18.97.14.80
dgid:
enl:
npi:0
/nas/content/live/oph360live/wp-content/themes/rd360/single.php
-Advertisement-
-Advertisement-
Conference Roundup
Video

Matching study assesses faricimab versus aflibercept in DME and AMD

Posted on

Jennifer Lim, MD, of the University of Illinois at Chicago, discusses her study presented at the 2024 AAO Annual Meeting, titled “Comparison of the Relative Effectiveness of Faricimab vs. Aflibercept 8mg in DME.”

Question:

Can you discuss the design of the study and its findings?

Jennifer Lim, MD:

It was my pleasure to present this comparison of faricimab to aflibercept 8 mg for both neovascular AMD and diabetic macular edema recently at the American Academy of Ophthalmology in Chicago this year.

We now have really great new drugs. We have faricimab 6 mg and aflibercept 8 mg that give us durability in the treatment of our patients with diabetic macular edema and neovascular age-related macular degeneration. These 2 drugs were approved based on the phase 3 clinical trials. In the case of faricimab for diabetic macular edema, these were the YOSEMITE and RHINE trials. For aflibercept 8 mg, this was based on the PHOTON study.

There was no direct comparison, however, of faricimab to aflibercept for diabetic macular edema, and we know cross-trial comparisons are difficult due to different treatment populations as well as different treatment paradigms.

However, there is a way to compare cross trials, which is a statistical method that mitigates cross-trial differences in terms of inclusion-exclusion criteria when comparing patient-level data from one study with the published aggregated data from the comparator treatment.

In this case, we are able, therefore, to use this matching-adjusted network meta-analysis, or NMA, to compare the results of YOSEMITE and RHINE to those of PHOTON, and this is what we did in my study.

In this study, we went ahead and matched the baseline characteristics for the YOSEMITE and RHINE trials to those of PHOTON. The way we do this is we match the baseline characteristics from the PHOTON study to those of the patient-level data in YOSEMITE and RHINE so we apply weights to these characteristics.

For example, in the PHOTON Study, the patients had, in general, thinner retinas in terms of CST, they had greater percentages of eyes that had had prior anti-VEGF therapy, and the DRSS levels were slightly lower in PHOTON compared to YOSEMITE and RHINE.

After we apply these characteristics to the entire YOSEMITE and RHINE database, we then create a new baseline set of patients, and then we reanalyze the data to see what were the best-corrected visual acuity outcomes and what were the CST outcomes for this new group.

Once that is done, we can then compare the mean differences for visual acuity and CST between the faricimab and the aflibercept-treated eyes.

When we did this matching adjusted network meta-analysis, we found in terms of best-corrected visual acuity that there was no difference between the faricimab-treated eyes and the aflibercept 8 mg-treated eyes.

However, when we looked at the central subfield thickness, we did find that the eyes treated with faricimab had better drying. That is, the CST was 19 microns thinner compared to the aflibercept 2 mg arm. It was also 19 microns thinner compared to the aflibercept 8 mg arm, and this was a statistically significant difference.

This analysis shows us that the faricimab-treated eyes were able to be drier compared to the aflibercept-treated eyes. This is only possible because both of the treatment comparisons in terms of the PHOTON study and then in terms of YOSEMITE and RHINE used a comparator arm of aflibercept 2 mg. That is the common anchor that allows this comparison to be done.

We were also curious to see how does faricimab 6 mg compare to aflibercept 8 mg for patients treated for neovascular AMD. In this case, we matched the TENAYA and LUCERNE-treated eyes to those of the PULSAR study. Once we did this, we were able to adjust for the fact that the PULSAR-treated eyes had slightly larger lesion sizes and had slightly smaller percentages of eyes that had absence of subretinal fluid or intraretinal fluid.

Once this matching was done, we again formed a new baseline collection of patients in the TENAYA and LUCERNE that matched the characteristics in the PULSAR study. Using the matching adjusted network meta-analysis, we showed that the faricimab-treated eyes had similar best-corrected visual acuity to those treated with aflibercept 8 mg.

However, when we looked at the central subfield thickness, we found that the faricimab-treated eyes, in general, had thinner retinas after the treatment was done. This difference was 18 microns in comparison to aflibercept 2 mg, and 17 microns in comparison to aflibercept 8 mg.

It’s very important, however, to realize that when we did this comparison, we compared at week 12. That is, the time point when the faricimab-treated eyes and the aflibercept-treated eyes had had the same number of treatments of their respective treatment arms so there was no difference in the number of treatments given for either drug.

In summary, this matching adjusted network meta-analysis did allow us to compare across the clinical trials for the faricimab and aflibercept-treated eyes in phase 3 trials and did show that the faricimab-treated eyes were able to be drier compared to the aflibercept 8 mg and even aflibercept 2 mg-treated eyes.

Question:

How do you think this research might impact treatment decision-making?

Jennifer Lim, MD:

I think it does lend to some evidence with good statistical comparisons that perhaps the faricimab-treated eyes do dry better than the aflibercept 8 mg, even though they were not directly compared. These network matching analyses are done in ophthalmology for glaucoma drugs. It’s also done across other areas of medicine so it really is a sound statistical method.

Now, of course, there are limitations because there may be some characteristics that were dissimilar between the 2 phase 3 trials that we are just not aware of, right? We can only correct for the characteristics that we know about. There may be some other factors that were not even accounted for based on that might be a difference and that would be a limitation.

The second limitation is that we use the aggregate group data in terms of the aflibercept 8 mg-treated eyes because we didn’t have the patient-level data to compare 1-on-1, and we use the patient-level data for the matching network meta-analysis. But again, this is a statistically sound method of doing this that’s used across multiple areas in the medical fields.

-Advertisement-
-Advertisement-
-Advertisement-
-Advertisement-
-Advertisement-
-Advertisement-