In vitro activity of apramycin (EBL-1003) in combination with colistin, meropenem, minocycline or sulbactam against XDR/PDR Acinetobacter baumannii isolates from Greece

Abstract Objectives To evaluate the in vitro activity of the combination of apramycin with colistin, meropenem, minocycline or sulbactam, against some well-characterized XDR Acinetobacter baumannii clinical isolates from Greece, to understand how apramycin can be best incorporated into clinical practice and optimize effectiveness. Methods In vitro interactions of apramycin (0.5×, 1× and 2× the MIC value) with colistin (2 mg/L), meropenem (30 mg/L), minocycline (3.5 mg/L) or sulbactam (24 mg/L) were tested using time–kill methodology. Twenty-one clinical A. baumannii isolates were chosen, exhibiting apramycin MICs of 4–16 mg/L, which were at or below the apramycin preliminary epidemiological cut-off value of 16 mg/L. These isolates were selected for a range of colistin (4–32 mg/L), meropenem (16–256 mg/L), minocycline (8–32 mg/L) and sulbactam (8–32 mg/L) MICs across the resistant range. Synergy was defined as a ≥2 log10 cfu/mL reduction compared with the most active agent. Results The combination of apramycin with colistin, meropenem, minocycline or sulbactam was synergistic, at least at one of the concentrations of apramycin (0.5×, 1× or 2× MIC), against 83.3%, 90.5%, 90.9% or 92.3% of the tested isolates, respectively. Apramycin alone was bactericidal at 24 h against 9.5% and 33.3% of the tested isolates at concentrations equal to 1× and 2× MIC, while the combination of apramycin at 2× MIC with colistin, meropenem or sulbactam was bactericidal against all isolates tested (100%). The apramycin 2× MIC/minocycline combination had bactericidal activity against 90.9% of the tested isolates. Conclusions Apramycin combinations may have potential as a treatment option for XDR/pandrug-resistant (PDR) A. baumannii infections and warrant validation in the clinical setting, when this new aminoglycoside is available for clinical use.


Introduction
Acinetobacter baumannii is an important cause of healthcareassociated infections, particularly in ICUs, 1 known for its ability to acquire resistance to several antimicrobial agents. 2 Most common A. baumannii infections consist of bloodstream infections (BSIs) and hospital-acquired pneumonia (HAP), including ventilatorassociated pneumonia (VAP). 3ecent data from the European Antimicrobial Resistance Surveillance Network (EARS-Net) show a large increase of 57% in Acinetobacter spp.BSI in the EU and European Economic Area (EEA) in the first years of the COVID-19 pandemic (2020- 21)   compared with 2018-19. 4Additionally, among isolates from ICUs in countries with high (≥50%) prevalence of carbapenemresistant Acinetobacter spp., carbapenem resistance increased from 73.2% to 85.6% over this 3 year period. 4ue to the lack of effective treatment options, infections caused by carbapenem-resistant A. baumannii (CR-Ab) are difficult to treat, with high mortality rates. 5The WHO has designated CR-Ab as a priority pathogen with an urgent need for additional research and development of new antibiotics. 6Apramycin (EBL-1003) has been proposed as a possible next-generation aminoglycoside, and it is currently the only new aminoglycoside in clinical development for the treatment of Gram-negative systemic infections. 7,8Based on its unique chemical structure, comprising an unusual bicyclic octose moiety, apramycin evades almost all clinically relevant aminoglycoside-modifying enzymes (AMEs) and is also unaffected by 16S rRNA-methyltransferase (RMTase)-mediated pan-aminoglycoside resistance. 9oreover, although apramycin is not available for human use yet, it has been used in veterinary infectious diseases since the 1980s and rare apramycin resistance has been reported due to aac(3)-IV. 9 Phase I trial has reported that apramycin in healthy volunteers is 'safe and well tolerated' and the PK/PD modelling studies predicted an 8-16 mg/L target attainment at a human dose of 30 mg/kg. 10,11 few studies have evaluated the in vitro activity of apramycin against A. baumannii isolates and showed that is highly active against MDR, XDR and pandrug-resistant (PDR) clinical isolates.9,[12][13][14][15][16] In time-kill studies, apramycin demonstrated rapid bactericidal activity within 1-2 h of antibiotic exposure at 2× and 4× MIC in three A. baumannii isolates with broth microdilution (BMD) MICs of 2, 16 and 64 mg/L, while antibiotic exposure at 1× MIC demonstrated rapid bactericidal activity only in the isolate with an MIC of 64 mg/L.17 In the present study, the in vitro activity of the combination of apramycin with colistin, meropenem, minocycline or sulbactam, against some well-characterized XDR A. baumannii clinical isolates from Greece was evaluated, to understand how apramycin can be best incorporated into clinical practice and optimize effectiveness.

Static-concentration time-kill experiments
In vitro interactions of apramycin with colistin, meropenem, minocycline or sulbactam were tested using time-kill methodology.Twenty-one clinical A. baumannii isolates were chosen, exhibiting apramycin MICs of 4-16 mg/L, which were at or below the apramycin preliminary epidemiological cut-off (ECOFF) value of 16 mg/L proposed by Juhas et al. 9 These isolates were selected for a range of colistin (4-32 mg/L), meropenem  1).
Apramycin was tested at concentrations of 0.5×, 1× and 2× the MIC value.Colistin was tested at a concentration of 2 mg/L (target steady-state colistin concentration when initiating therapy, achieved by 12 h dosing schedule). 20Meropenem was tested at a concentration of 30 mg/L, simulating the C max of a prolonged (3 h) infusion regimen of 1 g. 21Minocycline and sulbactam were tested at 3.5 and 24 mg/L, respectively, which represent the C max of a 200 mg oral single-dose tablet for minocycline and 1.0 g for three consecutive IV doses for sulbactam. 22,23ubes containing CAMHB (Becton Dickinson & Co., Sparks, MD, USA) and each antibiotic alone or in combination were inoculated with 10 5 cfu/mL of the studied strain and were quantitatively subcultured at 0, 1, 3, 5 and 24 h of incubation for viable colony counts.At the above time intervals, an aliquot of 0.1 mL was removed from each tube, serially diluted (seven 10-fold dilutions) and plated on MacConkey agar (Becton Dickinson) plates.Dilution was expected to minimize any probable antibiotic carry-over effect.The results were expressed as log 10 cfu/mL.A tube without antibiotic was also included in each experiment as a growth control.The lower limit of detection was 1.6 log 10 cfu/mL.
Synergy was defined as a ≥2 log 10 decrease in cfu/mL between the combination and the most active single agent at 24 h, with the number of surviving organisms in the presence of the combination being at least 2 log 10 cfu/mL below the number of organisms in the starting inoculum.Antagonism was defined as a ≥2 log 10 increase in cfu/mL between the combination and the most active single agent.All other interactions were characterized as indifferent.The bactericidal activity of single antibiotics or combinations was defined as a ≥3 log 10 reduction in the cfu/mL of the initial inoculum.All studies were conducted in duplicate and the combined data are presented as mean bacterial density (cfu/mL) for all isolates.

Results
MICs of apramycin, colistin, meropenem, minocycline and sulbactam, as determined by the BMD method, are presented in Table 1.Changes in bacterial density from 0 to 24 h for apramycin 0.5×, 1× and 2× the MIC value, colistin 2, meropenem 30, minocycline 3.5 and sulbactam 24 mg/L alone and in combination, for each isolate tested, are presented in Tables 2-5.
Apramycin alone was bactericidal at 24 h against two (9.5%) and seven (33.3%) tested isolates at concentrations equal to 1× and 2× MIC, respectively (Table 4), while at concentrations 4× and 8× MIC was bactericidal against 16 (76.2%)and all (100%) isolates (data not shown).Apramycin demonstrated rapid bactericidal activity within 3 h of antibiotic exposure at 2× MIC, exhibiting an average of Δlog 10 cfu/mL between initial and final inoculum of −3.19, which was increased to −3.47 at 5 h but decreased at −0.73 at 24 h; Tables S1-S5 and Figure S1 (available as Supplementary data at JAC Online).

Apramycin/meropenem
All strains were resistant to meropenem, according to the BMD method [MIC values of all strains >16 mg/mL, (Table 1)].The combination of apramycin with meropenem was synergistic, at least at one of the concentrations of apramycin (0.5×, 1× or 2× MIC), against 19 of the 21 isolates tested (90.5%) (Table 4).One of the two isolates exhibiting no synergy (AC-391), had an apramycin MIC of 16 mg/L.Bactericidal activity was observed with apramycin at 16 (1× MIC) and 32 mg/L (2× MIC) alone or with either of the combinations with meropenem.Apramycin at 8 mg/L (0.5× MIC) was synergistic with meropenem at 5 h but regrowth occurred at 24 h (data not shown).The second isolate (AC-709) exhibited a meropenem MIC of 64 mg/L and an apramycin MIC of 8 mg/L and no synergy or bactericidal/bacteriostatic activity was observed with the drugs alone or with either of the combinations with apramycin at 0.5× or 1× MIC.Bactericidal activity was observed for apramycin at 2× MIC alone and in combination with meropenem.The combination of apramycin at 0.5× MIC, 1× MIC or 2× MIC with meropenem was synergistic against 9 (42.9%), 17 (81.0%)or 14 (66.7%) of the tested isolates (Table 4), exhibiting a decrease in log 10 cfu/mL of 3.1-7.0(median 6.6), 2.0-7.2(median 6.5) or 3.3-6.8(median 5.9), respectively, compared with the most active compound.The difference between the mean starting inoculum and the mean viable cell count (Δlog 10 cfu/mL) over time of all A. baumannii isolates tested is presented in Figure S3.The combination of apramycin at 2× MIC with meropenem had bactericidal activity against all isolates tested (100%).

Apramycin/minocycline
Eleven strains were resistant to minocycline according to the BMD method [MIC values 8-32 mg/mL, (Table 1)].The combination of apramycin with minocycline was synergistic, at least at one of the concentrations of apramycin (0.5×, 1× or 2× MIC), against 10 out of 11 resistant isolates tested (90.9%), with the isolate exhibiting no synergy having a minocycline MIC of 32 mg/L.Minocycline at 3.5 mg/L was bactericidal against all isolates exhibiting MICs of ≤1 mg/L (data not shown).The combination of apramycin at 0.5× MIC, 1× MIC or 2× MIC with minocycline was synergistic against 7 (63.6%) of the tested isolates in all cases (Table 5), exhibiting a decrease in log 10 cfu/mL of 2.3-7.0 (median 5.0), 2.0-6.3 (median 5.3) or 3.3-6.6(median 5.3), respectively, compared with the most active compound.The difference between the mean starting inoculum and the mean viable cell count (Δlog 10 cfu/mL) over time of all A. baumannii isolates tested is presented in Figure S4.The combination of apramycin at 2× MIC with minocycline had bactericidal activity against 10/11 isolates tested (90.9%).

Discussion
As a consequence of limited therapeutic options for the treatment of A. baumannii infections, clinicians often resort to antibiotic combinations, in the hope of improving the activity of available agents.The combination of an aminoglycoside with another agent might promote the permeabilizing effect and enhance the periplasmic target site penetration of the second agent used in combination. 24,25Apramycin, the only new aminoglycoside in clinical development for the treatment of Gram-negative systemic infections, 7,8 retains significant activity against a broad spectrum of MDR and XDR clinical isolates, including CR-Ab, with the presence of various aminoglycoside resistance-conferring genes, such as AMEs and RMTase-coding genes, not adversely affecting the distribution of apramycin MICs. 9R-Ab is a critical Gram-negative organism included in the WHO priority pathogen list, with no optimal therapeutic strategy for the management of related infections, 26 with sulbactam, meropenem, minocycline, tigecycline and polymyxins, the last-resort antibiotics in recent decades, for critically ill patients. 27In a recent study, conducted in Greece, the dissemination of XDR and PDR bla OXA-23 -armA-harbouring A. baumannii isolates, corresponding to international clone (IC) II (87.8%), was associated with a dramatic decrease of in vitro activity of colistin to only 15.5%, leaving minocycline as the most active A negative sign denotes a reduction of inoculum compared with time 0; values in bold are consistent with bactericidal or bacteriostatic activity.b Apramycin at the concentration used was bactericidal after 24 h.c A reduction in the viable counts of bacterial cells of ≥2 log 10 cfu/mL was observed with the combination after 1 h, but apramycin was bactericidal at 3-24 h.
Apramycin combinations against A. baumannii agent, with 18.8% of the isolates exhibiting an MIC of ≤4 mg/L. 16nterestingly, a promising in vitro activity was verified for apramycin (EBL-1003), which exhibited MIC 50 /MIC 90 of 4/8 mg/L and 100% susceptibility according to the preliminary ECOFF (16 mg/L) defined by Juhas et al for A. baumannii. 9,16However, aminoglycoside monotherapy can lead to unfavourable clinical outcomes due to rapid emergence of resistance, while nephrotoxicity is a serious concern with prolonged use. 28,29n this study, excellent synergy was observed for apramycin and colistin, meropenem, minocycline or sulbactam combinations against A. baumannii XDR and PDR isolates despite resistance of the tested isolates to the second agent.The combinations of apramycin and colistin and apramycin and sulbactam exhibited bactericidal activity at all tested concentrations of apramycin, against XDR and PDR A. baumannii isolates.The bactericidal activity began as early as 3 h of incubation and continued up until 24 h.The combinations with meropenem and minocycline exhibited bactericidal activity at 1× and 2× MIC concentrations of apramycin from 3 to 24 h.
Literature review on apramycin combinations has revealed that only two laboratory research papers are currently available.The first laboratory study of Brennan-Krohn and Kirby 30 reported that apramycin/colistin combination was not synergistic against the PDR Klebsiella pneumoniae Nevada strain, which was an NDM-1-producing and colistin-resistant (MIC 16 mg/L) strain. 30n the second study, colistin/apramycin synergy was observed against 15.3% of carbapenemase-producing and colistinresistant K. pneumoniae strains, whereas other strains had antagonistic (30.7%) or additive (54%) effects. 31Additionally, the meropenem/apramycin combination was synergistic against 52%, while an additive effect was observed against 31% of the isolates tested. 31In both studies, synergy testing was conducted with the chequerboard microdilution method.
This study is a laboratory study that investigates the use of apramycin in combination with colistin, meropenem, minocycline or sulbactam against XDR/PDR A. baumannii blood isolates, with the time-kill method.
Proposed therapeutic approaches for CR-Ab severe infections include treatment with combination of two in vitro active agents, with preferable combinations, an aminoglycoside or a polymyxin with high-dose ampicillin/sulbactam or high-dose tigecycline or high-dose minocycline or in vitro synergy testing for selection of a combination scheme, when two in vitro active agents are not available. 32In the case of PDR CR-Ab infections, cefiderocol is recommended, if available, or the triple-drug combination of colistin, high-dose meropenem and high-dose ampicillin/sulbactam. 32 The recent ESCMID guidelines recommend double-agent combination therapy when CR-Ab is susceptible to more than one antibiotic, 33 while conditionally recommending against cefiderocol due to low certainty of evidence.
Aminoglycosides are among the active in vitro antibiotics commonly suggested for combination therapy against severe and high-risk CR-Ab infections.Thus, the high rate of in vitro activity of apramycin against the XDR/PDR armA-harbouring A. baumannii isolates is an interesting feature of this agent. 16lthough all A. baumannii strains (100%) were susceptible to apramycin, apramycin alone at 2× MIC exhibited bactericidal activity against only seven strains (33.3%) and only at 24 h.On the contrary, all (100%) colistin, meropenem and sulbactam combinations with apramycin at 2× MIC exhibited bactericidal activity at 24 h.The combination of apramycin (2× MIC) with minocycline

Galani et al.
was bactericidal at 24 h against 10 of the 11 strains tested (90.9%), with one of these isolates exhibiting a minocycline MIC of 32 mg/L.Synergy rates were high for all combinations, ranging from 83.3% (for colistin) to 92.3% (for sulbactam), while none of the apramycin combinations tested in this study was antagonistic towards the A. baumannii strains.
Overall, our results suggest that apramycin combinations may have potential as a treatment option for XDR and PDR A. baumannii infections and warrant validation in the clinical setting, when this new aminoglycoside is available for clinical use.

aA
negative sign denotes a reduction of inoculum compared with time 0; values in bold are consistent with bactericidal or bacteriostatic activity.b Apramycin at the concentration used was bactericidal/bacteriostatic after 24 h.c Sulbactam at the concentration used was bactericidal after 24 h.Apramycin combinations against A. baumannii

Table 1 .
Characteristics of the 21 blood A. baumannii isolates used in this study 14Sequence group/clonal lineages.14Galaniet al.

Table 2 .
Interactions of apramycin/colistin combinations and differences between the starting inoculums and the number of residual viable colonies (Δlog 10 cfu/mL) after 24 h of incubation with each combination APR, apramycin; SYN, synergy; IND, indifference.0.5×, 1×, 2×, apramycin concentration expressed as 0.5-, 1-or 2-fold the MIC; COL, colistin 2 mg/L.a A negative sign denotes a reduction of inoculum compared with time 0; values in bold are consistent with bactericidal activity.b Apramycin at the concentration used was bactericidal after 24 h.c Colistin at the concentration used was bactericidal after 24 h.

Table 3 .
Mean difference between viable counts in the presence of the combination and viable counts in the presence of each antibiotic alone at different timepoints (Δlog 10 cfu/mL) Each value represents the average of values calculated for that timepoint for A. baumannii isolates using time-kill studies.A negative sign denotes a reduction of viable counts in the presence of the combination compared with counts in the presence of antibiotic alone; values in bold are consistent with a ≥2 log 10 reduction.APR 0.5×, apramycin (EBL-1003) at 0.5× MIC (mg/L); APR 1×, apramycin (EBL-1003) at 1× MIC (mg/L); APR 2×, apramycin (EBL-1003) at 2× MIC (mg/L); CST, colistin; MEM, meropenem; MIN, minocycline; SUL, sulbactam.

Table 4 .
Interactions of apramycin/meropenem combinations and differences between the starting inoculums and the number of residual viable colonies (Δlog 10 cfu/mL) after 24 h of incubation with each combination a

Table 5 .
Interactions of apramycin/minocycline combinations and differences between the starting inoculums and the number of residual viable colonies (Δlog 10 cfu/mL) after 24 h of incubation with each combination APR, apramycin; 0.5×, 1×, 2×, apramycin concentration expressed as 0.5-, 1-or 2-fold the MIC; MIN, minocycline 3.5 mg/L; SYN, synergy; IND, indifference.a A negative sign denotes a reduction of inoculum compared with time 0; values in bold are consistent with bactericidal or bacteriostatic activity.b Apramycin at the concentration used was bactericidal after 24 h.