To evaluate the visual outcomes, number of injections, and direct medical cost of a treat-and-extend regimen in managing neovascular age-related macular degeneration with intravitreal bevacizumab.
Retrospective, interventional, consecutive case series.
Seventy-four eyes of 73 patients with treatment-naïve neovascular age-related macular degeneration from a single clinical practice were treated monthly with intravitreal bevacizumab until no intraretinal or subretinal fluid was observed on optical coherence tomography. The treatment intervals then were lengthened sequentially by 2 weeks until signs of exudation recurred and then were reduced accordingly to maintain an exudation-free macula. Main outcomes measured included mean change from baseline visual acuity, proportion of eyes losing fewer than 3 and gaining 3 or more Snellen visual acuity lines at 1 year of follow-up, annual mean number of injections, optical coherence tomography mean central retinal thickness change from baseline, mean maximum period of extension, adverse events, and mean direct annual medical cost.
The mean follow-up period was 1.41 years. Mean Snellen visual acuity improved from 20/230 at baseline to 20/109 at 12 months ( P < .001) and 20/106 at 24 months ( P < .001). The mean number of injections over the first year was 7.94. The mean optical coherence tomography central retinal thickness decreased from 316 to 239 μm at 12 months ( P < .001). The mean direct medical cost over the first year was $3493.85.
Eyes with neovascular age-related macular degeneration experienced significant visual improvements on average when managed with intravitreal bevacizumab using a treat-and-extend regimen with fewer patient visits and injections along with lower costs compared with a fixed, monthly dosing regimen.
The introduction of anti–vascular endothelial growth factor (VEGF) therapy has revolutionized the treatment of neovascular age-related macular degeneration (AMD). Ranibizumab has proven to be the first therapy for choroidal neovascularization (CNV) secondary to neovascular AMD to result in a significant improvement in visual acuity. Intravitreal bevacizumab currently is used as an off-label treatment for neovascular AMD, having been shown to be relatively safe and effective in several retrospective and prospective studies. The Comparison of AMD Treatments Trials (CATT), a randomized, prospective, head-to-head, noninferiority trial comparing ranibizumab with bevacizumab has shown that both anti-VEGF agents have equivalent effects on visual acuity when administered according to the same schedule.
Individualized dosing regimens using anti-VEGF therapy in the treatment of neovascular AMD have been evaluated to try to maximize visual acuity outcomes while minimizing treatment burden. The phase III Minimally Classic/Occult Trial of Anti-VEGF Antibody Ranibizumab in the Treatment of Neovascular AMD (MARINA) and Anti-VEGF Antibody for the Treatment of Predominantly Classic Choroidal Neovascularization in AMD (ANCHOR) study protocols mandated monthly intravitreal ranibizumab injections over a 2-year period. The Prospective Optical Coherence Tomography imaging of patients With Neovascular AMD Treated With Intraocular Ranibizumab (PrONTO) study was the first prospective trial evaluating the use of ranibizumab in a pro re nata (PRN) approach to treat neovascular AMD. The visual outcomes reported in the PrONTO study were comparable with those in the phase III studies using less than half the number of ranibizumab injections over the first year. However, according to the study protocol, patients were required to return for monthly examinations and optical coherence tomography (OCT) examinations as well as quarterly fluorescein angiography examinations. Since the PrONTO study, other studies with traditional PRN regimens using ranibizumab and bevacizumab have demonstrated mixed results. Those studies showing the most favorable results typically had very frequent follow-up with monthly ophthalmic examinations and OCT tests. Most recently, the CATT showed that at 1 year, ranibizumab given as needed was not inferior to ranibizumab given monthly. According to the study protocol, those patients in the as needed group returned for monthly and OCT examinations as well as fluorescein angiography analysis performed at the discretion of the treating ophthalmologist. However, even with monthly follow-up, patients treated in a traditional PRN fashion may have multiple recurrences and may lose vision over time.
The treat-and-extend regimen has been proposed as an alternative type of PRN regimen that aims to address the 2 short-comings alluded to above with traditional PRN regimens: minimizing the number of patient visits and associated testing and maintaining an exudation-free macula. The first retrospective review published evaluating the treat-and-extend regimen was in reference to retinal angiomatous proliferation and showed promising results. We recently published our results demonstrating the clinical and cost effectiveness of using ranibizumab in a treat-and-extend regimen in the management of new-onset neovascular AMD. Our goal was to maintain an exudation-free macula while individualizing treatment, and thereby minimizing the total number of injections, office visits, and ancillary tests. Overall, patients experienced significant visual acuity improvements on par with the phase III ranibizumab studies but with fewer encounters, fewer treatments, and lower costs. Similar to our study with ranibizumab, the purpose of the present study was to determine the visual outcomes, number of injections, and direct medical costs of managing treatment-naïve neovascular AMD with a treat-and-extend regimen using intravitreal bevacizumab.
Patient data from 2 treating physicians (R.S.K. and C.D.R.) in a single clinical practice were reviewed for this retrospective, interventional, consecutive case series.
All patient records were identified by the diagnosis of neovascular AMD (International Classification of Diseases 9th Edition code, 362.52) from 2006 through 2009. Patient medical records were reviewed and the following data were collected: age, sex, date of neovascular AMD diagnosis, best-corrected Snellen visual acuity using the most up-to-date distance correction at each visit, analysis of pretreatment fluorescein angiogram to establish CNV lesion type, size and presence or absence of hemorrhage, and OCT evaluation for central retinal thickness (CRT) and intraretinal or subretinal fluid, or both.
The inclusion criteria consisted of patients with treatment-naïve subfoveal CNV secondary to neovascular AMD who were managed with intravitreal bevacizumab using a treat-and-extend regimen with a minimum of 6 months of follow-up. All fluorescein angiographic CNV lesion subtypes and lesion sizes were eligible. Occult with no classic CNV eyes had to have signs of presumed disease progression with recent decreased visual acuity, known CNV enlargement, or signs of hemorrhage in the macula.
The exclusion criteria included eyes with any prior treatment for neovascular AMD, including laser photocoagulation, verteporfin photodynamic therapy, intravitreal pegaptanib sodium, or intravitreal ranibizumab. Other exclusion criteria included previous participation in a clinical trial for either eye for neovascular AMD or prior vitrectomy surgery in the study eye.
Optical Coherence Tomography Analysis
The Stratus OCT (Carl Zeiss Meditec, Dublin, California, USA) or RTVue-100 OCT (Optovue Corporation, Freemont, California, USA) was used for all evaluations. The internal limiting membrane, the retinal pigment epithelium, and the Bruch membrane were identified on all scans centered through the fovea as determined by the red-free image on the OCT scanner. The macular thickness map and radial line scan protocols were used on all Stratus OCT scans. This consisted of 6 radial line scans of 6-mm length with high-resolution 512 A-scans per line in 7.68 seconds of scanning. The MM5 and MM6/radial slicer scan protocols were used on all RTVue-100 OCT scans. The MM5 scan protocol consisted of 13 horizontal and vertical lines of 5-mm length with 807 A-scans per line and 8 horizontal and vertical lines of 4-mm length with 512 A-scans per line in 0.78 seconds of scanning. The MM6/radial slicer protocol consisted of 12 radial line scans 6-mm in length with 1024 A-scans per line in 0.27 seconds of scanning. All OCT images were examined qualitatively for the presence or absence of intraretinal and subretinal fluid. The CRT was obtained from the center subfield of the macular thickness map.
All patients initially were evaluated with slit-lamp biomicroscopy, fluorescein angiography, and OCT. Patients were then treated monthly (every 4 to 5 weeks) with intravitreal bevacizumab (Avastin; Genentech, Inc, San Francisco, California, USA) until no signs of macular hemorrhage were observed on slit-lamp biomicroscopic examination and no intraretinal or subretinal fluid was observed on OCT. The treatment intervals then were lengthened sequentially by approximately 2 weeks at each visit if there were no signs of recurrent exudation. OCT was used to confirm the presence or absence of exudation (intraretinal or subretinal fluid) at every visit. The follow-up period was shortened by 2 weeks if any sign of exudation or new macular hemorrhage was evident, or if the patient experienced a subjective decline in vision or worsening visual acuity. Fluorescein angiography was repeated at the discretion of the treating physician. The treatment interval also was shortened by 2 weeks if the fluorescein angiogram showed CNV lesion growth or leakage, even if the OCT images did not show intraretinal or subretinal fluid. Patients received an injection at every visit. However, if a patient’s treatment interval was lengthened successfully to 12 weeks, the patient then was given the option to continue treatments at 12-week intervals or to go without treatment and return earlier (8 weeks) for closer monitoring.
Snellen visual acuities were collected from patients wearing their most up-to-date distance correction and were converted into a logarithm of the minimal angle of resolution score for statistical analysis. A Student t test analysis with P < .05 was considered statistically significant. Clinical outcome measures included mean and median change from baseline visual acuity, proportion of eyes losing fewer than 3 Snellen visual acuity lines and gaining 3 Snellen visual acuity lines or more of visual acuity at 1 year follow-up from baseline, annual mean number of injections, OCT mean and median CRT change from baseline, mean maximum period of extension, and adverse ocular and systemic events.
A third-party insurer cost perspective using only direct medical costs was used. The health care costs arising from provider services were according to the 2009 Current Procedural Terminology data payments by the Centers for Medicare and Medicaid Services ( Table 1 ).
|CPT or J Code
|Ophthalmic examination and evaluation, intermediate
|Fluorescein angiography, fundus photographs, both eyes
|OCT evaluation, including technical and professional fees, 1 eye
|Intravitreal injection procedure
|Cost of bevacizumab injection
Direct medical costs of patients treated with intravitreal bevacizumab using a treat-and-extend regimen were compared with those treated with intravitreal ranibizumab using a treat-and-extend regimen, using the MARINA and ANCHOR study protocols, and using the PrONTO study protocol. The direct medical cost of MARINA and ANCHOR was determined per protocol. These studies included an ophthalmic examination, fluorescein angiography, OCT, and intravitreal ranibizumab at the initial evaluation. Subsequent monthly visits included an ophthalmic examination and intravitreal ranibizumab injection. The direct medical cost of the PrONTO study was similarly determined per protocol. An ophthalmic examination and an OCT were obtained at the initial visit and at monthly intervals thereafter. A fluorescein angiogram was obtained at the initial visit followed by at months 1, 2, 3, and every 3 months thereafter. Intravitreal ranibizumab injections were calculated based on the average of 5.6 injections over the subsequent 12 months. The cost of the protocol refraction was not included. It was assumed that the additional visits at days 14 and 45 as well as additional OCTs at days 2, 4, 7, and 14 were for study purposes as outlined in the PrONTO protocol and were not necessary for routine care in clinical practice, and therefore were not included in the cost analysis.
Seventy-four eyes of 73 patients met the inclusion and exclusion criteria. The mean age was 81.1 years (range, 53 to 98 years; standard deviation, 9.2 years), 54 patients were female (74%), and all participants were white. There were 41 right eyes (55.4%) and 47 pseudophakic eyes (63.5%). The baseline characteristics are listed in Table 2 and are comparable with those in our treat-and-extend ranibizumab study. The mean duration of visual symptoms before treatment was 94.5 days (median, 30 days; range, 1 to 365 days). Thirty-four eyes (45.9%) were treated within 1 month of symptom onset. The mean follow-up period was 1.41 years (range, 0.5 to 3.7 years). The mean CNV lesion size was 1.21 disc areas (range, 0.5 to 10 disc areas) with 33% (range, 0% to 100%) of the lesion having some degree of macular hemorrhage. The CNV subtype was occult in 34 eyes (45.9%), minimally classic in 24 eyes (32.4%), and predominantly classic in 16 eyes (21.6%). No adverse ocular or systemic events were reported throughout the study.