Abstract
p75 neurotrophin receptor (p75NTR) signaling pathways substantially overlap with degenerative networks active in Alzheimer disease (AD). Modulation of p75NTR with the first-in-class small molecule LM11A-31 mitigates amyloid-induced and pathological tau-induced synaptic loss in preclinical models. Here we conducted a 26-week randomized, placebo-controlled, double-blinded phase 2a safety and exploratory endpoint trial of LM11A-31 in 242 participants with mild to moderate AD with three arms: placebo, 200 mg LM11A-31 and 400 mg LM11A-31, administered twice daily by oral capsules. This trial met its primary endpoint of safety and tolerability. Within the prespecified secondary and exploratory outcome domains (structural magnetic resonance imaging, fluorodeoxyglucose positron-emission tomography and cerebrospinal fluid biomarkers), significant drug–placebo differences were found, consistent with the hypothesis that LM11A-31 slows progression of pathophysiological features of AD; no significant effect of active treatment was observed on cognitive tests. Together, these results suggest that targeting p75NTR with LM11A-31 warrants further investigation in larger-scale clinical trials of longer duration. EU Clinical Trials registration: 2015-005263-16; ClinicalTrials.gov registration: NCT03069014.
Main
Late-onset Alzheimer disease (AD) is the leading cause of dementia1,2. AD is a complex and heterogeneous disease in which multiple mechanisms become dysregulated to promote synaptic failure, degeneration and loss3,4. Two important approaches for disease-modifying AD therapies involve targeting the accumulation of pathological forms of amyloid-β (Aβ) or tau5,6,7,8. A limitation of these strategies is that they each target a narrow set of AD-related pathophysiological processes. An alternative pharmacological strategy is to target ‘deep biology’, that is, receptors and/or signaling networks that control manifold fundamental cellular pathways and may, therefore, be able to normalize multiple pathological processes underlying AD, particularly those relevant to synaptic resilience and degeneration9,10,11.
Over the past two decades, multiple lines of evidence have converged on the p75 neurotrophin receptor (p75NTR) as a promising deep biology target for modifying neuronal dysfunction and degeneration in AD. p75NTR is a member of the tumor necrosis factor family12. Although p75NTR has traditionally been known as a ‘death receptor’, more recent work has demonstrated that it can determine synaptic and cellular fate13. p75NTR is a coreceptor for sortilin and SorCS2. In its nonliganded state or when binding to proneurotrophin ligands, such as pro-nerve growth factor (pro-NGF) or pro-brain-derived neurotrophic factor (pro-BDNF), p75NTR promotes degenerative signaling that causes destabilization of dendritic spines, degeneration of synapses and neuronal death14,15,16,17. However, p75NTR can also bind mature forms of neurotrophins (such as NGF and BDNF) and can act as a tropomyosin receptor kinase (Trk) co-receptor, thereby promoting cell survival and synaptic plasticity through multiple pathways18,19,20. Thus, p75NTR acts as a potent and fundamental molecular signal switch for neuronal survival and synaptic integrity.
p75NTR regulates a broad intracellular signaling network that has considerable overlap with degenerative signaling networks active in AD, particularly those relevant to synaptic function and resilience18,21,22,23. Consistent with this overlap, p75NTR-mutant mice demonstrate resilience against Aβ-related neuronal degeneration17,24,25,26,27,28. In humans, polymorphisms in the genes encoding proneurotrophins and p75NTR coreceptors, including sortilin and SorCS2, are associated with altered AD risk29,30,31,32. Studies of patients with AD and tauopathy reported increased levels of p75NTR in brain tissue and elevated levels of pro-NGF in brain extracts and cerebrospinal fluid (CSF)33,34,35,36. In the adult human brain, the highest expression of p75NTR is observed in cell types that are among the earliest affected in AD, including cholinergic neurons of the basal forebrain and their cholinoreceptive target populations in the entorhinal cortex and hippocampus13,37,38. p75NTR is also expressed by cortical, hippocampal pyramidal and locus coeruleus neurons, with locus coeruleus neurons constituting another population involved in the earliest stages of AD pathology39. Within non-neuronal populations, p75NTR expression is upregulated in microglia and astrocytes in pathological settings, including AD and tauopathies15.
Taken together, these lines of evidence have motivated preclinical work examining the therapeutic potential for small-molecule modulation of p75NTR to downregulate its degenerative signaling40. One such candidate, LM11A-31, is a small molecule based on the structural, physical and chemical features of β-hairpin loop 1 of NGF, a domain of NGF that mediates interaction with p75NTR (ref. 41). LM11A-31 functions as a p75NTR modulator to downregulate its degenerative signaling and as an antagonist to pro-NGF-induced degeneration42,43. It was found to readily cross the blood–brain barrier following oral administration and was nontoxic in preclinical studies44,45. In AD and tauopathy mouse models, oral administration of LM11A-31 reduced excess activation of enzymes contributing to tau post-translational modifications, accumulation of multiple forms of pathological tau species and tau seeding activity, reduced elevations in multiple microglia and astrocyte markers, and decreased the loss of dendritic spines and synapses while improving performance on hippocampal-dependent memory tasks45,46. In β-amyloid precursor protein (APP)-transgenic mice, administration of LM11A-31 had no detectable effect on Aβ plaques or brain tissue-derived soluble Aβ levels47. These findings, along with in vitro studies demonstrating that LM11A-31 inhibits neurite and synaptic degeneration induced by oligomeric Aβ43, suggest that modulation of p75NTR confers resilience to Aβ.
Despite its fundamental functional role in neural and developmental cell biology, the therapeutic potential for targeted modulation of p75NTR in humans has not been tested. In this study, we report the application of a p75NTR-based therapy in a human disease setting through a 26-week randomized, double-blind, parallel-group phase 2a safety and exploratory efficacy trial of LM11A-31 in participants with mild to moderate AD dementia. On the basis of studies in preclinical AD-related mouse models and two prior safety and pharmacokinetic studies in healthy human participants (designated as phase 1 and 1b trials), we hypothesized that modulation of p75NTR using LM11A-31 in persons with AD would be well tolerated and would slow AD progression, as measured by biomarkers of synaptic function, degeneration and glial activation (CSF biomarkers, structural magnetic resonance imaging (sMRI) and [18F]-fluorodeoxyglucose positron-emission tomography ([18F]-FDG PET)). Consistent with phase 2a strategies in AD trials48, cognitive measures were included as secondary or exploratory outcomes for assessment of safety and nominal directionality; the study was not of sufficient duration or power to reliably assess effects on potentially slowing the loss of cognitive function.
Results
Participant disposition
A total of 316 participants were screened for inclusion; 242 were enrolled in the trial (safety population) and 241 were successfully randomized and accounted for in the intention-to-treat (ITT) population. The first participant was randomized in May 2017 and the last participant completed treatment in June 2020. Data lock was executed in November 2020. Of these individuals, 221 completed the study as outlined in the protocol and 211 completed the study at the 26-week visit (Fig. 1). Analyses of primary, secondary, prespecified exploratory and post hoc exploratory outcomes were based on the ITT dataset.
Baseline characteristics of the trial cohort are outlined in Table 1. All trial participants had a biologically confirmed AD diagnosis (CSF Aβ42 < 550 ng l−1 or ratio of Aβ42 to Aβ40 < 0.89). Participants in the twice-daily placebo, 200 mg LM11A-31 and 400 mg LM11A-31 groups did not differ with respect to any key subject variables such as age, sex, race, screening Mini Mental Status Exam (MMSE) score, screening CSF Aβ42 or use of acetylcholinesterase (AChE) inhibitors (AChEIs) (P > 0.1 for each; Table 1). There was a slightly higher proportion of carriers of pathogenic apolipoprotein E4 (APOE4) alleles in the 400-mg group, although differences between groups did not reach statistical significance (P = 0.09).
Primary outcome
This study reports the effects of the novel strategy of selectively targeting p75NTR in a human population with disease. Moreover, LM11A-31 constitutes a first-in-class therapeutic agent for p75NTR. As such, evaluation of safety and tolerability was of key importance. The study met its primary prespecified endpoint of demonstrating the safety and tolerability of LM11A-31.
In order, the most frequently observed adverse events (AEs) were nasopharyngitis, diarrhea, headache and eosinophilia (Table 2). In most cases, AEs were transient. Nasopharyngitis (17 participants) and diarrhea (13 participants) were significantly more commonly reported in the 400 mg LM11A-31 group compared to placebo (odds ratio (OR) with 95% confidence interval (CI): nasopharyngitis, 5.41 (1.15 to 25.52); diarrhea, 12.22 (1.54 to 97.00); P < 0.05 for each). Of these participants, two withdrew due to diarrhea and none withdrew due to nasopharyngitis. Headache was experienced by a total of 12 participants, with two in the placebo group, five in the 200-mg group and five in the 400-mg group (2.53 (0.48 to 13.44)). There were no discontinuations due to headache. There were more total discontinuations in the 400-mg group (12 participants) than in the 200-mg (3 participants) and placebo (5 participants) groups.
Eosinophilia occurred in ten participants, with five in the 200-mg group and five in the 400-mg group. Of these ten participants, three were permanently removed from the study. The study drug was discontinued temporarily in two participants. Eosinophil increases were asymptomatic and none were classified as serious AEs (SAEs). Four participants exhibited eosinophil increases to levels greater than 500 per mm3 above baseline. These values resolved to within the normal range by each participant’s next scheduled visit with a time range of approximately 1 month. In the six participants with lower levels of eosinophil elevation, four were found to return to a normal level at the 1-month follow-up and two participants discontinued the study before follow-up laboratory testing. Eosinophilia did not occur in the placebo group.
A total of 33 participants (14%) experienced AEs considered to be related to the study medications. Of these participants, 8 (10%) received placebo, 12 (15%) received 200 mg LM11A-31 and 13 (16%) received 400 mg LM11A-31 (Table 2). A total of 15 SAEs occurred in the study across 15 participants. Of these participants, two experienced an SAE before dosing and were considered screening failures. Of the remaining participants, four were in the placebo group, two were in the 200-mg group and seven were in the 400-mg group. One SAE (gastrointestinal bleeding) occurred after 16 consecutive days of dosing and was classified as possibly being related to LM11A-31 treatment. This participant withdrew from the study, was found by endoscopic exam to have a gastric ulcer of unknown duration and fully recovered. No other gastrointestinal bleeding was reported in the study.
Within the ITT population, the study medication was discontinued in 20 participants in total. Reasons for discontinuation were AEs (12 participants; placebo, n = 2; 200 mg LM11A-31, n = 2; 400 mg LM11A-31, n = 8), SAEs (4 participants; placebo, n = 1; 400 mg LM11A-31, n = 3) and withdrawal of consent (4 participants; placebo, n = 2; 200 mg LM11A-31, n = 1; 400 mg LM11A-31, n = 1). The most common reason for discontinuing the study was gastrointestinal symptoms (seven participants; placebo, n = 1; 200 mg LM11A-31, n = 1; 400 mg LM11A-31, n = 5) followed by eosinophilia (three participants; 200 mg LM11A-31, n = 1; 400 mg LM11A-31, n = 2). One participant died during the trial. This participant was in the placebo group and cause of death was pancreatic adenocarcinoma.
No significant abnormalities within the placebo or LM11A-31 groups were identified for participant vital signs (blood pressure, heart rate, respiratory rate and body temperature), 12-lead electrocardiogram or clinical laboratory assessment (hematology, biochemistry, coagulation, serology and urinalysis). MRI did not detect findings that raised concern regarding drug safety, including amyloid-related imaging abnormalities (ARIAs).
Assessment with the Columbia Suicide Severity Rating Scale detected no differences among treatment groups (P > 0.1).
Given that p75NTR may affect the vascular system49,50, it was of particular interest to analyze systolic and diastolic blood pressure values across the three treatment groups. No significant differences in these measures at screening were observed across the three groups (PKruskal–Wallis > 0.1 for each). No significant longitudinal differences in systolic blood pressure were observed with a Kruskal–Wallis test (P = 0.691). Longitudinal changes in diastolic blood pressure differed significantly among the three groups (P = 0.036). The median change in diastolic blood pressure was +1 mm Hg in the placebo group, 0 mm Hg in the 200 mg LM11A-31 group and −2 mm Hg in the 400 mg LM11A-31 group. Post hoc testing with Dunn’s test revealed that the median longitudinal change in diastolic blood pressure was significantly different in the 400 mg LM11A-31 group compared to the placebo group (P = 0.010). No other significant differences were detected among groups. The magnitude of longitudinal change in diastolic blood pressure was not clinically significant.
In all, the Data Safety Monitoring Board concluded that LM11A-31 caused no overall safety concerns and its safety profile was compatible with future larger-scale testing. Thus, the primary trial endpoint of safety was met.
Analysis of secondary and exploratory outcomes
All secondary endpoints were prespecified in the registrations (EU Clinical Trials: 2015-005263-16; ClinicalTrials.gov: NCT03069014). Prespecified exploratory outcome measures were determined on the basis of the results of preclinical studies43,46,47,51 and were described in the statistical analysis plan. Before assessing longitudinal treatment effects on the secondary and the prespecified exploratory outcome measures, we assessed the baseline characteristics of the clinical trial cohort on these measures. To do so, we computed pairwise Spearman correlations among CSF, imaging and cognitive data across all participants at baseline (Extended Data Fig. 1). Overall, the baseline interrelationships among the secondary and prespecified exploratory measures broadly recapitulated those found in prior AD biomarker studies52,53,54.
Having characterized the relationships among CSF biomarkers, neuroimaging biomarkers and clinical tests at baseline, we next examined whether longitudinal changes differed between placebo and LM11A-31. Results from preclinical studies and a prior phase 1b safety and CSF pharmacokinetic trial (F.M.L., unpublished data) suggest that both doses of LM11A-31 (200 mg and 400 mg twice daily) included in the present trial would reach brain exposure levels consistent with engagement of p75NTR-related mechanisms. Consistent with these observations, longitudinal changes in CSF, imaging region-of-interest analyses and cognitive tests did not differ between the two dose arms for 16 of the 17 variables assessed (Extended Data Table 1). Analyses of secondary and exploratory endpoints by dose group are presented in Extended Data Figs. 2–4. For further secondary and exploratory data analyses, we pooled participants from the 200-mg and 400-mg arms into a single LM11A-31 group.
For the analysis of secondary and prespecified exploratory endpoints, longitudinal changes in CSF biomarkers were quantified using an annual percent change calculation55 (Methods). Significant differences in median change between placebo and LM11A-31 groups were investigated using Wilcoxon rank sum tests with 95% bootstrap CIs from 5,000 bootstrap iterations.
Secondary outcomes
Secondary CSF outcomes consisted of the following core AD biomarkers: total tau (t-tau), tau phosphorylated at Thr181 (p-tau181), Aβ40 and Aβ42. LM11A-31 significantly slowed longitudinal increases in Aβ42 compared to placebo (Fig. 2a; Prank sum = 0.037). The difference in median annual percent change of Aβ42 in the LM11A-31 group relative to the placebo group was −6.98% (95% CI, −14.22% to −1.45%). LM11A-31 also significantly slowed longitudinal increases in CSF Aβ40 compared to placebo (Fig. 2a; Prank sum = 0.009). The difference in median annual percent change of Aβ40 in the LM11A-31 group relative to the placebo group was −8.96% (95% CI, −17.60% to −1.29%). Longitudinal changes in the ratio of Aβ42 to Aβ40 between the two groups were not significantly different (Prank sum = 0.952). The difference in median annual percent change of the ratio of Aβ42 to Aβ40 between LM11A-31 and placebo was −0.42 (95% CI, −2.90% to 2.49%). Overall, these findings indicate that longitudinal AD-related increases in CSF Aβ42 and Aβ40 were slowed or reversed by LM11A-31, while the ratio of Aβ42 to Aβ40 was unaffected….
