Neuromodulatory subcortical systems in Alzheimer's disease

Five years after Alois Alzheimer was buried in a Frankfurt cemetery, his contemporary, Karl Kleist, founded the Psychiatrischen und Nervenklinik (Psychiatric and Neurological Clinic) just 5 miles away across the city. Kleist's academic clinic formalized a proto‐framework for modern neuropsychiatry and behavioral neurology by establishing clinico‐anatomic approaches to neurologic and psychiatric diseases. ¹ Alzheimer's disease (AD) as a clinical‐pathological entity was, in those early decades, being delineated by multiple contemporaries—notably Oskar Fischer in Prague, whose 1907 description of miliary plaques in senile dementia paralleled Alzheimer's own. In 1928, a report from Kleist's clinic offered a notable reflection on the clinical and pathological observations of patients with AD. ² Its authors argued that non‐memory symptoms, particularly affective symptoms and sleep changes, had been under‐emphasized in the AD symptomatology initially published, for example:

Upon reviewing the cases reported in the literature and our own material, it appears that essential, recurring symptoms have not been given sufficient prominence. … First, it should be noted that affective disturbances are frequently the most prominent manifestations right at the onset of the illness.

Bei Durchsicht der in der Literatur mitgeteilten Fälle und unseres eigenen Materials scheinen seither wesentliche, immer wiederkehrende Symptome zu wenig in den Vordergrund gestellt worden zu sein. … Zuerst sei auf die Tatsache aufmerksam gemacht, daß häufig gleich zu Beginn der Erkrankung affektive Störungen im Vordergrund der Erscheinungen stehen.

Applying Kleist's clinico‐anatomic approach, the report extended its neuropathologic analysis beyond the cortical regions canonically associated with senile dementia to subcortical structures. The authors pointed to observed degeneration patterns in several “vegetative” nuclei, including the periventricular nucleus, tuberomammillary nucleus (at the time referred to as the tuberal/mamillary hypothalamic region), dorsal raphe nucleus, and locus coeruleus (LC), to justify the expanded symptomatology they proposed for AD.

Independent efforts spread across continents in the decades after 1928 repeatedly arrived at this same conclusion. Topographic mapping of AD‐type neurofibrillary tangles in the early 1960s broadened the anatomical frame of the disease beyond canonical cortical regions, ³ and by the mid‐1960s, systematic studies of the brainstem and hypothalamus had established neuromodulatory involvement as a reproducible neuropathologic phenomenon. ⁴

In the 1980s, this groundwork converged with parallel work across several continents. American groups documented degeneration of the nucleus basalis of Meynert and other basal forebrain cholinergic populations, ⁵ , ⁶ giving rise to the cholinergic hypothesis ⁷ and, in turn, to cholinesterase inhibitors—still part of symptomatic AD treatment today. ⁸ British groups produced the first stereologic studies of LC cell loss in senile dementia of the AD type, ⁹ with subsequent work mapping the magnitude, regional distribution, and morphological character of this loss. ¹⁰ , ¹¹ , ¹² Studies in Japan and the United States documented tangles and neuronal loss in the dorsal raphe nucleus, ¹³ anticipating a modern focus on the serotonergic system in AD. Other studies pointed to hypothalamic and brainstem nuclei relevant to autonomic, arousal, and affective symptoms. ¹⁴ , ¹⁵ The reports cited above are necessarily a partial sampling of a broader, internationally distributed effort whose very scatteredness across institutions, languages, and decades is part of why it did not coalesce into a coherent program until much later. With the publication of the amyloid cascade hypothesis in 1992, the field's mechanistic and therapeutic focus became overwhelmingly cortico‐centric—what Parvizi would later call “corticocentric myopia” ¹⁶ —even as neuropathologic reports of early neuromodulatory involvement, especially in the LC, continued to mount.

With earlier work establishing the involvement of neuromodulatory subcortical structures (NSSs) in AD, the sequence of degeneration remained less clear. Situating the NSSs within the temporal arc of AD required phosphorylated tau (p‐tau) immunohistochemistry, systematic sampling of brainstem and basal forebrain nuclei, and well‐characterized autopsy collections staged across Braak categories. As staged autopsy cohorts became more widely available, multiple groups placed NSS degeneration within the temporal sequence of AD ¹⁷ , ¹⁸ , ¹⁹ and in 2006, the Braak staging system incorporated the NSS as an early “Braak 0” phase of neurofibrillary tangle progression in AD. ²⁰ , ²¹ Quantitative work followed, including evidence that p‐tau changes in the dorsal raphe and many other NSSs may precede transentorhinal involvement. ²² The cover of this special issue features a neuronal cytoplasmic tau inclusion in the LC of a Braak stage “0” case from this latter study.

The 2010s saw an explosion of literature documenting the role of the NSSs beginning in the earliest phases of AD. ²³ These efforts culminated in neuropathologic studies across institutes and cohorts, examination of the NSS in laboratory models of AD, and clinico‐pathological studies. An international consortium was established in 2016 focused on developing biomarkers for the LC in AD, ²⁴ and magnetic resonance imaging measures of LC integrity are now routinely implemented in dementia research centers. This shift is increasingly visible in contemporary disease frameworks (e.g., the 2026 Lancet Seminar on AD) that incorporate subcortical neuromodulatory involvement in early phases. ²⁵

Degeneration of NSSs has broad implications for AD research and care. Because some of these nuclei are affected early, they are promising targets for biomarkers of the earliest disease phases; investment in NSS biomarkers may therefore expand the window for detection and intervention. NSS degeneration also carries therapeutic implications that are beginning to be clinically assessed. Management of neuropsychiatric symptoms and sleep dysfunction in AD should be informed by the lesion patterns underlying those symptoms. Sleep disturbance, in particular, is associated with faster decline in AD, raising the possibility that effective symptom management may not only improve quality of life for patients and caregivers, but also influence clinical trajectories. Further, loss of neuromodulatory transmitters can affect disease biology across the brain. Loss of noradrenergic input due to LC degeneration, for example, has been associated with exacerbated neuropathologic lesions and neuroinflammation. ²⁶ , ²⁷ , ²⁸ , ²⁹ , ³⁰ For these reasons, interventions that protect or restore neuromodulatory function may have effects beyond the nuclei themselves.

A related question for the field is whether mood, sleep, arousal, autonomic, and other neuromodulatory structure‐related symptoms in AD will come to occupy a role analogous to the non‐motor symptoms of Parkinson's disease: early clinical signals of vulnerable subcortical systems, detectable before the cortical syndrome becomes dominant. ³¹ , ³² If so, neuromodulatory subcortical systems research will reshape not only how AD is mapped pathologically, but also when it is recognized clinically.

A wide breadth of questions related to the NSS in AD is now under consideration. Recognizing that the integration of many technical disciplines and efforts across countries is necessary to advance the field, the Alzheimer's Association's International Society to Advance Alzheimer's Research and Treatment (ISTAART) formed the Neuromodulatory Subcortical Systems Professional Interest Area (NSS PIA) in 2020. For many years, neuromodulatory structure‐relevant work had been scattered across geography and disciplines creating disparate threads lost to more cohesive narratives such as the amyloid cascade hypothesis. The NSS PIA created a cohesive forum within ISTAART's larger infrastructure in which investigators could find one another, align methods, and develop collaborations. A 2023 publication from the NSS PIA describes the resulting research priorities in greater depth. ³³

This special issue in Alzheimer's & Dementia: The Journal of the Alzheimer's Association is a collection of review papers and primary research articles curated by the NSS PIA. The collection spans the breadth of contemporary neuromodulatory structure research, with papers addressing LC integrity and imaging, cholinergic and serotonergic mechanisms, hypothalamic involvement, sleep and thermoregulation, arousal and pupillary physiology, stress systems, sex differences, vascular contributions, animal models, and therapeutic interventions targeting neuromodulatory dysfunction. Together, they show that NSS research has moved from scattered historical observations to a coordinated, multidisciplinary field with direct implications for early detection, symptom biology, disease mechanisms, and therapeutic development.

It is now nearly 100 years since the 1928 report from Kleist's clinic first suggested that degeneration of NSSs is a clinically relevant aspect of the AD pathological process. The > 30 publications in this special issue reflect the field's overdue recognition that NSS dysfunction is not incidental in AD, but central to understanding its earliest biology and clinical heterogeneity.

Disclosures

All authors are current or former chairs of the Neuromodulatory Subcortical Systems Professional Interest Area of ISTAART and served as guest editors for this special issue of Alzheimer's & Dementia: The Journal of the Alzheimer's Association. A.J.E. sits on the advisory board for the Genomic Answers for Children's Health Alliance under the organization of Leavitt Partners. N.F. has received payment from Lilly. H.J. has received consulting fees as an ad hoc NIH‐NIH Study Section member and is a member of the ISTAART advisory board. L.T.G. has received payment or honoraria from Elsevier, Orator Inc., Springer Nature, Biogen, and Alamar as well as consulting fees from UCB. L.T.G. additionally sits in leadership positions for the Global Brain Health Institute and MSAG.

CONFLICT OF INTEREST STATEMENT

The authors declare no conflicts of interest. Author disclosures are available in the Supporting Information.

Supporting information

Supporting information: COI combined (2).pdf