DAVE study found little benefit of anti-VEGF/PRP for DME

April 12, 2017

The scientific community knows that vascular endothelial growth factor (VEGF) causes increased, vascular permeability, resulting in diabetic macular edema (DME) in the ischemic retina, but how to stop the VEGF drive remains the challenge.

The scientific community knows that vascular endothelial growth factor (VEGF) causes increased, vascular permeability, resulting in diabetic macular edema (DME) in the ischemic retina, but how to stop the VEGF drive remains the challenge.

“It makes sense, then, if we could obliterate the ischemic peripheral retina, that we could decrease the VEGF drive, and thus decrease the need for anti-VEGF injections,” said Rosa Kim, MD. Several published papers seem to suggest the benefit, “but none of these papers had a control group.”

Unfortunately, “with the information that we have, we would not–at this time–recommend targeted laser photocoagulation” as a means to reduce treatment burden in patients being treated with an anti-VEGF, said Dr. Kim, who is in private practice in Houston.

Study details

The Efficacy & Safety Trial of Intravitreal Injections Combined with PRP for CSME Secondary to Diabetes Mellitus (DAVE) study is an FDA-approved, randomized control trial with the following inclusion criteria:

Patients had to have center-involving DME, with a visual acuity between 20/32 and 20/400, and at least 50 disc areas of capillary non-perfusion. (Of note, neovascularization was not an exclusion criterion.) Patients were randomized to receive intravitreal ranibizumab 0.3 mg (Lucentis, Genentech) PRN (n = 20) or in combination with ultrawide 200° field angiography panretinal photocoagulation (PRP) (n = 20).

The monotherapy group received 4 monthly injections, was evaluated monthly, and received additional PRN injections only if the presence of DME on exam or through optical coherence tomography (OCT) confirmed. In the combination arm, patients received the same loading doses, but received targeted PRP at day 3, and possibly at month 6 and 18, if there were additional areas of ischemia on wide-angle angiography.

The primary outcome measures were the mean best-corrected visual acuity (BCVA) change from baseline and the difference in total number of intravitreal injections between the cohorts.

Baseline demographics were not different in either group. Mean age of the enrolled subjects was 55 years (ranging from age 31 to 75), the mean BCVA was 20/63 (Snellen equivalent on the ETDRS scale), and the mean central retinal thickness was 530 µm.

“There are two pertinent baseline points,” Dr. Kim said. “Patients’ mean hemoglobin A1c was 8.0 in the ranibizumab-only group and 8.8 in the combination group, which is an average test in diabetic population. These patients had significant DME, ranging from 496 µm in the combined group to 579 µm in the ranibizumab-only group.”

Dr. Kim said the investigator was initiating therapy based upon the DME, but not really determining how much non-perfusion there was. If the protocol did not specify foveal-involving DME as a criterion for re-treatment, it is possible more patients would have required injections.

Study outcomes

There were 16 patients in the ranibizumab-only group and 19 in the combined group with data available through 24 months. Out of the potential visits (492 in the ranibizumab-only group and 517 in the combined group), an average 10% were missed in each group. On OCT, both groups showed a typical anti-VEGF response with a reduction in edema; these differences were not statistically significant.

“Likewise, on average, patients gained 10 letters in both groups,” Dr. Kim said. “When we analyzed 15-letter gainers, 40% in monotherapy versus 30% in the combination arms gained 15 or more letters, but again this was not statistically significant.”

The group now has 80% of original subjects who have completed Year 3, Dr. Kim said.

But the “big” question was whether the targeted laser treatment made a difference in decreasing the number of injections needed to achieve those gains.

“The bottom line is, no!” Dr. Kim pointed out. “The combination arm actually had a slightly higher percentage of PRN injections than the monotherapy, demonstrating that adding laser treatment did not reduce the injection burden for DME.”

Patients in the monotherapy group received an average of 18.3 injections compared to an average of 19.7 injections in the combination group. In Year 3, the combination group had a 73% rate of re-injection (down from 80% in Year 2) compared to 64% in the ranibizumab-only group in Year 3 (down from 73% in Year 2).

There are some theories about why adding PRP did not reduce the burden, Dr. Kim said. For one, there is an average of 92 million photoreceptors in the human retina, a majority of which are located posteriorly.

“Since we don’t treat the posterior pole, we’re preserving about a half of photoreceptors,” Dr. Kim explained. “Even if you obliterate a significant amount of peripheral retina, about half the photoreceptors are unaffected. Perhaps to decrease the VEGF load, we would have to destroy valuable retina, which we’re not willing to sacrifice.”

However, the combined treatment was safe over the study timeframe, and did provide similar anatomic and visual acuity benefits. What that means for the role of PRP moving forward remains unanswered, she said.


Rosa Kim, MD

P: 713-524-3434

E: rosa.kim@houstonretina.com

Dr. Kim receives grant support from the National Eye Institute, Novartis, and Santen. This article was adapted from a presentation Dr. Kim delivered at Retina Subspecialty Day prior to the 2016 American Academy of Ophthalmology meeting.