To determine 3-year treatment outcomes after 1 to 3 years of ranibizumab or bevacizumab therapy using a treat-and-extend regimen in patients with neovascular age-related macular degeneration (AMD).
Retrospective, interventional, consecutive case series.
We treated 212 eyes from 196 patients diagnosed with treatment-naive neovascular AMD between January 2009 and March 2013; they were treated with either ranibizumab or bevacizumab for a minimum of 1 year, using a treat-and-extend regimen. The main outcome measures were change from baseline best-corrected Snellen visual acuity (BCVA), proportion of eyes losing <3 BCVA lines, proportion of eyes gaining ≥3 BCVA lines, change from baseline central retinal thickness, and mean number of injections at 1, 2 and 3 years of follow-up.
The mean follow-up period was 1.88 years (median, 2 years). At baseline, mean BCVA was 20/139; it improved to 20/79 ( P < 0.001) after 1 year of treatment and was maintained at 20/69 and 20/64 at 2 and 3 years follow-up ( P < 0.001), respectively. At baseline, mean central retinal thickness was 351 μm and significantly decreased to 285 μm, 275 μm and 276 μm at 1, 2 and 3 years of follow-up ( P < 0.001), respectively. Patients received, on average, 7.6, 5.7 and 5.8 injections over years 1, 2 and 3 of treatment, respectively. At final follow-up, 94% of eyes had lost <3 lines BCVA, and 34.4% of eyes had gained ≥3 lines BCVA.
The treat-and-extend regimen is effective in achieving and maintaining visual and anatomic improvements in patients with neovascular AMD for up to 3 years of treatment.
Antivascular endothelial growth factor (VEGF) injections were first established as the primary treatment for neovascular age-related macular degeneration (AMD) in the Minimally Classic/Occult Trial of Anti-VEGF Antibody Ranibizumab in the Treatment of Neovascular AMD (MARINA) and the Anti-VEGF Antibody for the Treatment of Predominantly Classic Choroidal Neovascularization in AMD (ANCHOR) trials. The Comparison of AMD Treatments Trials (CATT) and the Inhibit VEGF in Age-related Choroidal Neovascularization (IVAN) studies revealed similar visual outcomes in patients with neovascular AMD treated with either ranibizumab or bevacizumab over 2 years. An inherent limitation of a monthly regimen is that patients are subjected to monthly examinations, testing (eg, optical coherence tomography [OCT] scans) and intravitreal injections, which raise long-term economic and safety concerns.
Alternative management algorithms such as quarterly injections demonstrated inferior visual acuity gains compared to monthly injections. Individualized treatment regimens were also evaluated with patients’ receiving injections only if examination or OCT findings revealed choroidal neovascularization (CNV) activity. The Prospective OCT Imaging of Patients with Neovascular AMD Treated with Intraocular Lucentis (ranibizumab) (PrONTO) study revealed that an as-needed, or pro re nata (PRN), treatment algorithm achieved good visual acuity gains with reduced injection frequency, but patients are still subjected to monthly evaluations. Subsequently, the CATT and IVAN studies showed that the visual gains in an as-needed regimen were not equivalent to a monthly regimen at the end of 2 years of follow-up.
The treat-and-extend regimen allows for individualized anti-VEGF treatments. Prior studies have suggested that this algorithm may have the socioeconomic benefit of potentially limiting the number of injections, office visits and ancillary tests. Currently, the treat-and-extend regimen is the most commonly employed treatment approach in the United States, according to the American Society of Retina Specialists 2013 Membership Survey: Preferences and Trends, with approximately 77.8% of retina specialists favoring this style of treatment for neovascular AMD.
Although prior published studies have demonstrated good visual outcomes in a cost-effective manner with the treat-and-extend regimen, using either bevacizumab or ranibizumab, the studies are limited in terms of numbers of patients or follow-up. The purpose of our study was to determine 3-year visual and anatomic outcomes in patients with neovascular AMD using a treat-and-extend regimen involving bevacizumab and ranibizumab.
Patient Selection and Diagnosis
This study was a retrospective, consecutive case series that received approval from the institutional review board at Wills Eye Hospital. Billing records for patients diagnosed with neovascular AMD between January 2009 and March 2013 were identified by using the International Classification of Diseases-9 code 362.52. Patients’ charts were then reviewed for age, gender, date of diagnosis of neovascular AMD, best (available spectacle with/without pinhole) corrected (Snellen) visual acuity (BCVA) at each encounter, medical and ocular histories, anti-VEGF agent received, number of injections, and longest duration of successful extension. In addition, either Heidelberg spectral-domain optical coherence tomography (OCT) scans (Heidelberg Engineering; Heidelberg, Germany) or RTVue-100 SD-OCT scans (Optovue, Freemont, California, USA) were reviewed at the time of diagnosis and at 1-year intervals to document the presence or absence of intraretinal/subretinal fluid and to determine central retinal thickness. Due to variations in patients’ individual follow-up periods, spectral-domain OCT scans were evaluated within 6 weeks of the yearly target dates.
Patients were included in the study if they had treatment-naive AMD, had a minimum of 1 year of follow-up using a treat-and-extend regimen, and received treatment with either bevacizumab or ranibizumab. Patients were excluded if they had any 1 of the following: (1) concomitant retinal diagnoses, such as diabetic retinopathy, retinal vein occlusion or myopic degeneration; (2) Anti-VEGF treatment regimen other than treat-and-extend; (3) intravitreal injections other than bevacizumab or ranibizumab (specifically, patients treated with aflibercept were excluded because of limited follow-up and the lack of head-to-head studies comparing the efficacy and durability of aflibercept to those of ranibizumab and bevacizumab); (4) any prior treatments for neovascular AMD, such as laser photocoagulation or photodynamic therapy; and (5) vitrectomy surgery in the study eye.
All patients were evaluated at initial diagnosis using slit-lamp biomicroscopy and spectral-domain OCT. Patients were then treated with monthly (4–5 weeks) intravitreal injections of either bevacizumab (1.25 mg/0.05 mL; Avastin; Genentech, San Francisco, California, USA) or ranibizumab (0.5 mg/0.05 mL; Lucentis; Genentech, San Francisco, California, USA) until no signs of CNV activity were detected on slit-lamp biomicroscopy and spectral-domain OCT. (Signs of CNV activity include the presence of intraretinal or subretinal fluid according to spectral-domain OCT or hemorrhage in the macula.) Patient follow-ups and treatments were then extended by intervals of 2 weeks as long as no signs of CNV activity were present. However, if any signs of exudation (intraretinal/subretinal fluid or macular hemorrhage) were detected, treatment intervals would be subsequently shortened by 2-week intervals. Anti-VEGF treatment was administered at every visit, regardless of CNV activity.
Patients’ BCVAs were converted into logarithm of the minimum angle of resolution (logMAR) score for statistical analysis. Data for continuous variables were recorded as mean ± standard deviation (SD). Paired 2-tailed t test analysis with significance set at P < 0.05 was used to compare mean data points at baseline with years 1, 2 and 3 of follow-up and was carried out using GraphPad Software (GraphPad, La Jolla, California, USA). BCVA was also converted into approximate Early Treatment Diabetic Retinopathy Study (approx ETDRS) scores to determine the mean gain in letters from baseline at years 1, 2 and 3. The mean maximum period of successful treatment extension at each year of follow-up and the mean number of yearly injections were also calculated. Categorical variables were reported as proportions. They included data concerning spectral-domain OCT characteristics such as the presence of intraretinal or subretinal fluid, in addition to the proportion of eyes losing <3 BCVA lines and the proportion of eyes gaining ≥3 BCVA lines at yearly intervals.
A total of 212 eyes from 189 patients (123 female, 66 male) with treatment-naive neovascular AMD were included. The minimum follow-up of 1 year was completed by 212 eyes (100%), with 121 eyes (57%) and 59 eyes (27.8%) completing 2 and 3 years of follow-up, respectively. Similarly, spectral-domain OCT characteristics were obtained for 212 eyes (100%) at 1 year follow-up, and 121 eyes (57%) and 59 eyes (27.8%) provided spectral-domain OCT data at 2 and 3 years of follow-up, respectively. The mean age of included patients was 80.3 ± 9.0 years. Baseline mean BCVA was 0.832 ± 0.57 (Snellen equivalent, 20/136; median, 20/80), with a corresponding mean central retinal thickness of 350.8 ± 116.8 μm (median, 322.5 μm) ( Table ). The mean follow-up period was 1.88 years (median, 2 years).
|Baseline Characteristics||Treat-and-Extend Regimen (n = 212 eyes) a|
|Mean ± SD||80.3 ± 9.0|
|Median (min, max)||82 (54, 99)|
|Gender (n = 189)|
|Median (min, max)||20/80 (20/23, HM)|
|Mean ± SD||350.9 ± 116.8|
|Median (Min, Max)||322.5 (145, 869)|
a At baseline, 212 eyes were included in the study. At 1, 2 and 3 years of follow-up, 212, 121 and 59 eyes, respectively, were included.
Visual Acuity and Spectral-Domain Optical Coherence Tomography Outcomes at Years 1, 2 and 3
Mean visual acuity significantly improved at years 1, 2 and 3 compared to baseline. After 1 year of treatment, mean BCVA was 0.601 ± 0.47 (Snellen equivalent 20/79; median, 20/56; P < 0.001). For the 121 eyes that completed 2 years of follow-up, mean baseline BCVA improved from 0.749 ± 0.51 (Snellen equivalent 20/112) to 0.536 ± 0.40 (Snellen equivalent 20/69; median, 20/52; P < 0.001) at 2 years. Similarly, baseline BCVA was 0.777 ± 0.55 (Snellen equivalent 20/120) for the 59 eyes that completed 3 years of follow-up and reached 0.505 ± 0.37 (Snellen equivalent 20/64; median, 20/52; P < 0.001) following 3 years of anti-VEGF therapy. The mean change in BCVA corresponded to a gain of +11.6 ± 5.7 approx ETDRS letters at 1 year of follow-up, which was maintained at +10.7 ± 8.0 approx ETDRS letters at 2years and +13.6 ± 11.4 approx ETDRS letters at 3 years of follow-up ( Figure 1 ).