Diastology & Volume Assessment
The complexity of determining volume status and fluid responsiveness is aided greatly by having ultrasound at the bedside. However, the ability to integrate these ultrasound findings with the clinical physiology is essential in making accurate patient treatment decisions.
Physicians commonly ask whether a patient is “volume up or down”, “fluid overloaded or dehydrated”, or “ wet or dry”. However, in these situations, the correct clinical question is “What is the left atrial pressure (LAP) and if low, does it respond to volume?”
For decades, heart failure (HF) was seen as the result of a disease that caused left ventricular (LV) systolic function to “fail” and systolic function was measured as ejection fraction (EF), a measure of radial function. A large number of therapeutic trials were done in patients with “systolic heart failure” or what more recently is called “heart failure with reduced ejection fraction” (HFREF). However, EF was not the whole answer.
Many patients with fatigue, exertional dyspnea, and pulmonary congestion were identified who had normal EF. These patients had LVs that were normal to small in volume, “stiff”, and susceptible to elevated LV end-diastolic pressure when left-sided volume increased or tachycardia reduced the time the LV had to fill and empty. These patients often did poorly with the loss of atrial contraction during atrial fibrillation. The end result was elevated LAP, causing symptoms of pulmonary congestion.
For many years this disease was called “diastolic HF”, but more recently the term “HF with preserved ejection fraction” (HFPEF) is used. This is a disease of aging patients, particularly women and those who are obese, diabetic, or hypertensive. Amyloid, sarcoid, and Fabry disease are rare disorders that also cause HFPEF. Demographics indicate that HFPEF will be the dominant form of HF in primary care clinics. The BNP can be surprisingly low in these patients. If they have increased LV mass because of hypertension, the mass can decrease with hypertensive treatment, but the stiffness is more resistant. There is also evidence that stiffness occurs before hypertrophy in many chronic hypertension patients.
However, it is a mistake to think of HF patients as distinctly either HFREF or HFPEF. Most patients with HFREF also have abnormal diastolic function and most patients with HFPEF beyond the early stage have abnormal LV longitudinal systolic function. Thus, many patients with HF have a mixture of systolic and diastolic troubles.
A handful of medications improve survival for HFREF patients but this hasn’t been shown for HFPEF patients. A systematic review reported that mineralocorticoid antagonists (e.g. spironolactone) improved diastolic function in HFPEF without changing LV mass or dimensions. L-arginine was also reported to improve some patients with HFPEF. Nevertheless, for all HF patients, the clinical findings are predominantly determined by the LAP and optimizing this pressure with medications is a prime goal of HFREF and HFPEF treatment.
DON’T FORGET THE OTHER FINDINGS
Before discussing specific measures of diastolic function and LAP, there are three indirect markers of HFPEF severity to emphasize.
Interventricular septal (IVS) width: Patients with at least moderate LV hypertrophy (LVH) almost always have some diastolic dysfunction. However, non-hypertrophied LVs may still be stiff, so the lack of LV hypertrophy does not exclude HFPEF.
Left atrial size: The size of the left atrium (LA) has been called the “hemoglobin A1c of left-sided heart disease”. An enlarged LA doesn’t distinguish between HFREF, HFPEF, valvular disease, or chronic lone atrial fibrillation but indicates chronicity in any of these conditions. A patient may have diastolic dysfunction but a normal sized LA because the LAP has not been chronically elevated. Symptoms might only be present in such a patient with significant exertion, tachycardia, acute atrial fibrillation, or acute volume overload.
JVP: A large proportion of patients with chronic HF develop secondary pulmonary hypertension as the disease progresses. Studies in HFPEF show that an elevated JVP indicates a high LAP 90% of the time. However, the JVP is not sensitive for diastolic dysfunction and could be elevated from primary right heart conditions.
USING “DIASTOLOGY” MEASUREMENTS
Doppler techniques for measuring LV function were discussed in the LV chapter. This next section further integrates these findings to reach diagnostic conclusions.
In 2016, the American Society of Echocardiography and the European Association of Cardiovascular Imaging jointly published updated recommendations for the evaluation of left ventricular diastolic function (J Am Soc Echocardiogr 2016;29:277-314). These groups wanted to clarify and simplify this important assessment. Not all experts agree with everything in this consensus statement. This topic is being actively studied and changes will come.
A formal echocardiography laboratory measures some things that an IMBUS exam does not. However, conclusions from an IMBUS exam should adhere to the overall recommendations of the consensus statement.
Clinical conditions that make diastology measurements invalid or unreliable need to be emphasized. IMBUS diastology is unreliable in atrial fibrillation. Pacemaker patients also have special problems and should be avoided. Moderate or greater mitral annular calcification, mitral stenosis, mitral regurgitation, and eccentric aortic regurgitation directed toward the anterior mitral leaflet all confuse the measurements. Patients with mitral valve repair/replacement or basal septal wall motion abnormalities (e.g. from ischemia) should be avoided.
QUESTION 1: IS DIASTOLIC DYSFUNCTION PRESENT?
Is the heart “stiff” and not relaxing well? For IMBUS the answer comes from e’ with a little help from the e’/a’ ratio.
While the consensus statement only specified e’ being low for a patient’s age as the primary criterion for diastolic function, additional published work noted that the age-adjusted e’/a’ ratio may help evaluate LV stiffness because it is less variable than e’ to changes in LA volume/pressure. Unfortunately, as discussed in the Atria chapter, left atrial dysfunction with preserved sinus rhythm can cause the a’ velocity to fall, resulting in an increasing e’/a’ ratio that no longer reflects diastolic function.
When a patient has a clearly reduced e’ it is comforting to have confirmation of diastolic dysfunction from a similarly low e’/a’ ratio. However, if a low e’ patient has a normal to elevated e’/a’ ratio, there are several possibilities. A first is reduced left atrial volume and pressure, so other indicators of such a state would be sought. A second possibility is left atrial dysfunction, which would be further supported if the a’ velocity is abnormally low and the A-wave velocity is below normal for age. With the suspicion of left atrial dysfunction, the e’ velocity alone is used to evaluate diastolic dysfunction.
The majority of patients with HFPEF first develop stiffness of their LV, without other findings or symptoms. Does it matter if patients with stiff hearts are identified early? Would labeling such patients lead to different patient care decisions? These answers vary patient to patient and will change as treatment for HFPEF changes. For now, patients who have abnormally stiff hearts without LA enlargement, IVS hypertrophy, or any other marker of heart failure should probably be labeled as “diastolic dysfunction” with the term HFPEF being reserved for patients with additional findings. We are reluctant to overcall HFPEF but think that knowing a patient has diastolic dysfunction can be helpful in some situations. If the evidence accumulates that therapy benefits patients with early HFPEF, this diagnostic labeling approach could change. Significantly abnormal diastolic function without other findings must also raise the possibility of amyloid, sarcoid, and Fabry disease.
QUESTION 2: IS LAP ELEVATED?
If diastolic dysfunction is present, an estimate of current LAP gives the grade of dysfunction at that moment. Grade I is used for patients with normal LAP. Grade II is used with some evidence for elevated LAP and Grade III is used for strong evidence of high LAP. The grade of dysfunction is dynamic in every patient. LA size is a marker of the chronicity and severity of LAP increase but would always take grade II or III disease to create, unless atrial fibrillation was also present.
A. E/A: The velocity of the E wave is dynamic and varies with left sided preload, afterload, and LV contractile state. Thus, a patient with HFPEF can have E/A that is very low or very high at different times. E/A below the normal cutoff for age (old “impaired relaxation”) indicates low to normal LAP and is labeled Grade I HFPEF. An E/A in between the lower cutoff for age and twice this value (old “pseudonormal”) probably has some elevation in LAP and is labeled Grade II HFPEF. E/A > twice the lower cutoff for age (old “restrictive”) is very elevated LAP and Grade III HFPEF. The expert consensus gave simplified E/A cutoffs of < 0.8, 0.8 - 2.0, and > 2.0 for the three categories, but this ignores the variation in E/A with age and may “over grade” elderly patients. Alternatively, perhaps elderly patients more easily develop advanced grades of HFPEF.
Left atrial dysfunction is an important consideration in interpreting the E/A ratio. As LA function deteriorates, the velocity of the A-wave may fall, which increases the E/A ratio even though the LAP has not necessarily increased. While the frequency of this problem is unknown, it must be considered when Grades II and III are being assigned to a HFPEF patient. If the A-wave velocity is low for age, be cautious diagnosing elevated LAP.
An easy demonstration of the dynamic variation of E/A can be shown in patients with Grade II or III diastolic dysfunction who perform a Valsalva maneuver. In the following image from the expert consensus paper, a patient with Grade II diastolic dysfunction with an E/A of 1.3 in the left image performed a Valsalva maneuver for 10 seconds (resulting in decreased preload/LAP) with a resultant E/A of 0.6 on the right, which would be classified as Grade I.
B. E/e’: A 2016 systematic review concluded “There is insufficient evidence to support that E/e’ can reliably estimate left ventricular filling pressure (LAP) in HFPEF.” The specificity of an elevated ratio looks to be good (maybe 70-90%) but the sensitivity is much lower. Nevertheless, this popular ratio is supposed to give a category estimate of LAP. E/e’ < 8 indicates a normal LAP, E/e’ between 8 and 14 indicates elevating LAP, and E/e’ > 14 indicates clearly elevated LAP. We think the vagaries in the measurements of E and e’ account for the systematic review findings, not that the ratio is theoretically wrong as a measure of LAP. In particular, e’ should not be used in the various conditions that produce unreliable measurements (noted in the LV chapter) and E and e’ should be obtained at the same end-expiratory phase in the respiratory cycle. We think the E/A ratio should be the preferred measure of LAP in the IMBUS exam. But, because we have the data, E/e’ will be calculated and considered in our clinical conclusions. A very high ratio probably means elevated LAP if falsely low e’ can be excluded, but lower ratios may not be good for excluding elevated LAP.
C. “LAP Index”: The diastology assessment is a two-step process: determine the stiffness with e’ (with confirmation from e’/a’ ratio) and then use E/A ratio to categorize the LAP. Both ratios vary with age so each needs to be compared to age-adjusted norms.
It is mathematically logical that a combined (E/A)/(e’/a’) ratio might categorize LAP in a single number without needing to refer to age-adjusted norms. Across the age categories, it appears that this LAP index should be approximately 1.0 - 1.6 in patients with normal LAP and this is consistent with our experience. Ratios progressively greater than 1.6 are increasingly strong evidence of elevated LAP and ratios progressively less than 1.0 are stronger evidence of low LAP. However, this measurement is not yet formally recognized for clinical use so we are cautious. Patients with left atrial dysfunction should have lowered A-wave and a’-wave velocities, which then result in “falsely high” E/A and e’/a’ ratios. Theoretically, the (E/A)/(e’/a’) ratio might also normalize this change and perhaps correctly indicate the LAP category.
D. Tricuspid gradient: Chronically elevated LAP eventually causes some secondary pulmonary hypertension. Elevated JVP can become a marker of this process. As described in the tricuspid valve chapter, detecting pulmonary hypertension requires at least a small jet of tricuspid regurgitation (TR) in which to place the continuous wave Doppler (CWD) cursor to measure the peak velocity. The expert consensus stated that a TR jet velocity > 2.8 m/sec, which equates to a gradient of > 31 mmHg, indicates more severe grades of chronic diastolic dysfunction. Like E, A, e’, and a’, pulmonary pressure is at least partly dynamic, being influenced not only by changes in LAP but also by changes in the RV preload and contractility. But, the presence of an elevated tricuspid gradient certainly indicates important chronic LAP elevation, unless the elevated gradient is from an unrelated cause of pulmonary hypertension.
An 80-year-old man who lived alone was brought to clinic by his daughter because he was complaining of feeling weak and a little dizzy. The daughter had last seen him several weeks ago and thought that maybe he had been having some diarrhea. The physician was concerned about volume deficiency.
Clinic patients periodically present with lightheadedness, dizziness, weakness, reduced blood pressure, or increased creatinine with or without an orthostatic part to their story. They may or may not have other historical clues. The differential diagnosis includes cardiovascular, neurologic, metabolic, pharmacologic, and even psychologic conditions. Cardio-pulmonary IMBUS can help with the diagnosis.
Findings #1: IMBUS of the patient’s lungs showed only a few B-lines in the posterior inferior-lateral regions. There were no pleural effusions. This was consistent with low LAP.
If a patient has low LAP from isolated volume deficiency, there should not be a diffuse interstitial pattern, unless there is an unrelated subacute or chronic lung disease. Similarly, substantial pleural effusions should not be seen in the setting of low LAP unless a patient has an inflammatory pleural process with effusion (e.g. cancer, pneumonia), or had an elevated LAP for some duration that has resolved in the recent past without sufficient time to reabsorb the pleural fluid. Patients with cirrhosis and portal hypertension can have ascites and pleural effusions despite having low LAP. Thus, pulmonary IMBUS is important, but abnormal findings do not exclude low LAP, they only make the diagnosis more complicated. A patient may have more than one active disease.
Findings #2: In the 30 degree elevation position, the patient’s right IJ was mostly collapsed, with a longitudinal view “beak” barely above the clavicle with only blue flow coming from above with color Doppler. He was about 67 inches tall. This was consistent with at least low normal LAP.
JVP should not be elevated if a patient has low LAP from isolated volume deficiency. This average height patient with no IJ distention even at 30 degrees certainly has a peak JVP that is < 8 mmHg (normal) but “below normal” JVP cannot be measured. JVP also estimates right heart volumes and pressures, which can be dissociated from left heart processes. Patients with acute or chronic right heart failure can have an elevated JVP but still have a below normal LAP leading to low cardiac output. An elevated JVP would be very useful information in a patient like this, but it would not exclude low LAP.
Findings #3: The patient’s supine heart rate corrected carotid flow time (CFTc) was 290ms. When an assistant raised his legs to 45 degrees for a minute, the CFTc was 335ms. This was an increase of 15%. The examiner made sure the patient was relaxing his belly and breathing normally during the leg lift to be sure the patient wasn’t helping to lift his legs, which would cause a Valsalva maneuver, negating the auto-bolus. This low normal supine CFTc with a substantial increase with passive leg raise supported low LAP that was volume responsive.
CFTc should be performed on patients who may have low LAP. It is worth getting a second person to lift the legs to 45 degrees. As the CFTc increase with passive leg raise becomes progressively more than +5%, the chance of low LAP and volume responsive increases and when the increase is ≥15% it is highly likely.
Findings #4: The patient’s IVC was difficult to see, but didn’t seem to be distended. This was consistent with low LAP.
It may be difficult to see the IVC well, particularly when there is volume deficiency. Like the JVP, the IVC tells us about right atrial pressure and can be dissociated from LAP. However, an IVC that is well seen, small and completely collapsing with inspiration is good evidence for a patient being at least “fluid tolerant”. This helps support low LAP.
Findings #5: On cardiac IMBUS, the patient was mildly tachycardic in sinus rhythm and his chamber sizes were normal. Visually he had normal radial systolic LV function. His inter ventricular septum was 1.1 cm. There were no valve lesions. TDI of the mitral annulus showed an s’ = 10 (normal, confirming good LV longitudinal systolic function), an e’ of 8, and an e’/a’ ratio = 0.8 (both normal for age), indicating acceptable diastolic function. With continuous wave Doppler LV inflow analysis, his E/A ratio was 0.7 (mildly low for age). This pattern is suspicious because the e’/a’ ratio should be lower than the E/A ratio in most patients. The LAP index was 0.875, supporting low LAP.
Cardiac IMBUS can uncover many lesions and surprises, such as valve problems and heart failure that will change our diagnostic thinking. However, with or without structural heart disease, the issue of LAP remains. Classically, low LAP in a normal LV should produce a hyper-dynamic heart, although differentiating this from normal is not always easy. The normal e’ and e’/a’ ratio in this patient indicated acceptable diastolic function, but the lower E/A ratio suggested low LAP and the LAP index, which normalizes for age and possibly LA function, similarly indicated low LAP.
SUMMARY: This patient’s 15% change in the CFTc was an important early clue to low LAP. His age-normal e’ and e’/a’ ratio indicated acceptable diastolic function but his reduced E/A and LAP index indicated low LAP. Further questioning suggested that the patient had probably not been eating or drinking very well in the last few weeks after a bout of diarrhea. Further diagnostic studies were not pursued and his family assisted in rehydration at home.
A 70-year-old obese woman with hypertension had been feeling short of breath just walking out to the mailbox. She had bilateral posterior lower lobe diffuse B-lines without pleural effusions. Her JVP was elevated to 15 mmHg. Her IVC was normal size, but collapsed poorly with inspiration. Cardiac IMBUS was classic for HFPEF with some IVS hypertrophy and moderate enlargement of her LA. Her e’ = 6, e’/a’ = 0.6, E/A = 2.0, and LAP index = 3.3.
Treatment: The patient was already on excellent treatment for hypertension, confirmed by a recent 24 hour ABPM. A modest dose of furosemide was started at breakfast and dinner, hoping to avoid disturbing the patient’s sleep with the need to urinate. A low morning dose of spironolactone was also begun.
Follow up #1: She was seen briefly four days later and was already less short of breath. Her pulmonary IMBUS showed only mild interstitial pattern over her posterior lower lobes, an improvement from the initial exam.
The diffuse interstitial pattern of heart failure should progressively decrease as treatment is successful. The patient’s symptoms may improve before the interstitial pattern fully resolves.
Her JVP measured about 12 mmHg after four days of treatment, reduced from the initial 15, but still elevated. Her IVC looked about the same.
JVP elevation is probably a better marker of elevated RAP than the IVC, but neither of these is a direct estimate of LAP. Elevated JVP or dilated IVC should improve somewhat with HFPEF treatment.
Doppler evaluation showed unchanged e’ = 6 and e’/a’ = 0.6, but E/A had fallen to 1.5,, and the LAP index to 2.5. This showed a decrease in LAP, but not a return to normal. All the evidence supported the conclusion that the patient was symptomatically and hemodynamically improved. The treatment was left unchanged and the patient was re-evaluated 5 days later.
Follow up #2: Five days later she had no shortness of breath and was feeling pretty well. Her pulmonary IMBUS showed full resolution of the interstitial pattern. Her JVP was less than 10 mmHg. Her CFTc was 305ms supine and 320ms with a 45 degree leg raise (a 5% change). Her IVC measured a little smaller than before and seemed to be collapsing well with inspiration.
The key short-term goal with HFPEF or HFREF treatment is to reduce LAP so a patient is comfortable, but not to the point that low LAP is producing reduced cardiac output with weakness, fatigue, and rise in creatinine.
A 5% change in the CFTc is not definitive, but suggests beginning volume responsiveness, so caution is needed.
Doppler evaluation showed an e’ = 6, e’/a’ = 0.6, E/A = 0.8, and an LAP index = 1.3. These measurements, coupled with the other findings, indicated that LAP was about normal and should not be lowered further. To avoid cardiac output troubles, the furosemide dose was changed to once a day at dinner because urinating during the evening was preferable to her than during the morning.
Follow up #3: She was seen briefly a week later and was feeling well. Her JVP was less than 10 mmHg, her CFTc was 320ms supine with minimal change with passive leg raise, and her Doppler studies showed an e' = 6, e’/a’ = 0.6, E/A = 0.9, and LAP index = 1.5
SUMMARY: Patients with heart failure and volume overload can be followed at short intervals with relatively quick hemodynamic assessments that will show the progress of therapy but will also identify when the LAP might be getting lowered too much. The decrease in E/A and LAP index will be the most important parameters, but accompanying IMBUS data from the lungs, JVP, IVC, and CFTc will be helpful.