Social Pain Revisited: Opioids for Severe Suicidal Ideation

Does the pain of mental anguish rely on the same neural machinery as physical pain? Can we treat these dreaded ailments with the same medications? These issues have come to the fore in the field of social/cognitive/affective neuroscience.

As many readers know, Lieberman and Eisenberger (2015) recently published a controversial paper claiming that a brain region called the dorsal anterior cingulate cortex (dACC, shown above) is “selective” for pain.¹ This finding fits with their long-time narrative that rejection literally “hurts” — social pain is analogous to physical pain, and both are supported by activity in same regions of dACC (Eisenberger et al., 2003). Their argument is based on work by Dr. Jaak Panksepp and colleagues, who study separation distress and other affective responses in animals (Panksepp & Yovell, 2014).

Panksepp wrote The Book on Affective Neuroscience in 1998, and coined the term even earlier (Panksepp, 1992). He also wrote a Perspective piece in Science to accompany Eisenberger et al.'s 2003 paper:

We often speak about the loss of a loved one in terms of painful feelings, but it is still not clear to what extent such metaphors reflect what is actually happening in the human brain? Enter Eisenberger and colleagues ... with a bold neuroimaging experiment that seeks to discover whether the metaphor for the psychological pain of social loss is reflected in the neural circuitry of the human brain. Using functional magnetic resonance imaging (fMRI), they show that certain human brain areas that “light up” during physical pain are also activated during emotional pain induced by social exclusion [i.e., exclusion from playing a video game].

But as I've argued for years, Social Pain and Physical Pain Are Not Interchangeable. Whenever I read an article proclaiming that “the brain bases of social pain are similar to those of physical pain”, I am reminded of how phenomenologically DIFFERENT they are.

And subsequent work has demonstrated that physical pain and actual social rejection (a recent romantic break-up) do not activate the same regions of dACC (Woo et al., 2014). Furthermore, multivariate activation patterns across the entire brain can discriminate pain and rejection with high accuracy.² 

Modified from Fig. 3 (Woo et al., 2014). Differences between fMRI pattern-based classifiers for pain and rejection.

Feelings of rejection were elicited by showing the participants pictures of their ex-partners (vs. pictures of close friends), and physical pain was elicited by applying painful heat to the forearm (vs. warm heat).

Does this mean there is no overlap between brain systems that can dampen physical and emotional pain (e.g., endogenous opioids)? Of course not; otherwise those suffering from utter despair, unspeakable loneliness, and other forms of psychic turmoil would not self-medicate with mind-altering substances.


Separation Distress: Of Mice and Psychoanalysis

Although Panksepp has worked primarily with rodents and other animals throughout his career, he maintains a keen interest in neuropsychoanalysis, an attempt to merge Freudian psychoanalysis with contemporary neuroscience. Neuropsychoanalysis “seeks to understand the human mind, especially as it relates to first-person experience.” If you think that's a misguided (and impossible) quest, you might be surprised by some of the prominent neuroscientists who have signed on to this agenda (see these posts).

Prof. Panksepp is currently collaborating with Prof. Yoram Yovell, a Psychoanalyst and Neuroscientist at the Institute for the Study of Affective Neuroscience (ISAN) in Haifa. A recent review paper addresses their approach of affective modeling in animals as a way to accelerate drug development in neuropsychiatry (Panksepp & Yovell, 2014). Their view is that current models of depression, which focus on animal behaviors instead of animal emotions, have hindered new breakthroughs in treatments for depression. It’s actually a fascinating and ambitious research program:

We admit that our conceptual position may be only an empirical/ontological approximation, especially when contrasted to affective qualia in humans … but it is at least a workable empirical approach that remains much underutilized. Here we advance the view that such affective modeling can yield new medical treatments more rapidly than simply focusing on behavioral processes in animals. In sum, we propose that the neglect of affect in preclinical psychiatric modeling may be a major reason why no truly new psychiatric medicinal treatments have arisen from behavior-only preclinical modeling so far.

They propose that three key primal emotional systems³ may be critical for understanding depression: SEEKING (enthusiasm-exuberance), PANIC (psychic pain), and PLAY (joyful exuberance). If these constructs sound highly anthropormorphic when applied to rats, it's because they are!! Perhaps you'd rather “reaffirm classical behaviorist dogma” (Panksepp & Yovell, 2014) and stick with more traditional notions like brain reward systems, separation distress, and 50-kHz ultrasonic vocalizations (e.g., during tickling, mating, and play) when studying rodents.

Of interest today is the PANIC system (Panksepp & Yovell, 2014), which “mediates the psychic pain of separation distress (i.e. excessive sadness and grief), which can be counteracted by minimizing PANIC arousals (as with low-dose opioids).” Since low-dose opioids alleviate separation distress in animals (based on reductions in distress vocalizations), why not give them to suicidal humans suffering from psychic pain?

Well... because making strong inferences about the contents of animal minds is deeply problematic (Barrett et al., 2007). I've written about some of the problems with animal models of dread and despair. One might also question whether it's wise to give opioid drugs (even in very low doses) to severely ill people.


Low-Dose Buprenorphine for Suicidal Ideation

Recently investigators are increasingly entertaining the possibility of using ‘safe opioids’ for the treatment of depression, as well as the chronic ‘psychological pain’ that often promotes suicidal ideation. To be a ‘safe opioid’, the analgesic effects and the lethal (respiratory depression) effects of a particular opioid ligand need to be dissociated. Buprenorphine, a partial agonist at μ-opioid receptors (i.e. stimulating opioid receptors at low doses, but blocking them at high doses), is just such a drug.

Panksepp and Lovell's ideas led to a clinical trial (A Study of Nopan Treatment of Acute Suicidality) and a new paper in the American Journal of Psychiatry (Yovell et al., 2015). Nopan is sublingual buprenorphine hydrochloride 0.2 mg. At higher doses, buprenorphine is used as a treatment for opioid addiction, much like methadone.

Research on suicidal behavior is an important and tragically neglected topic, and many clinicians, organizations, and industry sponsors are reluctant to engage. So it's notable that the current study was funded by the Neuropsychoanalysis Foundation (which awards grants and sponsors the journal Neuropsychoanalysis), the Hope for Depression Research Foundation (whose Board is filled with some Heavy Hitters of Neuroscience – e.g., Akil, Mayberg, McEwen, Nestler, Hen), and ISAN.

It's interesting to track some of the changes in the study protocol and description over time. The initial ClinicalTrials.gov entry (dated 2010_01_11) dropped its psychoanalytic language on 2011_05_23:

The acutely suicidal patient presents a complex and dangerous clinical dilemma. Many suicidal patients receive antidepressant medications, but the onset of action of these medications is at least three weeks, and despite their established antidepressant effect, they have not shown a clear anti-suicidal benefit. Psychoanalysts hypothesized that depression (often leading to suicidality) shares important characteristics with the psychological sequelae of object loss and separation distress. Endogenous opioids (endorphins) have been implicated in mediating social bonding and separation distress in mammals. 

On the same date, the Secondary Outcome Measure (Reduction in psychache as measured by the Holden Psychache Scale) was replaced by a more standard and non-psychoanalytic instrument, the Beck Depression Inventory (Reduction in depression as measured by the BDI). Dr. Beck conceptualized depression in a cognitive framework.

On the other hand, “psychache” (coined by suicidologist Dr. Edwin Shneidman) means “unbearable psychological pain—hurt, anguish, soreness, and aching. ... Psychache stems from thwarted or distorted psychological needs . . . every suicidal act reflects some specific unfulfilled psychological need.”  Many of these views are at odds with neuropsychiatry (Schneidman, 1993):

Depression seems to have physiological, biochemical, and probably genetic components. The use of medications in treatment is on target. [so far so good] ... Suicide, on the other hand, is a phenomenological event... It is responsive to talk therapy and to changes in the environment. Suicide is not a psychiatric disorder. Suicide is a nervous dysfunction, not a mental disease.

But 90% of suicides are in people with clinically diagnosable psychiatric disorders; anxiety, depression, impulsivity, and alcohol abuse are major risk factors. While cases of psychache would certainly benefit from talk therapy and a change in environment, pharmacological (and/or brain stimulation) treatments seem to be essential. Which is the clearly the intention of Yovell et al. (2015), or else they wouldn't have conducted a drug study.

In short, I found it curious that the focus of their clinical trial changed so much mid-stream, and that the mental anguish of the original formulation is so completely and utterly human (given its genesis from the animal literature).

In the next post, I'll cover the actual study and the background on why anyone would think low-dose opioids are a good idea in cases of treatment-resistant depression and suicidality.


Read Part 2, Opioid Drugs for Mental Anguish: Basic Research and Clinical Trials.


Further Reading

Vicodin for Social Exclusion

Suffering from the pain of social rejection? Feel better with TYLENOL®

Existential Dread of Absurd Social Psychology Studies

Does Tylenol Exert its Analgesic Effects via the Spinal Cord?

The Mental Health of Lonely Marijuana Users

Tylenol Doesn't Really Blunt Your Emotions

Of Mice and Women: Animal Models of Desire, Dread, and Despair


Footnotes

¹ In contrast, based on years of detailed neuroanatomical and neurophysiological experiments, most neuroscientists think the dACC is a functionally heterogeneous region (e.g., Vogt et al., 1992). Shortly after the Lieberman & Eisenberger (2015) paper was published, a number of researchers expressed their vehement disagreement in blog posts: Yarkoni-1, Lieberman reply, Yarkoni-2, Shackman, Wager.

² In contrast to these results, an earlier study by this group claimed that social rejection shares somatosensory representations with physical pain. It's always nice to see examples where scientists update their own theories based on new evidence.

³ In Panksepp's scheme, there are seven basic or primal emotions that are subcortically based and evolutionarily conserved: SEEKING, RAGE, FEAR, LUST, CARE, PANIC/GRIEF, and PLAY. Needless to say, this model has not gone unchallenged (Barrett et al., 2007; LeDoux, 2015). Barrett and colleagues have argued that emotions are not natural kinds, but rather emergent psychological events constructed from core affect (positive or negative states) and a human conceptual system for emotion.


References

Barrett LF, Lindquist KA, Bliss-Moreau E, Duncan S, Gendron M, Mize J, Brennan L. (2007). Of Mice and Men: Natural Kinds of Emotions in the Mammalian Brain? A Response to Panksepp and Izard. Perspect Psychol Sci. 2(3):297-312.

Eisenberger NI, Lieberman MD, Williams KD. (2003). Does rejection hurt? An FMRI study of social exclusion. Science 302:290-2.

Panksepp, J., & Yovell, Y. (2014). Preclinical Modeling of Primal Emotional Affects (SEEKING, PANIC and PLAY): Gateways to the Development of New Treatments for Depression. Psychopathology, 47 (6), 383-393. DOI: 10.1159/000366208

Shneidman ES. (1993). Suicide as psychache. J Nerv Ment Dis. 181(3):145-7.

Woo CW, Koban L, Kross E, Lindquist MA, Banich MT, Ruzic L, Andrews-Hanna JR, & Wager TD (2014). Separate neural representations for physical pain and social rejection. Nature communications, 5. PMID: 25400102

Yovell, Y., Bar, G., Mashiah, M., Baruch, Y., Briskman, I., Asherov, J., Lotan, A., Rigbi, A., & Panksepp, J. (2015). Ultra-Low-Dose Buprenorphine as a Time-Limited Treatment for Severe Suicidal Ideation: A Randomized Controlled Trial. American Journal of Psychiatry DOI: 10.1176/appi.ajp.2015.15040535

This Week in Neuroblunders: fMRI Edition

My entire body of work has been called into question!

And what a fine week for technical neurogaffes it is. First was the threat that many trendy and important studies of neural circuits may need to be replicated using old-fashioned lesion methods, because of “off-target” effects:

Where do we go from here? Most acute manipulation studies that use optogenetics confirm, and so add valuable support to, existing hypotheses that were established in earlier studies. But for those studies that have proposed new circuit functions, it may be advisable to re-evaluate the conclusions using independent approaches.¹

Up next we have....

fMRI Neuroblunders in Brief
 
The most notable one of late is a new paper by Eklund et al. (2015), which demonstrated that common statistical tests used to analyze fMRI data can give wildly inflated false positive rates of up to 60%, as illustrated in the top figure.

“What they found is shocking”!

While voxel-wise error rates were valid, nearly all cluster-based parametric methods (except for FSL’s FLAME 1) have greatly inflated familywise Type I error rates. This inflation was worst for analyses using lower cluster-forming thresholds (e.g. p=0.01) compared to higher thresholds, but even with higher thresholds there was serious inflation. This should be a sobering wake-up call for fMRI researchers, as it suggests that the methods used in a large number of previous publications suffer from exceedingly high false positive rates (sometimes greater than 50%).

The problems (and recommended solutions) were expertly discussed already by Russ Poldrack, who is quoted above (see Big problems for common fMRI thresholding methods), and by Neuroskeptic (False Positive fMRI Revisited). I needn't belabor the issues any further.


Next question:

Is the ubiquitously activated dorsal anterior cingulate cortex (dACC) selective for pain (as opposed to conflict or cognitive control or salience)? That was the contention of a new paper by Lieberman and Eisenberger (2015) that made use of the Neurosynth meta-analytic framework developed by Tal Yarkoni.

It Depends on What “Selective” Means ²

A 15,000 word debate between Yarkoni (No, the dorsal anterior cingulate is not selective for pain) and Lieberman (Comparing Pain, Cognitive, and Salience Accounts of dACC) ensued, with no end in sight.

It Also Depends on What “Pain” Means

Social Pain and Physical Pain Are Not Interchangeable. This may sound obvious to you, but Eisenberger and Lieberman have argued otherwise, with their neural alarm view of dACC function. Neurosynth uses text mining and machine learning to build maps based on terms that appear in published papers, along with activation coordinates. So the map above doesn't distinguish between different types of experimentally-induced physical pain (heat, cold, pressure, etc.) vs. emotional pain or social exclusion in a video game.

This might be one of L&E's major points, but pain researchers aren't on board; many don't even think the dorsal posterior insula is a pain-specific region.


Footnotes

¹ If anyone can parse the bold red sentence that appears in the Nature commentary (immediately after the first quoted passage in the post), please let me know.

In the future, it might be helpful always to correlate acute and chronic manipulations of specific neurons. If results from acute and chronic manipulations are discrepant, analyses of circuits that act in parallel to the manipulated circuit, or of similar neurons that are activated by different stimuli, might be more likely to provide an explanation for the discrepancy than examination of chains of hierarchically connected neurons, because off-target effects probably propagate throughout neural circuits by spilling over into adjacent, connected circuits.

² Sam Schwarzkopf addressed this in his post, What is selectivity?

This Week in Neuroblunders: Optogenetics Edition

Recent technological developments in neuroscience have enabled rapid advances in our knowledge of how neural circuits function in awake behaving animals. Highly targeted and reversible manipulations using light (optogenetics) or drugs have allowed scientists to demonstrate that activating a tiny population of neurons can evoke specific memories or induce insatiable feeding.

But this week we learned these popular and precise brain stimulation and inactivation methods may produce spurious links to behavior!! And that “controlling neurons with light or drugs may affect the brain in more ways than expected”! Who knew that rapid and reversible manipulations of a specific cell population might actually affect (gasp) more than the targeted circuit, suggesting that neural circuits do not operate in isolation??

Apparently, a lot of people already knew this.

Here's the dire Nature News report:

...stimulating one part of the brain to induce certain behaviours might cause other, unrelated parts to fire simultaneously, and so make it seem as if these circuits are also involved in the behaviour.

According to Ölveczky, the experiments suggest that although techniques such as optogenetics may show that a circuit can perform a function, they do not necessarily show that it normally performs that function. “I don’t want to say other studies have been wrong, but there is a danger to overinterpreting,” he says.

But the paper in question (Otchy et al., 2015) was not primarily about that problem. The major theme is shown in the figure above — the difference between acute manipulations using a drug (muscimol) to transiently inactivate a circuit versus the chronic effects of permanent damage (which show remarkable recovery).¹ In the songbird example, acute inactivation of the nucleus interface (Nif) vocal control area (and its “off-target” attachments) warped singing, but the “chronic” lesion did not.²

In an accompanying commentary, Dr. Thomas C. Südhof asked:

How should we interpret these experiments? Two opposing hypotheses come to mind. First, that acute manipulations are unreliable and should be discarded in favour of chronic manipulations. Second, that acute manipulations elicit results that truly reflect normal circuit functions, and the lack of changes after chronic manipulations is caused by compensatory plasticity. 

But not so fast! said Südhof (2015), who then stated the obvious. “Many chronic manipulations of neural circuits (both permanent genetic changes and physical lesions) do actually produce major behavioural changes.” [as if no one had ever heard of H.M. or Phineas Gage or Leborgne before now.]

The acute/chronic conundrum is nothing new in the world of human neurology. But centuries of crudely observing accidents of nature, with no control over which brain regions are damaged, and no delineation of precise neural mechanisms for behavior, don't count for much in our store of knowledge about acute vs. chronic manipulations of neural circuits.

Let's take a look at a few examples anyway.

In his 1876 Lecture on the Prognosis of Cerebral Hæmorrhage, Dr. Julius Althaus discussed recovery of function:

Do patients ever completely recover from an attack of cerebral hæmorrhage?
This question used formerly to be unhesitatingly answered in the affirmative.
. . .

The extent to which recovery of function may take place depends—

1. Upon the quantity of blood which has been effused.  ...

2. Upon the portion of the brain into which the effusion has taken place. Sensation is more easily re-established than motion; and hæmorrhage into the thalamus opticus seems to give better prospects of recovery than when the blood tears up the corpus striatum.  ...

[etc.]

In his 1913 textbook of neurology (Organic and Functional Nervous Diseases), Dr. Moses Allen Starr discussed aspects of paralysis from cortical disease, and the uniqueness of motor representations across individuals: “Every artisan, every musician, every dancer, has a peculiar individual store of motor memories. Some individuals possess a greater variety of them than others. Hence the motor zone on the cortex is of different extent in different persons, each newly acquired set of movements increasing its area.”

In 1983, we could read about Behavioral abnormalities after right hemisphere stroke and then Recovery of behavioral abnormalities after right hemisphere stroke.

More recently, there's been an emphasis on connectome-based approaches for quantifying the effects of focal brain injuries on large-scale network interactions, and how this might predict neuropsychological outcomes. So the trend in human neuroscience is to acknowledge the impact of chronic lesions on distant brain regions, rather than the current contention [in animals, of course] that “acute manipulations are probably more susceptible to off-target effects than are chronic lesions.”

But I digress...

Based on two Nature commentaries about the Otchy et al. paper, I was expecting “ah ha, gotcha, optogenetics is a fatally flawed technique.” This Hold Your Horses narrative fits nicely into a recap of neurogaffes in high places. One of the experiments did indeed use an optogenetic manipulation, but the issue wasn't specific to that method.

Ultimately, the neuroblunder for me wasn't the Experimental mismatch in neural circuits (or a failure of optogenetics per se), it was the mismatch between the-problem-as-hyped and a lack of historical context for said problem.


Footnotes

¹ Here's a figure from the other experiment, which involved acute vs. chronic inactivation of motor cortex in rats. Basically, the tiny injection of muscimol impaired lever-pressing behavior (acutely), but the large lesion did not (chronically). Panel H shows a similar deleterious effect using optogenetic stimulation.

Modified from Fig. 1 (Otchy et al., 2015).

I can't stress this point enough – a human with a comparably sized lesion in primary motor cortex would not [likely] show that much spontaneous recovery of function in 5-10 days. Yes, of course there's plasticity in the central nervous system of adult humans, but I think Otchy et al. (2015) overstate the case here:

As in our experimental animals, patients with lesions to motor-related brain areas have motor deficits that resolve in the days and weeks following the injury. Aspects of this recovery are thought to be independent of rehabilitation, suggesting spontaneous processes at work.

² It isn't exactly true that the lesions had no effect on song: “A fraction of the initial post-lesion vocalizations were severely degraded and did not resemble pre-lesion song.”


References

Althaus J (1876). A Lecture on the Prognosis of Cerebral Haemorrhage. British medical journal, 2 (812), 101-4. PMID: 20748269

Otchy, T., Wolff, S., Rhee, J., Pehlevan, C., Kawai, R., Kempf, A., Gobes, S., & Ölveczky, B. (2015). Acute off-target effects of neural circuit manipulations. Nature DOI: 10.1038/nature16442

Reardon, S. (2015). Brain-manipulation studies may produce spurious links to behaviour. Nature DOI: 10.1038/nature.2015.19003

Südhof, T. (2015). Reproducibility: Experimental mismatch in neural circuits. Nature DOI: 10.1038/nature16323