In vivo dose response and efficacy of the β-lactamase inhibitor, durlobactam, in combination with sulbactam against the Acinetobacter baumannii-calcoaceticus complex

ABSTRACT Multi-drug resistant (MDR) Acinetobacter baumannii is emerging as a pathogen of increasing prevalence and concern. Infections associated with this Gram-negative pathogen are often associated with increased morbidity and mortality and few therapeutic options. The β-lactamase inhibitor sulbactam used commonly in combination with ampicillin demonstrates intrinsic antibacterial activity against A. baumannii acting as an inhibitor of PBP1 and PBP3, which participate in cell wall biosynthesis. The production of β-lactamases, particularly class D oxacillinases, however, has limited the utility of sulbactam resorting to increased doses and the need for alternate therapies. Durlobactam is a non-β-lactam β-lactamase inhibitor that demonstrates broad β-lactamase inhibition including class D enzymes produced by A. baumannii and has shown potent in vitro activity against MDR A. baumannii, particularly carbapenem-resistant isolates in susceptibility and pharmacodynamic model systems. The objective of this study is to evaluate the exposure-response relationship of sulbactam and durlobactam in combination using in vivo neutropenic thigh and lung models to establish PK/PD exposure magnitudes to project clinically effective doses. Utilizing established PK/PD determinants of %T>MIC and AUC/MIC for sulbactam and durlobactam, respectively, non-linear regressional analysis of drug exposure was evaluated relative to the 24-hour change in bacterial burden (log10 CFU/g). Co-modeling of the data across multiple strains exhibiting a broad range of MIC susceptibility suggested net 1-log10 CFU/g0 reduction can be achieved when sulbactam T>MIC exceeds 50% of the dosing interval and durlobactam AUC/MIC is 10. These data were ultimately used to support sulbactam-durlobactam dose selection for Phase 3 clinical trials.

higher ampicillin-sulbactam dose recommendation has emerged due to the dimin ished sulbactam activity against contemporary isolates of A. baumannii with the emergence of β-lactamase-mediated resistance (10,11).
Durlobactam is a non-β-lactam β-lactamase inhibitor with an extended spectrum of activity compared with other β-lactamase inhibitors demonstrating potent inhibition of Ambler class A, C, and D β-lactamases (10), although it has no significant intrin sic antibacterial activity against A. baumannii on its own (11).Unique to its activity and ultimate utility to address carbapenem resistance in A. baumannii, durlobactam inhibits class D carbapenemases of the OXA family which predominate in these resistant strains.In vitro studies have shown that durlobactam protects sulbactam from hydroly sis by β-lactamases and restores its activity against carbapenem-resistant and MDR A. baumannii (10).Against 4,038 recent, global, clinical isolates of A. baumannii, the addition of durlobactam restored the activity of sulbactam with a shift of the MIC 90 of sulbactam from 64 µg/mL to 2 µg/mL (12).Recent work utilizing in vitro static time kill studies and in vitro pharmacodynamic model systems has established exposure targets and PK/PD indices associated the activity of sulbactam and durlobactam with T>MIC in excess of 50% and an AUC/MIC of 10, respectively, associated with achieving a net 1-log 10 CFU reduction in bacterial burden over 24 hours (13).Utilizing the observed PK/PD indices characterized for sulbactam and durlobactam in vitro, the present study seeks to determine exposure magnitudes associated with achieving bactericidal activity in vivo using murine neutropenic thigh and lung infection models.

In vitro susceptibility testing
MICs of sulbactam and sulbactam in the presence of 4 µg/mL of durlobactam are summarized in Table 1 for the sulbactam-and carbapenem-susceptible A. baumannii ARC2058 and CRAB isolates used in the neutropenic thigh and lung infection models.A reduction in sulbactam MIC was observed for all strains in the presence of 4 µg/mL of durlobactam, with potentiated MICs (MICs of sulbactam in the presence of 4 µg/mL durlobactam) ranging from 0.5 to 16 µg/mL.

In vivo neutropenic lung and thigh infection model validation studies
Murine thigh and lung models were validated with meropenem vs. the carbapenemand sulbactam-susceptible A. baumannii ARC2058 isolate with the results summarized in Table 2. Exposure-response analysis of meropenem vs. A. baumannii ARC2058 in the neutropenic thigh and lung models are shown in Fig. S1.In the thigh model, the mean meropenem unbound percent time rates above MIC (%fT>MIC) associated with

In vivo neutropenic thigh and lung infection model studies with sulbactam in combination with durlobactam vs. CRAB strains
Individual strain %fT>MIC estimates for sulbactam to achieve PK/PD endpoints of 1-log 10 CFU/g reduction, 2-log 10 CFU/g reduction, and the EC 80 vs. CRAB strains are summarized in Table 3 for thigh and lung infection models utilizing a 4:1 dose titration of sulbac tam:durlobactam.Co-modeling of the %fT>MIC sulbactam exposure response data (when administered in combination with durlobactam) across multiple CRAB strains and the sulbactam susceptible strain ARC2058 is shown in Fig. 1 for thigh and lung models.Sulbactam %fT>MIC magnitudes associated with 1-log 10 CFU/g reduction, 2-log 10 CFU/g reduction, and the EC 80 of the co-modeled data are summarized in Table 3. Magnitudes of %fT>MIC were required for 1-log 10 , and 2-log 10 CFU reduction was nearly identical between the mean of the individual PK/PD endpoints determined across all the strains compared with the PK/PD endpoints determined from co-modeling the data.Co-modeling of the durlobactam exposure response data across multiple MDR A. baumannii strains and normalizing the AUCs by the potentiated MIC in the thigh and lung models resulted in Hill-type fits shown in Fig. 2. Correlation coefficients (R 2 ) of 0.86 and 0.91 were observed for thigh and lung models, respectively.For studies incorporat ing a fixed dose of sulbactam to keep %fT>MIC constant throughout the durlobactam dose range (Table S1), an R 2 of 0.82 was observed (Fig. 3).The PK/PD exposure endpoints for durlobactam when sulbactam was administered at a fixed dose are summarized in Table 4.The fAUC/MIC magnitudes were generally consistent in both lung and thigh models utilizing a 4:1 dose ratio as well as titrating durlobactam with a fixed dose of sulbactam.

DISCUSSION
The objectives of the present study were to determine the exposure magnitudes associated with achieving PK/PD endpoints of a net 1-log 10 and 2-log 10 CFU reduction over 24 hours in murine neutropenic thigh and lung models.Dose fractionation studies and dose range finding studies performed in vitro with sulbactam and durlobactam established the PK/PD indices most closely associated with the activity of each agent (13).Similar to most β-lactams, the time for which unbound concentrations of sulbactam exceed the MIC (%fT>MIC) was found to be the index correlated to its activity (14,15).For durlobactam, unbound fAUC/MIC correlated to its activity (13).As observed in the non-linear regression analysis using a Hill-type model with sulbactam and durlobactam exposures, good correlations were observed when co-modeling the data sets across the individual strains.Normalizing the exposure data by MIC, good correlations were observed in Hill-type model fits of the change in log 10 CFU/g from baseline at 24 hours vs. %fT>MIC for sulbactam and fAUC/MIC for durlobactam.Convergence of the dose  response with the addition of the MIC suggests a proportional relationship between inhibitor exposure and the degree of susceptibility.
Both thigh and lung neutropenic infection models were validated using the carbapenem-susceptible isolate A. baumannii ARC2058 and subcutaneous administra tion of meropenem.It was assumed that the %T>MIC parameter for meropenem best correlates with its activity as this driver has been shown to correlate best with efficacy (16).The magnitude of the EC 80 for meropenem versus A. baumannii ARC2058 matched the EC 80 for meropenem versus P. aeruginosa in the same model (17), as well as the projected magnitude of 40%T>MIC proposed for treating critically ill patient popula tions (18).For sulbactam treatment versus A. baumannii ARC2058, %fT>MIC magnitudes required for bactericidal activity (>1-log 10 CFU reduction over 24 hours) were consis tent with the estimated clinical exposures of sulbactam used to successfully treat A. baumannii infections based upon doses and MICs observed in human clinical studies (19)(20)(21)(22) and human PK parameters for sulbactam (23,24) to determine the PK/PD magnitudes associated with sulbactam efficacy in the clinic.An endpoint of stasis (no net reduction of log 10 CFU burden over 24 hours) was not considered in the exposureresponse analysis as bactericidal PK/PD endpoints of 1-and 2-log 10 CFU reduction over 24 hours would be more relevant for patients with these serious infections.For MDR A. baumannii isolates, exposure response data were obtained from thigh and lung studies performed in mice administered a 4:1 dose ratio of sulbactam:durlobactam.The rationale for the 4:1 dose ratio was to match exposures associated with achieving the required T>MIC and AUC/MIC requirements for sulbactam and durlobactam, respectively, that were associated with bactericidal activity within in vitro pharmacodynamic models (13).These exposures corresponded to sulbactam fT>MIC greater than ~70% and durlobac tam AUC/MIC > 10 which were shown to correlate to at least 1-log 10 CFU/mL reduction in 24 hours within the in vitro models.By comparison, in the present in vivo study, a 2-log 10 CFU reduction was achieved when sulbactam exceeded 50% fT>MIC vs. the sulbactam-susceptible A. baumannii isolate ARC2058.One weakness inherent in the study design using a titration of a fixed 4:1 dose of sulbactam:durlobactam was that both component exposures varied relative to the MIC across the dose range; however, %fT>MIC magnitudes were similar when sulbactam was evaluated vs. ARC2058 and when sulbactam was evaluated vs. CRAB isolates in the presence of durlobactam (Table 3).This consistency in sulbactam exposure targets in susceptible vs. CRAB strains suggests there was sufficient exposure of durlobactam to restore the activity of sulbactam.Magnitudes observed in vivo were slightly lower than targeted with %fT>MIC of 20.5 and 31.5 for 1-and 2-log 10 CFU/g reduction, respectively, in the neutropenic thigh model and 31.5 and 50.1 for 1-and 2-log 10 CFU/g reduction, respectively, in the lung model vs. the sulbactam-susceptible strain ARC2058.Exposure magnitudes were generally higher in the lung vs. thigh infection model.Attempts to target sulbactam exposure at a set duration above the MIC for six MDR strains evaluated in the thigh model were challenging given the variability encountered with establishing a modal MIC for these strains.One-dilution differences in microbroth MIC determinations (which are within the error of the MIC assay) combined with PK variability can lead to broad ranges of values observed for PK/PD endpoints such as %fT>MIC or AUC 0-24 /MIC ratio (25,26).Nevertheless, dose titration of durlobactam exposure with sulbactam at %fT>MIC exposures ranging from 18.3 to 43.7% showed fAUC/MIC 0-24 ratios of ~10 and ~30 for 1-log 10 and 2-log 10 CFU reductions from baseline at 24 hours, respectively, in the thigh model as well as in other in vivo and in vitro experiments (13).Interestingly, strains such as A. baumannii ARC5950 and ARC5955 harboring higher SUL-DUR MICs of 8 µg/mL, presumably due to mutations to PBP3 had similar sulbactam %fT>MIC requirements as observed in vitro (13).Although PK/PD exposures were considered in unbound plasma, both sulbactam and durlobactam demonstrated excellent distribution into lung epithelial lining fluid (ELF) of mice with penetration rates of 32% and 63%, respectively Table S2.These values differed from what has been observed clinically (27) where sulbactam and durlobactam penetration was determined to be 50% and 37%, respectively, based on total ELF and plasma AUC exposures.Based on this observation only unbound plasma PK/PD targets were considered for probability of target attainment analysis and Phase 3 dose justification.
In conclusion, sulbactam demonstrated robust bactericidal activity when adminis tered in combination with durlobactam to treat infections caused by CRAB isolates evaluated in murine neutropenic thigh and lung infection models.Using PK/PD indices established from in vitro pharmacodynamic model systems, %T>MIC and AUC 0-24 /MIC was highly correlated to the in vivo activity of sulbactam and durlobactam, respectively.Both murine neutropenic thigh and lung models suggested a 1-log 10 CFU reduction over 24 hours can be realized when sulbactam fT>MIC is greater than 50% and durlobactam fAUC/MIC 0-24 is ~10.Exposure requirements were slightly higher in the lung model vs. the thigh and were therefore ultimately used for the exposure targets.An additional log 10 CFU reduction can be realized when durlobactam fAUC/MIC 0-24 is ~30.These targets were used to support dose selection for the Phase 3 clinical trial.

Bacterial isolates
All strains with an "ARC" designation are part of the Entasis collection of clinical isolates.All isolates have been previously characterized by whole genome sequencing.

Antimicrobial susceptibility testing
Sulbactam and sulbactam-durlobactam broth MIC testing was performed according to the Clinical and Laboratory Standards Institute methodology (28).Sulbactam-durlobac tam MIC testing was performed as a titration of sulbactam in the presence of a fixed concentration of 4 µg/mL durlobactam.

Animal welfare statement
All in vivo procedures were completed in compliance with the Animal Welfare Act Regulations (9 CFR 3) under Entasis Therapeutics-reviewed Institutional Animal Care & Use Committee (IACUC)-approved protocols and under the supervision of a site-attend ing veterinarian.

In vivo neutropenic lung and thigh infection models
In vivo infection models of pneumonia and thigh tissue abscess were performed in CD-1 mice (15-18 g) rendered neutropenic by cyclophosphamide treatment prior to infection as previously described (29).In the lung model, mice were infected with A. baumannii isolates in 1% agar slurry via direct intratracheal instillation.In the thigh model, mice were infected intramuscularly in the right thigh.Infected mice were treated with either sulbactam alone or sulbactam combined with durlobactam at a constant 4:1 ratio.Dosing was initiated 2 hours after infection and administered as eight subcutaneous (SC) injections administered q3h or four SC injections administered q6h.Vehicle groups received a single SC injection, and positive control groups received either a single SC injection of colistin sulphate (maximum tolerated dose of 40 mg/kg) or two oral doses of levofloxacin (200 mg/kg bid) 2 hours after infection.Colistin sulphate administered as a single 40-mg/kg dose achieved 0.25 to 2.6 log 10 CFU/g reduction vs. MDR A. baumannii strains, and levofloxacin achieved −2.4 log 10 CFU vs. A. baumannii ARC2058 in the thigh model.In the lung model, colistin administration resulted in 1-log 10 CFU/g growth to 2.1 log 10 CFU/g reduction vs. MDR A. baumannii strains and levofloxacin administration resulted in 1.8 log 10 CFU/g reduction vs. A. baumannii ARC2058.The overall performance of positive controls was acceptable relative to isolate susceptibility.Efficacy was determined by harvesting infected tissue and determining viable bacte rial counts 24 hours after the start of treatment.A full dose response study utilized meropenem to validate the model system, where meropenem was administered SC on a q6h regimen from 1 to 300 mg/kg following uranyl nitrate pre-treatment (30) to extend drug exposure.Subcutaneous doses of sulbactam and durlobactam were administered at a 4:1 ratio at doses of 20:5, 40:10, 60:15, 80:20, and 160:40 mg/kg (SUL:DUR) with plasma PK sampling (n = 3 samples/timepoint) at 0.5, 1, 2, 3, 4, 6, 8, 10, and 12 hours post dose.These studies were completed in neutropenic CD-1 mice with active thigh infections.An additional PK and ELF distribution study was completed at a SUL:DUR dose of 100:25 mg/kg in neutropenic lung-infected animals with PK timepoints sampled at 0, 1, 3, 6, and 12 hours post dose.Blood was processed for plasma using microcon tainer tubes containing ethylenediamine tetraacetic acid (EDTA, Beckton Dickenson) and centrifugation for 5 minutes at 13,200 rpm.Plasma samples were treated 1:1 with a SigmaFAST protease inhibitor cocktail (Sigma-Aldrich, Cat No. S8820) prepared from one tablet dissolved in 10 mL of deionized water.The samples were then stored at −80°C until bioanalysis.Plasma samples were analyzed by LC-MS/MS using instrument parameters summarized in (Table S3) A murine unbound fraction of 100% was used to calculate free plasma concentrations for both sulbactam and durlobactam used in exposure response analyses.This value came from the mean of the mouse plasma protein unbound fractions determined ex vivo at a range of 0.4 µM to 50 µM (0.093 to 11.7 µg/mL for sulbactam and 0.111 to 13.9 µg/mL for durlobactam; data on file Entasis Therapeutics).

Pharmacokinetic-pharmacodynamic analysis
As the plasma PK of neutropenic lung-infected animals matched that of thigh-infected animals, the thigh-infected PK was used for the development of the population PK model and subsequent exposure estimates for doses used in the PK/PD exposure-response analyses.Pharmacokinetic models were fit to time vs. drug concentration profiles generated in infected mice using Phoenix6.2WinNonLin and NLME.For the purposes of estimating exposures across all doses used in dose response studies, population PK models were developed for sulbactam alone, for sulbactam when co-administered with durlobactam, and for durlobactam co-administered with sulbactam (dose ratio 4:1, sulbactam:durlobactam).Population PK parameter estimates were derived from a two-compartment PK model incorporating a log-additive error model constructed from fitting of sulbactam and durlobactam concentration vs. time data (Table S4).Represen tative model fit concentration vs. time profiles vs. observed data are shown in Fig. S1 for sulbactam at 320 and 40 mg/kg and durlobactam at 80 and 10 mg/kg.Predicted exposures were utilized in the PK/PD analyses as all dosing solutions tested were within the variability of the bioanalytical method (20%).
For pharmacokinetic-pharmacodynamic evaluation sulbactam %fT>MIC and durlobactam fAUC 0-24 /MIC were simulated for each dose using the population PK model.A Hill-type model was fit to the pharmacodynamic data generated from the dose response studies with linear least squares regression analysis of the relationships between change in log 10 CFU from baseline at 24 hours and sulbactam %fT>MIC and durlobactam fAUC/MIC.Magnitudes associated with 1-log 10 and 2-log 10 CFU reduction from baseline at 24 hours as well as the exposure required for an 80% reduction in bacterial burden (EC 80 ) were determined from the fitted function.

Distribution of sulbactam and durlobactam into mouse ELF
ELF penetration was studied for both sulbactam alone and sulbactam:durlobactam, where mice received a single subcutaneous injection of either 100 mg/kg sulbactam or a combination of 100 mg/kg sulbactam/25 mg/kg durlobactam.Administration was initiated 2 hours after intratracheal infection with A. baumannii ARC2058 in neutro penic mice.Whole blood samples were taken by cardiac puncture.ELF was collected immediately following blood collection via bronchoalveolar lavage (BAL), which was performed by inserting a catheter into the trachea and instilling 1 aliquot of 0.8 mL normal saline followed by immediate removal of the fluid.Four mice per time point were used.Plasma and BAL were separated using microcontainer tubes containing EDTA (Beckton Dickenson) by centrifugation for 5 minutes at 13,200 rpm, treated with 1:1 SigmaFAST protease inhibitor cocktail, and stored at −80°C until bio-analysis.The pre-diluted concentration of antibiotic in the BAL fluid was calculated by comparing the difference in amounts of urea measured (Bioassay Systems QuantiChrom Urea Assay Kit) in the simultaneously collected plasma and BAL fluid for each mouse.

Sulbactam and durlobactam concentration determination
Serial blood samples collected mice infected over a 24-hour time and administered doses from 50:12.5 mg/kg to 320:80 mg/kg sulbactam durlobactam were assayed by liquid chromatography -tandem mass spectrometry (LC-MS/MS) (Sciex API 5000) using a qualified non-GLP bioanalytical method (Table S3).Plasma samples were treated 1:1 with a SigmaFAST protease inhibitor cocktail (Sigma-Aldrich, Cat No. S8820) prepared from one tablet dissolved in 10 mL of deionized water.The samples were then stored at −80°C until bioanalysis.Prepared calibration standards in the assay (1, 2.5, 5, 10, 25, 100, 500, 1,000, 5,000, and 10,000 ng/mL) were made by serial dilution in 1:1 Mueller Hinton Broth 2:analytical sample.The analytical sample consisted of blank plasma or BAL diluted 1:1 in SigmaFAST prior to processing.Sample aliquots (50 µL) were added in the 96-well plates.Then, 150 µL of 100% acetonitrile and 0.1% formic acid containing 250 ng/mL of carbutamide as an internal standard was added to the plate.Plates were covered and centrifuged at 4,000 × g for 5 minutes.Supernatant (100 µL) was transferred to a clean collection plate.Samples were mixed well and analyzed by LC-MS/MS (Sciex API 5000-Analyst v1.6.1), and LC-MS/MS instrument parameters are summarized in Table S3.The assay had a lower limit of quantitation (LLOQ) of 1 ng/mL and a upper limit of quantitation (ULOQ) of 10,000 ng/mL.The assay performance is summarized in Table S5.

TABLE 1
MIC summary of A. baumannii isolates used for in vitro and in vivo PK/PD studies a Modal MIC.b MIC of sulbactam in the presence of 4 µg/mL durlobactam.c T, thigh model; L, lung model.d ND, not determined.

TABLE 2
Meropenem percentage unbound plasma time above MIC estimates to achieve PK/PD endpoints in neutropenic thigh model vs. A.

TABLE 4
Co-modeled durlobactam fAUC/MIC a ratio values associated with 1-log 10 CFU reduction from baseline, 2-log 10 CFU reduction from baseline, and EC 80 PK/PD endpoints a Estimate calculated from Emax fitting of net bacterial CFU change over 24 hours vs. AUC 0-24 and dividing by MIC of 4.0 µg/mL.b UL mean exposure of 32.9 to 43.5%T>MIC.c UL mean exposure of 43.9 to 81.5%T>MIC.d SUL exposure range of 18.2 to 52.7%T>MIC.