Tox and Hound – In the Beginning

by Diane Calello

Risk Assessment in Acetaminophen Toxicity

In the beginning God created the heavens and the earth. Darkness covered the formless earth, wind swept over the waters. And God said, “Let there be acetaminophen”.

And so it was. And it was good.

Then man discovered acetaminophen and, realizing he was naked without it, covered himself with an abundance of it. And there, in the Garden of Glutathione, arose a Serpent which man called “NAPQI”. Man quickly began to realize that wherever he went with his acetaminophen clothing, there the Serpent followed. Man felt tired, and his skin took on a yellowish cast.

A woman appeared out of nowhere with an IV bag and a potion of foul-smelling drink. She approached man, somehow plucked a rib out of his thorax, and called herself N-acetylcysteine (or NAC for short). Man felt better. The Serpent was foiled. And man followed NAC wherever she went, sometimes even when he didn’t need her.

And it was very good. Except, when it wasn’t.

Seemingly since the beginning of time (or at least since the 1970s) acetaminophen (APAP, or paracetamol to most of the world outside the US and Canada) has confounded us with its seemingly predictable mechanism of toxic injury but countless nuances on how to assess and treat patients with the age-old analgesic. As a wise colleague (&Howard) once said . . .

Fundamentals

We can agree on the fundamentals. N-acetyl-para-aminophenol (APAP) exerts its therapeutic mechanism of action via inhibition of prostaglandin E2 synthesis. This is similar to NSAIDs and salicylates, with the absence of the anti-inflammatory and anti-platelet effect. Under normal conditions of therapeutic dosing, 90% of APAP undergoes glucuronidation and sulfation to non-toxic, excretable metabolites. However, as the dose increases, the P450 isoenzymes (predominantly 2E1, but also 3A4, 2A6, 1A2) metabolize the excess APAP to NAPQI (N-acetyl-para-benzoquinone imine), causing free radical-mediated tissue damage, particularly to hepatocytes in Zone III. Endogenous detoxification by glutathione kicks in but then becomes depleted, and more tissue damage ensues. So starts an injury first heralded by transaminase elevation (usually within the first 24 hours, but sometimes as late as 36)^​1​, which can then progress to coagulopathy, hyperbilirubinemia, renal injury, and in severe cases, fulminant hepatic failure, cerebral edema and death. N-acetylcysteine counteracts this cascade by replacing glutathione, repairing oxidant tissue damage, and to a minor degree, inhibiting the metabolism of NAPQI in the first place. The Rumack-Matthew nomogram predicts which patients are at risk for hepatotoxic cellular damage based on an APAP concentration drawn between 4 and 24 hours after a single, acute ingestion.^​1,2​ Those over the line, e.g, >150 mcg/mL at 4 hours (in the US, Canada and Australia ), need NAC treatment, and those under do not. Treatment should be initiated within 8 hours of ingestion for maximum efficacy.

Easy. Except, when it isn’t.

For starters, let’s establish that the nomogram was developed to predict risk in single, acute APAP overdoses. Not chronic overdoses. Not staggered overdoses. Not repeated supra-therapeutic ingestions. Not the case of “I have no idea what happened because the history is completely unreliable, but I have a detectable level”.

Derived by charting plasma acetaminophen concentrations over time in untreated patients, and determining which ones developed “hepatotoxicity” (defined as AST>1000), the nomogram is trusty and reliable when assessing patients who take a lot of acetaminophen tablets all at once, or at least within 4 hours. Unfortunately, this reliability vanishes in other scenarios. One such example is chronic overdose, also called a repeated supra-therapeutic ingestion (RSTI). These patients take a little too much for a little too long – think 200mg/kg/day for 24 hours or 150mg/kg/day for 48 hours – and can clearly develop hepatic injury. This scenario is the most common cause of liver transplantation in the US. A serum APAP concentration in these cases is just supportive evidence of ingestion and may be falsely reassuring if low. The decision to treat is based on history. Similarly, the nomogram fails when there has been a staggered ingestion: multiple high doses over about 4-24 hours, either from intent to harm oneself or in response to severe acute pain (think toothache). The nomogram simply can’t help here, and attempts to wedge patients into the nomogram result in failures. Yes, the nomogram can fail.

Speaking of failures . . . In 2012, the UK Medications and Healthcare Products Regulatory Agency (MHRA) took a hard look at the behavior of the nomogram. This was in response to reported failures in patients whose initial APAP concentration was reassuring, treatment was not given, and they died. The use of the nomogram in the UK at the time was a bit different though. They utilized a 2-tiered approach with a “standard” risk threshold (200mcg/mL at 4 hours) and a “high” risk threshold (100mcg/mL). Risk factors placing the patient in the latter category included chronic alcohol abuse, concomitant medications inducing CYP2E1 and other enzymes, and starvation. Following an exhaustive review process, the MHRA recommended a revised use of the nomogram, now recommending treatment for any patient crossing 100mcg/mL at 4 hours, regardless of risk factors, or ANY patient with suspected staggered ingestion.^​3​

Should we do the same? Probably not, and for a few reasons. First, the nomogram is designed with sensitivity, not specificity, in mind – so we are already over-treating. Let’s not forget, the 150mcg/mL treatment line was arbitrarily lowered by 25% from the original 200mcg/mL threshold specifically to add a margin of safety. Approximately 60% of patients over the 150 mcg/mL treatment line would NOT develop hepatotoxicity if they were not treated. After adopting the new lower threshold in the UK, there was a predictable rise in admission rates, costs, and NAC adverse reactions.^​4​ A recent evaluation of this change estimated that adopting the lower threshold in the US would cost $45 million in extra admissions, for seemingly minimal (if any) benefit.^​5​

Nonetheless, this perspective should prompt a deeper dive into a patient’s history before applying this tried and true “charty graph thing”. It can be tremendously helpful. But it’s not a crystal ball. Use as directed.

Jumping the Gun: Don’t do it.

A 26-year-old patient presents to the ED shortly after a reported single overdose of 17g of acetaminophen alone. She appears well. An APAP concentration drawn approximately 2 hours after ingestion is 134 mcg/mL.

Levels obtained prior to the 4 hour post-ingestion mark are essentially noise. Although it is tempting to draw an APAP level along with all of the other initial labs in patients who present early, several studies have demonstrated that levels drawn prior to 4h have a false negative rate as high as 6% to predict hepatotoxicity.^​6,7​ The test characteristics (sensitivity and specificity) are particularly bad in the 0-1 h timeframe. The one useful nugget about an early level is that if it is undetectable after 2 hours, an at-risk acetaminophen exposure can probably be ruled out. Maybe. At least according to the study authors.^​6​ Otherwise . . .

What about co-ingestions?

The performance of the premature acetaminophen level plummets further in the context of coingestants causing delayed absorption – namely, opioids and antimuscarinics.^​8​ In one study, coingestants took the sensitivity of the nomogram from 96% and lowered it to 91% when the coingestant was an opioid and 86% when it was an anticholinergic.^​6​ Co-ingested opioids and anticholinergics also increase the likelihood of “line crossing”, the phenomenon where a patient with a 4-hour level below the treatment line has a subsequent level above the treatment line.^​9,10​ Groan. Although again there are no formal recommendations or data to support changing the approach, at the very least, co-ingestions like these should make us throw out those early levels even more.

Sustained Release Preparations and the Double Hump

Sustained release preparations, made up of a medication bilayer meant to deliver two doses over a longer period of time, come in different formulations with different dosing amounts. Sustained-release preparations can also diminish the sensitivity of levels obtained prior to 4 hours, and some patients do cross the nomogram.^​11,12​ In some cases, this can result in a second or late peak, or “double hump”^​13​, although the most common cause of this, aside from the Bactrian camel (above), is a truly massive ingestion (>50g) or coingestion.^​14​ Does SR-related nomogram crossing actually matter? Unclear, because all of the patients reported in the literature were treated once the nomogram crossing was discovered. The most sensible approach is to obtain a second level 4-6 hours after the first 4-hour level, if it falls below the nomogram. ^​12​

CYP Inducers

There is theoretical concern about patients chronically on CYP 2E1 inducers, specifically isoniazid, as well as non-specific CYP inducers such as ethanol, phenytoin, phenobarbital and carbamazepine. NAPQI production is increased in the presence of these substances, but clinical evidence to support a modified risk stratification does not exist.

Repeated Supra-therapeutic Ingestion

It’s hard to assess these, due especially to inter-individual variability in the development of hepatotoxicity. In general, the risk is proportional to the amount of acetaminophen taken daily over time. Conditions which favor increased production of NAPQI or depletion of glutathione stores do predispose to hepatotoxicity. Examples include the febrile infant, chronic alcoholics, patients in catabolic states such as starvation and post-surgery, and those on CYP inducers as described above. The best approach is to treat if there is any evidence of hepatic injury, and treat if the history is consistent with >200mg/kg/day for 24 hours, >150mg/kg/day for 48 hours, and less if the exposure is longer. In children, the proposed threshold is 100mg/kg/day for 72 hours. Of note, patients such as these with lower plasma APAP concentrations are also at higher risk of NAC-induced anaphylactoid reactions. ^​15,16​

The “Hell if I Know” history of ingestion

Totally unreliable history? If AST is elevated, treat. If APAP is detectable, treat. If neither are true, don’t treat. The same holds true for late-presenting overdoses.

Getting Fancy: Other Biomarkers

Several other adjunctive tests have been proposed to determine which patients are at risk, particularly those in the harder-to-define categories like chronic ingestion, or the patient with undifferentiated liver failure where it’s not clear acetaminophen is to blame. These include APAP-protein cellular adducts, micro RNA, High-Mobility Group Box 1 (huh?) Keratin-18, the elimination half-life of acetaminophen, and the ratio of APAP-protein adducts and alanine aminotransferase (ALT) at the time of presentation.^​16​ As exciting as having another measure of risk is, at present these are all either difficult to obtain routinely and/or do not have robust enough data to justify their use in clinical care. But hope springs eternal.

Conclusion

Acetaminophen toxicity has been with us seemingly since the origin of the world, with questions “more abundant than the sand on the seashore”. (Genesis 22:17) We see our nakedness. Our evolution continues.