OCTA may be harbinger of anti-VEGF efficacy

(FIGURE 2) The effect of bevacizumab (bottom row) on the CNVM complex is shown and compared with ranibizumab (top row). VA has not changed, but bevacizumab has little to no effect on the CNVM complex. (Image courtesy of P. Tornambe, MD)

Reviewed by Paul Tornambe, MD

Sometimes a single case example can provide food for thought and act as a springboard for future clinical studies. That’s exactly what Paul Tornambe, MD, of San Diego, hopes will happen from his findings on optical coherence tomography angiography (OCTA), choroidal neovascular membranes (CNVM), and treatment recommendations.

His case study results were presented at this year’s American Society of Retina Surgeons conference and suggested bringing a patient in 1 week after their first anti-vascular endothelial growth factor (VEGF) injection to undergo OCT can provide insights about the membrane and its likelihood to respond.

“The decision to start treatment, switch to another, or stop treatment altogether is sometimes arbitrary and capricious,” he said. “We really don’t know before starting treatment that this anti-VEGF drug works best for this specific neovascular membrane.”

Using OCT earlier

An 80-year-old Caucasian male complained of a new scotoma and distorted vision nasal to fixation in the right eye. Visual acuity was 20/25.

He had been treated with “all well-known treatment modalities for exudative macular degeneration,” Dr. Tornambe said-including 13 injections of bevacizumab, 13 of ranibizumab, five of aflibercept, two treatments of reduced fluence photodynamic therapy (PDT), and two intravitreal triamcinolone acetate injections- over the last decade and went into a remission for 18 months. The patient’s disease re-activated in Feb. 2017, where he was placed on a treat-and-extend every 2-3 months regimen. On August 19, 2017, he received PDT/aflibercept.

Dr. Tornambe said treatment was initiated after an OCT with an injection of aflibercept; the patient then had weekly OCTs for 3 weeks. The protocol was repeated every 4 weeks using ranibizumab and then bevacizumab. The patient agreed to monthly injections of a different VEGF inhibitor drug and weekly OCT A and B scans performed with an Optovue unit.

At week 1, OCT showed a “prompt effect of aflibercept on the temporal portion of the membrane and an almost complete resolution of subretinal fluid,” Dr. Tornambe said. By week 2, the neovascular membrane reperfused, but there was still no subretinal fluid. At week 4, visual acuity remained at 20/25. But OCT showed the temporal net had reperfused and subretinal fluid accumulated.

Reperfusion frequently develops before subretinal fluid accumulates, and the “very early effect of the drug at one week” is readily seen on OCT.

“This case clearly shows a neovascular membrane’s response to VEGF-inhibitors is not homogenous, occurs very early after the injections, and that only a portion of the membrane may be VEGF-inhibitor responsive,” he said. (See Figures 1 and 2.)

When the patient then underwent treatment with ranibizumab, the OCT showed the membrane to be the same as it was before the aflibercept injection. At week 1, there was less perfusion of the temporal complex, “but there was also persistent subretinal fluid,” Dr. Tornambe said. Weeks 2 and 4 showed approximately the same as when the patient had been given aflibercept.

“There appears to be no difference between these two drugs for this particular case,” Dr. Tornambe said. When the patient was injected with bevacizumab, however, there “appeared to be little to no effect on the neovascular complex for this specific case. Visual acuity did not change.”

In typical clinical settings, however, OCTs are often performed only at baseline at month 1/week 4, “So in this instance, there were no discernible difference between bevacizumab and ranibizumab, even though there was a significant ranibizumab effect at week 1,” Dr. Tornambe said.

Clinical relevance

“Mature membranes don’t respond to VEGF Inhibitors, just those that are growing and those that are immature. And that’s really important because it makes no sense to keep injecting a VEGF inhibitor if the membrane’s mature,” Dr. Tornambe said.

Further, there is only one way to tell which membranes are mature, and that is with an OCT in the first post-injection week.

The greatest drug effect can be seen consistently during the first week following an anti-VEGF injection, “and reperfusion may develop prior to subretinal fluid accumulation, possibly signaling the need for another injection, thus a possible marker to treat sooner.”

Week 1 is key

If there had been no OCT at week 1, “we would never have known there was any effect,” he said. “That is a very, very important finding because Centers for Medicare & Medicaid Services only pays for OCTs one month apart. Yet we’ve shown the 1-week time point is most clinically relevant.”

Dr. Tornambe’s group has now performed a case series and findings are mimicking this first case. If the 1-week OCT shows no effect on the membrane, it might suggest switching to another treatment or stopping injection treatments with close follow-up studies to search for recurrences elsewhere, he said.

“This methodology may prove to be the most patient-centered, customized, cost effective and rational way to determine drug selection, follow-up, and switching or stopping treatment for a specific neovascular membrane in a specific macular degeneration eye,” he said.

In the ongoing push towards treat-and-extend, “we are always looking for earlier markers for re-treatment,” he said. “In some of these cases, reperfusion has already occurred at week 3, and then 2-3 weeks later fluid returns. That’s what we’re trying to avoid.”

His advice is to take an OCT after week 1 (and if there is no change on OCT, change to a different anti-VEGF), and retreat when the vessel becomes reperfused, not when fluid returns.

Until there is a way to measure VEGF levels non-invasively, Dr. Tornambe believes early OCTs are the most effective way to measure treatment efficacy. He did caution that “we absolutely cannot generalize from this single case that bevacizumab is an inferior drug-it just did not work well on this particular CNVM. We must also ensure the OCT slices through the same area of the CNV complex, otherwise changes noted in the CNV perfusion may be due to an artifact.”

Disclosures:

Paul Tornambe, MDE: tornambepe@aol.com
Dr. Tornambe does not have any financial disclosures related to his comments.