Biomarkers for Chronic Pain and Analgesia. Part 2: How, Where, and What to Look for Using Functional Imaging
Abstract: Rapid advances in brain imaging chronic pain patients have yielded exciting data sets that could provide the basis for the development of chronic pain biomarkers that could increase the probability of success in analgesic drug development, aid clinicians in understanding, tracking, and treating disease, and link patients to the most effective therapies for their pain conditions. This review explores the potential of brain imaging techniques to detect functional, morphometric, and chemical changes that could serve as biomarkers for disease state and therapeutic efficacy. An important area for future research is to image clinical ongoing pain to further our knowledge of brain function in different pain states and the effects of treatment.
Chronic pain remains a significant public health problem. It has been estimated that currently prescribed analgesic medications are successful in treating chronic pain only 30-40% of the time (Woolf, 2010). Over the past 50 years, there has been a lack of a significant breakthrough in medications for chronic pain (Kissin, 2010). As a consequence, chronic pain affects quality of life and accounts for considerable direct and indirect healthcare costs associated with impaired daily functioning and loss of workplace productivity for many patients. The identification and development of biomarkers that could serve alongside patient reporting as objective quantitative measures of pain related processes and the impact of analgesics on pain would represent a major advance in the field. At present, as we have discussed in Part 1 of this review, pain assessment is based almost exclusively on patient self reports and visual analog rating scales that depend upon the relationship between biological nociceptive processes including higher cognitive and emotional functions and patients’ verbal and/or written descriptions on their pain. Indeed this was highlighted in the 2009 RFA from NIH that stated “Pain research has been greatly hampered by the unreliable nature of self-report based instruments. The establishment of objective affordable and reliable pain biomarkers and measurements would advance our understanding of pain mechanisms, provide a basis for improved clinical management of pain, help assess an individual’s risk for becoming addicted to opiates analgesics and establish much needed objective measures of treatment success or failure.” In the following sections we explore the use of brain imaging techniques including functional, morphometric (gray matter volume and white matter integrity), and chemical measures to detect changes in the brain of chronic pain patients and the effects of analgesics in order to evaluate their potential for use as pain disease state and therapeutic biomarkers (Figure 1).
How to Look — Brain Imaging Techniques to Detect Biomarkers of Chronic Pain (Disease State)
Functional imaging (BOLD and ASL)
fMRI measures of brain function, namely neuronal activity (Logothetis, 2002; Viswanathan and Freeman, 2007), may be obtained using Blood Oxygen Level Dependent (BOLD) measures (for a review see Brown et al., 2007), or change in cerebral blood volume (CBV) using arterial spin labeling (ASL) (for a review see Hernandez-Garcia, 2004), or a combination of these techniques. In both methods, changes in blood perfusion mirror neural system activation: BOLD imaging detects changes in the concentration ratio of oxy- and deoxy-hemoglobin as a result of changes in blood flow and volume (Ogawa et al., 1990; Bellevieau et al., 1991). BOLD signals can be both positive and negative in nature and it is thought that this directionality reflects the net consequence of neuronal activation and inhibition, respectively. BOLD imaging cannot therefore distinguish between the neural activity of excitatory versus inhibitory neurons such that it cannot delineate pathophysiological changes in particular neurotransmitter systems. Another confound is that blood flow and blood volume changes produce opposite effects in the net BOLD signal and potentially affect the relationship between neuronal firing and BOLD response. Nonetheless, BOLD imaging can be used to produce signatures of brain activity that reflect disease state and drug effects much in the same way expression profiling can be used to interrogate molecular changes in health and disease. In contrast to BOLD, ASL measures only blood flow changes through magnetic spin labeling or tagging of the iron in red blood cells in the proximal blood flow which for the brain is the carotid artery (Golay and Petersen, 2006; Detre et al., 2009). A control and spin labeled image is subtracted to produce a perfusion imaging map. The main advantage of ASL over BOLD is its ability to quantify blood flow but it has lower spatial resolution.
Anatomical circuit based approaches: pain circuit activity as biomarker
Compared with other brain diseases, the CNS circuits underlying pain are relatively well defined. For example, nociceptive pain activates a 3-neuron sensory circuit including the dorsal root ganglia, dorsal horn, thalamus, and cortex. Activation within this circuit can be quantitatively measured using fMRI (Borsook et al., 2004) in both acute (Borsook et al., 2003; DaSilva et al., 2002; Moulton et al., 2008) and chronic pain (Becerra et al., 2006b). The differences in activity detected across this circuit when a pain-free state is compared with acute or chronic pain provide a pain index that could be used as a biomarker for pain state.
Functional connectivity analysis: network activity as biomarker
Evaluating brain conditions in terms of networks seems more rational than evaluating individual brain regions for biomarker potential, given the complexity of the brain’s structural and functional connections that underpin behavior. Certainly in the context of chronic pain where multiple behaviors reflect alterations across numerous brain regions, it is appealing to think that a network-based approach, although potentially more complex to execute, would yield greater predictability. Functional connectivity analyses may therefore provide the best avenue to discover biomarkers of the brain in health and chronic pain (Bassett and Bullmore, 2009; Guye et al., 2008; Reijneveld et al., 2007).
Functional connectivity analysis is a tool that quantifies the relative synchrony or brain functional fluctuations among brain regions. Functional connectivity analysis yields information about changes in the strength of network connections between brain structures due to, for example, alteration of the connection or changes in their oscillatory characteristics. Consequently, functional connectivity analyses can assess whether disease or drug modulation alters the functional response relationship between brain regions. Data reported in non-pain models of connectivity using fMRI have indicated both reproducible and sensitive results for both disease state and pharmacological treatment (Rowe et al., 2010). Connectivity changes have been reported in diabetic neuropathic pain, where thalamocortical activity is altered in patients vs. controls (Cauda et al., 2009). Administration of naloxone has been shown to induce changes in the thalamo-temporal cortex connection (Patel et al., 2008). Functional connectivity gives the opportunity to interpret behaviors associated with drugs, pain, or pain testing of healthy volunteers and patients through up or down modulation of the strength of activity in defined neural circuits associated with the behavior. As a result, functional connectivity potentially can be used as a biomarker for drugs, pain, and pain related symptoms.
Default mode network (DMN) activity as biomarker
DMNs are a set of regional activation networks that reflect resting brain functions related to orientation and interpretation of environment, monitoring and reporting state of the self, among others (Gusnard and Raichle, 2001). Changes in these networks have been used to differentiate brain states, e.g., Alzheimer’s disease (Hou and Xu, 2010) and the effect of anesthetics (Martuzzi et al., 2010).
The investigational use of DMN/functional connectivity approaches to the discovery of brain biomarkers for chronic pain has a number of advantages: it is (1) well-founded theoretically; (2) highly reproducible (Deuker et al., 2009); (3) practical in that measurements are easily performed in a standardized fashion (both acquisition and analysis); (4) longitudinal such that multiple evaluations can be performed over time in single individuals for evaluation of disease state that may be altered by drug effects or disease progression; and (5) modular, thus giving a high possibility of usefulness to predict an individual’s response to treatment.
Volumetric biomarkers for chronic pain
The utility of MRI measures of anatomical variation in disease states is currently being evaluated for Alzheimer’s disease through the Biomarker Consortium Alzheimer’s Disease Neuroimaging Initiative (ADNI, www.adni-info.org), a joint program of NIH and the pharmaceutical industry, which aims to assess biomarkers for disease state and drug effects. This initiative will yield some generalizable operational observations about the reliability and reproducibility of MRI for determining longitudinal volumetric brain biomarkers. Chronic pain clearly changes brain morphometry (for a review see May, 2008), making it a promising candidate biomarker measure.
Chemical biomarkers for chronic pain
Several CNS neurotransmitters (e.g., glutamate, GABA, glycine) and brain metabolites (e.g., NAA, choline) that reflect excitatory and inhibitory neuronal activity and neuronal health status can be measured in vivo using Magnetic Resonance Spectroscopy (MRS) (Prost, 2008); its use in pharmaceutical research has been reviewed recently (Rodriguez et al., 2008). MRS has the potential to measure regionally specific chemical changes in the brain during chronic pain that reflect neuronal function and neuronal status to reveal: (1) abnormalities in disease compared with healthy controls; (2) measures of changes with therapy; and (3) potential predictive brain-based chemical markers of therapeutic outcome (Rohayem et al., 2008). MRS may be multiplexed to both functional and morphological studies of brain state to provide additional power to observations.
What to Look for and Where to Look
Functional imaging biomarkers — individual brain regions
While one structure obviously cannot fully represent the complex circuitry changes that underlie chronic pain, there are several structures that exhibit reproducible and specific changes that can be detected by neuroimaging during chronic pain. These include the somatosensory cortex, frontal lobe, nucleus accumbens, cingulate cortex, insula, and periaqueductal gray, and we consider, below, the changes in each of these regions for their potential as biomarkers for chronic pain. Other regions including the cerebellum (Borsook et al., 2008) and regions of the basal ganglia (Borsook et al., 2010) may also be involved in chronic pain, but their role is currently poorly understood, and so they are not reviewed here.
Somatosensory Cortex: Studies have shown that primary (SI) and secondary (SII) somatosensory cortices are activated in both healthy controls and chronic pain patients (Apkarian et al., 2005). SII appears to be more consistently activated than SI (Coghill et al., 1999; Peyron et al., 2000). SII has been observed to encode pain intensity (Kong et al., 2006), but the activation response seems to show a ceiling effect (Frot et al., 2007). This is a limitation, since drug effects might in some cases be undetected, making the SII’s use as a pain biomarker questionable. SI activation has been correlated with the subjective pain rating of experimental noxious stimuli (Becerra et al., 2001; Chen et al., 2002; Timmermann et al., 2001), suggesting that it might serve as a biomarker for pain intensity. Indeed, drugs known to affect sensory intensity processing, particularly opioids, show decreased signal intensity in SI (Becerra et al., 2006a; Leppa et al., 2006; Oertel et al., 2008). SI has, however, not been shown to be activated in all pain studies (Apkarian et al., 2005), perhaps due to differences in the experimental paradigms or data analysis methods used (Bushnell et al., 1999). More recent studies suggest that SI activation is best characterized by a more complex response model (Upadhyay et al., 2010) than is currently used in the field. The complex model was used in a study utilizing Near Infra-Red Spectroscopy (NIRS) to define pain intensity encoding in sensorimotor cortex that seems to clearly differentiate pain from no pain in the SI cortex (Becerra et al., 2009a).
The brain of chronic pain patients responds differently to stimuli than the normal brain. Results from amputees, chronic back pain patients, and patients with complex regional pain syndrome (CRPS) all show altered somatosensory cortical activation with anatomically “displaced” activation profiles dependent upon pain state. Some studies have reported in patients with chronic pain, brain activation in an abnormal location within the somatosensory cortex that shifts to the expected localization when pain is symptomatically abolished (Flor et al., 2006). Moreover, a pilot study of patients with and without phantom limb pain found that an abnormal misrepresentation in the somatosensory cortex was present only in patients with phantom pain but not in patients who had similar amputations and were pain free (Flor et al., 1995). Subsequent studies have confirmed consistent cortical reorganization in phantom limb pain and reversal of this reorganization upon its resolution (Birbaumer et al., 1997; Flor and Birbaumer, 2000). Similar functional cortical re-organization has been described in chronic low back pain (Flor et al., 1997). Mis-localization of responses to mechanical stimuli occurs in somatosensory systems of CRPS patients (Maihofner et al., 2006) and is reversed with clinical improvement lending further support to somatotopic activation mapping in the primary somatosensory cortex as a marker for clinical recovery (Maihofner et al., 2004). Functional imaging could be reasonably easily deployed to measure such changes in individual patients utilizing simple stimuli, such as brush or von Frey, to provide stimulation to the location affected by the pain compared with another, standard location. If displacement proves sufficiently reproducible, the shift from an abnormal to a normal location of activation could be used as a marker of treatment efficacy. Currently, however, there is no data to indicate whether commonly used analgesics for chronic pain modulate or reverse this cortical reorganization. The sensitivity of the method also remains to be defined, and it is not known whether functional correction or reorganization predates clinical improvement following standard therapeutic regimens (Maihofner et al., 2007). For example, in a study on fMRI changes in patients with pediatric CRPS, symptomatic recovery preceded the resolution of abnormalities in the cortical response to evoked stimuli, suggesting that the normalization of the brain’s response to stimulus lags behind symptomatic recovery (Lebel et al., 2008). An additional avenue to explore is whether combining the power of cortical activation imaging with structural imaging of cortical anatomy, to generate dynamic maps of plasticity in the somatosensory cortex, can improve the sensitivity and specificity of the technique.
Nucleus Accumbens (NAc) Activation: Pain and relief are clearly at opposite ends of the pain-or-analgesic spectrum, but also at opposite ends of the reward-aversion continuum. The NAc has been shown to exhibit opposite BOLD signals in response to reward and aversion (see Becerra and Borsook, 2008). Functional imaging of the NAc has demonstrated opposite BOLD activations to aversive (pain onset) vs. rewarding (pain offset) stimuli (Becerra and Borsook, 2008). A negative signal change is observed in the NAc contralateral to the stimulus with pain onset and a positive signal change with pain offset. Specific measures of NAc activity may provide a useful marker for the effects of pain and analgesia in healthy volunteers and perhaps in patients with chronic pain since its response to heat stimuli predicts differences between acute and chronic pain with 100% efficiency (Baliki et al., 2009). BOLD responses in the NAc may differentiate acute from chronic pain and are sensitive to opioid agonist and antagonist drugs (Becerra et al., 2006a; Borras et al., 2004), with what seems to be high sensitivity (Baliki et al., 2010), making it well positioned for further validation studies.
Prefrontal Regions: One of the major contributions of fMRI to the pain field has been the demonstration that prefrontal brain regions are involved in pain processing. Imaging data suggest that significant changes are present in frontal regions including the orbitofrontal gyrus (Gob), the dorsolateral prefrontal cortex (DLPC), and the medial prefrontal cortex (mPFC) during acute pain (Baliki et al., 2006; Becerra et al., 2001) and in multiple chronic pain conditions including neuropathic pain (Becerra et al., 2006b) and complex regional pain syndromes (CRPS) (Geha et al., 2008; Lebel et al., 2008; Maihofner and Handwerker, 2005). The activity in regions such as the medial prefrontal cortex is increased in chronic pain and attenuated by analgesics, and may provide a specific readout (Seifert et al., 2009). Although not as clear as the case for the NAc as discussed above, the role of these regions in pain and placebo (Craggs et al., 2007) and their modulation by analgesic drugs, including non-steroidal anti-inflammatory drugs (NSAID) (Lorenz et al., 2008), antidepressants (Morgan et al., 2005), and opioids (Becerra et al., 2006a), are notable and merit further evaluation of their usefulness as potential biomarkers for chronic pain and effective analgesic action.
Anterior (ACC) and Posterior (PCC) Cingulate Cortex: The cingulate cortex is involved in complex behaviors, including affective, cognitive, and autonomic functions (Devinsky et al., 1995). If data on lesions of the anterior cingulate cortex and chronic pain are correct (no controlled studies have been performed), it would seem that measures of specific changes in the ACC might well provide a functional marker for chronic pain. A complicating factor, however, is that specific regions of the ACC may be activated in diverse sensory, motor, and cognitive processes, making its specificity as a pain biomarker questionable unless studies are carefully designed to control for other aspects of ACC function, such as attention. In most studies of acute pain and in placebo effects, the ACC is activated (Apkarian et al., 2005). Interestingly, naloxone diminishes placebo effects and pain modulation in the ACC (Eippert et al., 2009). We are unaware of studies on the effect of analgesics on ACC activation in chronic pain, and this may be an area for future investigations.
Insula: Functional imaging studies in chronic pain have shown activation of the insula in pain processing where it is thought to be involved in the integration of the internal emotional cognitive and affective state (Sawamoto et al., 2000; Wiech and Tracey, 2009) with afferent sensory (posterior insula) function (Geha et al., 2008). Studies in two individuals with functional ablative lesions of the insula showed that they could evaluate pain and reported substantially higher pain intensity VAS ratings of acute experimental noxious stimuli than age-matched control subjects. Functional imaging detected elevation of activity of the primary somatosensory cortex ipsilateral to the lesioned insula. These findings suggest a role for the insula in modulating cognitive cortical functions involved in the appraisal of pain. The insula has also been shown to be involved in interoception (Craig, 2009), decision-making (Preuschoff et al., 2008), addiction (Goldstein et al., 2009), and autonomic responses to external stimuli (Ruggiero et al., 1987). Activity in the insula can be modified by the direct action of opioid analgesics (Leppa et al., 2006), and insula activation may be modulated by the opioid remifentanil during analgesic treatment of acute pain (Wise et al., 2002).
Periaqueductal Gray (PAG): Midbrain regions, including the periaqueductal gray and nucleus cuneiformis (Bandler and Shipley, 1994; Behbehani, 1995; Bouhassira et al., 1990), are known important output modulators of pain facilitation or inhibition. Among other things, these structures may play a role in central sensitization in humans when they do not inhibit afferent inputs through descending inhibition. Functional imaging has been used to show that processing of afferent and efferent signals (Becerra et al., 2001) by the PAG and specific drug effects exemplified by direct inhibitory changes (i.e., decreased BOLD signal) to opioids, including morphine (Becerra et al., 2006a) and naloxone modulation (Borras et al., 2004; Eippert et al., 2009). It is worth noting that opposite effects in PAG are observed for the µ agonist morphine and µ antagonist naloxone. A number of recent fMRI studies have shown that abnormalities of midbrain function occur in chronic pain conditions, including visceral pain, migraine, fibromyalgia, and complex regional pain syndrome (Berman et al., 2008; Moulton et al., 2008; Seifert et al., 2009; Staud, 2009).
Amygdala: Over the last 10 years, the amygdala has emerged as an important region in pain processing in humans probably because of the central role of emotion in pain perception. A reciprocal relationship exists between persistent pain and negative affective states such as fear, anxiety, and depression (Neugebauer et al., 2004). The central nucleus of the amygdala is involved in integrating nociceptive information with poly-modal information about the internal and external environment and so is involved in mediating the close association between anxiety and pain phenotypes. Decreasing activation to pain has been reported in the amygdala with a number of analgesics (Oertel et al., 2008). In contrast, opioids such as remifentanil appear to produce activation in this region (Leppa et al., 2006). Cannabinoids, which have analgesic action (Guindon and Hohmann, 2009), reduce amygdala activity to social signals, thereby suggesting an anxiolytic role of the drug in this context (Phan et al., 2008). Amygdala activation might be utilized as a cofactor to account for fear and anxiety (and potential modulation by drugs) and its influence in pain processing.
Hippocampal Formation: The hippocampal formation is involved in pain information processing (Liu and Chen, 2009) and is considered to be involved in understanding the context of the pain, including learning and memory in pain pathways (Apkarian, 2008; Sandkuhler, 2000). Imaging studies reveal activation in the hippocampus in acute (Ploghaus et al., 2001; Ziv et al., 2010) and chronic pain (Becerra et al., 2009b), thus it has potential as a functional biomarker as it has been observed to be activated in most chronic pain studies, and activation may change with therapy. This change may relate to alteration in pain memory. Since chronic stress (including pain) may modulate regions such as the hippocampus (McEwen, 2001), imaging studies may provide information on changes when pain/stress is relieved. For example, significant attenuation of hippocampal activation by thermal stimuli is observed following treatment of CRPS patients previously treated with ketamine (Becerra et al., 2009b).
Default mode networks and pain states
A number of recent studies have evaluated changes in default mode networks (DMN) in chronic back pain and in fibromyalgia. Chronic back pain results in reduced deactivation in key regions within the default mode network during a visual attention task (although patients and controls performed the task equally well) suggesting altered functioning is widespread in the brains of these individuals (Baliki et al., 2008). In non-task related evaluation of DMN in chronic back pain, abnormal correlations in back pain patients vs. healthy controls reported differences in regions that included bilateral insular cortex and regions in the middle frontal gyrus (Tagliazucchi et al., 2010). In a study of fibromyalgia patients, increased connectivity in DMN was observed that included the executive attention network (EAN) (Napadow et al., 2010). Such studies suggest that whole brain network approaches of brain function may be a way of defining a pain state and thus a basis for evaluating therapeutic interventions.
Volumetric change in chronic pain
Volumetric changes have been demonstrated in chronic back pain (Apkarian et al., 2004), trigeminal neuropathy (DaSilva et al., 2008), fibromyalgia (Kuchinad et al., 2007), migraine (Valfre et al., 2008), and several other pain disorders, as defined by DSM IV criteria (Valet et al., 2009). Volumetric changes in gray matter that are reversible with the control of pain in chronic pain patients are potential biomarkers of chronic pain and of recovery (see Rodriguez-Raecke et al., 2009). In adults with severe pain, hippocampal volumes are smaller and the decrease correlates with lower levels of N-acetylaspartate/creatine, which is associated with neuronal integrity (Zimmerman et al., 2009). Normal baseline ranges for candidate brain regions still, however, need to be established for comparison of healthy brain to disease and treatment states to further examine the utility of regional volumetric analyses to act as pain state biomarkers.
Chemical changes in chronic pain
To date, magnetic resonance spectroscopy (MRS) has been relatively underutilized in studying chronic pain and reports have been limited to chronic back pain (Fleisher et al., 2009), fibromyalgia (Harris et al., 2008), and migraine (Prescot et al., 2009). Recently, measurements of functional activity in the insula have, however, been linked to MRS studies of brain metabolites (Prescot et al., 2009), since the region is accessible to measurement and well defined anatomically.
Functional Imaging Techniques to Detect Analgesics (Drug Effect)
Analgesic drug targeting
There are well-documented expensive high failure rates for novel mechanism drugs such as analgesics that target CNS systems (DiMasi et al., 2003; Rawlins, 2004). More commonly in the field of chronic pain treatment are reverse translation, astute clinical observation, and repurposing. For example, gabapentinoids were found to have potential therapeutic utility for the treatment of chronic pain without conducting large clinical trials. In contrast, the NK-1 antagonist class failed in several clinical trials for chronic pain despite activity in preclinical assays (Duffy, 2004; Herbert and Holzer, 2002; Hill, 2000). A predictive biomarker for chronic pain might have allowed earlier detection of its likely lack of therapeutic efficacy (Chizh et al., 2009). These examples raise the question as to whether the use of neuroimaging biomarkers in small clinical studies in chronic pain patients can help focus development on the most promising drug candidates and mechanisms by providing robust early go/no go data for decision-making. Potential use of imaging in the clinical phase of drug development is shown in Figure 2.
A number of articles have reviewed the use of imaging to support analgesic drug development, but have not focused on the identification of predictive fMRI biomarkers (for analgesic efficacy see (Borsook et al., 2002; Borsook et al., 2006; Borsook and Becerra, 2006; Lawrence and Mackey, 2008; Schweinhardt et al., 2006; Borsook and Becerra, 2010). Some studies have used a whole brain activation approach to evaluate the effect of different analgesics on a thermal stressor response, and they have shown that this approach can differentiate between different drug classes (Borsook et al., 2007b). Few studies have used imaging to formally evaluate the effects of analgesic drugs in chronic pain (Scrivani et al., 2010), but techniques that catalogue a profile (”Brain Array”) of drug effects on the brain (Borsook et al., 2006) have promise for detection of unique signatures for effective and ineffective therapeutics within or between drug classes, in specific chronic pain disease states. Obvious approaches include evaluation of BOLD responses to drugs (Borsook et al., 2006), arterial spin labeling (ASL) (Detre et al., 2009), and effects of drugs on default mode networks. To date, few DMN/connectivity analyses have been reported using analgesics. Nevertheless, the approach appears to warrant investigation, especially for the determination of the overall and specific brain connectivity patterns in response or lack of response to therapeutics in chronic pain patients.
Cautions and Current Limitations on Brain Imaging Biomarkers of Chronic Pain
Despite considerable advances in the understanding of the peripheral pain-afferent pathway, the cortical and affective circuitry involved in pain processing and its regulation that influences the overall pain perception of a patient remains relatively ill defined compared to diseases such as Parkinson’s disease. Indeed many brain regions have alternative functions to pain processing — for example empathy can activate many of these areas without pain (absence of activity does assure there is no pain) — and understanding their influence on a patient’s subjective assessment of their pain sensation is still in its infancy with regard to being qualified and understood. Finally a note of caution, many of the BOLD imaging studies we have described compare an experimental pain model in patients with chronic pain versus controls and not activation in chronic pain vs. healthy controls. It is important to realize that in these studies the chronic pain patient’s brain was not constantly over-active in these areas, nor was there a proof of a correlation between this activity and the patient’s clinical pain. An important area for future research is therefore to image clinical ongoing pain in a longitudinal fashion in order to further our knowledge of brain function in different pain states and the effects of treatment.
Biomarkers are the essential elements of predictive, preventative, and personalized medicine. The prospects of developing CNS biomarkers for chronic pain may open new avenues for improved efficiency in the development of novel therapies for chronic pain by focusing on well characterized patient populations in smaller clinical trials, making decision-making less costly and matching treatments to the individual needs of patients. For clinicians, functional imaging biomarkers of chronic pain provide the opportunity to improve our understanding of pain processes and evaluate disease mechanisms and treatment responses particularly in patients whose self-reporting is unreliable. For chronic pain patients the benefit is obvious — the promise that functional imaging biomarkers can advance clinical care by enabling the discovery of novel therapeutic approaches with increased effectiveness that can improve their quality of life.
Supported by NINDS K24 (NS064050) to D.B. and the L. Herlands Fund to the P.A.I.N. group (D.B., L.B.).
The authors D.B. and L.B. do not have any financial disclosures or conflicts of interest to report. R.H. is a full-time employee of Merck and Co.
David Borsook, M.D., Ph.D., Director, P.A.I.N. Group, Center for Pain and the Brain, Athinoula A. Martinos Center for Biomedical Engineering, Massachusetts General Hospital, Charlestown, Massachusetts 02114, USA.
Apkarian AV. Pain perception in relation to emotional learning. Curr Opin Neurobiol 18(4):464-468, 2008.
Apkarian AV, Bushnell MC, Treede RD, Zubieta JK. Human brain mechanisms of pain perception and regulation in health and disease. Eur J Pain 9(4):463-484, 2005.
Apkarian AV, Sosa Y, Sonty S, Levy RM, Harden RN, Parrish TB, Gitelman DR. Chronic back pain is associated with decreased prefrontal and thalamic gray matter density. J Neurosci 24(46):10410-10415, 2004.
Baliki MN, Chialvo DR, Geha PY, Levy RM, Harden RN, Parrish TB, Apkarian AV. Chronic pain and the emotional brain: specific brain activity associated with spontaneous fluctuations of intensity of chronic back pain. J Neurosci 26(47):12165-12173, 2006.
Baliki MN, Geha PY, Apkarian AV. Parsing pain perception between nociceptive representation and magnitude estimation. J Neurophysiol 101(2):875-887, 2009.
Baliki MN, Geha PY, Apkarian AV, Chialvo DR. Beyond feeling: chronic pain hurts the brain, disrupting the default-mode network dynamics. J Neurosci 28(6):1398-1403, 2008.
Baliki MN, Geha PY, Fields HL, Apkarian AV. Predicting value of pain and analgesia: nucleus accumbens response to noxious stimuli changes in the presence of chronic pain. Neuron 66:149, 2010.
Bandler R, Shipley MT. Columnar organization in the midbrain periaqueductal gray: modules for emotional expression? Trends Neurosci 17(9):379-389, 1994.
Bassett DS, Bullmore ET. Human brain networks in health and disease. Curr Opin Neurol 22(4):340-347, 2009.
Becerra L, Borsook D. Signal valence in the nucleus accumbens to pain onset and offset. Eur J Pain 12(7):866-869, 2008.
Becerra L, Breiter HC, Wise R, Gonzalez RG, Borsook D. Reward circuitry activation by noxious thermal stimuli. Neuron 32(5):927-946, 2001.
Becerra L, Harris W, Grant M, George E, Boas D, Borsook D. Diffuse optical tomography activation in the somatosensory cortex: specific activation by painful vs. non-painful thermal stimuli. PLoS One 4(11):e8016, 2009a.
Becerra L, Harter K, Gonzalez RG, Borsook D. Functional magnetic resonance imaging measures of the effects of morphine on central nervous system circuitry in opioid-naive healthy volunteers. Anesth Analg 103(1):208-216, table of contents, 2006a.
Becerra L, Morris S, Bazes S, Gostic R, Sherman S, Gostic J, Pendse G, Moulton E, Scrivani S, Keith D, Chizh B, Borsook D. Trigeminal neuropathic pain alters responses in CNS circuits to mechanical (brush) and thermal (cold and heat) stimuli. J Neurosci 26(42):10646-10657, 2006b.
Becerra L, Schwartzman RJ, Kiefer RT, Rohr P, Moulton EA, Wallin D, Pendse G, Morris S, Borsook D. CNS Measures of pain responses pre- and post-anesthetic ketamine in a patient with complex regional pain syndrome. Pain Med, epub ahead of print, Feb. 25, 2009b.
Behbehani MM. Functional characteristics of the midbrain periaqueductal gray. Prog Neurobiol 46(6):575-605, 1995.
Belliveau JW, Kennedy DN Jr, McKinstry RC, Buchbinder BR, Weisskoff RM, Cohen MS, Vevea JM, Brady TJ, Rosen BR. Functional mapping of the human visual cortex by magnetic resonance imaging. Science 254(5032):716-719, 1991
Berman SM, Naliboff BD, Suyenobu B, Labus JS, Stains J, Ohning G, Kilpatrick L, Bueller JA, Ruby K, Jarcho J, Mayer EA. Reduced brainstem inhibition during anticipated pelvic visceral pain correlates with enhanced brain response to the visceral stimulus in women with irritable bowel syndrome. J Neurosci 28(2):349-359, 2008.
Birbaumer N, Lutzenberger W, Montoya P, Larbig W, Unertl K, Topfner S, Grodd W, Taub E, Flor H. Effects of regional anesthesia on phantom limb pain are mirrored in changes in cortical reorganization. J Neurosci 17(14):5503-5508, 1997.
Borras MC, Becerra L, Ploghaus A, Gostic JM, Dasilva A, Gonzalez RG, Borsook D. fMRI measurement of CNS responses to naloxone infusion and subsequent mild noxious thermal stimuli in healthy volunteers. J Neurophysiol 91(6):2723-2733, 2004.
Borsook D, Becerra L. Using NMR approaches to drive the search for new CNS therapeutics. Curr Opin Investig Drugs 11(7):771-778, 2010.
Borsook D, Upadhyay J, Chudler EH, Becerra L. A key role of the basal ganglia in pain and analgesia — insights gained through human functional imaging. Mol Pain 6:27, 2010.
Borsook D, Becerra L, Hargreaves R. A role for fMRI in optimizing CNS drug development. Nat Rev Drug Discov 5(5):411-424, 2006.
Borsook D, Becerra LR. Breaking down the barriers: fMRI applications in pain, analgesia and analgesics. Mol Pain 2:30, 2006.
Borsook D, Burstein R, Becerra L. Functional imaging of the human trigeminal system: opportunities for new insights into pain processing in health and disease. J Neurobiol 61(1):107-125, 2004.
Borsook D, Dasilva AF, Ploghaus A, Becerra L. Specific and somatotopic functional magnetic resonance imaging activation in the trigeminal ganglion by brush and noxious heat. J Neurosci 23(21):7897-7903, 2003.
Borsook D, Moulton EA, Tully S, Schmahmann JD, Becerra L. Human cerebellar responses to brush and heat stimuli in healthy and neuropathic pain subjects. Cerebellum 7(3):252-272, 2008.
Borsook D, Pendse G, Aiello-Lammens M, Glicksman M, Gostic J, Sherman S, Korn J, Shaw M, Stewart K, Gostic R, Bazes S, Hargreaves R, Becerra L. CNS response to a thermal stressor in human volunteers and awake rats may predict clinical utility of analgesics. Drug Develop Res 68(1):23-41, 2007.
Borsook D, Ploghaus A, Becerra L. Utilizing brain imaging for analgesic drug development. Curr Opin Investig Drugs 3(9):1342-1347, 2002.
Borsook D, Upadhyay J, Chudler EH, Becerra L. A key role of the basal ganglia in pain and analgesia — insights gained through human functional imaging. Mol Pain 6:27, 2010.
Bouhassira D, Bing Z, Le Bars D. Studies of the brain structures involved in diffuse noxious inhibitory controls: the mesencephalon. J Neurophysiol 64(6):1712-1723, 1990.
Brown GG, Perthen JE, Liu TT, Buxton RB. A primer on functional magnetic resonance imaging. Neuropsychol Rev 17(2):107-125, 2007.
Bushnell MC, Duncan GH, Hofbauer RK, Ha B, Chen JI, Carrier B. Pain perception: is there a role for primary somatosensory cortex? Proc Natl Acad Sci U S A 96(14):7705-7709, 1999.
Cauda F, Sacco K, D’Agata F, Duca S, Cocito D, Geminiani G, Migliorati F, Isoardo G. Low-frequency BOLD fluctuations demonstrate altered thalamocortical connectivity in diabetic neuropathic pain. BMC Neurosci 10:138, 2009.
Chen JI, Ha B, Bushnell MC, Pike B, Duncan GH. Differentiating noxious- and innocuous-related activation of human somatosensory cortices using temporal analysis of fMRI. J Neurophysiol 88(1):464-474, 2002.
Chizh BA, Priestley T, Rowbotham M, Schaffler K. Predicting therapeutic efficacy - experimental pain in human subjects. Brain Res Rev 60(1):243-254, 2009.
Coghill RC, Sang CN, Maisog JM, Iadarola MJ. Pain intensity processing within the human brain: a bilateral, distributed mechanism. J Neurophysiol 82(4):1934-1943, 1999.
Craggs JG, Price DD, Verne GN, Perlstein WM, Robinson MM. Functional brain interactions that serve cognitive-affective processing during pain and placebo analgesia. Neuroimage 38(4):720-729, 2007.
Craig AD. How do you feel — now? The anterior insula and human awareness. Nat Rev Neurosci 10(1):59-70, 2009.
Dasilva AF, Becerra L, Makris N, Strassman AM, Gonzalez RG, Geatrakis N, Borsook D. Somatotopic activation in the human trigeminal pain pathway. J Neurosci 22(18):8183-8192, 2002.
Dasilva AF, Becerra L, Pendse G, Chizh B, Tully S, Borsook D. Colocalized structural and functional changes in the cortex of patients with trigeminal neuropathic pain. PLoS One 3(10):e3396, 2008.
Detre JA, Wang J, Wang Z, Rao H. Arterial spin-labeled perfusion MRI in basic and clinical neuroscience. Curr Opin Neurol 22(4):348-355, 2009.
Deuker L, Bullmore ET, Smith M, Christensen S, Nathan PJ, Rockstroh B, Bassett DS. Reproducibility of graph metrics of human brain functional networks. Neuroimage 47(4):1460-1468, 2009.
Devinsky O, Morrell MJ, Vogt BA. Contributions of anterior cingulate cortex to behaviour. Brain 118( Pt 1):279-306, 1995.
Dimasi JA, Hansen RW, Grabowski HG. The price of innovation: new estimates of drug development costs. J Health Econ 22(2):151-185, 2003.
Duffy RA. Potential therapeutic targets for neurokinin-1 receptor antagonists. Expert Opin Emerg Drugs 9(1):9-21, 2004.
Eippert F, Bingel U, Schoell ED, Yacubian J, Klinger R, Lorenz J, Buchel C. Activation of the opioidergic descending pain control system underlies placebo analgesia. Neuron 63(4):533-543, 2009.
Fleisher AS, Donohue M, Chen K, Brewer JB, Aisen PS. Applications of neuroimaging to disease-modification trials in Alzheimer’s disease. Behav Neurol 21(1):129-136, 2009.
Flor H, Birbaumer N. Phantom limb pain: cortical plasticity and novel therapeutic approaches. Curr Opin Anaesthesiol 13(5):561-564, 2000.
Flor H, Braun C, Elbert T, Birbaumer N. Extensive reorganization of primary somatosensory cortex in chronic back pain patients. Neurosci Lett 224(1):5-8, 1997.
Flor H, Elbert T, Knecht S, Wienbruch C, Pantev C, Birbaumer N, Larbig W, Taub E. Phantom-limb pain as a perceptual correlate of cortical reorganization following arm amputation. Nature 375(6531):482-484, 1995.
Flor H, Nikolajsen L, Staehelin Jensen T. Phantom limb pain: a case of maladaptive CNS plasticity? Nat Rev Neurosci 7(11):873-881, 2006.
Frot M, Magnin M, Mauguiere F, Garcia-Larrea L. Human SII and posterior insula differently encode thermal laser stimuli. Cereb Cortex 17(3):610-620, 2007.
Geha PY, Baliki MN, Harden RN, Bauer WR, Parrish TB, Apkarian AV. The brain in chronic CRPS pain: abnormal gray-white matter interactions in emotional and autonomic regions. Neuron 60(4):570-581, 2008.
Golay X, Petersen ET. Arterial spin labeling: benefits and pitfalls o.f high magnetic field. Neuroimaging Clin N Am 16(2):259-268, 2006.
Goldstein RZ, Craig AD, Bechara A, Garavan H, Childress AR, Paulus MP, Volkow ND. The neurocircuitry of impaired insight in drug addiction. Trends Cogn Sci 13(9):372-380, 2009.
Guindon J, Hohmann AG. The endocannabinoid system and pain. CNS Neurol Disord Drug Targets 8(6):403-421, 2009.
Gusnard DA, Raichle ME. Searching for a baseline: functional imaging and the resting human brain. Nat Rev Neurosci 2(10):685-694, 2001.
Guye M, Bartolomei F, Ranjeva JP. Imaging structural and functional connectivity: towards a unified definition of human brain organization? Curr Opin Neurol 21(4):393-403, 2008.
Harris RE, Sundgren PC, Pang Y, Hsu M, Petrou M, Kim SH, Mclean SA, Gracely RH, Clauw DJ. Dynamic levels of glutamate within the insula are associated with improvements in multiple pain domains in fibromyalgia. Arthritis Rheum 58(3):903-907, 2008.
Herbert MK, Holzer P. [Why are substance P(NK1)-receptor antagonists ineffective in pain treatment?]. Anaesthesist 51(4):308-319, 2002.
Hernandez-Garcia L. Arterial spin labeling for quantitative functional MRI. Conf Proc IEEE Eng Med Biol Soc 7:5230-5233, 2004.
Hill R. NK1 (substance P) receptor antagonists — why are they not analgesic in humans? Trends Pharmacol Sci 21(7):244-246, 2000.
Hou B, Xu L. Values of default mode network to Alzheimer’s disease call for consilience of multimodal neuroimaging and genetics. J Neurosci 30(10):3553-3554, 2010.
Kong J, White NS, Kwong KK, Vangel MG, Rosman IS, Gracely RH, Gollub RL. Using fMRI to dissociate sensory encoding from cognitive evaluation of heat pain intensity. Hum Brain Mapp 27(9):715-721, 2006.
Kuchinad A, Schweinhardt P, Seminowicz DA, Wood PB, Chizh BA, Bushnell MC. Accelerated brain gray matter loss in fibromyalgia patients: premature aging of the brain? J Neurosci 27(15):4004-4007, 2007.
Kissin I. The development of new analgesics over the past 50 years: a lack of real breakthrough drugs. Anesth Analg 110(3):780-789, 2010.
Lawrence J, Mackey SC. Role of neuroimaging in analgesic drug development. Drugs R D 9(5):323-334, 2008.
Lebel A, Becerra L, Wallin D, Moulton EA, Morris S, Pendse G, Jasciewicz J, Stein M, Aiello-Lammens M, Grant E, Berde C, Borsook D. fMRI reveals distinct CNS processing during symptomatic and recovered complex regional pain syndrome in children. Brain 131(Pt 7):1854-1879, 2008.
Leppa M, Korvenoja A, Carlson S, Timonen P, Martinkauppi S, Ahonen J, Rosenberg PH, Aronen HJ, Kalso E. Acute opioid effects on human brain as revealed by functional magnetic resonance imaging. Neuroimage 31(2):661-669, 2006.
Liu MG, Chen J. Roles of the hippocampal formation in pain information processing. Neurosci Bull 25(5):237-266, 2009.
Logothetis NK. The neural basis of the blood-oxygen-level-dependent functional magnetic resonance imaging signal. Philos Trans R Soc Lond B Biol Sci 357(1424):1003-1037, 2002.
Lorenz IH, Egger K, Schubert H, Schnurer C, Tiefenthaler W, Hohlrieder M, Schocke MF, Kremser C, Esterhammer R, Ischebeck A, Moser PL, Kolbitsch C. Lornoxicam characteristically modulates cerebral pain-processing in human volunteers: a functional magnetic resonance imaging study. Br J Anaesth 100(6):827-833, 2008.
Maihofner C, Handwerker HO. Differential coding of hyperalgesia in the human brain: a functional MRI study. Neuroimage 28(4):996-1006, 2005.
Maihofner C, Handwerker HO, Neundorfer B, Birklein F. Cortical reorganization during recovery from complex regional pain syndrome. Neurology 63(4):693-701, 2004.
Maihofner C, Neundorfer B, Birklein F, Handwerker HO. Mislocalization of tactile stimulation in patients with complex regional pain syndrome. J Neurol 253(6):772-779, 2006.
Maihofner C, Ringler R, Herrndobler F, Koppert W. Brain imaging of analgesic and antihyperalgesic effects of cyclooxygenase inhibition in an experimental human pain model: a functional MRI study. Eur J Neurosci 26(5):1344-1356, 2007.
Martuzzi R, Ramani R, Qiu M, Rajeevan N, Constable RT. Functional connectivity and alterations in baseline brain state in humans. Neuroimage 49(1):823-834, 2010.
May A. Chronic pain may change the structure of the brain. Pain 137(1):7-15, 2008.
Mcewen BS. Plasticity of the hippocampus: adaptation to chronic stress and allostatic load. Ann N Y Acad Sci 933:265-277, 2001.
Morgan V, Pickens D, Gautam S, Kessler R, Mertz H. Amitriptyline reduces rectal pain related activation of the anterior cingulate cortex in patients with irritable bowel syndrome. Gut 54(5):601-607, 2005.
Moulton EA, Burstein R, Tully S, Hargreaves R, Becerra L, Borsook D. Interictal dysfunction of a brainstem descending modulatory center in migraine patients. PLoS One 3(11):e3799, 2008.
Napadow V, Lacount L, Park K, As-Sanie S, Clauw DJ, Harris RE. Intrinsic brain connectivity in fibromyalgia is associated with chronic pain intensity. Arthritis Rheum 62(8):2545-2555, 2010.
Neugebauer V, Li W, Bird GC, Han JS. The amygdala and persistent pain. Neuroscientist 10(3):221-234, 2004.
Oertel BG, Preibisch C, Wallenhorst T, Hummel T, Geisslinger G, Lanfermann H, Lotsch J. Differential opioid action on sensory and affective cerebral pain processing. Clin Pharmacol Ther 83(4):577-588, 2008.
Ogawa S, Lee TM, Kay AR, Tank DW. Brain magnetic resonance imaging with contrast dependent on blood oxygenation. Proc Natl Acad Sci U S A 87(24):9868-9872, 1990
Patel R, Borsook D, Becerra L. Modulation of resting state functional connectivity of the brain by Naloxone infusion. Brain Imaging Behav 2(1):11-20, 2008.
Peyron R, Laurent B, Garcia-Larrea L. Functional imaging of brain responses to pain. A review and meta-analysis (2000). Neurophysiol Clin 30(5):263-288, 2000.
Phan KL, Angstadt M, Golden J, Onyewuenyi I, Popovska A, De Wit H. Cannabinoid modulation of amygdala reactivity to social signals of threat in humans. J Neurosci 28(10):2313-2319, 2008.
Ploghaus A, Narain C, Beckmann CF, Clare S, Bantick S, Wise R, Matthews PM, Rawlins JN, Tracey I. Exacerbation of pain by anxiety is associated with activity in a hippocampal network. J Neurosci 21(24):9896-9903, 2001.
Prescot A, Becerra L, Pendse G, Tully S, Jensen E, Hargreaves R, Renshaw P, Burstein R, Borsook D. Excitatory neurotransmitters in brain regions in interictal migraine patients. Mol Pain 5:34, 2009.
Preuschoff K, Quartz SR, Bossaerts P. Human insula activation reflects risk prediction errors as well as risk. J Neurosci 28(11):2745-2752, 2008.
Prost RW. Magnetic resonance spectroscopy. Med Phys 35(10):4530-4544, 2008.
Rawlins MD. Cutting the cost of drug development? Nat Rev Drug Discov 3(4):360-364, 2004.
Reijneveld JC, Ponten SC, Berendse HW, Stam CJ. The application of graph theoretical analysis to complex networks in the brain. Clin Neurophysiol 118(11):2317-2331, 2007.
Rodriguez I, Perez-Rial S, Gonzalez-Jimenez J, Perez-Sanchez J, Herranz F, Beckmann N, Ruiz-Cabello J. Magnetic resonance methods and applications in pharmaceutical research. J Pharm Sci 97(9):3637-3665, 2008.
Rodriguez-Raecke R, Niemeier A, Ihle K, Ruether W, May A. Brain gray matter decrease in chronic pain is the consequence and not the cause of pain. J Neurosci 29(44):13746-13750, 2009.
Rohayem J, Bayle JF, Richa S. [Predictors of prophylactic response to lithium]. Encephale 34(4):394-399, 2008.
Rowe JB, Hughes LE, Barker RA, Owen AM. Dynamic causal modelling of effective connectivity from fMRI: Are results reproducible and sensitive to Parkinson’s disease and its treatment? Neuroimage 52(3):1015-1026, 2010.
Ruggiero DA, Mraovitch S, Granata AR, Anwar M, Reis DJ. A role of insular cortex in cardiovascular function. J Comp Neurol 257(2):189-207, 1987.
Sandkuhler J. Learning and memory in pain pathways. Pain 88(2):113-118, 2000.
Sawamoto N, Honda M, Okada T, Hanakawa T, Kanda M, Fukuyama H, Konishi J, Shibasaki H. Expectation of pain enhances responses to nonpainful somatosensory stimulation in the anterior cingulate cortex and parietal operculum/posterior insula: an event-related functional magnetic resonance imaging study. J Neurosci 20(19):7438-7445, 2000.
Schweinhardt P, Bountra C, Tracey I. Pharmacological FMRI in the development of new analgesic compounds. NMR Biomed 19(6):702-711, 2006.
Scrivani S, Wallin D, Moulton EA, Cole S, Wasan AD, Lockerman L, Bajwa Z, Upadhyay J, Becerra L, Borsook D. A fMRI evaluation of lamotrigine for the treatment of trigeminal neuropathic pain: pilot tudy. Pain Med 11(6):920-941, 2010.
Seifert F, Bschorer K, De Col R, Filitz J, Peltz E, Koppert W, Maihofner C. Medial prefrontal cortex activity is predictive for hyperalgesia and pharmacological antihyperalgesia. J Neurosci 29(19):6167-6175, 2009.
Staud R. Abnormal pain modulation in patients with spatially distributed chronic pain: fibromyalgia. Rheum Dis Clin North Am 35(2):263-274, 2009.
Tagliazucchi E, Balenzuela P, Fraiman D, Chialvo DR. Brain resting state is disrupted in chronic back pain patients. Neurosci Lett 485(1):26-31, 2010.
Timmermann L, Ploner M, Haucke K, Schmitz F, Baltissen R, Schnitzler A. Differential coding of pain intensity in the human primary and secondary somatosensory cortex. J Neurophysiol 86(3):1499-1503, 2001.
Upadhyay J, Pendse G, Anderson J, Schwarz AJ, Baumgartner R, Coimbra A, Bishop J, Knudsen J, George E, Grachev I, Iyengar S, Bleakman D, Hargreaves R, Borsook D, Becerra L. Improved characterization of BOLD responses for evoked sensory stimuli. Neuroimage 49(3):2275-2286, 2010.
Valet M, Gundel H, Sprenger T, Sorg C, Muhlau M, Zimmer C, Henningsen P, Tolle TR. Patients with pain disorder show gray-matter loss in pain-processing structures: a voxel-based morphometric study. Psychosom Med 71(1):49-56, 2009.
Valfre W, Rainero I, Bergui M, Pinessi L. Voxel-based morphometry reveals gray matter abnormalities in migraine. Headache 48(1):109-117, 2008.
Viswanathan A, Freeman RD. Neurometabolic coupling in cerebral cortex reflects synaptic more than spiking activity. Nat Neurosci 10(10):1308-1312, 2007.
Wiech K, Tracey I. The influence of negative emotions on pain: behavioral effects and neural mechanisms. Neuroimage 47(3):987-994, 2009.
Wise RG, Rogers R, Painter D, Bantick S, Ploghaus A, Williams P, Rapeport G, Tracey I. Combining fMRI with a pharmacokinetic model to determine which brain areas activated by painful stimulation are specifically modulated by remifentanil. Neuroimage 16(4):999-1014, 2002.
Woolf CJ. Overcoming obstacles to developing new analgesics. Nat Med 16(11):1241-1247, 2010.
Zimmerman ME, Pan JW, Hetherington HP, Lipton ML, Baigi K, Lipton RB. Hippocampal correlates of pain in healthy elderly adults: a pilot study. Neurology 73(19):1567-1570, 2009.
Ziv M, Tomer R, Defrin R, Hendler T. Individual sensitivity to pain expectancy is related to differential activation of the hippocampus and amygdala. Hum Brain Mapp 31(2):326-338, 2010.
[Discovery Medicine; ISSN: 1539-6509; Discov Med 11(58):209-219, March 2011.]