Let’s get this out of the way, shall we?
Not a word, you say? Look it up.
The exhortation to avoid tramadol is met either with a sage nodding of the head, or with one of two other responses, occasionally draped in hostility:
“But I’ve (taken it/prescribed it) and seen it work. Tylenol, ice packs and a dose of SuckItUp™ for everyone? Is that your plan, you anti-opioid zealot?”
“Why not? We need analgesic options. And besides, tramadol has two —TWO!— mechanisms of action. I learned that at a CME event last month (where the strip loin and Bordeaux were fantastic, btw …)
Let’s also get this out of the way: Tramadol can help some people (although “work” is the wrong construct, as we’ll discuss later). And yes, it does have two mechanisms of analgesic action.
So, why not use it? Read on.
Background and Pharmacology
Tramadol was developed in 1962 by the German pharmaceutical company Grünenthal, the same good people who graced humanity with thalidomide. For some reason it wasn’t licensed in Germany until 1977, and in the U.S. until 1995. But over the past decade or so, it has garnered its share of medical devotees. More than 40 million tramadol prescriptions are now issued each year in the United States alone.
Tramadol is generally viewed as a weak opioid, more or less on par with codeine. It’s therefore often perceived to be a more appropriate initial opioid than, say, morphine or oxycodone. But I like to think of tramadol as what would happen if codeine and Prozac had a baby, and that baby grew into a sullen, unpredictable teenager who wore only black and sometimes kicked puppies and set fires.
In fact, tramadol itself is hardly even an opioid. It binds to mu opioid receptors so weakly (with an affinity about 1/6000th that of morphine) that this doesn’t realistically contribute to its analgesic effect. What it does do, however, is inhibit the reuptake of serotonin and norepinephrine, more or less like venlafaxine. In fact, look at the structures, squint hard, rub your eyes, and you can start to see the resemblance:
And because tinkering with monoamines in the dorsal horn of the spinal cord modulates pain transmission, that is tramadol’s analgesic mechanism of action #1.
So why is tramadol lumped in with opioids? Because one of its metabolites (O-desmethyltramadol; ODT) is a genuine agonist at mu opioid receptors, binding with an affinity 300 times that of tramadol itself. (Note: ODT should not be confused with founding member of the Wu-Tang Clan, ODB (Ol’ Dirty Bastard), who collapsed and died in 2004 with cocaine and — I kid you not — tramadol in his system.)
What does all this mean?
What it means is that while tramadol is an SNRI (serotonin-norepinephrine reuptake inhibitor), its metabolite ODT (also sometimes referred to as “M1”) is an opioid. This, my friends, is the basis of the “dual mechanism of action” claim that has led so many clinicians to drink the tramadol Kool-Aid.
Except the story doesn’t end there. The conversion of tramadol to ODT happens in the liver – obviously. That’s where one of the cytochrome P450 enzymes (CYP2D6) removes a methyl group, magically converting an SNRI (tramadol) into an opioid (ODT).
And this is where tramadol starts falling apart.
Why? Because CYP2D6 activity varies tremendously among individuals and across ethnic groups. Somewhere between 3-10% of Caucasians are “poor metabolizers”, meaning they have no functional CYP2D6. Most exhibit an ‘intermediate” or “extensive” (that is, normal) metabolizer phenotype, while some (2-10% or more, depending on the country of origin) are “ultrarapid metabolizers”.
Check out how CYP2D6 genotype influences ODT after a fixed dose of tramadol. The poor metabolizers (“PMs” in the figure below) generate virtually none.
All this pharmaco-gobbledygook has an important clinical implication: When you give a patient tramadol, you have no idea whether you’re giving an SNRI or an SNRI-opioid combo. In other words, prescribing tramadol is like prescribing venlafaxine and morphine in an unknown ratio. And seriously, why do that?
Let’s step back for a moment and reflect on the goal of drug therapy for pain. When we prescribe an analgesic, the goal isn’t simply pain relief. The goal is to impart more benefit (including analgesia, improved function, improved quality of life) than harm. Sounds obvious, yes? Every time you start a new drug for pain, you’re conducting an experiment on an individual patient, aiming to help more than harm. Into that experiment, tramadol introduces needless pharmacologic unpredictability.
(The same is true, by the way, of codeine, which does nothing until CYP2D6 converts it to morphine. In other words, giving a known dose of codeine is the same as giving an unknown amount of morphine. It’s yet another inherently irrational opioid.)
Can tramadol “work” (meaning help more than harm)? Sometimes. But it seems to me that anyone willing to roll the tramadol dice hasn’t given the drug’s pharmacology quite enough thought.
But wait, there’s more!
As if its unpredictable kinetics weren’t enough, tramadol has plenty of other baggage worth mentioning. In no particular order:
Tramadol seems especially prone to cause seizures, presumably on account of the monoaminergic effects of the parent compound. In this series of patients presenting with a first seizure, tramadol was the leading drug-related cause, with half of patients taking 100 milligrams or less per day. And in this series of 71 tramadol overdoses (many involving coingestants), seizures occurred in 11%.
While tramadol causes seizures and an opioid toxidrome following overdose, it doesn’t seem to cause much in the way of serotonin toxicity. But combined with other serotonergic drugs, it can. The risk of this seems genuinely low, but some reports are quite compelling. To be fair, tramadol is hardly the only opioid associated with serotonin toxicity.
But it’s just one more thing to worry about, particularly given the prevalence of antidepressant use.
Other drug interactions
Imagine your patient is taking 150 milligrams of tramadol a day for chronic pain and is doing well. Then you start bupropion for smoking cessation. Or paroxetine for depression. Or terbinafine for a fungal nail infection. Or maybe celecoxib for ongoing pain. And a few days later, the patient is back in your office complaining of new-onset abdominal pain, diarrhea and insomnia. That’s because these new drugs inhibit CYP2D6 to varying degrees. In a patient on tramadol, they’ll effectively turn off ODT production and precipitate opioid withdrawal (assuming the patient has the genetic good fortune to convert tramadol to ODT in the first place). This phenomenon—turning a CYP2D6 extensive metabolizer (EM) or ultrarapid metabolizer (URM) into a poor metabolizer (PM)—is referred to as phenocopying, and it’s easy to mistake for something else.
I’m not exactly sure why tramadol causes hypoglycemia, but it does, as evidenced by at least one well-designed observational study and several well-documented case reports. Here’s the money shot from one particularly striking report:
In a series of almost 3000 patients, tramadol was associated with hypoglycemia in almost half of patients with DM1, about 17% of patients with DM2, and almost 5% of those without diabetes.
Interestingly, hypoglycemia has also been described with venlafaxine. What’s the mechanism? A quick literature review reveals a lot of hand-waving about increased peripheral glucose uptake and reduced hepatic gluconeogenesis, but honestly I have no idea. The important thing is to be aware of it, especially the next time you encounter refractory seizures in a patient on tramadol.
Dependence, addiction and death
It should come as no surprise that, as an SNRI and sometimes-opioid, tramadol can cause physical dependence. Since pain medicine is about experiments that yield anecdotes, here’s one of mine about a man who cold-called me at my office. He’d been on tramadol for more than a year, after opting for it over Percocet for a shoulder problem (he wanted to avoid opioids). Attempts to self-taper from 150 milligrams per day were met with crippling insomnia. Like countless other people on tramadol, he needed it simply because he’d been taking it. (The denouement? We worked with a compounding pharmacy, dropped his daily dose by 5 milligrams every week, and he took his last dose of tramadol about 6 months later. He tells me he’s never felt better.)
Tramadol addiction is a huge issue in many parts of the world, particularly the Middle East and Africa. If you happen to have access to the Wall Street Journal, this story tells the tale, but the upshot is that tramadol is not subject to the same international controls as, say, morphine. India and China export massive amounts of cheap, generic tramadol around the world. Here’s another story describing the growing popularity of tramadol in Egypt where, as with many other parts of Africa and the Middle East, CYP2D6 activity tends to be on the high side.
Tramadol-related deaths have been on this rise in many parts of the world, and unsurprisingly correlate with prescription volumes in developed countries. It’s worth taking a look at what happened to tramadol utilization and tramadol-related deaths in England and Wales after the government finally classified it as a scheduled drug in 2014:
What’s the Bottom Line?
The bottom line is that tramadol is a weird drug with unpredictable kinetics and a litany of dangerous toxicities and drug interactions.
- Can it sometimes help people with pain? Yes.
- Is it a rational drug to initiate? No.
- Does criticizing tramadol sometimes invite hostility? Yes.
And with that . . . Release the hounds!