Connective tissue diseases, like eosinophilic granulomatosis with polyangiitis (aka Churg-Strauss) and rheumatoid arthritis
Helminths, especially Strongyloides
Idiopathic, which is simply called hypereosinophilic syndrome
Neoplasms, like Hodgkin lymphoma, CML, and some solid-organ cancers
Allergies etc, which includes asthma and drug-induced eosinophilia (don’t forget about DRESS)
Also, Addison’s disease, for some reason
Of interest, helminth infections can cause a Loffler’s syndrome, also known as eosinophilic pneumonia, where you see a peripheral eosinophilia on the blood count and pulmonary infiltrates on the chest x-ray.
Read more: Eosinophils in Chapter 80: Disorders of Granulocytes and Monocytes, Harrison’s Principles of Internal Medicine 19e.
Heart muscle needs blood and oxygen like any other muscle. In the case of the myocardium, coronary perfusion occurs during diastole, when the muscle relaxes. The pressure gradient between the aortic and left ventricular pressures drives blood forward through the coronaries. The coronary perfusion pressure, in the context of chest compressions, is that difference between at end-diastole, for example.
In a normal heart, the gradient is present throughout diastole, so the myocardium gets lots of good, oxygenated blood throughout that part of the cycle (Figure 1). In severe aortic insufficiency, though, the blood from the aorta quickly rushes back into the ventricle during diastole, dropping the gradient between aortic and ventricular pressures to zero (Figure 2). The myocardium, therefore, only has a short period of time, in early diastole, to get that oxygenated blood.
Beta blockade prolongs diastole and therefore prolongs the time that the coronaries aren’t being perfused. Based on that, some cardiologists (including where I trained) avoid beta blockers with the thought that they decrease coronary perfusion in severe AI and therefore promote ischemia. However, beta blockers also decrease myocardial oxygen demand.
So, as with most things in medicine, it’s a balance, and there are few good clinical studies.
Read more: Aortic Regurgitation in Chapter 283: Aortic Valve Disease, Harrison’s Principles of Internal Medicine 19e.
The above Wiggers diagrams are modifications of: adh30 revised work by DanielChangMD who revised original work of DestinyQx; Redrawn as SVG by xavax – Wikimedia Commons: Wiggers Diagram.svg, CC BY-SA 4.0
In my simplified view, it comes down to GABA and NMDA receptors.
Alcohol promotes GABA receptors (which are neuroinhibitory receptors) and inhibits NMDA receptors (neuroexcitatory receptors). Chronic alcohol use upregulates both, but maintains a semblance of balance between the two.
Following alcohol withdrawal, the upregulated NMDA receptors are no longer inhibited by alcohol, so their overall activity increases. The GABA receptors are no longer being stimulated by alcohol. This imbalance of GABA and NMDA signals (with NMDA activity much greater than GABA activity) lowers the seizure threshold.
The acute management and prophylaxis of withdrawal-related seizures, therefore, is typically to promote GABA activity with benzodiazepines, thus restoring the balance between GABA and NMDA signals and increasing the seizure threshold.
Read more: Pharmacology and Nutritional Impact of Ethanol in Chapter 392: Alcohol and Alcoholism, Harrison’s Principles of Internal Medicine 18e.
None! Just kidding; it’s useful for evaluating hypoxia because it can easily rule out hypoventilation as the cause.
For background, the alveolar gas equation is a way of calculating what the level of oxygen is in the alveoli, given the atmospheric pressure (pATM, usually 760mmHg), the fraction of inspired oxygen (FIO2, which is 21% for room air and increases if they’re on supplemental oxygen), the pressure of water vapour in the lungs (pH2O, usually 47mmHg), the arterial CO2 (paCO2, taken from your ABG), and the respiratory exchange ratio (RER, the amount of oxygen exchanged for carbon dioxide in one breath, usually 0.8). The equation is:
The use of the alveolar gas equation is in the A-a gradient, the difference between what the alveolar gas equation says your alveolar oxygen is and what your ABG says your arterial oxygen is. If there’s lots of oxygen getting to the alveoli, then you should have lots of oxygen in the blood. A normal A-a gradient is approximately (age / 4) + 4, so it should be about 9 for a healthy young 20-year old and 24 for an 80-year old.
How is the A-a gradient useful? Well, there are only two things that cause hypoxia with a normal A-a gradient: hypoventilation (not moving enough air), and decreased PiO2 (that is, high altitude). Since I’m rarely doing my ABGs on a mountaintop, the A-a gradient is basically a quick and easy way to rule out hypoventilation as the cause of their hypoxia.
Read more:Adequacy of Ventilation in Chapter 306e: Disturbances in Respiratory Function, Harrison’s Principles of Internal Medicine 19e.
Supplemental oxygen can sometimes cause carbon dioxide to increase to dangerous levels, usually in patients with chronic lung diseases like COPD. I was originally taught that this was due to the extra oxygen blunting their respiratory drive, but it turns out that’s not the whole story. The mechanisms, in order of importance:
V/Q mismatch: Lungs autoregulate their circulation to match ventilation, so that low-oxygen blood only goes to the parts of the lung that have oxygen, which are usually the well-ventilated parts of the lung with lots of air moving in and out. If there’s extra oxygen diffusing to places that are poorly ventilated, it can cause vasodilation within that poorly-ventilated lung. As a result, blood is going to parts of the lung that aren’t well ventilated and can’t blow off CO2. Basically, it increases perfusion to physiologic dead space. Not good for getting rid of carbon dioxide.
Haldane effect: hemoglobin binds both oxygen and carbon dioxide in order to deliver oxygen from the lungs to the tissue and take CO2 from the tissue to the lungs. Unfortunately, when there’s high O2, the Haldane effect means that hemoglobin isn’t as good at carrying CO2. When there is also poor ventilation, this causes CO2 to build up in the blood.
Blunting of respiratory drive: respiratory drive is controlled by oxygen-sensing parts in the periphery and pH-sensing parts in the brain. It was once thought that chronic CO2 retainers lose their pH-based respiratory drive, and require their hypoxic drive to be working well in order for them to blow off any CO2. It turns out that this isn’t the case.
There are five basic processes that result in hypoxemia:
Ventilation-perfusion (V/Q) mismatch: air isn’t getting to the parts of the lung that the blood is passing through. Causes includes pneumonia, asthma, COPD, ARDS, pulmonary embolism, heart failure, and interstitial lung diseases. V/Q mismatches usually respond well to supplemental oxygen.
Right-to-left shunt: blood bypasses the lung altogether. This can happen due to an anatomic shunt in the heart itself as in an ASD, VSD, or PFO or in the lung vasculature through an AVM, or as a physiologic shunt due to severe pneumonia, ARDS, heart failure, or atelectasis. Because blood isn’t getting to the alveoli, supplemental oxygen doesn’t help–all it does it bring O2 to places without blood flow.
Hypoventilation: the patient just isn’t moving enough air. It’s associated with an increase in CO2, and causes include CNS causes (sedation, stroke, tumours), neuromuscular disorders, airway obstruction (COPD, asthma, laryngospasm), and dead space ventilation.
Diffusion defect: oxygen isn’t getting from the air to the blood. Causes include emphysema, PJP, atypical pneumonias, and pulmonary fibrosis.
Low inspired oxygen content: high altitude! And not much else.
If atrial natriuretic peptide (ANP) is released by distension of the atria, then brain natriuretic peptide (BNP) must be released by distension of the… ventricles? Don’t think about it too hard.
BNP and the N-terminal fragment of proBNP (NT-proBNP) are useful for diagnosing heart failure, especially when there’s uncertainty. In fact, at least one trial1 has suggested that it should be used for screening of high-risk patients to identify those who have preclinical heart failure (if such a thing can be said to exist).
Measurements of BNP and NT-proBNP can be used interchangeably, but have different cutoffs. The cutoffs stratify people into three categories: low likelihood of having heart failure (good LR-), high likelihood (good LR+), and intermediate (not terribly helpful).
Besides diagnosis, it can also be used for prognosis as well as for tracking response to treatment. That last one hasn’t made it into the guidelines yet.
As a medical student, I found the various interstitial lung diseases (ILDs) to be horribly confusing. The most common idiopathic ILD is idiopathic pulmonary fibrosis, which is often used interchangeably with usual interstitial pneumonia (UIP). Is there a difference?
Well, yes. UIP is a histopathological description of a lung biopsy that has a specific pattern of fibrosis. (It’s a horrible name, but I was recently told that they tried to change it a decade or two ago and couldn’t come up with anything better.) UIP might also be used to refer to specific findings on high-resolution CT that has a very high correlation with UIP on histopathology. High-res CT is now so good that you usually don’t need the biopsy to know that a patient has UIP.
IPF, on the other hand, is what you call someone with UIP in the lungs if you don’t know why they have it. There are many things that cause UIP on CT and biopsy, including chronic hypersensitivity pneumonitis, connective tissue disorders, and drugs. If the patient doesn’t have any of those diseases or exposures, then it’s said to be idiopathic, and you call it idiopathic pulmonary fibrosis.
This distinction is actually very important, because IPF has a very poor prognosis and has no good disease-modifying treatments, whereas some of the other causes of UIP can be treated.
And now that you understand the difference between IPF and UIP, I’d like to add cryptogenic fibrosing alveolitis to your vocabulary—the fancy British way of saying IPF.
Anion gap metabolic acidosis is one of those classic differentials that we all learn. My favourite approach is not the MUDPILES that I learnt in medical school, but rather a differential based on the four basic things that can cause an increase in unmeasured anions:
Ketoacidosis: this includes your DKA, starvation ketoacidosis, etc.
Lactic acidosis: often from poor perfusion, as in the various causes of shock.
Renal failure: kidney failure alone can raise your anion gap, particularly from decreased clearance of NH4+.
Toxins: ethanol, methanol, ethylene glycol, acetaminophen, and aspirin are some big ones here.
And, for interest’s sake: although the most common form of lactic acidosis is from L-lactate (that’s what the lab reports as “lactate”), there’s also the rare D-lactic acidosis. It’s caused by gut bacteria in patients with jejunoileal bypass, short bowel syndrome, and intestinal obstruction. The bacteria create D-lactate rather than our normal human L-lactate.
Read more: Metabolic Acidosis in Chapter 66: Acidosis and Alkalosis, Harrison’s Principles of Internal Medicine 19e.