Is combination therapy for carbapenem-resistant Klebsiella pneumoniae the new standard of care?

Abstract

Carbapenem-resistant Klebsiella pneumoniae causes serious nosocomial infections and therapeutic options are limited. There is increasing evidence suggesting that combination antibiotic therapy is more effective than monotherapy and leads to better outcomes. However, questions remain about which regimen is optimal and how to balance the potential benefits of combination therapy versus the risks and possible complications (e.g., toxicity, increased costs, Clostridium difficile infection). Well-designed randomized clinical trials are needed to clarify these issues.

Keywords:

• carbapenem-resistant Klebsiella pneumoniae
• combination antibiotic therapy

The emergence of carbapenem-resistant Klebsiella pneumoniae (CR-KP) poses a significant threat to global public health. Strains of CR-KP often cause serious infections with high morbidity and mortality such as bacteremia and hospital-acquired pneumonia [1]. Treatment options are limited and have included colistin, tigecycline, aminoglycosides and fosfomycin. The poor clinical outcomes associated with CR-KP infections have led to an interest in alternative treatment approaches, one of which is antibiotic combination therapy. A growing body of evidence suggests combination therapy is more effective than monotherapy, although most studies have been retrospective and no results from randomized prospective clinical trials have been reported.

Definitive combination antibiotic therapy has generally been accepted as evidence-based for only a few indications such as infective endocarditis (e.g., ampicillin and gentamicin for Enterococcus), Helicobacter pylori, and mycobacterial infections. The theoretical reasons for using more than one antibiotic include exploitation of synergistic activity between the drugs and decreasing the emergence of resistance while on therapy [2]. The administration of antibiotics in combination is also utilized when they are used as initial empiric therapy to broaden the coverage spectrum. The results of several older studies (conducted before the emergence of CR-KP) on the benefits of combination therapy for treating Gram-negative infections were mixed, although some experts recommended it for immunosuppressed patients and those severely ill with Pseudomonas aeruginosa, Klebsiella and Enterobacter [3]. A more recent report also recommended initial combination therapy for bacteremia in the intensive care unit when multidrug-resistant organisms are suspected [4]. Combination therapy does, however, have some disadvantages including a higher risk for side effects (e.g., seizures, rashes), toxicity (e.g., acute kidney injury), adverse drug–drug interactions, increased costs, alterations in the gut microbiota and possibly Clostridium difficile infection [5].

Antibiotics with the most in vitro activity against CR-KP include colistin, polymyxin B, tigecycline, gentamicin, and fosfomycin [6]. However, monotherapy with these agents has been shown in several studies to be inferior compared with combination therapy despite in vitro testing indicating drug susceptibility. In a retrospective cohort study involving patients with bacteremia due to K. pneumoniae carbapenemase (KPC)-producing K. pneumoniae, 13% who received combination therapy died within 28 days of the onset of bacteremia compared with 57.8% who received monotherapy (p = 0.01) [7]. The most common combination was a carbapenem with either colistin–polymyxin B or tigecycline. The isolates from seven patients who received colistin–polymyxin B monotherapy had MICs in the susceptible range of ≤0.25 to 0.5 μg/ml, but the mortality was 57.1% nevertheless. Another retrospective cohort study by Tumbarello et al. [8] also found a higher mortality rate among patients treated with monotherapy compared with combination therapy for KPC-producing K. pneumoniae bacteremia (54.3 vs 34.1%, respectively, p = 0.02). The triple-drug regimen of tigecycline, colistin and meropenem significantly reduced the risk for death. Indeed, 30-day survival increased from 69.7% in those given the combination of tigecycline plus colistin to 87.5% when meropenem was added. Daikos et al. [9] reported similar outcomes among 205 patients with CR-KP bacteremia. Significantly higher mortality was seen in patients who received monotherapy (44.4%) compared with combination therapy (27.2%; p = 0.018). However, patients with advanced age were more likely to have received monotherapy, which may have confounded this result. The lowest mortality rate (19.3%) was observed in patients given a carbapenem-containing regimen. The most active drugs were tigecycline and colistin, although 15.1% and 25.4% of the isolates were resistant, respectively. Among the patients who received a carbapenem, mortality increased from 19.3% for a carbapenem MIC ≤8 μg/ml to 35.5% for a MIC >8 μg/ml. Moreover, combination therapy was an independent predictor of survival and patients with advanced age were more likely to have received monotherapy. In contrast, another retrospective study that examined polymyxin B monotherapy for CR-KP infections found a treatment success rate of 73%, including 82% for bacteremia [10]. However, it is concerning that 3 out of 40 patients treated with polymyxin B developed breakthrough infections from polymyxin-resistant organisms while on therapy. Because these studies [7–10] were retrospective, the results may have been influenced by confounding factors like comorbidities, severity of sepsis and timing of the initiation of antibiotics.

In a recent report, Demiraslan et al. [11] compared colistin or tigecycline monotherapy to colistin plus tigecycline in a murine model of CR-KP sepsis. The mice were divided into two groups: one received methylprednisolone and other did not. Based on bacterial counts in the livers and lungs of the mice, the investigators found no significant differences between combination therapy and monotherapy. Whether these results are relevant to humans remains to be elucidated, especially since (in most cases) the risks of steroids outweigh the potential benefits in treating sepsis.

Another potential regimen is dual carbapenem therapy. Using a murine thigh model and KPC-producing K. pneumoniae isolates, Wiskirchen et al. [12] found that doripenem and ertapenem together reduced bacterial density more than doripenem monotherapy for two isolates with doripenem MIC ≤16 mg/l. A proposed mechanism for this observation is that ertapenem binds the carbapenemase with great affinity, thus protecting the more potent carbapenem (i.e., doripenem) from hydrolysis [6], but one study found the opposite with KPC-6 [13]. Among the novel agents in development for multidrug-resistant Gram-negative infections is the synthetic non-β-lactam, β-lactamase inhibitor avibactam. The combination of ceftazidime–avibactam has potent in vitro activity against KPC-producing K. pneumoniae [14]. Clinical trials have shown ceftazidime–avibactam to be effective in treating complicated urinary tract infections and complicated intra-abdominal infections, leading to its approval on 5 September 2014 by the US FDA for these indications [15].

Given the suboptimal designs of the aforementioned studies and paucity of strong evidence, should combination therapy for CR-KP be standard practice? Recently, some experts emphasized the weakness of the data and argued there is no evidence-based support for most combination therapies (including colistin/carbapenem) against carbapenem-resistant Gram-negative bacteria [16]. Indeed, the definitive answer will require a randomized clinical trial in which one group of subjects with CR-KP infections receives monotherapy and another receives combination therapy, with outcomes compared between the groups. In the interim, because of the high morbidity and mortality associated with these infections and the possible mortality benefit, we believe it is reasonable to treat serious infections (e.g., CR-KP bacteremia) with combination therapy. Although the optimal regimen is unknown, it should be based on antibiotic susceptibility results and we suggest, based on the studies by Tumbarello et al. [8] and Daikos et al. [9], tigecycline or colistin plus meropenem or all three together if the isolate is susceptible to both non-carbapenems. This combination avoids the high risk of nephrotoxicity from combining an aminoglycoside with colistin. One caveat is the pharmacokinetic and pharmacodynamic properties of colistin are complicated and optimal dosing can be a challenge. Although an interesting concept, there is not yet enough data to support dual carbapenem therapy. The potential for increasing antibiotic resistance with combination therapy is an important consideration and it must be used judiciously. Moreover, antibiotics alter the taxonomic, genomic and functional capacity of the human gut microbiota [17]. The impact may differ depending on the antibiotic combination used and this hypothesis requires further investigation. Hopefully, novel antibiotics with activity against CR-KP will be developed that obviate the need for combination therapy, thus eliminating the inherent risks that accompany it.

Financial & competing interests disclosure

RR Watkins has received research funding from Forest Pharmaceuticals and the Akron General Foundation. S Deresinski has consulted in the previous year for Cubist Pharmaceuticals and Bayer Pharmaceuticals. The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed.

No writing assistance was utilized in the production of this manuscript.