What an exciting year 2019 turned out to be for the management of retinal disease. As the year draws to a close, it’s an ideal time to take a look back at some of the biggest stories, including the introduction of a new US Food and Drug Administration (FDA)–approved pharmaceutical agent for neovascular age-related macular degeneration (nAMD), positive data on interventions capable of slowing geographic atrophy (GA) progression, intriguing trial results for patients at high risk for nonproliferative diabetic retinopathy (NPDR), and further exploration of the power of artificial intelligence (AI).
New Treatments for AMD
AMD remains a leading cause of blindness secondary to one of two advanced stages: nAMD and GA.[1] This year brought clinically relevant advances on both fronts.
nAMD
Brolucizumab (Beovu) received FDA approval for the treatment of nAMD in October 2019. Brolucizumab, a humanized, monoclonal single-chain antibody fragment, inhibits vascular endothelial growth factor (VEGF) A—and at 26 kD, it is smaller than the other anti-VEGF agents. At 6 mg per dose, the molar anti-VEGF binding capacity is substantially greater as well.
The FDA’s approval was based on the phase 3 HAWK and HARRIER double-masked, randomized trials, which used aflibercept (Eylea) as the comparator.[2] During the first 4 months of the 2-year studies, the matched phase, brolucizumab was found to achieve more complete drying of the retina. At week 16, 34% vs 52% of patients had residual fluid following treatment with brolucizumab and aflibercept, respectively. Subsequently, patients receiving brolucizumab were treated every 8 or 12 weeks depending on disease activity, with 39%-45% of these patients maintained at quarterly dosing through the week 96 endpoint.
GA
Characterized by loss of choriocapillaris, retinal pigment epithelium, and photoreceptors, GA represents the single greatest unmet medical need in patients with retinal disease,[3] with no approved therapeutic strategies to prevent onset or progression.
Strong genetic and physiologic data have implicated complement dysregulation in GA pathogenesis,[4] although targeting complement in an attempt to slow GA enlargement has yet to yield a validated pharmaceutical therapy.[5]
Renewed hope has been brought to the field, however, by two recent phase 2 clinical trials reporting positive data. First, prevention of complement C3 cleavage with pegcetacoplan resulted in statistically significant reductions in GA growth by 20%-29% at 1 year, which appeared to increase with longer drug exposure and be dose dependent.[6] Second, inhibition of the downstream complement component C5 with avacincaptad pegol (Zimura) similarly resulted in statistically significant reductions in GA by about 27% at 1 year.[7] Both of these molecules are being investigated in phase 3 trials.
The Moment for Gene Therapy Arrives
Since the 1990s, we’ve heard that the widespread clinical application of gene therapies and cell-based therapies was just around the corner, only to have our expectations dashed. It may be that 2019 will mark the milestone moment when our expectations were actually met.
Two gene therapies have already been approved in the United States: onasemnogene abeparvovec (Zolgensma), a gene therapy for the treatment of spinal muscular atrophy,[8] and voretigene neparvovec-rzyl (Luxturna), a gene therapy for biallelic RPE65-attributable Leber congenital amaurosis.[9] Although both are incredibly rare diseases (Zolgensma was essentially approved based on a trial involving just 15 patients), the implications of their approvals are enormous.
In the context of exudative retinal diseases, the lack of extended durability of current anti-VEGF agents is a major clinical limitation, leading to gross underdosing in patients that can lead to suboptimal visual outcomes. Employment of a gene therapy created to durably express an anti-VEGF agent intraocularly has the potential to dramatically reduce the need for ongoing, repeated anti-VEGF intravitreal injections and improve real-world outcomes.
Toward this aim, two major ongoing, multiple-cohort trials reported exciting, first-in-human data in 2019. REGENXBIO shared data from a phase 1/2a study assessing a single subretinal surgical delivery of RGX-314 (adeno-associated virus [AAV] serotype 8), which expresses a protein similar to ranibizumab.[10] Adverum Biotechnologies reported data from a phase 1 study assessing a single intravitreal injection of ADVM-022 (AAV.7m8), which expresses a protein similar to aflibercept.[11] Both trials enrolled treatment-experienced patients with nAMD and have reported reductions in the need for ongoing dosing while maintaining or improving visual and anatomic outcomes. This is new territory, and these trials have and will continue to face development hurdles, as evidenced by the clinical holds[12,13] both companies encountered through 2019.
The Great Promise of AI in Diabetic Retinopathy
Diabetic retinopathy is a leading cause of blindness globally and is rising in prevalence.[14] Diabetic retinopathy leads to vision loss primarily through the development of proliferative diabetic retinopathy (PDR) or diabetic macular edema (DME). Severe NPDR without DME affects about 3 million Americans, often with excellent vision. Although the risk of these patients developing PDR or DME is recognized,[15] there remains a high clinical hurdle to initiate intravitreal anti-VEGF injections or laser treatment to a largely asymptomatic population.
PANORAMA, the only modern prospective trial involving patients at high risk for NPDR without DME, provides valuable data to inform patient discussions and decision-making in the anti-VEGF era. Through 1 year, about 40% of untreated patients developed PDR or DME, a rate reduced by approximately 75% with fixed aflibercept dosing either every 2 or 4 months after loading doses.[16]
As our field continues to consider, and shift toward, treatment initiation at earlier stages of diabetic retinopathy, improved predictability is of paramount importance. Toward this end, I am committed to the promise of AI and deep-learning algorithms. It is logical to treat diabetic retinopathy before advanced stages of the disease are manifest and visual loss has occurred. The greatest challenge is not how to treat but whom to treat and when. Isolated fundus images can deduce age, sex, blood pressure, smoking status, and history of a major cardiac event with remarkable accuracy.[17] Similarly, algorithms capable of detecting diabetic retinopathy and its worsening severity as accurately as trained ophthalmologists have been reported.[18] These systems will unquestionably improve prognostic granularity as they are further refined and eventually introduced clinically.
Charles C. Wykoff, MD, PhD, is director of clinical research at Retina Consultants of Houston. His research interests pertain to angiogenesis and retinal vascular diseases, including age-related macular degeneration, diabetic retinopathy, and venous occlusive diseases, as well as vitreoretinal surgery topics, such as retinal detachment and macular surgery.
No comments:
Post a Comment
Note: Only a member of this blog may post a comment.