11/9 Presby report: a moste curious case of metformin-associated lactic acidosis

Late post because I’m on service (sorry!).

A huge thanks to recent Penn IM alum Geoff Bass for giving a great talk on a middle-aged diabetic man who came in with abdominal pain and was found to have metformin-associated lactic acidosis (MALA).

Screen Shot 2017-11-13 at 2.18.09 PM.png

Lactate metabolism is intertwined with that of glucose and pyruvate

Hyperlactatemia occurs via one of four mechanisms:

  • Tissue hypoxia
  • Aerobic glycosis (aka glycolysis that depends on factors other than tissue hypoxia (think high-epinephrine states)
    • Sepsis
    • Severe asthma + beta-agonist use
    • Cardiogenic or hemorrhagic shock
    • Pheochromocytoma, cocaine
  • Impaired oxidative phosphorylation
    • Drugs: ARVs, propofol (PRIS), metformin, toxic alcohols, salicylates, cyanide
  • Liver dysfunction
    • the liver normally clears ~70% of whole body lactate, so acute liver dysfunction can lead to elevated serum lactate; but chronic liver disease (ie cirrhosis) usually doesn’t (in the absence of sepsis or some other driver)

Metformin toxicity is generally rare, and usually only occurs in the context of other predisposing conditions: AKI/CKD, liver disease, alcohol use, concurrent critical illness, heart failure

It is a clinical diagnosis, as there is no one test that confirms the diagnosis; you can send off a metformin level, but it takes a while to come back and is generally unhelpful unless it’s totally negative (in which case you should look for other causes)

MALA is mostly treated with conservative measures:

  • Gastric decontamination
  • Supportive care
  • Hemodialysis
    • Consider if lactate >15-20, pH <7.1, or other conditions such as shock, renal failure, etc


  1. DeFonzo R, Fleming GA, Chen K, et al. Metformin-associated lactic acidosis: Current perspectives on causes and risk. Metabolism, 2016 Feb; 65(2): 20-29.

  2. Kraut JA, Madias NE. Lactic Acidosis. NEJM,  2014 Dec 11;371(24):2309-19.


10/16 HUP Report – The Opioid Epidemic

Today we were joined by Dr. Jeanmarie Perrone and Dr. Arun Gopal to discuss the opioid epidemic, its presence in Philadelphia, and how we can take steps to help our patients at HUP. Images in this post are courtesy of Dr. Perrone.

We all know that the opioid epidemic has become a very serious issue nationally. Here is some data about opioid deaths in Philadelphia:


As more of our patient population is effected by opioid use disorders, we need to give some thought to how to best care for them. One important point is that our goal on inpatient services does not always need to be to convince each and every patient to commit to abstinence, but rather we should support them through withdrawal to enable them to complete treatment for whatever complication is requiring inpatient care (rather than ignoring or incompletely treating their withdrawal, which frequently drives patients to leave against medical advice).

Buprenorphine is a medication that is a partial opioid agonist and partial antagonist, which helps to support patients comfortably and safely through withdrawal. Here is a graph showing the pharmacology:


Patients can be evaluated with the COWS score (see below) to quantify symptoms of withdrawal. Our hospital’s protocol suggests that if a patient has a COWS score of >/12, they should be given buprenorphine with further dosing pending repeat evaluation.


Appropriate treatment of opioid withdrawal is critical to the care of our patients, so please pay close attention to this protocol when dealing with this patient population!


8/8 (HUP): CCB toxicity

Thanks to Dr. Francis DeRoos for walking us through a case of a young woman who presented with shock, and was ultimately found to have overdosed on calcium-channel blockers.

First, we talked about the importance of (1) the physical exam and (2) the EKG in diagnosing and risk stratifying patients with toxic ingestions. Assessing the skin, pupils, and sweat/lack thereof can give you valuable clues. See the toxidromes post from a few months ago for more info.

Here’s a cute chart that illustrates that:

Screen Shot 2017-08-08 at 3.24.24 PM

Common toxidromes (source: http://www.sketchymedicine.com)

We also talked about various causes of drug-induced bradycardia:

Screen Shot 2017-08-08 at 3.43.41 PM.png

Lithium tends to cause bradycardia most commonly in patients with underlying cardiac disease. Amiodarone has particularly been known to cause bradycardia in concert with certain of the new HCV medications (sofosbuvir or daclatasvir), including slow VT!

CCB toxicity can cause a variety of ECG changes, including sinus bradycardia and varying degrees of AV blockade. Remember to look at the rhythm strip carefully to make sure you’re not missing complete heart block!

Treatment options for CCB toxicity

  1. Calcium: may improve inotropy and blood pressure. You can give either calcium gluconate (short acting, can give peripherally but lower Ca content) or calcium chloride (3x the Ca content of Ca gluconate, must give centrally given risk of tissue damage w/ extravasation).
  2. Pacing: unlikely to be of much benefit even in hemodynamically significant bradycardia, given that the CCB is still bathing cardiomyoctes with its negative chronotropic effect.
  3. Vasopressors: something (like norepinephrine) with inotropic and vasoconstrictive effects is best, but there’s no trial proving the superiority of one vs another. Large doses may be needed.
  4. High-dose insulin/euglycemic therapy: CCBs block the calcium-dependent release of insulin from pancreatic beta cells, while at the same time increasing glycogenolysis. Insulin is also postulated to have a positive inotropic effect, especially in these patients in whom CCBs are exerting a negative inotropic effect. Massive doses of insulin (to the tune of 0.5-1U/kg/hour) may be needed given (along with dextrose) to maintain euglycemia.
  5. Lipid formulations: intravenous lipid rescue (initially used for bupivicaine toxicity) can be used as a ‘lipid sink’ to bind up unbound CCB. This is not without risks, as it can cause lipemic serum and higher rates of pancreatitis!
  6. ECMO/CPB: with refractory shock, there is also a role for putting patients on ECMO while using adjunctive therapies (above) to support them

Lastly, the paper of the day– which suggests that earlier Lasix administration in patients with decompensated heart failure may have a mortality benefit! See their central illustration, which suggests that there is a critical period (<60 min) where early diuretic administration might be particularly beneficial:

Screen Shot 2017-08-08 at 12.12.45 PM.png


  1. Matsue, Y. et al. 2017. Time-to-Furosemide Treatment and Mortality in Patients Hospitalized With Acute Heart Failure. Journal of the American College of Cardiology. 69, 25 (2017), 3042–3051.
  2. Kerns, W. 2007. Management of β-Adrenergic Blocker and Calcium Channel Antagonist Toxicity. Emergency Medicine Clinics of North America. 25, 2 (2007), 309–331.
  3. Life in the fast lane. Calcium Channel Blocker Toxicity.


7/6 (HUP): common toxidromes

We went through an interesting case of a middle aged man found unconscious on the street, who was found to have a large anion and osmolar gap, and was ultimately found to have ethylene glycol toxicity. Along the way, we reviewed several important points:

  1. The importance of getting EKGs on anyone suspected of a toxic ingestion, as it can give you tons of valuable information and can help you diagnostically and therapeutically. You can and should trend EKGs to look for dynamic changes in QRS, QT and ST intervals!
    1. For example: terminal R wave in aVR –> very suggestive of TCA toxicityScreen Shot 2017-07-10 at 11.25.51 AM.pngScreen Shot 2017-07-10 at 11.37.26 AM.png
      Normal aVR
    2. QRS width in TCA poisoning is correlated with badness: QRS >100 is predictive of seizures, and >160 predictive of ventricular arrhythmias. You can and should trend QRS duration on serial EKGs to watch it shrink as you administer sodium bicarbonate!
  2. Toxic alcohols can create an anion and osmolar gap (EXCEPT isopropyl alcohol)

Screen Shot 2017-07-10 at 11.24.33 AM

Remember that sOsm and the anion gap can vary depending on when they’re drawn after the ingestion!

Screen Shot 2017-07-10 at 11.33.17 AM

Lastly, a note about the UDS

1. The ‘level 1’ drug screen (aka DA9) tests for many common drugs: amphetamines, benzos, cocaine, opiates, PCP, THC, methadone, barbiturates, and alcohol. Within those classes, certain drugs may not be tested for (for ex: fentanyl isn’t routinely tested for under opiates). Toxic alcohols are a send out.

2. If you’re looking for a specific drug, ask for a level 2 screen, which tests the sample against a library of 400-500 drugs. You have to have a specific drug(s) in mind- this isn’t just an automatic screen against every compound that wasn’t tested in the level 1!



  1. Yates, C. and Manini, A. 2012. Utility of the Electrocardiogram in Drug Overdose and Poisoning: Theoretical Considerations and Clinical Implications. Current Cardiology Reviews. 8, 2 (2012), 137–151.
  2. Kraut JAXing SX. Approach to the evaluation of a patient with an increased serum osmolal gap and high-anion-gap metabolic acidosis. Am J Kidney Dis 2011;58:480-484
  3. Burns, E. Tricyclic overdose (Sodium-channel blocker toxicity). Life in the Fast Lane. https://lifeinthefastlane.com/ecg-library/basics/tca-overdose/. Accessed July 2017.