Disrupted default mode network connectivity in migraine without aura

Migraine is a common and disabling primary headache disorder clinically characterized by episodic attacks of throbbing headache with specific features and associated symptoms [1]. A great body of studies has been conducted on patients with migraine, however, its pathophysiology is not completely understood. Indeed, currently, no integrative model has been formulated that accounts for all the factors that may play a role in migraine pathophysiology such as spreading depression, neurogenic inflammation, excitatory/inhibitory balance, genetic background and disturbed energy metabolism [2, 3]. More recently, converging evidence supports the role of a maladaptive stress response in migraine mechanisms [4, 5]. Resting-state fMRI (RS-fMRI) has allowed for the exploration of brain connectivity between functionally linked cortical regions, the so-called resting-state networks (RSNs) [6]. The most consistently reported RSN is the default mode network (DMN), which plays a relevant role in adaptive behavior other than in cognitive, emotional, and attention processes [7, 8]. In the last years, several RS-fMRI studies have identified functional connectivity changes in patients with migraine without aura (MwoA) [9]. To our knowledge, only a few of them have focused on DMN integrity in patients with migraine, reporting inconsistent results [10‐12]. For this reason, we investigated the DMN connectivity in patients with MwoA during the interictal period, by means of RS-fMRI using an independent component approach (ICA) to avoid any a priori hypothesis about the source of a possible functional disconnection [13]. In addition, we used Voxel Based Morphometry (VBM) to assess whether any between-group differences in resting-state functional connectivity were dependent on structural abnormalities, recently described in patients with migraine [14]. We hypothesized that DMN connectivity could be decreased in patients with MwoA in comparison to healthy controls (HC), supporting the current view that migraine may be related to a maladaptive stress response.

The present RS-fMRI study was designed to assess the functional integrity of DMN in patients with MwoA. Our findings demonstrate a reduced functional connectivity within the prefrontal and temporal cortices of the DMN in patients with MwoA during the interictal period. This altered functional connectivity was independent of structural abnormalities and not related to clinical or cognitive features of migraineurs. The DMN is a network highly relevant for cognitive processes and influences behavior in response to the environment in a predictive manner [7, 8]. In other terms, DMN represents a neural network related to individual stressful experiences and coping strategies to promote adaptation (i.e. allostasis) [19, 20]. This is done requiring most energy of brain baseline metabolic rate due to elevated levels of aerobic glycolysis required by the DMN [21]. Previous RS-fMRI studies investigating brain functional connectivity in patients suffering from different chronic pain conditions have already shown a dramatic alteration of DMN connectivity, suggesting that pain has a widespread impact on overall brain function, modifying brain dynamics beyond pain perception [22, 23]. Although several RS-fMRI studies, using different methodological approaches, have disclosed diffuse alterations in different brain areas and networks in patients with migraine [10‐12, 24, 25], only a few of them have specifically investigated DMN integrity [10‐12]. Xue and colleagues [10] have demonstrated an aberrant connectivity within the salience and executive networks in patients with MwoA; whereas DMN did not show any significant intra-network changes between patients with MwoA and HC. Nevertheless, an increased intrinsic DMN connectivity to brain regions outside the usual boundaries of this network (i.e. right insula) was reported. In another RS-fMRI study [11], the same group, using amplitude of low-frequency fluctuation and ROI-based functional connectivity analyses, has demonstrated a reduced DMN connectivity in left anterior cingulate cortex, bilateral prefrontal cortex and right thalamus. Furthermore, a significant decrease in regional homogeneity values has been observed in several brain areas involved in DMN in patients with MwoA [12]. DMN functional changes were negatively correlated only with disease duration [10‐12]. These conflicting data may be explained by the small sample size, patients clinical heterogeneity, and lack of consistent methodological approach. Furthermore, it is noteworthy that in those studies the cognitive profile of patients with migraine was not investigated, then behavioral correlates of the observed functional abnormalities are still unclear. In the present study, to specifically address this issue, we have performed a correlation analysis between clinical, cognitive and functional data, and we did not find any significant association. This is not surprising, considering our previous study [16] showing no correlation between executive network changes and neuropsychological data in patients with MwoA. Thus, taken together, our findings may suggest a possible alternative behavioral correlation of resting-state connectivity changes. One possibility is that the observed DMN dysfunction could underlie or be related to a maladaptive brain response to repeated stress [19, 20] which seems to characterize patients with migraine [4, 5]. Indeed, according to recent studies, recurrent migraine attacks alter both functional and structural brain connectivity [14], and these changes may disrupt mechanisms of stress response [4, 5]. When behavioral or physiological stressors are frequent or severe, allostatic responses can become maladaptive, leading, in a vicious cycle, to further allostatic load. Moreover, due to a high energetic demand, the observed DMN dysfunction may be associated with an impaired brain energy metabolism which has been demonstrated in previous MR spectroscopy studies in patients with migraine [26], likely due to an imbalance between ATP production and ATP use. In support of this notion, metabolic enhancers, such as riboflavin and coenzyme Q10 (both with a well-defined role in ATP generation), have shown effects in migraine prophylaxis [27]. In the present study, we identified two core regions of DMN [8], namely prefrontal and temporal areas, showing reduced functional connectivity in patients with MwoA. These areas have been demonstrated to be crucially involved in sensory-discriminative, cognitive and integrative pain functions within the so-called “neurolimbic pain network” [28]. In details, prefrontal cortex plays a specific role in mediating the attenuation of pain perception via cognitive control mechanisms [29, 30] whereas temporal cortex is involved in affective response to pain experience and its activation has been demonstrated both during pain experience [31] and migraine attacks [32]. Moreover, recent studies have reported both cortical abnormalities and microstructural changes of these regions in migraineurs [33‐35]. However, in the present study DMN connectivity disruption was detected in the absence of significant GM changes, possibly implying that functional changes may precede GM structural abnormalities. A few limitations of the current study should be considered. First, our methodological approach using ICA allows to evaluate functional interactions between brain areas but it does not provide information regarding causality and, consequently, it is still unclear whether functional changes are cause or consequence of repetitive migraine attacks. Second, we have studied a relatively small number of patients and further studies are needed to confirm our findings. Finally, the relationship between DMN functional changes and maladaptive brain response could be considered as a working hypothesis emerged from our work and future RS-fMRI studies are needed to further elucidate this potential correlation.

We believe that DMN connectivity changes may represent an early migraine biomarker, probably related to a maladaptive brain response. Future studies should examine other cortical resting-state networks and longitudinal studies are needed to evaluate the possibility that this modern neuroimaging approach can lead to the identification of different categories of patients or different timing of selective networks involvement.