Advertisement
Browse Subject Areas
?

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here.

  • Loading metrics

Pain Facilitation Brain Regions Activated by Nalbuphine Are Revealed by Pharmacological fMRI

  • Robert Gear,

    Affiliation Department of Oral & Maxillofacial Surgery, University of California San Francisco, San Francisco, California, United States of America

  • Lino Becerra,

    Affiliations Center for Pain and the Brain, Harvard Medical School, Boston, Massachusetts, United States of America, Brain Imaging Center, McLean Hospital, Belmont, Massachusetts, United States of America, Harvard Medical School, Boston, Massachusetts, United States of America, Departments of Radiology and Psychiatry, Massachusetts General Hospital, Harvard Medical School, Boston, Massachusetts, United States of America

  • Jaymin Upadhyay,

    Affiliations Center for Pain and the Brain, Harvard Medical School, Boston, Massachusetts, United States of America, Brain Imaging Center, McLean Hospital, Belmont, Massachusetts, United States of America, Harvard Medical School, Boston, Massachusetts, United States of America

  • James Bishop,

    Affiliations Center for Pain and the Brain, Harvard Medical School, Boston, Massachusetts, United States of America, Brain Imaging Center, McLean Hospital, Belmont, Massachusetts, United States of America, Harvard Medical School, Boston, Massachusetts, United States of America

  • Diana Wallin,

    Affiliations Center for Pain and the Brain, Harvard Medical School, Boston, Massachusetts, United States of America, Brain Imaging Center, McLean Hospital, Belmont, Massachusetts, United States of America, Harvard Medical School, Boston, Massachusetts, United States of America

  • Gautam Pendse,

    Affiliations Center for Pain and the Brain, Harvard Medical School, Boston, Massachusetts, United States of America, Brain Imaging Center, McLean Hospital, Belmont, Massachusetts, United States of America, Harvard Medical School, Boston, Massachusetts, United States of America

  • Jon Levine ,

    jon.levine@ucsf.edu

    Affiliations Department of Oral & Maxillofacial Surgery, University of California San Francisco, San Francisco, California, United States of America, Department of Medicine, University of California San Francisco, San Francisco, California, United States of America

  • David Borsook

    Affiliations Center for Pain and the Brain, Harvard Medical School, Boston, Massachusetts, United States of America, Brain Imaging Center, McLean Hospital, Belmont, Massachusetts, United States of America, Harvard Medical School, Boston, Massachusetts, United States of America, Departments of Radiology and Psychiatry, Massachusetts General Hospital, Harvard Medical School, Boston, Massachusetts, United States of America

Pain Facilitation Brain Regions Activated by Nalbuphine Are Revealed by Pharmacological fMRI

  • Robert Gear, 
  • Lino Becerra, 
  • Jaymin Upadhyay, 
  • James Bishop, 
  • Diana Wallin, 
  • Gautam Pendse, 
  • Jon Levine, 
  • David Borsook
PLOS
x

Abstract

Nalbuphine, an agonist-antagonist kappa-opioid, produces brief analgesia followed by enhanced pain/hyperalgesia in male postsurgical patients. However, it produces profound analgesia without pain enhancement when co-administration with low dose naloxone. To examine the effect of nalbuphine or nalbuphine plus naloxone on activity in brain regions that may explain these differences, we employed pharmacological magnetic resonance imaging (phMRI) in a double blind cross-over study with 13 healthy male volunteers. In separate imaging sessions subjects were administered nalbuphine (5 mg/70 kg) preceded by either saline (Sal-Nalb) or naloxone 0.4 mg (Nalox-Nalb). Blood oxygen level-dependent (BOLD) activation maps followed by contrast and connectivity analyses revealed marked differences. Sal-Nalb produced significantly increased activity in 60 brain regions and decreased activity in 9; in contrast, Nalox-Nalb activated only 14 regions and deactivated only 3. Nalbuphine, like morphine in a previous study, attenuated activity in the inferior orbital cortex, and, like noxious stimulation, increased activity in temporal cortex, insula, pulvinar, caudate, and pons. Co-administration/pretreatment of naloxone selectively blocked activity in pulvinar, pons and posterior insula. Nalbuphine induced functional connectivity between caudate and regions in the frontal, occipital, temporal, insular, middle cingulate cortices, and putamen; naloxone co-admistration reduced all connectivity to non-significant levels, and, like phMRI measures of morphine, increased activation in other areas (e.g., putamen). Naloxone pretreatment to nalbuphine produced changes in brain activity possess characteristics of both analgesia and algesia; naloxone selectively blocks activity in areas associated with algesia. Given these findings, we suggest that nalbuphine interacts with a pain salience system, which can modulate perceived pain intensity.

Introduction

In clinical studies of the analgesic efficacy of agonist-antagonist kappa opioids we found that all three clinically available agents in this class (viz., nalbuphine, pentazocine, and butorphanol) produce sexually dimorphic analgesia [1], [2], with males experiencing significantly less analgesia than females. Although the reasons for this sex difference have not been determined, a placebo-controlled dose response study of the analgesic effect of nalbuphine in patients experiencing moderate to severe post-operative pain showed that males, and not females, receiving the lowest dose of nalbuphine (5 mg), reported significantly greater pain than those receiving placebo [3], suggesting the presence of a pain-facilitating mechanism in males, not present or significantly diminished in females. Subsequently, we found that co-administration of the opioid antagonist naloxone, at low dose, abolished pain facilitation in males, resulting in profound analgesia that was very similar to that produced by nalbuphine in females [4]. Based on these findings we hypothesized that nalbuphine has pain-facilitating as well as analgesic effects produced by action at different receptors in the brain, an “analgesia” receptor, likely the κ-opioid receptor [5] and a “pain-facilitating” receptor, and that the latter, which is predominantly found in males, is more sensitive to naloxone. This model could explain the ability of naloxone to enhance analgesia by selectively blocking the pain-facilitating effect of nalbuphine.

The present study investigated putative analgesia and pain-facilitation brain circuitry affected by nalbuphine, in males. Since a radio-ligand for κ-opioid receptors is not currently available for use in human positron emission tomography (PET) or single positron emission tomography (SPECT) studies, we employed pharmacological magnetic resonance imaging (phMRI), a form of functional magnetic resonance imaging (fMRI), to identify regions of the brain whose blood oxygen level-dependent (BOLD) signals are increased or decreased by nalbuphine in subjects pretreated with either saline or naloxone. Important to the present study, phMRI (a measure of BOLD activation/brain function as a direct consequence of drug induced activity) is a technique that enables the measures of activity in brain regions affected by analgesics and other drugs without sensory stimulation [6], [7]. We have previously used phMRI to study other opioid agonists and antagonists including morphine [8] and naloxone [9], and most recently buprenorphine [10], demonstrating the utility of this approach to evaluate drug action on brain circuits.

In the present study male volunteers were administered nalbuphine preceded by vehicle saline or nalbuphine preceded by naloxone, in a double blind, crossover, two-session design to test the hypothesis that nalbuphine affects activity in brain regions and circuits related to pain-facilitation, as well as analgesia, and its effects on sites and circuits involved in possible pain facilitation are attenuated by naloxone pretreatment/co-administration.

Methods

To investigate the differences in phMRI brain activation produced by nalbuphine and nalbuphine in combination with naloxone, each subject was scanned twice in a cross-over double blind study. Thus, each subject received nalbuphine infusions on each trial, once with saline pretreatment (Sal-Nalb) and once with naloxone pretreatment (Nalox-Nalb).

Subjects

Recruitment.

The McLean Hospital Institutional Review Board approved the study. The study also complied with the guidelines of the Helsinki Accord and International Association for the Study of Pain (IASP) for experimental pain research in humans. Fifteen healthy male volunteers (age 24.07±2.56 (mean ± SD)) were recruited through advertisements in local newspapers, flyers on local college campuses, and on recruiting websites (Partners research site (CRNet), and Craig's List (http://boston.craigslist.org/). All subjects signed a written informed consent to participate in the study.

Enrollment.

Healthy right-handed male candidates 18–45 years of age attended a screening session where compliance with inclusion/exclusion criteria was evaluated and a full medical examination including was performed by a board certified neurologist. Inclusionary criteria. Candidates meeting any of the following criteria were excluded from participation in the study: current use of prescribed medications, BDI score >11, history of opioid abuse, history of smoking within the past year, claustrophobia, significant medical problems, significant alcohol intake (five or more glasses/week), history of allergy or untoward reaction to anticonvulsants, metal implants of any type, weight greater than 130 kg. An FDA-approved urine toxicology screen (7 Drug InstaStrip Drug Screen Test, Cortez Diagnostics, Calabasas, CA), which tested for barbiturates, benzodiazepines, amphetamine, cocaine, tetrahydrocannabinol, phencyclidine and opioids, was performed prior to each MR session. None of the subjects had a positive drug screen. Thirteen of the 15 subjects (three Asian, one black, 9 white, aged 24±2.56 years (mean ± s.d), weight 79±16.4 kg, completed both scanning sessions.

Subject Preparation.

All subjects were instructed to refrain from eating starting eight hours before the scan but were allowed to consume clear liquids. Subjects were scheduled for their second scan two weeks after the first.

Study Design

Drug Administration.

Naloxone and nalbuphine (both from Hospira Inc., Lake Forest, IL) were administered through contralateral intravenous lines placed in the antecubital region of each arm. Naloxone (0.4 mg) or saline was followed 5, 7, 9 and 11 minutes later by a divided dose of nalbuphine (5 mg/70 kg) administered through the access port in the left arm in 2 ml boluses at 0.1 ml/s (Fig. 1). Doses were chosen on the basis of our previous clinical studies [3], [4], [11]. Nalbuphine administration was carried out with an MRI-compatible microinjector (Medrad Spectris, Colombus, OH).

thumbnail
Figure 1. Drug administration protocol.

After the initial anatomical scans were obtained, either naloxone or saline was infused for Scan 1; for Scan 2 the other drug was infused. Nalbuphine infusion was then started five minutes later. One quarter of the nalbuphine dose was administered every other minute until the full amount was delivered by eight minutes.

https://doi.org/10.1371/journal.pone.0050169.g001

Randomization.

To neutralize any effect of treatment order, about half of the subjects received saline first and the other half received naloxone first, in a randomized, double-blinded, crossover design. Subjects as well as team members involved in drug administration and initial analysis were blinded to drug group assignment.

Scanning

Anatomical scans were acquired prior to drug administration (Fig. 1). Infusion scanning began immediately after infusion of naloxone or saline (five minutes before infusion of the first bolus of nalbuphine).

Data Acquisition.

Data were collected on a 3T Siemens Trio scanner with an 8-channel phased array head coil (Erlangen, Germany). phMRI data were collected using a gradient echo-echo planar pulse sequence (GE-EPI) at 3.5×3.5×3.5 mm3 resolution.

GE-EPI Parameters.

Time of Repetition (TR) = 2500 msec, Time of Echo (TE) = 30 msec, Field of View (FOV) = 224×224, Flip Angle (FA) = 90°, number of Slices = 41 axial slices and number of Volumes = 600. The acquisition time for the infusion scan was 25 min 5 sec. T1-weighted structural images were acquired using a 3-D magnetization-prepared rapid gradient echo (MPRAGE) sequence at a resolution of 1.33×1.0×1.0 mm3. MPRAGE Parameters: TR = 2100 msec, TE = 2.74 msec, Time of Inversion (TI) = 1100 msec, FA = 12°, number of Slices = 128 sagittal slices.

Ancillary Measures

Hedonic Ratings.

Hedonic ratings were recorded following nalbuphine administration. Prior to the scan subjects were instructed on how to rate their hedonic experiences using a dial controlled by the right hand and a visual analog scale (VAS) display that could be seen from within the bore of the magnet. The VAS anchors were ‘neutral’ in the middle of the scale, ‘max euphoria’ on the right, and ‘max dysphoria’ on the left.

Physiological Measures.

Throughout the period of the scans, pulse oximetry was employed to monitor heart rate (HR) and blood oxygen saturation (SpO2), and a nasal cannula was placed to monitor respiratory rate (RR) and end tidal CO2.

Imaging Data Analysis

Preprocessing.

Anatomical data: High resolution anatomical images for each subject were bias-corrected and segmented into white matter (WM), gray matter (GM) and cerebrospinal fluid (CSF) using FAST [12].

phMRI data: Preprocessing was performed using tools from the FMRIB Software Library (FSL) (http://www.fmrib.ox.ac.uk/fsl/). The first 2 volumes were removed from each subject's phMRI dataset to allow the MR signal to stabilize.

Preprocessing steps on the remaining 4D volume included: (1) motion-correction using MCFLIRT [13]; (2) brain extraction using BET [14]; (3) spatial smoothing using an isotropic Gaussian kernel of 5 mm FWHM; (4) grand mean intensity normalization of the entire 4D dataset by a single multiplicative factor; (5) registration of the functional space template to the anatomical space and the MNI 152 space using an affine transform with 12 degrees of freedom via FLIRT [13]; and (6) creation of the WM and CSF masks in functional space for each subject using the functional space affine transformation.

The phMRI data was not high pass filtered since our signal of interest is a low frequency infusion response. Preprocessed phMRI was inspected for scanner artifacts using a linear independent component analysis (ICA) based approach [15]. Confounding artifacts were removed from the phMRI before statistical analysis.

Infusion analysis

Single subject.

Analysis was performed using a generalized linear model (GLM) with Gaussian noise. The ideal infusion response (infusion EV) was modeled as a ramped unit step function with a 5 min baseline, a ramp going from 5 min to 13 min and a plateau thereafter. The infusion EV was not convoluted to a hemodynamic response function. The key statistic of interest is the coefficient/amplitude of the “infusion EV” in the GLM. Since the onset of the ramp can be variable depending on local pharmacodynamics, we included additional singular value decomposition regressors to enable unbiased estimation [16] of the infusion amplitude. Further, linear drift and subject specific average time courses extracted from the WM and CSF were included as covariates of no interest in the GLM. The statistical outputs from the single subject analysis were transformed to the MNI 152 space using precomputed affine transformations.

Group level.

For all group-level phMRI analyses, single subject GLM statistics for the amplitude of the “infusion EV” were entered into a paired mixed-effects group-level GLM analysis using a paired t-test design with FLAME [17]. The contrasts of interest were Sal-Nalb>Nalox-Nalb and Sal-Nalb<Nalox-Nalb.

Functional connectivity analysis

Differences in functional connectivity (Fc) between the Nalox-Nalb and Sal-Nalb conditions for specific brain structures were calculated. Only phMRI data from the last 5 min were analyzed for Fc because the drug effects were in a steady state during this time. A seed-based analysis was performed as described below:

Region of interest selection.

Biologically relevant regions of interest (ROIs) were defined in the standard space MNI 152 template and subsequently linearly transformed into the subject-specific fMRI space for functional connectivity analysis.

ROIs that showed significant differences in Fc were selected on the basis of evidence for such changes in observed phMRI infusion responses between Sal-Nalb and Nalox-Nalb treatments.

Single subject Fc analysis.

Functional connectivity analysis was performed utilizing a partial correlation coefficient approach between all brain voxels and the seed time course given the average WM, average CSF, and the linear drift as nuisance covariates. The seed time course was defined as the averaged time course of all the voxels within an ROI. We used partial correlation because it was recently shown to have good sensitivity in detecting network connections [18].

Group-level Fc analysis.

The raw partial correlation coefficient maps from single subject analysis were variance-stabilized using a Fisher z-transform and transformed to the MNI 152 space using precomputed affine transformations. The resulting z-transformed maps were entered into a random effects group-level GLM analysis using a paired t-test design. As in the case of infusion analysis, the contrasts of interest were Sal-Nalb>Nalox-Nalb and Nalox-Nalb->Sal-Nalb.

Statistical inference and thresholding.

The z-statistic maps for each contrast in the infusion analyses as well as the functional connectivity analyses were subjected to alternative hypothesis testing using Gaussian mixture modeling [19]. In each case, the mixture model was spatially regularized using a Markov random field (MRF) that was a soft-max prior on the class labels [20]. This prior encourages spatially neighboring voxels to have similar labels. The mixture model parameters as well as the MRF parameter were adaptively estimated from the data using iterated conditional modes (ICM) [20]. The posterior probability maps (PPMs) giving the “activation” probability of a voxel conditional on the estimated labels in its neighborhood and the observed data were created and thresholded at PPM>0.5 to detect “activation”.

PhMRI comparisons with morphine and naloxone

In order to infer biological relevance from the differences in the patterns of brain activation by Sal-Nalb and Nalox-Nalb, we compared the results of the current study with those of previous studies of morphine [8] and naloxone [9], which employed the same infusion protocol as that for the actual Nalbuphine administration.

Results

Psychophysical ratings and physiological measures

There were no significant differences between the two groups in euphoria-dysphoria scores, respiratory rate, end tidal CO2, or heart rate (data not shown).

phMRI infusion responses

PhMRI BOLD responses.

Sal-Nalb and Nalox-Nalb produced significant changes in activity in many brain regions. Representative examples of the time courses of changes in activation are shown in Fig. 2. In caudate and anterior insula Sal-Nalb induced greater signal changes than did Nalb-Nalox; in contrast, the effects were opposite in the amygdala and the middle cingulate cortex (MCC). Except for the MCC, these changes commenced soon after nalbuphine infusion started. These signal changes exemplify the data used to derive the statistically significant signal changes for regional activation shown in Tables 1, 2, and 3 and mapped in Fig. 3.

thumbnail
Figure 2. Examples of infusion-initiated BOLD signal changes.

Shown is an example in which the Nalb-Sal treatment induced greater activation (i.e., caudate), and an example in which the Nalb-Nalox treatment induced greater activation (i.e., amygdala).

https://doi.org/10.1371/journal.pone.0050169.g002

thumbnail
Figure 3. BOLD phMRI activation maps.

Top: BOLD activation maps for Nalb-Sal (see Table 1); middle: BOLD activation maps for Nalb-Nalox infusions (see Table 2). Bottom: Contrast maps for Nalb-Sal>Nalb-Nalox and Nalb-Nalox>Nalb-Sal (see Table 3).

https://doi.org/10.1371/journal.pone.0050169.g003

Effects of Sal-Nalb on brain activation.

Increased BOLD responses were observed in 60 brain regions including, the occipital cortex, temporal cortex (mostly inferior), anterior and posterior insula, thalamus, caudate, hippocampus, pons, and cerebellum. Decreased responses were observed in 9 brain regions, including the middle frontal cortex, inferior orbitofrontal cortex, post central parietal cortex, superior temporal pole, and cerebellum (Table 1, see Fig. 3 for maps). We interpret these changes to reflect de novo effects of nalbuphine.

Effects of Nalox-Nalb on brain activation.

Nalox-Nalb produced a very different pattern of changes in activity; fewer regions showed increased activity (n = 14) and also fewer regions showed decreased activity (n = 3) compared to Sal-Nalb. Regions with increased activity included the lingual cortex, the anterior insula but not the posterior insula, and the cerebellum. Decreased responses were only observed in the inferior orbital region (Table 2, Fig. 3). Thus, based on infusion activation patterns, large differences were apparent between the Sal-Nalb and Nalox-Nalb sessions.

Differences in brain activation.

Low-dose naloxone significantly blocked nalbuphine activation (i.e., Sal-Nalb>Nalox-Nalb) in superior medial and middle frontal cortex, postcentral parietal cortex (lateral aspect), occipital cortex (rolandic operculum), caudate, pons (trigeminal main sensory nucleus) and cerebellum (Table 3, Fig. 3). Addition of naloxone also induced activation of some areas significantly more than nalbuphine alone (i.e., Nalox-Nalb>Sal-Nalb); these included the superior and inferior orbital cortex (rostral), postcentral parietal cortex (superior aspect), occipital cortex (cuneus), temporal cortex (middle and superior poles), amygdala, putamen, pulvinar, and areas in the cerebellum.

Functional connectivity

Of the regions tested (e.g., thalamus, anterior insula, posterior insula, anterior cingulate, posterior cingulate, and caudate), only caudate was found to demonstrate significant Fc. Nalbuphine induced functional connectivity of the caudate with the following cortical areas: superior frontal, superior medial frontal, and middle frontal, middle orbitofrontal, inferior operculum, precentral frontal, superior and inferior parietal, calcarine and middle occipital, superior temporal and fusiform, middle cingulate, and anterior insula (Table 4, Fig. 4). Naloxone eliminated this connectivity between the caudate and all of these areas.

thumbnail
Figure 4. Functional Connectivity Maps.

Caudate functional connectivity induced by Nalb-Sal that was blocked by naloxone (see Table 4). There was no functional significant connectivity induced by the Nalb-Nalox treatment.

https://doi.org/10.1371/journal.pone.0050169.g004

Discussion

In this study we investigated the effect of nalbuphine on brain circuitry implicated in analgesia and pain. For this analysis we used phMRI to identify changes in BOLD activity following infusion of nalbuphine with and without naloxone pretreatment. Nalbuphine preceeded by saline (i.e., Sal-Nalb) administration significantly increased activity in 60 discreet brain regions and decreased it in 9 (Table 1); nalbuphine pretreated with naloxone (i.e., Nalox-Nalb) significantly increased activity in only 14 regions and decreased it in only 3 (Table 2). Given that nalbuphine is an opioid agonist and naloxone an opioid antagonist, these results should perhaps not be surprising; naloxone blocked many actions of nalbuphine. These results might help explain nalbuphine's pain-facilitating effect when administered alone and its analgesic effect when pretreated with naloxone. Importantly, although pain measures are not presented in the present study, the relevance of the various identified brain regions to analgesia/algesia has been well established in previous studies (see Table 5).

Actions of Nalbuphine (Nalb-Sal)

Because nalbuphine is widely employed as an analgesic, but also produces a marked pain-facilitation in males, its phMRI brain signature might be expected to show characteristics typical of both analgesia and pain, and this is indeed the case. Nalbuphine acted at sites associated with the production of analgesia in that it attenuated activity in some brain regions in which activity was also attenuated by morphine (see Table 5; data from our prior study [8]). However, it also activated many areas that are activated by noxious stimulation, areas that were not activated by either morphine or nalbuphine pretreatment with naloxone. Like morphine, nalbuphine attenuated activity in the inferior orbital cortex, but, like noxious stimulation, it increased activity in several regions in the temporal cortex, insula, thalamus (including pulvinar), caudate, and pons. In contrast to noxious stimulation, nalbuphine decreased activity in the primary somatosensory cortex (S1) and precuneous regions. In addition, nalbuphine induced functional connectivity of the caudate and multiple regions in the frontal, occipital, temporal, insular, middle cingulate cortices, the putamen, and many areas in the cerebellum.

Actions of Nalbuphine in the Presence of Naloxone (Nalb-Nalox)

Fewer brain regions were activated when nalbuphine was administered with naloxone, down from 60 to 14, and also fewer regions were deactivated, down from 9 to 3 (Table 5). Given that naloxone abolishes nalbuphine-induced pain-facilitation and induces significantly enhanced analgesia [4], these results suggest that naloxone blockade affects activity in areas that mediate pain facilitation. Consistent with this suggestion, regions in which naloxone blocked activity included pulvinar, which has been associated with increased activation in central sensitization [21], pons, which has been implicated in remifentanil withdrawal hyperalgesia [22], posterior insula, which is associated with pain-exclusive activations [23], and many areas in the cerebellum, which may have a role in pain processing [24]. Connectivity analysis showed that naloxone reduced nalbuphine-induced functional connectivity of caudate with all regions; that is, there was no significant connectivity in the presence of naloxone. Interestingly, some structures with reduced caudate connectivity (putamen, cingulate, temporal lobe, and cerebellum) showed significantly greater BOLD activation (Tables 3, 5). Increased activation of putamen was also observed with morphine [8], consistent with a role of this region in analgesia [25]. These results suggest that nalbuphine-induced activation of caudate (perhaps by disinhibition, as opioids are inhibitory) initiates pain-enhancing connectivity with other regions, and, although the receptor(s) involved is not known, blockade of this connectivity by naloxone abolishes this pronociceptive effect.

Drug Modulation of Interoception

The insula is postulated to be involved in interoception or a state of subjective feeling, defined as sensitivity to stimuli originating inside of the body [26]. The region, specifically the mid/posterior insula is connected with the posterior middle cingulate cortex (pMCC) that may “integrate interoceptive information with emotional salience to form a subjective representation of the body” [27]. Such ongoing subclinical processing may be altered by naloxone-nalbuphine vs. saline-nalbuphine effects. In essence, changes akin to those surmised in drug addiction may be present, albeit acutely [28]. A potential mechanism of the changes observed here (i.e., naloxone-induced inhibition of the insula cortex) may relate to naloxone effects on opioid receptors in the insula. The insula has high concentrations of opioid receptors [29] and naloxone may act to block an opioid-induced effect – i.e., the nalbuphine effects on the posterior insula. In support of this, naloxone infusion in healthy subjects, as measured by arterial spin labeling showed increases in insula blood flow following fentanyl, but not naloxone [30]. In a similar manner, the effect of naloxone may be to decrease effective action of nalbuphine.

Time course of effects

In males nalbuphine by itself induces early onset analgesia that descends into anti-analgesia between 40 and 90 minutes after administration, depending on dose [3]. For the 5 mg dose of nalbuphine, the approximate dose used in the current study, the duration of analgesia is shorter—about 30 minutes—and the appearance of anti-analgesia is earlier than for higher doses [3]. Thus, since the plasma half-life of naloxone is about an hour [31], the timing of the effects of naloxone observed in the present study are consistent with our clinical observations. Of note, it is likely that nalbuphine alone activates both analgesia and algesia mechanisms at the same time and that addition of naloxone eliminates that effect on the algesic circuitry.

Brain Generated Pain Enhancement

An important implication of these findings is that the brain can generate de novo pain enhancement elicited by the action of a drug such as nalbuphine. Aside from the clinical data arising from lesions of the nervous system (e.g., thalamic or brainstem stroke), few studies have investigated pain facilitation in the brain. Derbyshire and colleagues [32] showed that hypnotically-induced pain is accompanied by activation of classic pain regions, including anterior cingulate cortex (ACC), insula, and prefrontal cortex. Kong and colleagues [33] found that an expectation/conditioning manipulation model can produce a nocebo effect (expectation-induced pain enhancement, opposite of the placebo effect) correlated with activation of the medial pain system. Remifentanil offset or withdrawal is associated with acute opioid withdrawal hyperalgesia with increased activation in the mesencephalic pontine reticular formation [22]. We previously reported that naloxone alone, in contrast to its effect on nalbuphine-induced enhancement in pain circuits in the present study, increased brain activation induced by noxious thermal stimulation [9]. Thus, the brain appears to contain circuitry necessary for pain facilitation, a phenomenon distinct from pain induction or transmission, as nalbuphine alone does not induce pain.

Direct interventions in pain pathways can modulate nociceptive responses mediated by descending pathways. These include both non-opioid and opioid-based manipulations. Calejesan and colleagues [34] showed that either electrical stimulation or microinjection of a metabotropic glutamate receptor agonist into ACC facilitated the nociceptive tail-flick reflex (i.e., enhanced nociception), and these effects were mediated through the rostral ventral medulla (RVM). In contrast, electrical stimulation of the ACC reduces responses of dorsal horn neurons to noxious mechanical stimuli, suggesting an analgesic effect [35]. Although the findings of these studies disagree in terms of polarity, they agree that manipulation in the ACC, an important region in pain processing, can influence nociceptive responses at the level of the spinal cord. Moreover, there is evidence of connectivity between the ACC and the descending pain modulation system. Neural afferent projections from ACC to the periaqueductal gray (PAG) have been reported [36], [37], and functional connectivity based on fMRI data from 100 subjects has been reported between ACC, PAG, and RVM [38]. Interestingly, this study, which included equal numbers of males and females, also found sexually dimorphic functional connectivity at the mid-cingulate cortex. Others have reported sex differences in brain processing in a number of regions, including the perigenual cingulate cortex, in a study of equal pain experience induced by laser stimulation [39]. Whether these differences contribute to the sex differences in analgesic efficacy observed in our clinical studies [1], [2], [3], [4] remains to be investigated.

Limitations

A few caveats of this study need to be considered. (i) Although 12 subjects per cohort is a reasonable number for phMRI studies [40], smaller effects are probably not statistically robust enough to be detected. (ii) We did not observe significant differences in functional connectivity, the results probably were affected by the use of whole anatomical areas, rather than specific areas associated with significant differences. In order to use ROIs defined by statistical differences a different cohort of subjects would have been needed at it is not statistically correct to use the same cohorts to derive an ROI for secondary functional connectivity analysis. (iii) Prolonged effects on brain systems by low single dose naloxone is futher supported by prolonged (for hours) opioid receptor binding (as measured by [11C]-carfentanil displacement by naltrexone) [41]. In addition in our prior study of naloxone infusions in healthy men (albeit at higher doses), effects on brain system activation was observed within 10 minutes following the infusion and later (i.e., >20 min following the infusion) subjective measures of pain intensity and unpleasantness revealed significant differences even though no significant effects on BOLD were observed approximately 10 minutes after the initial infusion [9]. (iv) We did not perform a dose-response/multiple dose study across healthy subjects and patients that may provide insights into differences in response between the two groups. Such phMRI studies would help differentiate the effects of clinical pain on the phMRI signal. (v) Future studies evaluating effects in healthy and clinical groups of men and women would be needed to further dissect the pharmacological and psychometric differences across gender.

Conclusions

This study provides support for the use of phMRI in our understanding of the mechanism of action of CNS-acting analgesics and the development of more selectively targeted agents. The brain activity signature of nalbuphine demonstrates characteristics typical of both noxious stimulation and the opioid analgesic morphine, thus providing a neural basis for its paradoxical ability to produce both analgesia and pain-facilitation. Naloxone selectively blocks the actions of nalbuphine in brain regions associated with pain, leaving the analgesic-like actions intact. Given these findings, and that nalbuphine alone does not produce pain, it is possible that nalbuphine interacts with a pain salience system, which can modulate perceived pain intensity relative to the nature of the initial insult.

Although nalbuphine is the κ-agonist antagonist we have studied most extensively [2], [3], [4], we have shown that the other two clinical agents in this class, pentazocine and butorphanol, behave similarly [1], [2], strongly implying similar mechanisms of action of this class of opioid drugs. All three drugs in this class are known to act at multiple receptors; therefore, it will be important to identify the specific receptors mediating pain enhancement compared to those that mediate analgesia, in order to develop more efficacious analgesic treatments.

Author Contributions

Conceived and designed the experiments: RG LB JL DB. Performed the experiments: JU JB DB. Analyzed the data: LB JB DW GP. Wrote the paper: RG JU JL DB.

References

  1. 1. Gordon NC, Gear RW, Heller PH, Paul S, Miaskowski C, et al. (1995) Enhancement of morphine analgesia by the GABAB agonist baclofen. Neuroscience 69: 345–349.
  2. 2. Gear RW, Miaskowski C, Gordon NC, Paul SM, Heller PH, et al. (1996) Kappa-opioids produce significantly greater analgesia in women than in men. Nat Med 2: 1248–1250.
  3. 3. Gear RW, Miaskowski C, Gordon NC, Paul SM, Heller PH, et al. (1999) The kappa opioid nalbuphine produces gender- and dose-dependent analgesia and antianalgesia in patients with postoperative pain. Pain 83: 339–345.
  4. 4. Gear RW, Miaskowski C, Gordon NC, Paul SM, Heller PH, et al. (2000) Action of naloxone on gender-dependent analgesic and anti-analgesic effects of nalbuphine in humans. J Pain 1: 122–127.
  5. 5. Schmidt WK, Tam SW, Shotzberger GS, Smith DH Jr, Clark R, et al. (1985) Nalbuphine. Drug Alcohol Depend 14: 339–362.
  6. 6. Leslie RA, James MF (2000) Pharmacological magnetic resonance imaging: a new application for functional MRI. Trends Pharmacol Sci 21: 314–318.
  7. 7. Honey G, Bullmore E (2004) Human pharmacological MRI. Trends Pharmacol Sci 25: 366–374.
  8. 8. Becerra L, Harter K, Gonzalez RG, Borsook D (2006) Functional magnetic resonance imaging measures of the effects of morphine on central nervous system circuitry in opioid-naive healthy volunteers. Anesth Analg 103: 208–216.
  9. 9. Borras MC, Becerra L, Ploghaus A, Gostic JM, DaSilva A, et al. (2004) fMRI measurement of CNS responses to naloxone infusion and subsequent mild noxious thermal stimuli in healthy volunteers. Journal of neurophysiology 91: 2723–2733.
  10. 10. Upadhyay J, Anderson J, Schwarz AJ, Coimbra A, Baumgartner R, et al. (2011) Imaging Drugs with and without Clinical Analgesic Efficacy. Neuropsychopharmacology
  11. 11. Gear RW, Gordon NC, Miaskowski C, Paul SM, Heller PH, et al. (2003) Dose ratio is important in maximizing naloxone enhancement of nalbuphine analgesia in humans. Neurosci Lett 351: 5–8.
  12. 12. Zhang Y, Brady M, Smith S (2001) Segmentation of brain MR images through a hidden Markov random field model and the expectation-maximization algorithm. IEEE Trans Med Imaging 20: 45–57.
  13. 13. Jenkinson M, Bannister P, Brady M, Smith S (2002) Improved optimization for the robust and accurate linear registration and motion correction of brain images. Neuroimage 17: 825–841.
  14. 14. Smith SM (2002) Fast robust automated brain extraction. Hum Brain Mapp 17: 143–155.
  15. 15. Pendse G, Borsook D, Becerra L (2009) ADIS - A robust pursuit algorithm for probabilistic, constrained and non-square blind source separation with application to fMRI. ArXiv e-prints: ARXIV. pp. 44.
  16. 16. Pendse GV, Schwarz AJ, Baumgartner R, Coimbra A, Upadhyay J, et al. (2010) Robust, unbiased general linear model estimation of phMRI signal amplitude in the presence of variation in the temporal response profile. J Magn Reson Imaging 31: 1445–1457.
  17. 17. Woolrich MW, Behrens TEJ, Beckmann CF, Jenkinson M, Smith SM (2004) Multilevel linear modelling for FMRI group analysis using Bayesian inference. Neuroimage 21: 1732–1747.
  18. 18. Smith SM, Miller KL, Salimi-Khorshidi G, Webster M, Beckmann CF, et al. (2011) Network modelling methods for FMRI. Neuroimage 54: 875–891.
  19. 19. Pendse G, Borsook D, Becerra L (2009) Enhanced false discovery rate using Gaussian mixture models for thresholding fMRI statistical maps. Neuroimage 47: 231–261.
  20. 20. Besag J (1986) On the statistical analysis of dirty pictures. J R Stat Soc Ser B Meth 48: 259–302.
  21. 21. Burstein R, Jakubowski M, Garcia-Nicas E, Kainz V, Bajwa Z, et al. (2010) Thalamic sensitization transforms localized pain into widespread allodynia. Ann Neurol 68: 81–91.
  22. 22. Wanigasekera V, Lee MCH, Rogers R, Hu P, Tracey I (2011) Neural correlates of an injury-free model of central sensitization induced by opioid withdrawal in humans. J Neurosci 31: 2835–2842.
  23. 23. Oertel BG, Preibisch C, Martin T, Walter C, Gamer M, et al. (2011) Separating brain processing of pain from that of stimulus intensity. Hum Brain Mapp
  24. 24. Moulton EA, Schmahmann JD, Becerra L, Borsook D (2010) The cerebellum and pain: passive integrator or active participator? Brain Res Rev 65: 14–27.
  25. 25. Borsook D, Upadhyay J, Chudler EH, Becerra L (2010) A key role of the basal ganglia in pain and analgesia–insights gained through human functional imaging. Mol Pain 6: 27.
  26. 26. Craig AD (2010) The sentient self. Brain structure & function 214: 563–577.
  27. 27. Taylor KS, Seminowicz DA, Davis KD (2009) Two systems of resting state connectivity between the insula and cingulate cortex. Human brain mapping 30: 2731–2745.
  28. 28. Garavan H (2010) Insula and drug cravings. Brain structure & function 214: 593–601.
  29. 29. Baumgartner U, Buchholz HG, Bellosevich A, Magerl W, Siessmeier T, et al. (2006) High opiate receptor binding potential in the human lateral pain system. NeuroImage 30: 692–699.
  30. 30. Zelaya FO, Zois E, Muller-Pollard C, Lythgoe DJ, Lee S, et al. (2012) The response to rapid infusion of fentanyl in the human brain measured using pulsed arterial spin labelling. Magma 25: 163–175.
  31. 31. Berkowitz BA (1976) The relationship of pharmacokinetics to pharmacological activity: morphine, methadone and naloxone. Clin Pharmacokinet 1: 219–230.
  32. 32. Derbyshire SWG, Whalley MG, Stenger VA, Oakley DA (2004) Cerebral activation during hypnotically induced and imagined pain. Neuroimage 23: 392–401.
  33. 33. Kong J, Gollub RL, Polich G, Kirsch I, Laviolette P, et al. (2008) A functional magnetic resonance imaging study on the neural mechanisms of hyperalgesic nocebo effect. J Neurosci 28: 13354–13362.
  34. 34. Calejesan AA, Kim SJ, Zhuo M (2000) Descending facilitatory modulation of a behavioral nociceptive response by stimulation in the adult rat anterior cingulate cortex. Eur J Pain 4: 83–96.
  35. 35. Senapati AK, Lagraize SC, Huntington PJ, Wilson HD, Fuchs PN, et al. (2005) Electrical stimulation of the anterior cingulate cortex reduces responses of rat dorsal horn neurons to mechanical stimuli. J Neurophysiol 94: 845–851.
  36. 36. An X, Bandler R, Ongür D, Price JL (1998) Prefrontal cortical projections to longitudinal columns in the midbrain periaqueductal gray in macaque monkeys. J Comp Neurol 401: 455–479.
  37. 37. Floyd NS, Price JL, Ferry AT, Keay KA, Bandler R (2000) Orbitomedial prefrontal cortical projections to distinct longitudinal columns of the periaqueductal gray in the rat. J Comp Neurol 422: 556–578.
  38. 38. Kong J, Tu P-c, Zyloney C, Su T-p (2010) Intrinsic functional connectivity of the periaqueductal gray, a resting fMRI study. Behav Brain Res 211: 215–219.
  39. 39. Derbyshire SWG, Nichols TE, Firestone L, Townsend DW, Jones AKP (2002) Gender differences in patterns of cerebral activation during equal experience of painful laser stimulation. J Pain 3: 401–411.
  40. 40. Upadhyay J, Anderson J, Schwarz AJ, Coimbra A, Baumgartner R, et al. (2011) Imaging drugs with and without clinical analgesic efficacy. Neuropsychopharmacology: official publication of the American College of Neuropsychopharmacology 36: 2659–2673.
  41. 41. Rabiner EA, Beaver J, Makwana A, Searle G, Long C, et al. (2011) Pharmacological differentiation of opioid receptor antagonists by molecular and functional imaging of target occupancy and food reward-related brain activation in humans. Molecular psychiatry 16: 826–835, 785.
  42. 42. Becerra L, Breiter HC, Wise R, Gonzalez RG, Borsook D (2001) Reward circuitry activation by noxious thermal stimuli. Neuron 32: 927–946.