Expert perspectives on the evolution of retina practice, procedures, technologies and instrumentation.
Leo A. Kim, MD, PhD, FASRS; Joan W. Miller, MD
The origin of vascular endothelial growth factor (VEGF) as a target for pathologic ocular neovascularization dates back to at least 1948, when I.C. Michaelson first hypothesized about the existence of Factor X, a “factor in the retina capable of affecting the growth of vessels” and “associated with the metabolism of the retinal tissue.”[1] In 1971, Judah Folkman published his seminal theory that angiogenesis plays a role in tumor growth and metastasis[2]. He proposed the existence of a tumor angiogenic factor that mediated endothelial cell growth and vascular proliferation into the tumor. He further suggested that “anti-angiogenesis” may be used to restrict the blood supply to a tumor and induce dormancy of the tumor. This idea of anti-angiogenic therapy is the foundation of current treatment modalities for a wide array of ocular conditions characterized by aberrant angiogenesis.
In 1983, Harold Dvorak isolated vascular permeability factor (VPF), which was 50,000 more potent than histamine[3]. Then in 1989, Ferrara and Henzel cloned, sequenced, and characterized VEGF from pituitary follicular cells[4]. Concurrently, Connolly and colleagues identified human vascular permeability factor (hVPF) from a human histiocytic lymphoma cell line[5]. It was subsequently found that VEGF and hVPF shared the same cDNA sequence, and were in fact, the same molecule [6,7].
The elucidation of VEGF as Michaelson’s “Factor X” was the culmination of work from a group of researchers in Boston starting in the 1990s. First, Adamis et al. demonstrated that VEGF was secreted by the retinal pigment epithelium (RPE) in culture[8] and was responsive to hypoxia[9]. At the same time, Miller and colleagues identified VEGF in aqueous fluid samples from a nonhuman primate model of laser-induce retinal vein occlusion and iris neovascularization. This was the first demonstration in vivo of a correlation between VEGF protein levels and ocular angiogenesis[10]. Thereafter, this correlation was demonstrated in patients[10] and confirmed in a larger study that patients with active proliferative diabetic retinopathy, iris neovascularization, and central retinal vein occlusions, all had elevated expression of VEGF within intraocular fluid[11, 12]. Next, VEGF was shown to be an essential driver of neovascularization, with complete inhibition of neovascularization in the nonhuman primate model achieved using intravitreal injection of an anti-VEGF antibody (the precursor of bevacizumab). Finally, VEGF injection directly into normal nonhuman primate eyes led to iris neovascularization and neovascular glaucoma, recapitulating the pathology observed clinically with retinal ischemia[13]. VEGF alone was both sufficient and necessary for ocular neovascularization—we had solved for Factor X. Altogether; this body of work formed the foundational basis of a key role for VEGF in ocular neovascularization and the resulting development of successful anti-VEGF therapies.
Development of anti-VEGF therapies soon followed with studies investigating anti-VEGF antibodies[14], soluble VEGF receptor[15], and anti-VEGF aptamers[16]. Since it was postulated that full length antibodies might not cross the retina, experiments using intravenous delivery of fluoresceinated anti-VEGF antibodies demonstrated their localization to choroidal neovascularization (CNV) in nonhuman primate[17]. An anti-VEGF antibody fragment (rhuFab VEGF, a precursor to ranibizumab), was developed and its efficacy demonstrated in the laser-induced CNV model in 2002[18]. These studies supported clinical studies to further investigate the use of anti-VEGF therapy in patients.
The first anti-VEGF therapy approved for the eye was pegaptanib, an anti-VEGF aptamer developed by Eyetech Pharmaceuticals. Pegaptanib was shown in phase 2 and phase 3 trials to slow the progressive loss of vision associated with neovascular age-related macular degeneration (nvAMD)[19], and was approved by the FDA on December 17, 2004. Although they had lost the lead to be first to market, Genentech developed the anti-VEGF antibody fragment, ranibizumab, for intravitreal injection as they also developed a full-length anti-VEGF antibody, bevacizumab, for intravenous delivery in cancer. Ranibizumab was dramatically effective in the treatment of nvAMD with actual gains in vision as shown by the pivotal MARINA and ANCHOR trials[20, 21], the first time a treatment for nvAMD had been able to achieve improvement in vision. Based on these studies, ranibizumab was approved by the FDA on June 30, 2006 for the treatment of nvAMD. In 2005, at the time of the announcement of the positive ranibizumab results, Philip Rosenfeld presented a case study of a patient with nvAMD treated with intravitreal bevacizumab. The combination of remarkable success in trials of the anti-VEGF antibody fragment, combined with the promise of success with off-label use of the full-length antibody, led to widespread use of bevacizumab as an effective intravitreal anti-VEGF therapy. Subsequently, the large multicenter Comparison of Age-Related Macular Degeneration Treatments Trials (CATT) led by Dan Martin, demonstrated that ranibizumab and bevacizumab had similar efficacy for the treatment of nvAMD[22]. Other agents continue to be developed. Aflibercept, a chimeric fusion protein of VEGFR1 and VEGFR2 developed by Regeneron, functions as a decoy receptor with high affinity to VEGF[23]. Phase 3 results from the VIEW trials revealed that 2 mg of aflibercept every 2 months was not inferior to ranibizumab dosed monthly[24]. Aflibercept was approved by the FDA on November 18, 2011. Brolucizumab, a humanized single chain antibody fragment targeting VEGF was developed by Novartis. Based on the phase 3 results of the HAWK and HARRIER trials[25], the FDA approved brolucizumab for the treatment of nvAMD on October 8, 2019. However, post marketing reports have highlighted the risk of drug associated intraocular inflammation, retinal vasculitis, and retinal arterial occlusion tempering the use of brolucizumab[26]. Newer anti-VEGF gene therapies using AAV viral vectors to deliver anti-VEGF molecules are currently being developed in early clinical trials and larger trials are pending. RGX-314 developed by REGENXBIO uses AAV8 to deliver an anti-VEGF antibody fragment, similar to ranibizumab (NCT04832724). Similarly, ADVM-022 developed by Adverum Biotechnologies uses AAV.7m8 to deliver aflibercept (NCT04645212). Beyond nvAMD, the class of anti-VEGF drugs has now been expanded for the treatment of a wide array of neovascular conditions of the eye including diabetic retinopathy and retinal vein occlusions.
The VEGF story, from the discovery of Factor X to the widespread success of anti-VEGF therapy for so many retinal conditions, highlights the importance of basic and clinical research in improving the care of our patients. Prior to the development of anti-VEGF therapies, many of these conditions were often treated with ablative laser therapy or simply observed with a gradual decline in vision. The development of pharmacological treatments for retinal diseases has completely revolutionized the field of ophthalmology, improving vision outcomes and quality of life for millions of people. The importance of this therapeutic approach was recognized by the awarding of the 2014 Antonio Champalimaud Award Vision Award for the development of anti-angiogenic therapy for retinal disease[27].
(Milestone essay published 2021)