Is functional MRI feasible for multi‐center studies on multiple sclerosis?

Functional magnetic resonance imaging (fMRI) is an indirect measure of blood flow and neuronal activity based on changes in the local magnetic field (T2*) [1]. fMRI is mainly used to study the neuronal mechanisms of central nervous system functioning and to define abnormal patterns of brain activations resulting from disease. When neurons are active, there is an increase in blood flow to the region, which increases the amount of oxygenated hemoglobin in the capillary beds [2]. The amount of oxygen delivered by the hemodynamic response to neuronal activity exceeds the amount required by the tissue, thus increasing the ratio of oxygenated to deoxygenated hemoglobin in the venous beds compared with the resting state [3]. At rest, deoxygenated hemoglobin causes a slight disturbance in the local magnetic field, which is attenuated by an increase in the presence of diamagnetic oxygenated hemoglobin during neuronal activity, thereby causing a longer T2* and increased signal intensity. The signal change is very small, but is reliably measured by subtracting images collected at rest from images collected during activity. Unlike the other conventional and non-conventional MRI measures used in patients with multiple sclerosis (MS), fMRI is not used clinically for making diagnoses or monitoring the disease [1]. Rather it has been used in research settings to examine the neural correlates of known motor, visual and neuropsychological deficits in patients with MS. The most commonly used paradigms assess motor ability, processing speed and working memory. Several studies have found that patients with MS show increased bilateral frontal activation during working memory tasks, as compared with healthy controls who show unilateral activation when completing the same task. It has been suggested that increased activation in patients with MS is a compensatory mechanism. Movement-associated cortical changes, characterized by an increased recruitment of the contralateral primary sensorimotor cortex (SMC) during the performance of simple tasks, and by the recruitment of additional sensorimotor areas during the performance of more complex tasks, have been demonstrated in patients with all MS phenotypes using different fMRI paradigms. Although these functional changes are often interpreted as reflective of adaptive mechanisms that could limit the clinical expression of tissue damage, studies have not yet related movement-associated activation patterns in MS to clinical measures of disease progression. On the other hand, a number of MS studies have reported correlations between specific patterns of fMRI activation and MRI measures of disease burden [4] and axonal injury [5]. Given the limited number of MS patients and controls recruited in most of these studies and the potential complexity of the correspondence between neurophysiological and clinical measures generally limited to a particular disease sub-phenotype, it may not be surprising that consistent relations between fMRI metrics and clinical measures have not been reported. In the current issue of the European Journal of Neurology, Wegner et al. [6] conducted a prospective multi-center study using functional MRI to better characterize the relations between clinical expression and brain function in patients with MS. The study confirms that activation of movement-related tasks for MS patients demands significantly greater cognitive resource allocation in a large network of cortical and subcortical motor-related areas than that observed in healthy controls. A more bilateral pattern of activation across the motor system confirms previous findings from single-center fMRI studies that employed motor tasks in MS. In addition, the present study emphasizes age-related differences in brain responses to the disease in MS patients and controls. Greater fMRI activation with increasing age was found in the ipsilateral precentral and inferior frontal gyri, as well as in the thalamus. In this regard, longitudinal studies, carefully age-matched and designed to evaluate the effects of disease-modifying treatments on cortical plasticity, are needed to better understand the role of functional brain adaptation in relation to aging differences and response to therapy. In the past, there has been no consensus regarding use of common patterns that are consistently applied across patients for functional activation associated with different motor tasks. Some recent studies have provided more insight in addressing this important question in larger cohorts of patients [7]. The functional activation pattern [ block design sequence (ABAB), with six periods of a 30 s visual cue for hand movement (A) alternated with six periods of 30 s rest in the dark (B)], applied in the study by Wegner et al. [6], represents the first systematic effort to standardize a paradigm and stimuli pattern in a multi-center study. The same standardized hand frame was used at all sites to restrict the maximum amplitude of the finger extension to 3 cm and all centers were supplied with a metronome that was equipped with a red flashing LED to pace the movements at a 1 Hz frequency. Finally, the study by Wegner et al. [6] represents one of the largest cohorts (56 MS patients and 60 agematched, healthy controls) studied with fMRI to date [8]. It is well known that the quality of fMRI data