ARVO 2022 Day 3 – Topline Data from DAZZLE (KSI-301 for Wet AMD) and Interview with Carl Regillo

Retina Roundup DAZZLE KSI-301 Phase 3 Trial Summary
Blake Fortes, MD
PGY-3, Mayo Clinic

Dr. Carl Regillo from Wills Eye Hospital presented the one-year results of the DAZZLE trial, a prospective, randomized, multicenter, double-masked phase 2b/3 clinical trial investigating the use of KSI-301 for the treatment of neovascular age-related macular degeneration (nAMD). KSI-301 is a novel antibody biopolymer conjugate designed to provide long-lasting vascular endothelial growth factor (VEGF) suppression to reduce the treatment burden of patients undergoing treatment for nAMD. This study builds upon early clinical studies demonstrating efficacy and durability of up to 6 months for patients with nAMD.

This study recruited treatment-naïve patients with nAMD who were randomized in a 1:1 fashion into one of two treatment arms: KSI-301 (5 mg) or aflibercept (2 mg). Each patient received three monthly loading doses of their assigned treatment. Thereafter, subjects in the KSI-301 arm were treated every 3-5 months based on predefined disease activity assessments. After the first 3 months, patients in the aflibercept arm received every-other-month dosing.

Overall, 277 patients were randomized to KSI-301 treatment vs. 280 patients in the aflibercept group. Baseline characteristics were similar between both groups. However, there were a greater number of discontinuations in the KSI-301 group mainly driven by nAMD disease progression events possibly associated with relative undertreatment.

In terms of efficacy data, the study did not meet its primary endpoint of non-inferiority to aflibercept in best corrected visual acuity (BCVA), which was likely related to undertreatment of a minority of patients in the Q12 week KSI-301 treatment group. However, a majority of KSI-301 patients achieved a 20-week interval at year 1 with visual acuity gains comparable to the overall aflibercept group. Overall, 60% of patients treated with KSI-301 were treated with a 20-week injection interval, 10% were treated with a 16-week injection interval, and 30% were treated with a 12-week injection interval. A 12-week injection interval turned out to be insufficient for some patients and injection intervals less than 12 weeks were not allowed in this trial, which led to relative undertreatment in that group of patients.

Regarding safety, KSI-301 was well-tolerated. Common ocular adverse events were generally comparable to aflibercept with a few differences related primarily to undertreatment and disease progression. 3.2% of patients in the KSI-301 group experienced intraocular inflammation vs. 0% in the aflibercept arm (although prior phase 3 studies with aflibercept have demonstrated an IOI range of 1-2%). None of the intraocular inflammation events were serious and no patient discontinued the study due to the IOI event.

Five large-scale phase 3 studies are underway to compare KSI-301 vs. aflibercept for the treatment of nAMD (DAYLIGHT [NCT04964089]), diabetic macular edema (GLEAM [NCT04611152] and GLIMMER [NCT04603937]), non-proliferative diabetic retinopathy (GLOW [NCT05066230] and retinal vein occlusions (BEACON [NCT04592419]). These ongoing studies have a lower risk of undertreatment based on their designs allowing for more frequent KSI administration after the loading phase, while still providing an opportunity to show meaningful improvements in treatment durability.

Below please find further insight from Dr. Regillo from a post-presentation interview.

Interview with Dr. Carl Regillo 5/3/22

Question 1: Were there any factors that were found to be predictive of patients performing well on the Q20 week interval vs. those who performed poorly on the Q12 week interval?

Dr. Carl Regillo: That’s an analysis that will be conducted. This presentation that was conducted at ARVO was the first to the podium for this pivotal clinical trial and it consisted mainly of topline visual, anatomic and safety data. As I indicated in my conclusion slide of the talk, looking at factors that are predictive of a good, durable response is important and is planned. We really need to identify clinical features at baseline that help the clinician know where the drug works best such as with the 60% of patients in the trial who had good visual and anatomic results with KSI treatment successfully extended out to 20 weeks. Just as important, we also need to know in whom and why it didn’t perform as expected in terms on disease control and maintenance of visual gains such as in the 30% of patients who had vision loss over the year getting treatments at 12 week intervals. We may or may not be able to fully explain the marked differences in response. I’ve run such analyses before such as with the phase 3b HARBOR study which tested two doses and two treatment regimens of ranibizumab. Unfortunately, there weren’t a lot of patient characteristics at baseline that predicted dosing frequency in the PRN arms of that study. That being said, the case could be different with KSI in the DAZZLE study, and we hope to have the results of such an analysis in the near future.

Question 2: How did prior experience with postmarketing reports of brolucizumab causing intraocular inflammation (IOI) and retinal vasculitis impact the design of this study?

Dr. Carl Regillo: It didn’t necessarily impact the design, although for both this study and as you saw for the faricimab wet AMD trial results, we scrutinize all of the adverse events and characterize individual inflammatory events. On the podium, it is now routine and expected to show each inflammatory category (eg. anterior vs. posterior), the severity, and we report on any direct retinal or retinal vascular involvement, including vasculitis and any related vascular occlusions. In Dazzle, there was an imbalance in IOI between the two arms with a higher IOI rate with KSI-301 at 3.2% versus aflibercept at 0%. Here, the imbalance seemed even greater because aflibercept performed better in terms of safety than any other prior trial in wet AMD which had IOI rates in the 1-2% range. Furthermore, the good news for KSI-301 is that none of the IOI events were serious, and none had any associated direct retinal or retinal vascular effects.

Question 3: What role will KSI-301 have in your AMD armamentarium especially considering some of the other novel treatments, including faricimab and the port delivery system with ranibizumab?

Dr. Carl Regillo: KSI-301 didn’t meet its primary, noninferiority endpoint in DAZZLE and so an FDA approval for wet AMD is not possible with this study alone. Therefore, it will not be available to use in practice for the time being. However, how it performs in the other wet AMD trial called the DAYLIGHT study, where the drug is being dosed every 4 weeks, will give us important information about how it stacks up as an anti-VEGF agent to aflibercept and how it may be utilized in practice should FDA approval be possible based on meeting the primary endpoint of that study. Furthermore, it could be approved and potentially utilized in practice for other retinal conditions such as DME, RVO, and even DR without DME. It’s being tested for all those different indications at the phase 3 level at this time. Results of these other trials should be available over the coming year or so.