This blog is a classic example of a bait-and-switch. I drew you in with the non-intimidating cover picture, and I'll even start us off with an illustration about counting candy. However, anion gap acidosis has some heavy points to it, as you'll see. I've done my best to keep it centered on emergency medicine and critical care and give us, as brief of an overview of the things that can cause an elevated anion gap as I can. My hope is that even if all the points don't sink in right now, you can use this as an overview reference for elevated anion gaps.
(The Strawberry Gap)
Whenever I think about the anion gap, I picture a kid with candy. Imagine the parent calling the child over and inquiring about what they're holding. I'm going to use Tyler and Declan as examples in this blog because I can envision an exchange like this between those two.
Tyler: Hey Declan, whatcha got there?
Declan: Nerds - why?
Tyler: How many nerds?
Declan: I don't know.
Tyler: Well, let's count 'em.
Declan had separated his grape and strawberry nerds into his two hands. He had 140 grape nerds in one hand, and 130 strawberry nerds in the other hand. That's pretty close to being even! However, Tyler knows that the Wonka Factory is very meticulous in ensuring that there are always exactly the same number of strawberry and grape nerds in a box.
Tyler: Declan, did you eat any of these nerds yet?
Declan: No, I'm saving them.
Alright, so where did the other nerds go? Turns out, Declan was holding out and had put some nerds in his pockets for later. After counting all of the nerds in his pockets, along with the ones in his hands, a perfect balance of grape and strawberry was achieved! We could even go so far as to say Declan had a Strawberry Gap.
This is how we can look at ion charges - there are some that are easily seen and counted, but others that we have to do a little digging to find. However, one thing always holds true - what is it? The charges always equal out. Does this mean there is no such thing as an anion gap?
What really is an anion gap?
Technically, there is no such thing as an anion gap. This is because there is always the same number of cations as there are anions (this is called the principle of electrical neutrality). However, we don't routinely measure every single type of cation and anion when we get a chemistry panel on a patient. Since we're not counting every single type of ion there ends up being a gap between the cations and anions that we do measure. To sum that up, there is a perceived gap due to how we measure our cations and anions. However, if we were to count all of the ions in the body, there would be no gap (we just usually do not, or perhaps cannot, do that).
To bring this back to our illustration, we usually just look at the easily accessible ions, like the nerds in Declan's hands, where there ends up being a 'gap' between the strawberry and grape (and its usually about 12 if we're keeping things simple for now).
Here are a few methods of measuring cations and anions. Notice that in these common methods, you still end up with a little gap because we're accounting for more cations than anions (the stack on the left is always higher than the stack on the right).
MDCalc has a useful calculator for this - Click Here to check it out. They don't use a 'full' method, but they do throw in the albumin. If you're calculating anything besides the traditional anion gap, always use a calculator. Once you throw the albumin in there, it's not just simple addition and subtraction anymore. You can look under the "Evidence" tab on the MDCalc to see exactly how that's done. As you'll see, it's easy to make a mistake. The traditional formula is simply:
Sodium - (Bicarb+Chloride) = AG. Again, for anything more complicated than that, use a calculator. We typically say that a 'normal' anion gap is approximately 12, but always use a reference for whatever method you're using and what your lab includes in their formula.
Now we see that we never really account for all the positive and negative charges in the body, so we'll always have a gap in-between our stack of positive charges (cations) and negative charges (anions). We've also learned that when an anion gap is expressed when evaluating the anion gap clinically, it might be worth knowing if what we're looking at is a traditional, albumin-adjusted, or 'full' measurement.
As you probably know, our anion gap can grow out of proportion (the gap between measured cations and anions grows). Why does this happen?
Take a look at this elevated anion gap.
Something that we're not routinely measuring is causing there to be a larger difference than usual between the cations and anions we commonly measure (the anion gap has grown). The problem we now face is figuring out what is inside that unmeasured area. We have to account for the unmeasured anion(s) causing the gap to rise. This is where GOLDMARK comes in! Do you remember what each of the letters stands for? No? No one usually does - that's why we included it on our Lab Value Reference Card (shameless plug). Let's take a brief look at each one of these wild anions and get a little more familiar with them!
As we get into these specific GOLDMARK items, the majority of them are toxins. I'll mention their treatment, but I'm not going to go into deep detail. These are patients that should be cared for under the consultation of a toxicologist and with the direction of poison control (1-800-222-1222). The purpose of us going into more detail on these items as emergency and critical care clinicians is to recognize them (rapid recognition is a vital first step) and be at least familiar with the treatment that is indicated (or know where to find out quickly).
Here is how GOLDMARK appears on our lab card:
Ethylene Glycol is what's known as a 'parent alcohol'. The ethylene glycol gives off toxic metabolites, namely glycolate, glyoxylate, and oxalate. Once these metabolites build up in the body, we get end-organ damage (like in the renal system). The key to reducing end-organ damage is to stop the breakdown of ethylene glycol early in treatment (1). How do we do this?
The first step is recognition. How does a patient get ethylene glycol into their system? Here is a list of common agents, although this is far from an exhaustive list.
Automotive Antifreeze (important to differentiate if it was propylene glycol or ethylene glycol)
Car Wash Fluids
Vehicle Brake Fluids
If you're ever in doubt, see if you can identify a product that could have been ingested by the patient and check the label (while this is helpful - don't delay treatment). Bring the bottle with you if it is safe to do so.
The second step is noting the signs and symptoms that might be present.
Altered mental status (similar to ethanol intoxication)
Mydriasis (dilated pupils)
If the patient has anything like coma, seizures, pupillary changes, or Kussmaul respirations, they probably ingested a large amount of poison.
The final step is blocking alcohol dehydrogenase by administering fomepizole (this helps stop the ethylene glycol from being broken down into its toxic metabolites). Thiamine and pyridoxine may help as well. If you're thinking 'isn't this the one that we can give booze for??' - that's correct. However, fomepizole has replaced the need for giving someone a Kettle One IV infusion, since its dosing is much more predictable and standardized. Both fomepizole and Ethanol prevent alcohol dehydrogenase from making ethylene glycol into one of its toxic metabolites. Sodium bicarb can also help prevent metabolites from crossing cell membranes. Hemodialysis is also essential in more severe cases.
As a side note, one very interesting thing that I found in my research on this is that glycolate and glyoxylate (two of the byproducts of ethylene glycol breakdown) are structurally very similar to lactate. This can cause a lab to mistakenly interpret the lab result as lactic acidosis instead of ethylene glycol poisoning (2,3). The research seemed very lab and device-specific, so if this applies to you, you may want to inquire with your lab or manufacturer about this potential error.
"Oxoproline" refers to a structure called 5-oxoproline (Pyroglutamic acid). When we think of this in the context of GOLDMARK, we're actually usually referring to Acetaminophen overdose - why? It's been noted that ingestion of acetaminophen can lead to 5-oxoproline accumulation in the blood (along with elevated acetaminophen levels if the overdose is acute). Now, a toxicologist would probably tell you there are a lot of differences here between pyroglutamic acid and acetaminophen. Today, we're going to keep it simple and think of acetaminophen overdoses when we see this oxoproline in GOLDMARK. This treatment again focuses on limiting the body's ability to break down the acetaminophen into toxic byproducts by using a drug called N-acetylcysteine.
The first step in this process is again recognition, which we've pretty much covered as figuring out if the patient ingested acetaminophen or not (be aware some drugs contain multiple main ingredients and this could be one of them).
The second step is noting the signs and symptoms. Acetaminophen overdoses cause liver failure, but this takes time. Acute signs and symptoms might simply be nausea and vomiting. A patient may be asymptomatic for some time until the liver starts to fail. Lab values will also be run for end-organ damage and APAP levels (APAP is short for N-acetyl-para-aminophenol, which is acetaminophen). Liver failure in this instance is going to have 4 stages, which you can check out here.
The third step is administering N-acetylcysteine (also known as Mucomyst). N-acetylcysteine can be given in a variety of routes, and the severity of the APAP level will usually dictate how it's administered. Activated charcoal may also be given (4).
When you look at GOLDMARK, it initially seems a little strange that there are two different types of lactate mentions (L and D). This first one, L-Lactate, is the most common. This is the lactate that a point-of-care lactate monitor would check, and what is generally inferred when we say that someone has a 'lactic acidosis' (lactate is the anion that results from lactic acid). This form of lactate accumulates due to increased metabolism, or altered metabolism.
Anaerobic metabolism (lack of oxygen) will cause us to accumulate lactate. Also, glycolysis that occurs faster than the Krebs cycle will also cause us to accumulate lactate (due to increased endogenous or exogenous beta stimulation). We can see rises in this form of lactate from giving drugs like albuterol, epinephrine, dobutamine, dopamine, or terbutaline because they all have strong beta stimulation that causes glycolysis to occur more rapidly than the Krebs cycle can accept pyruvate (the pyruvate then turns into lactate - related to the Cori cycle). This starts to make a lot of sense as to why a patient with sepsis, for example, would have increased lactate - the endogenous beta stimulation is very high. This also tells us why patients on drugs that stimulate beta receptors tend to have rising lactates, and why they often fail to clear lactate - the exogenous beta stimulation is very high (5).
I would create or copy a list of what can cause high lactate, but it's probably better to just leave this at the general mechanisms. The individual processes that can cause high lactate comprise a very long list that probably is better left for another time.
Isomers are things that have the same general makeup, but are arranged differently - this is how you can think of L and D Lactate. D-Lactate is present in humans, but normally in very, very small amounts. There is, however, an illness called short bowel syndrome that can cause this D-Lactate to elevate to the point of an anion gap acidosis. Short bowel syndrome refers to the small intestine being reduced in length. This might be due to a bowel resection due to trauma, Crohn's disease, cancer, ischemic bowel, or jejunoileal bypass (a weight-loss surgery that has lost favor due to this complication) (6). If you're like me, you're probably wondering what having a short bowel has to do with D-Lactate accumulation.
Essentially, these patients have carbohydrate malabsorption, leading to carbohydrate fermentation in the gut. Those excess carbohydrates get turned into L and D Lactate. The body deals pretty well with the L-Lactate, but has trouble getting rid of the D-Lactate. This results in altered mental status due to the neurotoxicity of D-Lactate. Due to the method of testing for serum lactate levels, only L-Lactate is commonly measured, and a regular test for Lactate in the blood will likely not return any elevated results unless a specific test for D-Lactate is performed (6). Auto-brewery syndrome can be associated with D-Lactate accumulation as well.
Methanol is another parent alcohol just like ethylene glycol, but if gives off the byproducts formaldehyde and formic acid. While ethylene glycol was toxic to the renal system, methanol causes retinal injury which can even cause permanent blindness. However, they can share symptoms of toxicity such as altered mental status, nausea, vomiting, abdominal pain, coma, seizures, and renal injury (leading up to death) (7). Where is it mainly found?
Common sources of methanol include:
Windshield wiper fluid
Fuels and fuel additives
Some hand sanitizers
Copy machine fluid
All of the treatment algorithms will be the same for methanol as they were for ethylene glycol.
Salicylate toxicity is what we're getting at with this one. Other salicylates include asalicylic acid (acne / wart remover creams) and methyl salicylate, which is commonly supplied as Oil of Wintergreen (used for a variety of purposes). As you can probably guess, Aspirin takes up the vast majority of these toxic emergencies. What are the symptoms?
Altered mental status (ranging from agitation to lethargy)
Hyperpyrexia (Yes - HYPERpyrexa)
Noncardiac pulmonary edema
Salicylates are interesting in their mechanism on the acid-base status. Salicylates directly stimulate the brain stem (respiratory centers) so the patient may start out with respiratory acidosis, but then end up becoming more and more acidic in a metabolic sense. These patients can have huge minute volume demands as they try to compensate for their anion gap metabolic acidosis with a compensatory respiratory alkalosis (8). Beware not to inhibit their breathing, and support their minute volume by whatever means necessary to avoid intubation. To repeat that point, try NOT to intubate these patients.
Like a couple of our other anion gap acidosis issues, sodium bicarb is recommended to alkalinize the urine (watch for hypokalemia).
If the patient has an altered mental status, it's recommended that you have them dextrose even if the serum glucose reading is normal - "Salicylate poisoning inhibits Krebs cycle enzymes and uncouples oxidative phosphorylation. Under these circumstances, we hypothesize that CNS glucose supply is sometimes unable to keep up with demand resulting in hypoglycorrhachia and delirium even in the face of serum euglycemia" (9).
Hemodialysis may be the next step in treatment.
In renal failure, we typically think about the inability of the kidneys to excrete potassium (a sick renal failure patient should always make you suspect hyperkalemia). We also likely think about the inability of the renal system to produce bicarbonate, which leads to a typical metabolic acidosis. However, as renal function worsens (we're not expelling extra waste as efficiently), the body can start to retain some extra anions that we don't usually think about too often. Why are they?
Phosphate and sulfate sound familiar, but what are urate and hippurate? A high urate means there is a decrease in uric acid excretion. Increased hippurate is from an increased amount of Hippuric Acid, which is thought to have some type of antibacterial properties in the urine (I had to look those last two up...). Regardless of what these anions are used for, the important part for us today is that they're occupying space on the right side of our cation-anion columns - causing an elevated anion gap (10).
Patients with this degree of renal failure will likely need hemodialysis. Also, watch for the typical things we see with renal failure patients as well, such as other electrolyte abnormalities (hyperkalemia), and perhaps even the need to replace bicarbonate (10).
Finally, we're ending with a familiar one - ketones. This is pretty different than your friend who's on a ketogenic diet. Normally, ketones aren't a bad thing. A ketogenic process is our body's natural process for using stored energy (fat). However, ketone production can become severely pathological in states like type 1 diabetes, starvation, or alcoholic ketoacidosis. Ketosis occurs in the absence of insulin, which is why insulin is a cornerstone of treatment in these patients - it's all about getting them back to glucose metabolism and closing their anion gap. Caring for a ketoacidosis patient is complicated, so I won't go into all the details here, but here just to mention a few key treatments:
Fluids - these patients are usually profoundly hypovolemia (LR is the best fluid to use because it contains potassium).
Potassium replacement - Insulin lowers potassium, which these patients have usually excreted a large amount of through their urine.
Insulin - Needed to restore glucose metabolism.
Dextrose - Dextrose is usually needed concurrently with insulin to keep the serum osmolality from dropping too far and giving the patient cerebral edema/seizures.
It might seem silly to run insulin and dextrose at the same time, but we'll save that discussion for another time. If you're looking for more information on DKA, check out these blogs and podcasts and the references found in them for the information above:
Bicarb - The Weak Buffer
One thing you might be wondering as we look at all of these different types of anion gap acidosis examples is 'where did the bicarb go?' Good question. We noted at the beginning of the blog that strawberry and grape are always in a perfect balance. Even though some of the cations and anions are hard to find, we never have more of one than another. This begs the question of what happens to bicarbonate when we add anions to the column on the right (the anion side). So, what does happen?
The answer to that question comes from understanding strong and weak ions. Check out this illustration!
Bicarbonate is what is known as a weak ion - what does that mean? It means that bicarbonate can buffer when it needs to. We're intimately familiar with this buffering process:
By buffering through the carbonic acid buffering system, bicarbonate transforms itself and removes itself from our anion side of the equation. This is also why bicarbonate has little evidence (in the absence of a toxidrome or renal failure) to help in cases of anion gap metabolic acidosis (11). This is because the bicarbonate concentration is generally held in place by the other anions because they cannot buffer like bicarbonate can (that's one of the things that makes bicarbonate so special). If someone were to ask you what the buffering process is for chloride, or albumin, or phosphate is, what would you say? You wouldn't say anything, because there isn't one. For ions that cannot buffer, we call them strong ions - this is what is illustrated above. I made the strong ions metal, and the weak (able to buffer) ion wood. To take that illustration a bit further, you can rebuild wood even after it's been crushed. Just like wood, bicarb can be rebuilt after more room has been made for it by buffering the opposite way in the carbonic acid buffering system.
Before we conclude, I wanted to give an honorable mention to non-anion gap metabolic acidosis. This type of acidosis is also called hyperchloremic metabolic acidosis. As you can probably guess by its name, this acidosis results from an excessively high chloride in relation to sodium. Remember how your traditional anion gap calculation was the sodium minus the product of the chloride and bicarbonate? (Sodium - (Chloride + Bicarbonate) = Anion Gap.) To get something called a strong ion difference (SID), all we do is eliminate the bicarbonate from that equation. (Sodium - Chloride = SID.) If the SID is below 38, your chloride is high in relation to your sodium. Why does this matter? Because the more chloride we have, the less room we have for bicarbonate. As a rule, strong ions influence the action of weak ions. To relate this back to our discussion, imagine for a moment that this excess chloride is one of our GOLDMARK items - it has the same action of taking up space on the anion side. Whenever we have excess anions taking up room, what suffers? The other strong ions? No, because they cannot buffer. The only ion that can buffer is your bicarbonate, which is why excess chloride will cause bicarbonate to fall. The main cause of non-anion gap metabolic acidosis (hyperchloremia acidosis) is 0.9% NS administration. This is because 0.9% NS has equal parts sodium and chloride (ions charges must be balanced in a solution). 0.9% NS has 154 mEq of sodium, and 154 mEq of chloride. This is much different than our body, which has a normal sodium concentration of ~140 mEq, and a chloride concentration of ~100 mEq. As we add equal parts sodium and chloride to our body, it starts to shrink the gap between the sodium and chloride, creating this SID acidosis.
I hope this gave you a more intimate knowledge of how ions work, the disease processes that cause our anion gap to elevate, and why the increase in the anion gap can make us so sick. There is a lot more to know about all of those items in GOLDMARK, so please don't take this blog as adequate knowledge of each one of those (the purpose was just to cover the absolute basics of what those things actually are). Keep learning! Even if you don't have all of the tools to fix these patients, recognition and knowing how to navigate initial care will make a huge impact!
1. Iqbal A, Glagola JJ, Nappe TM. Ethylene Glycol Toxicity. [Updated 2020 Aug 14]. In: StatPearls [Internet]. Treasure Island (FL): StatPearls Publishing; 2021 Jan-. Available from: https://www.ncbi.nlm.nih.gov/books/NBK537009/
2. Morgan, T. J., Clark, C., & Clague, A. (1999). Artifactual elevation of measured plasma L-lactate concentration in the presence of glycolate. Critical care medicine, 27(10), 2177–2179. https://doi.org/10.1097/00003246-199910000-00017
3. Brindley, P. G., Butler, M. S., Cembrowski, G., & Brindley, D. N. (2007). Falsely elevated point-of-care lactate measurement after ingestion of ethylene glycol. CMAJ : Canadian Medical Association journal = journal de l'Association medicale canadienne, 176(8), 1097–1099. https://doi.org/10.1503/cmaj.061288
4. Agrawal S, Khazaeni B. Acetaminophen Toxicity. [Updated 2020 Nov 20]. In: StatPearls [Internet]. Treasure Island (FL): StatPearls Publishing; 2021 Jan-. Available from: https://www.ncbi.nlm.nih.gov/books/NBK441917/
5. Levy, B., Desebbe, O., Montemont, C., & Gibot, S. (2008). Increased aerobic glycolysis through beta2 stimulation is a common mechanism involved in lactate formation during shock states. Shock (Augusta, Ga.), 30(4), 417–421. https://doi.org/10.1097/SHK.0b013e318167378f
6. L-lactate and D-lactate - clinical significance of the difference. https://acutecaretesting.org/en/articles/l-lactate-and-d-lactate-clinical-significance-of-the-difference/.
7. Ashurst JV, Nappe TM. Methanol Toxicity. [Updated 2020 Jun 26]. In: StatPearls [Internet]. Treasure Island (FL): StatPearls Publishing; 2021 Jan-. Available from: https://www.ncbi.nlm.nih.gov/books/NBK482121/
8. Runde TJ, Nappe TM. Salicylates Toxicity. [Updated 2020 Jul 13]. In: StatPearls [Internet]. Treasure Island (FL): StatPearls Publishing; 2021 Jan-. Available from: https://www.ncbi.nlm.nih.gov/books/NBK499879/
9. Kuzak, N., Brubacher, J. R., & Kennedy, J. R. (2007). Reversal of salicylate-induced euglycemic delirium with dextrose. Clinical toxicology (Philadelphia, Pa.), 45(5), 526–529. https://doi.org/10.1080/15563650701365800
10. Csaba P Kovesdy, MD, FASN. (2021, April 6). Pathogenesis, consequences, and treatment of metabolic acidosis in chronic kidney disease. UpToDate. https://www.uptodate.com/contents/pathogenesis-consequences-and-treatment-of-metabolic-acidosis-in-chronic-kidney-disease.
11. BICAR-ICU: Sodium bicarbonate therapy for patients with severe metabolic acidaemia in the intensive care unit (BICAR-ICU): a multicenter, open label, randomized controlled, phase 3 trial. Lancet, 2018. PMID: 29910040