In Vitro Antibiofilm Activity of Fosfomycin Alone and in Combination with Other Antibiotics against Multidrug-Resistant and Extensively Drug-Resistant Pseudomonas aeruginosa

Background: Due to its rapid resistance development and ability to form biofilms, treatment of Pseudomonas aeruginosa infections is becoming more complicated by the day. Drug combinations may help reduce both resistance and biofilm formation. Methods: Using the microtiter plate assay, we investigated the in vitro inhibition of biofilm formation and the disruption of preformed biofilms in multidrug-resistant and extensively drug-resistant clinical isolates of P. aeruginosa in the presence of peak plasma levels of eight antipseudomonal antibiotics alone and in combination with fosfomycin: ceftazidime, piperacillin/tazobactam, cefepime, imipenem, gentamicin, amikacin, ciprofloxacin and colistin. Results: Combination therapy was significantly superior to monotherapy in its inhibition of biofilm formation. The highest inhibition rates were observed for combinations with colistin, cefepime and ceftazidime. Conclusion: Our results support fosfomycin combination therapy as an enhanced prophylactic option. Moreover, combinations with β-lactam antibiotics and colistin demonstrated a more potent inhibition effect on biofilm formation than protein synthesis inhibitors.


Introduction
Pseudomonas aeruginosa causes various infections, especially in individuals with impaired immune systems or underlying health conditions, such as respiratory tract infections; urinary tract infections; skin and soft tissue infections; wound, ear and eye infections; sepsis or bacteraemia.It is considered one of the most important hospital-acquired pathogens, and its treatment is becoming more difficult by the day as antibiotic resistance increases.Severe P. aeruginosa infections are link to high mortality rates and therefore require empirical treatment combinations with different mechanisms of action [1].P. aeruginosa is also known for its ability to form biofilms on human tissue, as well as on various surfaces, such as catheters and different prosthetic devices.Biofilms provide a safe harbour for bacteria, where they can withstand harsh conditions, escape the host's immune system and antibiotics and be the source of persistent and recurrent infections.Bacteria in biofilms exhibit different behaviour than their planktonic forms.Antibiotics inhibit bacterial growth or kill planktonic bacteria but are much less effective against sessile biofilm bacterial cells, so-called persisters [2].
One of the most important virulence factors of P. aeruginosa is biofilm formation [3].Biofilm elimination is very difficult and requires high doses of antibiotics, often in combination, with very uncertain outcomes [4,5].Nearly 80% of persistent bacterial infections are estimated to be linked to biofilms [6], and biofilm-related P. aeruginosa infections often lead to treatment failure [2].
Although antibiotics are more useful in preventing biofilm formation, some data also suggest a reduction in the biofilm matrix in the presence of antibiotics, proposing different mechanisms [7][8][9].Aminoglycosides showed a preventive effect on the early adhesion of P. aeruginosa isolates.Several antibiotics, including ceftazidime, piperacillin/tazobactam and colistin, have shown some effect against bacterial adhesion, motility and biofilm formation [2].Fosfomycin has recently reemerged as an interesting treatment option for antimicrobial-resistant infections, as no cross-resistance and low resistance rates are expected [10][11][12].It has been shown to penetrate very well into mature biofilms formed by P. aeruginosa [13,14] and to enhance the penetration of other antibiotics [15,16], presumably due to its low molecular weight [13,14,17,18].Fosfomycin has also been associated with biofilm reduction in uropathogenic E. coli [18].
Fosfomycin monotherapy is of concern because of rapid mutant selection; therefore, dual antimicrobial coverage is often advised [12,19,20].Moreover, since monotherapy is usually unsuccessful in eradicating biofilms, combination therapy is an important treatment alternative, and it is necessary to investigate the possible anti-biofilm activities of antibiotics, alone or in combination [3][4][5]16,21].Chronic P. aeruginosa infections in cystic fibrosis are very difficult to treat, but aggressive antibiotic therapy approaches have recently been proven successful in eradicating new P. aeruginosa colonisations [22].Several antimicrobial combinations have been shown to be synergistic against P. aeruginosa biofilms [4,5].However, studies on combination therapy with fosfomycin on P. aeruginosa biofilms are limited [21,[23][24][25][26][27][28] and have mostly been carried out on a few isolates, using a small number of antibiotic combinations [16,25,[27][28][29].Fosfomycin synergy against P. aeruginosa biofilms has mainly been recorded in combination with aminoglycosides and quinolones [21,[23][24][25][26][27][28].Biofilm eradication has also been reported in combination with tobramycin [16].Little is documented on the effects of combining fosfomycin with colistin and β-lactam antibiotics.Synergy has been observed in combination with high doses of imipenem, but this is of questionable clinical value due to the potential toxicity of the drug levels required for clinical effect [3].Recently, synergy has also been documented in combination with colistin against P. aeruginosa biofilms at concentrations achievable through the inhalation of nebulised drugs [30].However, to our knowledge, no reports covering combinations with the β-lactam antibiotics ceftazidime, cefepime or piperacillin/tazobactam against extensively drug-resistant (XDR) isolates have been published.
Therefore, the aim of our research was to analyse the in vitro inhibition of biofilm formation, as well as the disruption of preformed biofilms, in clinical isolates of P. aeruginosa with various resistance traits, including multidrug-resistant (MDR) and extensively drug-resistant (XDR) isolates, in the presence of peak plasma levels of eight routinely used antipseudomonal antibiotics alone or in combination with fosfomycin: β-lactam antibiotics, which inhibit peptidoglycan synthesis (ceftazidime, cefepime, piperacillin/tazobactam, imipenem); fluoroquinolones, which block DNA synthesis (ciprofloxacin); aminoglycosides, which prevent protein synthesis (gentamicin, amikacin) and polypeptides (colistin), interfering with outer membrane and cytoplasmic membrane function.

Inhibition of Biofilm Formation
For the inhibition of biofilm formation (IBF) test, 24 isolates were used, of which 13 were strong and 11 were moderate biofilm producers.Each isolate was exposed to peak plasma concentrations of the antibiotics alone and in combination.
Exposure to fosfomycin resulted in varying degrees of inhibition in 70.8% (n = 17) of the isolates and ranged from 2.09% to 90.43%.A higher OD 620 value after exposure to fosfomycin was observed in 29.2% of the isolates (n = 7) compared to the OD 620 values of the unexposed isolates; see Figure 1.These seven isolates were all resistant to fosfomycin (MIC ≥ 64 µg/mL).Five of these seven isolates were MBPs when not exposed to fosfomycin but were converted into SBPs upon exposure.The overall mean IBF value obtained was 14.88%; see Table 2.There was no statistically significant difference between the OD 620 values with and without exposure to fosfomycin (p = 0.069).

Inhibition of Biofilm Formation
For the inhibition of biofilm formation (IBF) test, 24 isolates were used, of which 13 were strong and 11 were moderate biofilm producers.Each isolate was exposed to peak plasma concentrations of the antibiotics alone and in combination.
Exposure to fosfomycin resulted in varying degrees of inhibition in 70.8% (n = 17) of the isolates and ranged from 2.09% to 90.43%.A higher OD620 value after exposure to fosfomycin was observed in 29.2% of the isolates (n = 7) compared to the OD620 values of the unexposed isolates; see Figure 1.These seven isolates were all resistant to fosfomycin (MIC ≥ 64 µg/mL).Five of these seven isolates were MBPs when not exposed to fosfomycin but were converted into SBPs upon exposure.The overall mean IBF value obtained was 14.88%; see Table 2.There was no statistically significant difference between the OD620 values with and without exposure to fosfomycin (p = 0.069).The average obtained IBF values of the antibiotics alone and in combination are listed in Table 2.The effect of single and combined antibiotic treatments on IBF is shown  The average obtained IBF values of the antibiotics alone and in combination are listed in Table 2.The effect of single and combined antibiotic treatments on IBF is shown graphically in Figure 2.While large differences were observed among the inhibition properties of individual antibiotics, a more uniform inhibition pattern was observed when these antibiotics were used in combination with fosfomycin.The highest inhibition rates were observed with colistin alone and in combination, a result that was somewhat expected because of its low resistance rates (14%), and with cefepime alone and in combination, despite high resistance rates to both cefepime (72%) and fosfomycin (51%).Although the highest values of IBF were recorded with combinations with colistin, with a mean inhibition percentage of 82.3%, no statistical significance was found compared to the use of colistin alone (p = 0.47).The use of colistin alone (75.6%) and in combination (82.3%) was superior to the use of fosfomycin (14.9%) alone in terms of the IBF.Nevertheless, comparison of the biofilm production categories showed the superiority of combination therapy, with all the isolates belonging to the NBP category because of the colistin-fosfomycin combination; see Figure 3. High IBF levels were also observed with cefepime alone and in combination.However, combination therapy was superior to the use of single antibiotics in terms of category change, as combination therapy resulted in a higher percentage of NBPs; see Figure 3.All the other antibiotic combinations, with ceftazidime, piperacillin/tazobactam, imipenem, gentamicin, ciprofloxacin and amikacin, resulted in significant IBF levels compared to single use (p < 0.05).In addition, all antibiotic combinations led to a higher proportion of NBPs; see Figure 3.Although the highest values of IBF were recorded with combinations with colistin, with a mean inhibition percentage of 82.3%, no statistical significance was found compared to the use of colistin alone (p = 0.47).The use of colistin alone (75.6%) and in combination (82.3%) was superior to the use of fosfomycin (14.9%) alone in terms of the IBF.Nevertheless, comparison of the biofilm production categories showed the superiority of combination therapy, with all the isolates belonging to the NBP category because of the colistin-fosfomycin combination; see Figure 3. High IBF levels were also observed with cefepime alone and in combination.However, combination therapy was superior to the use of single antibiotics in terms of category change, as combination therapy resulted in a higher percentage of NBPs; see Figure 3.All the other antibiotic combinations, with ceftazidime, piperacillin/tazobactam, imipenem, gentamicin, ciprofloxacin and amikacin, resulted in significant IBF levels compared to single use (p < 0.05).In addition, all antibiotic combinations led to a higher proportion of NBPs; see Figure 3. High average IBF values were also found for combinations with the β-lactam antibiotics ceftazidime (81%), piperacillin/tazobactam (78.8%) and imipenem (74.1%).The IBF of combinations with ceftazidime, piperacillin/tazobactam and imipenem was significantly higher compared to that with single use of antibiotics (56.5%, 36.8% and 15.2%, respectively), Table 2.Moreover, all the antibiotic combinations resulted in higher proportions of NBPs; see Figure 3.The use of imipenem alone resulted in the IBF in 58.3% of the isolates (n = 14), with a mean of 15.2%.The measured OD620 values of 10 isolates were higher when imipenem alone was used compared to the measured OD620 biofilm values of the unexposed controls, resulting in negative IBF values.Of these isolates, 70% (n = 7) had a MICIMI value of ≥64 and were thus exposed to subMIC concentrations of imipenem.A similar effect was observed when gentamicin, amikacin and ciprofloxacin were used alone.A total of 93.8% of isolates (15/16) with negative IBF values were exposed to subMIC concentrations of gentamicin, 100% of isolates (14/14) with negative IBF values were exposed to subMIC concentrations of colistin and 66.7% isolates (8/12) with negative IBF values were exposed to subMIC concentrations of amikacin.While the mean values of the IBF of gentamicin (-20.8%),amikacin (-18.3%) and ciprofloxacin (-24.2%) were low, combination therapy with fosfomycin resulted in significantly higher IBF values (39.9%, 68.1% and 52.8%, respectively).Additionally, these antibiotic combinations also led to a higher proportion of NBPs as well; see Figure 3.

Disruption of Preformed Biofilms
The disruption of biofilms was investigated in 24 h old biofilms from eight strong biofilm producers.The preformed biofilms were exposed to antibiotics for 6 h and 24 h.High average IBF values were also found for combinations with the β-lactam antibiotics ceftazidime (81%), piperacillin/tazobactam (78.8%) and imipenem (74.1%).The IBF of combinations with ceftazidime, piperacillin/tazobactam and imipenem was significantly higher compared to that with single use of antibiotics (56.5%, 36.8% and 15.2%, respectively), Table 2.Moreover, all the antibiotic combinations resulted in higher proportions of NBPs; see Figure 3.The use of imipenem alone resulted in the IBF in 58.3% of the isolates (n = 14), with a mean of 15.2%.The measured OD 620 values of 10 isolates were higher when imipenem alone was used compared to the measured OD 620 biofilm values of the unexposed controls, resulting in negative IBF values.Of these isolates, 70% (n = 7) had a MIC IMI value of ≥64 and were thus exposed to subMIC concentrations of imipenem.A similar effect was observed when gentamicin, amikacin and ciprofloxacin were used alone.A total of 93.8% of isolates (15/16) with negative IBF values were exposed to sub-MIC concentrations of gentamicin, 100% of isolates (14/14) with negative IBF values were exposed to subMIC concentrations of colistin and 66.7% isolates (8/12) with negative IBF values were exposed to subMIC concentrations of amikacin.While the mean values of the IBF of gentamicin (−20.8%),amikacin (−18.3%) and ciprofloxacin (−24.2%) were low, combination therapy with fosfomycin resulted in significantly higher IBF values (39.9%, 68.1% and 52.8%, respectively).Additionally, these antibiotic combinations also led to a higher proportion of NBPs as well; see Figure 3.

Disruption of Preformed Biofilms
The disruption of biofilms was investigated in 24 h old biofilms from eight strong biofilm producers.The preformed biofilms were exposed to antibiotics for 6 h and 24 h.Neither a statistically significant reduction in biofilms nor a change in OD 620 value was observed, regardless of the duration of antibiotic exposure; see Tables 3 and 4. Nevertheless, a category change from strong to moderate biofilm producers was observed in some isolates after 6 h of antibiotic exposure.This effect was observed when the P14 biofilms were exposed to fosfomycin alone, ceftazidime and ciprofloxacin in combination and cefepime, imipenem, gentamicin, colistin and amikacin alone and in combination.The same effect was also noted after the P46 biofilms were exposed to imipenem alone and in combination and after the P36 biofilms were exposed to amikacin in combination.Similarly, 24 h of antibiotic exposure resulted in equal category downgrading, from strong to moderate, when the P36 biofilms were exposed to amikacin, colistin and ciprofloxacin in combination and piperacillin/tazobactam alone.Abbreviations: optical density (OD), biofilm disruption (BD), standard deviation (SD), atb (antibiotic), fosfomycin (FOM), ceftazidime (CAZ), piperacillin/tazobactam (TZP), cefepime (FEP), imipenem (IMI), gentamicin (GM), amikacin (AM), ciprofloxacin (CIP), colistin (COL).Abbreviations: optical density (OD), biofilm disruption (BD), standard deviation (SD), atb (antibiotic), fosfomycin (FOM), ceftazidime (CAZ), piperacillin/tazobactam (TZP), cefepime (FEP), imipenem (IMI), gentamicin (GM), amikacin (AM), ciprofloxacin (CIP), colistin (COL).

Discussion
Our results showed neither an inhibitory nor a destructive effect of fosfomycin alone on MDR and XDR P. aeruginosa biofilms.Wang et al. showed a similar result, with no antibiofilm effects on fosfomycin-susceptible P. aeruginosa isolates, despite the high antibiotic concentrations used (up to 1.024 µg/mL) [21].Fosfomycin inhibits cell wall synthesis and works best against rapidly growing bacteria, which could partly explain these results.
Weak β-lactam antibiofilm activity was anticipated due to the slow growth of biofilm bacteria, which hinders β-lactam activity targeting the peptidoglycan synthesis of actively growing and dividing bacteria [31].Nevertheless, high IBF values were achieved in our study in combination with β-lactam antibiotics.This is the first report, to the best of our knowledge, of fosfomycin combinations with the β-lactam antibiotics ceftazidime, cefepime and piperacillin/tazobactam against P. aeruginosa biofilm.Here, we have shown, for the first time, that β-lactam antibiotics have a more potent IBF effect in comparison to protein synthesis inhibitors and quinolones, even when not combined with fosfomycin.Moreover, we observed that the IBF effect was most pronounced in combinations with β-lactam antibiotics (ceftazidime, cefepime) and polypeptides (colistin) despite high resistance rates to ceftazidime, cefepime and fosfomycin.This could indicate an interaction between the synthesis of peptidoglycan, the disruption of the cytoplasmic membrane and biofilm formation.Nevertheless, the IBF of all antibiotic combinations was superior to with single use of antibiotics.Additionally, the share of NBPs was higher.
Both aminoglycosides and quinolones are frequently used to treat infections caused by P. aeruginosa.The synergy of fosfomycin/aminoglycoside and fosfomycin/quinolone combinations against its biofilm has already been documented to some extent, using different biofilm models, antibiotic representatives of these groups and different isolate numbers, sometimes only one, on isolates that are predominantly sensitive to the tested agents [21,[23][24][25][26][27][28].Anderson et al. reported the eradication of P. aeruginosa biofilms using fosfomycin and tobramycin in combination [16].The suggested synergy mechanism of quinolones and aminoglycosides with fosfomycin against planktonic cells of P. aeruginosa is that fosfomycin amends P. aeruginosa membrane permeability, leading to increased fluoroquinolone/aminoglycoside uptake.Although fosfomycin, ciprofloxacin and gentamicin were found to penetrate well into biofilms of P. aeruginosa, the lowest IBF rates were recorded for combinations with aminoglycosides and ciprofloxacin [21].Nevertheless, combinations with both aminoglycosides and quinolones resulted in significantly higher levels of IBF.To further clarify the synergistic mechanism of these antibiotic combinations against biofilms, additional studies will be required.
We also observed a vast difference between the inhibitory properties of certain antibiotics, including fosfomycin, aminoglycosides, imipenem and ciprofloxacin, against different isolates, with higher OD 620 values after exposure to antibiotics.One explanation for this effect could be the biofilm induction potential following exposure to subMIC concentrations of some antibiotics such as aminoglycosides, β-lactams and fluoroquinolones in P. aeruginosa isolates [2,22,32].
While a downgrading of the biofilm category was observed for some isolates, no statistically significant effect on biofilm disruption was observed with either the sole or combined use of antibiotics, suggesting the weak disruption potential of antibiotics for mature biofilms.One limitation of antibiotic usage against biofilms is the fact that when the treatments are initiated, the biofilm is already formed, suggesting the importance of biofilm prevention.Furthermore, antibiotics are more likely to be effective when used in the initial biofilm formation stages, while the bacterial cells are more metabolically active and thus more accessible for antibiotic therapy.Unfortunately, low antibiotic activity was exerted on already formed biofilms.However, a category downgrade indicates antibiofilm activity with potential that needs further investigation.Effects of combination therapy have largely been observed in the prevention of biofilm formation, rather than the disruption of biofilms.The use of a static or closed in vitro biofilm model provides insights into biofilm physiology but is insufficient in controlling environmental factors like nutrient flux and the capacity to wash out metabolic products, as well as host factors [16,23,33].Dynamic or open systems provide conditions that better mimic those in vivo.Further studies using more clinical isolates and different biofilm models would provide additional insight into the impact of combination therapy in terms of biofilm prevention and disruption.Furthermore, a complete chequerboard analysis would provide a more comprehensive evaluation of antibiotic combination effects.
In conclusion, our results prove that the in vitro anti-biofilm effect of antibiotic combinations with fosfomycin, particularly in terms of the IBF, is considerably superior to the single use of antibiotics when using clinically achievable concentrations, even in situations where both antibiotics are not effective when used alone.Furthermore, according to these results, the IBF potential of combinations with β-lactam antibiotics and colistin is larger than that of combinations with protein synthesis inhibitors.The results obtained in this study could prove to be clinically valuable, particularly regarding biofilm prophylaxis of not only susceptible but also antibiotic-resistant isolates.

Bacterial Isolates
We collected forty-three clinical isolates of P. aeruginosa from various clinical specimens.Forty-one isolates were collected from six hospitals in Croatia: General Hospital Bjelovar, General Hospital Slavonski Brod, General Hospital Pula, University Hospital Centre Split, University Hospital Centre Osijek and University Hospital Centre Zagreb.Vitek 2 or MALDI-TOF MS was used to identify the isolates.Two of the isolates were kindly provided by G. M. Rossolini (Department of Microbiology and Virology, Careggi University Hospital, 50135 Florence, Italy).Antibiotic susceptibility was determined in previous studies carried out by Slade-Vitković et al. [34], and the isolates were categorized as multidrug-resistant (MDR) and/or extensively drug-resistant (XDR) according to Magiorakos et al. [35].To define resistance, the fosfomycin MIC breakpoint value was set at >64 µg/mL [36].Thirteen strong and eleven moderate biofilm producers were used to evaluate the anti-biofilm properties of fosfomycin alone and when combined with antipseudomonal antibiotics.All the isolates were MDR or XDR.

Quantitative Absorbance-Based Biofilm Measurement 4.2.1. The Microtiter Plate Assay
For quantitative detection of the biofilms, the microtiter plate assay was employed [37].Isolates were grown overnight on blood agar plates, and three to four colonies of each isolate were suspended in 5 mL of TSB (Tryptic Soy Broth).The cultures were incubated for 18 h at 36 • C. The tubes were shaken and diluted in 1:100 TSB.Then, 200 µL of the diluted culture was pipetted into 96-well plates and incubated at 36 • C for a further 24 h.After the incubation period, the medium was removed, and sterile distilled water was used to carefully wash the wells three times.The plates were air-dried at 56 • C for 60 min.To stain the biofilm, 250 µL of crystal violet solution was pipetted into each well.The dye was removed after 15 min, and the wells were cautiously washed using sterile distilled water.To dissolve the attached dye, 95% ethanol was added to each well, and the content was transferred into a flat-bottomed 96-well plate.Optical density was determined twice for each well using a microtiter plate reader at 620 nm.Sterile TSB was used as a negative control.Each experiment was performed in quadruplicate.The average OD 620 values were determined for the tested isolates and for negative controls.To define the cut-off value (OD c ), the following formula was used: OD c = averageOD negative control + (3 × SD negative control ).

Table 1 .
MIC values and resistance phenotypes of P. aeruginosa isolates.

Table 2 .
Average IBF rates of antibiotics alone and in combination, expressed as means and SD.

Table 3 .
Average calculated OD 620 values of 24 h old biofilms after 6 h of exposure to single and combined antibiotics and biofilm disruption rates, expressed as means with standard deviation.

Table 4 .
Average calculated OD 620 values of 24 h old biofilms after 24 h of exposure to single and combined antibiotics and biofilm disruption rates, expressed as means with standard deviation.