Sex Differences in Regional Brain Glucose Metabolism Following Opioid Withdrawal and Replacement

Abstract

Methadone and buprenorphine are currently the most common pharmacological treatments for opioid dependence. Interestingly, the clinical response to these drugs appears to be sex specific. That is, females exhibit superior therapeutic efficacy, defined as extended periods of abstinence and longer time to relapse, compared with males. However, the underlying metabolic effects of opioid withdrawal and replacement have not been examined. Therefore, using ¹⁸FDG and microPET, we measured differences in regional brain glucose metabolism in males and females following morphine withdrawal and subsequent methadone or buprenorphine replacement. In both males and females, spontaneous opioid withdrawal altered glucose metabolism in regions associated with reward and drug dependence. Specifically, metabolic increases in the thalamus, as well as metabolic decreases in insular cortex and the periaqueductal gray, were noted. However, compared with males, females exhibited increased metabolism in the preoptic area, primary motor cortex, and the amygdala, and decreased metabolism in the caudate/putamen and medial geniculate nucleus. Methadone and buprenorphine initially abolished these changes uniformly, but subsequently produced their own regional metabolic alterations that varied by treatment and sex. Compared with sex-matched control animals undergoing spontaneous opioid withdrawal, male animals treated with methadone exhibited increased caudate/putamen metabolism, whereas buprenorphine produced increased ventral striatum and motor cortex metabolism in females, and increased ventral striatum and somatosensory cortex metabolism in males. Notably, when treatment effects were compared between sexes, methadone-treated females showed increased cingulate cortex metabolism, whereas buprenorphine-treated females showed decreased metabolism in cingulate cortex and increased metabolism in the globus pallidus. Perhaps the initial similarities in males and females underlie early therapeutic efficacy, whereas these posttreatment sex differences contribute to clinical treatment failure more commonly experienced by the former.

Introduction

Opioid abuse is an urgent public healthcare concern. The 2014 National Survey on Drug Use and Health estimated that 4.3 million Americans used prescription pain relievers nonmedically (Center for Behavioral Health Statistics and Quality, 2015). In addition, although they constitute only 5.6% of the world’s populations, Americans consume ∼80% of the global opioid supply, 99% of the global hydrocodone supply, and two-thirds of the world’s illegal drugs (Manchikanti et al, 2010; NIDA, 1991). The surge in opioid abuse has been attributed to factors including the social acceptance of medications, the aggressive culture of pharmaceutical advertising, and the increasing availability of prescription analgesics (Chang et al, 2014; Daubresse et al, 2013). Subsequently, concerns have been raised over drug prescribing patterns (Compton and Volkow, 2006), the social and economic costs of opioid addiction (Connock et al, 2007; McCarty et al, 2010; Polsky et al, 2010; Schackman et al, 2012), as well as the safety (Maremmani and Gerra, 2010), availability (Bonhomme et al, 2012; Nosyk et al, 2013; Novick et al, 2015) and expansion of current treatment options, mainly methadone (METH) and buprenorphine (BUP) replacement (Bonhomme et al, 2012; Kraus et al, 2011; Wesson and Smith, 2010).

METH is an agonist of μ-opioid receptors. These G-protein-coupled receptors modulate dopamine, a catecholamine neurotransmitter that regulates reward-motivated behavior (Spanagel and Weiss, 1999). METH is a lipid-soluble compound, and although its analgesic duration of action is only 4 to 8 h, it has a relatively long half-life of 8 to 59 h and a high oral bioavailability. This allows for continuous receptor occupation and effective opioid maintenance (Ball and Ross, 1991; Dole, 1989; Eap et al, 2002; Ward et al, 1999). However, METH has been used and sold by opioid-dependent individuals seeking to subsidize the cost of illicit drug use (Davis and Johnson, 2008). Therefore, because of its potential for abuse, it must be administered in a supervised setting, thus creating logistic and compliance concerns (Haskew et al, 2008). In addition, METH detoxification can produce severe withdrawal lasting several days to weeks (Eap et al, 2002). In current clinical practice it is titrated to a therapeutic level that prevents withdrawal without requiring increased dosing. However, effective clinical dosing often varies substantially (Trafton et al, 2006). Emerging as an alternative to METH is Suboxone, which is BUP prepared with naloxone. This formulation enables its use as a prescription drug in primary care settings. When taken orally, its partial μ-opioid agonist activity allows for successful opioid replacement and maintenance. However, if Suboxone tablets are crushed, dissolved, and injected or insufflated, the antagonistic effects of naloxone in the preparation causes immediate opioid withdrawal, thereby limiting recreational diversion. This formulation has gained popularity in clinical detoxification settings because of its less intense withdrawal severity (Ling et al, 1996), but although Suboxone is effective for treating moderate opioid dependence, high levels of dependence usually require METH (Johnson et al, 2000).

Although these pharmacotherapies represent first-line treatments for opioid withdrawal, their efficacies are variable (Johnson et al, 2000). This is thought to be due in part to personal motivation and treatment adherence. However, there is an understudied observation that sex significantly affects therapeutic response, which is often measured as time to relapse. A 25-year follow-up of heroin-dependent patients prescribed METH indicated that surviving women were significantly more likely than men to have stopped heroin use (Jimenez-Trevino et al, 2011). Similarly, a 7-year follow-up study of heroin-dependent patients treated with BUP showed that, again, women were significantly more likely than men to have stopped heroin use (Ohlin et al, 2015). Finally, Sheynin et al (2016) observed that only males receiving replacement therapy (METH or BUP) demonstrated abnormal avoidance behavior.

Differences in rates of substance abuse and dependence have long been reported. Specifically, women are generally considered more susceptible to drug cravings and relapse (Fox et al, 2014; Hitschfeld et al, 2015; Kennedy et al, 2013; Kippin et al, 2005; Robbins et al, 1999; Rubonis et al, 1994). Men are more likely to abuse illicit drugs, but women are just as likely to develop drug dependence (Anthony et al, 1994). Despite these findings, there continues to be a marked paucity of studies focused on the development of potential sex-directed pharmacological interventions.

In an effort to better understand these reported sex-specific differences in treatment efficacy, we imaged adolescent male and female animals using micro-positron emission tomography (microPET) in combination with ¹⁸F-fluorodeoxyglucose (¹⁸FDG) following morphine cessation and subsequent METH or BUP replacement. For this study, BUP was chosen rather than Suboxone so as to not potentially confound the metabolic representation of the primary treatment modality. Adolescent animals were chosen as this age group is generally associated with incipient drug use, lifetime developmental consequences of drug exposure (Vassoler et al, 2014), as well as gestational complications following opioid abuse (Ross et al, 2015). Each of these factors may be due, at least in part, to neuronal pairing, incomplete maturation of frontal cortices, and elevated levels of brain glucose metabolism, all of which are hallmarks of the adolescent brain. We postulated that sex likely influences neural pathways associated with reward and addiction. Specifically, we hypothesized that male and female rats would respond uniquely to opioid withdrawal and replacement as represented by regional changes in brain glucose metabolism.

Materials and methods

Figure 1 provides a concise summary and timeline of the methods used in this study, as well as the number of animals per group. Drug-naive adolescent male and female Sprague-Dawley rats from Taconic Farms arrived on postnatal day (PND) 22. Animals were maintained on a 12 h light–dark cycle and received food and water ad libitum. Before each scan, animals were fasted for 12 h to ensure blood glucose stability (Fueger et al, 2006; Wong et al, 2011). Following an acclimation period (9 days), all animals received baseline ¹⁸FDG microPET scans on PND 31 (scan 1). On PND 35, animals commenced morphine treatment for 5 days at a dose of 10 mg/kg/day subcutaneously. This dose was selected based on data indicating that 10 mg/kg was adequate to achieve conditioned place preference within this time period (Lu et al, 2005; Raghavendra et al, 2004), and that a single dose was able to elicit conditioned place avoidance after a naloxone challenge (Araki et al, 2004). Morphine administration at this dose for 5 days also produced analgesic tolerance (Beaudry et al, 2009), and after only 4 days, produced withdrawal behaviors including increased defecation, urination, salivation, jumping, and wet dog shakes (Pinelli et al, 1997). Finally, this dosing schedule activated glial cells and enhanced proinflammatory cytokine expression in the spinal cord that has been implicated in morphine tolerance and withdrawal-induced hyperalgesia (Raghavendra et al, 2004).

Figure 1.

Experimental timeline. This flowchart provides a summary of the experimental timeline. From top to bottom, the progression of ¹⁸FDG microPET scans and treatment with morphine, methadone (METH), and buprenorphine (BUP) is delineated by postnatal day (PND). Scan 1 refers to the baseline (pretreatment) imaging time point. Scan 2 refers to images obtained after spontaneous withdrawal (control), subchronic METH replacement, and subchronic BUP replacement. Study group drug dosing, treatment duration, and sample size are noted in the figure.

Following day 5 of morphine treatment and 2 days of spontaneous withdrawal (PND 40–41), animals were divided into three treatment groups: (1) saline control, (2) METH, or (3) BUP. The mean baseline bodyweights for the saline control, METH, and BUP groups were 107.3, 127.3, and 109.3 g, respectively. Control animals received subcutaneous volume-matched saline injections and continued on in spontaneous withdrawal for 5 days (PND 42–46). The remaining two groups received either METH (1 mg/kg/day) or BUP (0.1 mg/kg/day) subchronically for the same 5 days. Following day 5 of opioid replacement treatment, and 3 days of washout (PND 47–49), ¹⁸FDG scans (scan 2) were again obtained (PND 50).

All images were acquired using a Siemens Inveon microPET. Before scanning, each animal received a single dose of ¹⁸FDG (1.8–2.0 mCi) via an intraperitoneal injection. The ¹⁸FDG doses used are consistent with both intraperitoneal and intravenous doses reported previously using both rats/mice and microPET (Chen et al, 2010; Parthoens et al, 2014; Yang et al, 2014). Perhaps most importantly, however, ¹⁸FDG doses were specifically designed to produce count rates that did not exceed the dead time correction capabilities of our scanner and images that could be reconstructed using an iterative method (ie, maximum a posteriori (MAP)). Relative to body weight, ¹⁸FDG is injected at significantly higher doses in rodents than in humans. These higher doses are necessary in order to achieve both sufficient counting statistics and maximal spatial resolution in these physically smaller rodent brains (Hildebrandt et al, 2008). In addition, published reports have established that higher doses are required for equivalent quality in these images. Specifically, Jagoda et al (2004) determined that roughly the same total amount of radiopharmaceutical used in humans should be used in rodents.

After radiotracer injection, animals were returned to their home cage and left undisturbed for 40 min to ensure radiotracer uptake. Afterwards, animals were transferred to a clear acrylic chamber, where isoflurane/oxygen was used to induce anesthesia. At 5 min after induction, each animal was transferred to the imaging platform and was secured. Continuous isoflurane/oxygen at 2–2.5% was administered via nasal cannula for the entire 10 min static scan. These imaging protocols have previously been shown to adequately reflect glucose metabolism in rodents (Marsteller et al, 2006; Patel et al, 2008; Schiffer et al, 2007).

All microPET images were corrected for attenuation and reconstructed using a MAP probability estimate with 20 iterations as described previously (Schiffer et al, 2009; Vo et al, 2014). Raw data files were uploaded into Pixel-wise Modeling Tool software (PXMOD version 3.3, PMOD Technologies LLC), and were aligned to a reference template created using the Paxinos and Watson Sprague-Dawley rat brain atlas. After being placed in anatomical space, images were corrected for injected dose to ensure that regional uptake values would be comparable and were skull-stripped to eliminate extraneous metabolic activity (Schiffer et al, 2006, 2007). Statistical Parametric Mapping (SPM5, Wellcome Trust Centre for Neuroimaging) was used for subsequent postprocessing steps including realignment to an atlas, normalization to a mean template, and smoothing. Between- and within-group comparisons were carried out using paired and two-sample T-tests, respectively. Postprocessed images were aligned to the Paxinos and Watson rat brain atlas (Paxinos et al, 1980) and significant regions were identified using x, y, and z coordinates. Increases and decreases in relative brain metabolism were visually represented using color mapping. Images were overlaid onto an anatomical cryostat template with metabolic increases set as hot, and metabolic decreases set as winter in the color selection options. The color scale used represents all T distributions achieving statistical significance (Carrion et al, 2009; Nie et al, 2014; Soto-Montenegro et al, 2009). All corresponding brain areas are significant at a value of p⩽0.001 (corrected) with a cluster-extent threshold of k=0 voxels. FSL (FMRIB Software Library, Oxford, UK) was used for extraction of significant brain regions using threshold values obtained in SPM (Jenkinson et al, 2012).

Results

Both longitudinal and cross-sectional comparisons within and between groups, respectively, were used to uncover metabolic alterations due to treatment and sex (Table 1). In reality, there are numerous interactions that could be examined. In this study, however, we chose to target our analysis to ¹⁸FDG microPET scans acquired at baseline and after treatment. Evaluation at these time points represents an examination of the effects of subchronic opioid exposure and replacement and, therefore, may be clinically relevant to the earliest stages of opioid addiction.

Table 1. Summary of Group Comparisons and Significant Brain Regions.

Description	Figure 2	Scan pair	Brain region	Change	ML	AP	DV
(A) Treatment effects
Effect of spontaneous opioid withdrawal	A1	Saline scan 1 vs saline scan 2	Angular thalamic nuclei	Increase	−1.6	−6	−6.6
			Dysgranular insular cortex	Decrease	−4.2	3.2	−5.2
			Agranular insular cortex	Decrease	4	3.2	−6
			Ventrolateral periaqueductal gray	Decrease	−0.6	−8	−6
Within-group effect of METH replacement	A2	METH scan 1 vs METH scan 2	Ventromedial thalamic nuclei	Increase	1	−3.2	−7.2
			Agranular insular cortex	Decrease	−3.6	3.4	−5.8
Within-group effect of BUP replacement	A3	BUP. scan 1 vs BUP. scan 2	Ventral posterior thalamic nucleus	Increase	2.6	−4.4	−6.4
			Anterior amygdaloid	Increase	−3.6	−1.6	−6.4
			Primary somatosensory cortex	Increase	−4.6	−1.2	−2.4
			Hippocampus (radiatum, CA3)	Decrease	−5	−4.6	−5.2
			Hippocampus (radiatum layer)	Decrease	4.8	−4.8	−6.4
			Reticular thalamic nucleus	Decrease	1.8	−1.4	−7
			Caudomedial entorhinal cortex	Decrease	−4.8	7.6	−7.4
Between-group effect of METH replacement	A4	Saline scan 2 vs METH scan 2	Ventral striatum	Increase	−2.4	0.8	−8.2
			Hippocampus (CA1/CA2)	Increase	4.2	−4.4	−7.8
Between-group effect of BUP replacement	A5	Saline scan 2 vs BUP scan 2	Ventral striatum	Increase	−2.4	1.2	−8.2
			Ventral striatum	Increase	1.4	1.6	−7.8
Between-treatment effect (metabolic changes in BUP relative to METH)	A6	METH scan 2 vs BUP scan 2	Dorsal raphe nucleus	Increase	0	−8.8	−7
			Medial geniculate nucleus	Increase	3	−5.4	−6.2
			Hippocampus (radiatum layer)	Decrease	5	−4.8	−7.2
(B) Within-sex treatment effects
Effect of METH replacement on female rats compared with spontaneous withdrawal	B7	Saline scan 2 female vs METH scan 2 female	Reticular thalamic nucleus	Increase	1.8	−1.4	−6.8
Effect of METH replacement on male rats compared with spontaneous withdrawal	B8	Saline scan 2 males vs METH scan 2 males	Caudate/putamen	Increase	5	−0.8	−5.6
			Caudate/putamen	Increase	−4.6	−0.8	−5.6
			Caudate/putamen (anterior)	Increase	3.8	1.6	−4.8
			Caudate/putamen (anterior)	Increase	−3.4	1.8	−4.4
Effect of BUP replacement in female rats compared with spontaneous opioid withdrawal	B9	Saline scan 2 females vs BUP scan 2 females	Primary motor cortices	Increase	2.4	0.4	−1.8
			Ventral striatum	Increase	0.8	1.8	−7.6
Effect of BUP replacement in male rats compared with spontaneous opioid withdrawal	B10	Saline scan 2 males vs BUP scan 2 males	Ventral striatum	Increase	1.8	1.6	−7.8
			Primary somatosensory cortex	Increase	−4.6	0.2	−2.6
			Dorsolateral entorhinal cortex	Decrease	−5.8	−8.2	−7.2
			Dorsal entorhinal cortex	Decrease	−6.8	−7.4	−6.4
(C) Between-sex treatment effects
Effect of female sex on METH replacement	C11	METH scan 2 males vs METH scan 2 females	Cingulate cortex	Increase	0.4	−0.4	−3.2
Effect of female sex on BUP replacement	C12	BUP scan 2 males vs BUP scan 2 females	Globus pallidus	Increase	3.4	−1.2	−7.2
			Cingulate cortex	Decrease	−0.4	−0.6	−2.4

This table provides a summary and description of microPET scan group comparisons organized by (A) treatment effects, (B) within-sex treatment effects, and (C) between-sex treatment effects. METH, BUP, and control (saline) group comparisons are presented. Scan 1 refers to group baseline scans (pretreatment). Scan 2 refers to the subchronic opioid replacement treatment scans or spontaneous withdrawal scans in the case of saline controls. Brain regions with significant metabolic increases and decreases in glucose metabolism are listed, as well as corresponding medial–lateral (ML), anterior–posterior (AP), and dorsal–ventral (DV) stereotaxic coordinates. All reported increases and decreases are significant at a strict p-value threshold of p⩽0.001 (corrected) with a cluster-extent threshold of k=0 voxels. A legend of corresponding images from Figure 2 is also shown. The metabolic changes reported are relative to the scan order in the scan pair column. For example, in the first row, saline scan 1 vs saline scan 2, the metabolic changes reported are the results of scan 2, relative to baseline scan 1. The description column is provided to summarize the interpretation of each scan pair.

Specific scan comparisons were utilized to assess (A) treatment effects, (B) within-sex treatment effects, and (C) between-sex treatment effects. Table 1 summarizes all group comparisons, lists significant increases and decreases in regional brain glucose metabolism, and provides a legend for corresponding microPET images highlighted in Figure 2. All reported increases and decreases are significant at a strict p-value threshold of p⩽0.001 (corrected) with a cluster-extent threshold of k=0 voxels. These constraints were chosen based on previous recommendations against liberal primary cluster extent thresholds. By eliminating large activations in overlapping anatomical areas, these parameters ensure the statistical validity of the reported regions of interest (Woo et al, 2014).

Figure 2.

¹⁸FDG microPET images highlighting significant brain regions. Coronal slices of (A) baseline brain metabolism between males and females, (B) changes in brain metabolism following acute withdrawal, and (C) changes in male brain metabolism. Slice distance in millimeters from bregma is noted in the upper right hand corner of each image. Significant increases and decreases in regional brain glucose metabolism are visually represented using hot (red–yellow) and winter (blue–green) color maps, respectively. MAX and MIN refer to the degree of regional radioisotope decay (percent injected dose per gram). The color scale used represents all T distributions achieving statistical significance. All corresponding brain areas are significant at a value of p⩽0.001 (corrected) with a cluster-extent threshold of k=0 voxels.

Spontaneous opioid withdrawal produced changes in both cortical and subcortical brain metabolism (Figure 2, A1). These bilateral changes were noted in the agranular thalamic nuclei (increase), insular cortex (decrease), and periaqueductal gray area (decrease). When disaggregated by sex, several differences were noted. Metabolism in the lateral preoptic area, primary motor cortex, and medial amygdaloid nucleus increased in females compared with males, whereas caudate nucleus, putamen, and medial geniculate nucleus metabolism decreased (images not shown). Subchronic treatment with METH or BUP abolished these withdrawal-associated changes in both sexes. However, both drugs increased ventral striatum metabolism (Figure 2, A4 and A5), consistent with their known effects on reward pathways. In addition, METH produced increased hippocampal metabolism (Figure 2, A4 and A6) and decreased insular cortex metabolism (Figure 2, A2), changes not seen in animals treated with BUP. METH and BUP also produced sex-specific changes that varied by drug. METH increased thalamic metabolism in females (Figure 2, B7) and caudate/putamen metabolism in males (Figure 2, B8). BUP increased motor cortex metabolism in females (Figure 2, B9), but decreased entorhinal cortex metabolism in males (Figure 2, B10). BUP also led to activation of the ventral striatum in both females (Figure 2, B9) and males (Figure 2, B10), compared with control subjects undergoing spontaneous opioid withdrawal. Finally, compared with males, females treated with METH expressed increased cingulate cortex metabolism (Figure 2, C11), whereas females treated with BUP expressed decreased cingulate metabolism and a concomitant increase in globus pallidus metabolism (Figure 2, C12).

Discussion

Opioid withdrawal produced both similar and unique alterations in regional brain metabolism in male and female animals. Specifically, similar changes were observed in the thalamus, insular cortex, and the periaqueductal gray. However, compared with males, female exhibited increased metabolism in the preoptic area, primary motor cortex, and amygdala, but decreased metabolism in the caudate/putamen and medial geniculate nucleus. Finally, methadone and buprenorphine abolished these changes, yet each produced their own regional metabolic alterations that varied by treatment and sex.

The animal model used in the present study produced distinct metabolic indicators of opioid withdrawal. Opioid withdrawal resulted in specific metabolic patterns in brain regions associated with sensory processing, salience modulation, reward, and memory. These changes are consistent with earlier reports using electrophysiological and behavioral techniques. Zhu et al (2016) recently demonstrated that activity in thalamic projections to the nucleus accumbens mediates behavioral aversion. Furthermore, using muscimol and morphine, Silva and Nobre (2014) demonstrated that GABA and opioid receptors of the periaqueductal gray impact the expression of unconditioned and conditioned fear responses in animals experiencing alcohol withdrawal. Interestingly, although brain lesions to the insula seem to interrupt addictive behaviors, functional neuroimaging studies show that addictive behaviors are associated with reduced insular cortex activity, a disparity that might be explained by regional heterogeneity (Droutman et al, 2015). Finally, opioid withdrawal produced increases in thalamic cyclic AMP, which is thought to play a role in the behavioral physiology of withdrawal (Sadava and Mack, 1986). Despite these data, our knowledge of the neural circuitry involved in opioid withdrawal remains incomplete. However, recent studies have shown that sex likely influences these pathways, and may actually affect treatment outcomes (Jimenez-Trevino et al, 2011; Ohlin et al, 2015).

Although we observed regional metabolic changes across sex, other alterations were unique to either males or females. Subchronic treatment with METH or BUP abolished these effects. It is important to note, however, that these drugs also produced their own sex-specific changes. Our findings are supported by previous studies revealing clinical and behavioral sex differences in opioid analgesia in the presence of pain and addiction (Becker and Hu, 2008; Craft, 2003; Fillingim and Gear, 2004; Terner et al, 2003). In line with these data, sex appears to alter the expression of spontaneous withdrawal, with males experiencing increased severity and length of withdrawal (Cicero et al, 2002). This might be explained by differences in receptor density as described by Vijay et al (2016), who recently reported that in a group of normal volunteers, males had a greater volume of distribution of κ-opioid receptors than females. The κ-opioid receptors in the accumbens shell have been shown to mediate aversive social motivation (Resendez et al, 2012). These findings might be related to neuroanatomical sex differences in the nucleus accumbens core and shell (Forlano and Woolley, 2010). As both METH and BUP bind to this receptor subtype, these findings could have clinical implications regarding the successful treatment of opioid withdrawal. Together, these data suggest that males and females respond to opioids differently. Therefore, they should be studied separately and managed uniquely in an effort to optimize treatment efficacy. Our data further support this notion.

Here, we observed that METH and BUP abolished the regional metabolic changes measured following spontaneous morphine withdrawal, regardless of sex. Interestingly, females expressed variable metabolism in the cingulate gyrus and increased metabolism within the globus pallidus following treatment. Furthermore, males exposed to BUP demonstrated decreases in entorhinal cortex metabolism compared with females. Frenois et al (2005) found that withdrawal memories drive neuronal activity in communicating limbic areas with known involvement in aversive motivational processes. Decreased μ-opioid neurotransmission in limbic and paralimbic circuits seems to correlate with negative affect (Zubieta et al, 2003), an emotional state that induces hyperalgesia in heroin withdrawal (Carcoba et al, 2011). Our data suggest that females preferentially activate limbic structures, regardless of drug treatment, compared with males, who deactivate limbic structures when treated with BUP. Perhaps the metabolic similarities observed following spontaneous withdrawal or opioid replacement produce the initial treatment efficacy experienced by both sexes, whereas the differences detected underlie the treatment failure more commonly observed in males.

Sanchis-Segura and Becker (2016) recently suggested that studies should examine both sex differences and sex similarities, so as to build a more accurate profile of male and female neurobiology and neurophysiology in drug-dependent states. In addition, in order to advance our understanding of drug dependence, studies should adopt a multidisciplinary approach, one that goes beyond the explanation of biological responses to opioids. It is important to not only consider the natural course of addiction between sexes, but also the comorbid psychiatric and sociocultural conditions that characterize males and females as well (Buccelli et al, 2016).

The purpose of this preliminary study was to specifically observe regional changes in brain glucose metabolism following opioid withdrawal and replacement. Thus, no formal assessment of locomotor activity or withdrawal behavior was assessed. Pairing this imaging strategy with verified behavioral measures may ultimately be useful in the development of effective pharmacological interventions. This experiment exemplifies how small animal imaging in combination with suitable animal models of opioid dependence and withdrawal may provide an experimental bridge between preclinical studies and human trials.

MicroPET and other imaging modalities offer insight into pharmacokinetics, central nervous system penetration, and dosing that can help accelerate drug development (Pien et al, 2005). The rodent model used in this experiment can be adapted to the study of other novel treatments for opioid abuse. These include BUP implants (Ling et al, 2010; Rosenthal et al, 2013), BUP–gabapentin combination therapy (Sanders et al, 2013), memantine–naltrexone therapy (Bisaga et al, 2011, 2014), slow-release morphine treatment (Jegu et al, 2011), herbal and plant remedies (Gao et al, 2014; Tabatabai et al, 2014), as well as potential immunotherapies including conjugate morphine–heroin vaccines (Li et al, 2011, 2015). Given the abundance of studies on therapeutic intervention, as well as epidemiologic data indicating escalating levels of opioid abuse, additional investigations into existing and novel treatments are warranted.

Current opioid replacement and maintenance therapies are initially effective in both sexes. Unfortunately, over time, they fail more commonly in males. Here, we report that spontaneous opioid withdrawal produced similar regional metabolic changes in males and females. METH or BUP exposure attenuated these changes equally. However, each drug produced unique sex-specific metabolic changes of its own. This mutual attenuation may underlie that period of time when therapeutic efficacy appears similar in both sexes. Concomitantly, these sex-specific differences may contribute to the long-term treatment failure often experienced by males.

These studies were conducted in the service of better understanding the impact of opioid withdrawal and treatment on regional brain glucose metabolism in males and females. These data suggest that novel sex-directed pharmacologic strategies may better serve this rapidly growing patient population. As a result, we may ultimately enhance the quality of life of those currently suffering from opioid dependence, while simultaneously protecting individuals from the potential consequences of its likely progression.

Funding and disclosure

The authors declare no conflict of interest.