I just read another paper about someone not waking up with naloxone after taking a synthetic opioid. The prospect of “narcan-resistant” opioids makes me think of David Letterman. When I was in elementary school I used to watch David Letterman’s show with my mother after school. This wasn’t normal, even for the 1980s, but it was something we could bond over. She set the VCR to record it every night, on the same part of the same VHS tape, and we would watch it together in the family room on a 32-inch square television tube while I simultaneously did my homework. (This explains a lot about my dry sense of humor and lackluster grade school performance.) Letterman’s innovative take on the talk-show format wasn’t raunchy, but much of it was over the head of a fifth grader, although you don’t have to be a sophisticated adult to understand why throwing a watermelon off a building is awesome. He had one recurring bit called “Is This Anything?”, where the curtains would open to one or many performers or juggling or hula hooping or using a grinder against a metal breastplate or all of the above. After a brief (usually less than a minute) performance, Letterman would skeptically ask Bandleader Paul Shaffer, “Is this anything?”
So, when I first read reports of “narcan-resistant” opioids, I had to ask, “Is this anything?” The way the story goes, there are now cases of opioid-poisoned patients who don’t wake up with naloxone. In most cases, the alleged culprit is a fentanyl analog, like carfentanil. The skeptic in me says this is not a thing that is happening. It’s a fact that we are in the midst of a big opioid epidemic that has morphed into a giant fentanyl epidemic. Fentanyl and its analogs have now surpassed heroin and prescription opioids as the leading drugs involved in US opioid deaths. But, the naloxone-resistant opioid story reminds me of other opioid bogeymen that have fizzled. (See: Flesh-eating Krokodil, One-Touch-Can-Kill Fentanyl.) I sought out to investigate whether there are “naloxone-resistant” opioids out there.
You know naloxone. Naloxone competes with agonists for the active binding site at delta, kappa, and- most important for mediating respiratory depression- the mu opioid receptors.
Here is it’s structure:
Naloxone reminds me of the structure of another opioid, oxymorphone. This shouldn’t surprise, because the name naloxone is short for “N-allylnoroxymorphone.” Naloxone is basically oxymorphone, with the addition of a large allyl sidechain on the “out-of-plane” fourth ring of naloxone, attached to the nitrogen
Oxymorphone, of course, is a semisynthetic opioid similar to morphine.
To bring it all back to another opioid that you know, add two acetyl groups to morphine and you have diacetylmorphine, aka heroin.
All of the above structures are often called phenanthrenes, but these are probably better described as “phenanthrene-like.” (A phenanthrene is three fused aromatic rings- rings that are flat with sharing of electrons. As you can see, morphine and naloxone have three in-plane rings, but only one is aromatic.)
The phenanthrenes have markedly different structures than the piperidine opioids (meperidine, fentanyl). We can make a few observations about the structures that explain clinical properties. First, notice how heroin (diacetylmorphine) – the original synthetic opioid – is just morphine with two acetyl groups. Those acetyl groups allow the drug to pass more easily through the blood-brain barrier, making heroin more potent than morphine and providing a better high. Those acetyl groups are removed during metabolism, resulting in 6-monoacetylmorphine, and then morphine. Heroin is metabolized to morphine, which is why heroin is detectable as an “opiate” on the your typical tox screen, unlike almost every other synthetic opioid.
You may also appreciate that naloxone has a large allyl side chain attached to the nitrogen. This chain is what makes naloxone an antagonist, as opposed to an agonist. The allyl chain prevents naloxone from folding the way that it needs to make the opioid receptor function.
Ok, enough with the structures. For a few paragraphs.
A naloxone displacement study shows that 13 mcg/kg of naloxone will bump off a labeled opioid and occupy 50% of brain mu receptors. This corresponds to a dose of about 1.04 mg in an 80 kg man. Like in baseball, 50% is really good. Reversing blockade of 50% of receptors is enough to keep your patient breathing.
As it turns out, 1 mg is a pretty solid clinical dose. The standard recommendations for intravenous naloxone administration have been remarkably stable over the years. The basic idea has always been to start low, and titrate up to a dose that restores breathing but doesn’t cause acute withdrawal or pulmonary edema. The most commonly recommended naloxone dosing progression (mg) was 0.04, 0.4, 2, 4, 10. Most recommended giving a “ceiling dose” of no more than 10 or 20 mg. The reasoning behind the ceiling dose isn’t that higher doses of naloxone in the non-responding patient are necessarily dangerous, just a waste of resources. If your patient isn’t breathing after a high dose of naloxone, move on to something else (probably intubation). Based on what we are seeing out in the world, is there any reason to change our dosing strategy?
Carfentanil is one of many “fentanils” that is showing up on our streets (Lewis Nelson and Jon Cole published a nice discussion of carfentanil that I recommend reading). Fentanyl and analogs are not phenanthrene-like. Fentanyl is a piperidine, because of the presence of a piperidine functional group. (The piperidine is the 6-membered ring with the nitrogen in the middle.)
Carfentanil is said to be 10,000 times as potent as morphine, but this may be true for pain control and not respiratory depression or response to naloxone. A teeny tiny 1 mcg therapeutic dose of carfentanil causes less respiratory depression than an equally therapeutic 10 mg dose of morphine. Opioid potency can be trickier to measure than you’d think. There is no single way to measure it and different endpoints don’t always increase in parallel. We can compare doses of opioids needed to make make mice respond to a hot plate or the dose needed to cause respiratory depression or even constipation.
Response to naloxone may have more to do with interactions with naloxone and the opioid agonist at the receptor. Do fentanyl and its analogs bind to the opioid receptor a lot more tightly than naloxone? The short answer is “no.” The long answer is “NOOOOOOOOOO.” (I make that joke with my kids all the time and they don’t think it’s funny either.) We can use a dissociation constant, Ki. This is the concentration of drug where exactly half of receptors will be occupied by drug.
There is an inverse association between Ki and affinity. The lower the Ki, the stronger a drug associates with the receptor. In many cases Ki does explain difference in potency. For example in one morphine had a Ki of 1.168 nM, while superpotent carfentanil’s Ki was measured as 0.06 nM and hydromorphone had Ki of 0.3654 nM. This seems to have validity, because we know that hydromorphone is several times more potent than morphine and carfentanil is way more potent than both of them. But, wait! The Ki of fentanyl is 1.346 nM- slightly less potent than morphine. This doesn’t fit our experience. We know that fentanyl is about 100-fold more potent than morphine.
The difference in potency between morphine and fentanyl probably isn’t caused by differences in binding at the receptor. The difference is determined by how much drug gets to the receptor. Fentanyl and its analogs pass through the blood brain barrier and get to the critical mu receptors much easier than morphine and phenanthrenes. We can estimate the ability to pass into the CNS by looking at something called the octanol-water partition coefficient. This laboratory measurement tells us about the comparative solubility of drugs in a hydrophilic solution compared to a hydrophobic solution. The higher this octanol-water partition coefficient, the more lipid soluble a chemical is and the more it penetrates the blood brain barrier. The octanol-water partition coefficient for fentanyl is 9950 and morphine is 6. (Carfentanil’s partition coefficient may be as high as 13,500!) The partition coefficient of naloxone is somewhere between that of fentanyl and morphine, and there is also evidence that naloxone is actively transferred into the brain by transport proteins.
Naloxone gets through the blood brain barrier. That’s cool – in a lab. How does this translate to actual real-life opioid reversal? Remember, we only need to block half the receptors with naloxone to restore breathing. There isn’t too much data out there comparing naloxone and carfentanil receptor affinity in vivo. Carfentanil is only approved for animal, not human, use (Elk can be readily reversed with a 2:1 ratio of naloxone to carfentanil). One researcher gave labeled carfentanil to healthy human volunteers. Administration of 2 mg of naloxone blocked 80.6% of receptors at 5 minutes. Administration of 2 mcg/kg blocked 42.6% at 5 minutes.
In the the one and only confirmed case of confirmed occupational exposure to carfentanil, a veterinarian splashed in the eyes and mouth with a dart containing 1.5 mg of carfentanil. After experiencing sedation, that man was treated with 100 mg of naltrexone and had no adverse outcome or return of sedation after being watched for a day. This was a hefty, but therapeutic naltrexone dose.
In an episode of confirmed reversal of fentanyl poisoning, a group of patients were poisoned by fake “hydrocodone” tablets that contained as much as 6.9 mg of fentanyl. (This is massive- remember that 100 mcg is a therapeutic dose for most adults.) In that group, the median naloxone dose needed for initial reversal was 2 mg, with one person getting as much as 8 mg. These doses are high, but within range of reasonable dosing levels. The most interesting thing about this epidemic is that one patient required a repeat dose of naloxone 8 hours after discontinuing it the first time. Fentanyl doesn’t typically last that long. So this may be an example of delayed absorption of the drug- a reminder that pharmacokinetics change in overdose.
A cluster of patients in British Columbia presented following adulteration of crack cocaine with furanylfentanyl. Some of these patients required as much as 3 mg of naloxone for reversal.
What about the reported cases of naloxone-resistant opioids? One 26-year-old intentionally chewed a few fentanyl skin patches. (Remember, even used fentanyl patches have a tremendous amount of drug left in them- a 25 mcg/hour patch has more than 2 mg of fentanyl remaining after it has been used for 72 hours.) He was barely breathing and didn’t respond to 2 mg of intranasal naloxone. However, when given 1 mg of IV naloxone, he woke up and breathed. This represents a failure of intranasal naloxone, probably because of unpredictable absorption.
The evidence supporting naloxone-resistant opioid is usually something like the poison center medical director who reported carfentanil-poisoned patients receiving “up to 18 mg” of naloxone. Poison center data comes from the real world, where details are sketchy. When the poison center gets a call from a hospital they are usually ultimately relying on patient’s history, and patient’s don’t always know what they took. It’s a literal game of telephone. Poison center data is good for spotting epidemiological trends and generating a hypothesis, but we can’t confirm an exposure with a phone call. After all, what if an alprazolam overdose was reported to a poison center as “carfentanil?” That patient may receive 10 or 20 mg of naloxone, but it won’t be effective. That alprazolam overdose will probably be recorded by the poison center as “carfentanil.”
The most convincing case of naloxone-resistance occurred in 2005-6, in Chicago. The authors describe 55 emergency department visits for fentanyl overdose. Patients were successfully treated with doses from 0.4 to as high as 12 mg. Twelve milligrams is a lot of naloxone, but it’s less than what you would give someone if you used the classic naloxone titration: 0.04 mg, 0.4 mg, 2 mg, 4 mg, 10mg.
In the most recent paper, a man was found unconscious after taking carfentanil. He was given 2 mg of IV naloxone in the field and another 2 mg in hospital. He was intubated and not given any further naloxone. He wasn’t able to breath on his own for 31 hours. Does this failure of naloxone represent naloxone-resistance?
I have a few thoughts.
- That is a high dose of naloxone
- I would have given more
- My plan probably wouldn’t have worked
First, 4 mg of naloxone is a substantial dose. In fact, that dose is probably too high for most. Most opioid overdose patients aren’t opioid naive, and 4 mg of naloxone will be a one-way ticket to Withdrawal Town, where you are the reluctantly-appointed mayor and your unfortunate residents all have diarrhea.
Four milligrams is a lot, but if you use a traditional serial dosing algorithm, and you see no response at 4 mg you may try 10 mg. I would have given 10 mg, but it probably wouldn’t have worked.
The reason is found right in the labs. This patient’s initial pCO2 was 138 mm Hg. Anyone that went to medical school or saw Apollo 13 knows that level in itself can cause a coma. This patient was down too long prior to EMS rescue to have much hope of waking up with naloxone. There are other metabolic reasons for coma in the severely opioid-poisoned patients, namely hypothermia. We have all seen a late-presenting opioid-poisoned patient, that we somberly shipped up to the ICU intubated and comatose, only to find out that they awoke three days later with normal mental status. Whether this is a function of hypercapnea, hypothermia, or a drug-induced torpor, these patients remind us to focus on the basics.
So let’s look back at our cases of “narcan-resistant” opioids. All but one of these patients woke up with a high, but not off-the-charts dose of naloxone. The last patient didn’t wake up, but by that point the reason didn’t have much to do with the opioid receptor.
Should the prospect “narcan-resistant opioids” keep us up at night? Letterman sidekick Larry “Bud” Melman said that insomnia was the product of a guilty conscience, caused by “hustling phony diplomas” or “selling tainted meat.”
So – is this anything?
My answer is this: We’re seeing something – a historic shift from heroin to fentanyl and fentanyl analogs. These drugs cause respiratory depression in small doses. However, naloxone has great blood-brain barrier penetration and most of the time the classic dosing strategy 0.04 / 0.4 / 2 / 4 / 10 will work just fine, even if you find yourself closer to the rightward side of that algorithm more often than you used to. You may also find yourself having to re-dose patients who become sedated earlier than you would have expected- especially if they took large oral ingestions.
The most important lesson is not to be so focused on the opioid receptor that you lose sight of the bigger picture. Patients may have several reasons for being unconscious. They may be hypercapnic. They may be cold. Our opioid overdose patients don’t need naloxone- they need to be warm and they need to breathe. Naloxone happens to be a handy, widely-available, syringe-sized means to that end, but it isn’t the only tool we have. We have warmers, bag-valve masks, endotracheal tubes, and mechanical ventilators. We have teams of nurses, techs, respiratory therapists, and physicians. We have our clinical judgement.
That is something.