Does Your Patient Really Have Diastolic Dysfunction?

Recently, while preparing an updated presentation to the Saint Louis University cardiology fellows on “diastolic dysfunction,” the skeptical cardiologist came across a document entitled “What Does Grade I Diastolic Dysfunction Really Mean: The Emperors of Diastology Have Very Flimsy Clothes.”

This appears to be something I wrote in 2018 but never finished or published anywhere. It addresses a widespread problem: the frequently inaccurate and constantly mutable diagnosis of diastolic dysfunction, something that has plagued echocardiography for 35 years.

I’m posting the document today, warts and all, in hopes it will provide succor for those patients who may have fretted upon encountering the seemingly worrisome diagnosis on their echo reports and for non-cardiologist ordering physicians who likely scratch their heads when they see the diagnosis on their patient’s echo report.

You order an echocardiogram for a 65-year-old woman with palpitations due to premature ventricular contractions (PVCs.) The report returns with all normal except that the left ventricle (LV) is described as having “Grade I Diastolic Dysfunction.” 

What should you tell your patient about this?

Is it something they should be concerned about? 

Should treatment be started? 

Is additional testing warranted?

Will it influence their life insurance?

The bewildering miasma

To properly answer the question I must provide you with a little background on the bewildering miasma that the echocardiographic diagnosis of diastolic function has become.

When I was undergoing my training in cardiology in the late 1980s at Saint Louis University, Doppler-echocardiography was rapidly emerging as an exciting tool for noninvasively measuring cardiac hemodynamics. After establishing that Doppler could accurately assess valvular heart disease and measure pulmonary pressures researchers in the field including yours truly (see here) realized that we could precisely record the pattern of diastolic filling of the left ventricle, something we automatically assumed would be the Rosetta Stone for the clinical diagnosis of “diastolic heart failure.”

Cardiologists have long relied on the ejection fraction (EF) to measure how the left ventricle squeezes/pumps/contracts (inotropy) during systole but equally important is how well the left ventricle relaxes and fills (lusitropy) during diastole. Patients who had heart failure with a normal EF were typically classified as having “diastolic heart failure” a term which is gradually being replaced by heart failure with preserved EF (HFpEF.)

Filling of the left ventricle in the normal young it was observed was very rapid in early diastole (high E velocity) but slowed during late diastole at the time of atrial systole (low A velocity) yielding a high E/A ratio. In older individuals and patients with LV hypertrophy and coronary artery disease (CAD), on the other hand, the reverse was seen with a low E/A ratio.

A low E/A ratio  was seized upon as a sign of diastolic dysfunction and patients began to be labeled as having diastolic dysfunction based on an E/A ratio of <1.0 (even though this is normal over the age of 65 years.)

Research (to which I and my colleagues contributed) fairly quickly showed that diastolic dysfunction was much more complicated and could not simply be diagnosed by a low E/A ratio but this criterion, perhaps due to its simplicity, persisted and to this day I see patients diagnosed at other facilities with diastolic dysfunction or “impaired relaxation” based on this criterion alone.

Over 3 decades and after thousands of research studies we have learned much more about the complexisites of diastole and a whole host of Echo/Doppler parameters have been introduced as potentially helpful. Some have  come and gone but many have persisted. The latest document with expert guidelines on diastolic function measurement by echo lists 16 of these parameters.

Understandably, general cardiologists (who read most echocardiograms), have difficulty keeping up with the ever-changing landscape of diastolic function. 

What Does Grade I diastolic dysfunction mean in an individual?

The first thing to recognize is that there is a high likelihood that your patient does not have Grade I diastolic dysfunction by current criteria. As I mentioned there are still readers basing their diagnosis of diastolic dysfunction or “impaired relaxation”  on information from 1988. 

Other readers may be trying to add in some of the various newer measurements they learned about at a cardiology conference. CME conferences and the presenters at these conferences take in considerable dollars and it behooves them to present the latest developments to maintain interest. Such developments are highly likely to be only evanescently important. Nevertheless, there are so many of them and with such changing application that the field of “diastology” has sprung up complete with conferences devoted entirely to the topic.

Over the years new diastolic parameters have included the velocity of M-mode color Doppler propagation into the LV (hopefully this is dying out) the difference in milliseconds between reverse atrial flow into the pulmonary veins and the forward atrial flow across the mitral valve (which persists despite my paper showing it is not reproducible). There are now so many of these parameters it is unclear to most nonacademic cardiologists how to use them. 

When I review echocardiograms from outside facilities it is very common for an inappropriate diagnosis of diastolic dysfunction to have been made.

There are also readers basing the diagnosis on the 2009 DF guidelines. The 2009 guidelines utilized a whole host of additional parameters to sort out patients including deceleration time, IVRT, pulmonary venous flow and tissue Doppler of the mitral annulus. The 2009 guidelines were so difficult to utilize they were replaced in 2016 by another set of guidelines which resulted in a markedly lower rate of Grade I diastolic dysfunction.

Here’s an example of the tortured and confused language in the 2009 guidelines:

In patients with mild diastolic dysfunction, the mitral E/A ratio is  0.8, DT is  200 ms, IVRT is  100 ms, predominant systolic flow is seen in pulmonary venous flow (S   D), annular e= is  8 cm/s, and the E/e= ratio is  8 (septal and lateral). These patients have reduced diastolic reserve that can be uncovered by stress testing. However, a reduced mitral E/A ratio in the presence of normal annular tissue Doppler velocities can be seen in volume-depleted normal subjects, so an E/A ratio   0.8 should not be universally used to infer the presence of diastolic dysfunction. In most situations, when the E/A ratio is  0.8, mean LA pressure is not elevated, except for some patients with severely impaired myocardial relaxation, as in long- standing hypertension or hypertrophic cardiomyopathy.

What is clear from the first sentence is that there is a lack of clarity: six different parameters are listed to diagnose “mild diastolic dysfunction.”  The first five are heavily influenced by normal aging. If they are not all abnormal it is unclear what diagnosis to make.

Current Guidelines Add More Confusion

In 2016 a joint task force of the American Society of Echocardiography (ASE) and the European Association of Cardiovascular Imaging (EACVI) proposed a new and “simpler” system for grading diastolic function (DF.) Despite reducing the number of diastolic parameters to be considered and improving reproducibility, these guidelines have only added to diastologic confusion and clinicians from the Mayo clinic have already called for a revision.

Let’s assume that your patient had his/her echocardiogram done in an echo lab where all the readers have been updated on the 2016 DF guidelines and all of the wall thickness measurements are made properly.

The first thing you should know is that EVERY patient with “myocardial disease” or a low ejection fraction is considered to have diastolic dysfunction (DDF) no matter what all the myriad Doppler parameters reveal. Some echo labs extend this to the mere presence of risk factors like hypertension or diabetes.

I wrote a letter to the JACC expressing my concern about this bizarre assumption and other problems with the DDF guidelines:

The creators of the guidelines responded:

In our paper  and in the guidelines, we discuss how patients can be identified as having cardiovascular disease based on clinical data, 2-dimensional findings, and specific Doppler signals. Dr. Pearson brings up important points about the clinical practice of some, where they interpret left ventricular (LV) diastolic function in the absence of clinical data and 2-dimensional echocardiography findings, including ignoring the presence of LV hypertrophy. These practices are discouraged. There are no good reasons why a physician would not seek clinical data before reading an echocardiogram or why LV wall thickness or LV mass are not measured. This is important because, as discussed in our paper, the accuracy of the algorithm for estimating LV filling pressure is likely to be lower for populations with lower prevalence of cardiac disease.

Having worked in and been the medical director of several non-academic and academic echocardiographic laboratories I can tell you that the vast majority of echocardiograms are read without detailed clinical information on the patient. It is simply too time consuming to browse through years of electronic medical records searching for relevant information and in many cases we have no access to the patient records.

This is an example of the guideline writers sitting in an academic ivory tower where someone else with more time on their hands (typically a cardiology fellow) is sifting through records for them.

According to the 2016 DF guideline writers, the mere presence of coronary artery disease from the clinical history confers the diagnosis of DDF. It’s not clear what should constitute a CAD diagnosis for DF grading purposes. For example, a non-zero coronary artery calcium scan indicates early plaque in the coronaries. Does that count? No one knows.

If all is normal on the echocardiogram, such patients will be given Grade I Diastolic Dysfunction or indeterminate DF.

Let’s say your patient had a cardiac catheterization 5 years ago that showed a 50% blockage in the mid LAD coronary artery and everything is normal on the current echocardiogram.

If the reader of your patient’s echo is fully aware of

1. the substance and nuance of the 2016 guidelines (and follows them rigidly)
2. AND has access to the patient’s records
3. AND manages to go back 5 years in the records to find the cardiac procedure

according to the 2016 guidelines, he/she should give the patient the diagnosis of “Grade I Diastolic Dysfunction.” If any of those three conditions are not met the patient should be reported as normal diastolic function.

How can a physician or patient know whether those 3 conditions were met when reading the echo report? It is not possible to know. Thus, the vast majority of isolated grade I diastolic dysfunction interpretations on echo reports are suspect.

In contrast, the identification of Grade I diastolic dysfunction simply because the patient had pathologic LVH or a low EF is basically meaningless. The diastolic indices do give us rough information on the LV filling pressure but that is a whole different kettle of fish which I shall fry in a future post.

Grade I diastolic function is functionally meaningless

To answer the questions I posed at the beginning of this post

What should you tell your patient about this? Tell them it is a meaningless diagnosis given the abject confusion in the field.

Is it something they should be concerned about? No

Should treatment be started? No

Is additional testing warranted? No

Will it influence their life insurance? Let’s hope not.

Lusitropically Yours,

-ACP

N.B. I have included the PDF of my presentation below