Limited Utility of Imaging-based (N) Markers in Predicting Cerebral Spinal Fluid Alzheimer's Disease Biomarker Positivity in Lecanemab-Eligible Mild Cognitive Impairment

Takeshi Kuroda^1*Yukiko Mori¹Daiki Shoji¹Satomi Kubota²Moeko Shinohara³Kenjiro Ono³Hidetomo Murakami¹

• ¹School of Medicine, Showa University, Tokyo, Japan
• ²Showa Daigaku Fujigaoka Byoin, Yokohama, Japan
• ³Kanazawa Daigaku, Kanazawa, Japan

Background: Early detection of mild cognitive impairment (MCI) due to Alzheimer's disease (AD) is essential for the timely initiation of anti-amyloid β (Aβ) therapy. In clinical practice, cerebrospinal fluid (CSF) testing and amyloid positron emission tomography (PET) are not feasible for all patients; therefore, initial screening sometimes relies on imaging-based neurodegeneration (N) markers within the amyloid-tau-neurodegeneration AT(N) framework. However, the diagnostic accuracy of imaging-based (N) markers remains uncertain in real-world MCI cohorts being evaluated for anti-Aβ antibody therapy. We aimed to assess the utility of imaging-based (N) markers in predicting AD pathology in patients with MCI who may be eligible for lecanemab. Methods: Thirty-six patients with MCI who were potentially eligible for lecanemab underwent CSF biomarker testing and were subsequently classified as MCI unlikely due to AD (MCI non-AD; n = 14) or MCI due to AD (MCI-AD; n = 22). Demographics, general risk factors, neuropsychological test scores, and imaging-based (N) markers— medial temporal atrophy (MTA) on magnetic resonance imaging (MRI) and regional cerebral blood flow (CBF) reductions on single-photon emission computed tomography (SPECT)—were compared between groups. Results: MCI-AD was diagnosed in 22 patients (61%). Demographics, neuropsychological test scores, and most general risk factors did not differ between groups, except for diabetes, which was more frequent in the MCI non-AD group. Overall, 34 of 36 patients (94%) were classified as (N)+ based on MRI or SPECT (91% MCI-AD; 100% MCI non-AD). There were no significant differences in mean MTA scores or the degree of CBF reduction between groups. In contrast, the proportion of MRI (N)+ patients was significantly higher in the MCI non-AD group than in the MCI-AD group, and two patients with MCI-AD were (N)− on both MRI and SPECT. Conclusions: Relying on imaging-based (N) markers to select MCI patients who may be eligible for lecanemab prior to CSF biomarker testing may lead to inefficient diagnostic pathways and may fail to identify patients who could benefit from anti-Aβ therapy.