Darin R. Goldman, MD

Choroidal neovascularization (CNV) may result from various pathological entities. The end result is an abnormal vascular complex that breaks through Bruch’s membrane into the subretinal space with consequent leakage of serous and hemorrhagic contents. When this occurs in proximity to the fovea, there may be significant deleterious effects to central visual acuity. Neovascular age-related macular degeneration (AMD) is the most common cause for CNV. However, there are numerous “other” pathological conditions that may result in CNV or CNV-like entities. These include pathological myopia, angioid streaks, traumatic sequela (ie, choroidal rupture), ocular histoplasmosis, polypoidal choroidal vasculopathy (PCV), central serous chorioretinopathy, hereditary macular dystrophies, inflammatory retinochoroidopathies, optic nerve drusen, and choroidal tumors.¹ Idiopathic CNV can also occur. Although these other causes for CNV have a much lower incidence than neovascular AMD, they tend to occur in younger patients, in whom their potential lifetime impact may be much greater.

Historically, various treatments have been employed to treat CNV with causes other than neovascular AMD. These include thermal laser photocoagulation,² verteporfin photodynamic therapy (PDT),^3–6 submacular CNV extraction,⁷ and intravitreal steroids.^8,9 With these treatment modalities, rates of vision loss could be slowed under the best of circumstances. None of these therapies resulted in an overall improvement in vision on average.

Vascular endothelial growth factor (VEGF) is integral to normal angiogenesis for a multitude of physiological processes.¹⁰ However, abnormally increased levels of VEGF can result in devastating pathological consequences. One of the major common denominators among all disease entities that cause CNV is abnormally high levels of VEGF. VEGF levels are highest in AMD, followed by myopic CNV, and then PCV.¹¹ Neovascular AMD is the most common cause of CNV and results in an enormous burden to society from a public health standpoint.¹² Over the last decade, anti-VEGF therapy has been revolutionary in the treatment of neovascular AMD. Anti-VEGF therapy with ranibizumab elevated treatment from historical benchmarks of decreasing the rate of vision loss to providing visual stabilization in the majority and visual improvement in a significant portion of eyes.13,14

Many non-AMD causes of CNV have experienced a similar revolution regarding treatment benefit due to the development of anti-VEGF agents.¹⁵ However, the lower prevalence of these other causes for CNV, in comparison to neovascular AMD, creates a significant barrier to the development of disease-specific treatment. As such, there remains no on-label anti-VEGF therapy for any of the non-AMD causes for CNV other than due to high myopia in the United States (US), even though they are used routinely in an off-label manner. This off-label approach to anti-VEGF treatment for these conditions began by extrapolating from the extensive scientific record pertaining to neovascular AMD. The body of literature devoted to non-AMD causes for CNV continues to grow but still does not compare in its rigor to those for AMD. Myopic CNV is the most frequently encountered non-AMD cause of CNV. Other entities that will be discussed in this chapter include PCV, angioid streaks, central serious chrioretinopathy, and presumed ocular histoplasmosis syndrome (POHS).

Myopic Choroidal Neovascularization

High myopia (> 6.0 diopters) along with its intrinsic axial elongation (> 25 mm to 26 mm) results in degenerative changes in various layers of the globe including Bruch’s membrane and the retinal pigment epithelium (RPE). CNV is the most common sight-threatening pathology associated with high myopia, present in more than 10% of highly myopic eyes¹⁶ and bilaterally in 15% of eyes.¹⁷ CNV develops spontaneously as an abnormal vascular complex that traverses through a weak area in Bruch’s membrane (lacquer crack). The CNV complex generally occurs in proximity to and underneath the fovea with resultant fluid, hemorrhage, and/or fibrosis that leads to vision loss or metamorphopsia. VEGF levels are significantly elevated in the presence of myopic CNV, although less so than in AMD.¹¹ These high VEGF levels are effectively suppressed with anti-VEGF therapy.¹⁸

The typical appearance of myopic CNV is that of a subfoveal grayish membrane. Associated exudation is usually minimal and, if present, hemorrhage is also typically minimal (Figure 8-1[A]). Fluorescein angiographic (FA) appearance is usually that of a classic, or type 2, CNV leakage pattern (Figure 8-1[B]). Optical coherence tomography (OCT) shows a CNV complex present between the neurosensory retina and the RPE, which is typically highly reflective, well-circumscribed and dome shaped, with minimal associated sub- or intraretinal fluid (Figure 8-2). The diagnosis of myopic CNV is made with the combination of clinical examination, FA, and OCT.

The natural history of CNV due to myopia is more favorable than that of neovascular AMD. Large disciform scarring is less common and the disease is not bilateral as often as AMD. However, overall the prognosis is still poor with 43% of eyes losing 2 or more lines of vision and 60% of eyes having final visual acuity of 20/200 or worse.¹⁹ The loss of vision in eyes with CNV from myopia is progressive over time²⁰ and contributes significantly as a cause for vision loss and blindness throughout the world.¹⁷ Specifically, the natural history of CNV in pathological myopia is significantly worse in patients older than 50 compared to those younger than 50.²¹ In a 3-year natural history study, 43% of patients younger than 40 maintained 20/40 or better visual acuity and only 11% had visual acuity worse than 20/200.22 In contrast, more than 50% of those aged over 40 had visual acuity worse than 20/200. Because of the more favorable natural history in younger patients, consideration should be given to a more conservative treatment approach in this demographic. Other factors that may play a role in the ultimate visual prognosis include presenting visual acuity, location of CNV, and degree of myopia. High myopia is increasing in prevalence with an estimated one billion people to be affected worldwide by 2050.²³ High myopia is more common in Asian countries, thought to be due to environmental factors such as decreased time spent outdoors and increased near work activities.²⁴

Until recently, the only Food and Drug Administration-approved treatment for myopic CNV in the US was PDT with Visudyne (Novartis AG), which is associated with less significant vision loss compared to sham at 2 years.⁴ However, this result was not statistically significant, with 64% of the PDT group avoiding loss of more than or equal to 8 lines of vision compared to 49% in the sham group. PDT has also been combined with intravitreal steroid, which allowed for fewer PDT treatment sessions, however, without any visual advantage and with the added risks of cataract and glaucoma.^8,25 Anti-VEGF therapy has emerged more recently as an effective off-label treatment for myopic CNV; available agents include bevacizumab, ranibizumab, and aflibercept. Intravitreal ranibizumab was approved for this indication in early 2017 and is currently the only FDA-approved treatment for myopic CNV in the US.

Intravitreal bevacizumab26–30 and ranibizumab^31–34 have been studied for the treatment of myopic CNV in a number of retrospective and prospective uncontrolled case series (Table 8-1). For bevacizumab, the percentage of eyes gaining 15 or more letters of vision ranged from 30% to 70%, the mean improvements in logMAR visual acuity averaged 0.24, and the mean number of injections ranged from 1.8 to 3.6 over one year. The best visual gains were noted in a study that had a high portion of treatment-naïve eyes in comparison to the other studies, with only 10% of eyes receiving prior PDT.²⁷ Although other studies found worse visual acuity outcomes with anti-VEGF therapy after PDT, a 2009 retrospective study did not find any differences in visual acuity outcomes whether PDT treatment was previously received or not.³⁰

For ranibizumab, the percentage of eyes gaining 15 or more letters of vision ranged from 25% to 40.3%, the mean gain in Early Treatment Diabetic Retinopathy Study (ETDRS) letters was 10, and the mean number of injections ranged from 2.3 to 4.1 over one year. The majority of visual gains were achieved after the first injection.³¹ Positive predictors of visual outcome included better baseline best-corrected visual acuity (BCVA) and extra-foveal CNV location. With continued treatment, BCVA continued to improve over time with 25%, 30%, and 35% of eyes experiencing BCVA gains of 3 or more lines of vision at 1, 2, and 3 years, respectively.³² In this study, a mean of 4.1, 2.4, and 1.1 injections were performed over the first, second, and third years, respectively, showing a trend toward less treatment being required over time. The REPAIR study showed that 14.4% of eyes improved from the worse-seeing eye to the better-seeing eye of the patient, and the greatest improvement in visual acuity occurred after the first month.³³

Furthermore, mean central foveal thickness (CFT) was consistently reduced with both bevacizumab and ranibizumab. Overall, patients younger than 50 years of age experienced better vision outcomes^27,29,30,32 and required fewer injections.³⁰ The percentage of eyes gaining 3 or more lines of vision (bevacizumab 30% to 70%, ranibizumab 25% to 40.3%) were superior when compared to 6% in the PDT-treatment group and 3% in the placebo-treatment group of the Verteporfin In Photodynamic Therapy Trial.³ Another study showed that more than 50% of eyes achieved final visual acuity of 20/40 or better and only 2% of eyes had final visual acuity worse than 20/200.²⁸ Six-line or better visual acuity gains were achieved in 13% of eyes compared to 0% in the Verteporfin In Photodynamic Therapy Trial.²⁸

Regarding treatment regimen, anti-VEGF dosing with 3 consecutive monthly injections at initiation of treatment has not been shown to be more effective compared to 1 initiation dose followed by an as-needed treatment approach (1-pro re nata [prn]).³⁴ One study retrospectively evaluated 46 treatment-naïve eyes with subfoveal myopic CNV treated with 0.5 mg intravitreal ranibizumab.³⁴ At 12 months, the mean number of injections given to the 1 initial injection group was 2.32 and to the 3 initial injections group was 3.57. No differences in visual outcome were found between the 2 groups. When 1-prn dosing is followed, required retreatment over the first year can range from 0.8 to 3.1 injections.^{28,30,32–34}

Aflibercept has also been evaluated for the treatment of myopic CNV and shown to be superior to sham treatment. In the MYRROR study, 122 patients with myopic CNV³⁵ were treated with either 2 mg intravitreal aflibercept at baseline and then as needed up to 44 weeks or sham treatment up to week 20 and rescue aflibercept treatment if needed at week 24 and beyond. The mean number of aflibercept injections in the treatment arm was 4.2 and in the sham arm was 3.0 at 48 weeks. At week 24, the intravitreal aflibercept group gained a mean of 12.1 letters of vision compared to 2 letters lost in the sham group, and 38.9% of eyes in the aflibercept group gained 3 lines or more compared to 9.7% in the sham group. Anatomical outcomes similarly favored the aflibercept treatment arm. At week 48 (crossover allowed in sham group at week 24), the aflibercept group gained 13.5 letters compared to 3.9 in the sham/aflibercept crossover group. These findings were all statistically significant. Another retrospective study showed that only 1 aflibercept injection was required to obtain CNV resolution in 55% of eyes and the mean number of injections over an 18-month study period was only 2.1, similar when compared to bevacizumab and ranibizumab. They also found that the mean number of aflibercept injections in patients younger than 50 years of age was 1.5 compared to 2.7 in patients 50 years or older at 18 months.36

Other agents have been compared to anti-VEGF in their effectiveness for the treatment of myopic CNV. In a study retrospectively evaluating 53 eyes with myopic CNV treated with either 20 mg sub-Tenon triamcinolone or 1 mg intravitreal bevacizumab,37 the mean number of treatments over 12 months was 1.3 in the triamcinolone group and 2.1 in the bevacizumab group. Visual acuity outcomes were superior in the bevacizumab group with 32% gaining more than 3 lines compared to 15% in the triamcinolone group. Mean BCVA changes at 12 months were 1.9 letters gained in the bevacizumab group compared to 0.3 letters lost in the triamcinolone group. These results support the superior efficacy of bevacizumab over sub-Tenon triamcinolone.

Two other studies compared patients treated with 0.5 mg ranibizumab to 1.25 mg bevacizumab at baseline and then as needed.^38,39 No statistically significant visual or anatomic difference was found between the 2 groups and the mean number of injections was also similar between the groups in both studies. BCVA improved significantly in all treatment arms, but there was a trend favoring more significant reduction of CFT in the ranibizumab group.³⁹

A prospective, multicenter study of 55 eyes compared PDT to intravitreal bevacizumab at 2 years.⁴⁰ Group 1 was treated with PDT at baseline and then every 3 months as needed. Group 2 was treated with 3 serial intravitreal bevacizumab injections for the first 3 months followed by as-needed treatment. At 2 years, there was a statistically significant improvement in BCVA in the bevacizumab group only. The mean number of PDT treatments was 1.8 over the first year and 0.2 over the second year. The mean number of intravitreal bevacizumab injections was 3.6 over the first year and 0.6 over the second year. This study supported the superiority of bevacizumab to PDT. Similar findings were reported for ranibizumab in RADIANCE, a 1-year randomized, controlled trial that enrolled 277 patients at 76 centers worldwide.⁴¹ Patients were randomized into 3 groups: group 1 received 0.5 mg ranibizumab on day 1, month 1, and then as needed; group 2 received 0.5 mg ranibizumab on day 1 and then as needed; group 3 received PDT on day 1 and then PDT and/or ranibizumab from month 3 at the discretion of the investigator. Both group 1 (+10.5 letters) and group 2 (+10.6 letters) achieved significantly better visual outcomes compared to group 3 (+2.2 letters) at 3 months, prior to ranibizumab rescue being allowed in group 3. Group 1 and group 2 achieved significantly better gains of 15 or more ETDRS letters compared to group 3 at three months (group 1 = 38.1%; group 2 = 43.1%; group 3 = 14.5%), and at one year, 53.3% of eyes in group 1 and 51.7% in group 2 gained 15 or more ETDRS letters, while only 32.7% of eyes in group 3 achieved the same improvement, even allowing for ranibizumab rescue beyond 3 months. The mean number of ranibizumab injections was 4.6, 3.5, and 2.4 in groups 1, 2, and 3, respectively. This study was the first randomized, controlled trial to demonstrate the superiority of ranibizumab over PDT for the treatment of myopic CNV and data from this study ultimately led to the FDA approval for ranibizumab in myopic CNV. Furthermore, even with rescue ranibizumab treatment after month 3, the visual outcomes in group 3 never caught up to group 1 or group 2, supporting the notion that early initiation of anti-VEGF treatment is consistent with best visual acuity outcomes.

The combination of PDT with anti-VEGF therapy has also been investigated to evaluate if there is a synergistic benefit to improve visual acuity outcomes and/or reduce treatment burden. Neither of these 2 outcomes has been definitively demonstrated. One retrospective study compared 3 groups of patients with myopic CNV totaling 79 eyes that were treated with either PDT, anti-VEGF (ranibizumab or bevacizumab), or a combination of PDT and anti-VEGF.42 Treatment protocol was induction of treatment by each method followed by as-needed treatment. They concluded the combination group was the only group to achieve a statistically significant BCVA improvement although this was likely flawed by a type 2 error. The visual acuity outcomes were quite similar between the anti-VEGF monotherapy group (0.16 logMAR gain) and combination therapy group (0.22 logMAR gain), both of which showed a trend toward improvement, whereas the PDT monotherapy group (0.09 logMAR gain) showed a trend toward stability. They did show, however, that the combination of PDT and anti-VEGF treatment may reduce the required anti-VEGF treatment burden.

However, combination PDT and anti-VEGF therapy did not appear to have an advantage over anti-VEGF monotherapy in other studies.⁴³ One report of 34 eyes treated with either combination PDT and intravitreal bevacizumab or bevacizumab monotherapy⁴³ showed both groups had a significant improvement in mean BCVA at one year and there were no statistically significant differences between groups regarding visual outcomes. The mean number of treatments required was significantly lower in the combination group, although this difference was easily accounted for by the required 3 monthly loading doses in the monotherapy group. Overall, it is not clear that PDT offers any benefit in reducing the required anti-VEGF treatment burden.

In summary, anti-VEGF therapy for myopic CNV has supplanted any previous treatment modality because of its superior efficacy. Both ranibizumab and aflibercept have gained on-label approval for myopic CNV in numerous countries around the world; ranibizumab was only recently FDA-approved and is the only on-label treatment available for the treatment of myopic CNV in the US.

Despite this, treatment of myopic CNV with anti-VEGF therapy has evolved into the standard of care.⁴⁴ Intravitreal anti-VEGF therapy for myopic CNV results in moderate visual acuity gains of 3 or more ETDRS lines at 1 year in 30% to 38.9% of eyes,^30,32,33,35 similar findings to those seen following anti-VEGF therapy for neovascular AMD (34% in ANCHOR,¹⁴ 40% in MARINA¹³). Level-one evidence exists for the effectiveness of both ranibizumab⁴¹ and aflibercept³⁵ for the treatment of myopic CNV, but there has not been an adequately powered study to sufficiently determine if one anti-VEGF agent is superior (or inferior) to the others. Other studies have shown anti-VEGF monotherapy superior to PDT monotherapy and there does not appear to be any added benefit of additional or combination PDT for treating myopic CNV. In addition, there is no clear benefit of 3 serial monthly loading doses followed by as-needed treatment compared to 1 initial dose followed by as-needed treatment, although the latter strategy does have the advantage of a reduced treatment burden.⁴⁵ Overall, earlier initiation of anti-VEGF therapy is associated with better long-term visual outcomes. We recommend prompt initiation of anti-VEGF therapy at the first sign of myopic CNV, followed by monthly initial follow-up and an as-needed treatment approach following the first injection. In patients older than 50 to 55 year, more aggressive therapy should be considered, as their visual prognosis is generally worse than younger patients.

Stay updated, free articles. Join our Telegram channel

Join