To evaluate the outcome of intravitreal injection of aflibercept in cases with exudative age-related macular degeneration, (AMD) resistant to injections of bevacizumab or ranibizumab.
Retrospective observational case series.
A retrospective chart review at a single institution was conducted to identify patients with exudative AMD and choroidal neovascularization (CNV) in 1 or both eyes resistant to treatment with ranibizumab or bevacizumab who were switched to treatment with at least 3 monthly injections of aflibercept. In total, 36 eyes from 31 patients were included. The demographic data, visual acuities, central macular thickness on optical coherence tomography (OCT), complications, and number of injections were reviewed.
The mean patient age was 79 years (range 60-88). There were 13 male and 18 female patients. The number of prior injections with either bevacizumab or ranibizumab ranged from 6-74. After 3 monthly injections of aflibercept, there was a reduction of either subretinal or intraretinal fluid in 18 of 36 (50.0%) of the treated eyes; the amount of fluid remained stable in 15 eyes (41.7%) and worsened in 3 eyes (8.3%). A significant average decrease was observed for the central macular thickness after 3 injections of 65 μm ( P = 2.9 × 10 -6 ), with no significant change in visual acuity.
Aflibercept therapy appears to be beneficial in a subset of patients with neovascular age-related macular degeneration who exhibit recurrent or resistant intraretinal or subretinal fluid following multiple injections with either bevacizumab or ranibizumab.
Age-related macular degeneration (AMD) is the most common cause of irreversible blindness in the industrialized world. Two landmark clinical trials using intravitreal injections of the humanized monoclonal antibody fragment ranibizumab (Lucentis) in eyes with new-onset neovascular AMD showed that over 90% of treated eyes lost fewer than 15 letters after 24 months of follow-up and the visual acuity could be improved in 33.3%-40%. This improvement was associated with the drug’s ability to reduce intraretinal edema, subretinal fluid, and hemorrhage. A recent randomized, multicenter, 2-year clinical trial sponsored by the National Institutes of Health revealed that bevacizumab, a humanized murine antibody to vascular endothelial growth factor (VEGF), was noninferior compared with ranibizumab for the treatment of exudative AMD. Bevacizumab and ranibizumab injections have had a significant impact on preserving vision and preventing blindness.
Some patients, however, have a good initial response with resolution of fluid, but then later become resistant to further treatment and develop recurrent exudation with vision loss. The mechanism of this resistance to treatment with these drugs is not known, but one possibility is tolerance or tachyphylaxis, manifested by a decreased response over time to repeated treatment with a medication. Tachyphylaxis sometimes can be reversed by increasing the dose or halting therapy for a period of time before reinstating the same treatment.
Tachyphylaxis has been suggested in patients treated for exudative AMD. Some eyes have shown better resolution of fluid after switching anti-VEGF treatment. Tachyphylaxis has been proposed for a number of other treatments, including brimonidine and infliximab. In a study performed at the National Eye Institute, Forooghian and associates observed 6 eyes that developed tachyphylaxis after repeated bevacizumab injections. Gasperini and associates reported a significant decrease of subretinal fluid in the majority of tachyphylactic patients after switching treatments from either ranibizumab or bevacizumab to the other medication. The study showed that 81% of the patients responded favorably after changing intravitreal injection medication.
Aflibercept is a novel VEGF inhibitor, with a high affinity for VEGF. It is a protein constructed by fusion of the second binding domain of the receptor VEGFR1 and the third binding domain of the receptor VEGFR2 to the crystalline portion of IgG1. This has resulted in aflibercept’s having a significantly higher affinity in vitro to VEGF compared with both bevacizumab and ranibizumab The intravitreal half-life of aflibercept is 4.7 days in rabbit eyes, which is longer than ranibizumab (2.9 days) and comparable with bevacizumab (4.3 days). The combined high affinity and longer half-life has led to a calculated duration of effect of a single intravitreal injection of 2 mg aflibercept of 48-83 days. Monthly treatment with aflibercept has been shown to improve the vision in exudative AMD in 2 clinical trials. The treatment with 3 initial monthly injections of 2 mg aflibercept, followed by injections every 8 weeks, has been shown to be noninferior compared with monthly treatments with either aflibercept or ranibizumab.
Bevacizumab is a humanized murine antibody to VEGF and ranibizumab is a humanized antigen-binding fragment to VEGF, whereas aflibercept is a fusion protein receptor with much higher binding affinity to VEGF than both bevacizumab and ranibizumab. Therefore, an important question is whether eyes that have become resistant to anti-VEGF molecules respond to aflibercept injections. The purpose of this study is to help answer this question by reporting on an initial cohort of eyes that became resistant to treatment with bevacizumab or ranibizumab and then were switched to injections of aflibercept.
A retrospective observational case series was performed to study the outcome in eyes with exudative AMD, resistant to the treatment of monthly injections with bevacizumab or ranibizumab, that were switched to at least 3 monthly injections of aflibercept therapy. A written consent was obtained from the patients prior to each injection treatment. The Institutional Review Board for human subjects research at the University of Iowa approved the study protocol for human subjects. The study was compliant with HIPAA (the Health Insurance Portability and Accountability Act of 1996) and adhered to the tenets of the Declaration of Helsinki. Charts were reviewed for patients treated in the Vitreoretinal Clinic at the University of Iowa Department of Ophthalmology and Visual Sciences. Intravitreal injections were performed using the same procedure in all patients, as previously described. Inclusion criteria were age >60 years and a diagnosis of exudative AMD, including presence of drusen, as well as pigment epithelial changes in combination with choroidal neovascularization, confirmed by fluorescein angiogram and optical coherence tomography (OCT) at the initial clinic visit. The criteria also included only eyes that had been previously injected with either 1.25 mg bevacizumab or 0.5 mg ranibizumab (Avastin and Lucentis, respectively; Genentech Inc, San Francisco, California, USA) and had an initial response (decrease of retinal edema and subretinal fluid), followed by a recurrent increase or persistent subretinal fluid or retinal edema on OCT. The intraretinal or subretinal fluid had to be present for a minimum of 3 months prior to the first aflibercept injection. This fluid had to be refractory to continued monthly intravitreal injections with the same drug or a change from either bevacizumab or ranibizumab to the other drug with 3 monthly injections administered the 3 months prior to the first aflibercept injection. The criteria for treatment with aflibercept were the same as retreatment criteria for bevacizumab or ranibizumab regarding presence of intraretinal or subretinal fluid. Finally, the last criterion was that at least 3 initial 2-mg aflibercept injections every 4-6 weeks had been administered, followed by a visit within 4-8 weeks after the third aflibercept injection. Patient charts were reviewed until 6 months after the first aflibercept injection. At each visit the visual acuity was measured using a standardized Snellen chart and a comprehensive eye examination, and OCT imaging was performed using Spectralis OCT (Heidelberg Engineering, Carlsbad, California, USA) or Cirrus HD-OCT (Carl Zeiss Meditech, Dublin, California, USA). With Spectralis, 20 × 15-degree volume scans containing 19 sections separated by 240 μm were used. The scans were acquired in high-speed mode. The central macular thickness (CMT) was equivalent according to the distance between the 2 algorithm lines. The Spectralis algorithm for placement of the segmentation line in an OCT with a vascular pigment epithelial detachment (PED) has been demonstrated to exhibit a significant difference compared with manual measurement of the CMT. Therefore, the CMT, central retinal thickness (CRT), and PED were also measured using the caliper function for the section corresponding to the foveal depression. CMT was measured from the foveal depression to the Bruch membrane; alternatively, where the PED obscured the outer retinal structures, it was measured to the estimated location of the Bruch membrane. CRT was measured from the foveal depression to the inner part of the PED along the CMT measurement line. The PED was calculated as the difference between the CMT and the CRT. The same numerical scan for each individual was analyzed at the different time points. With Cirrus, the 512 × 128 scan pattern was performed consisting of 128 horizontal lines of 512 A-scans per line.
The collected demographic data included patient age; sex; laterality of the treated eye; number of ranibizumab, bevacizumab, and aflibercept injections; the time between injections; and the response to treatment as determined by the OCT and visual acuity. The CMT on OCT measurements were collected from the visit of the first treatment visit with aflibercept and until 6 months after the first aflibercept injection. An assessment of the OCT appearance was performed for the images from the selected follow-up visits and compared with the OCT from the visit of the first aflibercept injection. A definite (qualitative visible) change in macular thickness or subretinal or intraretinal fluid was graded as improved, stable, or worsened, similarly as performed in a previously published study of the OCT response after switching anti-VEGF treatment. The visual acuity was collected from the first visit of any anti-VEGF injection, at the visit of the first aflibercept injection, at the visit after the third injection, and at the visit 6 months after the first aflibercept injection. The Snellen visual acuity was converted to logMAR units. Measurements for the eyes with a change in treatment regimen or eyes that were lost to follow-up after the third injection, but before the 6-month visit, were not included in data collection for the 6-month visit.
The occurrence of any severe postoperative complications, including infection, inflammation, arteriothrombotic events, or death within the 6 months after the first aflibercept injection was recorded. The significance of the difference of group means was tested with Student t test (2-tailed, equal variance).
A total of 36 eyes from 31 patients met the inclusion criteria. The mean patient age was 79 years (range 60-88). There were 13 men and 18 women. Of the 36 eyes, 22 were the right eye and 14 were the left eye. The study eyes had received an average of 25.6 injections (range 6-74) prior to treatment with aflibercept, with a mean of 15.5 bevacizumab and 12.8 ranibizumab treatments. Thirty-two eyes had at least 1 trial of switching between bevacizumab and ranibizumab during the period from the first anti-VEGF injection until the first aflibercept treatment, whereas 3 eyes received only bevacizumab and 1 eye received only ranibizumab ( Table 1 ). During the 6 months just prior to the change of treatment, an average of 5.2 injections (range 4-6) had been administered to the study eye. All eyes had received a series of 3 aflibercept injections given 4-6 weeks apart and the interval thereafter was extended at the discretion of the treating retina specialist. All 36 eyes had data collected for the visit after the third aflibercept injection. There were 27 eyes that continued aflibercept injections until at least 6 months after the first injection. The aflibercept treatment was stopped in 9 eyes prior to 6 months after the first injection. Four eyes in 4 patients were switched because of patient co-pay cost after the fifth injection, including 1 eye that in addition had increased subretinal fluid and worsened vision; 1 eye in 1 patient was lost to follow-up since the patient switched providers after the fifth injection; 1 patient was hospitalized after the fifth injection because of renal failure and was unable to return to the clinic for continued injections in the studied eye; 1 eye was treated with photodynamic therapy after the fourth injection because of persistent subretinal fluid; 1 eye had a change in treatment after the fifth injection because of patient preference; and 1 eye developed endophthalmitis after the fifth injection with coagulase-negative Streptococcus . The infected eye underwent vitrectomy and antibiotic injections, with the vision subsequently returning to 20/40, same as prior to the onset of aflibercept injections.
|Eye No.||Age/Sex||Eye||Total BZB or RZB Prior to AFL||BZB||RZB||CMT (μm) at Switch to AFL||CMT (μm) After AFL ×3||CMT Change (μm) at 3 Months||CMT (μm) After 6 Months.||CMT Change (μm) at 6 Months|
|P value a||2.9 × 10 -6||5.2 × 10 -6|
The manually measured, average CMT for the OCT at the time of switching therapies was 410 μm (range 174-1027, median 387). The average CMT at the visit after the third aflibercept injection was 346 μm (range 155-784, median 315), and after 6 months the average CMT was 296 μm (range 151-528, median 282). Compared with the CMT at the initial visit, there was a significant decrease in CMT after 3 months of 65 μm ( P = 2.9 × 10 -6 ) and after 6 months of 87 μm ( P = 5.2 × 10 -6 ) ( Table 1 ). This suggests that aflibercept therapy improved CMT in previously unresponsive eyes.
Using the standard Spectralis algorithm, the mean CMT on OCT prior to the first injection of aflibercept was 358 μm (range 181-579, median 335). The average CMT at the first visit after the third injection was 309 μm (range 162-623, median 274) and after 6 months was 298 μm (range 168-561, median 224). Compared with the visit at the first aflibercept injection, there was a significant average decrease in CMT after 3 injections of 48 μm ( P = 1.6 × 10 -4 ) and a decrease of 60 μm ( P = 9.1 × 10 -6 ) after 6 injections.
In 28 of the 36 eyes, a PED was observed at the visit of the first aflibercept injection. Manual measurement of the CRT and calculation of the PED were performed in the 28 eyes with a PED. At the visit for the first aflibercept injection, the average CRT was 295 μm, significantly decreasing by 45 μm, to 250 μm ( P = 2.0 × 10 -6 ), at 3 months and decreasing by 77 μm, to 220 μm ( P = 1.1 × 10 -5 ), at 6 months. The average PED was 153 μm at the switch, decreasing by 26 μm ( P = .015), to 127 μm, at 3 months and decreasing by 13 μm ( P = .14), to 101 μm, after 6 months ( Table 2 ). This indicates that aflibercept therapy reduced PED thickness after 3 months, but there were great fluctuations for the eyes observed at 6 months.
|Eye No.||CRT at Switch||CRT After 3× AFL||Change After 3× AFL||CRT After 6 Months||Change After 6 Months||PED at Switch||PED After 3× AFL||Change After 3× AFL||PED After 6 Months||Change After 6 Months|
|P value a||2.0 × 10 -6||1.14 × 10 -5||.015||.14|