Tolerance Troubles

[Not meant as a claim that science doesn’t know something. More of an admission that I don’t know some things, and a hope to be informed about them by someone who does.]

Everyone knows about tolerance. The first time you take heroin, you get really high. The second time you take heroin, you get slightly less high. The nth time you take heroin, you barely feel good at all – but if you stop taking heroin, then you feel miserable. Your body adjusts, the receptors desensitize, whatever.

This is so simple that it took me forever to ask the obvious next question – how come this doesn’t happen for everything else? Supposedly if you have ADHD you can just stay on Adderall forever. Nobody says “The first time you take Adderall you can concentrate really well, the second time you take it you’ll concentrate less well, and the nth time you take it you can barely concentrate better at all.”

The psychiatry textbooks contain a sentence or two saying that “some” patients “may” develop Adderall tolerance, but it’s not something that we’re trained to expect. There are a lot of anecdotal reports online, but there are also anecdotal reports of people who don’t develop any tolerance at all after years and years. Hmm.

Also, people who abuse Adderall develop tolerance all the time, and keep having to up the doses just like heroin abusers do. This is a little weird – my pet theory is that people only develop tolerance to drug effects that aren’t FDA-approved uses – though how your receptors know what the FDA says I haven’t quite figured out. More seriously, it may be an effect of method of use – taking a small amount daily versus snorting a large amount of crushed tablet whenever you feel like it. Or it may be that tolerance to euphoric effects is worse than tolerance to stimulant effects.

(This seems true in general – I get euphoric effects from caffeine when I drink it very rarely; if I drink it more often, the euphoria goes away and I just feel a little more awake. This is important since it suggests tolerance isn’t just your body metabolizing the drug better, but actually a matter of receptor-level action – something I think everyone agrees is true, but which it’s always nice to have independent confirmation for.)

Sometimes tolerance gets weird. Antipsychotics are supposed to block dopamine receptors. Too much dopamine can contribute to psychosis, but it can also screw up the basal ganglia’s modulation of movement and cause you to make repetitive jerking motions all the time. People who have been on antipsychotics for too long may remain protected from psychosis, but start making repetitive jerking movements in a way consistent with too much dopamine. The theory goes that the receptors involved in psychosis haven’t developed tolerance (for some reason), but the receptors involved in movement modulation have developed so much tolerance that they’ve overshot their baseline and become supersensitive to dopamine. So by taking a drug that lowers dopamine, you get higher dopaminergic effects. In the worst case scenario, you end up with a condition called tardive dyskinesia, which is permanent. The receptors stay supersensitive forever and you will always make repetitive jerking movements. If you stop the antipsychotic, that will just make it (temporarily) worse – now you have supersensitive dopamine receptors and you’re not blocking them, so that means lots of repetitive jerking movements.

In this case, giving someone a drug has caused them to develop not just tolerance but supertolerance, where they are permanently worse than before. It’s as if taking heroin for long enough made you permanently miserable.

…which actually isn’t totally hypothetical. Some percent of people who abuse opiates like heroin get what’s called a post-acute withdrawal syndrome (PAWS), meaning that they feel depressed for months or years after they stop using the opiates. I treated a patient like for a while. I tried pretty much every antidepressant on him without success. He was just miserable. He’d been clean for about six months and I told him that he might just have to wait it out – it usually goes away after a few months to a year or so.

There is a poorly-studied but anecdotally very helpful treatment for PAWS: low-dose naltrexone. Naively, this sounds like the stupidest possible thing to try. It’s an opiate antagonist, meaning that you’re taking somebody who is undersensitive to opiates and blocking the tiny number of functioning opiate receptors they already have. It should be the only thing capable of making this already bad condition worse. Yet people swear by it. The theory is supposed to be the tolerance reaction again. Your body reacts to this opiate-blocking agent by releasing more opiates. So we’re treating a condition in which drugs that increase opiates cause you to have fewer opiates, by giving you a drug that decreases opiates which will cause you to have more opiates. How annoying is that?!

(some people recommend that if you’re giving someone opiate painkillers, you can give them low-dose naltrexone at the same time to prevent development of tolerance. Giving someone an opiate and an opiate-blocker simultaneously seems kind of like the medical equivalent of digging holes and filling them back in again, but apparently it does something useful)

This means we have examples of all three of the following:

1. A drug that’s supposed to have effect X, and after a while it still has effect X (Adderall)
2. A drug that’s supposed to have effect X, but after a while it has no effect (heroin)
3. A drug that’s supposed to have effect X, but after a while it overshoots and has effect anti-X (Antipsychotics? Heroin? Naltrexone?)

You may notice that these are all three of the logical possibilities. So for example, if we gave someone a drug that was supposed to decrease anxiety, it might decrease anxiety, have no effect, or increase anxiety. If scientific hypotheses are about closing off parts of possibility-space, then the receptor sensitivity hypothesis isn’t doing a very good job.

But it’s actually worse than this, because I get the impression that different people will end up in different branches of this trilemma. Benzodiazepines are a special offender here. Some people can take Xanax once a day for anxiety, and it’s a perfect solution – it suppresses their anxiety, it never stops working, and they never become addicted – if twenty years later they get a good therapist who helps them treat their anxiety without drugs, they can stop the Xanax with just a couple of days of mild discomfort. Other people will lose all effect after a couple of weeks, up the dose, up the dose some more, and end up as total wrecks. I think this is much less common than most people say – my attending’s rule of thumb is “benzo tolerance develops for sleep but not anxiety” – but it certainly happens. And for that matter, I’ve met a few people who never seem to develop tolerance for benzodiazepine sleeping pills. You see this same pattern for opiates used as painkillers. I spent so many years confused about whether people develop tolerance to these or not, and my final conclusion is that some people do and some people don’t and if you try to find a coherent universal pattern here you will go insane.

And it’s actually worse than this. Drugs don’t just work differently in different people, sometimes the same person will cycle through totally different mechanisms of drug response. SSRIs have something called tachyphylaxis, where they’ll work really well for months and then suddenly stop working (the word means “fast protection”, ie you develop protection against the drug effects quickly). This is even more annoying than the other patterns – at least with heroin, it makes sense that the receptors will gradually lose their sensitivity. But here? In random people at random times, the drug just stops working suddenly. It might be after a month, it might be after a year, it might be after ten years. And then every so often you’ll try the drug again a decade later, and then it’ll work just fine. Why? Nobody knows.

Some skeptics have pointed out that this is exactly what you would expect if the drug had no real effect and it was just luck that people didn’t have depressive episodes while they were taking them, but we know SSRIs have some effect. And anyway, placebo tachyphylaxis isn’t any less weird than real tachyphylaxis.

One more weird thing: LSD users report very strong tolerance lasting about three days after a dose, to the point where a second dose the day after will do almost nothing. Rat experiments have shown this is definitely because of receptor downregulation and not just enzyme induction. Okay. But LSD is a pretty strong drug. If receptors are so down-regulated that you are essentially on negative one tabs of LSD, how are you remotely normal while the tolerance is in effect? Are people during their periods of LSD tolerance less crazy and creative than normal? What the heck is the 5-HT2A receptor even doing if decreased amounts of it sufficient to render LSD ineffective don’t have noticeable effects on consciousness?

This is my concern about naltrexone as well. Sometimes doctors give naltrexone to help with alcohol addiction, which usually works okay. The theory is that since naltrexone blocks opiates, and opiates power the endogenous reward system, alcohol will seem less rewarding. Fine. But shouldn’t everything seem less rewarding? I always worry that I’m just blocking my alcoholic patients’ ability to enjoy anything at all (of which enjoying alcohol is a subset), but that doesn’t seem to be how it works. This is about when I default to my theory of “receptors read the FDA labels for medications and make sure to only do what they’re supposed to”.

All of this annoys me for a few reasons.

First, psychiatry really doesn’t think about this enough (or sometimes at all). The pharmacology textbooks will tell you how effective a drug is, how long it lasts, how many side effects it has, et cetera, but not whether it’s going to produce tolerance or not. It’s mostly just assumed that it won’t.

Second, groups skeptical of psychiatry are always talking about tolerance and it’s hard to tell whether they’re right or wrong. For example, some people claim antidepressants cause tardive dysphoria – that like the antipsychotics that eventually give permanent repetitive jerking movements, antidepressants can make serotonin receptors permanently undersensitive (or something) and so make depression worse. Other people say that antipsychotics themselves can eventually screw up dopamine receptors in ways that make psychosis worse (though see here). My guess is that these problems don’t arise for most people, but I can’t explain why these things wouldn’t happen.

Third, I think something like this is involved in addiction. Addiction is highly genetic; some people can drink alcohol socially their entire lives and never become alcoholic; other people quickly get hooked. This seems related to the thing where some people are stable forever at their low dose of opiate painkiller, and other people quickly develop tolerance and need to keep increasing it. I’m sure there are other things involved in addiction, but this is probably one of them.

Fourth, how many interesting things are we missing because they’re stupid and make no sense? I don’t know who first discovered that low-dose naltrexone could help potentiate the effect of opiates, but there have got to be other things like that. Forcing your body to become more sensitive to its own chemistry seems like a good alternative to forcing more and more foreign chemicals into it.

Finally, the best drugs seem to be the ones we hesitate to use because they produce too much tolerance. Xanax, opiates, you name it. A version of Xanax that that didn’t produce any tolerance would be a holy grail of anxiety disorder pharmacology. Some way to switch off Xanax tolerance would be just as good.

And a tolerance-free version of heroin would be pretty interesting too – from a purely pharmacological perspective, of course.