Retina specialists frequently use expensive drugs to manage exudative AMD. The choice of drug is determined by clinical trial data, which might pool results from thousands of patients, and our personal experience. As physicians, we know that some drugs work better for one patient than for another. Is there a better way to select a drug for a specific eye? Can we customize treatment for better outcomes and perhaps reduce costs without jeopardizing quality?
By enabling visualization of the retinal vasculature noninvasively, OCT angiography (OCTA) opens the door to significant changes in patient management. Because OCTA is safe and efficient, it can be performed as frequently as desired. Furthermore, as I’m learning in my practice, the technology can be utilized to customize care. It provides a way to compare the effects of different VEGF inhibitor (VEGF-I) drugs on individual AMD patients and choroidal neovascular membranes (CNVMs), so the most effective drug can be chosen.
With OCTA, flow through choroidal neovascular vessels is monitored directly. This is, of course, different from traditional monitoring, which relies on the surrogate parameters of subretinal fluid on OCT B-scans or leakage on fluorescein angiography (FA). As I began following my AMD patients with serial OCTA scans, I realized I was observing that CNVMs don’t go away. Rather, they mature, stop leaking, and become unaffected by VEGF-I drugs. Even when a membrane is no longer detectable by FA, OCTA can verify that it’s still perfused. It can also show when new blood vessels begin budding off a mature vessel. I’ve found, too, that it’s the new vessels, which are visible as tiny buds at first, that leak as shown by FA. In addition, the new, immature vessels tend to re-perfuse earlier and sometimes grow far sooner than I would have previously expected after a VEGF-I injection.
The best time to determine how a specific VEGF-I affects a specific CNVM is within 1 week after an injection. Although OCTA scans within 1 week are unfortunately not afforded CMS coverage, I perform them anyway (at no charge), as part of the following protocol. I obtain OCTA images on the day of or a few days prior to a VEGF-I injection. I begin treatment with the least expensive drug, bevacizumab. During days 3 to 7, when the maximum drug effect can be seen, I use OCTA to determine which portion, if any, of the CNVM is VEGF-I responsive. If subretinal fluid is unchanged and I observe no effect on the membrane, I consider switching to a different VEGF-I (Figures 1 and 2). If I do see a response to the VEGF-I (Figures 3 and 4), I use OCTA at 4 weeks to assess the drug’s durability. If the subretinal fluid is gone and the VEGF-I-responsive portion of the membrane remains nonperfused, I inject the same drug again and check the response at 6 weeks. If at that time the CNVM is perfused and the fluid returns, I consider switching to a different VEGF-I, which might last longer, and repeat the process by checking the response by day 7 (to determine whether the new VEGF-I has an effect) and 4 weeks (to assess durability). If the CNVM is perfused but fluid hasn’t returned — which suggests the vessel might be maturing — I repeat OCTA at 6 weeks and then weekly until the fluid returns. If the fluid returns sooner than 2 months, I may add PDT or dexamethasone to the VEGF-I and follow the same timeline for monitoring. I schedule each of the monitoring visits as “fast track OCTA.” No pupil dilation is needed, and the patient sees only a technician, who performs the OCTA scans.
With this approach, I can provide truly customized care for each patient, which often translates to treating earlier (and presumably more effectively), treating less frequently, and stopping treatment (which would likely be ineffective) more often.
OCTA can be used to determine which VEGF inhibitor is most effective for a specific eye. OCTA scans obtained at 1 week following injection of a VEGF inhibitor show that monitoring choroidal neovascular membranes at 4-week intervals can be misleading.
Figure 1. Bevacizumab injected (L) and scan repeated 4 weeks later (R). CNVM same and SRF same ... no effect?
Figure 2. Bevacizumab injected (L) and scan repeated 1 & 4 weeks later (M,R). The 1-week scan shows no beneficial effect. Time to switch drugs.
Figure 3. Ranibizumab injected (L) and scan repeated 4 weeks later (R). CNVM same and SRF same ... no effect?
Figure 4. Ranibizumab injected (L) and scan repeated 1 & 4 weeks later (M,R). The 1-week positive effect is missed at 4 weeks; continue ranibizumab.
Dr. Tornambe, a frequent author and ophthalmic innovator, is founder and president of Retina Consultants San Diego.