seriously. i’m highly irritated with this supposed moral position that opiates should not be used in pain treatment. (really- where the hell does this opinion even come from?! that’s what opiates DO!) i think people shouldn’t inflict their lack of knowledge upon others. so i’m going to talk a little bit (looking back, i talk about more than just a little bit!) about opiates and how they work in pain transmission mechanisms, so that my readers will be educated if they should ever happen upon such an ignorant assertion based upon no facts and only personal judgment. opiates are very well known effective pain-modulating agents. their known issues with tolerance, dependence, and possible addiction, and their fairly easy availablility on the street give them a general bad name. this can be deserved or not, depending on the application.
let’s start with the opioid receptors.
yes, my astute readers, you have probably noticed i just switched to “opioid” instead of “opiate”- allow me to explain briefly. this is just a terminology issue. opiates are exogenous. opioids are endogenous. that’s all. now let’s continue.
we have three types of opioid receptors. they are the mu, kappa, and delta subtypes. (they are labeled as such because of the pharmacological ligand that was used to identify them, and of course they had to be greek letters, because science loves teh greek letterz.) each subtype has its own endogenous endorphin peptide ligand- beta-endorphin, dynorphin, and enkephalin, respectively. the major opioid target we have for pain is mu, with most of the well known agonists used in pain treatment binding as agonists to mu receptors. kappa receptors do play a role as well, but we have fewer kappa agonists; and since dysphoria is a side effect of kappa activation, it’s a presumably less desirable target. delta receptors produce very mild analgesia, and we have few well-understood agonists for these receptors. regardless, all of the opioid receptors are located postsynaptically to the endogenous opioid-producing neurons, and produce an inhibitory effect upon the postsynaptic cell via G protein coupled mechanisms.
opioid receptors are found throughout the brain. the real bitch about neuropharmacology is that, short of microinjecting something straight into a certain region of the brain (which you can imagine, is not done in general practice in people) you can not target regions of the brain specifically. so if you put an opiate into your brain, you’re affecting all opioid-mediated processes- including the respiratory depression, euphoria, and of course downstream dopaminergic stimulation/reward which can in some cases lead to addictive behavior. these are all things we would rather not mess with, that make the opiates dangerous when played with carelessly. this is why risk and benefit must be weighed, and appropriate medical supervision is required when opiate drugs are used. it’s also why we’re very glad that we have antagonists like naloxone that can save you from an overdose.
next: pain transmission- the ascending and descending pathways
now that we’ve got a basic understanding of the opioid receptors and how they function, we need to know about the pain pathways that they interact with.
there are two general pathways in pain transmission: the signal that goes from your sensory neurons to the spinal cord to your brain, and the feedback from your brain back to your spinal cord. these are the ascending (up to the brain) and descending (down from the brain) pathways. each direction of signal travels through varying channels:
the ascending pathways are comprised of multiple circuits. signal is sent through the spinothalamic tract- as you might guess from the name, this sends information from the level of your spinal cord to your thalamus, where sensory information (and much more!) is processed and sent along to other parts of your brain. this basically sends the signal that an insult to the body has occurred and the sense and location information of pain that accompanies it. some signal is also sent to the periaqueductal grey (PAG) and then to the thalamus through this pathway.
the spinomesencephalic (spinal cord to midbrain) tract sends information to the PAG and superior colliculus. these two regions are less directly involved in the actual sense of pain, and more involved in the modulation of perception and behavior (and emotion) related to pain. the involvement of these regions are probably the reason that pain is highly subjective and individuals have different experiences of pain. the PAG is also involved in the descending pathway, which i will get to briefly. (i know, that’s kind of confusing, right? stick with me here.)
there is also a spinoreticular tract- projecting to the reticular area (an evolutionarily very old part of your brain involved in basic functions, among them pain processing), then along to the thalamus. in another confusing switch, the reticular area may also regulate the descending pain pathway.
the descending pathway, then, takes information from the brain and send it back down to the spinal cord as a feedback mechanism. this is just another homeostatic pathway- the brain gets the signal that pain is present, but that signal can then in turn be dampened by the brain on its way up from the body. (imagine it as if you want to turn down some very loud music. you will still hear it once you turn it down, but you don’t need THAT much signal going into your ears. that is the best analogy i can come up with right now for this negative regulation.) the major descending pathway is from the PAG to the medulla, to the dorsal horn of the spinal cord. the descending pathway modulates the ascending pathways at the spinal cord level. also, i might mention that the PAG is full of opioid receptors. just keep that in mind.
so, let’s recap: ascending pathways send information about pain to the brain, where different brain regions process sensory, emotional and behavioral responses. descending pathways send signal from the brain back to the spinal cord, where modulation (dampening) of the ascending pain signal can occur.
where and how opioids act upon the pain pathways
alright, so now we know about the opioid receptors, and we know about the pain pathways. but what is so special about the location of opioid receptors that makes them so well-poised to modulate pain?
well, to cut to the chase here, opioid receptors in the spinal cord inhibit ascending pain signals. all ascending signals must pass from the sensory neurons, up the spinal cord, to the brain. when you throw an agonist on the mu receptors in the spinal cord- like, say, morphine or hydrocodone- they inhibit neuronal firing (via that inhibitory G protein coupled signaling i mentioned earlier) in the ascending pathway. the signal is not transmitted to the brain!
to make this a double-score for opiates, they stimulate descending pathways- the pathways that dampen pain signal transmission. this is a little more complex- you may be wondering now, since stimulating the opioid receptor is inhibitory, why this action stimulates the descending pathways? this is a little more complex, so i’m going to put that in its own section with its own set of explanations.
mechanism of opioid stimulation of descending pathways
remember that the brain is driven by, most basically, a series of additive binary signals. excitatory transmission leads to neuron firing, but inhibitory transmission makes it less likely that the neuron will fire. the sum of inputs determines the outcome- whether the neuron fires or not. keep this in mind as you read this next part.
the descending pathway is tonically inhibited by GABA in the PAG, recall this is the major inhibitory neurotransmitter in the brain. so under normal conditions, GABA is keeping the key neurons from firing. this is fine, not a problem, it’s the way things normally work. but the cool thing starts here. we have opioid receptors on the GABA neurons- so endogenous opioids and opiate drugs work by inhibiting the inhibition of the pathway. by removing the inhibition, now these pathway neurons are able to fire and the descending pathway can dampen the pain signal coming up to the brain. really, how amazingly brilliant is that?!
so to put it all together
ascending pathways are blocked by opiates. this keeps pain signal from reaching your brain. in addition, opiates indirectly stimulate the descending pathways, which dampen the pain signal on its way up to the brain. this is a double whammy against pain signaling, which is quite effective, as you might imagine.
but leigh, what about side effects? and addiction?
these are big issues. opiates are NOT wonder drugs. people become tolerant to the pain effects pretty rapidly, but tolerance to the respiratory depression effects doesn’t necessarily occur at the same rate. when you become tolerant to the pain relief, you might choose to take increasing doses of the drug to keep that pain relief effect going. however, this can put you dangerously close to respiratory failure. typically, opiate overdose becomes fatal due to respiratory suppression. it’s convenient then, that you never become tolerant to miosis (constricted pupils), because that is a good indicator that you’ve been taking opiates. in the case of overdose, if you’re really lucky and wind up in the ER, this might just get you identified as someone who stopped breathing due to opiate overdose and save your life. then, hopefully you’ve learned your lesson not to fool with the opiates.
the other side effects aren’t a picnic either. itching, insufficient thermoregulation (you can get really hot and sweaty- this is thanks to the opioid receptors in your brainstem), and constipation (due to the opioid receptors that regulate GI muscle motility). not to mention nausea and vomiting.
the biggest concern that people seem to have about these drugs is that you can get addicted to them. this is a legitimate concern, but not everyone who takes an opiate will become an addict.
addiction via the dopaminergic pathway
it turns out that opioid receptors also play a big role in regulating signaling in the dopaminergic system. go figure. remember the receptors are inhibitory. and we just talked about how they can indirectly stimulate a pathway by inhibiting GABA signaling. well, mu receptors are located on GABA neurons that are presynaptic to dopamine neurons! mu agonists then relieve inhibition on dopamine neurons, indirectly driving dopamine firing. that’s how mu agonists drive reward, and that property gives them addiction liability. remember that i mentioned kappa receptors had the side effect of dysphoria? they are located on dopamine neurons, and stimulating them with an agonist decreases dopamine firing. so as you might imagine, kappa agonists don’t have the same addiction liability that mu agonists do.
how to responsibly handle these drugs?
there are plenty of people who take opiates to manage pain. so how do they do it, with all the seeming risks? quite simply, they follow the instructions given by their doctor. when they become tolerant, their doctor should put them on a drug holiday to restore opioid receptor sensitivity and number at the membrane. (i’ve discussed the receptor-mediated mechanisms of tolerance previously if you’d like to catch up on that topic.) there is great responsibility for both doctor and patient to ensure that the drug is working to the benefit of the patient. a prevalent method in pain management clinics is to have a contract that outlines a very specific set of rules the patient must follow to ensure that the drugs are not being misused or redirected. in the case i am familiar with, the patient is evaluated by a psychiatrist prior to initiation of opiate therapy. then, they are subject to randomized drug testing at any time, must bring in their remaining pills for counting at every doctor visit, and must always use the same pharmacy. they are not allowed to obtain pain medications from any other doctor. any violation of these contractual rules leads to revocation of the opiate, and the patient is given drugs to counteract the effects of withdrawal. (this is important, since withdrawal can be very unpleasant indeed. and this is also one of the reasons it is just downright cruel to just randomly, senselessly deny somebody an opiate prescription when they are on long-term opiate pain maintenance.)
in summary
yes, opiates are quite effective in pain management. when used properly, they do their job well by hitting both ascending and descending pain pathways. there are risks, but the responsible use of opiates is key and falls on the shoulders of the doctor and the patient. misuse and recreational use have given these drugs a bad name- as our country struggles with this so-called war on drugs, people become increasingly judgmental of people who are taking opiate drugs. this is not necessary. a person managing pain with an opiate is not an addict or a criminal, they are simply using a very effective drug- likely because they’ve tried everything else without success. this does not constitute something that we can make a moral judgment upon. it does constitute the best use of medical knowledge that we have to improve the lives of others, which is what gets a lot of us out of bed in the morning even when work seems a little overwhelming.
and that’s all i have to say about that.

21 comments
Comments feed for this article
March 14, 2009 at 11:07 pm
whitecoattales
Beautiful summary, I love it.
As for the moral argument, I don’t understand it, and every time I ask someone about it, they never seem to be able to get past “well really, I just don’t know any medicine and I’m not qualified to have this discussion but I know morphine is bad!”
Now mind you, as a med student, I get stuck dealing with alot of Narcotic seeking. It sucks. Thats still no reason to deny actual pain control to legitimate pain.
March 14, 2009 at 11:15 pm
leigh
thanks for the comment, wct.
i can understand the need to use great caution in using opiates in medical practice. it’s kinda sad if you ask me, that doctors need to be so skeptical of their patients, but also reflective of reality today.
in the cases where pain is legitimate, the patient has a history of long term use, and maybe even that pain contract system is in place, though? there’s absolutely no evidence-based reason for a pharmacist to simply state “narcotics are bad!” and force their own opinion that the patient should be managing your pain another way.
March 14, 2009 at 11:32 pm
whitecoattales
In cases where pain is legitimate, etc, theres no reason, evidence based, or otherwise, for a pharmacist to deny filling a prescription. At best, they could justify calling the physician.
And frankly, a few bad apples ruin it for everyone. I’m on neurology now, and I’ve treated 15 people this month for intractable headaches. To my best guess, 3 were just drug seeking, 5 had legitimate pain BUT also had a likely narcotics addiction influenceing their behavior, and the rest just wanted their heads to stop hurting. I don’t know that those statistics are representative. Because of those 3 patients though, I’ve been darn suspicious of everyone, because I HATE feeling like a patient is playing me. So I can see a pharmacists suspicion gettig the better of him/her, but I don’t see the justification for anything else, its out of line to make medical decisions without gathering medical information.
March 15, 2009 at 9:22 am
Comrade PhysioProf
Nice post! There are also mu opioid receptors in the peripheral terminals of nociceptors, and they almost certainly play a role in opioid analgesia, particularly for neurogenic and neuropathic pain.
March 15, 2009 at 9:36 am
leigh
its out of line to make medical decisions without gathering medical information.
that’s exactly what pissed me off about the situation.
March 15, 2009 at 9:37 am
leigh
CPP- ooh, i didn’t catch that in my reading. thanks for pointing it out! i can see how that could definitely have a big impact, especially in neurogenic/neuropathic pain.
March 15, 2009 at 9:58 am
whitecoattales
Hmm, my understanding was that neuropathic pain was in large part not nociceptive, although what they teach us in med school is generally just enough to shut us up, rather than complete understanding.
In practice, I’ve noticed opioids aren’t great for neuropathic pain.
Better meds for that (at least in what I’ve seen) are pregabalin/gabapentin, duloxetine, sometimes topiramate. Now why those are specifically better medications, I’m not enough of a pharmacologist to tell you in any detail. Instead I shall wave my hands *wave* and say I think it has to do with those medications actions on ion channels (I think calcium channels in the case of pregabalin or gabapentin).
Please feel free to correct my understanding with science instead of my voodoo.
March 15, 2009 at 10:06 am
whitecoattales
Oo, tricyclics are good for neuropathic pain too… *handwave*
March 15, 2009 at 10:25 am
leigh
re: mu and neuropathic pain, i was able to find published evidence for peripheral mu involvement in a rat model of neuropathic pain: http://www.ncbi.nlm.nih.gov/pubmed/18276075
re: pregabalin/gabapentin, we don’t know exactly how that works. from what i’m gathering, in vitro data show that they inhibit calcium influx, which is a key step in vesicular neurotransmitter release. if i had to guess why this worked, i’d guess the specific subunit it binds to has to do with its selective activity and why it doesn’t just decrease neurotransmission through the brain. my spouse was on lyrica for about a month before he stopped taking it, because he had a hard time thinking clearly, and felt dumb. (this was not good, his job at the time required him to be on his toes at all times.)
my doc had me on tca’s a long time ago as my acute pain became chronic- that was one of the things that spiked my interest in pharmacology in the first place. several other lines of treatment had failed me at that point.
tca’s block reuptake of 5ht and norepi, which are major neurotransmitters in the descending pathway. but the anticholinergic side effects, they sucked. the pain still kept me up at night.
March 15, 2009 at 10:33 am
whitecoattales
Good to know!
I just discovered your blog through the ping from whitecoat underground, I’ve been backing my way through it, you’re writing is very enjoyable
.
March 15, 2009 at 1:52 pm
JLK
OOh…..you totally just made me want to do a post about the psychology of pain! Just added it to my to-do list!
March 15, 2009 at 10:53 pm
leigh
wct: why, thank you! i’m glad that you enjoy it.
JLK: i am so looking forward to your contribution from the psych side!
March 16, 2009 at 9:40 pm
scicurious
OOOOH!!! Leigh I am so mad at you now! I was totally going to cover this!! Argh. But you did WAY better than I could have. I love it. Awesome stuff.
March 17, 2009 at 2:00 am
Tuesday Happenings [Neurotopia]
[...] Finally, leigh has posted a GREAT post on opioid pharmacology. I’m now super pissed at her because she beat me to it. I planned to do that next week, [...]
March 17, 2009 at 4:15 pm
leigh
sci, you crack me up! thanks!
March 17, 2009 at 4:24 pm
AK
Great post.
I just discovered your blog (via scicurious) so I’m not sure how to react to this:
AFAIK this is extremely simplistic and unproven, unless you specify that the “sum of inputs” occurs at the level of small branches within the dendritic arbor. Am I behind times? Has simple addition across the neuron been demonstrated? Ref’s?
I have some other questions: How much cross-functionality is there between receptors and other receptors’ “endogenous endorphin peptide ligands”? Are there any other endorphins besides these three? Do any endorphins act as neurohormones or do they get attenuated enough away from the synapses that they have no effect? Are there any opioid receptors not associated with synapses?
I’d guess that’s one mark of a good post: that it leaves people with more questions than you answered.
March 17, 2009 at 11:37 pm
leigh
AK:
thanks for your comment.
of course the sum of inputs comment is simplistic, but that is the bare-bones basic description of inhibitory and excitatory inputs. since the voltage across the membrane is the key determining factor in whether an action potential fires or not, the result of excitatory (depolarizing) and inhibitory (hyperpolarizing) inputs do essentially determine whether the membrane depolarizes or not. given the intended level of my audience (i try to ask myself, would most of my non-science friends understand this if they tried reading it?), i chose to skip the hardcore neurophysiology to bring my point.
there is always the possibility of other opioid receptors we have not yet cloned and characterized. so far we know of these three and several other proposed (but i don’t think confirmed) receptors. research has determined the binding specificity of each ligand-receptor pair- the results indicate that, for example, beta-endorphin has high affinity for mu receptors, lower for delta, and very low affinity for kappa. basically, beta-endorphin is most effective at mu receptors, and mu receptors are located postsynaptically to beta-endorphin neurons. this leads us to our conclusions about mechanisms of action. there is another identified peptide, endomorphin, which is also active at mu receptors. again, this is our current state of knowledge and future work may give us more.
endorphin peptides have a short half-life, so i would not expect them to get too far outside the synapse. in the brain, these are definitely associated with synapses, since they are neurotransmitter receptors. in peripheral tissues, i haven’t done enough reading to give you a straight answer. i would guess they’re part of the pns, (given the time and my lack of a functional VPN to find a paper) but someone can correct me if i’m wrong.
i hope i’ve answered your questions!
March 18, 2009 at 9:55 am
AK
Yes you have, thanks Leigh.
I understand the issue with the hardcore neurophysiology; a possible problem with this is that may give the impression that the neuron is pretty “simple and dumb” in its computational activity. I remember a comment on HotWired (in ‘98, IIRC) that the neuron is “about as smart as my desktop computer”. My own readings (especially The Neuron Cell and Molecular Biology 2nd Edition by Levitan, Kaczmarek and The Synaptic Organization of the Brain 5th edition ed. by Gordon M. Shepherd) seem to have confirmed this. While such details may be beyond the audience, the simple fact that a single neuron can perform extremely complex analog computations on hundreds or thousands of inputs (even 100,000’s in the case of Purkinje cells) is easily grasped. The notion that the overall response to all these inputs is simple addition gives, IMO, an unfortunately simplistic view of the brain’s computational abilities.
The comment mentioned above occurred during a debate over some nonsense about quantum gravity waves traveling along micro-tubules as an “explanation” for human intelligence; a notion that could only (IMO) be justified by physicists’ misunderstanding of the abilities of even single neurons, much less networks of billions of them.
March 20, 2009 at 9:36 pm
juniorprof
gabapentin and pregabalin seem to act by inhibiting VGCCs through an action at the alpha2delta subunit. However, electrophys studies have also shown that they increase descending inhibition (similar to opiates). Why? no one knows. Tricyclics can increase spinal NE levels. This is also linked to descending inhibition because activation of descending inhibition circuits leads to increased levels of spinal NE. This is also part of the mechanism of intrathecal clonidine (alpha agonist). Interestingly, clonidine can also be used for opiate withdrawal and it inhibits some of the pain associated with opiate withdrawal. You might imagine how that works… Topiramate does all kind of stuff including inhibition of VGNaCs and inhibition of carbonic anhydrase.
Opiates are still effective for neuropathic pain but side-effects become strongly limiting because you have to take them over longer periods. There is also some indication that descending inhibitor circuits are dampened in neuropathic pain leading to a net increase of descending facilitation (which arises from the same areas) and making it harder for opiates to work. On the other hand, gabapentin is supposed to use this mechanism in neuropathic conditions (at least in part) so there is some hand waving there.
We place too much burden on pain patients to “prove” that they have pain in order to get their meds. I think this is bullshit. We know that unmananged pain makes pain become chronic more easily and leads to an increase in severity. NIH has spent millions so that we can learn this. You would think regulatory agencies would care about science and evidence based practice. I think they ignore this all too often, to the great detriment of health care practice.
March 22, 2009 at 9:47 pm
leigh
AK: point taken. i will be more careful to convey the vast complexity of neuronal networks next time.
JP: thanks for the awesome comment, it’s far more in depth of an answer than i could find in my brief dig around the literature. i agree with your opinion of the burden upon pain patients- i’ve got some incredible stories.
December 3, 2009 at 11:50 pm
it either targets ALL of the brain or NONE of the brain. « the path forward
[...] opiates are a pretty good example. they have powerful pain-relieving properties when they bind opioid receptors in the ascending and descending pain pathways. they also have powerful reinforcing properties when they bind opioid receptors in the ventral tegmental area and nucleus accumbens, and depress respiration when they bind opioid receptors in the brainstem. (i’m being general in this post, if you want more details on opiates i have a better explanation here.) [...]