When I talk about toxic alcohols causing anion gap metabolic acidosis I emphasize that these patients have large anion gaps. When you see an anion gap of 16 or 20 think uremia and lactic acid, not methanol. The cases I have seen have almost all had gaps greater than 25 and typically they run in the 30s.
But I have never heard of or seen a gap as big the one that came into the ICU this week-end:
To summarize the data from above. A patient was admitted with an anion gap of 65 that went up to 70 in the next 6.5 hours. I can visualize the ER doc reading the first chemistries, freaking out and re-ordering them, assuming that if the anion gap is greater than the chloride it must be a lab error, hence the repeat labs at 150 minutes after the first set.
Think about that, the anion gap was larger than the chloride concentration.
What kind of alien infestation causes numbers like that?
So when working up a large anion gap one tries to explain the anion gap. The lactic acid was only 34.5 mmol/L. So this patient has the highest lactic acid ever, yet it only covers half the gap. She also had serum ketones that remained positive at a 1:8 dilution and a creatinine of 14.
So is that it? Lactic acidosis, ketosis and uremia for an anion gap of 70? I sent off a D-lactic acid and a 5-oxoproline level, cause what the hell, when's the next time I'm going to see an anion gap of 70. A toxic alcohol screen was sent.
To compete the picture the ABG was: 6.95/13.4/187 with a measured bicarb of 3. With numbers this crazy a trip to the Henderson-Hasselbalch formula is probably not a bad idea:
So 6.97 is pretty close to the 6.95 measured, so no lab error at least in the ABG. Running Winter's formula (1.5 x HCO3) + 8 ±2 gives a predicted pCO2 of 13, so no respiratory component to the metabolic acidosis.
The next step is to look for toxic alcohols while waiting for the assay to come back from toxicology lab at Children's Hospital of Michigan. The osmolal gap calculation should be greater than 10 in the presence of methanol, isopropyl alcohol or ethylene glycol. The measured osmolality was 327.
|MedCalc also has a sweet osmolar gap calculator|
So no significant osmolal gap rules out a toxic alcohol. This was confirmed by the toxicology screens that eventually came back. Given the anuric renal failure, profound intractable acidosis and unknown anions still unaccounted for we initiated CVVHD. You can see the effect that had on her bicarbonate and creatinine.
The patient has since recovered and we have learned that she was on metformin so we are toying with metformin induced lactic acidosis as the etiology. Any other thoughts?