The ICU Gotchas: Pulmonary hypertension and right heart failure (part I)

[See the previous Gotcha here: peri-intubation arrest]

The patient is a previously healthy 55-year-old male who presented with shortness of breath and chest pain. A CTA in the ED reveals a large saddle pulmonary embolism. Although his vitals are relatively normal, his troponin and BNP are elevated, and the ECG shows signs of right heart strain. A formal transthoracic echocardiogram suggests grossly preserved right and left heart function, but an RVSP of 90 mmHg. He is admitted to the ICU for observation.

A few hours later, his respiratory status appears worse, and he becomes hypotensive, requiring a norepinephrine infusion. Overall, his clinical picture remains fair, so you decide to try him on BiPap.

Initially, his respiratory effort seems to improve. However, his blood pressure dips, and during an attempt at central line placement, his SpO2 falls to the 80% range. Before more measures can be taken, his BP suddenly bottoms out, and his pulse is lost. A quick bedside echo shows a massively enlarged right heart. CPR and ACLS are fruitless.

What happened?


It’s easy to read this story as a cautionary tale about submassive pulmonary embolism, and it is. But it also tells a larger story—a story about pulmonary hypertension, a condition that warrants a certain amount of fear in critical care. Let’s understand why.

Reviewing the big picture

A comprehensive look at pulmonary hypertension (PH) is a matter for another post—if not a book or two—but a general framework will be helpful.

The WHO classification system remains the most common taxonomy, and I actually find it to be helpful (a relative rarity). In brief, here’s an idiot’s guide to the five etiologies of PH.

Group 1

True Pulmonary Arterial Hypertension (PAH), the classic idiopathic flavor. PH due to connective tissue disease is also placed here, along with a few other things like HIV and liver disease, for reasons unclear to me.

Group 2

Cardiac in origin, usually from left heart failure, or occasionally from valvulopathy or anatomic abnormalities.

Group 3

Pulmonary in origin, most often from COPD, OSA, or OHS. Interstitial lung disease falls here as well.

Group 4

Thromboembolic disease, caused by acute or chronic pulmonary emboli.

Group 5

Miscellaneous. This category includes PH caused by sickle cell disease, metabolic diseases, etc.

True Group 1 PAH is its own beast and not a thing to be taken lightly. It’s a specialty disease that should be managed by specialists, in skilled centers, using drugs you can’t pronounce. These patients may be on exotic regimens—like continuous ambulatory epoprostenol pumps—which cannot be interrupted, even for minutes. If these people show up in your ICU, get help.

Contrariwise, the most common varieties of PH seen in the ICU are Groups 2, 3, and 4. In many cases, these are quite chronic, and may be noted incidentally during echocardiography. Although definitive diagnosis of PH requires right heart catheterization, pulmonary pressures can be estimated using echo if tricuspid regurgitation is present (in the presence of significant PH, it generally is). The doppler velocity of the regurgitant jet is taken, converted to a pressure, and added to the known or estimated CVP, resulting in a right ventricular systolic pressure (RVSP). The RVSP is usually a reliable surrogate for PA systolic pressure. Close enough for messy ICU purposes, anyway.

In the less acute setting, the finding of PH in a patient with, say, CHF (Group 2) or COPD (Group 3) is generally not a good thing, but also not a huge surprise or a matter for panic. The best treatment is usually to treat the underlying cause; PH-specific therapies such as pulmonary vasodilators are typically not indicated, and indeed may be harmful. (One drug, riociguat/Adempas, is available for chronic Group 4 PH.)

But what about in the ICU?

What’s the worry?

To understand the danger, we need to review the physiology of the pulmonary circulation.

Per cardiac cycle or over any unit of time, an equal volume of blood must pass through the left and right heart. Yet a normal pulmonary arterial pressure is only 25/10 mmHg, compared to a systemic pressure of 120/80. So it’s clear that the pulmonary circulation is a low-pressure system, and the right heart is accustomed to pumping against a low afterload. Inspect it using ultrasound or look at pathology specimens and you’ll see why this makes sense: its walls are thin and flimsy compared to the robust myocardium of the left heart.

So when the left heart faces an increased afterload, such as the acutely increased SVR of systemic vasoconstriction, it has the contractile reserve to push through and maintain cardiac output. The right heart? Not so much. In chronic cases, the RV does have some capacity to hypertrophy and adapt to higher pressures, but in virginal form, it has (as the sports media says) no “clutch,” no reserves. When push comes to shove, it simply fails.

And it fails in the most problematic way: a positive feedback loop. As pulmonary afterload increases, the right ventricle distends and dilates. As the septum bulges from the RV toward the LV, left heart output decreases as well, a phenomenon known as “interventricular dependence”—but if that concept is confusing, just remember that nothing can pump from the left heart that doesn’t first get pumped through the right heart. As CO falls, flow through the coronary arteries decreases, perfusion to the right heart becomes inadequate, and myocardial ischemia further degrades pump function. The increasing RV wall tension also mechanically obstructs coronary bloodflow, exacerbating this effect. As a third insult, the catecholamine surge from all of this stress constriction of the pulmonary arteries, further increasing RV afterload. With or without additional insults, such as hypoxia or acidosis… the end result is cardiac arrest from right heart failure.

(Remember, this is very different from right heart failure that occurs in the absence of pulmonary hypertension. Inferior STEMI involving the RCA, for instance, can knock down RV function, but cardiac output is usually preserved as long as preload as adequate; venous flow simply passes through the right heart like a conduit. But in that case, pulmonary arterial pressure is low, so the PAs are easy to fill, so long as the CVP is adequate. With a high PA pressure, you need the right heart to generate a pressure head sufficient to overcome it. If it’s unable to do so, forward flow ceases.)

The culprits

Hopefully it’s becoming clear that patients with significant pulmonary hypertension are at risk for hemodynamic deterioration in a way very different from left-sided failure. Acute LV failure results in pulmonary edema and cardiogenic shock. Acute RV failure results in sudden deadness from which you don’t recover.

The problem is that a patient may seem to be tolerating their PH reasonably well until some additional insult pushes their RV over the edge. What kind of insult? Anything that increases pulmonary afterload or worsens RV function.

  • Hypoxemia: Due to hypoxic vasoconstriction in the lungs, any global fall in oxygenation will cause an increase in PVR.
  • Hypercarbia and acidosis also cause pulmonary vasoconstriction, with the same deleterious result.
  • Systemic hypotension decreases coronary perfusion, further reducing RV function.
  • Positive pressure ventilation: Any increase in intrathoracic pressure will compress the pulmonary vasculature and increase its resistance.
  • Volume overload: Unlike the LV, whose Starling curve levels out with sufficient preload but does not reverse (a plateau, not a horseshoe), the RV’s systolic function actually decreases when it becomes overloaded. Too much preload will result in worsening RV failure.
  • Vasopressors: Vasoconstriction can contract the pulmonary arteries, increasing pressure and right-sided afterload.

So what should we do about all this? Tune in for part 2 to find out.