Can Anti-VEGF Trigger GA?

The issue is spurring a lively debate.

Jerry Helzner, Senior Editor

One of the most puzzling and controversial findings to come out of the two-year CATT study data is the presence of geographic atrophy (GA) in 30% of the patients who received monthly doses of Lucentis during the trial. Some patients who received Avastin monthly or Lucentis or Avastin PRN also showed signs of GA but not at the high levels seen in the Lucentis monthly arm of the trial.

The association of GA with the most potent dosing regimen of CATT has led at least two highly respected retina specialists, Drs. Philip Rosenfeld of Miami and Lawrence Singerman of Cleveland, to express their concern that repeated anti-VEGF injections can serve as a trigger mechanism for GA. Dr. Rosenfeld has called the GA findings "one of the most interesting outcomes" from the entire CATT trial. But other equally highly respected retina specialists are reserving their opinion on whether anti-VEGF can trigger GA pending further corroborative data.

Dr. Rosenfeld told Retinal Physician he had suspected a possible anti-VEGF/GA association while following patients in the Lucentis trials, including his own Pronto study. In his retrospective review of all patients who participated in the pivotal ANCHOR and MARINA trials for Lucentis and published last year in the journal Ophthalmology (Ophthalmology 2011;118:523-530), Dr. Rosenfeld explored the causes of vision loss among the 10% of the patients in these trials who lost at least 3 lines of vision at the 24-month mark. He found that retinal pigment epithelium (RPE) abnormalities indicative of GA were associated with this vision loss. At the time, Dr. Rosenfeld says he thought the changes were most likely signs of normal disease progression once the CNV was controlled; however he says in light of the CATT results, the other possibility that needs to be considered is that too much VEGF inhibition might promote the formation and growth of GA.

In his 2011 article, Dr. Rosenfeld concluded that "vision loss after two years of monthly ranibizumab therapy was associated with lesion characteristics commonly associated with suppressed CNV, such as pigmentary abnormalities, atrophic scar, and the absence of leakage. He added that "future VA improvements in patients receiving ranibizumab therapy may require preservation of photoreceptor and RPE function rather than strategies that target CNV."

As a result of his analysis of the MARINA, ANCHOR and the results of the CATT, Dr. Rosenfeld has come to the conclusion that, given the current lack of an effective treatment for GA, anti-VEGF therapy should be used judiciously following a "treat and extend" protocol. In this way, he says the significant and immediate visual benefits of anti-VEGF can still be achieved. He sees treat-and-extend dosing as possibly a necessary real-world trade-off to reduce the possible risk of GA until researchers know whether the GA results of the CATT trial can be reproduced.

"While blockage of VEGF can stop the growth of blood vessels and improve vision in the short-term, too much VEGF inhibition may cause the dry AMD to progress," Dr. Rosenfeld states. "I refer to treat-and-extend dosing as the Goldilocks approach; we don't want too much VEGF inhibition, we don't want too little, we want just the right amount of inhibition."

Though almost all retina specialists tend to support the "treat and extend" protocol as an effective and patient-sparing regimen, the possible role of anti-VEGF injections in triggering GA remains highly controversial. Dr. David Boyer of Los Angeles calls the theory "speculation." He notes that "we should be aware that GA was not diagnosed by autoflourescence and may be related to the difference of drying that Lucentis was shown to do more frequently than Avastin. GA can also be measured by OCT but is very difficult to measure when you have glial scarring."

Others have gone so far as to say the GA images are meaningless artifacts that carry no real indications of new disease. However, mouse studies conducted by Patricia D'Amore, PhD, of Schepens Eye Institute and Martin Friedlander, MD, PhD, of the Scripps Research Institute, were early in indicating that attacking VEGF aggressively could produce the unintended consequence of producing GA.

Asked by Retinal Physician to comment on meaning of the CATT findings, a spokesperson for Genentech e-mailed that "we believe the GA finding in CATT needs to be corroborated with other phase 3 wet AMD studies. We are in the process of analyzing our own data to determine whether we see the same GA finding or not." Also adding that "the finding is not definitive and requires further research."

A spokesperson for Regeneron told Retinal Physician that the incidence of GA was not evaluated during the pivotal VIEW trials but that the company is open to further evaluation of the VIEW trials, assuming that any relevant information could be gleaned from the fundus photography and angiography used in that trial.

If anti-VEGF therapy is proven to be associated with the formation and progression of GA, it would be a boost to those companies pursuing alternatives to anti-VEGF or effective treatments for GA.