CHF exacerbation
Lesson Outline
Background
Baseline cardiac dysfunction may be defined in various ways
Ejection fraction - HFrEF (EF <40%) vs HFpEF
Now also expanded to include HF with recovered EF, HF with mid-range EF)
Formerly called systolic vs diastolic dysfunction, which is still essentially the idea
An extremely coarse separation; hearing thes labels should usually prompt further exploration of the actual physiology
Common to note most recent EF in HFrEF as a rough marker of severity
Cause of dysfunction
E.g. HFrEF: ischemic vs non-ischemic cardiomyopathy
Note that valvular pathology may be an underlying driver, such as aortic stenosis
Degree of left vs right-sided failure may be important although not usually captured in labels (significant right-sided failure usually earns the additional label of pulmonary hypertension)
Severity by NYHA or ACC/AHA classifications
Many are ischemic in nature, and hence travel alongside new or historical diagnoses of CAD (left heart caths, stents, CABG, etc)
Caring for heart failure nowadays is an extremely evidence-based field
Along with symptom management (dyspnea, weakness, lightheadedness, etc), goal is generally to reduce hospital admissions, disease progression, and most of all mortality
Cardiology community keeps performing massive RCTs on different agents, showing that they improve mortality when introduced to the package of drugs, and adding them to "GDMT" (guideline-directed medical therapy) - aka the bucket of drugs they'd like HF patients to be receiving, if possible.
ACEI, ARB, or ARNI (angiotensin receptor-neprilysin inhibitor - sacubitril-valsartan aka Entresto)
Beta blocker (carvedilol, metoprolol XL, or bisoprolol)
MRA (mineralocorticoid receptor antagonist), aka potassium-sparing diuretic - spironolactone or eplerenone
SGLT2 inhibitor (-flozins, usually empagliflozin [Jardiance] or dapagliflozin [Farxiga])
Antithrombotics and statins for underlying CAD
Diuretics for symptom management
Much of this is less relevant for acute exacerbations, especially causing critical illness, which are usually respiratory in nature
Flavors of exacerbation
First paradigm: Helpful to define exacerbations along two axes - venous congestion and perfusion
Congestion, or "backwards" problem: is there volume overload causing organ dysfunction (especially but not limited to pulmonary edema)?
Perfusion, or "forward" problem: is there hypoperfusion causing organ dysfunction (i.e. shock)?
Well compensated: warm and dry
Most common phenotype for hospitalization: warm and wet
Generally responsive to volume management, such as diuresis
Acuity depends on respiratory stability - many are treated on the floor - some need non-invasive support - the worst need intubation
Scary and generally critically ill: cold, usually wet
Essentially cardiogenic shock
A late stage of disease; perfusion preserved (especially with appropriate care) until quite severe heart failure
If refractory to medical optimization, may require more complex care such as mechanical circulatory support
Note: congestion can also contribute to organ ischemia in severe cases (impairs gradient to perfusion)
Uncommon: cold and dry
Uncommon because the natural state of cardiogenic shock typically involves congestion
But can occur especially in "unnatural" states, e.g. after diuresis
Appropriate: optimized volume status
Inappropriate: overshot into hypovolemia
Second useful paradigm: understanding loading
Cardiac function must always be understood as three interacting variables: contractility, preload, and afterload
Preload: the pressure filling it before it squeezes (end diastole). Approximated by CVP on the right, PCWP on the left.
Contractility: how hard the heart squeezes (relevant on either side)
Afterload: the pressure opposing chamber emptying. Approximated by arterial BP or SVR on the left, PAP on the right.
How much blood is pumped forward or congested backwards always depends on how these variables interact.
All things being equal, higher afterload = more cardiac work = less cardiac output and more congestion
All things being equal, higher preload = more cardiac output initially, then more venous congestion at higher filling pressures (Starling curve)
So optimal loading conditions for the heart are generally the lowest preload and afterload that do not cause hypoperfusion (from hypotension and hypovolemia), along with optimizing other variables such as heart rate, rhythm, etc.
Example: if you cut the SVR in half, a weak LV will pump twice as much blood as before, without any changes to the myocardium itself.
Example: if you pump enough fluid into any circulation, pulmonary edema will develop, even with a normal heart. Conversely, if you pull enough fluid out, any patient will have dry lungs, even with terrible LV function. There is no such thing as volume status or cardiac function in isolation of each other.
Example 1: the warm, wet floor patient
Patient with HFrEF (35%) presents after 1 week of increasing dyspnea and orthopnea, increased peripheral edema, found hypoxemic and admitted to the floor. BP 135/85, requires 3L nasal cannula, otherwise stable.
Home furosemide increased, diureses net negative 1 liter daily for three days, breathing improves, weaned to room air, discharged.
Example 2: the cold, wet ICU patient
Same patient presents with dyspnea, weakness, dizziness, found obtunded. In ED, noted hypotensive to 60/40, pale and cold extremities, lethargic, SpO2 60% on room air.
Placed on BiPap, intubated. POCUS shows EF 15%, started on dobutamine and norepinephrine. BP improves, extremities become warm and pink, patient rouses. Cardiology called to evaluate for underlying etiology and next steps.
Example 3: "flash" pulmonary edema
First patient admitted to floor. On day 1, was interacting with family, became anxious, nursing called and found him hypertensive to 220/110, pale, diaphoretic, diffuse crackles in lungs, SpO2 70% on 6L NC. Placed on BiPap, started on nitroglycerine drip with rapid improvement.
4 hours later, BP normalized, weaned off NIPPV and nitro.
"Flash" pulmonary edema an acute episode of abnormal loading driven by afterload. Tends to occur suddenly and episodically as a positive feedback loop.
Increased afterload -> Increased pulmonary edema -> Hypoxemia -> Increased catecholamine tone -> Increased afterload
No changes on contractility or preload necessary (patient need not be volume overloaded, although that will make it more likely)
Presentation
May be protean, but patients requiring ICU will generally be short of breath, with a prodrome usually days to weeks
Shorter (hours) if there is a discrete provocation, such as dietary
Shorter (minutes) in flash pulmonary edema
May or may not note weight gain, increase in peripheral edema, cough, dietary indiscretions, or changes in medication use (intentional modifications, non-compliance due to access issues, interruption due to illness)
Respiratory failure, primarily hypoxic
Generally some response to FiO2, but incompletely (shunt physiology, but less dense than in other cases like pneumonia, more in the V/Q mismatch spectrum)
Responds much better to positive pressure
Generally much easier to recruit than exudative processes
Hypercarbia may be present but usually not significant without another secondary process (COPD, OHS, etc)
May be secondary symptoms suggesting a provoking trigger
Infection not uncommon
Clinical evaluation
Assess preload, afterload, and contractility
Preload
Physical exam
JVD
Auscultation
Peripheral edema
Body weight
Diagnostics
BNP
Chest imaging (CXR, CT, ultrasound)
Echo - IVC, chamber dilation
Invasive
CVP
PCWP
Afterload
LV
BP
Skin
SVR (via PA catheter)
RV
PAP (via PA catheter)
RVOT acceleration time or notching on echo
Contractility
Echo, echo, echo
PA catheter (cardiac output)
Non-invasive cardiac output monitors (PiCCO, FloTrac, etc)
Perfusion
Skin, including cap refill
Mentation
Lactate
Kidney and liver function tests
Reminder: these are all in context of each other. Does cold extremities mean high afterload or low contractility? Both - the afterload is too high for the current contractility, or contractility is too weak for the current afterload.
Treatment
In everyone: Reduce afterload
Hypotension harmful, but any afterload will generally worsen cardiac output and congestion
MAP goal barely >65 or SBP 90-120 probably reasonable targets
BP "normal" to high? Start arterial vasodilators
Acute/unstable: Nitroglycerine, nicardipine, clevidipine, or nitroprusside
More stable, or weaning: ACEI/ARB, alpha blockers (carvedilol, etc), nitrates, hydralazine, dihydropyridine calcium channel blockers (e.g. amlodipine)
The wet patient: Treating congestion
Assess volume status as best as possible - historical clues, physical exam, labs and imaging
Remember that volume is interpreted in context. A patient may be 5 liters "positive" but intravascularly hypovolemic at the moment. As the heart and other conditions change, your decision may change.
Most commonly pulmonary edema, but any organ can suffer from edema and congestion, including kidneys, liver, abdominal compartment syndrome secondary to ascites. Lungs are usually most obvious though.
If there is venous congestion, attempt to remove intravascular volume.
First line for removal: diuresis
See Diuresis lessons in Foundations
Titrate aggressiveness to stability. A little congestion but stable? Can go slowly. In severe respiratory distress? Be aggressive (although other measures will be more relevant acutely).
Will they tolerate it hemodynamically?
Consider Starling curve. If LV is not on the steep portion, removing preload should have little effect on cardiac output.
However, in severe shock, even this small difference could precipitate instability
RV generally appreciates euvolemia - not hypovolemia, but may precipitously fail from overdilation as well
In severe shock, transient hypotension from boluses of diuretics may be a problem; consider a flatter response from something like a drip.
Supporting breathing
Cardiogenic pulmonary edema responds somewhat to FiO2, but much more so to positive pressure
NIPPV first line
Improves preload, improves afterload, helps drive fluid out of alveoli
BiPAP is fine, but CPAP works too - the EPAP (PEEP) is the main tool
Okay to temporize with non-rebreather or similar if awaiting positive pressure, but not the same
If in extremis, use BVM to rescue - must use PEEP valve with tight mask seal (two-handed seal preferred).
HFNC a distant second line
Provides a little PEEP but not much and not titratable
More of a fallback for patients who cannot tolerate a mask
Intubate if failing
Remember when weaning: the positive pressure is always present on the vent even at low settings, so it may be easy to overestimate readiness for extubation (they will breathe worse off the vent).
The cold patient: Treating cardiogenic shock
Use POCUS to understand the source of the problem
LV vs RV failure. (Note severe LV failure will usually entail RV failure, but not vice versa.)
Rule out obstructive causes of shock
Rule out significant hypovolemia (i.e. hypovolemic shock, not cardiogenic).
Consider underlying cause as this will significantly impact care
Some causes are non-acute and not emergent, e.g. chronic cardiomyopathies
The shock may still be acute, but the base cause is not
Some will not be fixable without identifying the cause, such as obstructive shocks
Some will progress and cause permanent damage without identification, namely ACS
Evaluate every patient in new cardiogenic shock for coronary ischemia, whether or not there is a known baseline
12-lead ECG and troponin in everyone
Very low threshold to involve Cardiology
Optimize loading conditions
Target euvolemia
Volume overload worsens congestion, but may worsen forward flow as well, particularly for RV failure
Increase contractility (AFTER optimizing loading)
First line if hypotensive: norepinephrine
More of a vasopressor than an inotrope, but creates a stable hemodynamic baseline and ensures coronary perfusion.
Proceed to specific inotropes. No right choice, but consider:
Dobutamine if BP normal/high
Milrinone if significant RV failure and pulmonary hypertension
Caution in renal failure
Epinephrine if hypotensive
Avoids vasodilation and possible hypotension
Favor low doses for inotropy (<0.1 mcg/kg/min)
Titrate to markers of perfusion
Do not order as nurse-titratable, unless you have a concrete number for them to target (e.g. PA catheter)
Titrate to skin findings, echo, perfusion labs (lactate, etc)
Don't forget the heart rate
Assume that higher HR = higher cardiac output up to maybe 120-ish (eventually, higher rate may entail lower stroke volume)
If bradycardic, try to drive up rate with inotropes without causing arrhythmia
If shock persists despite optimizing loading conditions and reasonable medical therapy, consider mechanical circulatory support
Nuanced decision, heavily driven by underlying cause (reversibility) and local resources
Involve Cardiology
Most centers can place a balloon pump
Many can place Impellas or other minimally-invasive support
Some can proceed to VA ECMO
Goal is generally to bridge to recovery, to transplant, occasionally to prognostication or a decision (buying time), or perhaps to durable support like a VAD
If you do not consider the possibility, it won't happen; don't just flog a patient with drugs that aren't working when they might be a candidate for a device.
The stabilized patient
Goal is to maintain stability, tight cardiopulmonary homeostasis, while transitioning to a medication regimen that can be used out of ICU, and prioritizing GDMT drugs that optimize outcome
Reasonable approach:
Diurese until euvolemic, then transition to a maintenance oral diuretic regimen
Start beta blocker if not on inotropes and well perfused
Favor metoprolol (if tenuous BP) or carvedilol (if BP okay/high)
Titrate up as HR/BP tolerate
Transition to long-acting formulations when sufficiently stable
Note short-acting metoprolol not proven to improve outcomes; use as a titration tool but not an end goal
Add ACEI/ARB/ARNI
Classic approach: Rapid captopril uptitration (can increase every dose)
Many now favor an ARB over ACEI, which is fine too
Cardiology may be comfortable proceeding directly to ARNI, but most of us in ICU would rather start on more familiar ground and have them transition later
Throw on spironolactone, SGLT2 inhibitors, and non-GDMT drugs to lower BP if still elevated
This order is generally reasonable and safe in the ICU setting; Cards may be more aggressive or use other tactics on the floor or in clinic where priorities are different
Hemodynamics are the most common limiter to GDMT; BP is a finite bucket and you may not have room to optimally uptitrate everything to goals
Approaching in a reasonable order helps increase your chances of meeting as many of your goals as possible
Cards will fix your regimen later if needed, but a thoughtless approach in the ICU makes it harder, and maybe increases risk of recurrent complications in the meantime (e.g. using amlodipine before an ACEI)
Special situations
Diastolic dysfunction
Generally, treat HFpEF similarly to HFrEF (partly because we don't really understand another way to do it). Differences:
Increasing contractility will not help, so less role for inotropes - fortunately, an uncommon isolated cause of cardiogenic shock (usually causes congestion), although it might occur in conjuction with other processes
Slower rates usually help (takes time to fill)
May require somewhat higher preload
Too much still means congestion and edema, but too little may quickly mean hypoperfusion, so there is a narrow window here
Severe/critical aortic stenosis
Also a narrow window for volume - lower preload may precipitously drop cardiac output
Less filling -> Worsens stenosis -> Less cardiac output -> Less coronary perfusion -> Less cardiac output
Similar to diastolic failure, may be hard to find a therapeutic window for volume; be cautious
Also like diastolic failure, lower rates may be helpful
LVOTO
Dynamic obstruction of the LV outflow tract during LV systole
Usually in setting of thick septum (especially with a sigmoid shape, i.e. thickening towards base)
Usually exacerbated by higher rates (less filling), lower preload, lower afterload
Septic or hypovolemic patient with pre-existing LVH would be a classic situation
Causes a self-perpetuating limitation to stroke volume, as well as mitral regurgitation
LV outflow tends to catch the anterior leaflet of mitral valve, causing SAM (systolic anterior motion) and mitral regurgitation
If LVOTO noted on echo, even if mild, or anatomy consistent with risk for it, be cautious with hemodynamic approach
Like diastolic failure and LVOTO, favor lower heart rates and higher preload
Inotropes that cause tachycardia may worsen hemodynamics, not help (phenylephrine is a classic choice)
Diuresis may precipitate hemodynamic decompensation
