Competing/conflicts of interest: R Troutbeck and A van Heerdon were previous Novartis medical retinal fellows, and R Bunting is the current Novartis medical retinal fellow. R Guymer is on the Novartis Australia advisory board.
Ranibizumab therapy for choroidal neovascularization secondary to non-age-related macular degeneration causes
Version of Record online: 23 DEC 2011
© 2011 The Authors. Clinical and Experimental Ophthalmology © 2011 Royal Australian and New Zealand College of Ophthalmologists
Clinical & Experimental Ophthalmology
Volume 40, Issue 1, pages 67–72, January/February 2012
How to Cite
Troutbeck, R., Bunting, R., van Heerdon, A., Cain, M. and Guymer, R. (2012), Ranibizumab therapy for choroidal neovascularization secondary to non-age-related macular degeneration causes. Clinical & Experimental Ophthalmology, 40: 67–72. doi: 10.1111/j.1442-9071.2011.02719.x
Funding sources: CERA receives Operational Infrastructure Support from the Victorian Government.
- Issue online: 5 FEB 2012
- Version of Record online: 23 DEC 2011
- Accepted manuscript online: 17 OCT 2011 08:49AM EST
- Received 17 July 2011; accepted 22 September 2011.
- choroidal neovascular membrane;
- multifocal choroiditis;
Background: To investigate the efficacy of ranibizumab therapy for choroidal neovascular (CNV) membranes secondary to conditions other than macular degeneration.
Design: Prospective case series conducted at the Royal Victorian Eye and Ear Hospital.
Participants: Twelve-month follow-up data for 41 patients with CNV recruited from the outpatient clinic from May 2008 to April 2010 is presented. Fifteen patients had myopia, seven had multifocal choroiditis, and eight had other primary causes.
Methods: All patients had visual acuity, fluorescein angiogram and optical coherence tomography performed at the initial visit (baseline). Ranibizumab was injected with a standard sterile technique. Patients were reviewed after 1 month, and further injections were given at the treating doctors' discretion.
Main Outcome Measures: Change in visual acuity and central macular thickness at 12 months was compared with baseline for each of the groups. Local and systemic adverse outcomes were recorded.
Results: Analysis was stratified by primary pathology. On average, 40%, 43% and 25% of patients with myopia, multifocal choroiditis and ‘other’ pathologies, respectively, experienced a three or more line improvement in vision. The average number of injections in 12 months was 4.2 for the entire group. Central macular thickness significantly decreased in the 12-month period for the combined group (P = 0.03). No patient had an adverse systemic side-effect; however, there was one case of endophthalmitis.
Conclusions: Ranibizumab is an effective treatment for CNV secondary to non-age-related macular degeneration causes, with most patients gaining an improvement in the first 2 months following injection.
Choroidal neovascular (CNV) disease is a common cause of profound vision loss globally. In general, CNV represents a wound repair response that involves new blood vessel growth from the choroid, traversing Bruch's membrane and extending above or below the retinal pigment epithelium (or sometimes both). This abnormal vascular proliferation has the potential to affect vision through exudation of blood and fluid into the subretinal and intraretinal spaces, producing tissue fibrosis.1 Although CNV is a hallmark of age-related macular degeneration (AMD), it is less frequently implicated in other diseases, such as pathological myopia, angioid streaks and inflammatory conditions of the posterior pole such as multifocal choroiditis (MFC).2 Although these conditions are far less prevalent than AMD, taken together, they account for a significant burden of visual morbidity in society, particularly as they commonly occur in younger patients.3–5
The poor natural history of CNV has led to several treatment options, including photodynamic therapy, thermal laser, intravitreal triamcinolone and submacular surgery; however, none of these has proven consistent in their therapeutic benefit.6–9 More recently, treatment of subfoveal CNV in AMD has been revolutionized by anti-vascular endothelial growth factor (anti-VEGF) agents.10 Although the mechanism of VEGF upregulation in CNV is not clear, inhibition of this growth factor has been effective in closing the membrane, restoring retinal architecture and improving visual acuity (VA).10 Within this class of agents, ranibizumab is a monoclonal antibody fragment that binds and inhibits all subtypes of VEGF-A. Large randomized control trials have demonstrated the success of ranibizumab for the treatment of subfoveal CNV in AMD,10–12 an indication that is now subsidized by Medicare in Australia. Small studies have suggested that anti-VEGF agents are useful for treatment of CNV because of causes other than AMD.13–18 Most of these, however, have been conducted with bevacizumab (a non-selective VEGF inhibitor), given both its accessibility and significantly reduced cost to the patient. In this study, we set out to examine patients with CNV secondary to causes other than AMD and subsequently treated with ranibizumab in order to determine its efficacy and safety profile in a non-AMD cohort.
Patients were recruited from the Royal Victoria Eye and Ear Hospital, Melbourne, Australia, between May 2008 and April 2010. The study was conducted in accordance with the Declaration of Helsinki, and written informed consent was obtained from all participants. To be eligible for inclusion, patients were required to have a subfoveal CNV secondary to causes other than AMD. Patients were excluded if they had received any prior treatment for CNV or had any active inflammation. Cases were classified into three groups broadly based on pathology for purposes of analysis: myopia, MFC and other causes (peripapillary CNV, angioid streaks, central serous chorioretinopathy, macular telangectasia, idiopathic CNV). It was decided to group the latter causes together as several categories that contained too few cases for sufficient statistical power.
Baseline characteristics regarding age, sex, and relevant ocular and medical history were recorded. VA, central macular thickness (CMT) determined by optical coherence tomography (OCT) and fluorescein angiography were assessed at the first visit, and intravitreal ranibizumab was subsequently given. The injection was given following the placement of povidone-iodine on the lids and ocular surface, and a speculum was used. Post-treatment chloramphenicol drops were given for 3 days. All patients were reviewed after 1 month, with ensuing reviews and injection schedules determined at the treating doctor's discretion (patients were retreated when there was evidence of new leakage on OCT or fluorescein angiogram, or new retinal haemorrhage). Patients had at least 12 months of follow-up, and no concurrent treatment was given in this time.
The change in VA and CMT measured from baseline to 12 months for all participants represents the primary outcome measures in this research. Vision change was also quantified as a proportion of those who achieved a gain/loss of ±3 lines of acuity from baseline, respectively. Worsening of VA was defined as a loss of three lines or more on a logMAR chart, improvement as greater than three lines, and stable as within this range.
VA measurements are presented as logMAR equivalent scores. In cases of poor acuity, measurements were scored accordingly; counting fingers – 2.0, hand movements – 2.3. Differences in VA and CMT at baseline compared with 12 months post treatment were evaluated using a paired t-test (P < 0.05 was considered significant). Analyses were conducted in R.19
Data were collected for 30 patients diagnosed with non-AMD associated CNV: 15 with myopia, 7 with MFC and 8 with other causes (peripapillary CNV – 2, angioid streaks – 2, central serous retinopathy (CSR) – 1, macular telangectasia – 1 and idiopathic CNV – 2). Approximately one quarter of participants were male, and the mean age for the cohort was 54 (range 25–86) (Table 1).
|Number of patients||15||7||8|
|Sex (M : F)||3:12||1:6||4:4|
|Mean age (range)||61 (38–86)||48 (25–68)||54 (40–68)|
VA and CMT outcomes comparing baseline and 12-month measurements for each group were presented in Table 2. Although the myopic group had the lowest baseline VA, 93% demonstrated stable or improved vision after 12 months (average of 3.5 injections). However, only 13% of patients obtained a final VA of 6/12 or better. Patients grouped by MFC did not have active uveitis during the study and, in fact, the majority presented initially with symptoms of CNV. These patients showed better baseline vision and final VA compared with the myopic group, with 57% achieving ≥6/12 (average of 3.4 injections) at final follow-up. Of the remaining patients, the number of injections required was greater than the myopia and MFC groups (average 6.25). This may reflect the poorer natural history of some of these diseases such as angioid streaks. As a whole, this group showed stable or improved vision in 75% of patients.
|Myopia n = 15||Multifocal choroiditis n = 7||Other n = 8|
|Mean VA baseline (logMAR) (±SD)||1.00 (±0.53)||0.69 (±0.41)||0.67 (±0.47)|
|Mean VA 12 months (logMAR) (±SD)||0.80 (±0.48)||0.49 (±0.50)||0.54 (±0.43)|
|P value (95% CI for Diff Means)||0.09 (−0.11–0.52)||0.24 (−0.45–0.86)||0.28 (−0.38–0.64)|
|Mean CMT baseline (µm) (±SD)||334 (±166)||328 (±82)||405 (±106)|
|Mean CMT 12 months (µm) (±SD)||258 (±112)||326 (±205)||273 (±101)|
|P value (95% CI for Diff Means)||0.15 (−83–235)||0.49 (−118–192)||0.006 (41–223)|
|Mean number of injections/patient in 12 months||3.5 (1–8)||3.4 (1-9)||6.25 (2–11)|
|Total number of injections in 12/12||52||24||50|
|% of injections given in first 3/12||58%||70%||56%|
|% of injections given in first 6/12||84%||92%||70%|
|Number of patients with final visual acuity better than 6/12||2 (13%)||4 (57%)||5 (%)|
|Gain of three or more lines||6 (40%)||3 (43%)||2 (25%)|
|Stable||8 (53%)||3 (43%)||4 (50%)|
|Loss of three or more lines||1 (7%)||1 (14%)||2 (25%)|
CMT for the entire cohort decreased significantly at the 12-month end-point (P = 0.03) (Fig. 2). However, considering the groups, this was not significant for both the myopic and the MFC group. This result was distorted by one patient in each of the MFC and myopic groups, who at 12 months had a significant reactivation of the CNV and had a CMT of 700 microns and 528 microns, respectively.
A total of 126 injections were given during the study period. One patient developed post-injection endophthalmitis (Streptococcus salivarius) and had a poor outcome (>3 lines of letters lost); there were no other recorded local or systemic complications.
Figure 1 compares the vision over the 12-month time period for each group and combined results. In general, patients had most vision improvement within the first 2 months and tended to stabilize thereafter.
In this study, we investigated the efficacy of ranibizumab therapy for causes of CNV other than AMD. Our results suggest that CNV arising from these pathologies may be successfully treated with ranibizumab, with on average, fewer injections required compared with AMD. Although not statistically significant, the general trend showed improvement in patient vision at 12 months compared with baseline, with most recovery occurring within the first 2 months. Overall, 37% of patients achieved a final VA of ≥6/12. In relative terms, 37% gained vision, 50% remained stable, and 13% lost vision after 12-month follow-up. Improvements in macula thickness were similarly noted with a significant difference between baseline and 12-month measurements for the entire cohort.
The average number of injections for all causes of CNV was 4.2 over 12 months. This is less than figures quoted for AMD, where it has been shown in the past that best results are obtained with monthly injections.10,11 The recently published Comparison of Age-Related Macular Degeneration Treatments Trial (CATT) study, comparing ranibizumab and bevacizumab for neovascular AMD, found that ranibizumab given monthly was equivalent to an as-needed regime (6.9 injections given over 12 months).20 In our study, the majority of injections were given in the first 3 months (60%), and 80% were given in the first 6 months. The younger age of our patients and the general state of their retinal pigment epithelium compared with that in AMD may account for the lower number of injections given in the 12 months.
The natural history of CNV in pathological myopia is poor; Hampton et al. found that final vision was ≤6/60 in 60% of patients, and Tabandeh et al. reporting similar VA in 73% of patients at 49 months.21,22 Conventional management of myopic CNV, including thermal laser photocoagulation, photodynamic therapy and submacular surgery, is associated with significant complications, and thus alternative therapies are being increasingly employed.23
In a recent study, Gupta et al. considered data regarding CNV for causes other than AMD.14 They examined 17 reports of anti-VEGF therapy (15 – bevacizumab, 2 – ranibizumab) utilized for the treatment of myopic CNV, and found that at 12-month, visual outcomes were better than other current treatments available. Additionally, injection frequency did not appear to affect results. Patients treated in an ‘as required’ fashion showed an improvement in vision from 0.91 to 0.66 at 6 months compared with those treated with monthly injections for the first 3 months, whereby vision improved from 0.66 to 0.33. Our patients in the myopic group had a worse baseline VA and did not gain as much vision. Several patients presented late and did not improve because clinically, they developed a stable fibrotic scar or atrophy, and further injections were considered redundant. The patient in the myopic group who had a decrease in VA greater than three lines developed endophthalmitis post injection.
All patients with MFC in this study had inactive inflammatory disease at the time of treatment. Patients in this group were unaware of any previous inflammatory ocular disease until they presented with subfoveal CNV. No patient required concurrent treatment with steroids. The trend for improved vision with ranibizumab was also seen in this group. This trend has also been reported elsewhere in the literature,14 although the number of cases was small and comparisons are complicated by the underlying pathology and the location of the CNV (juxtafoveal vs. subfoveal).
For all other causes of CNV, group numbers were too small for meaningful independent analysis and thus were grouped together. Two patients had CNV related to angioid streaks, with both demonstrating significantly decreased vision in the contralateral eye also because of CNV (treatment options did not include anti-VEGF). One of these patients had improved vision over the 12-month study period, and the other lost vision (baseline logMAR 1.0 and remained with a stable scar at logMAR 1.3). The other patient who had significant deterioration in vision over 12 months had macular telangectasia.
Sawa et al. reported on long-term treatment results of angioid streaks with bevacizumab.16 In their study, patients were treated on an as-required basis, and over the average 19 months of follow-up required 4.5 injections. At final follow-up, 87% showed stable or improved vision. The authors highlight that recurrence of CNV in eyes with angioid streaks is a major problem over time, although, in comparison with laser photocoagulation or photodynamic therapy (PDT), further damage to Bruch's membrane with an anti-VEGF seems unlikely, and therefore visual outcomes tend to be better. It would appear that we are injecting our patients more frequently than others in the literature. This is because of the lack of strict criteria with regard to timing of injections, which was left to the treating clinician's discretion.
Patients with peripapillary CNVs were also included in our study if the membrane extended subfoveally and they had no drusen or polypoidal disease. Two patients were in this group, and they required an average of nine injections. Both patients improved their vision over the study period by more than three lines. Indocyanine green angiography was not performed to exclude polyps in these cases.
The use of ranibizumab for non-AMD causes of CNV appears relatively safe, with no systemic adverse events recorded during the study. However, one case of endophthalmitis occurred with Streptococcus salivarius being cultured. This patient had a poor visual outcome. This reminds us of the seriousness of potential complications associated with intravitreal injections. Additionally, three quarters of the patients in our cohort were female, and some were of child-bearing age. There is some concern regarding potential teratogenesis associated with anti-VEGF agents and a possible effect on future fertility. Ranibizumab is known to have a shorter half-life than bevacizumab, and is therefore favoured by some clinicians for female patients of child-bearing age, although patients are counselled not to become pregnant during the course of treatment.
A major limitation of this work was the small sample size. Even though the findings for VA were not statistically significant because of the small numbers, they appear to be clinically relevant and are supportive for a recommendation that intravitreal ranibizumab be considered in the treatment of these conditions.
In conclusion, we have shown that ranibizumab is an effective therapy for both stabilizing and remediating vision loss because of CNV from pathology other than AMD. It is hoped that the results of this study will help facilitate a wider, funded use of the drug than its current approval permits.
- 7Surgical removal vs observation for subfoveal choroidal neovascularization, either associated with the ocular histoplasmosis syndrome or idiopathic: I. Ophthalmic findings from a randomized clinical trial: Submacular Surgery Trials (SST) Group H Trial: SST Report No. 9. Arch Ophthalmol 2004; 122: 1597–611., , et al.
- 19R Development Core Team. R: A Language and Environment for Statistical Computing. Vienna, Austria: R Foundation for Statistical Computing, 2010; ISBN 3-900051-07-0. Accessed January 2011. Available from: http://www.R-project.org