Endophthalmitis Following Intravitreal Injection of Anti-VEGF Agents

Although rare, infection remains a concern.

STEPHEN G. SCHWARTZ, MD, MBA • HARRY W. FLYNN, JR., MD • INGRID U. SCOTT, MD, MPH

The marked increase in the use of anti-VEGF agents for posterior-segment diseases has resulted in a corresponding increase in intravitreal injections over the past decade.¹ Perhaps the most serious complication following intravitreal injections is infectious endophthalmitis, which can lead to severe visual loss, even with prompt and appropriate therapy (Figure 1, page 54).²

Figure 1. Anterior-segment photograph, left eye, of a patient who developed endophthalmitis following an intravitreal ranibizumab injection. The cultures isolated coagulase-negative Staphylococci.

Endophthalmitis following intravitreal anti-VEGF injection has several characteristics distinct from other categories of endophthalmitis. Compared to endophthalmitis following cataract surgery, endophthalmitis following intravitreal injection generally presents earlier, is more likely caused by Streptococcus species, and has generally poorer visual outcomes.³

VARYING RATES REPORTED

Multiple large series have reported the per-injection endophthalmitis risk to be 0.03% or less (Table 1).^4-11 Similarly, a population-based study estimated a per-injection endophthalmitis risk of 0.025%.¹²

However, a review of the CMS claims database reported a per-injection endophthalmitis rate of 0.09%, suggesting that the true incidence may be higher than the rates reported in clinical series.¹³

Most patients undergoing anti-VEGF therapy receive a series of injections over months or years. Rates of endophthalmitis per injection, therefore, will necessarily be lower than the cumulative rates of endophthalmitis per treated eye over the course of therapy.

The prospective CATT study reported per-patient endophthalmitis rates of 0.7% with ranibizumab (Lucentis, Genentech, South San Francisco, CA) and 1.2% with bevacizumab (Avastin, Genentech) over two years; this difference was not statistically significant.¹⁴

The prospective Randomised Controlled Trial of Alternative Treatments to Inhibit VEGF in Age-Related Choroidal Neovascularization (IVAN), did not specifically report endophthalmitis but reported “severe uveitis” in one of 610 patients at one year (0.16%).¹⁵

COMPOUNDING THE PROBLEM

Ophthalmologists use intravitreal bevacizumab off-label, and compounding pharmacies generally prepare the intravitreal doses, incurring additional risks of endophthalmitis (Table 2).

A series of eight patients treated with a combination of compounded bevacizumab and triamcinolone acetonide developed fungal endophthalmitis with Bipolaris hawaiiensis.¹⁶ A series of 12 eyes treated with intravitreal bevacizumab developed endophthalmitis with a common strain of Streptococcus mitis/oralis, with generally poor outcomes (Figure 2, page 56).¹⁷

Figure 2. Anterior-segment photograph, right eye, of a patient who developed endophthalmitis following an intravitreal bevacizumab injection that was contaminated with Streptococcus mitis/oralis.

Noninfectious intraocular inflammation may also follow intravitreal injections, and this condition may be very difficult to distinguish from infectious endophthalmitis. One series reported culture-negative endophthalmitis in 14 consecutive patients treated with a single batch of bevacizumab.¹⁸

Another series reported that six of eight patients treated with bevacizumab from the same vial developed culture-negative inflammation.¹⁹ A third series reported culture-negative endophthalmitis in four of 15 eyes treated with one batch of bevacizumab and seven of 18 eyes from a second batch.²⁰

Eighty patients of 116 injected from three vials of counterfeit bevacizumab developed culture-negative endophthalmitis.²¹ In a review of 1,584 injections, severe intraocular inflammation was 12 times more common after treatment with bevacizumab than after treatment with ranibizumab.²²

Cunningham et al reported sequential noninfectious endophthalmitis in one patient following treatment with bevacizumab and then following treatment with ranibizumab.²³

RISK REDUCTION

Multiple strategies have emerged to reduce the risk of endophthalmitis after anti-VEGF injections. It is not currently possible to eliminate the risk of postinjection endophthalmitis, which should not be considered a “never event.”²⁴

However, we can minimize the risk by applying established guidelines for injection. Traditionally, the use of an eyelid speculum was recommended, but a recent review of 10,164 consecutive injections performed without an eyelid speculum, using an assistant’s fingers to retract the lids manually, reported only three cases of endophthalmitis (0.030%).⁸

Oral Flora and Facemasks

Oral flora appear to represent a relatively greater risk factor for endophthalmitis following intravitreal injection than other categories of endophthalmitis. In a series of 88 cases of postsurgical endophthalmitis and 21 cases of postinjection endophthalmitis, the study authors isolated Streptococcus viridans in a nonsignificantly higher proportion of postinjection patients (14.3% vs 4.5%).²⁵

A meta-analysis of 52 endophthalmitis cases out of 105,536 injections (0.049%) reported that Streptococcus isolates comprised 30.8% of culture-positive cases, which was significantly greater than the proportion of Streptococcus isolates reported in multiple published series of postsurgical endophthalmitis.²⁶

The relatively more important role of oral flora, including Streptococcal species, in endophthalmitis following intravitreal anti-VEGF injections raises the question of whether facemasks may be helpful in reducing the incidence of infection. Although some physicians use facemasks, no randomized clinical trials have evaluated their use.

In two small prospective trials using simulated intravitreal injections, bacterial dispersal significantly decreased with the wearing of a facemask or by not speaking.^27,28 Notably, one series of 15,144 injections performed using facemasks reported zero cases of endophthalmitis.⁹

Antibiotic Prophylaxis

In the early years of anti-VEGF injections, physicians commonly employed pre- or postinjection topical antibiotics, but this practice has declined in recent years. Several groups have reported that antibiotic use does not reduce rates of endophthalmitis and may actually increase the endophthalmitis risk.

In a series of 5,901 treated eyes, 68% were treated with antibiotics, and six endophthalmitis cases (0.10%) emerged, all in antibiotic-treated eyes.¹⁰ In a series of 15,895 anti-VEGF and triamcinolone injections, the authors reported higher rates of endophthalmitis in antibiotic-treated patients.²⁹

The Diabetic Retinopathy Clinical Research Network reported 8,027 intravitreal injections with a nonsignificantly higher incidence of endophthalmitis with the use of topical antibiotics.³⁰ A recent series of 117,171 injections reported a nonsignificantly higher rate of endophthalmitis in antibiotic-treated eyes.¹¹

The Emergence of Resistance?

If antibiotics do increase the risk of endophthalmitis following intravitreal anti-VEGF injections, then the explanation for this increase is unknown. Repeated use of topical antibiotics following repeated intravitreal injections may alter the conjunctival flora.

In a prospective nonrandomized study, the authors treated 84 patients with topical moxifloxacin (Vigamox, Alcon, Fort Worth, TX) after monthly injections and 94 patients were not treated with antibiotics; the treated patients showed statistically significant increases in minimal inhibitory concentrations of ocular surface flora, indicating increased resistance to moxifloxacin and ceftazidime (Fortaz, GlaxoSmithKline, Brentford, United Kingdom).³¹

Alternatively, another prospective study of 104 patients reported that repeated exposure to topical fluoroquinolones did not alter baseline fluoroquinolone resistance in conjunctival and nasal flora. ³²

A prospective series of 13 patients reported that repeated use of topical povidone-iodine 5% alone, without postinjection topical antibiotics, did not alter the conjunctival flora or promote bacterial resistance.³³

Notably, a prospective study of 131 eyes reported that topical povidone-iodine 5% did not cause a statistically significant reduction in the number of conjunctival bacterial colony-forming units after 15 seconds of exposure, but it did cause a statistically significant reduction after 30 seconds of exposure.³⁴

A Change of Venue?

Several investigators have studied the use of an operating room, rather than a clinic, for injections. One series of 12,249 injections reported a significantly lower endophthalmitis rate when the injections were performed in an operating room.³⁵ Similarly, a second series of 20,293 anti-VEGF injections performed in an operating room reported zero infections.³⁶ In contrast, a third series of 11,710 injections reported a nonsignificantly higher endophthalmitis rate in the operating room than in the clinic.³⁷

Following some intravitreal injections, a small amount of liquid may reflux back under the conjunctiva, creating a temporary bleb. It is unknown whether the fluid in this bleb represents medication or vitreous, but one study reported that vitreous reflux was a risk factor for endophthalmitis.³⁸ An animal study using radiolabelled anti-VEGF agents demonstrated that the refluxed material was most likely medication and not vitreous.³⁹

RECOMMENDATIONS

At this time, no consensus exists regarding the preferred management of endophthalmitis following intravitreal anti-VEGF injections. The Endophthalmitis Vitrectomy Study did not enroll patients with postinjection endophthalmitis, and its findings cannot be directly applied to these patients.

In a series of 23 patients with endophthalmitis following intravitreal anti-VEGF injection, vitreous tap and inject was associated with more favorable treatment outcomes than pars plana vitrectomy.⁴⁰

Given this information, certain recommendations become evident (Box, page 55). Retina specialists should continue to inform their patients that endophthalmitis remains a serious, and potentially vision-threatening, risk following intravitreal anti-VEGF injections.

The per-injection endophthalmitis risk, based on multiple retrospective and prospective studies, appears to be in the range of 0.03%, or 1:3,000.

Facemasks may be beneficial, although discouraging talking may be similarly helpful. Although facemasks pose limited additional cost and essentially no additional risks, consensus is lacking regarding their use, and they should not be considered standard practice at this time.⁴¹

In contrast, it is becoming increasingly apparent that topical antibiotics do not reduce the rate of endophthalmitis following intravitreal anti-VEGF agents and that antisepsis, rather than topical antibiotics, is preferable.^42-44

Traditionally, the use of an eyelid speculum was recommended. A recent large series reporting no apparent increase in the rate of endophthalmitis without the use of an eyelid speculum⁸ is intriguing and merits further study. However, at this time we continue to use speculums.

Bevacizumab, generally prepared by a compounding pharmacy, is associated with additional risks. A root-cause analysis has suggested using a single vial of bevacizumab for each eye, which substantially increases costs, or strictly adhering to US Pharmacopeia Chapter 797 requirements to use one vial for multiple injections,⁴⁵ which may not be practical outside the United States.⁴⁶

CONCLUSION

By following evidence-based guidelines, the rate of endophthalmitis following intravitreal anti-VEGF injections may be reduced, but not eliminated. Despite precautions, some patients will still develop endophthalmitis, which may progress to severe visual loss despite prompt and appropriate therapy. RP

REFERENCES