If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
C1q, the initiating molecule of the classical complement cascade, is involved in synapse elimination and neuronal loss in neurodegenerative diseases including glaucoma. Here we report an evaluation of the safety, tolerability, and ocular pharmacokinetics and pharmacodynamics of intravitreal (IVT) injections of ANX007, an anti-C1q monoclonal antibody fragment that blocks activation of the classical complement cascade.
Design
An open-label, single-dose-escalation phase 1a study followed by a double-masked, randomized, sham-controlled, repeat-injection phase 1b study.
Participants
A total of 26 patients with primary open-angle glaucoma.
Methods
Nine patients with primary open-angle glaucoma (mean Humphrey visual field deviation between –3 and –18 dB) were enrolled in phase 1a and received single doses of ANX007 (1.0 mg, n = 3; 2.5 mg, n = 3; or 5.0 mg, n = 3). Seventeen patients (mean Humphrey visual field deviation between –3 and –24 dB) were enrolled in phase 1b and randomized to 2 monthly doses of ANX007 (sham, n = 6; 2.5 mg ANX007, n = 6; or 5 mg ANX007, n = 5).
Main Outcome Measures
Safety and tolerability (including laboratory evaluation of urinalysis, complete blood count, and serum chemistries), ANX007 pharmacokinetics, target engagement, and immunogenicity.
Results
The mean age overall was 70 years in phase 1a and 68 years in phase 1b. In both studies, no serious adverse events were observed, no non-ocular treatment-emergent adverse events (TEAEs) attributable to study drug were reported, and ocular TEAEs were mild. IOP returned to normal levels for all patients within 45 minutes of IVT injection. No clinically significant deviations in laboratory results were observed. In the phase 1b study, C1q in the aqueous humor was reduced to undetectable levels in both the 2.5 mg and 5 mg cohorts 4 weeks after the first ANX007 dose.
Conclusions
In these studies, single and repeat IVT ANX007 injections were well tolerated and demonstrated full target engagement 4 weeks after dosing with both low and high doses, supporting monthly or less-frequent dosing. Further investigation in neurodegenerative ocular diseases is warranted.
The irreversible loss of vision associated with glaucoma results from the loss of synapses and neurons (retinal ganglion cells [RGCs]) within the retina and degeneration of the optic nerve.
Aging and increased intraocular pressure (IOP) are key risk factors for the development of glaucoma, and current medical management consists of IOP-lowering agents.
Patients experiencing progressive degeneration in glaucoma or who are not sufficiently managed with standard of care IOP-lowering medical therapy require laser or incisional surgery to lower IOP and preserve sight.
For these patients, for those who want to avoid the increased risk of morbidity associated with surgery and for those who progress despite surgery, alternative treatment modalities are needed to block neurodegeneration independently of IOP lowering.
C1q, the initiating molecule of the classical complement pathway, has emerged as a target for treatment of neurodegenerative disorders, including ocular diseases.
In normal embryonic development, C1q and the classical complement pathway drive synaptic pruning to shape neuronal circuitry. C1q recognizes and binds to ‘weaker’ synapses and activates the classical cascade, tagging them with complement components for phagocytosis by microglial cells, thereby preserving more-active synaptic pathways.
In neurodegenerative disorders, including ocular diseases, changes on the surface of synapses due to aging and other insults, appear to re-activate this embryonic pathway, leading to the elimination of functioning synapses and neuronal damage. For example, in glaucoma, there is increased deposition of C1q and the downstream complement activation product C3b within the cellular and synaptic layers of the retina.
In the spontaneous, chronic model of glaucoma in the DBA/2J mice, C1q deposition has been demonstrated in the inner plexiform, RGCs, and retinal nerve fiber layers prior to overt retinal pathology.
Further, in a photo-oxidative model of photoreceptor cell loss, a model of age-related macular degeneration, IVT administration of anti-C1q is protective against photoreceptor cell loss and preserves retinal function.
These observations support a role for C1q in the pathophysiology of neurodegenerative disorders of the retina and the idea of testing anti-C1q as a therapeutic approach for the treatment of such disorders.
ANX007 is an investigational, humanized antibody antigen-binding fragment (Fab) that recognizes C1q and blocks its initiation of the classical complement pathway (Figure 1).
By selectively targeting C1q, ANX007 inhibits the classical pathway while leaving the lectin and alternative complement pathways intact for normal immune function and tissue homeostasis. Here we report the results of phase 1 single- and repeat-dose studies evaluating the safety and tolerability of intravitreally administered ANX007, along with ocular pharmacokinetics (PK) and pharmacodynamics (PD) in patients with primary open-angle glaucoma.
Figure 1C1q drives activation of the classical complement cascade and becomes a driving force in neurodegenerative disease.
The phase 1a study was a first-in-human, open-label (no masking), dose-escalation safety study that evaluated 3 dose levels of ANX007 administered as single IVT injections (NCT03488550). The phase 1a study was conducted at a single site in the United States. The phase 1b study was a double-masked, randomized, sham-controlled trial evaluating 2 dose levels of ANX007 (2.5 mg and 5 mg) to determine the safety and tolerability (primary objectives) of 2 repeat monthly IVT injections (NCT04188015). The phase 1b study was conducted at 2 sites in the United States. Institutional Review Board (IRB)/Ethics Committee approval and informed consent were obtained from all patients prior to screening in accordance with the Declaration of Helsinki, ICH E6 Good Clinical Practice guidelines, and other applicable local regulations.
Participants
In the phase 1a study, eligible patients were aged 18 years or older, of either gender, with primary open-angle glaucoma on stable IOP-lowering medications, with a mean deviation of –3 to –18 dB on a reliable visual field test using the Humphrey field analyzer (HFA) in the study eye, and an IOP <21 mm Hg for at least 4 weeks prior to dosing, with no change to the IOP treatment regimen anticipated throughout study participation. In the phase 1b study, eligible patients were aged 18 years or older, of either gender, with primary open-angle glaucoma with a mean deviation of –3 to –24 dB on a reliable visual field test using the HFA in the study eye as determined at screening, and an IOP <21 mm Hg at screening and day 1, with no change to the IOP treatment regimen anticipated throughout study participation. For both studies, key exclusion criteria included intraocular inflammatory or infectious eye disease; uncomplicated cataract surgery less than 3 months prior to injection; trabeculectomy, iridotomy, vitrectomy, or other incisional ocular procedures in the study eye that could affect drug distribution and excretion; and history of autoimmune or neurologic disease. A detailed list of inclusion and exclusion criteria is available on ClinicalTrials.gov (NCT03488550 and NCT04188015).
Randomization and Treatments
In the dose-escalation phase 1a study, patients were enrolled into the lowest open cohort in an open-label manner and received a single dose of ANX007 at 1 of 3 dose levels (1 mg, 2.5 mg, or 5.0 mg), with cohort dose escalation in the absence of any dose-limiting toxicity. Dose-limiting toxicity criteria were defined as any of the following adverse events (AEs) occurring within 2 weeks after an IVT injection and considered to be related to ANX007: intraocular inflammation of at least grade 2, using the Standardization of Uveitis Nomenclature criteria, that was unresponsive to topical corticosteroid therapy; increase in IOP >30 mm Hg that was unresponsive to topical or oral IOP-lowering medications; or loss of best corrected visual acuity of >3 Snellen lines from baseline, excluding a transient loss immediately after IVT injection. No dose-limiting toxicities were observed.
In the repeat-dose phase 1b study, patients were randomized to receive 2 injections 4 weeks apart (day 1 and day 29) of sham (no ocular penetration) or 2.5 mg ANX007 or 5.0 mg ANX007. Doses were administered at volumes of 20 μL (1 mg dose), 50 μL (2.5 mg dose), and 100 μL (5 mg dose), respectively.
Intravitreal Injections
Intravitreal injection was performed under sterile conditions. Topical proparacaine was used for anesthesia. Lid margins were cleaned with Povidone-iodine 10% and povidone-iodine 5% was applied to the conjunctival surface, including the injection site. For ANX007 injection, the needle was inserted perpendicular to the sclera, 3.5 to 4 mm posterior to the limbus, between the vertical and horizontal rectus muscles. A sterile cotton-tip applicator was applied over the injection site immediately following removal of the needle to reduce vitreous reflux. For the sham injection, the preparation was identical to injection with ANX007. The sham injection was performed by applying pressure to the eye at the location of a typical intravitreal injection using the blunt end of an empty syringe without a needle. Intraocular pressure was checked post-injection. Topical antibiotics were not used post-injection.
Outcomes
In the dose-escalation phase 1a study, patients were followed for a total of 8 weeks for safety, tolerability, PK (unbound ANX007 serum concentration), PD (concentration of unbound C1q in serum, as a measure of “target engagement”), and immunogenicity (emergence of serum antidrug antibodies [ADAs]).
In the repeat-dose phase 1b study, patients were followed for a total of 12 weeks for safety, tolerability, PK (unbound ANX007 concentration in serum and aqueous humor), PD (concentrations of unbound C1q in serum and aqueous humor, as a measure of “target engagement”), and immunogenicity (emergence of serum ADAs).
Safety Assessments
The safety assessments included adverse events, vital signs, physical examination, urinalysis, complete blood count, serum chemistries, best corrected visual acuity assessment, slit-lamp examination, IOP measurement, and dilated fundoscopy.
Pharmacokinetics and Immunogenicity Analysis
Blood samples were collected throughout the study for determination of unbound ANX007 concentrations, unbound C1q concentrations, and anti-ANX007 antibodies in serum. In the single-dose study, samples for PK and PD analyses were obtained pre-injection and 3 hours post-injection on day 1, and on days 2, 4, 8, and 15. Samples for the evaluation of serum ADAs were obtained pre-injection, on study day 29, and at the follow-up visit on day 56. In the repeat-dose study, samples for PK and PD analyses (serum and aqueous fluid) were obtained pre-injection on study days 1 and 29, and serum samples were also collected 3 hours post-dose on each dosing day. Samples for the evaluation of serum ADAs were obtained pre-injection on study days 1 and 29 and at follow-up visits on days 56 and 84.
All serum PK, C1q, and ADA samples were assayed by Charles River Laboratories (Skokie, IL, USA) using GLP validated methods. The method for the quantitative determination of unbound ANX007 in human serum used an electrochemiluminescence immunoassay with a quantification range of 1.25 ng/mL to 250.0 ng/mL, run at a minimum required dilution of 40-fold. Unbound C1q in human serum used an enzyme-linked immunosorbent assay (ELISA) with a quantification range of 2.5 ng/mL to 80.0 ng/mL, run at a dilution of 2500-fold. A bridging electrochemiluminescence assay was used for quantitative determination of anti-ANX007 antibodies in human serum.
Aqueous humor was collected only in the phase 1b study per investigator standard of care into a tube provided for the study. Aqueous humor levels of unbound ANX007 and unbound C1q were measured by Annexon Biosciences using research-based assays. Unbound ANX007 in aqueous humor (diluted 50 to 250-fold) was measured by capture onto human C1q-coated ELISA plates, with a quantification range of 0.02 to 50 ng/mL. Plate-captured ANX007 was detected using an anti-Fab alkaline phosphatase (AP) conjugate antibody. Unbound C1q was measured in aqueous humor (diluted 100-fold) by capturing with monoclonal mouse anti-C1q (abcam71940)–coated ELISA plates (antibody specific to collagen tail of C1q), with a quantification range of 0.01 to 30 ng/mL. Plate-captured C1q was detected using a monoclonal mouse anti-C1q antibody-AP conjugate that binds to the same site as ANX007, thus enabling measurement of free C1q levels. Plates were washed at the end of each of the steps prior to the next step, including antibody capture, standard or sample addition, and detection antibody. Plates were finally developed using an AP substrate and signal was measured using a luminometer. Standard curves were fit using a 4PL logistic fit and concentration of unknowns determined and corrected for dilution.
Statistical Analyses
No formal comparisons or hypothesis testing were planned for comparisons between dosing levels evaluated in either study, and data are presented descriptively throughout. Descriptive statistics for categorical variables included counts and percentages. Descriptive statistics for continuous variables included means, medians, standard deviation (SD), and minimum and maximum values. Sample sizes were determined by clinical and practical, rather than statistical, considerations. Pooled data are reported by dose level of ANX007 by study and non-study eye, where applicable. Patients who received at least 1 injection of ANX007 were assessed for safety and immunogenicity endpoints. The incidences of treatment-emergent AEs were tabulated by system organ class and preferred term, as well as by ocular or systemic intensity, or Common Terminology Criteria for Adverse Events grade, and by relationship to ANX007.
RESULTS
Demographics and Baseline Characteristics
In the phase 1a study, 9 patients were enrolled and received a single IVT dose of ANX007 of 1.0 mg (n = 3), 2.5 mg (n = 3), or 5.0 mg (n = 3). The majority of patients were male (n = 6, 66.7%) and White (n = 6, 66.7%), with a mean age (SD) of 69.9 (7.4) years (Table 1). All 9 patients completed the study; there were no early terminations.
Table 1Phase 1a: Demographics and Baseline Characteristics.
In the phase 1b study, 18 patients were included in the study and 17 received study treatment. One patient was randomized to receive ANX007 2.5 mg but terminated participation in the study and did not receive any study treatment due to elevated IOP on day 1. Six patients received sham IVT injection, 6 patients received ANX007 2.5 mg, and 5 patients received ANX007 5.0 mg. The majority of patients were female (n = 9, 52.9%) and white (n = 13, 76.5%), with a mean age (SD) of 67.8 (17.2) years (Table 1). One patient with pseudo-exfoliative glaucoma in the sham group terminated the study early after receiving the day 1 injection due to protocol violation (pseudo-exfoliation glaucoma was not recognized as included in primary open angle glaucoma inclusion criteria).
Single Ascending Doses of ANX007 (Phase 1a)
Safety and tolerability. There were no serious adverse events (SAEs), treatment-emergent adverse events (TEAEs), or fatal AEs reported following single IVT injections of ANX007 at any of the dose levels evaluated. Eight ocular TEAEs were reported in the study eye, including ocular irritation, conjunctival hemorrhage, and conjunctival hyperemia, and 1 systemic TEAE was reported (sinusitis) which was not assessed to be related to ANX007 (Table 2). No ocular TEAEs were reported in the non-study eye for any patient. All AEs were mild in severity and resolved by the end of the study and there were no clinically significant laboratory results at any time during the on-therapy period.
Table 2Phase 1a Summary of Adverse Events for Each Treatment Group Dose Level.
IOP returned to normal levels (defined as within 5 mm Hg of baseline or ≤21 mm Hg) at 30 minutes for all patients. Mean IOP measurements were within normal limits (21 mm Hg or less) at all time points (Table 3).
Table 3Phase 1a: Mean and Median Intraocular Pressure Over Time for Each Treatment Group.
Mean IOP (SD) Median IOP (min, max)
Phase 1a Single Ascending Dose Study
ANX007 1 mg
ANX007 2.5 mg
ANX007 5 mg
Baselinea
17.7 (1.15) 17 (17, 19)
16.3 (3.06) 17 (13,19)
15.0 (2.65) 14 (13, 18)
30 minutes post-injection
19.3 (2.89) 21 (16, 21)
17.3 (4.16) 16 (14 ,22)
18.0 (3.61) 19 (14, 21)
Day 2
16.3 (3.79) 18 (12, 19)
10.7 (3.06) 10 (8, 14)
13.3 (2.52) 13 (11, 16)
Day 4
16.0 (2.65) 17 (13, 18)
13.0 (5.20) 10 (10, 19)
11.7 (2.31) 13 (9, 13)
Day 8
17.0 (3.61) 16 (14, 21)
13.0 (1.00) 13 (12, 14)
13.0 (2.00) 13 (11, 15)
Day 15
18.3 (2.89) 20 (15, 20)
13.3 (2.52) 13 (11, 16)
14.0 (1.73) 13 (13, 16)
Day 29
18.3 (1.53) 18 (17, 20)
17.0 (1.00) 17 (16, 18)
15.0 (5.29) 13 (11, 21)
Day 56
15.3 (1.15) 16 (14, 16)
14.0 (1.00) 14 (13, 15)
14.3 (3.21) 13 (12, 18)
IOP, intraocular pressure; SD, standard deviation
aBaseline IOP was defined as the last non-missing value collected prior to the first dose of study drug on day 1.
Serum pharmacokinetics and pharmacodynamics. Serum unbound ANX007 was not detected, and serum unbound C1q levels were unchanged over the period of the study.
Immunogenicity. No ADAs positivity was detected during the study at any ANX007 dose level at any of the timepoints assessed.
Randomized Multiple Dosing of ANX007 (Phase 1b)
Safety and tolerability. No SAEs, severe TEAEs, or fatal AEs were reported in any group at any time during the study. A total of 6 TEAEs were reported, 2 events in each dose level. TEAEs included conjunctival hemorrhage, conjunctival hyperemia, eye irritation/pain, foreign body sensation in the eyes, ocular hyperemia, blurry vision, and 1 systemic TEAE of bronchitis, which was not assessed to be related to ANX007 (Table 2). All TEAEs were mild in severity and none resulted in study discontinuation. All resolved by study closure. No drug-specific safety signals were observed, and there were no clinically significant laboratory results during study.
IOP returned to normal levels (≤21 mm Hg) within 45 minutes for all patients and within 30 minutes for 16/17 (94.1%) patients. Mean IOP measurements were within normal limits (21 mm Hg or less) at all time points (Table 4).
Baseline BCVA LogMAR transformed results were available for all participants for both the study and non-study- eyes. Between Baseline and Day 84, the mean (SD) LogMAR results for the study eye decreased (representing improved visual acuity) in the Sham (-0.073±0.0485) and ANX007 2.5 mg (-0.032±0.1001) groups but increased (representing worsened visual acuity) in the ANX007 5.0 mg group (0.024±0.0485). In the non-study eye, there were decreases in mean (SD) LogMAR between baseline and Day 84 for the Sham group (-0.20±0.0885) and ANX007 2.5 mg group (-0.21±0.0479), and no change in the ANX007 5.0 mg group. Details are provided in Appendix C
Pharmacokinetics and pharmacodynamics. Mean (SD) and median (min, max) ANX007 levels in aqueous humor on day 29 were 911 ng/mL (846) and 631 (225, 2539) ng/mL in the ANX007 2.5 mg dose group and 5345 ng/mL (6351) and 3111 (413, 12512) ng/mL in the ANX007 5.0 mg dose group (Fig 2). Correspondingly, mean (SD) and median (min, max) unbound C1q at day 1 (pre-injection) was 100.1 (99.9) and 51.6 (16, 254) ng/mL in sham, 110.0 (109.5) and 65.7 (23, 293) in the ANX007 2.5 mg dose group, and 110.0 (83.1) and 86.8 (14, 222) in the ANX007 5.0 mg dose groups. Unbound C1q was reduced to undetectable levels at day 29 (<0.6 ng/mL) in aqueous humor at both dose levels in all subjects (Fig 3).
Figure 2Phase 1b aqueous humor ANX007 levels following a single intravitreal (IVT) dose of ANX007 (data are mean +/- SEM and n=3-6). Dose-dependent increase in aqueous unbound ANX007 levels 29 days after a single IVT dose.
Figure 3Phase 1b aqueous humor C1q levels following a single IVT dose of ANX007 (data are mean +/- SEM and n=3-6). Significant reduction in aqueous unbound C1q levels 29 days after a single IVT dose at both dose levels
Immunogenicity. Low levels of ADA positivity were detected in 7 patients (1 patient in the sham group, 3 patients in the ANX007 2.5 mg group, and 3 patients in the ANX007 5 mg group). The signal in the ADA assay was marginally above the cut points. Of these, 5 patients (1 patient in the sham group, 1 patient in the ANX007 2.5 mg group, and 3 patients in the ANX007 5 mg group) were ADA-positive at all time points, including the pre-dosing period, without any dose-related changes. These patients likely represent false positives in the ADA assay, as there was no change in titer over time and ADA positivity was detected prior to dosing in the ANX007 groups. These data suggest an unclear relationship between ADA measurement and ANX007 dosing.
DISCUSSION
Overall, the results of the phase 1a and phase 1b evaluation of single and repeated IVT dosing with ANX007 demonstrated the safety and tolerability of the dose levels tested. All AEs were mild in intensity and were resolved by the end of the respective studies, and no SAEs, no severe TEAEs and no clinically significant laboratory results were reported. The results observed in the treated eye were similar to those seen in the sham. These data suggest that C1q does not play a critical role in normal vision, since short-term inhibition achieved in this study did not lead to any visual compromise.
Animal models have demonstrated an intimate role of C1q in the pathologic processes that drive neurodegenerative diseases, including ocular diseases such as glaucoma (Fig 1).
IVT administration of an antibody against C1q (a mouse precursor antibody to ANX007) protected RGCs following acute induction of increased IOP (Annexon, unpublished data). Complement pathway activation may lead to RGC degeneration either through increased neuroinflammation and tissue damage, or through targeted elimination of synapses and subsequent neurodegeneration, or a combination of these mechanisms. Consistent with a role for complement in glaucoma, a recent analysis of proteomic studies of the aqueous humor revealed a particular enrichment for complement protein changes in glaucoma.
The PK and PD analyses conducted in the phase 1b study demonstrated detectable levels of ANX007 and corresponding complete reduction of free C1q in the aqueous humor following IVT injection with ANX007 at both the 2.5 and 5.0 mg dose levels through 29 days after the first injection. Importantly, no PK or PD changes were observed in the systemic circulation. Full engagement of C1q in the aqueous humor at day 29 in all treated subjects supports monthly or less-frequent dosing of ANX007 in future clinical studies.
Inhibition of C1q in the aqueous humor is anticipated to represent inhibition of C1q in relevant sites of neurodegenerative ophthalmic disease within the retina. While ocular PK/PD sampling is generally limited to the aqueous humor in clinical trial subjects, additional ocular fluid and tissue sampling has been conducted in nonhuman primates administered ANX007 IVT. In these animal studies, complete C1q target engagement in the aqueous humor reflected complete target engagement in the vitreous and retinal tissue, with measurable unbound drug at these sites in the back of the eye (Annexon, data on file).
ANX007 exhibits binding affinity for C1q (KD= 10 pM) and is a potent inhibitor of the classical complement cascade.
Therefore, IVT delivery of ANX007 for the treatment of ocular neurodegenerative diseases may provide superior retinal tissue penetration and C1q inhibition compared to that achieved through systemic inhibitor administration, since the retina is behind the blood-brain barrier. The roles that C1q and the classical cascade play in neurodegenerative disease suggests that C1q inhibition may have clinical relevance for a range of ocular disorders, including glaucoma and geographic atrophy.
When considering the results of these data in a broader context, it should be noted that the relatively small sample size and descriptive nature of the analyses preclude formal clinical endpoint comparisons between the sham and treatment groups in the phase 1b study and also between the dose levels evaluated. Additional data from larger cohorts will add to our understanding of safety and efficacy at different dosing levels.
The results of these 2 studies provide safety and target engagement data and support the further development of ANX007 for patients with ophthalmic neurodegenerative disorders such as glaucoma and geographic atrophy.
Acknowledgements
The authors would like to thank Anita Grover, PhD, and Sanjay Keswani, MBBS, FRCP for their contributions. Y.S. and J.L.G. gratefully acknowledge departmental support from Research to Prevent Blindness, Inc.
Meeting presentation: These data were presented at the American Academy of Ophthalmology Annual Meeting 2020.
Financial support: These studies and medical writing support were funded by Annexon Biosciences
Conflict of interest disclosures:
Yang Sun: Received grant support from Annexon Biosciences.
David Wirta: Received grant support from Annexon Biosciences.
Vidhu Mathur, Sethu Sankaranarayanan, Lori K. Taylor, Ted Yednock, Donald S. Fong: Full-time employees of Annexon and may hold Annexon stock and/or stock options.
Wendy Murahashi: Has provided consulting services and holds Annexon stock options.
Jeffrey L. Goldberg: Consultant for Annexon Biosciences.
Precis
Single and repeat intravitreal injections of the anti-C1q antibody fragment ANX007 were well tolerated and demonstrated full target engagement 4 weeks after dosing in patients with primary open-angle glaucoma.
00022-2&id=gr1.jpg){kind=link}
00022-2&id=gr2.jpg){kind=link}
00022-2&id=gr3.jpg){kind=link}
00022-2&id=gr4.jpg){kind=link}