PharmTox Lab Rats Volume 5: Pain Science Behind the Scenes
Patients come into the ER with broken bones and leave with opioid addictions. That’s not fair.
Pain labs are trying to stop these patients from ever becoming addicts in the first place, by developing new, non-addictive pain relievers. Here is a behind-the-scenes look at how new pain medicines are tested in the lab.
(Check out PharmTox Lab Rats Volume 2 for information about opioid pharmacology).
A few years ago, I sprained my ankle. The next day, it swelled to the size and coloring of mango. The emergency room doctor gave me Vicodin, an opioid pain reliever, to take twice daily for the pain. I took my first dose at the hospital, then had no recollection of the rest of the day. When I checked my computer the next morning, I discovered that I had, apparently, been emailing people in Swedish. Remarkably, I had used perfect grammar, punctuation, and dotted my å’s and ö’s, to go on a rant about reindeer cheese toothpaste.
I decided that it was in my best interest not to take the rest of my prescription.
For a week, I was unable to walk due to pain. Fortunately, I was able to work from home during that period. But what if I had a job where being on crutches was not an option, like being a server or contractor? I think I would have decided to continue using opioids for pain, if it meant that I could keep my job. However, this line of thinking is part of what’s driving the opioid epidemic: people with pain are taking their medication as prescribed so that they can function in their daily lives, but in doing so, are accidentally transitioning into addicts.
In response to the exponential increase in opioid-related deaths this decade, The National Institutes of Health has proposed the HEAL Initiative: $1 billion for research to prevent pain patients from becoming opioid addicts, and to treat individuals who are currently opioid addicts.
How do we prevent a pain patient from becoming an opioid addict?
Don’t give them opioids. The patients can’t become addicted if they never try opioids in the first place.
True as that reasoning might be, we have a moral dilemma here. Opioids (morphine, hydrocodone, fentanyl) are the only effective medication for severe, acute pain. If you break your leg, aspirin is not going to make you feel better. Anticonvulsant (anti-seizure) drugs are sometimes prescribed for chronic, neuropathic pain, but only a small subset of patients report that anticonvulsants improve their pain, and of patients with pain relief, anticonvulsants only provide partial pain relief; the rest of the patients have mental fog and sleepiness. Antidepressants sometimes improve chronic pain, slightly. If a patient has severe acute pain, withholding pain relief borders on cruel.
Behavioral pharmacologists are hard at work trying to find analgesics better than opioids. When testing experimental pain relievers on animals, here are three broad points to consider:
- Does the animal have pain?
- How do we measure this pain?
- Does the experimental medicine have abuse potential?
1. Does the animal have pain?
What is pain? Take a moment and try to give me a definition. You’ve surely experienced pain in your life. Why is it so hard to put that feeling into words?
An hour after injuring my ankle, it did not feel like a specific sensation I could describe, but rather like a feeling of something being very wrong. I walked my bicycle home because I could not figure out how to ride it. I did not even think that I had pain, because I was able to walk, but for some reason, I could not lift my bicycle up the 2" step into my house. I rationalized that I was struggling because I was somehow severely dehydrated.
According to the International Association for the Study of Pain, pain is:
“an unpleasant sensory and emotional experience associated with actual or potential tissue damage, or described in terms of such damage.”
In the case of the injured ankle, I was having an unpleasant emotional experience associated with actual tissue damage. By the next day, the swelling and unpleasant sensory experience set in. The ER doctor wondered why I had waited a whole day to seek treatment for something that looked very painful.
Here is a happy golden retriever. Specifically, its facial expression and posture make it look happy.
Here is another golden retriever. Its facial expression and posture make it look sad. The x-ray shows that he has osteoarthritis. In humans, osteoarthritis is an unpleasant sensory and emotional experience. Is the sad-looking golden retriever in pain?
As a dog lover, I would say, “Yes, he is in pain.”
As a scientist, I would say, “No. Technically, he is not in pain.”
The reason why the scientist in me says “no” is because pain is sensory and emotional, or described by the person experiencing it as pain. Unlike your great-aunt, the dog cannot say that the humidity is making his osteoarthritis act up today, putting him in a bad mood.
Instead, when scientists talk about pain in animals, they use the term “nociception”, which means the animal’s ability to feel something that typically causes humans to report pain. Nociception is often accompanied by a reflex to avoid the thing causing the sensation. For example, if you touch a hot stove, you will have a reflex to move your hand away. If an animal’s tail is put in hot water, it will have a reflex to pull its tail out of the hot water. Instead of treating “pain”, scientists try to determine if their experimental medicines will treat “nociception” in animals, typically be testing if the medicine will prevent these reflex behaviors.
A possible exception to the rule that animals cannot feel “pain” is Koko the gorilla. Koko communicated with sign language that she had high pain and pointed to her tooth. A dentist observed that the tooth was decayed, which does cause excruciating pain in humans.
2. How do we measure this pain (or nociception)?
Here are some common laboratory techniques used to measure nociception.
von Frey Test
The von Frey test for mechanical hypersensitivity is used in the clinic and the lab. In this experiment, a person is poked with a tool that looks like a toothbrush with one bristle. The person says whether they can feel it, and if the sensation is painful. If a person has pain in a particular area, they will be more sensitive to the bristle touching them in that spot.
Remember that sunburn you got at the spring break concert? The slightest touch hurt. Technically, you had mechanical hypersensitivity.
Here is what the von Frey test looks like in a rat:
Instead of a self report of pain, we look at how much poking force is required to make the rat move his paw. Rats with nerve damage in their legs, similar to humans with neuropathy, are extra sensitive to the slightest touch. When we treat these rats with morphine or anticonvulsants, they stop being as sensitive to touch. Because the rat has “nociception” instead of “pain”, morphine is not producing “pain relief”, but rather “antinociception” (inability to feel the noxious sensation).
When I performed the von Frey test on mice with nerve damage in their legs, I saw that the mice who were treated with the anticonvulsant gabapentin did not flinch with any amount of poking force. It looked like gabapentin worked well to block their nociception. Except, I poked one mouse’s foot so hard that he tipped over on his side, and stayed there. He was either deep asleep or paralyzed, able to feel everything. Whichever one was problematic.
Years later, when I went to see the neurologist about chronic nerve pain in my legs, he prescribed gabapentin. I told him about the mouse. He recommended taking amphetamine along with the gabapentin to cancel out the side effects. I respectfully declined treatment.
Acid stretch test
Have you ever gotten a burning sensation in your calves after running? When muscles are performing strenuous exercise, they produce lactic acid. This temporary buildup of acid can be painful. Your first instinct was probably to rub your calves. Rodents do the same thing: when you inject lactic acid into the belly of a mouse, he will try to rub his belly by stretching his torso across the floor.
Here is a mouse performing the acid stretch test:
When mice are pre-treated with drugs that fight pain in humans, the mice will stretch less in response to an acid injection. Scientists interpret this reduction of stretching behavior as antinociception.
However, there is a flaw to this logic. What if the mouse stops flinching his paw in the von Frey test and stops stretching in the acid test, not because the experimental drug treats pain, but because the mouse is paralyzed, stoned out of his mind, or asleep?
To make sure that the drug does not impair the mouse’s ability to move, scientists use the rotarod test. For this test, rodents are injected with the candidate pain reliever, then placed on a spinning rod. If the new pain reliever impairs motor coordination (or makes them sleepy), they will fall off the rotarod faster.
Here are a group of rats performing the rotarod test:
FYI, lumberjacks compete in their own rotarod test, log rolling:
Nesting assay
Another method to determine whether an experimental pain reliever works without causing motor impairment is to measure behaviors laboratory animals typically do, that they stop doing when they have pain.
For example, when I’m getting cozy in bed, I like to fluff up my five pillows and two feather comforters, and go to sleep in a pillow fort. Rodents like fluffy beds, too. If you give a rat pieces of compressed cotton, he will shred them to create his own pillow fort.
Here is a before-and-after picture of a rat nesting:
If you inject the rat with acid in his belly, he will be too busy focusing on how the acid makes him feel (and stretching his belly) to build his bed. If the rat is given a pain reliever, the acid will not bother him as much, and he will go to work making his bed.
An ideal pain reliever would not let pain get in the way of doing chores around the house (building pillow forts) or going to work.
3. Does the experimental medicine have abuse potential?
The experiments I showed you are screening tools, to find candidate pain relievers. However, just because a drug passes those tests does not mean that it is safe. Morphine passes all of those tests, but we know that morphine is addictive. Therefore, we need to include another category of testing: abuse potential.
Let’s divide abuse potential testing into two general categories:
- How much does the animal “like” the experimental medicine?
- How much does the animal “want” the experimental medicine?
“Liking”: Conditioned Place Preference
Conditioned place preference uses Pavlovian conditioning. In Pavlovian conditioning, something that makes your body involuntarily react in a certain way (seeing tasty food makes you drool) is accompanied by something else that you don’t particularly care about (a bell ringing). If there is always a bell ringing when you eat the tasty food, in the future, when you hear the bell, you will involuntarily start drooling.
We use this same principle to train rodents (or humans!) to associate how they feel when given the experimental medicine with the room that they are in. After repeatedly treating the laboratory animals with the experimental pain reliever and putting them in a hideously-decorated compartment of a multi-compartment cage, the animals learn that they feel a certain way in that compartment, and they feel neutral (after getting a placebo) when put in a different compartment. Later, the animals have access to all of the compartments in the cage, and can choose to stay in the hideous compartment, or go to the other compartments. Essentially, we are asking the animals if they like the way they feel in the hideous compartment, or if that compartment makes them feel unhappy and they want to leave.
Conditioned place preference is sort of like if you go to the same bar every week with friends, and the ugly wallpaper starts to make you happy. The alcohol and social interaction you received while in that room gave you conditioned place preference for that ugly wallpaper.
Morphine produces conditioned place preference in rats with and without nociception. If the experimental medication produces preference in everyone, then it might be a drug of abuse. Ideally, the experimental medicine would not cause conditioned place preference; rats wouldn’t care one way or the other whether they got drug or placebo. However, if a rat has prolonged nociception from an injury, and the experimental medicine is antinociceptive, then the rat would show conditioned place preference for “pain” relief. An important aspect of this test is that the rats only show preference for the room associated with the medicine when in a pain state.
“Wanting”: Self Administration on a Progressive Ratio
On April 10th every year, Ben & Jerry’s gives away ice cream. If you’re like me, you’ll be in line for a free scoop. While Ben & Jerry’s is my favorite brand of ice cream (high liking), I’m not going to spend $5 on a scoop of ice cream on April 11th, or $9 for a scoop when I go to the beach. I might buy some pints of Ben & Jerry’s if they are on clearance at the grocery store, though (low wanting).
How badly do you want a scoop of Ben & Jerry’s ice cream right now? Would you eat it if it were free? If it were on sale? If it were full-price? If it were tourist price? If it cost $1,000?
Scientists do a similar experiment to determine if a new medicine will have abuse potential. How much will a rat pay for the drug?
Instead of paying with money, rats can pay by doing a task: pressing a lever.
Initially, the drug is cheap, where pressing the lever once results in one injection of the drug. However, the drug gets more expensive over time, so that more work is required in order to get the next injection. Eventually, it might cost the rat 1,000 lever presses to get the drug. For something with high abuse potential, like heroin, the rats will gladly keep lever pressing (high wanting). For drugs without abuse potential, the rat stops lever pressing as soon as he realizes that the drug does not produce a pleasurable outcome. However, if a rat has nociception, he might pay a little more for an experimental therapeutic that is antinociceptive.
Self-administration studies are not 100% fool proof, though; they can give false negatives and false positives. For example, there are many experimental design details that must be just right in order to get a rodent to lever press for nicotine, even though rodents show conditioned place preference (“liking”) for nicotine. It is near impossible to get rodents to lever press for THC (the active ingredient in marijuana; check out PharmTox Lab Rats Volume 4 for more information about THC). On the other hand, rodents will lever press for caffeine, even though caffeine is not an illicit drug of abuse.
Summary
- Pain is complicated. It is hard for humans to describe, and there are challenges in identifying pain in animals.
- Candidate pain relievers must be tested in many different types of experiments in order to determine if they work. The drug must block excessive pain sensation, and also block pain from getting in the way of performing tasks of daily living.
- Abuse potential testing is critical. An ideal drug would have no abuse potential, no addiction risk.
Discovery of new pain relievers could prevent millions of pain patients from becoming addicted to opioids. Hopefully, we find something that works, soon.
References:
Childs, E., & de Wit, H. (2009). Amphetamine-Induced Place Preference in Humans. Biological Psychiatry , 65(10), 900–904. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2693956/pdf/nihms115427.pdf
Klinck, M.P., Mogil, J.S., Moreau, M., Lascelles, B.D.X., Flecknell, P.A., Poitte, T., & Troncy, E. (2017). Translational pain assessment: could natural animal models be the missing link? Pain, 158(9): 1633–1646. https://www.ncbi.nlm.nih.gov/pubmed/28614187
Negus, S.S. (2018). Addressing the opioid crisis: the importance of choosing translational endpoints. Trends in Pharmacological Sciences, 39(4): 327–330. https://www.ncbi.nlm.nih.gov/pubmed/29501211
Wall, P. D., McMahon, S. B., & Koltzenburg, M. (2013). Wall and Melzack’s textbook of pain. Philadelphia: Elsevier/Churchill Livingstone.
