Introduction, results and discussion

Non tuberculous mycobacteria (NTMs) are emerging pathogens in patients with cystic fibrosis (CF), recently estimated to affect approximately 12% of patients in Western countries [1]. Of NTM pulmonary infections, Mycobacterium abscessus infection is considered especially concerning as it is associated with increased morbidity and mortality, and is a poor prognostic factor even following lung transplantation [2–4]. Treatment of M. abscessus infections is challenging due to antibiotic resistance and tolerance mechanisms. Despite prolonged courses of multiple antibiotic treatments, long-term clearance of M. abscessus from respiratory airway in patients with CF is rarely successful [1]. Multi-bacterial infections, specifically of M. abscessus, Pseudomonas aeruginosa and sometimes Staphylococcus aureus are especially challenging and difficult to treat. Non-CF patients also suffer from M. abscessus infections, many times related to chronic lung diseases, plastic surgery, foreign bodies and other clinical situations.

Most β-lactam antibiotics, bar carbapenems and cefoxitin, are ineffective against M. abscessus, as it harbors BlaMab [5], a potent β-lactamase able to degrade both β-lactams and β-lactamase inhibitors. Avibactam is a new, non-β-lactam, β-lactamase inhibitor, active against BlaMab [5]. Data from zebrafish suggests avibactam can enhance the activity of ampicillin against M. abscessus, but ampicillin is inactive against P. aeruginosa. No data exists on whether it can augment the efficacy of the antipseudomonal drug piperacillin, enabling a single agent use against both Pseudomonas aeruginosa and M. abscessus, a co-infection often found in patients with CF [2].

In this study, we aimed to evaluate the effect of piperacillin/avibactam against Mycobacterium abscessus, in vitro and in vivo, using our recently established infection model in Galleria mellonella larvae [6], Coupled with a luminescent M. abscessus mutant (mDB158) [6].

To test the susceptibility of M. abscessus to piperacillin/avibactam in vivo, we used mDB158 [6], treated by meropenem, piperacillin, or ampicillin, each one with and without the addition of avibactam. Bacterial growth was assessed by luminescence measurement. Using 96 well plates, 5*103 CFU of mDB158 were cultured with serial 1:2 dilutions of meropenem (50 to 0 mg/L), piperacillin (800 to 0 mg/L) and ampicillin (200 to 0 mg/L), alone or with the addition of 4 mg/L of avibactam. Following 72 h of incubation at 37 °C, luminescence was measured using the SpectraMaxi3® microplate detection system. At 72 h, luminescence consistently increased by 103 fold in wells without antibiotics. The luminescence minimal inhibitory concentration (lu-MIC) was thus defined as the concentration in which luminescence remained similar to baseline value or increased no more than 3 fold compared to baseline. As expected, avibactam did not enhance the antibacterial activity of meropenem, as carbapenems are not considerably degraded by the β-lactamase of M. abscessus. The lu-MIC of ampicillin was reduced approximately 16 fold when augmented with avibactam, showing activity at 3.125 mg/L of ampicillin. Most importantly – the activity of piperacillin was enhanced 16–32 fold when combined with avibactam, also showing substantial antibacterial activity at 3.125 mg/L. Avibactam alone did not inhibit M. abscessus growth.

All the experiments/concentrations were done in triplicates, and repeated 3 times with similar results. For visual representation of these findings we analyzed a single 96-well plate with a single well for each concentration using the IVIS® imaging system (Fig. 1) with the above antibiotic combinations.

Fig. 1
figure 1

Avibactam lowers MIC of piperacillin for M. abscessus. A broth dilution assay was performed using 5*103 CFU of a luminescent M. abscessus mutant with serial 1:2 dilutions of meropenem (50 to 0 mg/L), piperacillin (800 to 0 mg/L) and ampicillin (200 to 0 mg/L) alone or with the addition of 4 mg/L of avibactam. Image showing luminescence demonstrated by IVIS® following 72 h of incubation. Mero – meropenem, pip – piperacillin, amp – ampicillin, avi – avibactam. Dotted white line borders the test area. White numbers show mero/pip/amp concentrations

Full size image

To test if this combination is also effective in-vivo, we used our previously described G. mellonella larvae as a model of M. abscessus infection [6]. We inoculated 60 larvae with luminescent M. abscessus mDB158 on day 0 and kept them at 37 °C. On days 1 and 2 we treated larvae with 40 μg (200 mg/kg) meropenem, 100 μg (500 mg/kg) piperacillin, 0.2 μg (1 mg/kg) avibactam alone, or piperacillin combined with avibactam (100 μg/0.2 μg), approximating two daily doses of antibiotics. Using IVIS® Lumina Series III (Caliper LifeSciences), we measured infection progression in live infected larvae on day 3 (We previously showed RLU correlates well with CFU – [6]). Larvae treated with either meropenem or piperacillin/avibactam had a significantly lower infection burden compared to untreated controls (p < 0.0001 and p = 0.004 respectively). Piperacillin and avibactam alone had no significant inhibitory effect (Fig. 2). A second experiment with only one injection of antibiotics on day 1 showed similar results.

Fig. 2
figure 2

Piperacillin/avibactam is effective in treating Mycobacterium abscessus in a Galleria mellonella infection model. We inoculated 60 G. mellonella larvae with luminescent M. abscessus on day 0, and treated larvae with meropenem, piperacillin, avibactam alone, or piperacillin/avibactam on days 1 and 2. Using IVIS® imaging, we measured infection progression in live infected larvae on day 3. “Healthy” are un-infected larvae

Full size image

It is well established that pulmonary infection with M. abscessus is a poor prognostic factor for patients with CF, independently associated with a progressive decline in lung function [2–4, 7]. Treatment necessitates prolonged multi-drug regimens including a carbapenem backbone [1]. Use of a narrower spectrum β-lactam backbone has so far been hindered due to the M. abscessus potent β-lactamase BlaMab [5]. Avibactam was recently shown to inhibit BlaMab, yet its role in treating this infection is unclear. Some in vitro data and in vivo zebrafish data demonstrated an ampicillin/avibactam combination to have an anti-mycobacterial effect [5]. Unfortunately, as patients with CF suffer from multi-bacterial infections including Pseudomonas aeruginosa, such a combination would not adequately target their pathogenic respiratory flora.

In our study, we showed piperacillin, an antipseudomonal β-lactam, to have a substantial effect against M. abscessus when augmented by avibactam. In vitro, 4 mg/L avibactam enhanced the activity of piperacillin 16–32 fold. This data suggests avibactam lowers the piperacillin MIC to a clinically-relevant range. In vivo, we also showed piperacillin/avibactam is able to treat M. abscessus infection in G. mellonella larvae, similarly to meropenem.

Our data suggests piperacillin/avibactam is a promising novel combination for patients with CF, targeting M. abscessus. Although the spectrum of piperacillin/avibactam is only mildly narrower than that of carbapenems (especially for Acinetobacter), it does spare the use of meropenem, slowing the development of carbapenem-specific resistance mechanisms. Use of piperacillin/avibactam may be especially useful for treating patients suffering from with M. abscessus and P. aeruginosa co-infections. Further in vivo studies are needed to establish efficacy, pharmacodynamics and pharmacokinetics of this combination.