Playing Russian Roulette with Alzheimer’s Disease Patients: Do the Cognitive Benefits of Lecanemab Outweigh the Risk of Edema,Stroke and Encephalitis?

As recently reported, lecanemab has passed FDA efficacy and safety criteria for accelerated approval. Despite the improved safety of this monoclonal antibody as compared to treatment with previous “mAbs” to amyloid, approximately 1 in 8 patients treated with lecanemab will still develop amyloid-related imaging abnormalities (ARIA) [1]. Increased numbers of both ARIA-H (microhemorrhages, macrohemorrhages, and/or superficial siderosis; 8.3%) and ARIA-E (vascogenic edema and/or sulcal hyperintensity; 10.9%) were identified in subjects treated with lecanemab as compared with placebo. Adjusting for concurrent cases of ARIA-E and ARIA-H of 7.2% results in 12.0% of the study population developing one or both forms of ARIA. Moreover, those taking blood thinners or clot-busting drugs appear to be at even greater risk of hemorrhage, stroke, and death (https://www.alzforum.org/news/research-news/should-people-blood-thinners-forego-leqembi#comment-form).

As predicted two decades earlier, the removal of amyloid-β from the brain has major ramifications on the integrity of the vascular, connective, supportive, and functional tissues of the brain [2 –4]. This ‘vascular scab’ theory of amyloid-β was developed based on our understanding of the biochemistry and pathology of amyloid-β as a barrier protein [3, 5], and has since been supported by numerous histological [6, 7] and neuroimaging [8 –10] studies. Indeed, the identification of ARIA following the removal of amyloid-β from the brain starting with bapineuzumab, and as also seen with gantenerumab, aducanumab, and lecanemab [1, 10] clearly demonstrates not only the efficacy of the drug in removing various amyloid species from the brain but also its destabilizing nature on the vasculature.

Amyloid-β as a barrier protein that both seals the vasculature to prevent hemorrhage and edema [3] and provides antimicrobial activity against bacteria, yeast, and viruses [11 –13], is a crucial mechanism for maintaining structural integrity as well as warding of pathogens within the brain (as reviewed in [5, 14]). While the evolution of amyloid-β monoclonal antibodies towards the removal of specific amyloid-β species such as the lower molecular weight (neurotoxic) oligomeric species (lecanemab) appears to lessen ARIA, it does nonetheless open the possibility of decreasing the innate immune functions of these oligomeric species that together with copper and zinc allow the binding, entrapment, and killing of pathogens. Based on the biochemistry of this protein, in addition to the increased rate of hemorrhage and edema observed in patients treated with lecanemab compared with untreated patients [1], one might also predict an increased risk of meningitis and/or encephalitis. Extrapolation of the amyloid load curve presented in Fig. 2B [1] indicates further decreases in amyloid load are likely with continued treatment (beyond the ∼50% reduction at 18 months) and that the risk for vascular compromise and associated conditions could increase.

While it is hard to deny an AD patient a treatment with the prospect of slowing their declining cognition, the risk of vascular compromise cannot be ignored [4]. The cognitive benefits of lecanemab although significant, are in reality quite small and their clinical relevance doubtful (0.45 points in Clinical Dementia Rating– Sum of Boxes score versus placebo at 18 months) [15]. It should also be noted that there is a relatively high frequency of infusion-related reactions (26.4% of the participants). If a treatment were able to completely halt (or reverse) AD progression, we might agree that significant side effects are justified. Unfortunately, like all of its monoclonal antibody precursors, lecanemab at best belongs to a greatly flawed class of drugs, based on a flawed hypothesisof AD.

Given the liabilities of attempting to modify a normal physiological response to damage, greater attention should be focused on the upstream causes of AD, and not the downstream sequalae of the disease like amyloidosis and neurofibrillary tangle formation. Specifically, the events that promote the age-related re-entry of neurons into an aberrant cell cycle leading to neuron loss and dysfunction in late-onset AD (i.e., hormonal dysregulation of AβPP processing), and in early-onset AD (familial mutations that alter AβPP processing; reviewed in [16]) are avenues for developing real disease modifying therapies for this disease [17 –19].

We would like nothing more than to see an efficacious and safe treatment for AD as we have afflicted family members. However, the odds of serious adverse events with lecanemab (with close to 1 in 3 homozygote APOE ɛ4 carriers developing ARIA) [1] together with the limited cognitive benefit calls into question which patient populations, if any, should receive treatment. It will be argued that ARIA is clinically manageable and that only 3.5% of patients are symptomatic. However, there will be significant adverse events and loss of life with continued use of this drug. Only long-term studies of lecanemab in discrete populations, and over longer time periods will ultimately determine the cost-benefit of this anti-amyloid therapy. While Leqembi might be the current darling of AD therapeutics, its luster is sure to tarnish as adverse events mount.