To our knowledge, this is the first study to evaluate the activity of sulbactam/durlobactam against A. baumanni isolates in Türkiye. In this study, the in vitro activity of sulbactam/durlobactam was also compared with that of sulbactam, and the possible mechanisms of resistance in sulbactam/durlobactam resistant isolates were investigated.

The sulbactam MICs for isolates carrying blaOXA−23 plus blaPER-1 were higher than those observed for isolates carrying blaOXA−23 (Table 2). Sulbactam is not an inhibitor of OXA-23 [24]. Instead, it is susceptible to hydrolysis by OXA-23 enzyme, as sulbactam is a β-lactam itself [25]. On the other hand, sulbactam is known to be both a substrate and inhibitor of serine β-lactamases [25] including the ESBL enzyme PER-1 [26].

It was reported that many sulbactam molecules are needed to inactivate a single molecule of a serine β-lactamase. For example, the TEM-1 β-lactamase first hydrolyses ~7000 molecules of sulbactam and then sulbactam functions as an inhibitor of enzyme [27]. Although we could not find a data regarding the number of sulbactam molecules to inactivate PER-1 enzyme, it is likely that large quantities of sulbactam molecules are used for the inactivation of PER-1. Thus, in isolates carrying blaOXA−23 plus blaPER-1, lower number of free sulbactam molecules available for the inhibition of PBPs may be the cause of increased sulbactam MICs when compared with isolates that do not carry blaPER-1.

In a systematic review, the molecular epidemiology of carbapenem-resistant A. baumannii isolates in Türkiye was evaluated by analyzing original research articles published between January 2010 and January 2022 [28]. The incidence rates of carbapenemase genes blaOXA−23 and blaOXA−58 in carbapenem-resistant A. baumannii isolates were reported as 76.4% and 8.4%, respectively. Also, the prevalences of both these genes increased with time. The average positivity rates of blaNDM and the ESBL gene blaPER-1 was reported as 0.1% and 9.7%, respectively [28]. The strains in our study was isolated in 2020 and 2021. The prevalence of OXA-type carbapenemases in A. baumannii may have increased from 2020 to 2026. Despite a possible increase in the prevalence of OXA-type carbapenemases, sulbactam/durlobactam may still show high activity against carbapenem-resistant A. baumannii isolates as durlobactam is an inhibitor of Ambler class D β-lactamases. However, it is unknown whether the PBP3 mutation patterns or the prevalence of isolates with PBP3 mutations have changed in Türkiye over time. Changes regarding PBP3 may directly influence the activity of sulbactam/durlobactam and may limit the generalizability of our findings from earlier isolates.

In sulbactam/durlobactam combination, sulbactam is known to be the active component against Acinetobacter spp. Durlobactam is reported to have no intrinsic activity against A. baumannii [11]. Durlobactam protects sulbactam from hydrolysis by β-lactamase enzymes. However it does not inhibit metallo-β-lactamases. For the two sulbactam/durlobactam resistant isolates in our study, β-lactamases likely had very limited effect on sulbactam and sulbactam/durlobactam resistance and other factors were present; because adding durlobactam to sulbactam did not restore sulbactam susceptibility and these isolates did not carry metallo-β-lactamase genes. In one of the isolates, the addition of durlobactam led to a very limited reduction in sulbactam MIC value (from 24 to 12 mg/L). The isolate with high level sulbactam resistance (MIC > 256 MIC mg/L) still exhibited high level resistance (MIC > 64 mg/L) after the addition of durlobactam. In both of the isolates, mutations in PBP3 gene were detected. The first isolate encoded the A515V variant of PBP3 gene. A515V substitution was reported to be associated with sulbactam/durlobactam resistance [7]. However, some authors suggest that A515V substitution is not sufficient to confer sulbactam/durlobactam resistance and additional factors are probably needed [9]. The first isolate had an additional mutation (C1546T) in PBP3 gene. Furthermore, the AdeN (the repressor of AdeIJK efflux system) was absent in this isolate. Durlobactam was reported to be a substrate for the AdeIJK efflux system [9]. The overexpression of this efflux system and the additional mutation in PBP3 gene may also have contributed to sulbactam/durlobactam resistance.

In the second isolate, insertion of an additional amino acid at position 374 (-374D) in PBP3 was detected. Unlike A515V and C1546T, this mutation may be associated with high-level resistance to sulbactam and sulbactam/durlobactam because this isolate had sulbactam MIC of >256 MIC mg/L and sulbactam/durlobactam MIC of >64 mg/L. This mutation may have caused a substantial reduction in the affinity of PBP3 for sulbactam.

In our study, sulbactam/durlobactam resistant isolates were also found to encode three silent mutations in PBP3 gene. A silent mutation alters the nucleotide sequence but not the amino acid sequence. The effect of silent mutations has traditionally been considered to be either neutral or nearly neutral. However, recent evidence has demonstrated that they can affect protein levels or conformation [29].

Although we included a small number of isolates in our study, it may be useful to indicate the sulbactam/durlobactam susceptibility rate (91.7%). In a study including 5032 clinical isolates of A. baumannii-calcoaceticus complex from 33 countries, sulbactam/durlobactam MIC values for 98.3% of isolates were ≤4 mg/L (susceptible) [8]. In another study including 1722 clinical isolates of A. baumannii-calcoaceticus complex from 31 countries, 97.7% of isolates had sulbactam/durlobactam MIC of ≤4 mg/L [11]. However, the collections in these two studies comprised both carbapenem-susceptible and carbapenem-resistant isolates [8, 11].

In the study conducted by Seifert et al., 237 (96.3%) of 246 carbapenem-resistant A. baumannii isolates from 37 countries had sulbactam/durlobactam MIC of ≤4 mg/L [30]. The sulbactam/durlobactam susceptibility rate in our set of isolates (91.7%) was similar to that reported by Segatore et al. (92.2%) who evaluated the in vitro activity of the sulbactam/durlobactam against 141 carbapenem-resistant A. baumannii isolates collected from six clinical microbiology laboratories in Italy [31].

Studies which report lower sulbactam/durlobactam susceptibility rates (below 90%) among carbapenem-resistant A. baumannii isolates also exist. Petropoulou et al. tested the in vitro activity of sulbactam/durlobactam against 190 carbapenem-resistant A. baumannii isolates collected from 11 hospitals throughout Greece and found a susceptibility rate of 87.9% [3].

This study has some limitations. First, we used gradient diffusion method instead of the reference broth microdilution method for the determination of sulbactam and sulbactam/durlobactam MICs. Broth microdilution is known to provide more reliable and accurate results when compared to gradient diffusion method. Second, sulbactam breakpoints for A. baumannii have not been formally established by CLSI or EUCAST. As sulbactam is known to be the active component of ampicillin/sulbactam combination against Acinetobacter spp., we used the susceptibility breakpoint of 4 mg/L for sulbactam, based on the ampicillin/sulbactam susceptibility breakpoint of 8/4 mg/L established by the CLSI. Although several studies adopted this approach, using a surrogate breakpoint may influence the interpretation of comparative activities of sulbactam and sulbactam/durlobactam against isolates [8, 31, 32]. Third, we could not include detailed clinical information regarding patients because we couldn’t access the complete medical records of patients.

In conclusion, sulbactam/durlobactam demonstrated high activity against carbapenem-resistant A. baumannii strains in our study and restored sulbactam susceptibility in ~90% of sulbactam-non-susceptible isolates. Presence of blaPER-1 in addition to blaOXA−23 led to an increase in sulbactam MICs but not in sulbactam/durlobactam MICs. Mutations observed in the PBP3 gene of sulbactam/durlobactam resistant isolates may have led to the development of resistance by reducing affinity of PBP3 for sulbactam. Also, lack of the repressor of AdeIJK efflux system in one of the resistant isolates may have contributed sulbactam/durlobactam resistance by causing the overexpression of efflux pumps.