RETINA CONVERSATIONS: A Conversation With Philip J. Rosenfeld, MD, PhD

A retina specialist dedicated to research into therapies for AMD.


Dr. Rosenfeld visits the US Capitol to speak about return on investment in OCT development.

Philip J. Rosenfeld, MD, PhD, of the Bascom Palmer Eye Institute spoke with Retinal Physician about recently published results of a return-on-investment (ROI) study sponsored by The Association for Research in Vision and Ophthalmology (ARVO), and some current directions in research.

Q. The American Journal of Ophthalmology just published some very impressive results of an ROI study, for which you were one of the coauthors.1 This type of study seems particularly important today, when funding for basic science can be scarce. Could you describe the study?

A. ARVO deserves enormous credit for supporting this research, which unequivocally demonstrated the importance of basic science funding, and the return-on-investment for optical coherence tomography (OCT). Members of Congress immediately understood the importance of our message. The impetus for the study was the 25th anniversary in 2016 of the invention of OCT in 1991 and the first publication on OCT that came out in Science. This was a big deal for ARVO, patients, and clinicians because OCT truly was a wonderful basic science discovery that had real clinical benefits.

We set out to show that the discovery resulted in a significant return on the initial investment in the technology. Based on our calculations, using Medicare databases from 2008 through 2015, we found an initial investment in OCT technology of about $400 million, which resulted in an $11.2 billion return to government and patients from the use of OCT to manage patients with exudative AMD. The savings amounted to a $9 billion return to Medicare and a return of approximately $2.2 billion dollars to patients, because Medicare pays 80% of the cost of any clinical expenditure and patients cover the remaining 20%. Our study was based on 100% of the fee-for-service beneficiary databases, which comprises about 70% of Medicare-eligible patients. Once patients get into a Medicare HMO or Advantage plan, they don’t appear in the Medicare databases, they appear in the databases of the private insurers. So, if we assume that OCT-guided therapy was used similarly when treating the 30% of Medicare patients in Advantage plans, then the overall cost savings is closer to $15 billion.

We examined how much anti-VEGF therapy was used in the real world by looking at how many patients were diagnosed with wet AMD, the number of injections, and how much money was spent by Medicare. Then, we looked at those patients and asked, “How much would Medicare have spent if these patients had received ranibizumab or bevacizumab monthly, or aflibercept every other month as recommended by the industry? We assumed ranibizumab and bevacizumab would have been dosed monthly. Then, we compared this amount with the amount Medicare spent using OCT-guided therapy, which is the real-world amount. We found that the difference in these 2 figures amounted to about $11.2 billion. The details of our study and calculations are explained in the paper.

I hope our paper will inspire others to pursue this type of research. A similar ROI strategy could be followed for the treatment of diabetic macular edema, or for retinal vein occlusion. In fact, we just submitted a follow-up paper looking at the cost savings from bevacizumab between 2008 and 2015. The message is that investment in basic science is critical regardless of whether we can demonstrate a corresponding reduction in health care costs — but we were fortunate that we could easily show an enormous cost benefit from the basic science support of OCT research.

Q. Those are stunning results. And you took a trip to Washington, DC in December to share the results of the study?

A. We went to Washington to speak to our elected officials in Congress and present the results of the ROI study. The trip was sponsored by ARVO, and we were warmly received on Capitol Hill. Our message received bipartisan support.

Q. You have discussed expanding opportunities for medical retina specialists. Now it seems that there may be new revolutions on the horizon. You must be excited about getting closer to a treatment for dry AMD.

A. Now I’m totally focused on nonexudative or dry AMD. About 10 years ago, it became obvious to me that patients with wet AMD undergoing anti-VEGF therapy were losing vision from the underlying dry AMD, not the exudation. Therefore, the only way to prevent long-term vision loss was to treat the dry AMD. Right now, I’m very excited about the Apellis [Pharmaceuticals] drug known as APL-2, which is a complement inhibitor. I’ve been involved with this research for more than 10 years. I was a phase I investigator when this research started on the drug, then known as POT-4, and the company was called Potentia. The drug was later licensed by Alcon, and then Alcon was acquired by Novartis. When Novartis returned the drug to Potentia, the company had renamed itself to Apellis and had begun studying complement inhibition in nonophthalmic diseases. Apellis had retained the rights to studying the drug for everything outside the eye. At that point, I was involved in helping to raise financial support for Apellis, and I became an investor. So, because I’m now an investor in this company, the university has asked me to step back from the clinical research.

Q. But you are no doubt following the progress closely.

A. Yes, and I can provide technical support in the trial by injecting the drug. The phase 2 study in patients with geographic atrophy (GA) showed positive results, and these results were announced last fall. This is the first drug that has truly been shown to slow the enlargement of GA lesions in eyes with dry AMD. We now anticipate that a phase 3 study will start in late spring 2018.

Q. There has been speculation that GA is triggered by anti-VEGF injections. For example, the CATT data seem to indicate that more injections lead to a higher rate of GA. What are your thoughts on that?

A. That is a question without an answer right now. When we treat exudative AMD with injections, we are converting exudative AMD to nonexudative AMD, but we are not treating the underlying disease. Whether the injections promote atrophy or whether the underlying disease promotes atrophy is a question that cannot be easily answered. When nonexudative or dry AMD progresses along its normal course, we are finding that nonexudative neovascular lesions arise months before exudation, and these nonexudative neovascular lesions may protect against the formation of GA. So, it’s unclear whether we want neovascularization to go away or just stop leaking. When neovascularization goes away, we’re left with atrophy. That seems to be the end stage of AMD, and someone with the disease who lives long enough will almost always develop atrophy. So, does anti-VEGF therapy cause atrophy? Most likely, it’s the AMD that causes atrophy, and anti-VEGF therapy prevents the catastrophic vision loss that occurs from exudation, bleeding, and scar formation. We don’t have a choice — we have to treat with anti-VEGF therapy. Now, if we have a treatment to prevent atrophy, then we can preserve vison long term.

Q. And what are your thoughts about gene therapy? Spark Therapeutics just had success with a gene therapy.

A. Yes, that’s a remarkable achievement. But, of course, it’s not relevant to AMD. While many human diseases, such as Alzheimer disease, AMD, and multiple sclerosis are late-onset diseases that are thought to have a strong genetic component, they are not amenable to gene therapy. However, they are amenable to targeted therapies such as complement inhibition. Genome-wide association studies have clearly implicated the complement cascade as playing a very important role in causing AMD, and this association is the strongest association of any human locus with a complex human disease.

Q. So you were excited when you learned complement was associated with AMD.

A. When I heard that complement was associated with AMD, it became immediately obvious to me that there was no better disease than AMD to demonstrate the usefulness of these genome-wide association studies. The holy grail of genetic research is to identify a gene or a pathway involved in human disease and then to target that pathway in the hopes of modifying diseases severity. For late-onset complex human diseases, this has never been achieved, and complement inhibition in AMD may be the first example of achieving this milestone.

Q. There also has been some exciting progress in the study of macular telangiectasia type 2 (MacTel). You are one of the primary investigators, and that research is moving toward a phase 3 trial, correct?

A. Yes, the progress of that study was recently outlined in a Retina supplement. The story behind this research is wonderful, and the Lowy Institute, which purchased Neurotech, has supported the entire international endeavor. We’ve discovered over time that this disease is not quite as rare as we initially thought. We have more than 100 patients in our natural history clinical trial at the Bascom Palmer Eye Institute with new patients being discovered all the time. This study has been a wonderful example of how clinical research can be used to identify unique features of a disease and then design clinical trials with novel endpoints to test different therapies. It has been a wonderful convergence of basic science research and clinical research to address an unmet need.

Q. I want to ask you your thoughts about what may be a less promising direction in research, at least at the moment, and that is stem cell therapy. You were one of the coauthors of the New England Journal of Medicine article last year, which revealed some pretty devastating results of stem cell procedures.2

A. That paper described a rogue organization that was engaging in what appeared to be criminal activity. Consequently, stem cell therapy has become a controversial topic. We have to move forward with the research, but I’m less than optimistic about some avenues of the research. In macular degeneration, tissues in the back of the eye — not only cells, but also the supporting vasculature — are disappearing. The remedy for the disease has to involve more than just replacing cells — we have to replace the entire complex of the supporting structure and the vasculature. I like to use the analogy of a sink hole. We have a lot of sink holes here in Florida — our state is built on porous lime stone. A loss of water and limestone can suddenly cause the ground to cave in, and sink holes can swallow roads and buildings. In AMD, there’s a sink hole developing in the macula, and using stem cells is like knowing that a sink hole is developing and building a house right on top of it. The house is not going to survive.

Q. What would make research in this area more promising?

A. There are projects under way that involve growing cells on a scaffold. That provides a lasting structural support for the cells. That method has a greater likelihood of success, compared with just putting cells in the back of the eye.

Q. And for one last question, what directions do you think your own research will take in the future?

A. I’m interested in looking at a range of therapeutic strategies for macular diseases. What I love to do is help scientists and companies move from late-stage preclinical work, and lab work, into early-stage human clinical trials. I enjoy developing proof-of-concept clinical trial designs for unmet clinical needs. Right now, I’m working with a number of companies on early-stage molecules, and most of the rest of my time is occupied exploring the enormous potential of using swept-source OCT angiography for understanding disease progression in AMD. This will allow us to develop better clinical trial endpoints for testing novel therapies. RP


  1. Windsor MA, Sun SJJ, Frick KD, Swanson EA, Rosenfeld PJ, Huang D. Estimating public and patient savings from basic research-a study of optical coherence tomography in managing antiangiogenic therapy. Am J Ophthalmol. 2018;185:115-122.
  2. Kuriyan AE, Albini TA, Townsend JH, et al. Vision loss after intravitreal injection of autologous “stem cells” for AMD. N Engl J Med. 2017;376(11):1047-1053.