Failed lampalizumab studies offer numerous lessons about GA
Although the phase III Chroma and Spectri studies for lampalizumab (Genentech), an investigational compound for the treatment of geographic atrophy secondary to age-related macular degeneration (AMD), failed to meet their primary endpoints, numerous lessons can be learned from the study data.
By Michelle Dalton, ELS
Although the phase III Chroma and Spectri studies for lampalizumab (Genentech), an investigational compound for the treatment of geographic atrophy secondary to age-related macular degeneration (AMD), failed to meet their primary endpoints, numerous lessons can be learned from the study data, according to two experts.
Jeffrey S. Heier, MD, and Frank G. Holz, MD, presented comprehensive data on lampalizumab, including genetics and long-term visual function outcomes that will help advance the understanding of geographic atrophy. Both presented their data at the 41st Annual Macula Society Meeting, held in February 2018,
Dr. Heier is in private practice at Ophthalmic Consultants of Boston. Dr. Holz is chairman and professor of ophthalmology, University of Bonn, Germany.
Lampalizumab is a selective inhibitor of complement factor D. There has been genetic evidence the implicates the complement cascade dysregulation in geographic atrophy/AMD, Dr. Heier said, noting complement factor D is a rate-limiting enzyme of the alternate complement pathway with the lowest plasma concentrations of complement proteins. In a phase II study, geographic atrophy area progression was reduced by 20% from baseline to 18 months in the lampalizumab group compared to the sham group.
Warrant move to phase III
That was a substantial enough difference to warrant moving lampalizumab into phase III studies (Spectri and Chroma, which enrolled 975 and 906 patients, respectively). These randomized, double-masked, sham-controlled trials enrolled patients “with well-demarcated geographic atrophy secondary to AMD with no proliferative choroidal neovascularization in both eyes,” Dr. Heier said.
Best-corrected visual acuity (BCVA) had to be Snellen equivalent of 20/100 or better, and if vision was 20/25 or better, the geographic atrophy lesion had to be within 250 μm of the foveal center. Total lesion size had to be between 1 to 7 disk areas.
Patients were randomized in a 2:1 ratio to either 10 mg lampalizumab or sham injection every 4 weeks in Spectri (lampalizumab q4), and to either 10 mg lampalizumab or sham injections every 6 weeks in Chroma (lampalizumab q6).
For ease of data analysis, both sham arms were pooled. The primary outcome for both arms was the change in geographic atrophy area from baseline at week 48, as assessed by fundus autofluorescence.
Baseline characteristics “were well balanced across study arms and in alignment with AMD clinical trials,” Dr. Heier said.
Results from Spectri, Chroma
In the 2 studies, geographic atrophy area growth from baseline to week 48 was 2.09 mm2 in Spectri, 2.02mm2 in Chroma, and 1.93mm2 in the Spectri sham group and 2.04 mm2 in the Chroma sham group. There was no statistical difference between the arms.
Both Spectri and Chroma were designed to test whether or not complement factor I (CFI) biomarker profile status would affect the response rate to lampalizumab.
Previous studies had shown “genetic variants of CFI have been linked to an increased risk of advanced AMD development,” Dr. Heier said. “The phase II Mahalo study suggested a differential treatment response based on CFI-profile biomarker status.”
No new safety signals were observed in the phase III studies compared with Mahalo. There were 5 cases of endophthalmitis (out of 12,447 injections, or 0.4 events/1,000 injections), and a total of 5 events in 1,252 treated patients (0.4%).
In 2.9% of the lampalizumab-treated patients, there was an increased intraocular pressure. Nonocular adverse events were similar across all treatment arms.
The results of Spectri and Chroma “do not support the CFI-profile genetic marker as a predictor of geographic atrophy lesion growth,” Dr. Heier said.
Secondary outcomes
Dr. Holz discussed the secondary outcome results, including BCVA, low luminence visual acuity (LLVA), number of scotomatous points, binocular reading speed, National Eye Institute Visual Field Questionnaire-25 (NEI VFQ-25) composite scores, and functional reading independence (FRI) Index score. The correlation of geographic atrophy area with these outcome measures were analyzed at baseline and calculated at week 48 (again, sham patient results were pooled).
All groups lost about 1 line, with the lampalizumab q4 losing 4.1 ETDRS letters, compared to lampalizumab q6 and the pooled sham arms each losing 4.9 ETDRS letters. The lampalizumab q4 and sham groups lost about the same number of LLVA letters (-2.5 and -2.7, respectively), while the lampalizumab q6 group lost 3.2 letters.
Scotomatous points also increased in all 3 groups in whom the outcome was measured.
Fluent reading speed is typically 80 words/min (wpm) or better. Those patients in the sham groups had a decrease of 17.09 wpm, compared to -16.91 wpm in the lampalizumab q6 group and -13.55 wpm in the lampalizumab q4 group.
NEI VFQ-25 scores decreased in all groups: -1.72 in the sham groups, -1.94 in the lampalizumab q6 group, and -1.22 in the lampalizumab q4 group. Similar declines were noted in the FRI index: -0.11 in the sham groups, -0.14 in the lampalizumab q6 group, and -0.12 in the lampalizumab q4 group.
In all variables, there was a negative correlation between geographic atrophy area and functional outcomes, with the exception of microperimetry outcomes.
The “rich dataset, including genetics and long-term visual function outcomes,” will help advance the understanding of geographic atrophy pathophysiology and natural history, Dr. Heier concluded.
Jeffry S. Heier, MD
p. 617-314-2694
Frank G. Holz, MD
This article was adapted from presentations that Drs. Heier and Holz presented at the 2018 Macular Society meeting. Dr. Heier is a consultant and receives grant and research fees from Genentech/Roche. Dr. Holz is a consultant for Genentech/Roche.
