardless of specimen source or geographical region (data not shown).

The activity of AZD0914 against Enterococcus faecalis, 16 (1.3%) isolates of which were vancomycin resistant, is shown in Table 3. The overall MIC₉₀ of AZD0914 against E. faecalis was 1 mg/liter, with 98.8% of the isolates giving an AZD0914 MIC of ≤2 mg/liter. With only a 2-fold dilution difference observed in the MIC₅₀s and MIC₉₀s between vancomycin-susceptible and -resistant isolates, vancomycin status did not affect the activity of AZD0914. The MIC₉₀ for levofloxacin was >8 mg/liter against E. faecalis, with 70.4% of the isolates being susceptible. AZD0914 was equally active against the levofloxacin-susceptible and levofloxacin-resistant E. faecalis isolates. Compared to E. faecalis, AZD0914 and comparator antibiotics demonstrated reduced activity against Enterococcus faecium (Table 3). The overall MIC₉₀ of AZD0914 against E. faecium was 8 mg/liter. Similar activity was observed in the vancomycin-susceptible and -resistant populations, with AZD0914 MIC₉₀s of 8 and 16 mg/liter, respectively. The MIC₉₀ of levofloxacin was >8 mg/liter for all E. faecium isolates tested, irrespective of vancomycin susceptibility, with only 17.5% of vancomycin-susceptible isolates and none of the vancomycin-resistant isolates being susceptible to levofloxacin. AZD0914 had similar activity against the levofloxacin-susceptible and levofloxacin-resistant E. faecium isolates, with MIC₉₀s of 8 mg/liter for both groups.

As shown in Fig. 1, the AZD0914 MIC distributions against three key organism groups (MRSA, S. pneumoniae, and E. faecalis) were quite narrow, with the majority of MICs ranging from 0.12 to 0.25 mg/liter for both MRSA and S. pneumoniae, and from 0.5 to 1 mg/liter for E. faecalis. These results further reflect that regardless of geographic or specimen source, the activity of AZD0914 was consistent with narrow MIC distributions.

The in vitro activities of AZD0914 and comparator antibiotics against fastidious Gram-negative respiratory pathogens are shown in Table 4. The MIC₉₀ of AZD0914 against Haemophilus influenzae was 0.5 mg/liter (MIC range, 0.03 to 2 mg/liter), which included 60 (17.0%) β-lactamase-positive isolates. Most isolates (99.2%) were susceptible to levofloxacin, which had an MIC₉₀ of 0.015 mg/liter. AZD0914 MIC₉₀s were 0.5 mg/liter for isolates from all geographical regions. Among 70 Haemophilus parainfluenzae isolates, the MIC₉₀ of AZD0914 was 2 mg/liter. All isolates were susceptible to levofloxacin (MIC₉₀, 0.5 mg/liter). The overall MIC₉₀ of AZD0914 against Moraxella catarrhalis was 0.25 mg/liter (Table 4). Only two isolates were identified with MICs of 0.5 mg/liter, one β-lactamase-negative isolate and one β-lactamase-positive isolate. The MIC₉₀s for levofloxacin and moxifloxacin were 0.12 and 0.06 mg/liter, respectively.

This is the first comprehensive analysis of the in vitro activity of AZD0914 against a global collection of pathogens associated with numerous infection types. This novel agent exhibited excellent in vitro activity against all species tested, with the exception of E. faecium.

Fluoroquinolones are among the most common antimicrobial agents prescribed for treating patient infections. The use of fluoroquinolones has prompted resistance to develop rather quickly among several highly important bacterial pathogens, both Gram-positive (S. aureus) and Gram-negative (E. coli and N. gonorrhoeae [20–24]). Other species, such as Haemophilus spp. and β-hemolytic streptococci, have been less prone to becoming fluoroquinolone nonsusceptible but are emerging at high rates in some countries (25–31). Hampering the advancement of fluoroquinolones to include other more potent agents has been difficult, due to the safety profile observed for this class. Antimicrobial agents that represent a new class (spiropyrimidinetrione) and target a similar mechanism, namely, enzymes involved with DNA replication, may provide alternative treatment options for bacterial species that have become resistant to current fluoroquinolones.

AZD0914 represents the first compound in a new class of gyrase/topoisomerase inhibitors with potent in vitro activity relative to levofloxacin and moxifloxacin against medically important Gram-positive and fastidious Gram-negative bacterial species, including fluoroquinolone-resistant clinical isolates. Cross-resistance between AZD0914 and other important Gram-positive class agents, including fluoroquinolones, β-lactams, macrolides, and glycopeptides, was not observed. The current indication being pursued for AZD0914 is uncomplicated gonococcal infections with additional in vitro activity demonstrated against other organisms responsible for sexually transmitted infections (STI), including C. trachomatis (13–16).

The findings of this present study, based solely on in vitro activity, indicate that AZD0914 is a strong candidate for continued advancement as a therapeutic agent for infections caused by staphylococci, streptococci, enterococci, E. faecalis, M. catarrhalis, and H. influenzae. The spectrum of activity of AZD0914 suggests that it may have utility beyond STIs.

AZD0914 has the potential to become an important novel antimicrobial agent within a new class that has broad coverage against important Gram-positive and fastidious Gram-negative pathogens isolated from multiple sources of infection.

AstraZeneca Pharmaceuticals provided financial support for this investigation, and all authors provided analysis input and have read and approved the final manuscript.

D.J.B., D.F.S., and M.D.H. are employees of International Health Management Associates, Inc. None of the IHMA authors have personal financial interests in the sponsor of this paper (AstraZeneca Pharmaceuticals). P.A.B. and B.L.M.D.J. are employees and stock holders of AstraZeneca Pharmaceuticals LP.

We gratefully acknowledge the contributions of the clinical trial investigators, laboratory personnel, and all members of the global surveillance program that contributed isolates and information for this study.