Depression in late-life is associated with greater likelihood of development of all-cause dementia, including Alzheimer’s and vascular dementia (Diniz etal., Reference Diniz, Butters, Albert, Dew and Reynolds2013). Cognitive impairment is frequently observed in late-life depressed patients and the deficits often persist, despite remission of depressive symptoms (Bhalla etal., Reference Bhalla2009; Koenig etal., Reference Koenig2015). The development of effective treatments for cognitive impairment associated with late-life depression might potentially slow or prevent cognitive decline. An understanding of the neurobiology of cognitive impairment in late-life depression would inform treatment development. Structural, functional, and molecular imaging studies in late-life depressed patients have implicated changes in neural circuits associated with mood symptoms and cognitive deficits, molecular mechanisms including oxidative stress and mitochondrial dysfunction and neurochemical changes and, in some studies, Alzheimer’s Dementia (AD) pathology (Manning and Steffens, Reference Manning and Steffens2018; Mathias etal., Reference Mathias, Monette, Harper and Forester2017; Hirao and Smith, Reference Hirao and Smith2014). The integration of neuroimaging methods into clinical trials can identify molecular mechanisms and predictors of persistent cognitive deficits.

Lavretsky and colleagues conducted a 6-month, randomized, double-blind, placebo-controlled trial of combined escitalopram and memantine for older adults with depression and subjective memory complaints (Lavretsky etal., Reference Lavretsky2020). While both groups treated with escitalopram demonstrated significant improvement in mood, the group treated with escitalopram and memantine versus escitalopram and placebo showed significantly greater improvement in delayed memory and executive function. The improvement was observed after 6 months of naturalistic follow-up after the 6- month clinical trial was completed. Krause-Sorio and colleagues reported the results of the subgroup of individuals in the clinical trial who underwent positron emission tomography (PET) scans with a radiotracer for AD pathology ([¹⁸ F]-FDDNP) to determine the relationship between AD pathology and the mood and cognitive responses to treatment (Krause-Sorio etal., Reference Krause-Sorio2020). In both treatment groups, greater frontal cortical AD pathology (higher volume of distribution ratio of [¹⁸ F]-FDDNP) was associated with greater improvement in executive function, not mood, at 6- month follow-up. No correlations between AD pathology in other cortical regions (medial temporal or parietal cortices) and cognitive domains (memory) were observed. As the authors discuss, the small sample size of the PET substudy may have limited the ability to detect differences between the treatment groups, including mood and other cognitive outcomes.

Several aspects of the experimental design are noteworthy. The selection of patients with depression and subjective memory complaints is likely to include patients vulnerable to cognitive decline with modest levels of AD pathology, who may benefit to a greater extent from treatment compared to patients with more sever pathology. The treatment paradigm employed, escitalopram combined with memenatine, is innovative as combined treatment may have a synergistic neurobiological effect on primary (serotonin or glutamate) or downstream neurobiological mechanisms (i.e. promoting synaptic plasticity; Rogawski and Wenk, Reference Rogawski and Wenk2003). It is important to note that the PET radiotracer used, [¹⁸ F]-FDDNP, binds to multiple AD pathologies, including beta-amyloid and tau, compared to PET radiotracers that have been developed more recently that are more selective for beta-amyloid or tau (Villemagne etal., Reference Villemagne, Doré, Burnham, Masters and Rowe2018). Both PET imaging and neuropathological studies of AD pathology in late-life depression have shown modest increases in patients compared to controls, except in patients who meet criteria for mild cognitive impairment (Butters etal., Reference Butters2008; Sweet etal., Reference Sweet2004). The clinical and neurobiological correlates of AD pathology in patients with late-life depression are not well understood. Specifically, an understanding of the synaptic changes associated with AD pathology may play an important role in the heterogeneity of mood and cognitive responses to antidepressant treatment in late-life depression (Colom-Cadena etal., Reference Colom-Cadena2020).

The results reported by Lavretsky etal. and Karuse-Sorio etal., as well as preclinical and human data, support further investigation of combined escitalopram/memantine treatment in patients with late-life depression and AD pathology. In amyloid transgenic mice (5XFAD), memantine improved memory performance in mice with lower compared to higher beta-amyloid concentrations (Devi and Ohno, Reference Devi and Ohno2016). Several pilot studies of combined citalopram and memantine treatment in late-life geriatric depressed patients with cognitive impairment and in patients with AD showed improvement in mood and cognition (Pelton etal., Reference Pelton, Harper, Roose, Marder, D’Antonio and Devanand2016; Zhou etal., Reference Zhou, Wang, Xin, Kong and Wang2019). While these studies did not include imaging of AD pathology, it is likely that AD pathology was present in these cognitively impaired patients. Future studies might employ other molecular imaging measures such as magnetic resonance spectroscopy or PET neurotransmitter imaging to understand the neurobiological mechanisms of combined escitalopram and memantine treatment, as well as to develop predictors of mood and cognitive response.

The approach employed by Krause-Sorio and colleagues to incorporate neurobiological measures into randomized clinical trials, especially of novel interventions, is highly informative. Valuable information is obtained about more effective treatments and predictors of treatment response, as well as neurobiological mechanisms. Such studies are challenging but may ultimately inform the development of more effective treatments to improve cognition and to potentially slow or prevent cognitive decline in patients with either late-life depression or with depressive symptoms in neurodegenerative disease.