TMS as a Treatment for Apathy in Alzheimer's Disease

Apathy is among the most common neuropsychiatric symptoms in Alzheimer's Disease (AD), with prevalence estimates up to 72%. Apathy in AD is associated with significant morbidity, high caregiver burden, and is correlated with disease severity. While other behavioral symptoms are clinically addressable in AD, there are currently no effective treatments for apathy.

Executive aspects of apathy involve deficits in goal-directed planning, rule maintenance, cognitive flexibility, and set-shifting. Motivational aspects of apathy involve reward anticipation, valuation of the effort needed to achieve a goal, and the consummatory experience once a goal is achieved. Importantly, these discrete cognitive-behavioral dimensions can be mapped onto discrete fronto-striatal circuits. For example, executive deficits are likely to map onto dorsolateral prefrontal-striatal loops (a dorsal apathy-related circuit). In contrast, motivational processing is localizable to projections between the ventral striatum (VS) and the anterior cingulate cortex (ACC), ventromedial prefrontal cortex (vmPFC) and orbitofrontal cortex (OFC) (a ventral apathy-related circuit).

Dopamine (DA) supports processing in both the dorsal and ventral circuits, and there is voluminous evidence that DA plays roles in encoding reward prediction/valuation and motivation. For example, the ventral circuit receives dense monosynaptic dopaminergic projections from the ventral tegmental area, and several studies have invoked this pathway in reward processing. DA also plays a role in executive function. Moreover, apathy is observed in patients with disorders of DA availability, in patients taking DA antagonists and with lesions in the meso-cortico-limbic pathway. More specifically, AD patients exhibit decreased striatal DA availability. Taken together, decreased DA transmission in the dorsal and ventral circuits represent putative neurochemical mechanisms for apathy in AD.

Repetitive transcranial magnetic stimulation (rTMS) is capable of modulating fMRI resting-state functional connectivity (FC) in a circuit-specific manner. For example, our group, as well as others, have demonstrated the feasibility of modulating network FC through stimulation of a cortically accessible network node. Other studies have demonstrated behavioral changes with circuit-specific neuromodulation. Prefrontal rTMS can also modulate monoaminergic neurotransmitter release in a regionally specific way. For example, intermittent theta burst stimulation (iTBS, a form of rTMS) increases tonic endogenous DA release in animal models, and this effect can be blocked by DA receptor antagonists. There is also evidence in humans that topographically specific changes in DA release can be elicited based on the circuit stimulated. For instance, Strafella and colleagues used an rTMS-11C- raclopride paradigm to show regionally specific changes in DA availability following dorsolateral prefrontal (dlPFC) stimulation and primary motor cortex stimulation, respectively. We recently extended this work by employing a combined MRI-PET paradigm with 11C-raclopride. In that protocol, three subjects underwent iTBS to two adjacent dlPFC targets. These targets were functionally identified by the maximal FC with two dorsal striatal regions: the left caudate and left putamen, and these were seeded based on coordinates derived from large datasets. Targets were stimulated based on modeled E- field distributions. As predicted, iTBS induced regionally specific changes in striatal DA availability, with changes in 11C- raclopride binding potentials (BPND) occurring at the sites of maximal FC with the respective iTBS targets.

Aim 1: To optimize rTMS targeting of apathy-relevant circuits in AD patients with apathy: Hypothesis 1.1: FC estimates, acquired with accelerated multi-band imaging, will identify, on an individualized basis, rTMS accessible cortical targets (e.g., in vmPFC/mOFC and in dlPFC) which are most functionally coupled to relevant striatal regions (e.g., the VS and dorsal caudate) in a group of AD patients with apathy. Hypothesis 1.2: State-of-the-art E-field modeling will optimize coil position and orientation needed to best stimulate the apathy-relevant circuits. Modeling will be informed by (a) structural estimates of cortical atrophy, (b) diffusion MRI tractography estimates of white matter fiber bundles from cortical targets.

Aim 2: To use rTMS to differentially modulate FC in apathy-relevant circuits in AD patients with apathy: Hypothesis 2.1: rTMS to the dorsal circuit will increase FC specifically in the dorsal circuit. Hypothesis 2.2: rTMS to the ventral circuit will increase FC specifically in the ventral circuit. Hypothesis 2.3: Sham rTMS will not affect FC in either circuit.

Aim 3: To use rTMS to differentially modulate DA in apathy-relevant circuits in AD patients with apathy: Hypothesis 3.1: rTMS to the dorsal circuit will increase DA availability, (as measured by changes in 11C- raclopride binding potential (BPND)), specifically in the dorsal circuit. Hypothesis 3.2: rTMS to the ventral circuit will increase DA availability specifically in the ventral circuit. Hypothesis 3.3: Sham rTMS will not affect DA availability in either circuit.

Aim 4: To assess behavioral changes induced by modulation of apathy-relevant circuits in AD patients: Hypothesis 4.1: rTMS to the dorsal circuit will selectively improve executive measures of goal-directed behavior, as assessed by fMRI task performance on planning portions of the PACT. Hypothesis 4.2: rTMS to the vental circuit will selectively improve motivation and reward anticipation, as assessed by fMRI task performance and activation on motivational parts of the PACT, and on mood scales. Hypothesis 4.3: Sham rTMS will have no impact on fMRI task performance or activation, nor on mood scales.