original clinical study Optimal Duration for the Use of.5% Levofloxacin Eye Drops Before Vitreoretinal Surgery Xiaoxin Li, MD,* Xiaoling Liang, MD, Luosheng Tang, MD, Junjun Zhang, MD, Lijun Shen, MD, Guanfang Su, MD, and Xiaorong Li, MD** Purpose: To investigate various regimens of prophylactic antibiotic therapy for vitreoretinal surgery. Design: This study compared different prophylactic therapies with.5% levofloxacin eye drops. Methods: Two hundred nine patients from 7 hospitals scheduled to undergo vitreoretinal surgery were randomized into 3 groups to receive.5% levofloxacin eye drops for 1 day (6 times/d), 2 days (3 times/d), or 3 days (3 times/d) before surgery (groups 1D, 2D, and 3D, respectively). All patients received 3 applications of levofloxacin eye drops at 15-minute intervals beginning 1 hour before surgery and conjunctival sac disinfection with 5.% povidone-iodine 15 minutes before surgery. Conjunctival swabs were cultured before levofloxacin therapy (T), on the morning of surgery (T1), after povidone-iodine disinfection (T2), immediately after surgery (T3), 1 day postoperatively (T4), and 1 week postoperatively (T5). Results: The positive bacterial culture rates in groups 1D, 2D, and 3D fell, respectively, from 37.33%, 3.77%, and 31.88% at T to 1.67%, 12.31%, and 11.59% at T1 and 1.33%, %, and % at T2. At each time point (T T5), there were no significant differences among the groups in positive bacterial culture rate. The bacterial eradication rates in groups 1D, 2D, and 3D were, respectively, 1%, 94.74%, and 9.% at T1 (after levofloxacin) and 1% in all groups at T2 (after povidone-iodine). Conclusions: The efficacy of levofloxacin in preventing postoperative infection was similar in the 3 treatment groups. It is recommended that.5% levofloxacin be used for only 1 day before vitreoretinal surgery (6 times/d) to minimize the use of prophylactic antibiotics. Key Words: antibiotic, levofloxacin, prophylaxis, vitreoretinal surgery (Asia-Pac J Ophthalmol 217;1:4 44) The number of vitreoretinal surgeries has continued to increase in recent years. However, postoperative infection remains a serious potential complication of vitreoretinal surgery, From the *Peking University People s Hospital, Beijing; Yat-Sen Ophthalmic Center of Sun Yat-Sen University, Guangzhou; Xiangya Second Hospital of Central South University, Changsha; Sichuan University Huaxi Hospital, Chengdu; Eye Hospital of Wenzhou Medical University, Wenzhou; Second Hospital of Jilin University EyeHospital, Changchun; and **Eye Center of Tianjin Medical University, Tianjin, China. Received for publication December 29, 215; accepted April 6, 216. The study was sponsored by Santen Pharmaceutical Co, Ltd, Osaka, Japan. The sponsor or funding organization had no role in the design or conduct of this research. The authors have no other funding or conflicts of interest to declare. Reprints: Xiaoxin Li, MD, Department of Ophthalmology, Peking University People s Hospital, No. 11 Xizhimen South Avenue, Xicheng District, Beijing 144, China. E-mail: lixiaolin1512@126.com. Copyright 217 by Asia Pacific Academy of Ophthalmology ISSN: 2162-989 DOI: 1.2268/APO.215197 with a reported incidence of.3% to.13%. 1,2 The incidence of endophthalmitis varies among and within countries because of different perioperative asepsis techniques and antibiotic use. 3 Studies of human endophthalmitis have shown that the causative pathogens are primarily from the eyelid and conjunctival sac flora 4 ; hence, preoperative sterilization of the conjunctival sac is considered key to successful surgery. It is generally accepted that the preoperative use of antibiotics in patients undergoing intraocular surgery can effectively reduce the bacteria-positive rate of the conjunctival sac. 5 A study by the European Society of Cataract and Refractive Surgeons (ESCRS) showed that the use of perioperative antibiotics could reduce the rate of postoperative infection. The ESCRS 27 guidelines cite preoperative topical antibiotic prophylaxis as an option to consider. 6 In a study of patients receiving preoperative ofloxacin to prevent postoperative infection after cataract surgery, Ta et al 7 reported that the bacteria-positive rate fell from 68% to 19% in patients receiving the antibiotic for 3 days before surgery and from 6% to 42% in patients receiving ofloxacin 1 hour before surgery (with a significant difference between the 2 groups). A subsequent study by Inoue et al 5 determined that the use of levofloxacin for 3 days before surgery (3 times/d, plus once at 1 hour before surgery) was superior to its use for only 1 day before surgery (3 times/d, plus once at 1 hour before surgery) and to a single administration 1 hour before surgery. In contrast, a single-center study by Ta et al 8 concluded that the use of levofloxacin for 3 days before surgery (4 times/d) resulted in a reduction in the bacteria-positive rate (from 74% to 34%) that was not significantly different from that in response to therapy for only 1 day before surgery (from 82% to 46%). Although these studies concur that a single administration of antibiotic 1 hour before surgery is inferior to medication for 3 days before surgery, there remains debate as to whether preoperative antibiotic use for 3 days offers benefits over its use for only 1 day before surgery. It should be noted that the frequency of administration differed between the studies of Inoue et al 5 (3 times/d) and Ta et al 8 (4 times/d), raising the possibility that this could have influenced the observed efficacy of antibiotic therapy for 1 day before surgery. Although it is critical that the conjunctival sac is cleared of bacteria before surgery, it is also important to decrease the possibility of bacterial drug resistance by minimizing the preoperative use of antibiotic eye drops. Therefore, we designed and conducted a randomized, open, parallel, multicenter study to measure the rate of eradication of conjunctival sac bacteria in response to the use of.5% levofloxacin eye drops before intraocular surgery to determine whether there are any differences in effectiveness among preoperative administration for 1 day (6 times/d), 2 days (3 times/d), or 3 days (3 times/d). Our novel findings provide a reference that will help to optimize the prophylactic use of preoperative antibiotics in patients undergoing vitreoretinal surgery. 4 www.apjo.org
Levofloxacin Use Before Vitreoretinal Surgery materials and methods Patient Selection and Randomization The present study enrolled consecutive patients attending 7 hospitals between January 212 and July 212. The inclusion criteria were as follows: (1) male or female, aged 18 to 7 years; (2) undergoing vitreoretinal surgery for the first time; (3) had not participated in any other clinical trial during the previous 3 months; and (4) voluntary participation, with informed written consent provided. The following patients were excluded: (1) those with diabetes, immunologic diseases, or ocular autoimmune diseases such as retinal vasculitis; (2) those who had received previous vitreoretinal surgery; or (3) those in whom the use of intravitreal silicone oil tamponade was planned. The study was approved by the ethics committees of the hospitals involved and was conducted in accordance with the tenets of the Declaration of Helsinki and its subsequent revisions (213). The included patients were allocated into 1 of 3 groups using random numbers produced by SAS version 9.2 (SAS Institute Inc, Cary, North Carolina): group 1D, to receive.5% levofloxacin eye drops for 1 day (6 times/d) before surgery; group 2D, to receive.5% levofloxacin eye drops for 2 days (3 times/d) before surgery; and group 3D, to receive.5% levofloxacin eye drops for 3 days (3 times/d) before surgery. At each administration, 1 drop of.5% levofloxacin was applied to the conjunctival sac. Beginning 1 hour before surgery, patients in all groups received 3 applications of.5% levofloxacin eye drops (1 drop per application) at 15-minute intervals, along with 5% povidone-iodine disinfection liquid applied to the conjunctival sac 15 minutes before surgery. After surgery, all patients were administered.5% ofloxacin eye drops for 7 days (3 times/d, 1 drop per administration). Conjunctival Sac Bacterial Specimens Conjunctival swabs were taken before the administration of levofloxacin eye drops (T), early in the morning on the day of surgery (T1), after preoperative disinfection of the eye with povidone-iodine (T2), immediately after surgery (T3), 1 day after surgery (T4), and 1 week after surgery (T5). Bacterial specimens were collected by appropriately trained medical staff. With the patient looking upward, the lower eyelid was everted and the dome of the conjunctiva swabbed from one end to the other in a horizontal direction, with care taken not to miss the inner canthus; the swab was then rotated and the process repeated with the other side of the swab. The collected specimens were immediately inoculated on blood agar, chocolate agar, and MacConkey agar plates, and the agar plates were cultured at 35 C with 5% CO 2 to test for aerobic and facultative anaerobic bacteria. Bacteriologic Efficacy Indexes The positive bacterial culture rate and the bacterial eradication rate were used as evaluation indexes. The positive bacterial culture rate at each time point was calculated as the number of patients with positive bacterial cultures divided by the total number of patients. The bacterial eradication rate at a given time point was calculated as the number of positive bacterial cultures at the previous time point minus the number of positive bacterial cultures at the present time point, divided by the number of positive bacterial cultures at the previous time point [ie, (positive T-1 positive T ) / positive T-1 ]. Statistical Analysis Statistical analyses were performed using SAS 9.2 (SAS Institute Inc) and Stata 12 (StataCorp, College Station, Texas) software. Data are presented as the mean ± SD, n (%), or median (range) as appropriate. Statistical comparisons among groups were carried out using analysis of variance for continuous data, the χ 2 test or Fisher exact test (for data with a sample size of no more than 5) for categorical data, and the Kruskal-Wallis test for descriptive statistics data. All statistical tests were 2-sided. P <.5 was considered to indicate statistical significance. RESULTS Demographic and Clinical Characteristics of the Patients A total of 29 patients were enrolled: 75 in group 1D, 65 in group 2D, and 69 in group 3D. Of these, 78.5% underwent vitrectomy, and 21.5% received a scleral buckling operation. The baseline demographic and clinical characteristics of the included patients are presented in Table 1. There were no significant differences among the 3 groups in any of these characteristics, including patient age, sex (% female), body mass index, blood pressure, eye involved (% left eye involvement), intraocular pressure of affected eye, visual acuity of affected eye, or positive bacterial culture rate. Species of Bacteria Cultured Before Surgery Figure 1 shows the distribution of the main bacterial strains Table 1. Baseline Demographic and Clinical Characteristics of the Patients Included in the Study Characteristic Total (N = 29) 1D (n = 75) 2D (n = 65) 3D (n = 69) P Age, y Sex, female, n (%) Body mass index, kg/m 2 Blood pressure, mm Hg Systolic Diastolic Left eye lesion, n (%) Intraocular pressure of affected eye, mm Hg Visual acuity of affected eye Positive bacterial culture rate, n (%) 53.61 ± 13.2 98 (46.89) 23. ± 3.27 132.25 ± 15.52 79.49 ± 1.76 93 (44.5) 12.98 ± 3.51 (, 1.4) 7 (33.49) 52.51 ± 14.35 42 (56.) 22.86 ± 2.93 13.18 ± 12.23 79.19 ± 9.42 28 (37.33) 12.62 ± 3.23.1 (, 1.4) 28 (37.33) 54.22 ± 12.52 28 (43.8) 22.67 ± 3.19 132.65 ± 18.84 78.49 ± 11.91 28 (43.8) 13.26 ± 3.43 (,.6) 2 (3.77) 55.72 ± 12.53 28 (4.58) 23.46 ± 3.68 134.12 ± 15.18 8.74 ± 11. 37 (53.62) 13.1 ± 3.88 (, 1.2) 22 (31.88).352.137.382.39.463.14.63.142.673 www.apjo.org 41
Li et al Table 3. Intragroup Comparisons of Positive Bacterial Culture Rates at Different Time Points P for Comparisons of Positive Bacterial Culture Rates Within Each Treatment Group 1D 2D 3D A T1 vs T T2 vs T1 T3 vs T2 T4 vs T3 T5 vs T4 <.1.2.8.564.2.564 to T2 was observed (Table 3). The number of new bacterial strains was higher at T1 (9, 7, and 6 for the 3 treatment groups, respectively) than at T2 to T5. B figure 1. Distribution of the bacterial strains cultured from conjunctival swabs. A, Distribution of bacterial strains before the start of therapy with levofloxacin eye drops (T). B, Distribution of bacterial strains after therapy with levofloxacin eye drops, early in the morning on the day of surgery (T1). At both time points, the most commonly identified strain was S. epidermidis, followed by S. hominis. Strains comprising less than 2% of the total are collectively referred to as Others. cultured from the conjunctival swabs of patients before the start of antibiotic therapy (T) and on the morning of the vitreoretinal surgery (T1). At both T and T1, the most common strain cultured was Staphylococcus epidermidis, followed by Staphylococcus hominis. Positive Bacterial Culture Rates Table 2 compares the positive bacterial culture rates for all patients and for the individual treatment groups. The positive bacterial culture rates in groups 1D, 2D, and 3D fell, respectively, from 37.33%, 3.77%, and 31.88% at T to 1.67%, 12.31%, and 11.59% at T1 and 1.33%, %, and % at T2. The positive bacterial culture rate dropped to very low levels (<1.5%) at T2, and low levels (<5%) were maintained from T3 to T5 (Table 2). Intergroup comparisons revealed no significant differences among any of the groups at any time point. However, within each treatment group, there was a significant reduction in the positive bacterial culture rate from T to T1, and for the 1D group, a further significant decrease in the positive bacterial culture rate from T1 Conjunctival Sac Bacterial Eradication Rate The bacterial eradication rates in the 3 treatment groups at various time points (T T5) are shown in Table 4. The bacterial eradication rate in response to the administration of levofloxacin eye drops (T1) exceeded 9% in all 3 study groups and reached 1% after povidone-iodine administration just before surgery (T2) (Table 4). Importantly, there were no significant differences in the bacterial eradication rates among the groups at any time point. Adverse Events A drug-related adverse event was observed in only 1 (.48%) of the 29 patients included in the study. This drug-related adverse event occurred in a patient in the 2D group who was diagnosed with endophthalmitis that occurred after vitrectomy (the organism isolated was S. epidermidis). Drug-related adverse eventswere not observed in the 1D and 3D groups. DISCUSSION The results of this study indicate that the preoperative use of.5% levofloxacin eye drops had significant antibacterial effects, irrespective of whether the eye drops were administered for 1, 2, or 3 days before surgery. Inoue et al 5 reported that the application of levofloxacin (3 times/d) for 3 days could achieve a higher disinfection rate than a 1-day application. In contrast, Ta et al 8 failed to find any superiority of a 3-day application over a 1-day application, but their protocol used a higher daily frequency of levofloxacin application (4 times/d), raising the possibility that the effectiveness of a 1-day prophylaxis may depend on the number of times the antibiotic eye drops are administered. In the present study, there were no significant differences among the 3 treatment Table 2. Comparison of Positive Bacterial Culture Rates Among Study Groups Total (N = 29) 1D (n = 75) 2D (n = 65) 3D (n = 69) P T positive bacterial culture rate T1 positive bacterial culture rate T2 positive bacterial culture rate T3 positive bacterial culture rate T4 positive bacterial culture rate T5 positive bacterial culture rate 7 (33.49%) 24 (11.48%) 1 (.48%) 4 (1.91%) 7 (3.35%) 4 (1.91%) 28 (37.33%) 8 (1.67%) 3 (4.%) 2 (3.77%) 8 (12.31%) 3 (4.62%) 2 (3.8%) 1 (1.54%) 22 (31.88%) 8 (11.59%) 2 (2.9%) 2 (2.9%).673.954.117.839 42 www.apjo.org
Levofloxacin Use Before Vitreoretinal Surgery Table 4. Comparison of Bacterial Eradication Rates Among Study Groups 1D (n = 29) 2D (n = 2) 3D (n = 22) P T1 bacterial eradication rate T2 bacterial eradication rate T3 bacterial eradication rate T4 bacterial eradication rate T5 bacterial eradication rate 27 9 1 1 3 27 (1%) 9 (1%) 1 (1%) 1 (.%) 3 (1%) 19 8 2 2 18 (94.74%) 8 (1%) 2 (.%) 2 (1%) 2 8 3 18 (9.%) 8 (1%) 3 (1%).264 groups in the antibacterial effect of levofloxacin. A notable difference between the present investigation and those of Inoue et al 5 and Ta et al 8 is that the frequency of levofloxacin application in this study was higher in the 1D group (6 times/d) than in the 2D and 3D groups (3 times/d). Therefore, it may be the case that a 1-day course of levofloxacin is as effective as a 3-day course only if a sufficient number of doses (at least 4) are administered each day. Further studies are merited to explore this possibility in more detail. The ESCRS 27 guidelines recommended prophylaxis including perioperative antisepsis and antibiotics. 6 Preoperative prophylaxis can include intravenous, oral, and topical antibiotics. Topical antibiotic use is common in most countries. 3,9 The duration of topical antibiotic use ranges from 1 to 3 days. 5,7,8 Based on the present findings and those of related studies, it is recommended that the administration of.5% levofloxacin eye drops should be initiated 1 day before surgery. This protocol would shorten the duration of prophylactic antibiotic use, potentially reducing the development of bacterial drug resistance. Although the use of.5% levofloxacin eye drops was associated with a high rate of bacterial eradication from the conjunctival sac at T1 (Table 4), the positive bacterial culture rate did not show a corresponding reduction, mainly due to the emergence of new strains. The distribution of the strains (Fig. 1) showed that most were gram-positive bacteria (such as S. epidermidis and S. hominis) normally present in the conjunctival sac and meibomian glands. The distribution of new bacterial strains detected at T1 to T5 was similar (Fig. 1), with S. epidermidis and S. hominis most common. These distributions were consistent with the ES- CRS 27 guidelines, in which coagulase-negative staphylococci were most common. The finding of a positive bacterial culture rate of greater than 1% on the morning of the day of the procedure indicates that, although antibiotic eye drops can eliminate bacteria, a fully sterile status cannot be achieved, necessitating the use of povidone-iodine disinfection. Indeed, preoperative povidone-iodine administration was associated with very low positive bacterial culture rates at T2. Povidone-iodine is an effective disinfectant that can rapidly kill the majority of microorganisms. Studies have shown that the use of 5.% povidone-iodine is an effective approach to reducing postoperative endophthalmitis 1 and that the combination of povidone-iodine and topical antibiotics is more effective than either alone for eradicating bacteria. 8 These findings are consistent with those of the present study. The design of this study did not include a direct comparison between topical antibiotic use (such as.5% levofloxacin eye drops) and povidone-iodine disinfection; hence, additional studies are merited to explore this further. One drug-related adverse event (endophthalmitis after vitrectomy due to S. epidermidis) was recorded among the 29 patients in this study. Eyelid edema occurring after retrobulbar anesthesia had compressed the meibomian gland, leading to postoperative infection. Bacterial culture confirmed that the organism was levofloxacin resistant. In the present study, the positive bacterial culture rate before the administration of levofloxacin (ie, at T) was approximately 32% to 37%. It is notable that the corresponding rates reported by previous studies have varied widely: 74% to 82%, 8 6% to 68%, 7 65%, 11 45.8%, 12 4.5%, 13 4%, 14 and 18.4%. 15 Nonetheless, the positive culture rates in this study were broadly in keeping with these previously published findings. The reasons underlying this wide variation among investigations remain unknown, although factors reported to be associated with conjunctival bacterial colonization include spring/summer season and showering on the morning of the operation. 13 However, it cannot be ruled out that differences among studies in the sensitivities of the bacterial culture methods may have contributed to the variation in positive bacterial culture rates. The present study is not without limitations. First, considering the low rate of endophthalmitis after vitreoretinal surgery, the sample size was too small to allow comparison of the rates of endophthalmitis among the treatment groups. Second, a control group that did not use levofloxacin prophylaxis was not included, so it was not possible to determine the benefits of preoperative levofloxacin on bacterial eradication. Third, the frequency and total dosage of levofloxacin administration differed among the groups; the inclusion of additional groups to enable more direct comparisons to be made (eg, 3 times/d for 1 day vs 3 times/d for 2 days vs 3 times/d for 3 days to compare different regimens using the same daily frequency and 2 times/d for 3 days vs 3 times/d for 2 days vs 6 times/d for 1 day to compare different regimens using the same total dosage) would have provided a more detailed assessment of the optimal prophylactic regimen. In conclusion, this study has shown that there were no significant differences in the sterilization effects of.5% levofloxacin administered preoperatively for 1 day (6 times/d), 2 days (3 times/d), or 3 days (3 times/d). Therefore, it is recommended that.5% levofloxacin eye drops should be administered for only 1 day before vitreoretinal surgery to reduce the duration of prophylactic antibiotic use and decrease the risk of antibiotic resistance without compromising medical efficacy. ACKNOWLEDGMENTS The authors thank Santen Pharmaceutical (China) Co, Ltd, for providing the study sample and financial support. 1. references Foster RE, Rubsamen PE, Joondeph BC, et al. Concurrent endophthalmitis and retinal detachment. Ophthalmology. 1994;11:49 498. www.apjo.org 43
Li et al 2. 3. 4. 5. 6. 7. 8. Shimada H, Nakashizuka H, Hattori T, et al. Incidence of endophthalmitis after 2- and 25-gauge vitrectomy: causes and prevention. Ophthalmology. 28;115:2215 222. Behndig A, Cochener B, Guell JL, et al. Endophthalmitis prophylaxis in cataract surgery: overview of current practice patterns in 9 European countries. J Cataract Refract Surg. 213;39:1421 1431. Speaker MG, Milch FA, Shah MK, et al. Role of external bacterial flora in the pathogenesis of acute postoperative endophthalmitis. Ophthalmology. 1991;98:639 649; discussion 65. Inoue Y, Usui M, Ohashi Y, et al. Preoperative disinfection of the conjunctival sac with antibiotics and iodine compounds: a prospective randomized multicenter study. Jpn J Ophthalmol. 28;52:151 161. Barry P, Behrens-Baumann W, Pleyer U, et al. ESCRS guidelines on prevention, investigation and management of post-operative endophthalmitis. 27. Ta CN, Egbert PR, Singh K, et al. Prospective randomized comparison of 3-day versus 1-hour preoperative ofloxacin prophylaxis for cataract surgery. Ophthalmology. 22;19:236 24; discussion 24 231. Ta CN, Sinnar S, He L, et al. Prospective randomized comparison of 1-day versus 3-day application of topical levofloxacin in eliminating conjunctival flora. Eur J Ophthalmol. 27;17:689 695. 9. Fay A, Nallasamy N, Bernardini F, et al. Multinational comparison of prophylactic antibiotic use for eyelid surgery. JAMA Ophthalmol. 215; 133:778 784. 1. Ciulla TA, Starr MB, Masket S. Bacterial endophthalmitis prophylaxis for cataract surgery: an evidence-based update. Ophthalmology. 22; 19:13 24. 11. Leong JK, Shah R, McCluskey PJ, et al. Bacterial contamination of the anterior chamber during phacoemulsification cataract surgery. J Cataract Refract Surg. 22;28:826 833. 12. Yang W, Kuang L, Deng F, et al. Study on bacterial strains in conjunctival sac before and after intraocular surgery [translated from Chinese]. Yan Ke Xue Bao. 1999;15:267 269. 13. Halachmi-Eyal O, Keness Y, Lang Y, et al. Risk factors for conjunctival microorganism colonization in adults undergoing intraocular surgery. OJOph. 212;2:26 3. 14. Halachmi-Eyal O, Lang Y, Keness Y, et al. Preoperative topical moxifloxacin.5% and povidone-iodine 5.% versus povidone-iodine 5.% alone to reduce bacterial colonization in the conjunctival sac. J Cataract Refract Surg. 29;35:219 2114. 15. Herde J, Tost M, Wilhelms D, et al. Perioperative conjunctival flora [translated from Chinese]. Klin Monbl Augenheilkd. 1996;29:13 2. 44 www.apjo.org