Carbapenem-resistant (CR) Gram negative bacteria (GNB) - which are resistant to carbapenems (a last-line potent antibiotic with a high therapeutic index) - are also resistant to all other beta-lactam antibiotics. Most CRGNB are also extensively-drug resistant (XDR) (resistant to all classes of antibiotics except polymyxins and/or tigecycline) or pan-drug resistant (PDR) (resistant to all antibiotics), resulting in a dearth of effective options against these life-threatening infections.

Against CRGNB, standard therapy includes monotherapy (using polymyxins or tigecycline) or unguided antibiotics combination (polymyxins + carbapenem). Unfortunately, CRGNB can develop resistance after antibiotic monotherapy, resulting in the further development of pan-drug resistance. Unguided antibiotic combinations, selected anecdotally based on past experience, are also unlikely to be useful in our local setting, as effective antimicrobial combinations are bacterial-strain specific due to large variation in molecular mechanisms of resistance.Hence, the investigators propose to evaluate the efficacy of a novel treatment strategy using in vitro antibiotic combination testing (iACT) to guide antibiotic combinations in the management of patients with CRGNB infections in a randomised controlled trial (RCT).

The health problem - Antimicrobial resistance (AMR) in GNB and the dearth of therapeutic options AMR, particularly in GNB infections (e.g. Pseudomonas aeruginosa, Enterobacteriaceae and Acinetobacter baumannii), means resistance to multiple or even all available antibiotic classes. The growing incidence of CRGNB and XDRGNB is an urgent global healthcare challenge today. These pathogens are classified as "priority 1" (utmost critical) pathogens by the World Health Organization (WHO). Hsu et al from Singapore has published recently on the estimated prevalence of carbapenem-resistant Enterobacteriaceae (CRE) and carbapenem-resistant A. baumannii (CRA).

Antibiotics are the "foundation" of modern medicine, without which key medical procedures including surgery and chemotherapy for cancer are rendered too dangerous to perform. With the advent of CRGNB, effective antibiotic treatment strategies are now limited. Currently, approximately 700,000 people die from resistant infections every year in the world. It is estimated that by 2050, 10 million lives (1 in 3 The people) per year and a cumulative USD $100 trillion of economic output are at risk due to drug resistant infections, if resistance rates continue to rise unabated. Patel and co-workers observed mortality rates of 40-50% in patients with bloodstream infections caused by CR Klebsiella pneumoniae. Most CRE infections and the transmission of CRE are predominantly linked with healthcare exposure.

Combination therapy - a current mainstay therapeutic option against CRGNB Given the current resistance landscape, coupled with a continued dwindling pipeline of drugs to treat these infections, physicians in Singapore now face a severe lack of effective antimicrobial therapies to treat infections caused by CRGNB. While polymyxin B, an old antibiotic previously forsaken due to an allegedly high rate of toxicities, has been resurrected for such infections, the presence of polymyxin B heteroresistance limits its utility as a monotherapy in severe or deep-seated infections. Currently, antibiotic combination therapy has been adopted by most infectious diseases (ID) physicians locally, as the mainstay treatment against CRGNB infections, because of the theoretical benefits conferred by antibiotic combination therapy (e.g. synergism, prevention of development of resistance). In addition, there is increasing clinical evidence suggesting that antibiotic combination therapy may improve prognosis.

Unfortunately to date, the mortality reduction associated with antibiotic combination therapy in the treatment of CRGNB infections has not been consistently observed across all studies. For example, in a prospective comparison of 55 patients treated for XDR A. baumannii infections, the authors found that the combination of colistin + tigecycline resulted in higher mortality than colistin + a carbapenem when the tigecycline minimum inhibitory concentration (MIC) of the causative pathogen was >2 mg/L (HR = 6.93, 95% CI 1.61-29.78; P= 0.009), suggesting that caution is needed when developing polymyxin combination regimens.

In our local retrospective study, the Investigators have shown that when compared to polymyxins monotherapy or unguided antibiotic combination therapy, the guided selection of a rationally optimised antibiotic combinations was associated with significantly lower rates (by 8 times) of infection-related mortality in patients with XDRGNB infections. In most previous combination therapy studies in the clinical setting, the studies are typically limited by small sample sizes and retrospective nature. To address the ambiguity surrounding the use of polymyxin combinations, three large prospective RCTs are currently underway in Europe and the USA (ClinicalTrials.gov IDs NCT01732250, NCT0159797, NCT03159078)). In these trials, a fixed combination (polymyxins + carbapenem) will be compared to a fixed type of monotherapy (polymyxins alone).

Molecular diversity and complexity in local CRGNB and the implications in antibiotic combination selection Although the aforementioned overseas trials will attempt to address the ambiguity surrounding the use of polymyxin combinations by comparing a fixed combination (colistin + meropenem) to a single monotherapy, the findings of the trials will likely not be generalizable to or applicable in many geographical areas in Asia, especially our local setting, due to the diversity and complexity in the molecular mechanisms of resistance in our local CRGNB strains. Therefore, additional clinical trials will be needed to assess synergy between polymyxins and other agents for specific type pathogens. Such an approach to determining effective antibiotic combinations will be too time-consuming and certainly not cost-effective.

Unlike countries in the United States and Europe, where a predominant K. pneumoniae clonal (ST-258) and resistance type (KPC) is observed, there is greater diversity in Singapore. This greatly complicates management strategies, including the selection of effective combinations for clinical use.

Our local CRE is associated with a variety of resistance mechanisms (e.g. various carbapenemases production in IMPs, KPCs, NDMs, OXA-48, OXA-181, OXA-232, and dual carbapenemases production with/without porin down regulation). The local K. pneumoniae appeared to be of highly varied sequence types (STs) (ST 11, 14, 15, 17, 29, 42, 48, 147, 163, 231, 237, 273, 437, 568, 841, and 885). In addition, the investigators have also detected the presence of mcr-1 novel plasmid that confers resistance to the polymyxins in our CRE isolates that are KPC- or NDM-producers in SGH. For CR A. baumannii, three clones (IC I, IC II, and IC2 II, respectively) from the three outbreaks in 1996, 2001, and 2006 were observed. While these clones harboured carbapenemase producing genes - blaOXA-64, blaOXA-66, and blaOXA-51-like, respectively, blaIMP-4, ISAba1-blaOXA-23-like and blaOXA-58-like have also been described. For CR P. aeruginosa, our local isolates did not belong to any major international clonal complexes such as ST111 and ST235. While carbapenemase production in CR P. aeruginosa is less common, blaIMP and blaVIM have been described. Non-carbapenemase-mediated mechanisms (veb-1 gene +/- efflux pumps +/- porin losses) accounted for carbapenem resistance in local P. aeruginosa strains.

The heterogeneous mechanisms of resistance, coupled with the interplay of multiple resistance mechanisms in local CRGNB, impact on therapy selection decisions and presents as a huge challenge to clinicians locally. The presence of porin mutations and varying levels of porin expression further complicates the pathogen's responses to antibiotic therapy, including combination therapy. For instance, one of the postulated synergy mechanisms is that membrane permeabilisation mediated by polymyxins may enhance the access of carbapenems to their target sites, hence reducing the chances of carbapenem hydrolysis by carbapenemases. This synergistic activity may be diminished in the presence of decreased porin-related permeability, and it is demonstrated that combinations of colistin-doripenem/ertapenem were not synergistic and not bactericidal against isolates with low porin expression. Locally, the investigators have also previously shown that, due to the variability of resistance mechanisms and the interplay between multiple mechanisms, the types of effective antimicrobial combination are bacterial strain-specific. Given the wide variety and permutations of resistance mechanisms, effective antimicrobial combinations are highly varied, and therefore strain-specific testing is required to guide selection of strain-specific combinations in our local setting.

Solving the health problem - Strain-specific combination therapy as a precision therapy against CRGNB In view of the problem described above, the application of "precision medicine" through individualised antimicrobial chemotherapy will be the crux in our combat against CRGNB infections. Strain-specific antibiotics combination testing, in particular, should be performed to guide selection of drugs for combined therapy for the following purposes: Rule out antagonistic antibiotic combinations and ascertain the right and effective antibiotic combination that is synergistic and bactericidal for a specific CRGNB. Currently, in a routine hospital setting, the therapeutic selection for CRGNB is guided by in vitro susceptibility testing (e.g. disk diffusion, Vitek 2), and at best coupled with the determination of MICs (that may require an additional day for results to be available). Currently, this constitutes standard care or best available therapy. However, such traditional single- antibiotic susceptibility results have limited utility in the guiding the selection of combination therapy against CRGNB, as MICs alone are not useful in the prediction of synergistic/bactericidal effect and guide the selection of antibiotics combination therapy, which is required in CRGNB infections where nearly all single antibiotics are ineffective.

In light of the clinical need in our local setting, a prospective in vitro antibiotic combination testing (iACT), with a rapid turn-around time of less than 24 hours (to know an antibiotics combination that is at least inhibitory), was developed to guide physicians in managing patients with CRGNB/XDRGNB infections. The iACT and its workflow were developed at the urgent request of our Infectious Diseases (ID) physicians, after extensive research about the use of iACT for management of patients with CRGNB/XDRGNB infections. Antibiotic combinations, that are shown to be at least inhibitory against the growth of pathogen, will be made known to the requesting ID physician within 18-24 hours. After an additional 20 hours, the bactericidal antibiotics combination will be confirmed with the ID physician. The objective of the iACT is to guide physicians in the selection of an individualised and rationally optimised combination therapy, taking into account the in vitro combination testing results of each strain and the patient's clinical and pharmacokinetic (PK) parameters. To date, the role and feasibility of the iACT, as well as the clinical utility of the service, have been published in a number of retrospective studies. However, iACT is not officially accepted to be used in routine practice in SGH, or anywhere else globally. SGH Translational Medicine Office has recommended that strong evidence with high quality data from prospective, randomised control trials is needed to change current practice.

To fully evaluate the clinical utility of iACT guided therapy, the Investigators propose to conduct an RCT comparing the therapy strategy incorporating iACT to guide selection of antibiotic combinations with current standard therapy. The iACT RCT will overcome the limitations that are inherent in retrospective studies and above-mentioned clinical trials (ClinicalTrials.gov IDs NCT01732250, NCT01597973, NCT03159078) conducted overseas.

Firstly, the RCT design will minimise confounding and allocation bias between treatment arms. Secondly, the use of our novel iACT platform will allow the strain-specific selection of antibiotic combinations against each CRGNB strain in a timely manner, as opposed to using only a single fixed combination in these trials (ClinicalTrials.gov IDs NCT01732250, NCT01597973, NCT03159078). Thirdly, through detailed workup of the presence of various carbapenemases, efflux pumps, and porin loss detection, the Investigators will have detailed description of the molecular mechanisms mediating carbapenem resistance. Such description of the molecular mechanisms will document that our proposed guided therapy strategy is suitable for more than one type of CRGNB infections with variety of resistance mechanisms, and capable of combating against local and global molecular epidemiology of CRGNB infections, strengthening the external validity and applicability of this novel therapy strategy.

Developing new knowledge with scientific and clinical applications Currently, CRGNB are listed as the utmost critical pathogens by the World Health Organization (WHO), for which new and effective treatment strategies against CRGNB is urgently needed. Once completed, our study would represent the first randomised controlled trial that examines the utility of strain-specific antibiotic combination testing in the treatment of CRGNB infections. If our hypothesis is proven, this guided therapy strategy can change current practice paradigms in the treatment of CRGNB both locally and internationally, and provide a novel solution to the management of CRGNB infections. Given the heterogeneity of patients with CRGNB infections, our proposed RCT unequivocally addresses the challenges in treating CRGNB infection in ways that is otherwise not possible (e.g. ClinicalTrials.gov IDs NCT01732250, NCT01597973, NCT03159078). Through the description of the molecular epidemiology of CRGNB infections in our study, the Investigators will provide new insight into the complicated resistance mechanisms common in our local CRGNB strains. The investigators anticipate that the findings of this study will drive further research, particularly translational research, to further enrich knowledge, and improve management, of CRGNB infections.

Inclusion Criteria:

1. Inpatient at the time of enrolment.
2. Age ≥16 years.
3. An ongoing infection as defined by the published Centers for Disease Control and Prevention (CDC) /National Healthcare Safety Network (NHSN) or Infectious Diseases Society of America (IDSA) guidelines; Section 16.1-16.6 Appendix specifies the most common examples expected in this study.
4. Positive culture of CRGNB isolates from relevant clinical sites (i.e. samples that are not obtained for surveillance purposes, such as rectal swabs)
5. No more than 5 calendar days has elapsed since the first positive culture collection.

Exclusion Criteria:

1. Unable to provide consent AND have no legal representative (LR).
2. Subjects on palliative care or with less than 24 hours of life expectancy (as discussed with their primary physicians).
3. Colonisation only, which is defined as positive isolation of CRGNB isolated at screening sites (e.g., rectal swabs) only
4. Prior recruitment into this study.