Ocular flora in patients with intravitreal injections and topical antibiotic prophylaxis: antibiotic resistance patterns and susceptibility to antiseptic picloxydine

Repeated courses of topical ophthalmic antibiotics result in the increased resistance rates and changes in the composition of conjunctival flora with the significant increase of Staphylococcus epidermidis, the main causative agent of ocular infections. Antiseptic povidone-iodine is considered as a “gold standard” for prevention of infectious complications in eye surgery. Effectiveness of pre- and post-injection prophylaxis with topical antiseptics is under question. effectiveness studied against and resistant

with 20 or more IVI and repeated courses of antibiotic eye drops; without IVI and ophthalmic operations in history (control group); with antibiotic eye drops Tobrex or antiseptic eye drops Vitabact (picloxydine) applied 3 days before the first IVI and 5 days after it. In the last two groups swabs were taken at baseline and after the treatment.
Bactericidal effectiveness of picloxydine was studied in vitro against antibiotic sensitive and resistant conjunctival isolates. Minimal inhibition concentration was determined with microdilution test.

Results
Two out of three potent ophthalmic patients showed conjunctiva bacterial contamination.
Along with few Staphylococcus aureus and Gram-negative isolates susceptible to most antibiotics, the majority (71 -77%) were coagulase-negative staphylococci (CoNS), 40-50% of which multidrug resistant (MDR). Eye disinfection in the operating room and periinjection courses of Tobrex or Vitabact resulted in total elimination of isolates found at baseline. However, in 10 and 20% patients, respectively, recolonization of the conjunctiva with differing strains occurred. In patients with repeated IVI and Tobrex/Maxitrol treatment, the conjunctival flora showed high resistance rates: 90% of CoNS were MDR. In the in vitro study, picloxydine showed bactericidal effect against staphylococci isolates regardless of their antibiotic sensitivity with MIC ≥ 13,56 µg/ml. Incubation of bacteria for 15 min in Vitabact eye drops, commercially available form of picloxydine, 434 µg/ml, showed total loss of colony forming units of all tested isolates including Pseudomonas aeruginosa.

Conclusions
The confirmed bactericidal efficacy of picloxydine against conjunctival bacterial isolates and the presence of its commercial form, 0,05% eye drops, convenient for use by patients before and after injection, makes this eye antiseptic promising for prophylaxis of IVIassociated infectious complications.

Background
Intravitreal injections (IVIs) are one of the effective, widespread and minimally invasive methods of treatment of various retinal diseases. The effectiveness of such therapy is observed in treatment of exudative age-related macular degeneration, edema, associated with diabetic retinopathy or retinal vein occlusions. Due to a noticeable increase in the incidence of diabetes mellitus and cardiovascular diseases it is expected that the number of IVIs will steadily continue to rise. Generally, IVI is a safe procedure. However, like any surgical intervention, it carries the risk of potential complications. Infectious complications, associated with IVIs, occur, when a pathogen from the eye surface in the site of the injection penetrates the vitreous capacity. The most dangerous complication is the infectious endophthalmitis with visual impairment threat even in a case of proper and early treatment [1,2].
The importance of antimicrobial treatment accompanying the IVI procedure is obvious.
However, to date, there is no single approach to the management of patients regarding the use of antibacterial eye drops before and and/or after IVI as a prevention method of inflammatory infectious complications. In 2004 when the practice of IVI was just introduced, ophthalmic antibiotics were widely used for this purpose [3]. But unlike other ocular surgeries, conducted once or twice in a patient's life, where topical antibiotics may be an appropriate prophylactic measure, IVIs are often repeated to the same eye [4]. In such patients, short-term repeated courses of topical antibiotics accompanying IVI may not only reduce the risk of infectious complications but actually enhance it by increasing antibiotic resistance of conjunctival flora. In recent years, this has been confirmed in several studies.
Coagulase-negative staphylococci (CoNS), the most common bacteria on the eye surface [5], demonstrated increased rates of resistance to fluoroquinolones if isolated from eyes repeatedly exposed to one of the following ophthalmic antibiotics ofloxacin/gatifloxacin/moxifloxacin hydrochloride [6]. As shown in the study, CoNS isolates from azithromycin-exposed eyes were characterized by increased macrolide resistance.
The predominant CoNS strain Staphylococcus epidermidis developed coresistance to trimethoprim/sulfamethoxazole, gentamicin and clindamycin in fluoroquinolone-exposed eyes and to trimethoprim/sulfamethoxazole and doxycycline in azithromycin-exposed eyes [7]. Milder et al. [8] and Dorrepaal et al. [9] also found the increased antibiotic resistance of conjunctival flora due to repeated use of fluoroquinolone drops. The selection of resistant bacteria requires not so much time. Thus, bacteria colonies with high resistance to gatifloxacin were isolated from the eye of the patient who received only 3 IVI with a prophylactic use of this topical antibiotic [9].
Along with the increased resistance, repeated courses of topical eye antibiotics cause changes in the composition of conjunctival flora with the significant increase in the percentage of S. epidermidis [10]. The authors note the clinical significance of this fact because S. epidermidis is the main causative agent of ocular infections.
Grzybowski et al. [5] consider IVI to be "a prime example where unnecessary and/or improper use of antibiotics may have serious consequences". An alternative may be the use of antiseptics with efficacy comparable to antibiotics such as povidone-iodine or biguanides. Barkana et al. [11] proved that there was no significant difference between povidone-iodine, chlorhexidine (cationic biguanide) and ofloxacin in terms of reduction of conjunctival flora, 91.2%, 87.6% and 85.6%, respectively. Merani et al. [12] showed that aqueous chlorhexidine used as an antiseptic drug before IVI was well tolerated and effective in terms of low endophthalmitis rate. In conjunctival samples after the treatment with chlorhexidine 0,05%, there was a significant reduction in the total bacterial load (82%) and even greater results were observed for CoNS (90%). In this study by Gili et al. [13], no povidone-iodine was administered to the patient, eye irrigation was performed only using 0,05% chlorhexidine solution.
When comparing povidone-iodine and chlorhexidine, first is still considered as a "gold standard" for prophylaxis of infectious complications in eye surgery [12]. Along with efficacy, application of brown colored povidone-iodine, in contrast of colorless chlorhexidine, is easier for surgeon due to the visible areas of irrigation. However, povidone-iodine has its deficiencies. There is a cohort of patients with povidone sensitivity, not true immunoglobulin E-mediated allergy [14]. Sensitivity can be expressed in conjunctival hyperemia, irritation (mild to severe) and pain. In these cases, surgeon should consider using another antiseptic drug, for example chlorhexidine. Thus, Oakley and Vote [15] switched povidone-iodine to 0,1% chlorhexidine solution in patients reporting high levels of pain. As the result, the average pain score decreased from 8 of 10 points to 3 of 10.
Another antiseptic from the group of biguanides is picloxydine dihydrochloride commercially available as Vitabact, 0,05%. These approved eye drops could be useful in pre-and post-injection prophylaxis of eye infections. However, we found no data on the use of Vitabact in the management of patients with IVI.
In the study, we confirmed the increased resistance of conjunctival flora in patients with multiple IVI and antibiotic eye drops courses in anamnesis. We compared effectiveness of antiseptic Vitabact and antibiotic Tobrex eye drops in the eye surface decontamination. In the in vitro experiments, the bactericidal effect of Vitabact eye drops was proved against both antibiotic sensitive and resistant conjunctival bacterial isolates.

Methods
This was a prospective case-control study comparing 4 groups of patients, 20 patients in each group (n=80): 1) patients who received 20 or more IVI and the concomitant courses of antibiotic eye drops Tobrex based on aminoglycoside tobramycin, in some courses it was replaced by Maxitrol containing aminoglycoside neomycin, polymyxin B and dexamethasone (group 1); 2) patients of the control group, comparable in age, without IVI and ophthalmic operations in history (group 2); 3) patients with the first IVI who have applied antibiotic eye drops Tobrex 3 days before the IVI and within 5 days after it (group 3); 4) patients with the first IVI who have applied antibacterial eye drops Vitabact (picloxydine) 3 days before the IVI and within 5 days after it.
Exclusion criteria for all groups were the following: age younger 50 years old, use of systemic antibiotics within 3 months, use of ocular hypotensive drops for the glaucoma management; besides that, use of antibiotic drops and ocular surgery were exclusion criteria for the second, third and fourth groups.
All patients were recruited from the Research Institute of Eye Diseases in Moscow, Russia.
The Local biomedical ethics committee of the Research Institute of Eye Diseases approved the protocol (protocol № 49/4). Informed written consent was obtained from each patient before participation in the study.
In the standard IVI procedure, the eyelid skin and the area around the eye were treated with a 10% solution of iodopyrone. Next, eyelid speculum was applied. Local anesthetic drops of Alcaine were instilled in the conjunctival sac. The conjunctival cavity was irrigated for 30 seconds with 2,0 ml of 5% povidone-iodine and then with saline solution to wash away povidone residue.
Conjunctival swabs were performed by using sterile disposable tampons using a standard procedure (from lateral to medial angle of the eye) in the lower conjunctival fornix to the Amies transport system, which maintains the viability of microorganisms from the time of the material collection to the beginning of the study. Care was taken to minimize the contact with lashes, eyelids and skin. In the groups of patients 3 and 4 conjunctival swabs were taken both before (at baseline) and the next day after the end of post-injection treatment with Tobrex or Vitabact.
In positive swabs, the isolated microorganisms were identified and tested for antibiotic susceptibility by BD Phoenix 100 automated identification and susceptibility testing system. Statistical analysis was conducted with SPSS via contrasting the respective 95% and 99% confidence intervals (based on the estimates of group means and standard deviations).
In the group of 20 patients with IVI and antibiotic eye drops Tobrex prophylaxis, the swabs were obtained before (20 eyes) and after (20 eyes) this treatment. Like in the control group, at baseline before the treatment microflora growth was observed in 14 swabs (70%, 14/20).
In one case, Gram negative Enterobacter cloacae was isolated ( Table 1). The remaining 13 cases represented different types of staphylococci, including 3 S. aureus (23,08%, 3/13) and 10 CoNS (76,92%, 10/13). One of the S. aureus isolates was antibiotic sensitive, the second resistant to penicillin G and the third to tobramycin and tetracycline. As for 10 CoNS, 2 isolates were sensitive to all tested drugs, 3 isolates showed resistance to drugs from 2 antibiotic classes and 5 were MDR (Table 2). Among CoNS isolates, 30% (3/10) belonged to MRS, 20% (2/10) were gentamicin and tobramycin and 20% (2/10) -ciprofloxacin resistant ( Fig. 1). This corresponded to the control group. However, in the groups of ophthalmic patients before IVI and Tobrex treatment, the percentage of erythromycin resistant CoNS reached 70% (7/10) that was much higher as compared with the control group. Doxycycline resistant isolates were found in group of patients before IVI and Tobrex treatment but not in the control group (Fig. 1). In the same group of 20 patients after IVI and treatment with Tobrex (20 eyes), only two conjunctival swabs were positive. In one patient, no growth was observed in swabs taken at the first visit, and after the treatment with Tobrex and IVI, sensitive S. hominis was isolated (Table 1). In the swabs of the second patient S. haemolyticus resistant to erythromycin, chloramphenicol and fosfomycin (includes glucose-6-phosphate) was isolated in the first visit, and S. lugdunensis resistant to fosfomycin was obtained at the second visit.

Gramnegative
Enterobac ter cloacae As for antibiotic sensitivity in the group with multiple IVI and antibiotic eye drops treatment, we found no CoNS sensitive or resistant to 1 drug. Only one isolate was resistant to antibiotic of two classes ( Table 2). MRS were 50% (5/10) of CoNS isolates. Moreover, the conjunctival flora in such patients was characterized by an increase in the number of strains resistant to a wide range of antibiotics. Nine of 10 CoNS were MDR (90%). Among these, we found S. epidermidis isolate resistant to 11 antibiotic classes. In this group of patients who have received repeated courses of aminoglycoside-containing eye drops Tobrex/Maxitrol, 90% (9/10) CoNS and the single isolate of S. aureus were gentamicin and tobramycin resistant (Fig. 1). This threatening situation encouraged us to try antiseptic picloxydinecontaining eye drops Vitabact in peri-injection antimicrobial prophylaxis.
In the group of 20 patients who were prescribed antiseptic eye drops Vitabact 3 days before the first IVI (20 eyes) and within 5 days after it (20 eyes), the swabs of 6 (30%, 6/20) patients were negative at baseline and after the treatment. was negative before the treatment, but after the treatment antibiotic-sensitive E. coli was isolated, which can also be explained by secondary infection due to eye hygiene breaches.
In general, 80% of swabs taken next day after the end of the treatment with Vitabact were negative.  isolates were also included in the study: P. aeruginosa, P. luteola and E. coli.
After 24 hours, we did not detect growth of any staphylococci in series of liquid growth media containing double-diluted Vitabact (from 1:2 to 1:32) that corresponded to decrease in picloxydine concentration from 217,00 to 13,56 µg/ml. Control suspensions without Vitabact showed bacteria growth, as observed by absorbance increase. In Table 4 we summarized the growth parameters of staphylococci isolates in nutrient broth without addition or in the presence of Vitabact in its lowest concentration (1:32 dilution) tested in our study. In order to identify possible differences in the Vitabact effect on MDR strains, coagulase-negative staphylococci that made up the most of isolates were grouped by their sensitivity. S. aureus formed one group, as among the few S. aureus MDR strains were not isolated. We found no differences (at any reasonable statistical significance level) in picloxydine inhibitory effect on growth of isolates, MDR or resistant to no more than two drugs, as well as CoNS and S. aureus.
After 96 h growth, the absorbance changes were insignificant as compared with 24 h for all groups apart from S. aureus (24 h vs 96 h). For the latter, there is an indication of statistical difference at 5% level, although the groups are not statistically different at 1% level. Only one isolate, S. epidermidis resistant to clindamycin, chloramphenicol and erythromycin, showed the growth of absorbance to the value about 0,25. Its contribution in the average MDR group absorbance after 96 h growth is seen from increased standard deviation ( Table   4). The results again did not give a difference depending on antibiotic sensitivity and staphylococci species (Vitabact effect was statistically strong in all isolates at 1% level).
Probes of each staphylococci culture grown in nutrient broth for 24 or 96 h with or without Vitabact dilutions were cultivated further on agar plates within 24 h. Those taken from picloxydine-containing liquid media samples showed no growth except one mentioned isolate with the lowest picloxydine concentration tested (13,56 µg/ml). These data confirmed bactericidal properties of Vitabact. Thus, the picloxydine minimal inhibitory concentration (MIC) for staphylococci conjunctival isolates ≥ 13,56 µg/ml.
Among the Gram-negative bacteria, the most resistant was P. aeruginosa isolate. It grew in liquid medium even with 217,00 µg/ml picloxydine content (1:2 deluted Vitabact). E. coli and P. luteola were much more sensitive, the minimal picloxydine concentrations that inhibited the growth of these isolates for 24 h were 54,25 and 13,56 µg/ml, respectively.

Discussion
In our study, at least two out of three potent ophthalmic patients showed conjunctiva bacterial contamination: in the groups 3 (n=20) and 4 (n=20) the baseline swabs of 14 and 13 patients were positive. Consistent with other studies, reviewed in [5], the majority were CoNS (71 -77%) followed by S. aureus (15 -21%) and single Gram-negatives. Among CoNS the most frequently isolated was S. epidermidis (70 -100%). S. epidermidis is believed to prevent the colonization of conjunctiva by more serious pathogens [16]. CoNS (93%, 87% of these S. epidermidis) and S. aureus (7%) constituted the ocular flora in the control group 1 (n=20).
Antibiotic resistance rate of ocular flora and especially CoNS in potent ophthalmic patients (groups 3 and 4, preoperatively) requires increased attention when prescribing pre-and post-injection prophylactic antimicrobials. Along with S. aureus and Gram-negative isolates susceptible to most antibiotics, 40-50% CoNS in our study were MDR. In view of the continuing practice of antibiotic eye drops prophylactic peri-injection treatment in Russian Federation, we should mention that CoNS resistant to gentamycin/tobramycin were found in 20% and to moxifloxacin in 10% of potent ophthalmic patients.
In patients who have received 20 or more IVI and concomitant prophylactic courses of antibiotic therapy, a significant increase in the resistance of the conjunctival flora to a wide range of antibiotics was observed. Nowadays asepsis and antisepsis are beneficial rather than use of topical antibiotics as a prevention method of post-injection complications [4]. Aseptic and antiseptic techniques include performing injections in the operating rooms, application of povidone-iodine and peri-injection prophylactic treatment with antiseptic drops. Ultraclean air and good ventilation are required in operating rooms [2]. In order to reduce the risk of infection, the spread of pathogens from the oral cavity of patients and medical personnel should be minimized [17] by using sterile masks by surgeons and nurses and sterile adhesive eye drapes that isolate patients' nasopharyngeal area and periocular region [18]. It is important to prevent the contact of eyelashes and eyelid margins from the injection site and the needle, through which the medication is injected into the vitreous cavity. This can be achieved by using an eyelid speculum which remove the eyelashes, potential source of infection of the needle tip [17]. The preparation of the ocular surface should include irrigation with a solution of povidone-iodine for at least 30 seconds [19]. Precisely, irrigation of the conjunctiva is needed, not a drop application of the solution, that corresponds to Safar and Dellimore [20] findings. A single application of povidone-iodine demonstrates a bactericidal effect, equivalent to the 3-day course of local antibiotics [21].
Several studies have shown that resistance to povidone-iodine does not develop [22,23] unlike the reported reduced levels of susceptibility to chlorhexidine [24], so we can safely continue using povidone-iodine solution in the operating rooms.
As to pre-and post-injection prophylaxis, in patients, who used the Vitabact antibacterial eye drops 3 days before the first IVI and within 5 days after it, 80% of swabs taken the day after the end of the treatment were negative. Thus, the effectiveness of a single pre-and post-injection course of Vitabact was close to that of Tobrex (90% negative swabs). In positive conjunctival swabs taken the day after the end of post-injection treatment with Tobrex or Vitabact, 2 and 4 isolates were found, respectively. However, these isolates differed in their resistance to certain antibiotics from those found at baseline in the same patients. This means that the isolates found at baseline were eliminated with the povidoneiodine irrigation before the injection and/or eye drops treatment. Most likely, after the end of the eye drops post-injection treatment a rapid, within one day, recolonization of the conjunctiva occurred.
In the in vitro study with microdilution test, picloxydine had bactericidal effect against 39 CoNS, 5 S. aureus, E. coli and P. luteola isolates regardless of their antibiotic sensitivity.
The picloxydine MIC for staphylococci ≥ 13,56 µg/ml. Another test with incubation of bacteria 15 min in Vitabact eye drops, commercially available form of picloxydine with concentration 434 µg/ml, resulted in total loss of CFU of 10 conjunctival Gram-positive isolates, both antibiotic sensitive or MDR, and Gram-negative E. coli and P. aeruginosa.

Conclusions
The confirmed bactericidal efficacy of picloxydine against conjunctival bacterial isolates before and after injection, makes this eye antiseptic promising for prophylaxis of IVIassociated infectious complications. Availability of data and material The datasets used and/or analysed during the current study are available from the corresponding author (Khalatyan Anait S.) upon reasonable request.

Competing interests
The authors declare that they have no competing interests.

Funding
None.
Authors contributions BMV, SMG and ZVG carried out the study concept and design. BMV, KAS, SMG and ZVG collected and interpreted the data, researched relevant literature. KAS performed the statistical analysis. All authors wrote the manuscript and approved the final manuscript.