Researchers identify set of five cognitive phenotypes in MS

An international group of researchers developed a set of five cognitive phenotypes for patients with MS that could provide “a more meaningful measure” of cognitive status in these patients, according to findings published in JAMA Neurology.

The novel approach for measuring cognitive status in patients with MS could improve clinical decision-making and allow for more tailored rehabilitation strategies, according to the researchers.

Ermelinda De Meo, MD

“This research was prompted by the need to go beyond the traditional preserved/impaired classification dichotomy conventionally adopted for cognitive functioning in MS. Indeed, the application of this dichotomous classification in previous studies has led to the inclusion of heterogeneous groups of patients with variable cognitive profiles, thus preventing a clear assessment of neuroanatomical substrates and personalized rehabilitation strategies,” Ermelinda De Meo, MD, of the neuroimaging research unit in the division of neuroscience at IRCCS San Raffaele Scientific Institute in Milan, told Healio Neurology. “In the present study, by adopting a data-driven approach in a large and heterogeneous cohort of patients with MS, we identified five cognitive phenotypes.”

De Meo and colleagues performed a multicenter, cross-sectional study in which they consecutively screened patients with MS who were clinically stable as well as healthy controls from eight MS centers in Italy. Inclusion criteria included no use of psychoactive drugs and no history of other neurological or medical disorders, learning disability, severe head trauma and alcohol or drug abuse. The study enrolled patients between Jan. 1, 2010 and Oct. 31, 2019.

Participants received a neurological examination and a cognitive evaluation using the Rao Brief Repeatable Battery and Stroop Color and Word Test. The researchers also performed a brain MRI in a subgroup of participants. They used latent profile analysis on cognitive test z scores to identify cognitive phenotypes and linear regression and mixed-effects models to characterize clinical and MRI features of each phenotype.

The study included 1,212 patients with MS (mean age, 41.1 years; 64.7% women) and 196 healthy controls (mean age, 40.4 years; 66.3% women).

Deo Meo and colleagues identified five cognitive phenotypes, including preserved cognition (n = 235; 19.4%), mild–verbal memory/semantic fluency (n = 362; 29.9%), mild–multidomain (n = 236; 19.5%), severe–executive/attention (n = 167; 13.8%) and severe–multidomain involvement (n = 212; 17.5%).

“[The first phenotype was] characterized by preserved functioning in all cognitive tests. The second phenotype [was] characterized by mildly reduced performance in verbal learning and memory and in semantic fluency, likely due to impaired common semantic clustering strategies and lexical access modalities,” De Meo said. “A third phenotype … showed mildly reduced cognitive performance in verbal learning and memory, executive/attention functioning and information processing speed. A fourth phenotype [was] characterized by decreased performance on all tests, with more severe involvement of attention and aspects of executive functions, such as cognitive interference inhibition. A fifth phenotype [was] characterized by severely reduced performance in all cognitive tests.”

The researchers found that patients with preserved cognition and mild–verbal memory/semantic fluency were younger (mean age, 36.5 years and 38.2 years, respectively) with a shorter disease duration (mean duration, 8 years and 8.3 years, respectively) compared with patients with mild–multidomain (mean age, 42.6 years; mean disease duration, 12.8 years; P < .001), severe–executive/attention (mean age, 42.9 years; mean disease duration, 12.2 years; P < .001) and severe–multidomain (mean age, 44 years; mean disease duration, 13.3 years; P < .001).

The fifth phenotype occurred more frequently in the later stages of MS, according to De Meo, which correlated with end-stage cognitive failure in the population.

MRI findings demonstrated that, compared with patients with preserved cognition, patients with mild–verbal memory/semantic fluency showed reduced mean hippocampal volume (5.42 mL vs. 5.13 mL; P = .04). Patients with mild–multidomain showed reduced mean cortical gray matter volume (687.69 mL vs. 662.59 mL; P = .02) and patients with severe–executive/attention had greater mean T2-hyperintense lesion volume (51.33 mL vs. 99.69 mL; P = .04). Patients with the severe–multidomain phenotype had widespread

brain damage, with reduced volume in all brain structures that were examined except for the nucleus pallidus, amygdala and caudate nucleus.

“[By] using MRI, we were able to identify distinct underlying neuroanatomical substrates, supporting data-driven cognitive findings with a biological basis,” De Meo said. She added, “Given that volume loss in a specific gray matter region reflects demyelination and loss of neurons, synaptic trees and supporting cells, the finding of reduced volume in a region with known functional relevance in a given phenotype can represent an important biological validation of the data-driven classification.”

The findings could have implications for the clinical management of MS and related decision-making, according to De Meo.

“This categorization of cognitive features could help plan rehabilitative strategies. Moreover, the patient’s transition to a more severe phenotype may provide support for clinical decisions about switches in pharmacological treatment,” she said. “The employment of these cognitive phenotypes can also represent a step forward in research, allowing a better selection of candidates for cognitive rehabilitation trials, as well as fostering future studies on the pathophysiology of cognitive changes in MS by using more advanced MRI techniques and deep learning approaches.”

Next steps in this research involve the development of an automated and unbiased classifier in which neuropsychologists provide patients’ scores for each different test and a classifier returns the appropriate phenotype. This would “facilitate the introduction of this classification into clinical practice,” De Meo noted.