Andres M. Lozano, MD, PhD, FRCSC, FRSC, FCAHSAndres M. Lozano, MD, PhD, FRCSC, FRSC, FCAHS
An estimated 40 to 50 million individuals worldwide are living with Alzheimer disease (AD) and other dementias.1 Characterized by progressive memory impairment and deterioration in quality of life, the disease is expected to pose a significant public health burden in the coming decade. With the isolation and characteriza-tion of the amyloid plaque core protein,2 investigators have focused their search on therapeutics that can combat and clear the plaques. Unfortunately, most of the investigational agents have not demonstrated clear efficacy in reversing the progression or cognitive symp-toms of the disease.3-5 Neuromodulation techniques, however, may provide a promising treatment avenue, with the goal of modifying nerve activity via pharmaceutical or electrical stimulus.

Deep brain stimulation (DBS) is a well-established surgical neuromodulation technique that has been used in the treatment of Parkinson disease (PD),6,7 essential tremor,8 obsessive-compulsive disorder,9 and epilepsy.10 It also has potential uses in obesity,11 eating disorders,12 depression,13 and Tourette syndrome.14 DBS is an invasive procedure administered under local anaesthesia. Physicians make a small opening in the skull through which they insert an electrode. Digital brain mapping technologies deter-mine the precise locations for stimulation, and an implanted pulse generator (magnet) attached to the electrode can then regulate the timing and strength of stimulatory signals. According to estimates, over 160,000 patients around the world have undergone a DBS implant.15

DBS and Memory

Patients with seizures who have received DBS have demonstrated improvement in memory,16 and results of several studies have shown a potential improvement in visual memory in patients with epilepsy undergoing DBS.17-19 In rodent studies, DBS improved spatial memory and reversed the negative memory effects of scopolamine.20,21 In addition, in preclinical studies in both animal models and humans, DBS has demonstrated neuroprotective effects.22 Encouraged by these findings, practitioners are now exploring DBS in the treatment of AD. “With DBS, we believe that we are able to adjust the activity of brain circuits that are malfunctioning,” Andres M. Lozano, MD, PhD, FRCSC, FRSC, FCAHS, Dan Family Chair in Neurosurgery at the University of Toronto in Ontario, Canada, told NeurologyLive®. “In the case of Alzheimer disease, this could apply to circuits involved in memory and cognitive function.”

The Memory Circuit in ADResearch indicates that the cognitive impairment associated with AD results from widespread neuronal network dysfunction in the brain.23-25 The Papez circuit (or medial limbic circuit) involving the hippocampus, the fornix, the mammillary bodies, the anterior thalamic nuclei, and the posterior cingulate region26 is implicated in episodic memory and is therefore a viable candidate for study in AD.

The potential role of DBS in AD was a serendipitous discovery. Lozano’s team was exploring the use of DBS in the treatment of obesity, when hippocampal stimulation—particularly of the fornix—resulted in unexpected memory retrieval.27 Stimulation of these memory circuits in rats stimulated neurogenesis in the hippocampus.28 Subsequently, forniceal DBS resulted in the expression of neurotrophic factors and markers of synaptic plasticity involved in memory processing.29

The ADvance phase 1 study (NCT01608061) of DBS in AD included 6 patients with mild to moderate AD.30 DBS of the fornix improved cognitive function and increased cerebral glucose metabolism and hippocampal volume. Although DBS was generally well tolerated, 2 of the 6 patients did report worse prognoses after the surgery.30 Following 1 year of DBS, the patients presented with increased cerebral glucose metabolism and neural connectivity, in contrast with the decline associated with AD.31 Furthermore, fornix DBS appears to slow down the hippocampal atrophy associated with AD.32 

Phase 2 of the ADvance study, comprising 42 patients with mild AD, found an increase in cerebral glucose metabolism, which became nonsignificant at 12 months. In addition, investigators observed no significant improvement in cognitive outcomes. The long-term safety profile of fornix DBS was similar to that observed in the treatment of movement disorders.33 Four patients required addi52tional surgery for infection34 and the repositioning and drainage of chronic subdural hematoma.35

Age-Related DBS Effects

An interesting finding of the ADvance study was the heterogeneity in cognitive decline between patients ≥65 and <65 years. Patients ≥65 years displayed an average reduction of 4.1 points in cognitive deterioration, as assessed via the AD Assessment Scale–Cognitive Subscale. Furthermore, cerebral glucose metabolism increased by 14% to 20% after 12 months of DBS in those patients; the cohort of patients <65 years uniformly exhibited declined glucose metabolism.36 “However, none of the 42 patients displayed any permanent neurological adverse effects from DBS,” Lozano said.

A 2-year follow-up study that sought to examine the long-term safety of fornix DBS also demonstrated no serious adverse effects of treatment. Younger patients manifested a rapid cognitive decline irrespective of treatment. However, older patients (>65 years) receiving active stimulation deteriorated less compared with their nonstimulated counterparts.37 

The cause of this heterogeneity is not fully clear. “Younger patients with Alzheimer disease may have a very malignant, severe form of Alzheimer that deteriorates very rapidly with or without DBS, whereas the older patients have a less malignant course that responds to DBS,” Lozano speculated. Another possible explanation could be that the autosomal dominant mutations that characterize early-onset AD might render the patients with an aggressive prognosis that is not alleviated with DBS.38 “Based on these findings, we have now decided to only operate on older patients because their illness is not progressing that rapidly,” Lozano said.

The phase 2b/3 trial is a 12-month, sham-controlled, double-blind study examining the safety and efficacy of fornix DBS in patients with mild AD. “We have just begun enrolling,” Lozano told NeurologyLive®. This study will be the largest trial of DBS in North America and Europe, exceeding previous trials of DBS for depression and epilepsy. The investigators intend to examine DBS efficacy and safety for up to 3 years after stimulation.

Nucleus Basalis of Meynert

Other research points to another brain target for DBS: the nucleus basalis of Meynert (NBM). Although PD is generally characterized as a motor syndrome resulting from nigrostriatal dopaminergic denervation, recent studies have begun to explore nonmotor features such as cholinergic denervation. Evidence indicates that NBM, the source of cholinergic innervation in the cerebral cortex, may atrophy in PD, leading to protein aggregation and cognitive impairment.39,40 

Subsequently, DBS stimulation of the NBM in a 77-year-old patient with PD dementia improved apraxia.41 Interestingly, in a 71-year-old man suffering from PD dementia syndrome, DBS of the NBM markedly improved cognitive function. The patient reported improvement in attention, concentration, alertness, and drive, all of which resulted in improved quality of life.42 These findings formed the rationale for evaluating NBM DBS as a therapeutic strategy for AD.

Jens Kuhn, MD, of the University of Cologne in Germany, has pioneered the use of NBM DBS in the treatment of AD. Results of a phase 1, double-blind, sham-controlled study showed a reduction in cognitive decline in 4 of 6 patients. In contrast with results of fornix DBS, younger patients (<65 years old) with mild AD benefited more from NBM DBS. Furthermore, imaging studies revealed that patients with AD with less atrophy benefited more from the surgery.43,44 However, a limitation of the study is that for ethical reasons, all the patients also continued acetylcholinesterase inhibitor therapy. Therefore, the effect of DBS on the cholinergic system may have been disguised.

Mechanism of Action

The pathology of AD has evolved past the singular view of amyloid plaque aggregation. The disease is characterized by multiple neuro-pathological changes that are thought to interfere with the function of memory networks. Many studies have demonstrated that dysfunction of synapses in the hippocampus precedes amyloid plaque aggregation and neuronal cell death.45,46 Furthermore, level of cognitive impairment appears to directly correlate with synaptic loss.47 The hippocampus also demonstrates a proportional decrease in glucose metabolism and gray matter volume with AD-related atrophy.48

DBS is believed to regulate abnormal circuitry by generating action potentials, releasing neurotransmitters, and modulating sodium channels. In rodents, DBS also stimulates hippocampal neurogenesis.28 However, the precise mechanism of cognitive improvement triggered by DBS remains unknown.22 

The proposed etiology of the effect of DBS on AD may include an increase in cerebral glucose metabolism, stimulating hippocampal acetylcholine release, increasing nerve growth factor, increasing neuronal and synaptic activity, regulating neural oscillations, or resetting of theta activity.30,36,49 In rodents, DBS has been found to reduce deposition of amyloid plaques via lysosomal pathways.50 DBS-stimulated reduction in synaptic and neuronal loss also led to an increase in hippocampal volume, but the neuroprotective mechanism is not fully understood.51

The Future of DBS for AD

Despite the effects of DBS in animal studies, clinical studies have not replicated the pronounced clinical effects. A possible expla-nation for this discrepancy could be the complex, interconnected nature of neural networks that regulate cognition. Subsequently, in patients with AD, fornix DBS induced visuospatial function that was preserved for 3 years despite a rapid decline in cognitive function.32 Notably, animals and pharmacological models of AD may not accu-rately represent the pathophysiology of AD in humans.

Clinical doses of AD DBS have so far been guided by similar treat-ments of movement disorders. Therefore, the optimal dose remains unknown. However, the current ADvance II study (NCT03622905) will examine the difference between high and low frequencies of DBS.15 Other modifiable parameters that investigators should evaluate include pulse frequency, width, and voltage of the signal.

Another factor to consider is the effect of continuous versus acute stimulation. Continuous stimulation could lead to depletion of neurotransmitters and synaptic exhaustion. Therefore, a need exists for optimal stimulation strategies that regulate memory circuits without exhausting them.

Ethical Considerations of DBS

Experts are debating the ethics of conducting DBS research from many perspectives, especially as AD is a progressive disease associated with cognitive decline, which raises many questions about obtaining consent. Research fellow and ethicist Merlin Bittlinger, MA, of the Berlin Institute of Health, told NeurologyLive® that “DBS for geriatric patients with severe Alzheimer disease is a no-go. Severe AD is characterized by dementia, which hampers autonomous-in-formed consent. Without autonomous-informed consent, participation in research that implies more than minimal risk is unethical and would cross the red line.”52

Consequently, current DBS AD studies do not include patients with severe AD. Furthermore, Lozano said that because “patients with mild AD could become cognitively impaired or are already cognitively impaired, we insist on having double consent where both the patients and their families must consent. In this way, if the patient loses the ability to consent because of a decline in the illness, then the family can continue to give consent on their behalf.”

Bittlinger also insisted that from an ethical standpoint, the progres-sive nature of AD and the lifelong follow-up required of DBS necessitates a more robust proof of research rationale. “Animal experiments must be well designed and demonstrate that findings are relevant for translating the intervention to humans. They must allow an informed decision on whether or not the risk of serious unintended effects is unlikely to outweigh any potential clinical benefit.”

DBS studies in humans have also been limited by the period of follow-up. To date, the longest follow-up included only 7 patients for 2 years.37Investigators may identify more adverse events from larger cohorts with longer stimulation periods. “We know from previous research53 on DBS for PD that psychological adverse events are likely to be detected as soon as the number of patients receiving DBS for AD increases into the hundreds or thousands, compared to the limited number of patients who have undergone DBS so far,” Bittlinger said. “However, irreversible psychological effects are not the most pressing ethical problem. Most psychological effects will likely be manageable by adjusting stimulation parameters.”

Noninvasive Brain Stimulation

Noninvasive brain stimulation (NIBS) is a new tool being examined for regulating the function of neuronal networks. Repetitive tran-scranial magnetic stimulation (rTMS) and transcranial direct current stimulation (tDCS) have been shown to induce long-lasting neurological changes. Several studies in patients with AD have demon-strated a beneficial effect on cognitive function.54-56 Like DBS, NIBS also shows better efficacy in patients with mild and moderate AD compared with a severe prognosis.

The adverse effect profile of NIBS appears favorable. Effects associated with rTMS and tDCS have included neck pain, headaches, tinnitus, itching, mild psychiatric effects, and flashes of light.57NIBS remains in the preliminary stages of exploration, with its mechanisms not yet fully understood. Additionally, in rodents, chronic stimulation in the trough has been associated with long-term depression. Therefore, future studies in the field should include a determination of optimal doses and patterns of dose delivery.58

Clinical studies examining the use of NIBS in AD have produced heterogeneous results, partially because of varied stimulation protocols. A need exists for larger, randomized clinical trials that could establish the safety and efficacy profile of NIBS.


In the past decade, the development of novel stimulation technologies has opened many exciting avenues for the treatment of neurological disorders. A potential future treatment strategy could include stimulation of multiple target sites in the brain that engage in cognitive function. Such a synergistic stimulation could exceed the efficiency of lone target stimulation. However, practitioners must also exercise caution in deriving overarching interpretations from these preliminary studies. Efforts should also focus on elucidating the etiology of neuropathological changes induced by DBS and NIBS. Despite their promise, neuromodulation technologies still have a long way to go before practi-tioners can widely apply them.