Vascular endothelial growth factor (VEGF) inhibitors are the current standard for treatment of cystoid macular edema (CME) associated with central retinal vein occlusion (CRVO). Post-injection uveitis or a sterile inflammatory response to both aflibercept (Eylea; Regeneron, Tarrytown, NY) and ranibizumab (Lucentis; Genentech, South San Francisco, CA) has been separately reported.1–5
Herein, we report two patients who developed sequential, sterile intraocular inflammatory responses associated with separate sequential injections of intravitreal aflibercept (2 mg/0.05 mL) and ranibizumab (0.5 mg/0.05 mL).
An 89-year-old pseudophakic, hypertensive female was undergoing anti-VEGF therapy for CRVO-associated CME in her left eye (24 prior aflibercept injections and six prior ranibizumab injections without adverse events). Following the twenty-fifth aflibercept injection in that eye, she developed decreased visual acuity to count fingers from her baseline of 20/32, photosensitivity, and pain without tenderness. Examination revealed trace conjunctival injection, 4+ anterior chamber cell and flare, 1 mm hypopyon, and 4+ vitreous haze. B-scan showed moderate vitreous opacities. A vitreous tap grew rare Propionibacterium acnes, which was deemed a contaminant.6 She received empiric intravitreal vancomycin (1 mg/0.1 mL) and ceftazidime (2.25 mg/0.1 mL) and was started on hourly topical prednisolone acetate 1% and a cycloplegic. One month following treatment, her inflammation resolved and vision improved to 20/64.
After a 10-week hiatus from intravitreal injections, she received intravitreal ranibizumab for recurrent CME. Five days later, she presented with a second episode of painless decreased vision to 20/100, trace conjunctival injection, 1+ anterior chamber, and 2+ vitreous cells. Anterior chamber tap, culture, and gram stain were negative. She received intravitreal vancomycin and was started on hourly topical prednisolone acetate 1% and a cycloplegic. After 4 weeks, her vision improved to 20/40, inflammation resolved, and there was minimal residual cellular debris in the anterior vitreous. The patient has since been observed without further injections and minimal recurrent CME.
A 56-year-old phakic, healthy male had received six prior ranibizumab injections for CME associated with a perfused CRVO in his right eye without adverse events and a vision of 20/20 with treatment. Three days after his seventh ranibizumab injection, he presented with pain, photosensitivity, and decreased vision to 20/200. Exam showed 3+ conjunctival injection, corneal haze, 4+ anterior chamber cell with fibrin, and a 3 mm hypopyon (Figure). B-scan demonstrated moderate vitreous opacities. He was treated empirically with intravitreal vancomycin and ceftazidime, hourly topical difluprednate, a cycloplegic, and oral prednisone. An aqueous and vitreous tap was deferred by the treating physician.
Slit-lamp photograph (A) of Case 2 presenting with diffuse conjunctival injection and a 3-mm layered hypopyon, 4+ cell, and fibrin in the anterior chamber, with B-scan (B) showing vitritis and moderate hyperechoic vitreous opacities 3 days after ranibizumab injection to the right eye. The patient, a 56-year-old phakic, healthy male, had received six prior ranibizumab injections for cystoid macular edema associated with a perfused central retinal vein occlusion in his right eye without adverse events. He was treated empirically with intravitreal vancomycin and ceftazidime, hourly topical difluprednate, a cycloplegic, and oral prednisone. Following an 8-week injection hiatus, he was switched to intravitreal aflibercept, following which he developed a second inflammatory response. Four weeks after similar treatment, the inflammation had resolved (C).
The inflammation resolved in 2 weeks, and vision improved to 20/80. Following an 8-week injection hiatus, he was switched to intravitreal aflibercept for recurrent CME. Eight days later, he presented with a second episode of inflammation with pain, redness, and decreased vision to 20/400. He was treated with intravitreal vancomycin, ceftazidime, and dexamethasone and started on hourly topical prednisolone acetate 1%, a cycloplegic, and oral prednisone. Four weeks later, the inflammation had resolved with improvement in vision to 20/126 and residual vitreous opacities. Anterior chamber tap grew a single colony of yeast, which was deemed a contaminant. He has not received further injections.
A sterile inflammatory response has been separately associated with intravitreal injections of bevacizumab (Avastin; Genentech, South San Francisco, CA), ranibizumab, and aflibercept.1–5,7–9 In this report, we describe two cases that were being treated for CRVO-associated CME and developed a sequential sterile inflammatory response to both aflibercept and ranibizumab following temporally separated injections and despite multiple prior injections without adverse effects. To the best of our knowledge, such a sequential sterile inflammatory response in the same patient with both ranibizumab and aflibercept, agents with different molecular structures, has not been previously described. Although the minimal inoculum of a low-grade pathogenic yeast noted in the culture results could conceivably contribute to these episodes, it is unlikely. Neither received antifungal treatment.
The immunologic stimulus may be different among the anti-VEGF drugs, and clinicians often switch to a different anti-VEGF agent following a sterile inflammatory response.10 Possible risk factors for sterile endophthalmitis are a history of uveitis, pseudophakia, and prior vitrectomy.8 There was no difference in intraocular inflammation assessed by anterior chamber flare following aflibercept or ranibizumab injections.11 Sterile endophthalmitis following aflibercept has a reported incidence ranging from 0.16% to 0.37%.2–4,12 Rates of sterile inflammation after prior injections with ranibizumab and bevacizumab range from 0.02% to 0.1%1,3,7,14 and up to 1.9% to 7.2% in some retrospective series with bevacizumab.15,16
Using post-marketing surveillance data, the American Society of Retina Specialists Therapeutic Surveillance Committee reported that aflibercept-related sterile inflammation had typical findings of loss of vision, redness (13%), pain (46%), or hypopyon (< 1%). There was no difference in outcomes among patients undergoing topical or intravitreal therapy, and the overall prognosis was favorable with resolution of inflammation usually within 4 weeks.5 Most cases had a history of prior aflibercept injections, and some cases were successfully rechallenged with aflibercept. In the phase 3 aflibercept studies, the pretreatment and 1-year post-treatment incidence of immunoreactivity to aflibercept was 1% to 3% across treatment groups, and there were no differences in efficacy or safety between patients with or without immunoreactivity.17 In the phase 3 ranibizumab trials, the pretreatment incidence of immunoreactivity to ranibizumab was 0% to 5% across treatment groups.18 After 6 months to 24 months of monthly dosing, this increased to approximately 1% to 9%, suggesting that multiple prior injections may increase the risk of immunogenic reactions.18
Various causative factors have been suggested for sterile inflammation including manufacturing contamination, patient-specific immune reactions, or physician-related procedural issues.5 Our investigative efforts did not reveal a specific cause for this sequential inflammatory response. Both ranibizumab and aflibercept are delivered in single-dose, sterile glass vials without preservatives and are drawn into an individual patient syringe at the time of the injection. Injections were performed with standard techniques, including an eyelid speculum, povidone-iodine before injection to both lids and conjunctiva, topical non-gel anesthesia, use of the syringe included in the aflibercept and ranibizumab packaging with a 30-gauge needle, and administration in the inferotemporal quadrant for all injections. Gloves were used for all injections. There was no sign of infection at the time of injection in either patient. The supply chain within the practice from delivery, internal distribution and storage, and day of use was intact; refrigeration conditions were assessed at each stage and for each office and found to be within the manufacturer's recommended range (36 °F to 46 °F).17 The lot numbers of both aflibercept and ranibizumab used in the two patients were different. Speculation regarding protein denaturation of aflibercept, perhaps owing to the more viscous nature of the molecule, compared to bevacizumab or ranibizumab has been suggested but not validated as a possible etiology.4 The only common component of ranibizumab (ranibizumab; 10 mg/mL in 10 mM histidine-HCl, 10% alpha, alpha-trehalose dihydrate, 0.01% polysorbate 20, pH 5.5) and aflibercept (40 mg/mL aflibercept in 10 mM sodium phosphate, 40 mM NaCl, 0.03% polysorbate 20, 5% sucrose, pH 6.2) is polysorbate-20, which is unlikely to elicit ocular hypersensitivity but a postinjection toxic reaction to this excipient cannot be excluded.
In conclusion, although several questions remain unanswered, clinicians should be aware that it is possible to develop a sequential inflammatory response to consecutive injections of ranibizumab and aflibercept. Both patients had received multiple prior anti-VEGF injections uneventfully and, unique to this report, developed a sequential inflammatory response with both ranibizumab and aflibercept. Distinguishing sterile inflammation and infectious endophthalmitis at the time of presentation may often be difficult, and cautious evaluation and management of these patients is warranted. It may be prudent to assume an infectious cause and consider empiric treatment for such cases.2,5 Long-term monitoring for inflammatory events that may represent cross-reactivity to anti-VEGF agents might help in understanding the complex proteomic milieu of such episodes of intraocular inflammation.