ELIMINATING ANTIBIOTIC PROPHYLAXIS FOR INTRAVITREAL INJECTIONS A Consecutive Series of 18,839 Injections by a Single Surgeon ABDHISH R. BHAVSAR, MD,* DANIELLE R. SANDLER, BA† Purpose: By optimizing the protocol for intravitreal injections, the risk of endophthalmitis can be minimized. This study assesses the incidence of endophthalmitis and other complications after a consecutive series of intravitreal injections where all antibiotics were excluded. Methods: Injections were performed from August 1, 1997 to October 31, 2012 in outpatient examination rooms at the Retina Center of Minnesota by a single retinal surgeon, the lead author. Most injections were performed to treat exudative age-related macular degeneration. Other reasons included diabetic macular edema, cystoid macular edema because of retinal vein occlusions, cytomegalovirus retinitis, and severe uveitis. Injections were given with topical povidone–iodine, proparacaine, and tetracaine, a sterile eyelid speculum, and clean nonsterile gloves, but without any antibiotics. Data were retrospectively analyzed using billing codes from a computer database system. Results: A total of 18,839 injections were given. Of these, the following injections were administered: bevacizumab, 15,479 (82.16%); ranibizumab, 1,669 (8.86%); triamcinolone acetonide (Kenalog-40), 1,014 (5.38%); pegaptanib sodium, 370 (1.96%); aflibercept, 148 (0.79%); dexamethasone implant, 88 (0.47%); triamcinolone acetonide (Triesence), 32 (0.17%); dexamethasone, 29 (0.15%); and ganciclovir, 10 (0.05%). There was one case of postinjection endophthalmitis. The incidence of endophthalmitis per injection was 0.0053%. Conclusion: A low incidence of endophthalmitis can be achieved when topical antibiotics are omitted. RETINA 35:783–788, 2015
T
24 hours after intraocular surgery. Conversely, polymyxin B sulfate–neomycin sulfate–gramicidin, a broad-spectrum topical antibiotic, has a significantly shorter duration of bactericidal activity.3 Thus, there is evidence that povidone–iodine can more effectively eradicate pathogens than topical antibiotics. Given this, it is logical to inquire whether topical antibiotics are a necessary addition to povidone–iodine. Isenberg et al2 found that there may be a synergistic effect of topical antibiotics and povidone–iodine in reducing the culture positive rate of the preoperative conjunctival surface in patients undergoing intraocular surgery. Moss et al,4 however, reached a contrasting conclusion in a related study on patients undergoing intravitreal injections. They concluded that the addition
here is a great deal of controversy about the efficacy of topical antibiotics in the prevention of endophthalmitis after intravitreal injections. Topical antibiotics are routinely administered as part of the intravitreal injection procedure for endophthalmitis prophylaxis. However, to our knowledge, there has never been a study proving that any antibiotic can effectively reduce the risk of postinjection endophthalmitis. Nevertheless, there is evidence that topical povidone–iodine alone reduces the risk of endophthalmitis after intraocular surgery.1 According to Isenberg et al,2 one application of povidone–iodine has a bactericidal effect equivalent to a 3-day course of topical antibiotics. In addition, the antimicrobial effect of a drop of povidone–iodine 5% has been shown to last at least 783
784
RETINA, THE JOURNAL OF RETINAL AND VITREOUS DISEASES 2015 VOLUME 35 NUMBER 4
of a preinjection course of antibiotics to povidone– iodine does not reduce the conjunctival bacterial counts any more than the immediate preinjection application of povidone–iodine alone. As a result, they discouraged the administration of antibiotics on the days preceding intravitreal injections. There is additional evidence suggesting that antibiotics given immediately preceding injections provide no additional benefit in reducing rates of endophthalmitis. Topical antibiotics have significantly longer kill times than povidone–iodine. Certain antibiotics such as moxifloxacin that are administered directly before an injection likely do not have a sufficient amount of time to produce the intended bactericidal effect.5,6 Also, when topical fluoroquinolones are administered, the resulting antibiotic levels in the vitreous may be too negligible to effectively combat the bacterial species most commonly responsible for postinjection endophthalmitis.7–9 For this reason, the administration of postinjection antibiotics might not be an effective prophylactic strategy either. There are other abnormal consequences to the routine use of topical antibiotics, including cost, allergic reactions, toxicity, and a potential effect on normal ocular flora allowing for the selection of more virulent and antibiotic-resistant bacterial strains.10–12 In a longitudinal controlled study, Dave et al investigated the role of antibiotic resistance in the context of intravitreal injections. With a cohort of 24 patients, the authors obtained conjunctival cultures from 24 noninjected (control) eyes and 24 contralateral eyes that received a series of consecutive monthly injections and an assigned postinjection antibiotic. The authors isolated Staphyloccocus epidermidis from the cultures and found that, after 4 injections, it was resistant to 3 or more antibiotics in 90% of the treated eyes when compared with 69% of the control eyes. Likewise, they found that, after 4 injections, S. epidermidis was resistant to 5 or more antibiotics in 71% of the treated eyes when compared with 48% of the control eyes. This study demonstrates that repeated antibiotic use is associated with an increase in conjunctival bacterial resistance.11 A related study by Kim and Toma12 also reaches this same conclusion. Both of these studies From the *Retina Center of Minnesota, Minneapolis, Minnesota; and †Midwestern University, Glendale, Arizona. Posters presented in part at the AAO Annual meeting, 2006 Las Vegas, Nevada, 2010 Chicago, Illinois, 2012 Chicago, Illinois. None of the authors have any financial/conflicting interests to disclose. A. R. Bhavsar had full access to all of the data in the study and takes responsibility for the integrity and accuracy of the data analysis. Reprint requests: Abdhish R. Bhavsar, MD, Retina Center, 710 East 24th Street, Suite 304, Minneapolis, MN 55404; e-mail: [email protected]
show that repeated antibiotic administration promotes for the selection of coresistant bacterial strains. Moreover, coresistance can make endophthalmitis much more challenging to treat. Accordingly, the literature offers reasons to both support and oppose the administration of topical antibiotics for intravitreal injections. This study provides additional evidence on this controversial topic. In the following large series of intravitreal injections, the incidence of endophthalmitis and other adverse events are reported for patients who received no topical antibiotics. Methods This study is a retrospective review of 18,839 consecutive intravitreal injections performed by a single retinal surgeon, the lead author, without the use of topical antibiotics. All injections were performed at the Retina Center of Minnesota in a private-practice clinical setting. This study was reviewed and found to be exempt from approval by the Western Institutional Review Board. Retrospective data were collected from August 1, 1997 to October 31, 2012. Every patient who had at least 1 CPT billing code for an intravitreal injection (67,028) was selected for inclusion in the study. Amongst this group, there were no exclusions. Data were compiled using Current Procedural Terminology billing codes and International Classification of Diseases-9 diagnostic codes obtained from the clinic’s practice management computer database software (Cerner PowerWorks PM; Cerner Inc, North Kansas City, MO). The database was searched using the following ICD-9 codes: 363.20 (endophthalmitis); 361.01 to 361.06 (retinal detachment); 364.03 to 364.04 (secondary uveitis); 365.12, 365.60, and 365.62 (secondary glaucoma); and 365.03 (steroid responder glaucoma). Various intravitreal medications were administered. They included bevacizumab (Avastin; Genentech, Inc, South San Francisco, CA), ranibizumab (Lucentis; Genentech, Inc), triamcinolone acetonide (Kenalog40; Bristol–Myers Squibb, Princeton, NJ), triamcinolone acetonide (Triesence; Alcon Laboratories, Inc, Fort Worth, TX), dexamethasone sodium phosphate (APP Pharmaceuticals, LLC, Schaumburg, IL), dexamethasone implant (Ozurdex; Allergan, Inc, Irvine, CA), pegaptanib sodium (Macugen; Eyetech Pharmaceuticals, New York, NY), aflibercept (Eylea; Regeneron Pharmaceuticals, Tarrytown, NY), and ganciclovir (Cytovene; Genentech, Inc). The administered dosages were 1.25 mg (0.05 mL) of bevacizumab, 0.5 mg (0.05 mL) of ranibizumab, 4 mg (0.1 mL) of triamcinolone acetonide, 0.4 mg (0.1 mL) of dexamethasone sodium phosphate, 0.7 mg NOVADUR solid polymer
ELIMINATING ANTIBIOTICS FOR INJECTIONS BHAVSAR AND SANDLER
dexamethasone implant, 0.3 mg (0.09 mL) of pegaptanib sodium, 2 mg (0.05 mL) of aflibercept, and 2 mg (0.1 mL) of ganciclovir. Before each intravitreal injection, the eye was prepared using a standardized protocol, which involved the use of a sterile eyelid speculum and clean nonsterile gloves, and the administering of 0.5% topical proparacaine (Akorn, Inc, Lake Forest, IL), 0.5% topical tetracaine (Bausch & Lomb, Tampa, FL), and several drops of 5% povidone–iodine (Catalent Pharma Solutions, LLC, Woodstock, IL). Surgical masks and eyelid drapes were not used. Special attention was given to avoid contact between the injection needle and eyelashes or lid margins. Surgical masks and eyelid drapes were not used. All injections were performed in the outpatient office setting in normal exam rooms with carpet. In addition, injections were given without cessation of talking, although the injecting physician minimized talking during the injection and routinely counted down “Hold steady, 3,2,1, don’t move. . .1001 and all done” while at arms length away from the site of the injection. An anterior chamber paracentesis with a 30 G needle and tuberculin syringe was typically performed before intravitreal injections of dexamethasone sodium phosphate and triamcinolone acetonide to prevent elevated intraocular pressure. Ranibizumab, dexamethasone sodium phosphate, triamcinolone acetonide (Kenalog-40), triamcinolone acetonide (Triesence), and aflibercept were drawn from a single use vial at the time of the injection, and then injected using a 30 G needle. Dexamethasone implant was injected using preloaded injector devices. Pegaptanib was injected using prefilled single-dose syringes. Bevacizumab was injected using prefilled syringes that were aliquoted by an offsite compounding pharmacy. An additional drop of 5% povidone–iodine was placed over the intended injection site after inserting the eyelid speculum. The injection was then performed at approximately 3.0 mm posterior to the limbus in the inferotemporal quadrant. Immediately after the injection, a drop of 5% povidone–iodine was reapplied to the conjunctival surface. It is essential to note that topical antibiotics were not administered before, during, or after the intravitreal injections.
Results Over a period of 15 years and 3 months, a total of 18,839 intravitreal injections were given to 3,457 patients. There were 1,251 males and 2,206 females in this study. The ages of the patients ranged from 16 years to 108 years. The mean age was 81 years. The mean number of injections per patient was 5.5. Of
785
these injections, 15,479 (82.16%) were bevacizumab, 1,669 (8.86%) were ranibizumab, 1,014 (5.38%) were triamcinolone acetonide (Kenalog-40), 370 (1.96%) were pegaptanib sodium, 148 (0.79%) were aflibercept, 88 (0.47%) were dexamethasone implant, 32 (0.17%) were triamcinolone acetonide (Triesence), 29 (0.15%) were dexamethasone sodium phosphate, and 10 (0.05%) were ganciclovir. The largest number of injections was performed to treat exudative agerelated macular degeneration. Other common reasons for the injections included diabetic macular edema and cystoid macular edema because of retinal vein occlusions. The less common reasons were posterior uveitis and cytomegalovirus retinitis. Of the 18,839 injections, there was 1 case of postinjection endophthalmitis. Therefore, the incidence of endophthalmitis from each intravitreal injection was 0.0053% (1 of 18,839). Secondary uveitis with aqueous and vitreous cells developed in 52 eyes (52 of 18,839 [0.28%]). All cases of uveitis were treated with topical prednisolone acetate 4 times daily, which was tapered over several weeks. No antibiotics were administered, and no procedures were performed on these patients (such as anterior chamber or vitreous taps). All cases of uveitis resolved with prednisolone alone. There were 8 cases of retinal detachment (8 of 18,839 [0.04%]). Steroid responder glaucoma occurred in 74 eyes after an injection of triamcinolone acetonide (Kenalog-40) (74 of 1,014 [7.30%]). Secondary glaucoma developed in 1 eye after bevacizumab (1 of 15,479 [0.0065%]). It may be theoretically possible that some patients could have experienced endophthalmitis and sought treatment for it elsewhere. However, this is very unlikely. In our clinical experience, patients who have major side effects and/or adverse outcomes immediately after our medical interventions are eager to seek prompt treatment with us. Furthermore, their referring physicians are extremely likely to contact us immediately if a major medical issue occurs. The only patient developing postinjection endophthalmitis was a 58-year-old man with diabetes mellitus Type 2 who had received an injection of bevacizumab to his left eye for diabetic macular edema. He returned 2 days postinjection with symptoms of decreased vision, discomfort, and floaters in his left eye. He had a hypopyon, anterior chamber cell and fibrin, vitreous debris, and a visual acuity of hand motion in his left eye. He underwent a pars plana vitreous biopsy with a Visitec vitrectomy unit (Insight Instruments, Inc, Stuart, FL), vitreous and aqueous cultures, and intravitreal injections of 1 mg (0.1 mL) vancomycin, 2.25 mg (0.1 mL) ceftazidime, 1 mg (0.1 mL) clindamycin, and 0.4 mg (0.1 mL) dexamethasone. The vitreous
786
RETINA, THE JOURNAL OF RETINAL AND VITREOUS DISEASES 2015 VOLUME 35 NUMBER 4
tested positive for coagulase-negative Staphylococcus species. In addition, it is important to note that the patient inserted an extended-wear contact lens into the same eye immediately after the bevacizumab injection without knowledge of the lead author. It is conceivable that this could have contributed to the development of endophthalmitis. At 2 months posttreatment, the visual acuity measured 20/100. Approximately 1 year later, the patient underwent a pars plana vitrectomy with internal limiting membrane dissection for an epiretinal membrane in the left eye. At his 1 month postoperative visit, his visual acuity improved to 20/60.
Discussion The incidence of endophthalmitis per injection in this study was 0.0053%. In the literature, the incidence of postinjection endophthalmitis is reported to range from 0.02% to 1.9%.13–16 Without the use of pre-, peri-, or post-injection topical antibiotics, a low incidence of endophthalmitis was achieved, albeit lower than one might expect from the literature. Results of this study are consistent with those from several recent, large prospective randomized multicenter clinical trials by the Diabetic Retinopathy Clinical Research Network (DRCR.net). In the two DRCR.net Laser-Ranibizumab-Triamcinolone clinical trials, there were 3 cases of endophthalmitis of 3,838 injections (3 of 3,838 [0.078%]). In all 3 cases, the patients had received a several day course of topical antibiotics after their injection but none before it. There were no cases of endophthalmitis in the group of patients who did not receive any topical antibiotics.17 In a recent report of the DRCR.net trials, there were a total of 8,027 intravitreal injections. Of these, 4,694 injections were administered with preinjection and/or postinjection antibiotics and 3,333 were administered without any topical antibiotics. In the cohort that received antibiotics, there were 6 cases of endophthalmitis (6 of 4,694 [0.13%]). However, in the cohort that did not receive antibiotics, there was only 1 case of endophthalmitis and a smaller overall incidence of endophthalmitis per injection (1 of 3,333 [0.03%]).18 Therefore, a lower incidence of endophthalmitis was achieved in the DRCR.net trials when topical antibiotics were omitted. Nonetheless, topical antibiotics are commonly used for intravitreal injections. Surveys conducted in 2008 by the American Society of Retina Specialists (ASRS) indicated that 40% of retina specialists applied topical antibiotics before anti-vascular endothelial growth factor intravitreal injections and 86% applied them
after injections. The 2009 ASRS surveys suggested that over 80% of retina specialists administered antibiotics either before or after anti-vascular endothelial growth factor injections. The 2011 ASRS surveys suggested that 27% of retina specialists administered preinjection antibiotics and 62% administered postinjection antibiotics. Although there seems to be a decline in the use of topical antibiotics, a large percentage of retina specialists still administer them. Part of the reason that topical antibiotic use is so prevalent may be due to the fact that many contemporary clinical trials have required it.19–21 As a result, the Food and Drug Administration labeling has included topical antibiotic prophylaxis for a number of medications. DRCR.net, however, has recently “moved to a more minimalist approach for intravitreal injection technique.” DRCR.net first removed the requirement for preinjection topical antibiotics from their protocol, and then later removed periinjection antibiotics from their protocol as well.22 In the DRCR.net Laser-Ranibizumab-Triamcinolone clinical trials, no topical antibiotics were required, but they could be administered on the discretion of the investigator.17 A low incidence of endophthalmitis has been achieved in the DRCR.net clinical trials without topical antibiotics. The results of other studies in the literature also support the findings of this study. Englander et al23 recently reported a low incidence of endophthalmitis (0.029%) in a study involving 10,208 intravitreal injections, where 5 of 9 ophthalmologists elected to administer a course of postinjection antibiotics, but none of the ophthalmologists administered preinjection antibiotics. Moshfeghi et al also reported a low incidence of endophthalmitis (0.02%) in a study involving 60,322 anti-vascular endothelial growth factor intravitreal injections. In this large case series, preinjection antibiotics were omitted, and periinjection and postinjection antibiotics were administered at the physician’s discretion.24 Falavarjani et al described a case series of 5,901 injections, where 68% of the patients received only postinjection antibiotics and 34% received none. There were six cases of postinjection endophthalmitis, all of which occurred in patients who had received antibiotics.25 Comparative case series also suggest a limited usefulness of topical antibiotics. Cheung et al presented a retrospective comparative case series where anti-vascular endothelial growth factor injections were performed on 2 different cohorts of patients who received topical antibiotics (n = 10,629) and a third cohort (n = 5,266) that received no antibiotics. The incidence of endophthalmitis was actually higher in the cohorts that received antibiotics (0.066%) when compared with the cohort that did not (0.038%).26 In
ELIMINATING ANTIBIOTICS FOR INJECTIONS BHAVSAR AND SANDLER
a study by Bhatt et al27 involving 4,767 injections, the incidence of endophthalmitis was similar in the group of patients that received a course of postinjection antibiotics (0.22%) when compared with the group that did not (0.20%). Thus, low rates of postinjection endophthalmitis have been reported in a number of series where antibiotic use has been omitted. Conversely, the results of a study on the incidence of endophthalmitis in the United Kingdom contradict the present findings. Of an estimated total of 186,972 intravitreal injections, Lyall et al reported 47 cases of endophthalmitis. To obtain these data, the authors collected survey information from ophthalmologists involved in those cases as well as for 200 control patients. They concluded that failure to administer topical antibiotics immediately after intravitreal injections is a significant risk factor for endophthalmitis.28 However, the survey that they used was not validated, the control group was inappropriately selected, and the use or omission of povidone–iodine at the time of the injection was not included in the study. For these reasons, the conclusions of their study may be invalid.29 There are a few other studies that have also reported a very low incidence of endophthalmitis. Two recent studies found no cases of endophthalmitis,30,31 but both were conducted in an operating room setting and had highly involved protocols. Shimada et al conducted 15,144 intravitreal injections where preinjection and postinjection antibiotics were administered. The surgeons wore surgical masks, placed facial drapes on patients, and irrigated the eye with 0.25% povidone–iodine.30 It is difficult to assess the role of antibiotics in their study because of these extraneous factors. Likewise, Brynskov et al31 performed a series of 20,293 injections in an operating room with positive pressure ventilation. The doctors wore surgical masks, placed facial drapes, and irrigated with 5% povidone– iodine. A drop of an antibiotic was also administered immediately after all injections, and a course of antibiotics was given for 13,293 of the injections. This operating room setting and ostensibly expensive protocol would not be practical or cost-effective in many regions of the world. Limitations of this study include its retrospective nature and noncomparative design, as well as the lack of randomized prospective data. In addition, because the series involves only a single surgeon, it is not known whether these results could be reproduced by others. However, the use of data from a single surgeon with a consecutive series of injections provides a consistent protocol. One reason this study is unique is because the lead author has omitted topical antibiotics for intravitreal injections since 1997. Few retina specialists were
787
withholding antibiotics at that time. Results from the initial 1,000 injections of this study were presented at Association for Research in Vision and Ophthalmology in 2008 when no cases of postinjection endophthalmitis had occurred. This presentation and several other subsequent presentations by the lead author initiated discussion and prompted others to publish about the efficacy of antibiotics in preventing postinjection endophthalmitis. The findings of DRCR.net studies have also significantly contributed to awareness of this concern within the retina community. Although there are now many publications on this subject, this study is one of only a few in which patients received no antibiotics at all.17,18,25,26 To our knowledge, this is the first large-scale consecutive case series to demonstrate that intravitreal injections, in the absence of all topical antibiotics, can be administered safely with a very low incidence of endophthalmitis in an outpatient office setting. Of note, this was accomplished with topical povidone–iodine, a sterile eyelid speculum, and nonsterile gloves. Topical antibiotics are not necessary for achieving a low incidence of endophthalmitis after intravitreal injections. Key words: antibiotics, bevacizumab, complications, eliminating, endophthalmitis, intravitreal injection, omitting, ranibizumab, triamcinolone, vascular endothelial growth factor. References 1. Speaker MG, Menikoff JA. Prophylaxis of endophthalmitis with topical povidone-iodine. Ophthalmology 1991;98:1769– 1775. 2. Isenberg SJ, Apt L, Yoshimori R, Khwarg S. Chemical preparation of the eye in ophthalmic surgery. IV. Comparison of povidone-iodine on the conjunctiva with a prophylactic antibiotic. Arch Ophthalmol 1985;103:1340–1342. 3. Apt L, Isenberg SJ, Yoshimori R, et al. The effect of povidoneiodine solution applied at the conclusion of ophthalmic surgery. Am J Ophthalmol 1995;119:701–705. 4. Moss JM, Sanislo SR, Ta CN. A prospective randomized evaluation of topical gatifloxacin on conjunctival flora in patients undergoing intravitreal injections. Ophthalmology 2009;116: 1498–1501. 5. Hyon JY, Eser I, O’Brien TP. Kill rates of preserved and preservative-free topical 8-methoxy fluoroquinolones against various strains of Staphylococcus. J Cataract Refract Surg 2009;35:1609–1613. 6. Callegan MC, Novosad BD, Ramadan RT, et al. Rate of bacterial eradication by ophthalmic solutions of fourth-generation fluoroquinolones. Adv Ther 2009;26:447–454. 7. Puustjarvi T, Terasvirta M, Nurmenniemi P, et al. Penetration of topically applied levofloxacin 0.5% and ofloxacin 0.3% into the vitreous of the non-inflamed human eye. Graefes Arch Clin Exp Ophthalmol 2006;244:1633–1637. 8. Sakamoto H, Sakamoto M, Hata Y, et al. Aqueous and vitreous penetration of levofloxacin after topical and/or oral administration. Eur J Ophthalmol 2007;17:372–376.
788
RETINA, THE JOURNAL OF RETINAL AND VITREOUS DISEASES 2015 VOLUME 35 NUMBER 4
9. Yalvac IS, Basci NE, Bozkurt A, Duman S. Penetration of topically applied ciprofloxacin and ofloxacin into the aqueous humor and vitreous. J Cataract Refract Surg 2003;29: 487–491. 10. Yin VT, Weisbrod DJ, Eng KT, et al. Antibiotic resistance of ocular surface flora with repeated use of a topical antibiotic after intravitreal injection. JAMA Ophthalmol 2013;131: 456–461. 11. Dave SD, Toma HS, Kim SJ. Ophthalmic antibiotic use and multidrug-resistant Staphylococcus epidermidis: a controlled, longitudinal study. Ophthalmology 2011;118:2035–2040. 12. Kim SJ, Toma HS. Antimicrobial resistance and ophthalmic antibiotics: 1-year results of a longitudinal controlled study of patients undergoing intravitreal injections. Arch Ophthalmol 2011;129:1180–1188. 13. Mason JO III, White MF, Feist RM, et al. Incidence of acute onset endophthalmitis following intravitreal bevacizumab (Avastin) injection. Retina 2008;28:564–567. 14. Pilli S, Kotsolis A, Spaide RF, et al. Endophthalmitis associated with intravitreal anti-vascular endothelial growth factor therapy injections in an office setting. Am J Ophthalmol 2008;145:879–882. 15. Scott IU, Flynn HW Jr. Reducing the risk of endophthalmitis following intravitreal injections. Retina 2007;27:10–12. 16. Meyer CH, Mennel S, Eter N. Incidence of endophthalmitis after intravitreal Avastin injection with and without postoperative topical antibiotic application [in German]. Ophthalmologe 2007;104:952–957. 17. Bhavsar AR, Googe JM Jr, Stockdale CR, et al. Risk of endophthalmitis after intravitreal drug injection when topical antibiotics are not required: the diabetic retinopathy clinical research network laser-ranibizumab-triamcinolone clinical trials. Arch Ophthalmol 2009;127:1581–1583. 18. Bhavsar AR, Stockdale CR, Ferris FL III, et al. Update on risk of endophthalmitis after intravitreal drug injections and potential impact of elimination of topical antibiotics. Arch Ophthalmol 2012;130:809–810. 19. Gragoudas ES, Adamis AP, Cunningham ET Jr, et al. Pegaptanib for neovascular age-related macular degeneration. N Engl J Med 2004;351:2805–2816.
20. Rosenfeld PJ, Brown DM, Heier JS, et al. Ranibizumab for neovascular age-related macular degeneration. N Engl J Med 2006;355:1419–1431. 21. Kaiser PK, Brown DM, Zhang K, et al. Ranibizumab for predominantly classic neovascular age-related macular degeneration: subgroup analysis of first-year ANCHOR results. Am J Ophthalmol 2007;144:850–857. 22. Bhavsar AR. Avoiding endophthalmitis following intravitreal injections. Retinal Physician 2013. Available at: http://www. retinalphysician.com/articleviewer.aspx?articleID=108234. Accessed April 3, 2014. 23. Englander M, Chen TC, Paschalis EI, et al. Intravitreal injections at the Massachusetts Eye and Ear Infirmary: analysis of treatment indications and postinjection endophthalmitis rates. Br J Ophthalmol 2013;97:460–465. 24. Moshfeghi AA, Rosenfeld PJ, Flynn HW Jr, et al. Endophthalmitis after intravitreal vascular [corrected] endothelial growth factor antagonists: a six-year experience at a university referral center. Retina 2011;31:662–668. 25. Falavarjani KG, Modarres M, Hashemi M, et al. Incidence of acute endophthalmitis after intravitreal bevacizumab injection in a single clinical center. Retina 2013;33:971–974. 26. Cheung CS, Wong AW, Lui A, et al. Incidence of endophthalmitis and use of antibiotic prophylaxis after intravitreal injections. Ophthalmology 2012;119:1609–1614. 27. Bhatt SS, Stepien KE, Joshi K. Prophylactic antibiotic use after intravitreal injection: effect on endophthalmitis rate. Retina 2011;31:2032–2036. 28. Lyall DA, Tey A, Foot B, et al. Post-intravitreal anti-VEGF endophthalmitis in the United Kingdom: incidence, features, risk factors, and outcomes. Eye (Lond) 2012;26:1517–1526. 29. Bhavsar AR, Sandler DR, Gentile RC. Intravitreal injections, antibiotics and endophthalmitis. Eye (Lond) 2013;27:1426–1427. 30. Shimada H, Hattori T, Mori R, et al. Minimizing the endophthalmitis rate following intravitreal injections using 0.25% povidone-iodine irrigation and surgical mask. Graefes Arch Clin Exp Ophthalmol 2013;251:1885–1890. 31. Brynskov T, Kemp H, Sørensen TL. No cases of endophthalmitis after 20,293 intravitreal injections in an operating room setting. Retina 2014;34:951–957.
T
24 hours after intraocular surgery. Conversely, polymyxin B sulfate–neomycin sulfate–gramicidin, a broad-spectrum topical antibiotic, has a significantly shorter duration of bactericidal activity.3 Thus, there is evidence that povidone–iodine can more effectively eradicate pathogens than topical antibiotics. Given this, it is logical to inquire whether topical antibiotics are a necessary addition to povidone–iodine. Isenberg et al2 found that there may be a synergistic effect of topical antibiotics and povidone–iodine in reducing the culture positive rate of the preoperative conjunctival surface in patients undergoing intraocular surgery. Moss et al,4 however, reached a contrasting conclusion in a related study on patients undergoing intravitreal injections. They concluded that the addition
here is a great deal of controversy about the efficacy of topical antibiotics in the prevention of endophthalmitis after intravitreal injections. Topical antibiotics are routinely administered as part of the intravitreal injection procedure for endophthalmitis prophylaxis. However, to our knowledge, there has never been a study proving that any antibiotic can effectively reduce the risk of postinjection endophthalmitis. Nevertheless, there is evidence that topical povidone–iodine alone reduces the risk of endophthalmitis after intraocular surgery.1 According to Isenberg et al,2 one application of povidone–iodine has a bactericidal effect equivalent to a 3-day course of topical antibiotics. In addition, the antimicrobial effect of a drop of povidone–iodine 5% has been shown to last at least 783
784
RETINA, THE JOURNAL OF RETINAL AND VITREOUS DISEASES 2015 VOLUME 35 NUMBER 4
of a preinjection course of antibiotics to povidone– iodine does not reduce the conjunctival bacterial counts any more than the immediate preinjection application of povidone–iodine alone. As a result, they discouraged the administration of antibiotics on the days preceding intravitreal injections. There is additional evidence suggesting that antibiotics given immediately preceding injections provide no additional benefit in reducing rates of endophthalmitis. Topical antibiotics have significantly longer kill times than povidone–iodine. Certain antibiotics such as moxifloxacin that are administered directly before an injection likely do not have a sufficient amount of time to produce the intended bactericidal effect.5,6 Also, when topical fluoroquinolones are administered, the resulting antibiotic levels in the vitreous may be too negligible to effectively combat the bacterial species most commonly responsible for postinjection endophthalmitis.7–9 For this reason, the administration of postinjection antibiotics might not be an effective prophylactic strategy either. There are other abnormal consequences to the routine use of topical antibiotics, including cost, allergic reactions, toxicity, and a potential effect on normal ocular flora allowing for the selection of more virulent and antibiotic-resistant bacterial strains.10–12 In a longitudinal controlled study, Dave et al investigated the role of antibiotic resistance in the context of intravitreal injections. With a cohort of 24 patients, the authors obtained conjunctival cultures from 24 noninjected (control) eyes and 24 contralateral eyes that received a series of consecutive monthly injections and an assigned postinjection antibiotic. The authors isolated Staphyloccocus epidermidis from the cultures and found that, after 4 injections, it was resistant to 3 or more antibiotics in 90% of the treated eyes when compared with 69% of the control eyes. Likewise, they found that, after 4 injections, S. epidermidis was resistant to 5 or more antibiotics in 71% of the treated eyes when compared with 48% of the control eyes. This study demonstrates that repeated antibiotic use is associated with an increase in conjunctival bacterial resistance.11 A related study by Kim and Toma12 also reaches this same conclusion. Both of these studies From the *Retina Center of Minnesota, Minneapolis, Minnesota; and †Midwestern University, Glendale, Arizona. Posters presented in part at the AAO Annual meeting, 2006 Las Vegas, Nevada, 2010 Chicago, Illinois, 2012 Chicago, Illinois. None of the authors have any financial/conflicting interests to disclose. A. R. Bhavsar had full access to all of the data in the study and takes responsibility for the integrity and accuracy of the data analysis. Reprint requests: Abdhish R. Bhavsar, MD, Retina Center, 710 East 24th Street, Suite 304, Minneapolis, MN 55404; e-mail: [email protected]
show that repeated antibiotic administration promotes for the selection of coresistant bacterial strains. Moreover, coresistance can make endophthalmitis much more challenging to treat. Accordingly, the literature offers reasons to both support and oppose the administration of topical antibiotics for intravitreal injections. This study provides additional evidence on this controversial topic. In the following large series of intravitreal injections, the incidence of endophthalmitis and other adverse events are reported for patients who received no topical antibiotics. Methods This study is a retrospective review of 18,839 consecutive intravitreal injections performed by a single retinal surgeon, the lead author, without the use of topical antibiotics. All injections were performed at the Retina Center of Minnesota in a private-practice clinical setting. This study was reviewed and found to be exempt from approval by the Western Institutional Review Board. Retrospective data were collected from August 1, 1997 to October 31, 2012. Every patient who had at least 1 CPT billing code for an intravitreal injection (67,028) was selected for inclusion in the study. Amongst this group, there were no exclusions. Data were compiled using Current Procedural Terminology billing codes and International Classification of Diseases-9 diagnostic codes obtained from the clinic’s practice management computer database software (Cerner PowerWorks PM; Cerner Inc, North Kansas City, MO). The database was searched using the following ICD-9 codes: 363.20 (endophthalmitis); 361.01 to 361.06 (retinal detachment); 364.03 to 364.04 (secondary uveitis); 365.12, 365.60, and 365.62 (secondary glaucoma); and 365.03 (steroid responder glaucoma). Various intravitreal medications were administered. They included bevacizumab (Avastin; Genentech, Inc, South San Francisco, CA), ranibizumab (Lucentis; Genentech, Inc), triamcinolone acetonide (Kenalog40; Bristol–Myers Squibb, Princeton, NJ), triamcinolone acetonide (Triesence; Alcon Laboratories, Inc, Fort Worth, TX), dexamethasone sodium phosphate (APP Pharmaceuticals, LLC, Schaumburg, IL), dexamethasone implant (Ozurdex; Allergan, Inc, Irvine, CA), pegaptanib sodium (Macugen; Eyetech Pharmaceuticals, New York, NY), aflibercept (Eylea; Regeneron Pharmaceuticals, Tarrytown, NY), and ganciclovir (Cytovene; Genentech, Inc). The administered dosages were 1.25 mg (0.05 mL) of bevacizumab, 0.5 mg (0.05 mL) of ranibizumab, 4 mg (0.1 mL) of triamcinolone acetonide, 0.4 mg (0.1 mL) of dexamethasone sodium phosphate, 0.7 mg NOVADUR solid polymer
ELIMINATING ANTIBIOTICS FOR INJECTIONS BHAVSAR AND SANDLER
dexamethasone implant, 0.3 mg (0.09 mL) of pegaptanib sodium, 2 mg (0.05 mL) of aflibercept, and 2 mg (0.1 mL) of ganciclovir. Before each intravitreal injection, the eye was prepared using a standardized protocol, which involved the use of a sterile eyelid speculum and clean nonsterile gloves, and the administering of 0.5% topical proparacaine (Akorn, Inc, Lake Forest, IL), 0.5% topical tetracaine (Bausch & Lomb, Tampa, FL), and several drops of 5% povidone–iodine (Catalent Pharma Solutions, LLC, Woodstock, IL). Surgical masks and eyelid drapes were not used. Special attention was given to avoid contact between the injection needle and eyelashes or lid margins. Surgical masks and eyelid drapes were not used. All injections were performed in the outpatient office setting in normal exam rooms with carpet. In addition, injections were given without cessation of talking, although the injecting physician minimized talking during the injection and routinely counted down “Hold steady, 3,2,1, don’t move. . .1001 and all done” while at arms length away from the site of the injection. An anterior chamber paracentesis with a 30 G needle and tuberculin syringe was typically performed before intravitreal injections of dexamethasone sodium phosphate and triamcinolone acetonide to prevent elevated intraocular pressure. Ranibizumab, dexamethasone sodium phosphate, triamcinolone acetonide (Kenalog-40), triamcinolone acetonide (Triesence), and aflibercept were drawn from a single use vial at the time of the injection, and then injected using a 30 G needle. Dexamethasone implant was injected using preloaded injector devices. Pegaptanib was injected using prefilled single-dose syringes. Bevacizumab was injected using prefilled syringes that were aliquoted by an offsite compounding pharmacy. An additional drop of 5% povidone–iodine was placed over the intended injection site after inserting the eyelid speculum. The injection was then performed at approximately 3.0 mm posterior to the limbus in the inferotemporal quadrant. Immediately after the injection, a drop of 5% povidone–iodine was reapplied to the conjunctival surface. It is essential to note that topical antibiotics were not administered before, during, or after the intravitreal injections.
Results Over a period of 15 years and 3 months, a total of 18,839 intravitreal injections were given to 3,457 patients. There were 1,251 males and 2,206 females in this study. The ages of the patients ranged from 16 years to 108 years. The mean age was 81 years. The mean number of injections per patient was 5.5. Of
785
these injections, 15,479 (82.16%) were bevacizumab, 1,669 (8.86%) were ranibizumab, 1,014 (5.38%) were triamcinolone acetonide (Kenalog-40), 370 (1.96%) were pegaptanib sodium, 148 (0.79%) were aflibercept, 88 (0.47%) were dexamethasone implant, 32 (0.17%) were triamcinolone acetonide (Triesence), 29 (0.15%) were dexamethasone sodium phosphate, and 10 (0.05%) were ganciclovir. The largest number of injections was performed to treat exudative agerelated macular degeneration. Other common reasons for the injections included diabetic macular edema and cystoid macular edema because of retinal vein occlusions. The less common reasons were posterior uveitis and cytomegalovirus retinitis. Of the 18,839 injections, there was 1 case of postinjection endophthalmitis. Therefore, the incidence of endophthalmitis from each intravitreal injection was 0.0053% (1 of 18,839). Secondary uveitis with aqueous and vitreous cells developed in 52 eyes (52 of 18,839 [0.28%]). All cases of uveitis were treated with topical prednisolone acetate 4 times daily, which was tapered over several weeks. No antibiotics were administered, and no procedures were performed on these patients (such as anterior chamber or vitreous taps). All cases of uveitis resolved with prednisolone alone. There were 8 cases of retinal detachment (8 of 18,839 [0.04%]). Steroid responder glaucoma occurred in 74 eyes after an injection of triamcinolone acetonide (Kenalog-40) (74 of 1,014 [7.30%]). Secondary glaucoma developed in 1 eye after bevacizumab (1 of 15,479 [0.0065%]). It may be theoretically possible that some patients could have experienced endophthalmitis and sought treatment for it elsewhere. However, this is very unlikely. In our clinical experience, patients who have major side effects and/or adverse outcomes immediately after our medical interventions are eager to seek prompt treatment with us. Furthermore, their referring physicians are extremely likely to contact us immediately if a major medical issue occurs. The only patient developing postinjection endophthalmitis was a 58-year-old man with diabetes mellitus Type 2 who had received an injection of bevacizumab to his left eye for diabetic macular edema. He returned 2 days postinjection with symptoms of decreased vision, discomfort, and floaters in his left eye. He had a hypopyon, anterior chamber cell and fibrin, vitreous debris, and a visual acuity of hand motion in his left eye. He underwent a pars plana vitreous biopsy with a Visitec vitrectomy unit (Insight Instruments, Inc, Stuart, FL), vitreous and aqueous cultures, and intravitreal injections of 1 mg (0.1 mL) vancomycin, 2.25 mg (0.1 mL) ceftazidime, 1 mg (0.1 mL) clindamycin, and 0.4 mg (0.1 mL) dexamethasone. The vitreous
786
RETINA, THE JOURNAL OF RETINAL AND VITREOUS DISEASES 2015 VOLUME 35 NUMBER 4
tested positive for coagulase-negative Staphylococcus species. In addition, it is important to note that the patient inserted an extended-wear contact lens into the same eye immediately after the bevacizumab injection without knowledge of the lead author. It is conceivable that this could have contributed to the development of endophthalmitis. At 2 months posttreatment, the visual acuity measured 20/100. Approximately 1 year later, the patient underwent a pars plana vitrectomy with internal limiting membrane dissection for an epiretinal membrane in the left eye. At his 1 month postoperative visit, his visual acuity improved to 20/60.
Discussion The incidence of endophthalmitis per injection in this study was 0.0053%. In the literature, the incidence of postinjection endophthalmitis is reported to range from 0.02% to 1.9%.13–16 Without the use of pre-, peri-, or post-injection topical antibiotics, a low incidence of endophthalmitis was achieved, albeit lower than one might expect from the literature. Results of this study are consistent with those from several recent, large prospective randomized multicenter clinical trials by the Diabetic Retinopathy Clinical Research Network (DRCR.net). In the two DRCR.net Laser-Ranibizumab-Triamcinolone clinical trials, there were 3 cases of endophthalmitis of 3,838 injections (3 of 3,838 [0.078%]). In all 3 cases, the patients had received a several day course of topical antibiotics after their injection but none before it. There were no cases of endophthalmitis in the group of patients who did not receive any topical antibiotics.17 In a recent report of the DRCR.net trials, there were a total of 8,027 intravitreal injections. Of these, 4,694 injections were administered with preinjection and/or postinjection antibiotics and 3,333 were administered without any topical antibiotics. In the cohort that received antibiotics, there were 6 cases of endophthalmitis (6 of 4,694 [0.13%]). However, in the cohort that did not receive antibiotics, there was only 1 case of endophthalmitis and a smaller overall incidence of endophthalmitis per injection (1 of 3,333 [0.03%]).18 Therefore, a lower incidence of endophthalmitis was achieved in the DRCR.net trials when topical antibiotics were omitted. Nonetheless, topical antibiotics are commonly used for intravitreal injections. Surveys conducted in 2008 by the American Society of Retina Specialists (ASRS) indicated that 40% of retina specialists applied topical antibiotics before anti-vascular endothelial growth factor intravitreal injections and 86% applied them
after injections. The 2009 ASRS surveys suggested that over 80% of retina specialists administered antibiotics either before or after anti-vascular endothelial growth factor injections. The 2011 ASRS surveys suggested that 27% of retina specialists administered preinjection antibiotics and 62% administered postinjection antibiotics. Although there seems to be a decline in the use of topical antibiotics, a large percentage of retina specialists still administer them. Part of the reason that topical antibiotic use is so prevalent may be due to the fact that many contemporary clinical trials have required it.19–21 As a result, the Food and Drug Administration labeling has included topical antibiotic prophylaxis for a number of medications. DRCR.net, however, has recently “moved to a more minimalist approach for intravitreal injection technique.” DRCR.net first removed the requirement for preinjection topical antibiotics from their protocol, and then later removed periinjection antibiotics from their protocol as well.22 In the DRCR.net Laser-Ranibizumab-Triamcinolone clinical trials, no topical antibiotics were required, but they could be administered on the discretion of the investigator.17 A low incidence of endophthalmitis has been achieved in the DRCR.net clinical trials without topical antibiotics. The results of other studies in the literature also support the findings of this study. Englander et al23 recently reported a low incidence of endophthalmitis (0.029%) in a study involving 10,208 intravitreal injections, where 5 of 9 ophthalmologists elected to administer a course of postinjection antibiotics, but none of the ophthalmologists administered preinjection antibiotics. Moshfeghi et al also reported a low incidence of endophthalmitis (0.02%) in a study involving 60,322 anti-vascular endothelial growth factor intravitreal injections. In this large case series, preinjection antibiotics were omitted, and periinjection and postinjection antibiotics were administered at the physician’s discretion.24 Falavarjani et al described a case series of 5,901 injections, where 68% of the patients received only postinjection antibiotics and 34% received none. There were six cases of postinjection endophthalmitis, all of which occurred in patients who had received antibiotics.25 Comparative case series also suggest a limited usefulness of topical antibiotics. Cheung et al presented a retrospective comparative case series where anti-vascular endothelial growth factor injections were performed on 2 different cohorts of patients who received topical antibiotics (n = 10,629) and a third cohort (n = 5,266) that received no antibiotics. The incidence of endophthalmitis was actually higher in the cohorts that received antibiotics (0.066%) when compared with the cohort that did not (0.038%).26 In
ELIMINATING ANTIBIOTICS FOR INJECTIONS BHAVSAR AND SANDLER
a study by Bhatt et al27 involving 4,767 injections, the incidence of endophthalmitis was similar in the group of patients that received a course of postinjection antibiotics (0.22%) when compared with the group that did not (0.20%). Thus, low rates of postinjection endophthalmitis have been reported in a number of series where antibiotic use has been omitted. Conversely, the results of a study on the incidence of endophthalmitis in the United Kingdom contradict the present findings. Of an estimated total of 186,972 intravitreal injections, Lyall et al reported 47 cases of endophthalmitis. To obtain these data, the authors collected survey information from ophthalmologists involved in those cases as well as for 200 control patients. They concluded that failure to administer topical antibiotics immediately after intravitreal injections is a significant risk factor for endophthalmitis.28 However, the survey that they used was not validated, the control group was inappropriately selected, and the use or omission of povidone–iodine at the time of the injection was not included in the study. For these reasons, the conclusions of their study may be invalid.29 There are a few other studies that have also reported a very low incidence of endophthalmitis. Two recent studies found no cases of endophthalmitis,30,31 but both were conducted in an operating room setting and had highly involved protocols. Shimada et al conducted 15,144 intravitreal injections where preinjection and postinjection antibiotics were administered. The surgeons wore surgical masks, placed facial drapes on patients, and irrigated the eye with 0.25% povidone–iodine.30 It is difficult to assess the role of antibiotics in their study because of these extraneous factors. Likewise, Brynskov et al31 performed a series of 20,293 injections in an operating room with positive pressure ventilation. The doctors wore surgical masks, placed facial drapes, and irrigated with 5% povidone– iodine. A drop of an antibiotic was also administered immediately after all injections, and a course of antibiotics was given for 13,293 of the injections. This operating room setting and ostensibly expensive protocol would not be practical or cost-effective in many regions of the world. Limitations of this study include its retrospective nature and noncomparative design, as well as the lack of randomized prospective data. In addition, because the series involves only a single surgeon, it is not known whether these results could be reproduced by others. However, the use of data from a single surgeon with a consecutive series of injections provides a consistent protocol. One reason this study is unique is because the lead author has omitted topical antibiotics for intravitreal injections since 1997. Few retina specialists were
787
withholding antibiotics at that time. Results from the initial 1,000 injections of this study were presented at Association for Research in Vision and Ophthalmology in 2008 when no cases of postinjection endophthalmitis had occurred. This presentation and several other subsequent presentations by the lead author initiated discussion and prompted others to publish about the efficacy of antibiotics in preventing postinjection endophthalmitis. The findings of DRCR.net studies have also significantly contributed to awareness of this concern within the retina community. Although there are now many publications on this subject, this study is one of only a few in which patients received no antibiotics at all.17,18,25,26 To our knowledge, this is the first large-scale consecutive case series to demonstrate that intravitreal injections, in the absence of all topical antibiotics, can be administered safely with a very low incidence of endophthalmitis in an outpatient office setting. Of note, this was accomplished with topical povidone–iodine, a sterile eyelid speculum, and nonsterile gloves. Topical antibiotics are not necessary for achieving a low incidence of endophthalmitis after intravitreal injections. Key words: antibiotics, bevacizumab, complications, eliminating, endophthalmitis, intravitreal injection, omitting, ranibizumab, triamcinolone, vascular endothelial growth factor. References 1. Speaker MG, Menikoff JA. Prophylaxis of endophthalmitis with topical povidone-iodine. Ophthalmology 1991;98:1769– 1775. 2. Isenberg SJ, Apt L, Yoshimori R, Khwarg S. Chemical preparation of the eye in ophthalmic surgery. IV. Comparison of povidone-iodine on the conjunctiva with a prophylactic antibiotic. Arch Ophthalmol 1985;103:1340–1342. 3. Apt L, Isenberg SJ, Yoshimori R, et al. The effect of povidoneiodine solution applied at the conclusion of ophthalmic surgery. Am J Ophthalmol 1995;119:701–705. 4. Moss JM, Sanislo SR, Ta CN. A prospective randomized evaluation of topical gatifloxacin on conjunctival flora in patients undergoing intravitreal injections. Ophthalmology 2009;116: 1498–1501. 5. Hyon JY, Eser I, O’Brien TP. Kill rates of preserved and preservative-free topical 8-methoxy fluoroquinolones against various strains of Staphylococcus. J Cataract Refract Surg 2009;35:1609–1613. 6. Callegan MC, Novosad BD, Ramadan RT, et al. Rate of bacterial eradication by ophthalmic solutions of fourth-generation fluoroquinolones. Adv Ther 2009;26:447–454. 7. Puustjarvi T, Terasvirta M, Nurmenniemi P, et al. Penetration of topically applied levofloxacin 0.5% and ofloxacin 0.3% into the vitreous of the non-inflamed human eye. Graefes Arch Clin Exp Ophthalmol 2006;244:1633–1637. 8. Sakamoto H, Sakamoto M, Hata Y, et al. Aqueous and vitreous penetration of levofloxacin after topical and/or oral administration. Eur J Ophthalmol 2007;17:372–376.
788
RETINA, THE JOURNAL OF RETINAL AND VITREOUS DISEASES 2015 VOLUME 35 NUMBER 4
9. Yalvac IS, Basci NE, Bozkurt A, Duman S. Penetration of topically applied ciprofloxacin and ofloxacin into the aqueous humor and vitreous. J Cataract Refract Surg 2003;29: 487–491. 10. Yin VT, Weisbrod DJ, Eng KT, et al. Antibiotic resistance of ocular surface flora with repeated use of a topical antibiotic after intravitreal injection. JAMA Ophthalmol 2013;131: 456–461. 11. Dave SD, Toma HS, Kim SJ. Ophthalmic antibiotic use and multidrug-resistant Staphylococcus epidermidis: a controlled, longitudinal study. Ophthalmology 2011;118:2035–2040. 12. Kim SJ, Toma HS. Antimicrobial resistance and ophthalmic antibiotics: 1-year results of a longitudinal controlled study of patients undergoing intravitreal injections. Arch Ophthalmol 2011;129:1180–1188. 13. Mason JO III, White MF, Feist RM, et al. Incidence of acute onset endophthalmitis following intravitreal bevacizumab (Avastin) injection. Retina 2008;28:564–567. 14. Pilli S, Kotsolis A, Spaide RF, et al. Endophthalmitis associated with intravitreal anti-vascular endothelial growth factor therapy injections in an office setting. Am J Ophthalmol 2008;145:879–882. 15. Scott IU, Flynn HW Jr. Reducing the risk of endophthalmitis following intravitreal injections. Retina 2007;27:10–12. 16. Meyer CH, Mennel S, Eter N. Incidence of endophthalmitis after intravitreal Avastin injection with and without postoperative topical antibiotic application [in German]. Ophthalmologe 2007;104:952–957. 17. Bhavsar AR, Googe JM Jr, Stockdale CR, et al. Risk of endophthalmitis after intravitreal drug injection when topical antibiotics are not required: the diabetic retinopathy clinical research network laser-ranibizumab-triamcinolone clinical trials. Arch Ophthalmol 2009;127:1581–1583. 18. Bhavsar AR, Stockdale CR, Ferris FL III, et al. Update on risk of endophthalmitis after intravitreal drug injections and potential impact of elimination of topical antibiotics. Arch Ophthalmol 2012;130:809–810. 19. Gragoudas ES, Adamis AP, Cunningham ET Jr, et al. Pegaptanib for neovascular age-related macular degeneration. N Engl J Med 2004;351:2805–2816.
20. Rosenfeld PJ, Brown DM, Heier JS, et al. Ranibizumab for neovascular age-related macular degeneration. N Engl J Med 2006;355:1419–1431. 21. Kaiser PK, Brown DM, Zhang K, et al. Ranibizumab for predominantly classic neovascular age-related macular degeneration: subgroup analysis of first-year ANCHOR results. Am J Ophthalmol 2007;144:850–857. 22. Bhavsar AR. Avoiding endophthalmitis following intravitreal injections. Retinal Physician 2013. Available at: http://www. retinalphysician.com/articleviewer.aspx?articleID=108234. Accessed April 3, 2014. 23. Englander M, Chen TC, Paschalis EI, et al. Intravitreal injections at the Massachusetts Eye and Ear Infirmary: analysis of treatment indications and postinjection endophthalmitis rates. Br J Ophthalmol 2013;97:460–465. 24. Moshfeghi AA, Rosenfeld PJ, Flynn HW Jr, et al. Endophthalmitis after intravitreal vascular [corrected] endothelial growth factor antagonists: a six-year experience at a university referral center. Retina 2011;31:662–668. 25. Falavarjani KG, Modarres M, Hashemi M, et al. Incidence of acute endophthalmitis after intravitreal bevacizumab injection in a single clinical center. Retina 2013;33:971–974. 26. Cheung CS, Wong AW, Lui A, et al. Incidence of endophthalmitis and use of antibiotic prophylaxis after intravitreal injections. Ophthalmology 2012;119:1609–1614. 27. Bhatt SS, Stepien KE, Joshi K. Prophylactic antibiotic use after intravitreal injection: effect on endophthalmitis rate. Retina 2011;31:2032–2036. 28. Lyall DA, Tey A, Foot B, et al. Post-intravitreal anti-VEGF endophthalmitis in the United Kingdom: incidence, features, risk factors, and outcomes. Eye (Lond) 2012;26:1517–1526. 29. Bhavsar AR, Sandler DR, Gentile RC. Intravitreal injections, antibiotics and endophthalmitis. Eye (Lond) 2013;27:1426–1427. 30. Shimada H, Hattori T, Mori R, et al. Minimizing the endophthalmitis rate following intravitreal injections using 0.25% povidone-iodine irrigation and surgical mask. Graefes Arch Clin Exp Ophthalmol 2013;251:1885–1890. 31. Brynskov T, Kemp H, Sørensen TL. No cases of endophthalmitis after 20,293 intravitreal injections in an operating room setting. Retina 2014;34:951–957.
![[Intravitreal injections: Satisfaction and preference].](/assets/img/hold.png)
