Purpose
To compare time to retreatment and visual function between patients with treatment-naïve neovascular age-related macular degeneration (AMD) treated with either intravitreal ranibizumab (IVR) or intravitreal aflibercept (IVA) in routine clinical practice.
Design
Retrospective, interventional comparative case series.
Participants
A total of 200 eyes of 197 patients with neovascular AMD.
Methods
A total of 99 patients in the IVR group and 101 patients in the IVA group who met the inclusion criteria with 12 months of follow-up were included in the present study. All patients received 3 consecutive monthly injections of 0.5 mg/0.05 mL ranibizumab or 2.0 mg/0.05 mL aflibercept as loading doses. Retreatment was allowed if evidence of clinical deterioration or the presence of intraretinal edema or subretinal fluid on spectral-domain optical coherence tomography examination performed at the 1-month follow-up was noted. The time to retreatment after the third injection during the loading phase to the first recurrence during the maintenance phase was compared between treatments using the Kaplan-Meier analysis. Functional and anatomic outcomes were also compared between the IVR and IVA groups.
Results
The median time to retreatment after the last induction dose was 5 months in both groups. The proportion of IVR patients who required injection retreatment was not significantly higher than that of IVA patients (67.7% and 63.4%, respectively, at the 12-month follow up; log-rank test, P = .554). In both groups, significant improvements in postoperative best-corrected visual acuity (BCVA) compared with preoperative visual acuity was observed over the 12-month follow-up period ( P < .05 for both). Central foveal thickness (CFT) decreased from the baseline values in both groups during the follow-up period ( P < .001 for both). Although there was a trend toward greater BCVA improvements in the IVA group, no significant differences in BCVA or CFT were observed between the treatment groups.
Conclusions
Both IVR and IVA were well tolerated and demonstrated efficacy in improving the visual acuity in treatment-naïve patients with AMD. Despite a trend toward greater BCVA improvements in the IVA group, a similar injection burden was observed following the loading phases of both ranibizumab and aflibercept.
Age-related macular degeneration (AMD) is most frequently treated with anti–vascular endothelial growth factor (VEGF) agents that have demonstrated efficacy in improving visual acuity outcomes. Ranibizumab (Lucentis; Novartis Pharma AG, Basel, Switzerland and Genentech Inc, South San Francisco, California, USA) is a recombinant humanized anti-VEGF antibody fragment targeting all isoforms of VEGF-A that has been extensively used for the treatment of AMD. The MARINA and ANCHOR trials demonstrated that monthly injections of ranibizumab resulted in excellent visual outcomes. However, frequent injections and financial cost represent significant limitations for patients and medical staff. To reduce these treatment burdens, a variable-dosing protocol of intravitreal injections during the maintenance phase with monthly monitoring, known as the pro re nata (PRN) regimen, has become the accepted management strategy for AMD.
Alternatively, aflibercept (Eylea; Regeneron Pharmaceuticals Inc, Tarrytown, New York, USA and Bayer Healthcare Pharmaceuticals, Berlin, Germany), a recombinant fusion protein that binds to isoforms of VEGF-A, VEGF-B, and placental growth factor, has also been approved for the treatment of AMD. Bimonthly aflibercept injections in the maintenance phase are associated with similar visual outcomes compared with monthly intravitreal ranibizumab (IVR), thereby indicating the potential to decrease the treatment burden. However, the efficacy of treatment with intravitreal aflibercept (IVA) according to the PRN regimen after the loading dose remains uncertain.
The choice between IVR and IVA during the PRN regimen for the treatment of neovascular AMD represents a clinically important issue. A previous study reported the utility of a modified PRN regimen in decreasing the number of bimonthly injections of IVA compared with that of IVR required to improve or maintain best-corrected visual acuity (BCVA). However, a comparison of the injection burden after the loading phase between IVR and IVA has not yet been reported. The purpose of the present study was to compare time to retreatment and visual function between IVR and IVA patients with treatment-naïve neovascular AMD treated according to the PRN regimen in routine clinical practice. We further evaluated these outcomes in patients with typical AMD and polypoidal choroidal vasculopathy (PCV).
Methods
We retrospectively studied Japanese patients aged ≥50 years who were newly diagnosed with neovascular AMD and who had provided written informed consent for treatment. All patients were initially treated at Yokohama City University Medical Center between January 2012 and December 2014 and were followed up for 12 months. Treatment modalities were administered during different time periods. Ranibizumab was administered from January 1, 2012 to February 28, 2013, and aflibercept was administered from February 18, 2013 to December 31, 2014. The present study was conducted according to the principles of the Declaration of Helsinki and was approved by the Ethics Committee of Yokohama City University Medical Center. This study was registered in ClinicalTrials.gov (ID number NCT02126904 ).
The inclusion criteria of the present study were as follows: a diagnosis of neovascular AMD, including typical AMD and PCV, based on clinical, spectral-domain optical coherence tomography (SDOCT), and angiographic findings; and a baseline BCVA of 20/400 or better.
Because the prevalence of retinal angiomatous proliferation (RAP) is low in Japanese patients with AMD, patients with RAP were not included in this study. Patients who had previously received treatment for AMD (ie, laser photocoagulation, submacular surgery, photodynamic therapy, or intravitreal injections of other anti-VEGF agents) or who underwent vitrectomy were also excluded from the present study. Furthermore, patients with choroidal neovascularization (CNV) as a result of high myopia, angioid streaks, hereditary disorders, uveitis, or any other secondary diseases were also excluded.
All patients received 3 consecutive monthly injections of 0.5 mg/0.05 mL ranibizumab or 2.0 mg/0.05 mL aflibercept as induction therapy. During the maintenance phase, IVR or IVA administrations were repeated in the following situations: persistent or recurrent subretinal fluid or intraretinal fluid accumulation on SDOCT imaging; clinically detectable hemorrhage; or visual acuity loss, as judged by an evaluating physician at the 1-month follow-up examination. Patients were retreated within a week.
The primary outcome measure of the present study was the comparison of the time to retreatment, defined as the time from the last injection of the loading phase to the first retreatment in the maintenance phase, in patients requiring retreatment between the IVR and IVA groups using Kaplan-Meier analysis. The secondary outcome was the comparison of the change in BCVA and central foveal thickness (CFT) between the 2 groups. We further evaluated these outcomes in patients with typical AMD and PCV.
Digital simultaneous fluorescein angiography and indocyanine green angiography using a confocal scanning laser ophthalmoscope (SPECTRALIS Product Family Version5.3; Heidelberg Engineering Inc, Dossenheim, Germany) were performed in a standard manner for the diagnosis of lesion subtypes. SDOCT (Cirrus HD-OCT model 5000; Carl Zeiss Meditec, Dublin, California, USA) was used to evaluate lesion changes during the study follow-up period.
For statistical analyses, baseline characteristics were compared using the unpaired t test and Fisher exact tests. The proportions of patients requiring retreatment with IVR or IVA over time were estimated using the Kaplan-Meier method and compared using the log-rank test. Best-corrected visual acuity and CFT values before and after treatment were compared within each group using the paired t test. Multivariate analyses were also used to adjust for the baseline differences. All statistical analyses were performed using Ekuseru-Toukei 2012 (Social Survey Research Information Co, Ltd, Tokyo, Japan). P values <.05 were considered statistically significant.
Results
Patient Characteristics
Patient baseline characteristics and clinical data are shown in Table 1 . Among the 98 patients in the IVR group, 69 were male and 29 were female, with patient age ranging from 61 to 92 years (mean age ± standard deviation (SD), 76.6 ± 6.7 years). Among the 99 patients in the IVA group, 66 were male and 33 were female, with patient age ranging from 50 to 95 years (mean age ± SD, 71.8 ± 9.1 years).
| IVR Group | IVA Group | P Value | |
|---|---|---|---|
| Number of eyes | 99 | 101 | |
| Number of patients | 98 | 99 | |
| Male/female | 69/29 | 66/33 | .607 a |
| Age, mean ± SD, y (range) | 76.6 ± 6.7 (61–92) | 71.8 ± 9.1 (50–95) | <.001 a |
| Lesion subtypes (typical AMD/PCV) | 55/44 | 53/48 | .673 b |
| Baseline logMAR BCVA | 0.44 ± 0.33 | 0.37 ± 0.37 | .030 a |
| LogMAR BCVA change, mean ± SD | 0.08 ± 0.32 | 0.12 ± 0.28 | .136 a |
| Mean CFT ± SD (μm) | |||
| Baseline | 340 ± 135 | 338 ± 116 | .945 a |
| Month 12 | 237 ± 62 | 243 ± 71 | .467 a |
| Number of injections, mean ± SD | 4.9 ± 2.2 | 5.0 ± 2.4 | .874 a |
a P value calculated using the unpaired t test.
In the IVR group, 55 eyes (55.6%) were diagnosed with typical AMD and the remaining 44 (44.4%) with PCV. Among the 101 eyes in the IVA group, 53 (52.5%) were diagnosed with typical AMD and the remaining 48 (47.5%) with PCV. Both groups were compared based on sex, lesion subtypes, and baseline CFT ( P > .05 for all). However, significant differences in age and baseline BCVA were observed between the 2 groups ( P < .001 and P = .030, respectively).
Comparison of Time to Retreatment
The proportions of patients requiring retreatment during the maintenance phase were 67.7% (67/99) in the IVR group and 63.4% (64/101) in the IVA group at 12 months after the initial treatment. No significant difference in the proportion of patients requiring retreatment was observed between the IVR and IVA groups ( P = .554). The median time to retreatment in both groups was 7 months after the initial injection, which is 5 months after the last loading phase injection ( Figure 1 ). We also evaluated the time to retreatment between the 2 groups using Cox proportional-hazard regression for adjustment of age and baseline BCVA. No significant difference in the proportion of patients requiring retreatment were observed between the IVR and IVA groups ( P = .904; Table 2 ).
| Variable | Univariate Analysis | Multivariate Analysis | |||
|---|---|---|---|---|---|
| P Value a | Hazard Ratio | 95% CI | P Value b | ||
| Total | Anti-VEGF agents (ranibizumab or aflibercept) | .554 | 0.978 | 0.685–1.397 | .904 |
| Age | 1.023 | 0.999–1.475 | .053 | ||
| Baseline BCVA | 0.789 | 0.481–1.293 | .347 | ||
| Typical AMD | Anti-VEGF agents (ranibizumab or aflibercept) | .550 | 0.946 | 0.589–1.521 | .819 |
| Age | 1.016 | 0.985–1.049 | .310 | ||
| Baseline BCVA | 0.534 | 0.449–1.513 | .825 | ||
| PCV | Anti-VEGF agents (ranibizumab or aflibercept) | .965 | 1.021 | 0.584–1.785 | .943 |
| Age | 1.026 | 0.989–1.063 | .176 | ||
| Baseline BCVA | 0.645 | 0.272–1.530 | .320 | ||
In patients with typical AMD, the proportions of patients requiring retreatment in the maintenance phase were 78.2% (43/55) in the IVR group and 69.8% (37/53) in the IVA group at 12 months after the initial treatment. No significant differences in the proportions of patients with typical AMD requiring retreatment were observed between the 2 groups; however, there was a trend toward a lower proportion of patients requiring retreatment in the IVA group ( P = .550; Figure 2 ). In contrast, the proportions of patients with PCV requiring retreatment during the maintenance phase were 54.5% (24/44) in the IVR group and 56.3% (27/48) in the IVA group at 12 months after the initial treatment, with no significant difference observed between the 2 groups ( P = .965). Cox proportional-hazard regression also showed no significant differences between the 2 groups in patients with both typical AMD and PCV ( P = .819 and P = .943, respectively; Table 2 ).
Number of Treatments
During the 12-month study period, the mean numbers of ranibizumab and aflibercept injections administered were 4.9 ± 2.2 and 5.0 ± 2.4 in the IVR and IVA groups, respectively, with no significant difference observed between the 2 groups ( P = .874; Figure 3 ).
In patients with typical AMD, the mean numbers of ranibizumab and aflibercept injections administered during the 12-month study period were 5.1 ± 2.1 and 5.2 ± 2.5 in the IVR and IVA groups, respectively, with no significant difference observed between the 2 groups ( P = .733). In patients with PCV, the mean number of ranibizumab and aflibercept injections administered during the 1-year follow-up period was 4.7 ± 2.2 in both groups, with no significant difference observed between the 2 groups ( P = .935).
Visual Acuity Outcomes
The mean logarithm of the minimum angle of resolution (logMAR) visual acuity at baseline was 0.44 ± 0.33 in the IVR group and 0.37 ± 0.37 in the IVA group. The mean logMAR BCVA at 3, 6, and 12 months after the initial injection was 0.34 ± 0.36, 0.34 ± 0.38, and 0.37 ± 0.41 in the IVR group and 0.26 ± 0.37, 0.25 ± 0.38, and 0.25 ± 0.40 in the IVA group, respectively. In both groups, the postinjection BCVA was significantly improved when compared with the baseline visual acuity throughout the 12-month period ( P < .05 at 3, 6, and 12 months in both groups; Figure 4 ).
The clinical courses of patients with each disease subtype are also shown in Figure 4 . Although postinjection BCVA improved in both groups during the follow-up period, BCVA tended to decrease after the loading phase in patients with typical AMD in the IVR group.
In total, the mean differences in preinjection and postinjection logMAR BCVA values at 12 months after initial treatment were 0.08 ± 0.32 in the IVR group and 0.12 ± 0.28 in the IVA group, respectively. No greater degree of improvement in the visual acuity was shown between the 2 groups (95% confidence interval [CI] [−0.044∼0.123], P = .355). The mean differences in preinjection and postinjection BCVA values at 12 months after initial treatment in each subtype were 0.08 ± 0.30 in the IVR group and 0.12 ± 0.27 in the IVA group in patients with typical AMD and 0.08 ± 0.34 in the IVR group and 0.11 ± 0.30 in the IVA group in patients with PCV, respectively. Despite a trend toward greater BCVA improvement in the IVA group among patients with both typical AMD and PCV, no significant difference in BCVA improvement was observed between the groups at 12 months (typical AMD, 95% CI [−0.063∼0.152], P = .415; PCV, 95% CI [−0.100∼0.168], P =.617; Table 3 ).
