Endophthalmitis After Intravitreal Injections in Patients with Self-Reported Iodine Allergy

Endophthalmitis After Intravitreal Injections in Patients With Self-reported Iodine Allergy

BOBECK S. MODJTAHEDI, TAVE´ VAN ZYL, HEMANG K. PANDYA, ROBERT E. LEONARD, II, AND DEAN ELIOTT

PURPOSE: To present cases of endophthalmitis IMITING THE RISK OF POSTINJECTION ENDOPHTHAL- following intravitreal injections where povidone-iodine mitis is an area of considerable practical and aca- (PI) was not used as part of the surgical preparation. demic interest, especially in the era of regular L DESIGN: Retrospective case series. intravitreal injections. Although the incidence of postin- METHODS: All cases of presumed injection-related jection endophthalmitis is low (0.056% per injection in a endophthalmitis presenting to the Massachusetts Eye recent meta-analysis),1 the risk to an individual patient is and Ear Infirmary between June 2008 and November magnified by the recurrent nature of the procedure. Using 2014 and Dean McGee Eye Institute between January povidone-iodine (PI) for surgical antisepsis is well estab- 2010 and January 2015 were identified. Patients who lished and is the only preoperative measure shown to did not receive PI preparation owing to documented reduce the risk of endophthalmitis in patients undergoing self-reported allergy to iodine, iodine-containing contrast intraocular surgery.2 The importance of PI application material, or shellfish were identified and their injection prior to intravitreal injection has been observed by histories and clinical courses reviewed. Nentwich and associates,3 Bhavsar and Sandler,4 and Bryn- 5 RESULTS: The combined rate of postinjection endoph- skov and associates, the latter of whom described no cases thalmitis at these 2 centers was 0.019%. Among 42 of endophthalmitis after 20 293 injections. Patients with a patients with postinjection endophthalmitis, 5 (11.9%) reported allergy to iodine or iodine-containing agents pre- did not receive PI prophylaxis. The mean number of sent a clinical challenge and some clinicians modify the intravitreal injections without PI before the development standard ophthalmic preparation to avoid PI because of of endophthalmitis was 10.6 with a 9.4% rate of endoph- perceived iodine allergy. This study presents a series of thalmitis (5 cases per 53 injections). All patients under- cases of intravitreal injection–related endophthalmitis in went tap-and-inject procedures with vancomycin 1 mg patients who underwent modified surgical preparation and ceftazidime 2 mg. Two patients did not receive PI owing to reported iodine allergy. at the time of tap and inject; 1 of these patients required subsequent pars plana vitrectomy for worsening clinical course. Cultures were positive in 4 of 5 cases; all positive cultures grew coagulase-negative Staphylococcus. All METHODS patients who received subsequent intravitreal injections received PI prophylaxis without allergic reactions, thus INSTITUTIONAL REVIEW BOARD (IRB) APPROVAL FOR A demonstrating a lack of true PI allergy. retrospective chart review was obtained from the Massa- chusetts Eye and Ear Infirmary Human Studies Committee CONCLUSIONS: Avoiding PI owing to self-reported iodine ‘‘allergy’’ risks substantial ocular morbidity. and the University of Oklahoma Health Sciences Center Allergy testing can be pursued per patient request or in IRB. All cases of presumed intravitreal injection–associ- rare cases of suspected true PI allergy; however, in cases ated endophthalmitis diagnosed at 2 institutions—Massa- where delayed treatment would adversely affect visual chusetts Eye and Ear Infirmary (MEEI, Boston, outcome, the clinician should feel confident that minimal Massachusetts), between June 1, 2008 and November 1, allergic risk exists. (Am J Ophthalmol 2016;170: 2014, and Dean McGee Eye Institute (DMEI, Oklahoma 68–74. Ó 2016 Elsevier Inc. All rights reserved.) City, Oklahoma), between January 1, 2010 and January 1, 2015—were identified using International Classification of Disease (ICD)-9 billing codes. Patients who had received ophthalmic care by outside doctors and subsequently Supplemental Material available at AJO.com. Accepted for publication Jul 12, 2016. presented to either institution were included, in addition From the Retina Service, Department of Ophthalmology, to those who were established patients at the study centers. Massachusetts Eye and Ear Infirmary, Harvard Medical School, Boston, Patients with documented self-reported allergy to iodine, Massachusetts (B.S.M., T.v.Z., D.E.); and Retina Service, Dean McGee Eye Institute, University of Oklahoma College of Medicine, Oklahoma iodinated contrast material, or shellfish were identified City, Oklahoma (H.K.P., R.E.L.). and their charts further reviewed to determine whether Inquiries to Dean Eliott, Retina Service, Department of PI was instilled as part of the antiseptic protocol prior to Ophthalmology, Massachusetts Eye and Ear Infirmary, Harvard Medical School, 243 Charles St, Boston, MA 02114; e-mail: Dean_Eliott@meei. the injection that preceded endophthalmitis as well as sub- harvard.edu sequent tap and inject or pars plana vitrectomy.

68 Ó 2016 ELSEVIER INC.ALL RIGHTS RESERVED. 0002-9394/$36.00 http://dx.doi.org/10.1016/j.ajo.2016.07.010 Additionally, complete procedure history was reviewed, injections without PI prior to development of endophthal- including all intravitreal injections before and after mitis, which corresponds to an 11% rate of endophthalmi- endophthalmitis as well as further demographic and clin- tis and a mean of 8.8 injections without PI before ical characteristics including age, sex, medication injected, endophthalmitis. At MEEI, the median interval between ocular condition being treated, interval between last injec- injection and presentation with endophthalmitis was tion and endophthalmitis presentation, microbiology 4 days and 3.5 days for those receiving and not receiving results, visual acuity (at presentation and last follow-up), PI, respectively. At DMEI the median number of days and number of injections (before and after endophthalmi- between injection and presentation was 5 days in both tis) with and without preinjection PI antisepsis. groups. Patients presented for evaluation between 0 and 2 days after symptom onset. All 5 patients who did not receive PI at the time of their injection were managed with tap-and-inject procedures RESULTS and 3 received 5% PI antisepsis as part of the tap-and- inject procedure protocol, without complications. One INJECTION PROTOCOLS AT BOTH MEEI AND DMEI WERE patient (Case 5) did not receive PI at the time of tap and reviewed. The procedures were office-based at both institu- inject and required pars plana vitrectomy with intravitreal tions. To achieve anesthesia, proparacaine and lidocaine antibiotics within 24 hours because of worsening examina- were employed. Topical lidocaine was the norm at DMEI, tion findings. This surgery was done without preoperative whereas physicians were divided evenly between topical PI preparation. Cultures were positive in 4 of 5 cases: all and subconjunctival lidocaine use at MEEI. Topical PI grew coagulase-negative Staphylococcus. After resolution was applied to the ocular surface; eyelid scrubs were not of endophthalmitis, 4 of the 5 patients went on to receive typically used. Masks and eyelid drapes were not used but subsequent intravitreal injections, all with PI antisepsis and speculum, gloves, and caliper use were the norm at both without allergic reactions. One of these 4 patients received institutions. Topical antibiotics were typically not given af- a modified PI preparation, where a sterile cotton-tipped ter injections; however, 2 patients (Cases 2 and 3) who did applicator soaked in PI was applied directly over the con- not receive PI at the time of their injection received post- junctiva at the planned injection site immediately prior injection topical antibiotics, and 1 patient who did receive to injection, and ultimately went on to have another PI and went on to develop endophthalmitis (at MEEI) also episode of endophthalmitis. The 1 patient who did not received topical antibiotics. None of the patients with receive any post-endophthalmitis intravitreal injections endophthalmitis (with or without PI use at the time of did not have a reaction to PI use at the time of her tap injection) at DMEI received postinjection antibiotics. and inject. The patients’ individual case histories are There were 30 046 injections performed at MEEI during described below and summarized in Table 1. Table 2 the study period and 6 of these resulted in endophthalmitis, provides a summary of their injection histories. giving a 0.020% rate of postinjection endophthalmitis. There was a 0.018% rate of postinjection endophthalmitis INDIVIDUAL CASE HISTORIES: Case 1. An 86-year-old for patients injected at DMEI (33 699 injections, 6 infec- woman with a history of neovascular age-related macular tions). The combined rate of endophthalmitis was 0.019% degeneration (AMD) of the right eye, treated previously at both institutions. There were 21 postinjection endoph- with a total of 12 intravitreal aflibercept injections, thalmitis cases from outside doctors who referred the presented 5 days following the last injection with patients to MEEI and 9 cases referred to DMEI; however, worsening eye pain and nausea. She was found to have because the total number of injections in the patient catch- hand motion (HM) visual acuity (decreased from 20/32 ment area is unknown, the regional rate of endophthalmitis at last visit), a 1.5 mm hypopyon, and diffuse mobile outside these institutions could not be calculated. In total, vitreous debris. Unlike her prior 11 intravitreal 42 cases of presumed intravitreal injection–related endoph- injections, the patient’s most recent intravitreal injection thalmitis were identified: 27 cases at MEEI and 15 at DMEI. was performed with topical preinjection gatifloxacin Five of the 42 patients (11.9%) were identified as having 0.5% instead of PI antisepsis owing to the patient’s report a history of self-reported prior allergic reaction to iodine- of a recent unknown adverse reaction to iodinated related agents and did not receive PI antisepsis at the contrast material. She underwent an anterior chamber time of the injection immediately preceding presentation paracentesis (and vitreous aspirate, which was dry) and with endophthalmitis. These patients underwent 53 injec- intravitreal injection of vancomycin (1 mg/0.1 mL) and tions without PI (including bilaterally in Case 3) before ceftazidime (2 mg/0.1 mL) with use of 5% PI prior to the developing endophthalmitis in the affected eye, which rep- procedure without adverse reaction. Anterior chamber resents a 9.4% rate of endophthalmitis and a mean of 10.6 culture was positive for S lugdunensis, a coagulase- injections before endophthalmitis. When excluding injec- negative Staphylococcus (sensitive to vancomycin). Visual tions without PI to the eye that did not go on to develop acuity improved to 20/80 at last follow-up. She did not endophthalmitis (the left eye of Case 3), there were 44 receive further intravitreal injections.

VOL. 170 POSTINJECTION ENDOPHTHALMITIS AND IODINE ALLERGY 69 TABLE 2. Injection History of Patients With Reported Iodine Allergy Who Developed Endophthalmitis After Not Receiving Follow-up VA at Last Povidone-Iodine With Their Injection

Injections Endophthalmitis Procedural Without Injections Total Rate Without VA on Patient Eye Iodine With Iodinea Injectionsa Iodine (%) Presentation

1 OD 1 11 12 100

2 20/200 20/25 2 OS 12 6 18 8.33 þ 3OUb 36 72c 108 2.8 VA Before 4 OD 1 2 3 100 Endophthalmitis 20/3220/25 HM 20/80 20/60 HM 20/70 Unknown HM 20/160 5OS3 3 6 33

aIncludes injections done after endophthalmitis. bPatient developed endophthalmitis in the right eye.

Bacteriology cIncludes injections with ‘‘modiﬁed’’ povidone-iodine prep. S lugdunensis S epidermidis -S epidermidis 20/50 HM 20/125 S epidermidis

Case 2. A 71-year-old woman with a history of neovascular AMD of the left eye, treated previously with photodynamic therapy and a total of 12 ranibizumab injections, presented 33 days after her last Management

PPV 48 hours later injection with 2 days of pain and photosensitivity.

/ Visual acuity was 20/200, reduced from 20/25þ prior to injection. The eye had 4þ anterior chamber cells with a I: Vancomycin 1 mg/0.1 mL I: Vancomycin 1 mg/0.1 mL and I: Vancomycin 1 mg/0.1 mL I: Vancomycin 1 mg/0.1 mL I: Vancomycin 1 mg/0.1 mL and ceftazidime 2 mg/0.1 mL ceftazidime 2 mg/0.1 mL and ceftazidime 2 mg/0.1 mL and ceftazidime 2 mg/0.1 mL and ceftazidime 2 mg/ 0.1 ml þ þ þ þ þ 0.75 mm hypopyon and vitreous opacities. She had received gatiﬂoxacin 0.5% without PI prior to all previous injections owing to a history of respiratory

Days to distress and hives with intravenous iodine contrast. Presentation She underwent an intravitreal tap and inject with vancomycin (1 mg/0.1 mL) and ceftazidime (2 mg/0.1 mL); 5% PI was used prior to the procedure without adverse visual acuity.

¼ reaction. Vitreous cultures were positive for S epidermidis (sensitive to vancomycin). Visual acuity improved to 20/25 Alternative Agent Used Gatiﬂoxacin 0.5% 33 T Shurclens 4 T at last follow-up and she has received PI with her subsequent 6 injections without incident.

tap and inject; VA Case 3. A 62-year-old woman with a history of ¼

I bilateral neovascular AMD presented 3 days after þ bilateral intravitreal aﬂibercept injections with pain, redness, and photophobia in the right eye. Vision was

Summary of Cases of Endophthalmitis After Intravitreal Injections Without the Use of Povidone-Iodine HM, a decline from 20/50 prior to injection. The right Reported Allergy eye had 4þ anterior chamber cells, <1 mm hypopyon, and 3–4þ vitreous cell/debris. The patient underwent TABLE 1. respiratory distress and hives anterior chamber paracentesis (and vitreous tap, which Iodine contrast—hives Gatifloxacin 0.5% 5 T Iodine contrast—hives and Topical iodine—eyelid swelling Gatifloxacin 0.5% 3 T Topical iodine—eyelid swelling Anaphylaxis to iodine Moxifloxacin 0.5% 3was T dry) and intravitreal injection of vancomycin pars plana vitrectomy; T

¼ (1 mg/0.1 mL) and ceftazidime (2 mg/0.1 mL) without PI owing to patient refusal. Anterior chamber culture was negative, without organisms on Gram stain. Visual aflibercept ranibizumab aflibercept bevacizumab bevacizumab acuity at last follow-up was 20/125. Prior to presentation she had received 31 intravitreal ranibizumab injections with PI in the right eye, after which she developed hand motions; PPV intolerable eyelid swelling and then refused PI. The ¼ subsequent treatment course included 20 intravitreal HM Patient Age/Sex1 Procedure 86/F Intravitreal 2 71/F3 Intravitreal 60/F Intravitreal 4 65/F Intravitreal 5 86/F Intravitreal ranibizumab and 7 intravitreal aflibercept injections in the right eye (as well as 4 intravitreal ranibizumab and

70 AMERICAN JOURNAL OF OPHTHALMOLOGY OCTOBER 2016 5 aflibercept injections in the left eye) with gatifloxacin as a DISCUSSION substitute for PI prep. After her episode of endophthalmitis, the patient was counseled about the importance of PI prep SELF-REPORTED ALLERGY TO IODINE PRESENTS A SIGNIFI- and was asked to reconsider injections with PI. She cant challenge to clinicians treating patients who require ultimately consented to a modified PI prep by which a ophthalmic procedures. Intravitreal injection patients are sterile cotton-tipped applicator soaked in PI was applied exposed to a recurrent risk of endophthalmitis, given the re- to the conjunctiva directly over the planned injection petitive nature of the procedure. Preoperative PI has been site for 20–30 seconds. The injection protocol otherwise found to reduce the risk of endophthalmitis after cataract included topical lidocaine, no lid scrubs with iodine, and surgery2 and is similarly thought to hold significant value no masks. She had 20 injections of intravitreal aflibercept in other ocular procedures, including intravitreal injec- in the right eye (and 21 in the left eye) with this tions.2–7 Clinicians are faced with 3 common challenging modified PI prep and developed another episode of scenarios: reported or observed ‘‘allergy’’ to topical iodine endophthalmitis in the right eye. She was treated with preparation, adverse reaction to iodinated radiocontrast, tap and inject but insisted that the doctor use the and allergy to seafood. True allergy to iodine is not modified PI prep. Vitreous cultures were positive for S believed to be possible. The element is not sufficiently epidermidis (sensitive to vancomycin). complex to induce allergenicity and is essential for thyroid function and normal biologic processes.8,9 Most Case 4. A 65-year-old woman with a history of salt in the United States is supplemented with iodine as a neovascular AMD in the right eye presented 4 days public health measure.8 Simply, an allergy to elemental following her first injection of intravitreal bevacizumab iodine is not compatible with life. with pain, redness, and decreased vision from 20/60 to A common misperception is that a reaction to iodinated HM. Examination was notable for corneal edema, contrast media is due to an allergy to iodine. The mechanism dispersed anterior chamber fibrin with 2 mm hypopyon, behind this adverse event is poorly understood. Nonallergic and moderate vitritis. Owing to reported allergy to mechanisms, such as the release of basophil and mast cell topical PI consisting of swelling and hives, she had mediators via direct cell membrane effects or indirect com- received Surclens (20% solution of poloxamer 188 in plement activation, have long been postulated to play the sterile water) as a substitute to PI antisepsis prior to primary role in these reactions.10 An immunologic/allergic injection. She underwent a tap and inject with vancomycin component likely plays a role in some patients. Prieto- (1 mg/0.1 mL) and ceftazidime (2 mg/0.1 mL); 5% PI was Garcia and associates found that 10.4% of patients with a used for this procedure. Vitreous cultures were positive for S history of immediate hypersensitivity reactions to iodinated epidermidis (sensitive to vancomycin). Visual acuity contrast media had positive intradermal tests to their offend- improved to 20/70 at last follow-up and she has received ing contrast media. Seven patients were subsequently PI with her subsequent 2 injections without incident. exposed to a different iodinated contrast medium (to which they had negative skin tests) and tolerated it well, indicating Case 5. An 86-year-old woman with a history of that iodine (which is universal in iodinated contrast media) neovascular AMD in the right eye presented with floaters was not the offending agent.10 The chemical compound and decreased vision 3 days following her third responsible for sensitization to iodinated contrast media is intravitreal bevacizumab injection. She had HM vision, a not known and several compounds may be responsible hypopyon, and vitreous debris. Owing to patient-reported owing to different patterns of cross-reactivity in some pa- anaphylaxis to iodine-containing agents, she had been tients to different contrast media.11 Iodine is not felt to be receiving topical moxifloxacin as a substitute to PI the eliciting factor in iodinated contrast media reactions antisepsis prior to intravitreal injections. She underwent and, as such, topical PI can safely be used in patients with tap and inject with vancomycin (1 mg/0.1 mL) and a history of adverse events to contrast.8,9 ceftazidime (2 mg/0.1 mL); 5% PI was not used prior to Seafood allergy is commonly incorrectly thought to the procedure because of the severity of the reported confer a risk of allergy to iodine-containing compounds. allergy and patient request. Vitreous cultures were Seafood allergy is the response of an IgE-mediated reaction positive for Sepidermidis(sensitive to vancomycin). to parvalbumins in scaled fish and tropomyosin in shellfish. Owing to worsening of symptoms and visual acuity over Therefore, even though seafood has relatively high iodine the next 48 hours, she underwent pars plana vitrectomy content, an allergy to seafood does not represent a risk with intravitreal injection of vancomycin, ceftazidime, for allergic reaction to iodine-containing compounds and dexamethasone. Again, PI was not used as part of such as PI or iodinated contrast.8,9 Radiologists typically presurgical preparation. Visual acuity improved to 20/120 do not factor shellfish allergy into their risk assessment at last follow-up; she received 3 subsequent intravitreal for iodinated contrast reactions. injections of ranibizumab with PI preparation without This study presents 5 cases of injection-related endoph- complication. thalmitis among patients who did not receive PI prior to

VOL. 170 POSTINJECTION ENDOPHTHALMITIS AND IODINE ALLERGY 71 intravitreal injection owing to presumed iodine allergy. in vivo.17 Shimada and associates had no cases of endoph- Together, these patients consisted of 11.9% of all thalmitis after 15 144 injections using 10% concentration injection-related endophthalmitis cases seen during the on the skin/lashes and 0.25% in the eye.18 Anaphylaxis study period at MEEI and DMEI collectively. Endophthal- to topical exposure of PI is exceedingly rare.19,20 There mitis occurred after an average of 10.6 injections in are fewer than 10 cases of IgE-mediated allergic reactions patients without PI use, corresponding to a rate of 9.4% that have been reported after external (topical, vaginal, (5 cases of endophthalmitis per 53 injections without PI). rectal) PI exposure.20 There are no reports of anaphylaxis One patient had a second episode of endophthalmitis following ocular PI use.9 Allergic reactions to PI are with use of a modified PI prep; otherwise, there were no sub- derived from the povidone (polyvinylpyrrolidone), which sequent cases of endophthalmitis in these patients once PI is a widely used polymer and is used as an excipient.19,20 was used. The failure to use PI was a risk factor leading to Anaphylaxis has been seen after exposure to povidone the development of endophthalmitis in these cases. Most alone, including in a child with prior reaction to PI.19,21 cases of postinjection endophthalmitis are referred to our Additionally, IgE antibodies to povidone have been institutions from outside doctors, and as a result, statistical demonstrated in the serum of allergic individuals.22 Povi- analysis comparing the rates of endophthalmitis with and done is a near-ubiquitous compound found in a wide range without PI is not possible owing to missing data regarding of commonly encountered objects, from drug tablets to the number of injections done by referring doctors, their hairspray.19,20 A true allergy to this compound would preinjection protocol, and how many of their patients have potentially debilitating consequences. presented elsewhere with endophthalmitis. Nonetheless, Alternatives to PI are limited. Chlorhexidine-based considering the low rate of postinjection endophthalmitis preparations may be more effective in certain nonophthal- at our institutions (6 out of 30 046 injections, 0.020%, at mic situations than PI23; however, their ocular use is MEEI; and 6 out of 33 699 injections, 0.018%, at DMEI; limited by endothelial toxicity.9,24 Additionally, for a total of 12 out of 63 745 injections, 0.019%) and an chlorhexidine is often not readily available in most approximate risk of 0.056%1 following intravitreal injec- ophthalmology offices and is usually combined with tions in the literature, it can be inferred that patients alcohol, which risks corneal epitheliopathy and would who did not receive PI had a disproportionate risk of dis- make intraocular visualization difficult during surgery. ease. This notion is supported by a recently published, large Amukine 0.06% solution (isotonic hypochlorite sodium retrospective case series of 316 576 intravitreal injections solution) has been found to have a good ocular safety from 25 ophthalmic centers in France in which prophylaxis profile (based on assessment of keratitis and conjunctival with an antibiotic (topical 1.5% azithromycin) or anti- hyperemia) compared with PI, but its efficacy for septic (topical 0.05% picloxydine) in the absence of antisepsis is unclear.25 Topical antibiotics may help to PI was statistically associated with an increased incidence decrease conjunctival bacterial load and have been used of endophthalmitis (incidence rate ratio ¼ 2.77, 95% CI, as monotherapy in patients with presumed PI allergy. 1.54–5.00, P ¼ .001).12 Ciprofloxacin was found to result in more bacterial eradica- Topical reactions to iodine preparations are overwhelm- tion when instilled once on the day prior to surgery and 4 ingly irritant in nature but are frequently confused with times during the morning of surgery (separated by 15 allergy.13 Lachapelle found that only 0.4% of patients minutes) when compared with ofloxacin (instilled in a (2/500) had true allergic contact dermatitis to PI, although similar fashion) or PI (with application 15 minutes before 2.8% (14/500) had a positive initial patch test, which was cultures). Small sample size (n ¼ 164 eyes) limits the gener- due to the irritant properties of PI.14 This demonstrated the alizability of this study.26 The high rate of resistance of importance of following a positive patch test with a confir- coagulase-negative Staphylococcus to topical fluoroquino- matory examination such as a repeat open application test lones makes their use suboptimal. The relationship (ROAT) to prove true allergic reaction.14 Many reports of between postinjection antibiotic use and endophthalmitis allergic contact dermatitis in the literature do not use rate remains controversial, with some authors finding a nonirritant preparations and, as such, many of these reports non–statistically significant trend toward increased inci- represent irritation rather than true allergy.15 Lee and asso- dence in patients receiving antibiotics.27 In this study, ciates found that PI less than or equal to 10% concentra- there were 3 patients who developed endophthalmitis after tion was the least irritating form of iodine when being given topical antibiotics: 2 who did not receive PI at compared with petrolatum and isopropyl alcohol prepara- the time of their injection and 1 who did. PI possesses tions.15 Lengthy contact duration may worsen the irritant activity against a wider spectrum of microbes compared response,9,14 and although a contact time of at least 30 with antibiotic drops, including against fungi, viruses, seconds is necessary to achieve sufficient antimicrobial and spores.26 The addition of PI has been found to signifi- activity from PI,16 it may be rinsed from the eye following cantly reduce conjunctival flora in patients treated with the injection to mitigate ocular irritation. In vitro studies topical levofloxacin.28 de Kaspar and associates found have suggested that more dilute PI may paradoxically be that eyes pretreated with levofloxacin prior to 1% PI irriga- more effective, although the opposite trend has been shown tion had less conjunctival bacteria than those treated with

72 AMERICAN JOURNAL OF OPHTHALMOLOGY OCTOBER 2016 only 1% PI; however, 1% PI has been found to be inferior to of discomfort after PI use, which can erroneously be 5% PI, which makes these results difficult to interpret.29 PI attributed to allergy, and may have an independently has poor intraocular penetration and lacked efficacy increased risk of endophthalmitis.31 compared with moxifloxacin and ofloxacin once bacteria There are several limitations to this study, including the entered the anterior chamber in a rabbit model.30 The ben- small sample size and retrospective design. Additionally, efits of PI are numerous. It is the least expensive option, is many physicians use a template for injection procedure readily available, and is immune to the development of notes and it is foreseeable that iodine use may have errone- antimicrobial resistance.6 As a result, it remains the gold ously been checked off or included on these forms even standard for preoperative ophthalmic preparation.6 The when it was not used in some patients. A thorough rinsing importance of PI use for injections is underscored by the of the ocular surface after PI use can minimize irritation and requirement to use PI with all injections in the Diabetic patient discomfort. In some cases a few drops of PI on the Retinopathy Clinical Research Network’s protocol for in- patient’s forearm convinced the patient that he or she jections to reduce the risk of endophthalmitis.4 was not allergic to PI and the patient then permitted its In summary, avoiding PI because of presumed or self- use without subsequent allergic reaction after the proced- reported iodine ‘‘allergy’’ risks substantial ocular ure. Some patients are resolute in their refusal to undergo morbidity. Most reported allergic reactions to PI are irri- PI antisepsis prior to intravitreal injection; in such cases, tant in nature and should not dissuade the clinician or when suspicion of true allergy to PI exists, patch testing from using it to prevent endophthalmitis. This study pre- (with confirmatory ROAT testing) may be employed. sents 5 patients who underwent intravitreal injections Patch testing delays treatment by approximately 1 week without the use of PI owing to self-reported allergy who (with another week for ROAT testing, if needed), which developed postinjection endophthalmitis. None of these represents an unacceptable delay in treatment for certain patients demonstrated allergic reactions when PI was conditions, including neovascular glaucoma, viral retinitis, used for injections after they developed endophthalmitis endophthalmitis, and (to a lesser extent) neovascular (including tap and inject), which supports the notion AMD; therefore, such ancillary testing should be reserved that these patients did not have a true allergy. The for other conditions, such as cystoid macular edema, where patients presented herein were erroneously assumed to treatment delay is unlikely to permanently impact visual have PI allergy and the resultant lack of PI prophylaxis acuity. In patients who report a history of PI ‘‘allergy,’’ it was likely responsible for the development of endophthal- is recommended to offer reassurance and to proceed with mitis after the injection. Patients with blepharitis and PI use, as its antiseptic properties far outweigh its potential associated evaporative dry eye are more likely to complain irritant properties.

FUNDING/SUPPORT: NO FUNDING OR GRANT SUPPORT. FINANCIAL DISCLOSURES: DEAN ELIOTT REPORTS PERSONAL FEES from Acucela, personal fees from Alcon, personal fees from Alimera, personal fees from Allergan, personal fees from Arctic, personal fees from Avalanche, personal fees from Bausch & Lomb, personal fees from Biogen, personal fees from Dutch Ophthalmic, grants from Ocata, personal fees from Ophthotech, personal fees from Thrombogenics, personal fees from MacuLogix, and grants from Neurotech, outside the submitted work. The following authors have no ﬁnancial disclosures: Bobeck S. Modjtahedi, Tave´ van Zyl, Hemang K. Pandya, and Robert E. Leonard II.

REFERENCES 5. Brynskov T, Kemp H, Sorensen TL. No cases of endophthalmitis after 20,293 intravitreal injections in an operating room 1. Fileta JB, Scott IU, Flynn HW Jr. Meta-analysis of setting. Retina 2014;34(5):951–957. infectious endophthalmitis after intravitreal injection of 6. Ahmed Y, Scott IU, Pathengay A, Bawdekar A, Flynn HW Jr. anti-vascular endothelial growth factor agents. Povidone-iodine for endophthalmitis prophylaxis. Am J Ophthalmic Surg Lasers Imaging Retina 2014;45(2): Ophthalmol 2014;157(3):503–504. 143–149. 7. Schwartz SG, Flynn HW Jr. Endophthalmitis associated with 2. Speaker MG, Menikoff JA. Prophylaxis of endophthalmitis intravitreal anti-vascular endothelial growth factor injec- with topical povidone-iodine. Ophthalmology 1991;98(12): tions. Curr Ophthalmol Rep 2014;2(1):1–5. 1769–1775. 8. Schabelman E, Witting M. The relationship of radiocontrast, 3. Nentwich MM, Yactayo-Miranda Y, Schwarzbach F, Wolf A, iodine, and seafood allergies: a medical myth exposed. J Emerg Kampik A, Mino de Kaspar H. Endophthalmitis after intravi- Med 2010;39(5):701–707. treal injection: decreasing incidence and clinical outcome-8- 9. Wykoff CC, Flynn HW Jr, Han DP. Allergy to povidone- year results from a tertiary ophthalmic referral center. Retina iodine and cephalosporins: the clinical dilemma in 2014;34(5):943–950. ophthalmic use. Am J Ophthalmol 2011;151(1):4–6. 4. Bhavsar AR, Sandler DR. Eliminating antibiotic prophy- 10. Prieto-Garcia A, Tomas M, Pineda R, et al. Skin test-positive laxis for intravitreal injections: a consecutive series of immediate hypersensitivity reaction to iodinated contrast 18,839 injections by a single surgeon. Retina 2015;35(4): media: the role of controlled challenge testing. J Investig Aller- 783–788. gol Clin Immunol 2013;23(3):183–189.

VOL. 170 POSTINJECTION ENDOPHTHALMITIS AND IODINE ALLERGY 73 11. Brockow K, Romano A, Aberer W, et al. Skin testing in patients urticaria-angioedema after topical application of povi- with hypersensitivity reactions to iodinated contrast media - a done-iodine. Allergol Immunopathol (Madr) 1998;26(1): European multicenter study. Allergy 2009;64(2):234–241. 23–26. 12. Dossarps D, Bron AM, Koehrer P, et al. Endophthalmitis after 23. Darouiche RO, Wall MJ Jr, Itani KM, et al. Chlorhexidine- intravitreal injections: incidence, presentation, manage- alcohol versus povidone-iodine for surgical-site antisepsis. ment, and visual outcome. Am J Ophthalmol 2015;160(1): N Engl J Med 2010;362(1):18–26. 17–25.e1. 24. Phinney RB, Mondino BJ, Hofbauer JD, et al. Corneal edema 13. Lachapelle JM. A comparison of the irritant and allergenic related to accidental Hibiclens exposure. Am J Ophthalmol properties of antiseptics. Eur J Dermatol 2014;24(1):3–9. 1988;106(2):210–215. 14. Lachapelle JM. Allergic contact dermatitis from povidone- 25. Auclin F, Rat P, Tuil E, et al. [Clinical evaluation of the iodine: a re-evaluation study. Contact Dermatitis 2005;52(1): ocular safety of Amukine 0.06% solution for local application 9–10. versus povidone iodine (Betadine) 5% solution for ocular irri- 15. Lee SK, Zhai H, Maibach HI. Allergic contact dermatitis gation in preoperative antisepsis]. J Fr Ophtalmol 2002;25(2): from iodine preparations: a conundrum. Contact Dermatitis 120–125. 2005;52(4):184–187. 26. Coskun M, Altintas AG, Anayol MA, Raza S, Celikbilek N, 16. Friedman DA, Mason JO 3rd, Emond T, McGwin G Jr. Povi- Simsek S. Evaluation of efficacy of topical povidone-iodine done-iodine contact time and lid speculum use during intra- and different types of fluoroquinolones in the sterilization of vitreal injection. Retina 2013;33(5):975–981. bacterial flora on the conjunctiva. J Ocul Pharmacol Ther 17. Ferguson AW, Scott JA, McGavigan J, et al. Comparison of 2011;27(6):589–592. 5% povidone-iodine solution against 1% povidone-iodine 27. Merani R, Hunyor AP. Endophthalmitis following intravi- solution in preoperative cataract surgery antisepsis: a prospec- treal anti-vascular endothelial growth factor (VEGF) injec- tive randomised double blind study. Br J Ophthalmol 2003; tion: a comprehensive review. Int J Retina Vitreous 2015; 87(2):163–167. 1(1):1–19. 18. Shimada H, Hattori T, Mori R, Nakashizuka H, Fujita K, 28. Ta CN, Sinnar S, He L, Myung D, Mino De Kaspar H. Yuzawa M. Minimizing the endophthalmitis rate following Prospective randomized comparison of 1-day versus intravitreal injections using 0.25% povidone-iodine irriga- 3-day application of topical levofloxacin in eliminating tion and surgical mask. Graefes Arch Clin Exp Ophthalmol conjunctival flora. Eur J Ophthalmol 2007;17(5): 2013;251(8):1885–1890. 689–695. 19. Pedrosa C, Costa H, Oliveira G, Romariz J, Praca F. Anaphy- 29. de Kaspar M, Kreutzer TC, Aquirre-Romo I, et al. A prospec- laxis to povidone in a child. Pediatr Allergy Immunol 2005; tive randomized study to determine the efficacy of preopera- 16(4):361–362. tive topical levofloxacin in reducing conjunctival bacterial 20. Gray PE, Katelaris CH, Lipson D. Recurrent anaphylaxis flora. Am J Ophthalmol 2008;145(1):136–142. caused by topical povidone-iodine (Betadine). J Paediatr Child 30. Kowalski RP, Romanowski EG, Shanks RM, Mah FS. The Health 2013;49(6):506–507. comparison of fluoroquinolones to nonfluoroquinolone 21. Michavila-Gomez AV, Moreno-Palanques MA, Ferrer- antibacterial agents for the prevention of endophthalmitis Vazquez M, Ferriols-Leisart R, Bartolome B. Anaphylactic in a rabbit model. J Ocul Pharmacol Ther 2012;28(6): reaction to povidone secondary to drug ingestion in a young 604–608. child. Allergol Immunopathol (Madr) 2012;40(4):259–261. 31. Lyall DA, Tey A, Foot B, et al. Post-intravitreal anti-VEGF 22. Lopez Saez MP, de Barrio M, Zubeldia JM, Prieto A, endophthalmitis in the United Kingdom: incidence, features, Olalde S, Baeza ML. Acute IgE-mediated generalized risk factors, and outcomes. Eye (Lond) 2012;26(12):1517–1526.

74 AMERICAN JOURNAL OF OPHTHALMOLOGY OCTOBER 2016