Introduction

Trigeminal neuropathic pain (TNP) disorders, such as classical, atypical and postherpetic neuralgias, are persistent pain conditions that can be either spontaneous, or elicited by light touch to the face [1]. The fact that the current therapeutic modalities that focus only on peripheral mechanisms (e.g. microvascular decompression and percutaneous stereotactic rhizotomy) do not provide long lasting relief for these frequently treatment-refractory patients raises the possibility that the causes for the chronicity of those debilitating disorders may also be related to central nervous system alterations. In fact, cortical thickness changes were found in TNP patients, which co-localized with functional (de)activation following allodynic stimulation (brush induced pain) [2]. Furthermore, those neuroplastic changes in the TNP patients were confined to cortical systems associated with pain experience and modulation, especially associated with the μ-opioidergic system, arguably one of the mechanisms centrally involved in the regulation of multiple aspects of the pain experience [3].

Studies with positron emission tomography (PET) using non-selective (e.g., 11C]diprenorphine) radiotracers or 11C]carfentanil, a selective μ-opioid receptor (μOR) radiotracer, have shown reduced opioid receptor availability in chronic pain syndromes such as rheumatoid arthritis [4], neuropathic pain [5], fibromyalgia [6] and complex regional pain syndrome [7]. Such findings might represent either greater occupancy of opioid receptors by their endogenous ligands, down-regulation of opioid receptors after persistent activation during pain, or both [6]. At the present time, it is unknown whether the μ-opioid system is involved in TNP, and the clinical consequences of that involvement. A down-regulation of μORs could explain hypersensitivity (e.g., allodynia, comorbidity with other pain disorders) and frequent treatment refractoriness, including that to opiate medications. Hence, in this preliminary study we investigated changes in the baseline μORs BPND in patients diagnosed with TNP when compared to age-matched pain-free healthy subjects. Based on the existing literature, we hypothesized that patients suffering from TNP would show reduced μOR BPND in regions related to pain regulation, possibly representing persistent activation of the endogenous opioid system and subsequent dysregulation due to the ongoing pain. To our knowledge, this is the first study investigating changes in the endogenous μ-opioid system of patients with TNP in vivo.

Materials and methods

Subjects

We recruited four right-handed refractory TNP patients (three males and one female; mean age = 50.5 ± 16.5), and eight gender and age-matched healthy subjects (six males and two females, mean age = 44.1 ± 14.9). The selection of TNP patients met the criteria defined by the International Headache Society (IHS) [8], American Academy of Orofacial Pain (AAOP) [9] and International Association for the Study of Pain (IASP) Terminology [10]. We only included patients with: A) TNP for at least six months not adequately controlled by previous medicine therapies; B) Minimal average pain score of 4 (moderate to severe) in the visual analogue scale (VAS); C) Unilateral pain; D) Orofacial allodynic region to mechanical (light touch) or thermal stimulation (heat or cold) and E) ages from 18 to 65. The exclusion criteria included: A) Evidence of other local pathology (e.g., orofacial lesion); B) Recent unrelated orofacial surgery or trauma (< 6 months); C) History of systemic disorders (e.g., multiple sclerosis) or D) Chronic pain other than TNP (e.g., back pain or migraine); E) Use of narcotic analgesics (< 6 months); F) Major psychiatric illnesses (current schizophrenia, major depression with suicidal ideation, or substance abuse within two years); and G) Contra-indications to PET. All healthy controls were: right handed; with ages between 18 and 65 years old; with no history of chronic medical illnesses. Approximately 43 subjects applied to participate in the TNP group. However, only four were considered eligible and completed the study. The reasons that determined the exclusion of subjects for this group were: age (1), overweight (9), presence of other chronic pain disorders (18), multiple sclerosis (4) and use of opioid medication (7). Patients in opioid therapy were not recruited for this study. However, the use of other types of medications (e.g. analgesics, anticonvulsants and antidepressants) was not part of the exclusion criteria. This research investigation was carried out in accordance with the bioethical rules for studies involving human beings of the WMA (World Medical Association)––Declaration of Helsinki (1990), and all of the procedures applied were approved by the University of Michigan Investigational Review Board for Human Subject Use, and the Radioactive Drug Research Committee of the US Food and Drug Administration. All subjects gave written informed consent prior to the participation in the study.

Clinical assessment

All subjects were initially screened by obtaining the medical history, and performing a clinical orofacial pain exam by a pain specialist. During this visit each subject was asked to complete the McGill Pain Questionnaire (MPQ) [11], which provided quantitative measures of clinical pain. This questionnaire has three major classes of pain descriptors (sensory, affective and evaluative), used to measure the subjective pain experience. It also provides the total Pain Rating Index (PRI), based on the rank values of the words selected as descriptors and the Present Pain Intensity (PPI), a 0–5 intensity scale. In addition to the MPQ, patients with TNP were requested to complete an craniofacial pain map [12]. Implemented in an in-house mobile application (PainTrek®, University of Michigan), this method provides a 3D head and facial map based on a squared grid system with vertical and horizontal coordinates using anatomical landmarks. Each quadrangle, measuring approximately 1.6 cm × 1.6 cm, frames well-detailed craniofacial and cervical areas and can be filled by the patient to express his/her exact pain location. The method allowed the investigators to precisely localize and measure the total pain area, as well as the dermatomes involved in each TNP patient (Figure 1).

Neuroimaging

A T1-weighted anatomical MRI scan was acquired on a 3 Tesla scanner (General Electric, Milwaukee, WI). The MRI acquisition utilized the following sequence parameters: axial spoiled-gradient recalled (SPGR) 3D acquisition, 15.63 bandwidth, repetition time [TR] = 9.2 ms, echo time [TE] = 1.9 ms, inversion recovery preparation 500 ms, flip angle = 15°, 25/26 FOV, number of excitations [NEX] = 1, 144 contiguous slices, 1.0 mm slice thickness, 256 x 256 matrix. VBM8 toolbox ( http://dbm.neuro.uni-jena.de/vbm.html) in SPM8 (Wellcome Department of Imaging Neuroscience Group, London, UK) was used for the normalization of the MRI data to MNI (Montreal Neurological Institute) space. PET scans (HR+ scanner; Siemens, Knoxville, TN) were acquired in 3D mode (reconstructed full-width/half-maximum resolution 5.5 mm in-plane and 5.0 mm axially), with septa retracted and scatter correction. Tracer quantity 11C]carfentanil was administered (10–15 mCi, ≤ 0.03 μg/kg) through an intravenous line (50% in initial bolus and remainder continuously infused to more rapidly achieve constant plasma concentrations). 11C]carfentanil was synthesized at high specific activity (> 2000 Ci/mmol) by the reaction of 11C]methyliodide and a non-methyl precursor as previously described [13, 14]. Images were decay-corrected and reconstructed, and the dynamic frames were coregistered to each other and transformed into tracer transport (K1 ratio) and receptor-related (BPND, binding potential) measures. To avoid the need for arterial blood sampling, these measures were calculated using a modified Logan graphical analysis [15], with occipital cortex (a region devoid of μ- opioid receptors) as the reference region. After 5–7 minutes of radiotracer administration, the Logan plot becomes linear with slope = BPND + 1, which is proportional to μOR concentrations (Bmax)/receptor radiotracer affinity (Kd) (Bmax/Kd ≈ BPND). PET images were co-registered to the individual T1-weighted MRI and then normalized by applying the deformation matrix obtained from the normalization of the T1-weighted MRI data to μOR binding maps.

The control subjects and patients with TNP were compared voxel-by-voxel using unpaired t-test on μOR BPND data. In view of the small sample size of this pilot study, we set significance at p ≤ 0.001, uncorrected with a priori hypothesis (regions involved in pain modulation). Cluster size is reported for voxels with p < 0.01 in the area of statistical significance.

Results

No age differences were found between groups (TNP and healthy controls, unpaired t-test, t = 0.676, p = 0.86). TNP patients had average pain duration of 5.4 ± 1.6 years; two patients presented with either spontaneous or evoked pain, while the other two presented both pain modalities. On a scale from 0 to 10 (zero representing no pain and ten representing worst pain possible) the average of spontaneous pain intensity (n = 3) was 5.5 ± 1.3, while the average of evoked pain (n = 3) was 5.4 ± 3.9. All TNP patients had previously tried medications for pain control, including: carbamazepine (subject TNP 3), oxcarbazepine, gabapentin (subjects TNP1 and TNP4), pregabalin (subjects TNP 2 and TNP 4) and amitriptyline (subject TNP 2). Subjects TNP1, TNP2 and TNP3 were under pain control medication during the study. As described before, patients taking opioids were not recruited for this study. The Figure 1 summarizes the main clinical characteristics from each TNP patient included in this study.

TNP patients displayed a significant reduction of the μOR BPND in the ventral striatal area (z = 2.98, cluster size 989 mm3, p = 0.001) of the basal ganglia, with the greatest difference being found in the left nucleus accumbens (NAc) (peak MNI coordinates x/y/z: -4/4/-5). (Figure 2A-D).

Figure 1
figure 1

Clinical profile of Trigeminal Neuropathic Pain patients recruited for this study (for more information, refer to the text). The pain location was accessed by a craniofacial pain map of the face [12]. This method (PainTrekW, University of Michigan) provides a 3D map of orofacial pain.

Full size image

The correlation analysis between clinical pain levels and μOR BPND availability in TNP patients showed that McGill Total PRI scores coincided with reduced μOR in the left NAc. Reduced μOR BPND in the left NAc was also associated to higher McGill sensory subscale scores (Figure 2E).

Discussion

The present study compared the μOR BPND of TNP patients to age-matched healthy controls. We found a significant decrease of μOR BPND in the basal ganglia, located mainly in the left NAc of patients with TNP, when compared to controls. Other groups have found reductions in opioid receptor BPND measured with the non-selective radiotracer 11C]diprenorphine in other chronic pain disorders, such as rheumatoid arthritis [4], complex regional pain syndrome [7], cluster headache [16] and neuropathic pain [5, 17, 18]. Those changes involved brain regions known to play important roles in pain processing, such as the insula, anterior cingulate cortex, amygdala, hypothalamus, caudate nucleus and NAc. A significant increase in the volume of distribution of 11C]diprenorphine (availability of opioid receptors) was also shown after treatment with radiofrequency thermocoagulation (RFTC) in patients with trigeminal pain [19]. Our findings reinforce the concept that the ongoing pain experience in TNP is linked to the persistent activation of endogenous opioid neurotransmission and the subsequent downregulation of μ-opioid receptors. Considering the limited number of patients recruited for this study, it is possible that increasing the sample size might expand our results to a broader set of brain structures.

Figure 2
figure 2

Changes in the μ-opioid receptor availability in Trigeminal Neuropathic Pain. A-C, Representation of decreased μOR BPND in the left NAc in axial (A), sagittal (B) and coronal (C) planes (T = 3.2). D, Plots of individual μOR BPND extracted from the left NAc. Each TNP patient is represented in a black circle and each healthy subject in a white circle. E, Correlations between MPQ scores (PRI and Sensory) and μOR BPND in the left NAc of TNP patients.

Full size image

In our study, the most significant peak in binding potential reduction of μ-opioid receptors was in the left nucleus accumbens (NAc) of TNP patients when compared to age-matched healthy subjects (Figure 2D). The NAc is located in the ventral striatum, at the interface of sensorimotor and limbic systems, and is part of the circuit involved in the integration of cognitive, affective and motor responses, which is modulated by the endogenous opioid system [20–22]. It receives inputs from limbic areas, such as amygdala and prefrontal cortex and projects to different structures, including brainstem and ventral pallidum [23]. It is largely recognized to be involved in reward and aversive behaviors, and in placebo response [24]. However, there is substantial evidence from both animal [25–27] and clinical studies [3, 24, 28] that the NAc is involved in pain processing, including TNP [29]. More recently, functional changes in the NAc signal were observed in a model of peripheral nerve injury [27], and a decrease in the NAc gray matter volume was demonstrated in TNP patients [29]. Along similar line, NAc was the only brain region differentiating healthy volunteers and chronic low back pain patients in an fMRI study examining regional brain activations related to acute painful thermal stimulation [30, 31]. Regarding its relationship to the endogenous μ-opioid system, the NAc was one of the areas where reductions in μOR BPND were identified in fibromyalgia patients when compared to healthy controls. Conversely, short- and long-term increases in the μOR BPND were observed in the same area after traditional chinese acupuncture, which were associated with improvements in clinical pain ratings in fibromyalgia patients [32]. The results of our study support the evidence of the NAc participation in the pain processing, previously proposed in the Motivation-Decision Model of pain [33, 34].

Differences in function of NAc μ-opioid receptors could ultimately contribute to the clinical pain in trigeminal neuropathic pain disorders. A negative relationship was detected between the μOR BPND in the left NAc and the McGill scores (MPQPRI and MPQ sensory) in sample of four TNP patients (Figure 2E). Subjects with higher McGill scores exhibited lower μOR BPND, and subjects with lower McGill scores displayed higher μOR BPND. Although the reduced sample size limits the conclusions about this relationship, the results suggest a relationship between the clinical presentation of TNP disorders, and μ-opioid neurotransmission in the NAc, an area related to both pain processing and motivational mechanism. Based on our results, it is possible to hypothesize that the chronicity of trigeminal neuropathic pain is also related to central nervous system molecular neuroplasticity at the level of the μORs in the limbic area of the basal ganglia. This initial proof of concept study supports the initiation of further studies to examine the central molecular mechanisms, such as endogenous opioid neurotransmission, that may influence the clinical course and treatment responses of patients afflicted with persistent TNP.