Contribution of Positron Emission Tomography (PET) for Understanding Neuronal Activation and Neurotransmission in Pain

FIGURE 1 This figure visualizes findings of ligand-PET studies in chronic pain states: alterations of the opioidergic binding potential (BP) were observed in various pain processing and modulating brain regions including the anterior (ACC) and middle cingulate cortex (MCC), thalamus (Thal), insula (Ins), and amygdala (Amy). Patients suffering from trigeminal neuralgia, rheumatoid arthritis, fibromyalgia, central post-stroke pain (CPSP) as well as from chronic regional pain syndrome (CRPS) show a reduced opioidergic receptor availability in brain areas encoding the sensory-discriminative and/or the emotional-affective component of pain.

As an example, the lower left part of the figure (A) depicts the results of ligand-PET demonstrating decreased BP in MCC, PAG and amygdala in fibromyalgia. The lower right picture (B), shows a strong correlation between the BP in the MCC and individual rating scores of the sensory and affective quality of pain [12].

As opioids have a strong analgesic effect, these findings may explain some aspects of the pathogenesis of chronic pain.

Compared to healthy control subjects, another PET study found reduced glucose uptake in cluster headache patients in the pACC, prefrontal and orbitofrontal cortices. Interestingly, this reduction of metabolism was detectable not only during the acute cluster period (“in bout”), but also during the remission period (out of the bout). This study also revealed that the glucose uptake was significantly higher in patients in bout compared to out of the bout period in brain areas involved in top-down pain modulating circuits such as the pACC and the frontal brain structures. These findings emphasize notions about an altered endogenous pain control system in cluster headache patients [36].

Furthermore, a tracer study explored the opioidergic system in cluster headache patients and found an opioidergic dysfunction in the pineal gland [38]. This part of the brain produces the hormone melatonin that modulates circadian rhythms. At the molecular level, melatonin affects the beta-endorphin level, which has a strong analgesic effect, and as opioids can stimulate the secretion of melatonin, an opioidergic dysfunction in cluster headache may cause an impaired melatonin homeostasis in the pineal gland, which could explain some rhythmic characteristics of the disorder. With longer disease duration, the authors of this study further observed that the receptor availability decreased in the ACC and in the hypothalamus [38].

Migraine

Although the pathophysiology of migraine is not yet completely understood, several PET studies have substantially improved our knowledge about underlying mechanisms of the disorder. Through the means of H₂¹⁵O-PET, hyperactivity in the rostral part of the pons was consistently shown in migraineurs, both during headache attacks and even after the injection of sumatriptan (and resulting pain relief), but not during the interictal interval. This finding suggests that the activity of the brainstem is not the result of pain perception, but might be in fact related to the migraine attack itself. It would appear as though this hyperactivity persists for the entire length of the migraine attack and does not decrease before the attack is remitting. This part of the brainstem has thus been suggested to be a ‘migraine generator’, although clearly more evidence is necessary to substantiate this concept [3,4,43]. A seminal ligand-PET study demonstrated alterations of the 5HT (5-hydroxytryptamine) synthesis in migraine patients in several cortical regions as well as in the dorsal part of the brainstem [5], where a large amount of the neurotransmitter serotonin is localized (in the dorsal raphe nucleus and the nucleus raphe magnus). The highest ligand binding values were detected during acute migraine attacks, the lowest binding values were detected after the administration of sumatriptan and intermediate values were observed during the pain-free period. As serotonin levels in migraineurs were similar during attacks to those of healthy controls (but decreased interictally), these findings point to the contribution of reduced serotonin levels in the pathogenesis of migraine [32].

Increased blood flow was further observed in the hypothalamus not only during the headache phase of migraine attacks [6], but already at the earliest clinical stage of an attack, the so-called premonitory phase, when patients experience symptoms such as yawning, tiredness, reduced concentration or cravings [35]. Hypothalamic dysfunction would be in line with the clinical picture of the disease as non-headache symptoms often occur before (e.g. fatigue), during (e.g. nausea, vomiting) and after headache attacks (e.g. hyperuricemia). Moreover, many trigger factors of migraine attacks are related to hypothalamic (dys-)functions, namely hormonal fluctuations, alterations of the sleep-wake cycle, delaying a meal, or stress.

Using FDG-PET a further study was able to examine glucose metabolism in migraine patients during the pain-free interval and reported reduced glucose uptake in numerous pain processing brain areas such as the ACC, the insula, and the posterior, prefrontal and primary somatosensory cortices. Interestingly, the abnormal hypometabolism corresponds to an increasing disease duration and attack frequency. The authors thus speculated that an altered metabolism in the pain processing network may be the result of recurrent head pain and might, therefore, be reversible [18]. In line with this finding, a further FDG-PET study showed that the process in which medication overuse transforms episodic migraine into chronic daily headache is generally reversible. Before withdrawal of analgesics, hypometabolism was present in the thalamus, the orbitofrontal, the anterior cingulate, the inferior parietal cortices, the insula, and in the striatum, whereas hypermetabolism was observed in the cerebellar vermis. Except for the orbitofrontal cortex (OFC), metabolism normalized in each of these regions after withdrawal (some months later). The authors hypothesized that the persistent hypometabolism in the OFC is related to an abuse trait that most likely persists even after 3 month of withdrawal and increases the risk for recurrent overuse of analgesic medications [9].

LONGITUDINAL PET-STUDIES FOR ASSESSING TREATMENT OUTCOMES

The efficacy of pain treatments is usually assessed using subjective verbal reports of the patients. The next paragraph describes examples of how imaging methods such as PET can be used to objectively characterize dynamic changes in the neurotransmission following various interventions.

Motor Cortex Stimulation

In two seminal studies, motor cortex and spinal cord stimulation induced changes of brain activation were studied in patients suffering from neuropathic pain with H₂¹⁵O-PET [19,30]. The authors were able to show decreased pain intensity ratings in response to both stimulation techniques. In the brain, this was reflected by increases of cerebral blood flow in specific regions involved in the descending pain modulatory system (e.g. the ACC).

A further PET study explored the effect of motor cortex stimulation in neuropathic pain on the opioidergic receptor occupancy utilizing the tracer ¹¹C-diprenorphine. After several months of invasive motor cortex stimulation a decrease of the ligand binding potential was observed in brain areas belonging to the endogenous pain control system, i.e. in the prefrontal and the cingulate cortex as well as in the PAG. These findings were accompanied by reduced pain intensity ratings. Thus, motor cortex stimulation likely induces an enhanced release of endogenous opioids. The data provide further evidence for the relevance of an impaired endogenous opioidergic system in chronic pain [22]. Using the ligand ¹¹C-diprenorphine, another recent PET study further explored longitudinal effects of motor cortex stimulation induced by implanted electrodes in refractory neuropathic pain patients. PET scans were assessed in both preoperative and postoperative sessions and it was shown that the level of preoperative opioid binding in some key regions (i.e. the insula, thalamus, PAG, ACC, and the orbitofrontal cortex) was positively correlated with postoperative pain relief after 7 months, i.e. patients with the lowest pre-operative opioid receptor density benefited the least from the procedure. The authors of the study argued that these findings may aid clinicians in selecting appropriate candidates for surgery who likely benefit from the invasive procedure [23].

Deep Brain Stimulation

Only gold members can continue reading. Log In or Register to continue