Purpose
To determine visual acuity (VA) and spectral-domain optical coherence tomography (OCT) outcomes with intravitreal ranibizumab for diabetic macular edema (DME) in a United Kingdom National Health Service clinical setting.
Design
Retrospective interventional case series.
Participants
Consecutive patients with DME, treated with the first ranibizumab injection between August 2013 and March 2014 across 4 sites of Moorfields Eye Hospital, London.
Methods
Two hundred eyes of 164 consecutive patients with center-involving DME and VA ≤79 ETDRS letters, central subfield macular thickness (CST) ≥350 μm on Topcon 3D OCT 2000, initiated on a loading phase of 3 intravitreal ranibizumab injections and who had at least 6 months follow-up were reviewed. Subsequent retreatment was guided by VA and OCT with the aim of treating to stability. VA, OCT CST, and macular volume (MV) were recorded at baseline and monthly to 12 months.
Results
The mean VA, mean CST, and mean MV at baseline were 54.4 (± 15.26) letters, 490.16 (± 116.54) μm, and 10.46 (± 2.28) mm 3 . The mean VA change at 12 months was +6.6 (± 13.35) letters ( P = .0003). A total of 40.3% of patients (n = 77) gained ≥10 letters and 25.1% (n = 48) gained ≥15 letters; 8.9% (n = 17) lost ≥10 letters and 6.3% (n = 12) lost ≥15 letters. At 12 months, the mean change in CST and MV were −133.9 (± 160.12) μm ( P = .0001) and −1.5 (± 1.96) mm 3 ( P = .0001), respectively. An average of 7.2 (± 2.3) injections were given over 12 months.
Conclusions
Outcomes with 3 loading injections of 0.5 mg ranibizumab given monthly followed by pro re nata retreatment in a clinical setting are comparable with outcomes from clinical trials.
Diabetic macular edema is one of the leading causes of visual impairment in the working-age population in developed countries. In the United Kingdom, the prevalence of clinically significant macular edema is 15.8%–18.1% with central macular involvement in 8.7%–10.0% of eyes. The prevalence of diabetes is steadily increasing, with 285 million people living with diabetes globally in 2010, and this number is projected to increase to 539 million by 2030. Diabetic macular edema will therefore pose a significant burden on future public health systems worldwide.
In the past, treatment options for diabetic macular edema were limited to macular laser photocoagulation, intravitreal triamcinolone, and pars plana vitrectomy in addition to adequate systemic control of diabetes and hypertension. Macular laser therapy led to stabilization of visual acuity, with only up to 28% achieving a 10-letter gain. In recent times evidence from several randomized controlled clinical trials of vascular endothelial growth factor (VEGF) inhibitors, including the DRCR.net Protocol I, RESTORE, RELIGHT, READ-2, RESOLVE, RISE & RIDE, RESTORE extension, BOLT, REVEAL, VIVID and VISTA, and DRCR.net Protocol T, demonstrated an improvement in visual acuity in diabetic macular edema patients. Therefore, patient and clinician expectations of visual outcome in diabetic macular edema have increased. Indeed, these clinical trials indicate that 50% of patients treated with these agents can expect a 2-line gain and 30% a 3-line gain in best-corrected ETDRS visual acuity.
Ranibizumab (Lucentis; Novartis Pharma AG, Basel, Switzerland, and Genentech Inc, South San Francisco, California, USA) was the first licensed VEGF inhibitor approved for the treatment of diabetic macular edema. Ranibizumab is a humanized monoclonal antibody Fab fragment (without the Fc portion) specifically designed for ocular use. It binds to VEGF-A with high affinity and inhibits all isoforms of VEGF-A.
Ranibizumab was European Medicines Agency licensed for diabetic macular edema in 2011 based on prospective phase II and III randomized clinical studies. At 12 months, these studies showed that intravitreal ranibizumab provided superior visual acuity gains compared with controls including sham and laser therapy. Long-term data from the RESTORE extension and Protocol I studies demonstrated that individualized ranibizumab therapy based on strict retreatment criteria including changes in visual acuity and optical coherence tomography (OCT) central macular thickness resulted in maintenance of year 1 visual acuity gains with reducing frequency of injections in years 2 and 3 and thereafter.
The National Institute for Health and Care Excellence (NICE), the regulatory body that provides national evidence-based guidance for health and social care in the United Kingdom, approved of the use of ranibizumab for the treatment of diabetic macular edema on the National Health Service in August 2013. However, this approval was only for eyes having a central retinal thickness of ≥400 μm on OCT. This guidance was based on the RESTORE sub-analysis, which showed that eyes with central retinal thickness ≥400 μm on OCT benefited the most from injections.
With the introduction of ranibizumab into clinical practice, it is important to establish the effectiveness of the therapy in real-life settings in the National Health Service in the United Kingdom, where multiethnic populations with varying levels of diabetic control, multiple comorbidities, missed or delayed appointments, and under-treatment can influence outcomes. This study compares real-life outcomes of pro re nata (PRN) ranibizumab treatment after 3 loading doses with those from the tightly controlled environment of clinical trials. The main study objectives were to evaluate mean change from baseline in visual acuity, OCT central subfield thickness, and total macular volume at 12 months. Baseline factors that influenced outcome were also assessed.
Methods
Study Design
This is a retrospective interventional case series studying the 12-month outcome of consecutive patients with diabetic macular edema in 1 or both eyes, treated with the first ranibizumab injection between August 2013 and March 2014. The study was approved prospectively by the Clinical Audit and Assessment Committee of Moorfields Eye Hospital and registered with the trust clinical audit department (reference no: CA13/MR/45). Patients who had consented to imaging and anonymized data collection and analysis of outcomes as part of their clinical care were included and the study followed the tenets of the Declaration of Helsinki.
Patients
The study enrolled patients aged ≥18 years of age with either type I or type II diabetes mellitus. Patients with center-involving diabetic macular edema having a visual acuity of ≤79 letters with the Early Treatment Diabetic Retinopathy Study (ETDRS) visual acuity chart (6/9) at baseline, with a central subfield thickness ≥350 μm on Topcon 3D OCT 2000 (Topcon, Tokyo, Japan) and a foveal avascular zone greatest linear diameter ≤1500 μm on fluorescein angiography (based on evidence at that time that showed good visual gains in patients with no macular ischemia), who had at least 6 months of follow-up visits after having received ≥3 intravitreal ranibizumab injections were included in the study. Eligibility was determined based on these criteria at the clinic visit when the decision to treat was made and then future appointments for initiation and continuation of injection therapy were arranged at a standardized injection clinic where visual acuity assessments and Topcon 3D OCT acquisition were performed. Because we used Topcon 3D OCT for image acquisition, the inclusion criterion of ≥350 μm on the Topcon 3D OCT (which approximately equates to 400 μm on OCT Spectralis; Heidelberg Engineering, Heidelberg, Germany) was used to ensure compliance with national guidelines. If both eyes were receiving treatment, both eyes were included in the analysis. All grades of diabetic retinopathy were included, as well as vitrectomized eyes, eyes with vitreomacular interface abnormalities, and eyes with previous macular laser or panretinal photocoagulation (PRP). These patients were included because this reflects our routine clinical practice in the real-life situation where the full range of patients with diabetic macular edema, including vitrectomized eyes and those with mild or early epiretinal membranes that did not require surgery, were treated. Key exclusion criteria included a history of an acute coronary event or stroke in the previous 3 months, pregnancy or lactation, active infection or intraocular inflammation in either eye, poor view of the fundus, other causes for macular edema, anti-VEGF treatment for any other condition, and other macular diseases present at baseline that might confound the outcomes, such as coexistent retinal vein occlusion.
Assessments
Visual acuity was assessed using ETDRS charts at 4 meters at baseline and at all injection visits and an ETDRS letter score equivalent to the Snellen score estimated for other visits using a conversion table. The primary efficacy endpoint was the change in mean visual acuity from baseline to 12 months. The secondary efficacy endpoint was the proportion of patients gaining ≥10 and ≥15 letters at month 12.
OCT scans were performed at every visit using only the Topcon 3D OCT 2000. The efficacy endpoints included the change in mean central subfield thickness (defined as the thickness of central 1000 μm on the ETDRS grid) and the change in mean total macular volume from baseline to month 12.
The number of injections received at the end of 12 months and prevalence of infectious endophthalmitis were also noted. Data collected at baseline included age, sex, ethnicity, type of diabetes, and previous treatment history.
Treatment Protocol
Patients were treated according to the hospital-trust protocol, which was in accordance with the 2011 EU summary of product characteristics label for ranibizumab. The initial mandated “loading” phase of injections involved 3 consecutive monthly intravitreal injections of 0.5 mg ranibizumab, which was then followed by a PRN phase. The retreatment criteria in the PRN phase were determined by the treating clinician with the aim of treating toward visual acuity and OCT stability and were not based on strict guidelines as in clinical trial protocols.
Treatment was withheld upon stabilization of vision—that is, when there was no further improvement in visual acuity attributable to ranibizumab treatment and/or the macular fluid was stable on 2 consecutive visits at the discretion of the treating clinician. Once stable, patients were monitored twice monthly, and if a decrease in visual acuity due to disease activity was observed, ranibizumab treatment was resumed until visual acuity and macular fluid were judged stable.
Laser therapy was allowed throughout the study according to ETDRS guidelines, at the treating clinician’s discretion. If both ranibizumab and laser were given on the same day, then ranibizumab was administered at least 30 minutes after laser treatment. Macular laser could be repeated after a minimum interval of 90 days. Panretinal photocoagulation was permitted throughout the study period.
Statistical Analysis
Sample measurements have been summarized with mean value and standard deviation. Comparison between baseline and final parameters were done by paired t test. Correlation between parameters was evaluated using correlation coefficient and a correlation coefficient of at least 0.30 was interpreted as clinically significant.
A complete model with eye, visit, and eye-by-visit interaction factors was considered. Data collection and analyses were done using Microsoft Excel 2010 (Microsoft Corporation, Washington, USA) and SPSS 22.0 (SPSS Inc, Chicago, USA). A P value of <.05 was interpreted as statistically significant.
Results
Patient Disposition and Baseline Characteristics
A total of 968 eyes were screened for eligibility and 200 consecutive eyes of 164 patients were included in this study. Baseline characteristics and demographic data for the cohort are summarized in Table 1 .
| Characteristic | Result |
|---|---|
| Mean age | 65.11 (± 10.83) years |
| Sex | |
| Male | 63.4% (n = 104) |
| Female | 36.6% (n = 60) |
| Ethnicity | |
| White | 36% (n = 59) |
| Southeast Asian | 28.7% (n = 47) |
| Afro-Caribbean | 9.2% (n = 15) |
| Other | 15.8% (n = 26) |
| Unknown | 10.3% (n = 17) |
| Diabetes mellitus | |
| Type 1 | 2.4% (n = 4) |
| Type 2 | 97.6% (n = 160) |
| HbA1c | 8.51 (± 5.49)% (n = 123) |
| Prior treatment | |
| Macular laser | 65.5% (n = 131) |
| PRP | 37% (n = 74) |
| Vitrectomy | 8.5% (n = 17) |
| Anti-VEGF injections | 11.5% (n = 23) |
| Mean washout period after prior anti-VEGF injection | 152.91 (± 110.55) days |
| Mean baseline visual acuity | 54.4 (± 15.26) letters (6/24) |
| Mean baseline central subfield thickness | 490.16 (± 116.54) μm |
| Mean baseline macular volume | 10.46 (± 2.28) mm 3 |
A total of 21.95% of patients (n = 36) had bilateral injections. Table 2 summarizes demographic data in unilaterally vs bilaterally treated eyes. The mean baseline visual acuity was 55.4 (± 13.63) (6/24) in unilateral cases and 52.1 (± 18.23) (6/30) in bilateral cases. One hundred ninety-five eyes (97.5%) were followed up for 12 months, with 4 of these eyes having missing visual acuity records at this time. Five eyes (2.5%) were lost to follow-up at 12 months.
| Unilaterally Treated (128 Patients) | Bilaterally Treated (36 Patients) | P Value | |
|---|---|---|---|
| Mean age | 65.55 (± 10.83) years | 64.33 (± 10.25) years | .56 |
| Sex | |||
| Male | 64.84% (n = 83) | 58.33% (n = 21) | .56 |
| Female | 35.16% (n = 45) | 41.66% (n = 15) | |
| Ethnicity | |||
| White | 35.94% (n = 46) | 36.11% (n = 13) | .96 |
| Southeast Asian | 29.69% (n=38) | 25% (n = 9) | |
| Afro-Caribbean | 9.38% (n=12) | 8.33% (n = 3) | |
| Other | 14.84% (n = 19) | 19.44% (n = 7) | |
| Unknown | 10.16% (n = 13) | 11% (n = 4) | |
| Diabetes mellitus | |||
| Type 1 | 1.56% (n = 2) | 5.56% (n = 2) | .21 |
| Type 2 | 98.44% (n = 126) | 94.44% (n = 34) | |
| HbA1c | 7.89 (± 1.31) (n = 90) | 8.33 (± 1.46) (n = 32) | .27 |
Efficacy
Visual Acuity
The mean visual acuity change at 1 year (n = 191 eyes) was +6.6 (± 13.35) letters ( P = .0003). 40.3% (n = 77) gained ≥10 letters and 25.1% (n = 48) gained ≥15 letters, while 8.9% (n = 17) lost ≥10 letters and 6.3% (n = 12) lost ≥15 letters; of those, there were cases of central retinal artery occlusion (n = 1), retinal re-detachment (n = 1), Fuchs endothelial dystrophy (n = 1), progression of cataract (n = 1), and incorrect visual acuity recording (n = 1).
Among those who had bilateral treatment, the eye treated first (or the right eye, in cases where both eyes had the first injection on the same day) had a mean visual acuity gain of 7.42 (± 15.74) letters ( P = .0077) and the eye treated second (or the left eye, in cases where both eyes had the first injection on the same day) showed a visual acuity change of +9.33 (± 13.85) letters ( P = .0003). There was no statistically significant difference in the visual acuity change between these 2 groups ( P = .5864).
In subgroups of baseline visual acuity, the mean visual acuity change from baseline to 1 year was +8.8 (± 14.54) in those with baseline visual acuity ≤60 letters ( P = .0001), +4.2 (± 8.38) in those with baseline visual acuity between 61 and 73 letters ( P = .001), and −4.3 (± 8.44) in those with baseline visual acuity >73 letters ( P = .22) ( Figure 1 ).
