Introduction

Bevacizumab (Avastin®; Genentech, Inc., South San Francisco, CA, USA) is a full-length humanized monoclonal antibody against vascular endothelial growth factor-A (VEGF-A) [1].

Despite being off-label for intraocular injection, bevacizumab has been extensively used in the treatment of vascular leakage and neovascular diseases such as age-related macular degeneration (AMD), retinal vein occlusion (RVO), proliferative diabetic retinopathy (PDR), and diabetic macular edema (DME) [2–4]. Since 2005, after Rosenfeld et al. published the results of their pilot study, intravitreal injection of bevacizumab has been one of the most widespread minor procedures in ophthalmology worldwide [5–8].

Bevacizumab is injected directly into the vitreous cavity and is generally well-tolerated with a low overall complication rate [9, 10]. Nevertheless, intravitreal injection of bevacizumab is not entirely free of adverse effects, most of which are related to the administration procedure than to the drug per se [11]. Endophthalmitis following intravitreal injection, a rare but feared complication, has been reported to have an incidence of 0.02–0.05 % [9, 10]. In published data from the Comparison of AMD Treatments Trials (CATT) study, six cases of endophthalmitis (0.05 %) were observed after 10,957 intravitreal injections of anti-VEGF drugs [12]. Sustained elevation of intraocular pressure (IOP) and severe intraocular inflammation have been increasingly reported following IVB injection [13–17]. Although the exact cause is yet to be determined, it is believed that mishandling of the drug during storage or repackaging as well as allergic reactions might be involved [13, 14, 18]. Recently, several outbreaks of bacterial and fungal infections, or sterile inflammations, have raised concern about the safety of the drug-preparation process [19–24]. Included among these was a report on a series of 12 cases of Streptococcus endophthalmitis after intravitreal injection of bevacizumab [20] and a large series of 80 cases of culture-negative endophthalmitis after intravitreal injection of counterfeit bevacizumab [24].

In the present case series, we report a large outbreak of acute-onset bacterial endophthalmitis following intravitreal injection of counterfeit bevacizumab, and will describe the clinical characteristics and outcomes of 28 cases of endophthalmitis at the same time period at two hospitals in Tehran in which the administered drug came from two different batches.

Methods

Twenty-eight eyes of 21 patients presented with acute-onset bacterial endophthalmitis following IVB injection between February 24 and February 26, 2015, in two eye hospitals simultaneously. Medical records and results of microbiologic evaluation of these cases were retrospectively reviewed. This study was approved by the institutional review board of the Ophthalmic Research Center of Shahid Beheshti University of Medical Sciences, Tehran, Iran.

All injections were done using a sterile technique. In the first hospital, all injections were performed at the end of cataract procedures or vitrectomy from a previously unused single vial of bevacizumab. In the second hospital, all intravitreal injections were carried out in an operating room designated for intravitreal injections and again from a single, new vial. These two vials showed different batch numbers. Before the procedure, physicians scrubbed their hands and wore masks and sterile gloves. After cleaning the skin around the eye with a solution of povidone-iodine 10 % and instilling a single drop of povidone-iodine 5 % (Betadine; Alcon Laboratories, Inc., Fort Worth, TX, USA) in the cul-de-sac, a sterile lid speculum was placed. Topical proparacaine 0.5 % drops were used for local anesthesia. Prophylactic antibiotic eye drops were prescribed three times daily for 3 days post-injection.

The collected data included age, sex of each patient, the affected eye and its lens status, the date of the previous bevacizumab injection, the underlying diagnosis and the indication for IVB injection, and the pre-injection visual acuity. The presentation and diagnostic data for endophthalmitis were also recorded including the date of presentation, visual acuity and intraocular pressure, the presenting symptoms and signs, and the initial management as well as the microbiologic results. In addition, the clinical course of post-injection endophthalmitis was recorded, including visual acuity, routes of treatment with antibiotics and steroids, pars plana vitrectomy and other surgical interventions, as well as visual acuity and the status of the eye at the last follow-up examination.

To present the data, we used mean, standard deviation, frequency, and percent, with a 95 % confidence interval (95 % CI). To evaluate the change of best corrected visual acuity (BCVA) we used Wilcoxon's signed–rank test. All statistical analysis was performed by SPSS software (IBM Corp. Released 2013. IBM SPSS Statistics for Windows, Version 22.0. Armonk, NY: IBM Corp.) P-values less than 0.05 were considered as statistically significant.

Results

In total, 28 eyes of 21 patients were included in this study. Among them, 11 (52.4 %) were male and 10 (47.6 %) were female. Fourteen right eyes (50 %) and 14 left eyes (50 %) were involved. Of 28 eyes, nine (32.1 %) were infected with E. coli in the first hospital, and 19 eyes (67.9 %) were infected by Citrobacter in the second hospital. E. coli was cultured from the same batch that was used in the first hospital. Citrobacter was cultured from the same batch that was used in the second hospital. The infected vials caused endophthalmitis in all treated patients. All patients were followed for 3 months (Table 1).

Table 1 Specifications of all affected cases
Full size table

Symptoms and signs of a fulminant endophthalmitis was observed in nine eyes of eight patients after intravitreal injection of bevacizumab at the end of cataract or vitreous surgery in the operation room from a single batch in the first hospital, and 19 eyes of 13 patients from the other single batch in the second hospital within 12–48 (20.2 ± 9.7) hours following the IVB injection.

All patients were admitted with the diagnosis of endophthalmitis. Anterior chamber and vitreous tap was performed followed by intravitreal injection of vancomycin (1 mg) + ceftazidime (2.25 m g).Pars plana vitrectomy was performed on 24 (85.7 %) eyes. Topical steroid, fortified and systemic vancomycin, and ceftazidime were started for all patients; a systemic steroid was prescribed after 24 hours and continued for 10 days. Repeat vitrectomy and silicone oil injection was done for five eyes; in two of them it was performed in combination with penetrating keratoplasty due to severe keratitis. Conjunctival flap and amniotic membrane graft was performed in two other eyes for the prevention of corneal melting.

The mean BCVA was 1.27 ± 0.89 LogMAR before bevacizumab injection, 2.80 ± 0.45 LogMAR after presentation of endophtha,lmitis and 2.12 ± 0.97 LogMAR 3 months after intravitreal injections of bevacizumab. Final visual acuity was no light perception in four eyes. All eyes were saved without a need for enucleation or evisceration.

Discussion

The 3-month clinical outcomes in this largest outbreak of bacterial endophthalmitis following intravitreal injection of counterfeit bevacizumab have been poor. Only four patients (14.2 %) recovered useful vision (>1/10) in the affected eye. Previous studies have reported poor visual outcomes in post-intravitreal injection endophthalmitis, with 59–80 % of cases having hand-motions or worsened final visual acuity.10–25 The outcomes in this case series are consistent with the results of prior published reports [10, 25].

Our patients developed the acute and fulminant form of post-IVB injection endophthalmitis. Mean time between intravitreal injections of bevacizumab and presentation of endophthalmitis was 20.2 hours (less than 24 hours), which is unusual for an infectious endophthalmitis. The fulminant course of the disease and positive cultures in all cases and vials with severe endophthalmitis in all treated patients indicate a highly contaminated drug. Although there was no case of enucleation or evisceration in this outbreak, at three 3 months post-presentation, 16 eyes (57 %) had the visual acuity of hand-motions or worse. Final visual acuity was NLP in four (14.2 %) eyes and LP in six (21.4 %) eyes. The mean BCVA decreased significantly from 1.27 ± 0.89 LogMAR to 2.12 ± 0.97 LogMAR (p = 0.002) before and 3 months after intravitreal injections of bevacizumab, respectively. There were seven cases of bilateral endophthalmitis that were injected from a single batch. In bilateral cases, final visual acuity was NLP in both eyes of one patient and LP in both eyes of another patient, which was a disastrous outcome. All affected cases were treated with topical, intravitreal, and intravenous vancomycin and ceftazidime, but this treatment resulted in infection control only in four eyes (14.2 %); urgent PPV was required in 24 eyes (85.7 %) to control endophthalmitis (at post-IVB injection day 2). Despite reports about the effects of silicone oil as an antimicrobial or globe-salvaging agent, final visual outcomes in this outbreak, even in the silicone oil-filled eyes were poor [26–29].

Vascular endothelial growth factor (VEGF) inhibitors such as bevacizumab and ranibizumab are frequently used in the treatment of neovascular age-related macular degeneration (AMD), as well as retinal vascular diseases, including retinal vein occlusion (RVO) and diabetic retinopathy [30-33]. Endophthalmitis after intravitreal injection of anti-VEGF agents is rare and occurs between 1 in 2,000 and 1 in 5,000 injections according to the previously reported studies. Half of the cases are culture-negative, while those that are culture-positive most commonly grow Staphylococcus and Streptococcus species [9, 10]. A published data from the CATT study showed that ranibizumab and bevacizumab have similar efficacy in the treatment of neovascular AMD with comparable rates of endophthalmitis after intravitreal injection of these two drugs [12]. Reusing the same bevacizumab vial for multiple injections seems not to increase its frequency in comparison to single-use ranibizumab vials. To date, other large series found no difference in the endophthalmitis risk in patients receiving bevacizumab as compared to ranibizumab [12, 34-38]. Bevacizumab is used as an “off-label” drug; in most countries it is prepared by compounding pharmacies for intravitreal use. The involvement of a compounding pharmacy introduces an additional step between the drug manufacturer and the treating physician, and this additional step may result in increased risk of contamination of the medication. Most of the previously reported culture-positive outbreaks occurred secondary to repackaging in compounding pharmacies [39-41]. The Pan-American Collaborative Retina Study Group (PACORES) reported more frequent endophthalmitis (6/1833 injections, 0.33 %) in eyes injected using previously compounded aliquots than in eyes given injections from the same multidose vial (1/2470 injections, 0.04 %) that was reused appropriately [42]. A series of 12 patients in Florida treated with bevacizumab from a single compounding pharmacy developed endophthalmitis with a common strain of Streptococcus mitis/oralis [20].

In our country, ophthalmologists do not employ compounding pharmacies as a source for individual doses of bevacizumab. Alternatively, they draw the dose directly from the bottle and use it for about 15–20 eyes. Bottles are obtained from the distributors, which in turn claim to buy the drug from the manufacturer. Since the process involves the drug changing hands twice, it is vulnerable to the introduction of the counterfeit drug. Unfortunately, these hospitals had bought two counterfeited bevacizumab vials from invalid distributors.

Counterfeit medicines are fake products that are deliberately and fraudulently manufactured to be mistaken for legitimate drugs. They may be contaminated or contain the wrong or inactive ingredient. They could also have the right active ingredient but at the wrong dose. Some counterfeit bottles contain not just water, but ethanol, citrate, and polyethylene glycol, which are added to the content to inhibit microbial growth [43]. Because the markup on counterfeit drugs is so high, they have become attractive to organized crime groups.

Bevacizumab is distributed as a 4- or 16-mL preservative-free single-use vial. Counterfeit bevacizumab poses a significant risk for ophthalmology and oncology patients. An outbreak of endophthalmitis following the counterfeit bevacizumab has previously been reported in China. Eighty out of 116 patients that received intravitreal injections from three vials of counterfeit bevacizumab developed culture-negative endophthalmitis, and the endotoxin was identified in the vitreous samples [24].

To our knowledge this report is the largest culture-positive outbreak of endophthalmitis after intravitreal injection of counterfeit bevacizumab. In the present case series, the first counterfeit batch was contaminated with E. coli and the second one was contaminated with Citrobacter. E. coli and Citrobacter are Gram-negative bacteria in the Enterobacteriaceae family. These bacteria can be found almost everywhere in contaminated soil, water, wastewater, etc.

There is no current consensus on the preferred treatment of post-injection endophthalmitis. The Endophthalmitis Vitrectomy Study did not enroll patients with post-injection endophthalmitis, so its findings cannot be directly applied to these patients. Topical, intravitreal, and systemic antibiotic therapy in mild cases and PPV with or without silicone oil injection in severe cases have been reported as treatment of post-injection endophthalmitis [44].

In summary, in this outbreak 28 eyes were affected by virulent strains of E. coli or Citrobacter after receiving an intravitreal injection of contaminated counterfeit bevacizumab at the same time at two separate hospitals in Tehran. This outbreak, like in previously reported studies, confirmed that bilateral injections should be done from separate batches. According to the acute-onset course and poor outcomes of post-intravitreal injection endophthalmitis, we recommend prompt treatment with vitrectomy. Expensive drugs like bevacizumab are at risk of being replaced with counterfeit drugs, so vigilance is needed on the part of physicians, drug distributors, healthcare providers, and governments to prevent such a disastrous event in the future.