Article Date: 9/1/2009

Avoidance of Endophthalmitis After Intravitreal Injections

Avoidance of Endophthalmitis After Intravitreal Injections


Intravitreal administration of therapeutic agents has come to play an increasingly important role in the treatment of ocular disease. Further, in the anti-VEGF era, patients are receiving injections more frequently and over a longer period of time. In the medical profession, we are acutely aware that “an ounce of prevention equals a pound of cure.” Though few studies have focused directly on the effect of injection procedure technique on complication rates, this subject warrants attention, given the potentially devastating consequences of poor technique. In this review, we provide information regarding the incidence and pathogenesis of infectious endophthalmitis following intravitreal injection and then consider guidelines for prophylaxis as supported by published data.


Intravitreal injection as a treatment for retinal disease has come to play an increasingly important role over the past decades. Management of retinal detachment was the first described indication for intravitreal injection, utilizing intraocular tamponade, first with air (1911), then with silicone oil (1962).1 Studies evaluating intravitreal penicillin and sulfanilamide for endophthalmitis were published in the 1940s. Though efficacy was poor, these studies helped further establish intraocular injection as a safe and feasible treatment. Intravitreal ganciclovir sodium was described in 19872 as a treatment for CMV retinitis in the setting of AIDS. Almost a decade later, the FDA approved fomivirsen sodium for intravitreal injection for the same indication.

Other treatment applications for intravitreal injection include 5-fluorouracil for PVR, dexamethasone for diabetic retinopathy, tPA for submacular hemorrhage, and methotrexate for intraocular lymphoma. The advent of anti-VEGF therapies for neovascular age-related macular degeneration (AMD) catalyzed a new era characterized by an explosion in the use of intravitreal injections. The use of intravitreal injections for the treatment of ocular disease promises to continue to increase as new agents are introduced and new applications for current therapies are described.

Nathan C. Steinle, MD, Aimee V Chappelow, MD, and Rishi P. Singh, MD, practice ophthalmology at the Cole Eye Institute of the Cleveland Clinic. Drs. Steinle and Chappelow have no relevant financial disclosures. Dr. Singh has received honoraria from the following companies: Alcon Inc, Genentech, Oraya Therapeutics, Inspire Pharmaceuticals, and Lifecore Pharmaceuticals. Dr. Singh has received research grant support from J&J Vistakon and Alcon. Dr. Singh can be reached via e-mail at


Analyses of complication rates of intravitreal injections have been carried out as a part of numerous trials conducted primarily to evaluate treatment efficacy. In such studies, one cannot differentiate risks related to specific agents from risks related to the procedure of intravitreal injection; further, each study population possesses a different set of risk factors for the development of endophthalmitis. Thus, we partition our discussion with respect to three therapeutic categories: antiviral agents (ganciclovir, fomivirsen, and cidofovir); triamcinolone acetonide; and anti-VEGF agents (pegap-tanib sodium, bevacizumab, and ranibizumab).

Anti-viral Agents. In 1996, a case series of 156 eyes with CMV retinitis treated with 2,890 injections of ganciclovir reported four cases of endophthalmitis (0.1% incidence).3 A more recent clinical trial evaluating intravitreal fomivirsen for the treatment of CMV retinitis reported two cases of endophthalmitis among 1,791 injections (0.1% incidence).4 Given the immunosuppressed state of the treated population, one might expect an even lower risk of endophthalmitis in the general population.

Triamcinolone Acetonide. Initial small-case series evaluating intravitreal triamcinolone for neovascular AMD found no cases of endophthalmitis.5-7 However, two larger subsequent retrospective case series8,9 together identified 17 cases of endophthalmitis, 10 of which were culture positive, following 1,362 injections of triamcinolone for various off-label indications (1.2% incidence). If only the 10 cases that were culture positive are considered, the incidence is 0.7%, still several times that observed with antiviral injections. One of these case series8 identified history of non-insulin–dependent diabetes (n=5), injection agent obtained from a multiuse vial (n=2), and history of filtering bleb (n=1) or blepharitis (n=1) as possible risk factors for development of endophthalmitis.

Roth et al.10 reported seven cases of culture-negative or “sterile” endophthalmitis that occurred within a five-week time frame during which 104 injections were administered at a single institution. Likely, these cases occurred as a reaction to a contaminant in the medication vials, rather than to an infectious etiology.

Anti-VEGF. Pegaptanib is the first FDA-approved therapy for neovascular AMD to specifically target VEGF. A pegylated oligonucleotide aptamer, it was designed to have a high binding specificity for only the VEGF165 isoform, known to stimulate pathologic neovascularization in the eye. The VISION trials11-13 consisted of two multicenter, controlled replicate trials that randomized a total of 1190 patients, regardless of angiographic lesion subtype, to receive one of three doses of pegaptanib or sham injection every six weeks for 54 weeks.

Pegaptanib and photodynamic therapy (PDT) outcomes were similar, but pegaptanib was useful for all CNV compositions so a wider range of subjects could be treated. Twelve cases of endophthalmitis were reported in conjunction with a total of 10,208 injections administered over two years, resulting in an incidence of 0.1% per injection. The investigators emphasize that nine of 12 of the cases were associated with violations of the injection protocol (ie, failure to use an eyelid speculum).13 During the clinical trial, preoperative topical fluoroquinolones were prescribed three days prior and for seven days after the procedure.

The incidence of endophthalmitis was even lower in the ANCHOR and MARINA trials that evaluated intravitreal ranibizumab for treatment of predominantly classic and minimally classic/occult CNV, respectively, in AMD. The ANCHOR trial,14 a prospective double-masked study, randomized 423 patients in a 1:1:1 ratio to receive either monthly intravitreal ranibizumab (0.3 mg or 0.5 mg) or verteporfin PDT on study entry and every three months thereafter according to accepted guidelines. Though not specifically stated, a total of 3,324 injections would have been given. Investigators reported two cases of endophthalmitis, one of which was culture-positive, resulting in an incidence of 0.06%.

The MARINA trial15 randomized 716 patients to receive one of two doses of intravitreal ranibizumab or sham injections every month for 24 months, with the goal of evaluating the safety and efficacy of ranibizumab in minimally classic and occult subfoveal CNV Researchers gave 10,443 doses and five cases of endophthalmitis were reported (incidence 0.05%), only one of which was culture-positive. Again, preop topical fluoroquinolones were prescribed three days prior and for seven days after the procedure.

Mason et al.16 reported a 0.019% rate of endophthalmitis (n=5233) as a complication of bevacizumab following povidone-iodine preparation ± prophylactic antibiotic. Another series of 10,254 injections of anti-VEGF agents (pegaptanib, bevacizumab, or ranibizumab) delivered without preinjection antibiotic prophylaxis resulted in three cases of endophthalmitis, all culture-negative.17 The 0.029% rate of endophthalmitis reported by the authors compares favorably with that reported by the major clinical trials.

The conclusion from the numerous registration trials for anti-VEGF drugs is that topical antibiotic therapy and povidone iodine preparation were indicated by the sponsor prior to intravitreal injections for these pivotal FDA trials.


Cultures taken from the 420 patients enrolled in the Endoph thalmitis Vitrectomy Study (EVS), which examined only cataract surgery-related endophthalmitis, established gram-positive, coagulase negative micrococci (predominantly Staphylococcus epidermidis) as the most common cause of acute postoperative endophthalmitis.18 Staphylococcus aureus, Streptococcus species, Enterococcus species, and miscellaneous gram-positive species were established as causative in a minority of cases, and gram-negative organisms represented only 5.9% of isolates. Polymicrobial cases accounted for 9.3% of cases.18

A subset of EVS patients (n=105) had both eyelid and intraocular (vitreous or aqueous) bacterial isolates submitted for culture. In 67.7% of cases, the eyelid and intraocular isolates were indistinguishable.19 A smaller series20 established genetically identical bacterial isolates in both external and intraocular cultures in 14 of 17 (82%) cases of acute postoperative endophthalmitis. Thus the patient's own eyelid flora serves as an important source of infecting bacteria.

As such, it is extrapolated from the EVS data that minimizing access of periocular flora to the injection site and needle is imperative in the prevention of postinjection endophthalmitis. However, more studies of intravitreal injection practices have directly linked bacterial entry to the injection procedure. In one recent study, investigators cultured both injection needle and bulbar conjunctiva of 107 eyes immediately following antibiotic plus povidone-iodine prep and injection with the aid of an eyelid speculum. Culture of two needles and 10 conjunctival swabs grew coagulase-negative Staphylococcus.


Povidone-iodine. Application of povidone-iodine to the conjunctiva represents arguably the most crucial step in the prevention of endophthalmitis prior to intravitreal injection. Though studies evaluating its efficacy specifically in the setting of intravitreal injection are lacking, data have consistently demonstrated that povidone-iodine reduces conjunctival bacterial load.21,22

Topical Antibiotic Drops. Ophthalmologists routinely use topical antibiotic therapy before and after intraocular surgery. For decades, debate has continued as to whether this practice improves outcomes. Fortunately, endophthalmitis is a rare postoperative complication. However, the low incidence of endophthalmitis renders even moderately large trials inadequately powered to detect a difference in this outcome measure. Thus, we have been forced to adopt surrogates for risk of endophthalmitis, such as conjunctival culture posifivity.

It has been suggested that topical antibiotics have a synergistic effect with preoperative topical povidone-iodine preparation with regard to conjunctival culture posifivity. Isenberg et al.23 compared the number of sterile conjunctival cultures following either three times daily polymyxin/ neomycin/gramicidin ophthalmic solution for three days preoperatively, 5% povidone-iodine solution, or a combination of both prior to ophthalmic surgery. Eighty-three percent of conjunctival cultures were negative following both preoperative topical antibiotic and povidone-iodine preparation; however, preoperative topical antibiotic and povidone-iodine alone resulted in 31% and 40% culture negativity, respectively.23

Studies conducted without antibiotic prophylaxis (either preoperatively or at any time perioperatively) have consistently reported low rates of endophthalmitis. Topical antibiotics were not prescribed prior to the day of injection for both the Diabetic Retinopathy Clinical Research Network Macular Edema Trial and SCORE clinical trial. Instead, topical gafifloxacin was given three times at least 15 minutes prior to the procedure.

Following 1,378 intravitreal triamcinolone injections administered between the two studies, only one case of endophthalmitis was reported (0.05% incidence; 95% CI: 0.001% to 0.277%).24 One series reported no cases of endophthalmitis following 1,000 consecutive injections of various agents (bevacizumab, pegaptanib, triamcinolone, gancyclovir, and ranibizumab), all using povidone-iodine prep without topical antibiotic prophylaxis pre-, peri-, or postinjection.25

Widespread use of broad-spectrum antibiotic therapy contributes to the development of bacterial antibiotic resistance. The medical community has witnessed and suffered the consequences of this phenomenon with regard to oral and intravenous antibiotic therapy. Resistance of periocular flora to many commonly used antibiotic eyedrops is well documented.26 Further, patients with local risk factors (blepharitis, conjunctivitis) and systemic risk factors (diabetes, autoimmune disease, skin disorders, asthma, treatment with immunosuppressant medication) are statistically more likely to carry multi-drug-resistant bacteria.27

Given the increase in number of patients receiving frequent topical antibiotic therapy associated with repeated anti-VEGF injections, the rare case of endophthalmitis may present as more virulent and less responsive to treatment. However, we lack evidence to confirm that increasing antibiotic resistance confers harm to the patient population as a whole.

If the surgeon chooses to employ topical antibiotic prophylaxis, fourth-generation fluoroquinolones and azithromycin may offer an advantage in the setting of increasing bacterial antibiotic resistance. Kowalski et al.28 reported that moxi-floxacin and gafifloxacin (both fourth-generation fluoroquinolones) demonstrated increased susceptibility to some S. aureus isolates that were resistant to levofloxacin (a third-generation quinolone) and ciprofloxacin and ofloxacin (second-generation quinolones). Moxifloxacin 0.5% and gafifloxacin 0.3% achieved statistically similar vitreous concentrations when administered either four times daily for three days preoperatively or one drop every 15 minutes for a total of three doses preoperatively.29

However, the vitreous concentration of both moxifloxacin and gafifloxacin was below the minimum inhibitory concentration (MIC90) for common endophthalmitis pathogens. In fact, azithromycin achieves higher conjunctival concentrations than moxifloxacin for inhibition of common lid pathogens (Figure 1). Further, resistance to fourth-generation fluoroquinolones is increasing. A study of coagulase-negative Staphylococcus isolates recovered from patients with clinical endophthalmitis revealed 96.6% sensitivity to gatifloxacin and moxifloxacin between 1990 and 1995, which dropped to only 65.4% between 1999 and 2004 (P <.05).30 A comparison of topical agents is given in Table 1.

Figure 1. Conjunctival concentration of azithromycin vs moxifloxacin.


Though we lack data confirming efficacy of prophylactic topical antibiotics in preventing endophthalmitis following intravitreal injection, many studies continue to use topical antibiotic prophylaxis either before or after the injection. A review of 17 studies that evaluated intravitreal injection of various agents prior to 2004 revealed that four (23.5%) utilized preoperative prophylactic antibiotics, whereas 10 (58.8%) utilized antibiotic drops postoperatively31 Further, most major anti-VEGF agent trials mentioned above utilized topical prophylactic antibiotic drops.

In summary, no data exist that directly support or oppose the use of prophylactic topical antibiotic therapy before, during, or after intravitreal injection. Despite concerns about multi-drug-resistant bacterial strains, many major trials, academic centers, and private practitioners continue to employ topical antibiotic prophylaxis.


In 2004, Jager et al.31 conducted a review of over 14,866 injections in 4,382 eyes performed by various investigators dating back to 1965. The 14 investigators then convened for a roundtable discussion with the objective of developing guidelines for intravitreal injection, seeking consensus when possible and citing relevant objective evidence where available. In this section, we highlight salient points of these pre-, peri-, and postinjection guidelines and refer the reader to the original publication32 for a more detailed discussion of proper intravitreal injection technique. To our knowledge, there are no recent data that dispute these guidelines.

Preinjection. Prior to injection, the ophthalmologist should identify and treat patients with comorbidities that predispose to endophthalmitis or for whom the injection procedure may exacerbate the pre-existing condition. Specifically, patients with blepharitis should first be treated with topical and/or systemic antibiotics for several days. Recent studies utilizing azithromycin showed significant improvement in many of the signs and symptoms of anterior and posterior blepharitis measured in two separate studies.33,34 Severe glaucoma should be under good control prior to injection due to the transient increase in IOP experienced with intravitreal injection. Of note, true allergy to povidone-iodine is exceedingly rare; patients with a distant history to povidone-iodine should be patch tested and then treated (even if some redness occurs with the patch test). Further, history of systemic iodine allergy is usually not a contraindication for topical povidone-iodine. As discussed, antibiotics are optional, though they have been employed in the majority of large clinical trials.

Lid scrubs can liberate bacteria from the lashes and meibomian glands into the conjunctival sac and should be avoided preoperatively Preinjection tissue softening (massage) to facilitate injection is optional, though care should be taken not to concurrently apply pressure to the eyelids or adnexa.

Peri-injection. Pupillary dilatation and application of topical anesthetic (either by sterile cotton swab or subconjunctival injection) are recommended. Subconjunctival injection was allowed at investigator's discretion in the and SCORE trials; however, some worry that use of a needle for anesthetic administration introduces the risk of globe perforation.

Without dispute, the guidelines recommend application of povidone-iodine directly to the eyelid margins, eyelashes, and ocular surface, followed by placement of an eyelid speculum. It is suggested that a drop or two of iodine is instilled after placing the speculum. The importance of the lid speculum was highlighted by change in endophthalmitis occurrences seen after speculum use was enforced during the course of the study12 Injection is given with a 27-gauge instrument or smaller, 0.5- or 0.625-inch length needle 3.5 to 4 mm posterior to the limbus in the inferotemporal quadrant. One investigator recommends rolling a sterile cotton tip over the injection site immediately following withdrawal in attempt to minimize reflux.

Postinjection. As previously discussed, topical antibiotic prophylaxis following injection is optional. Intraocular pressure should be monitored in some way — either by direct measurement or through verification of light perception vision or direct visualization of the optic nerve head in order to verify central retinal artery perfusion.

Discbarge Instructions. Instructions given on discharge following intravitreal injection should focus on early recognition of endophthalmitis, retinal detachment, and hemorrhage. Important signs and symptoms of endophthalmitis that should alert the patient to contact his or her ophthalmologist include eye pain (especially if persistent or worsening), increased sensitivity to light, and decreased vision. With respect to endophthalmitis, therapeutic success depends in part upon prompt diagnosis. If evaluation is delayed until vision has dropped to light perception only, the risk of permanent vision loss increases twofold.35 Though almost half of cases treated for postoperative endophthalmitis will eventually recover vision >20/50, about 15% will progress to no light perception vision.31


As intravitreal injections becomes increasingly important in our arsenal of treatment techniques, we must continue to strive to minimize complications. Evidence-based review has shown us the following:

Endophthalmitis, perhaps the most feared complication, can occur regardless of injected agent. However, we must remember what the VISION trial32 taught us: Most cases of endophthalmitis occur in the setting of protocol violation. Thus, the adoption of and adherence to evidence-based guidelines should take precedence in everyday practice. RP


  1. Cibis PA, Becker B, Okun E, Canaan S. The use of liquid silicone in retinal detachment surgery. Arch Ophthalmol. 1962;68:590-599.
  2. Henry K, Cantrill H, Kish MA. Intravitreal ganciclovir for patients receiving zidovudine. JAMA. 1987;257:3066.
  3. Baudouin C, Chassain C, Caujolle C, Gastaud P. Treatment of cytomegalovirus retinitis in AIDS patients using intravitreal Injections of highly concentrated ganciclovir. Ophthalmologlca. 1996;210:329-335.
  4. Safety of Intravitreal fomivirsen for treatment of cytomegalovirus retinitis in patients with AIDS. Am J Ophthalmol. 2002;133:484-498.
  5. Danis RP, Ciulla TA, Pratt LM, Anliker W. Intravitreal triamcinolone acetonlde in exudative age-related macular degeneration. Retina. 2000;20:244-250.
  6. Gillies MC, Simpson JM, Luo W, et al. A randomized clinical trial of a single dose of Intravitreal triamcinolone acetonlde for neovascular age-related macular degeneration: one-year results. Arch Ophthalmol. 2003;121:667-673.
  7. Penfold PL, Gyory JF, Hunyor AB, Billson FA. Exudative macular degeneration and intravitreal triamcinolone. A pilot study. Aust N Z J Ophthalmol. 1995;23:293-298.
  8. Moshfeghi DM, Kaiser PK, Scott IU, et al. Acute endophthalmitis following Intravitreal triamcinolone acetonlde injection. Am J Ophthalmol. 2003;136: 791-796.
  9. Nelson ML, Tennant MT, Sivalingam A, Regillo CD, Belmont JB, Martidls A. Infectious and presumed noninfectious endophthalmitis after intravitreal triamcinolone acetonide Injection. Retina. 2003;23:686-691.
  10. Roth DB, Chieh J, Spirn MJ, Green SN.Yarian DL, Chaudhry NA. Noninfectious endophthalmitis associated with intravitreal triamcinolone injection. Arch Ophthalmol. 2003;121:1279-1282.
  11. Chakravarthy U, Adamis AP, Cunningham ET, Jr., et al. Year 2 efficacy results of 2 randomized controlled clinical trials of pegaptanib for neovascular age-related macular degeneration. Ophthalmology. 2006;113:1508 e1-e25.
  12. Gragoudas ES, Adamis AP, Cunningham ET, Jr., Felnsod M, Guyer DR. Pegaptanib for neovascular age-related macular degeneration. N Engl J Med. 2004;351:2805-2816.
  13. D'Amico DJ, Masonson HN, Patel M, et al. Pegaptanlb sodium for neovascular age-related macular degeneration: two-year safety results of the two prospective, multicenter, controlled clinical trials. Ophthalmology. 2006;113:992-1001 e6.
  14. Brown DM, Kaiser PK, Michels M, et al. Ranlbizumab versus verteporfln for neovascular age-related macular degeneration. N Engl J Med. 2006;355:1432-1444.
  15. Rosenfeld PJ, Brown DM, Heler JS, et al. Ranlbizumab for neovascular age-related macular degeneration. N Engl J Med. 2006;355:1419-1431.
  16. Mason JO, 3rd, White MF, Felst RM, et al. Incidence of acute onset endophthalmitis following Intravitreal bevacizumab (Avastin) Injection. Retina. 2008: 28:564-567.
  17. Pilli S, Kotsolis A, Spalde 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.
  18. Han DP, Wisniewski SR, Wilson LA, et al. Spectrum and susceptibilities of microbiologic Isolates in the Endophthalmitis Vitrectomy Study Am J Ophthalmol. 1996;122:1-17.
  19. Bannerman TL, Rhoden DL, McAllister SK, Miller JM, Wilson LA. The source of coagulase-negative staphylococci In the Endophthalmitis Vitrectomy Study A comparison of eyelid and Intraocular Isolates using pulsed-field gel electrophoresis. Arch Ophthalmol. 1997; 115:357-361.
  20. Speaker MG, Milch FA, Shah MK, Eisner W, Kreiswirth BN. Role of external bacterial flora In the pathogenesis of acute postoperative endophthalmitis. Ophthalmology. 1991;98:639-649; discussion 650.
  21. Caldwell DR, Kastl PR, Cook J, Simon J. Povidone-iodine: its efficacy as a preoperative conjunctival and periocular preparation. Ann Ophthalmol. 1984;16: 577, 580.
  22. Apt L, Isenberg S, Yoshimori R, Paez JH. Chemical preparation of the eye In ophthalmic surgery. III. Effect of povldone-iodlne on the conjunctiva. Arch Ophthalmol. 1984;102:728-729.
  23. Isenberg SJ, Apt L, Yoshimorl R, Khwarg S. Chemical preparation of the eye In ophthalmic surgery. IV Comparison of povldone-iodlne on the conjunctiva with a prophylactic antibiotic. Arch Ophthalmol. 1985;103:1340-1342.
  24. Bhavsar AR, lp MS, Glassman AR. The risk of endophthalmitis following Intravitreal triamcinolone Injection In the DRCRnet and SCORE clinical trials. Am J Ophthalmol. 2007;144:454-456.
  25. Bhavsar AR. A consecutive series of 1000 Intravitreal Injections without topical antibiotic prophylaxis. Poster presented at: Annual Meeting of the American Academy of Ophthalmology; November 12-13, 2006; Las Vegas, NV
  26. Kunimoto DY, Das T, Sharma S, et al. Microbiologic spectrum and susceptibility of Isolates: part I. Postoperative endophthalmitis. Endophthalmitis Research Group. Am J Ophthalmol. 1999;128:240-242.
  27. Mino de Kaspar H, Shriver EM, Nguyen EV, et al. Risk factors for antibiotic-resistant conjunctival bacterial flora in patients undergoing Intraocular surgery Graefes Arch Clin Exp Ophthalmol. 2003;241:730-733.
  28. Kowalski RP, Dhaliwal DK, Karenchak LM, et al. Gatifloxacin and moxifloxacin: an in vitro susceptibility comparison to levofloxacin, ciprofloxacin, and ofloxacin using bacterial keratitis Isolates. Am J Ophthalmol. 2003;136:500-505.
  29. Costello � Bakri SJ, Beer PM, et al. Vitreous penetration of topical moxlfloxacln and gatifloxacln in humans. Retina. 2006;26:191-195.
  30. Miller D, Flynn PM, Scott IU, Alfonso EC, Flynn HW, Jr. In vitro fluoroquinolone resistance In staphylococcal endophthalmitis Isolates. Arch Ophthalmol. 2006: 124:479-483.
  31. Jager RD, Aiello LP, Patel SC, Cunningham ET, Jr. Risks of Intravitreal injection: a comprehensive review. Retina. 2004;24:676-698.
  32. Alello LP, Brucker AJ, Chang S, et al. Evolving guidelines for Intravitreal injections. Retina. 2004;24:S3-S19.
  33. John T, Shah AA. Use of azithromycin ophthalmic solution In the treatment of chronic mixed anterior blepharitis. Ann Ophthalmol. 2008;40:68-74.
  34. Luchs J. Efficacy of Topical azithromycin ophthalmic solution 1% In the treatment of posterior blepharitis. Advances of Therapy. 2008;25:858-870.
  35. Lemley CA, Han DP. Endophthalmitis: a review of current evaluation and management. Retina. 2007;27:662-80.

Retinal Physician, Issue: September 2009