UPFRONT

Protocol TBD?

Peter K. Kaiser, MD

In our diabetic retinopathy issue, it is only fitting that we finally have the results of the DRCRnet’s Protocol T (see Subspecialty News, page 11). But the conclusions of the paper were very strange. The mean changes in vision were 13.3, 9.7, and 11.2 for Eylea, Avastin, and Lucentis, respectively, after one year of a protocol PRN treatment regimen.

Overall, the differences were statistically significant for both comparisons with Eylea (P<.001 and P=.03). The difference in the number of injections was also significant (P=.045), and the number of rescue laser treatments was significantly different (P<.001).

I won’t discuss safety, which was also significantly different, because the study was not powered to find a safety difference and I think the difference found was due to chance and not a real difference in adverse events, given the results of other comparative studies to date.

So we have one drug that had better visual outcomes and that required fewer injections and rescue laser treatments, yet the conclusions of the paper do not state these conclusions. Rather, the paper states, “the magnitude of the greater effect of aflibercept lacked clinical applicability because it was dependent on initial visual acuity.” Huh?

In prespecified subgroup analysis, there was an insignificant difference seen in patients with good vision but a large and very significant difference in patients who started with poor vision (20/50 or worse) or thicker retinas (>400 μm center subfield).

Thus, the guidelines set forth in the accompanying editorial seem sound — start with Avastin in patients with good vision and with Eylea in those with poor vision. But what do you do in a patient with good vision and retinal thickness greater than 400 μm?

Moreover, how does visual acuity measured by Snellen notation compare to the forced choice ETDRS protocol vision performed in the study? In short, what truly is “good vision” in clinical practice?

There are other questions that will remain well beyond the study. The most obvious is that the study used 0.3-mg Lucentis, which is the approved dose in the US, but outside the US, the approved dose is 0.5 mg, so how applicable are the results worldwide? Is the efficacy of 0.3 mg and 0.5 mg the same? The RISE and RIDE studies suggested that they are, but remember these trials used fixed dosing.

Finally, we are faced with the dilemma regarding when to switch medications in patients with good vision responding poorly on Avastin or in those with bad vision responding poorly to Eylea.

I look forward to more in-depth analysis of the Protocol T data set and the papers that will surely be coming. It is a groundbreaking study, and I am glad that we finally have guidance regarding what we should choose to best treat our patients with DME.