M Mode query

RE: M Mode query

Roger Wilkinson — Thu, 23 May 2019 08:47:55 GMT

I'm doing:

LVEDV-LVESV = total stroke vol

Ao CSA at sinotubular junction x VTI at same place from L apical = forward stroke vol

obvs regurg = TSV - FSV

I was impressed in...

https://onlinelibrary.wiley.com/doi/full/10.1111/jvim.15461

..that the mean difference between clinicians in measuring RF was 1% (at least I think that's what their numbers mean).

In terms of echo geekery this is also a nice thing. Can check ones technique quite nicely by doing those measures on dogs without MR and see if TSV =FSV.

I'm glad you think CHF not easy either Dave! I'm pretty sure you don't often see that admitted by authors writing textbooks

RE: M Mode query

Dave Dickson — Wed, 22 May 2019 18:24:57 GMT

There's also a great set of human echo guidelines on it: https://www.escardio.org/static_file/Escardio/Subspecialty/EACVI/position-papers/eacvi-recommendations-valvular-regurgitation-summary.pdf

RE: M Mode query

Dave Dickson — Wed, 22 May 2019 18:21:36 GMT

Hi Roger

Do you measure regurg fraction by comparing jet area to atrial area? I never have, but mainly because I've been put off by colleagues! How do you use it in the clinical setting?

I think Julia Sargent looked at this a few years ago and what I remember was that it wasn't great.

https://www.ncbi.nlm.nih.gov/pubmed/26473746

I know others have looked and found it to be useful.

https://www.ncbi.nlm.nih.gov/pubmed/17669041

I might start having a play, though typically I find what affects my decision making is my assessment of left atrial size.

Dave

RE: M Mode query

Roger Wilkinson — Wed, 22 May 2019 17:41:46 GMT

This is fab Dave.....much appreciate you sharing your thoughts so generously.

Regurg fraction?....been playing with that recently and I think I can measure it fairly repeatably. Do you think that's useful at all? ...obviously it's interesting to have new numbers :), but does it change decision-making (like when to start pimo?) if it's 60% as opposed to 30%?

RE: M Mode query

Dave Dickson — Wed, 22 May 2019 08:10:19 GMT

Hi Alan

I played around with atrial areas and volumes and gave up. We did a whole study comparing areas to linear measurements (volumes are just a calculation of areas so they are essentially the same thing from a diagnostic point of view). Some studies, like the one you mentioned, find that areas are superior but generally that's because the authors want them to be so the studies are designed to show up areas/volumes in their best light. Whereas in practice, in my hands, areas are hard to measure and really variable. In our study of LA areas vs. linear dimensions, we found that areas added nothing to the clinical picture - you knew it had bad disease from the LA:Ao and measuring an area didn't tell you anything useful. Also judging mitral valve disease severity based on the regurg severity score is fraught with danger - how many dogs do you see with palpable murmurs yet B1 disease? I see plenty! Their valves look awful and they have loads of prolapse, but a normal LA. So I'm wary of these papers. Until people compare measured echo volumes with actual volumes or estimates from other modalities like MRI, I take echo estimates of LA volume with a pinch of salt. Just my opinion though.

In the same breath though, using a single linear dimension of the atrium in borderline cases is also far too simplistic. You need to take 2 or even 3 different views of the atrium, as well as your subjective assessment, and put that all into the mix. And there are still cases that will not fit the general pattern.

As to when to start pimo, it's a minefield. I've seen dogs with an LA:Ao 1.6 and LVDDN 1.8 go into CHF in 6 months. And I've seen dogs with the same numbers unchanged after 2 years. They are all different. In an ideal world, we need to prove progression, but many owners can't afford all the rechecks. In these cases, I discuss the pros and cons with the owner, the risks/benefits, and the costs, then let them choose. ultimately, I see my job is offering the choices and the owners job is to choose.

I'll do another post on how I do echo repeatability studies as a few people have asked. I'll turn you all into echo geeks if you aren't already!

Dave

RE: M Mode query

Dave Dickson — Wed, 22 May 2019 08:00:52 GMT

Hi Roger

I agree with you. EPIC has created some dilemmas when it comes to treatment and I see plenty of cases who are borderline and I never know which way to go. I'm not sure about the argument that the later the stage of B2, the stronger the case - if you want to delay onset to CHF, you want to start as early as possible. Some folks use the KM curve from EPIC to argue this point but I think that is a statistically invalid argument. That isn't to say that starting them really early, ie in late B1/early B2, is a good idea, but I'm not convinced by the argument that they need to be "at least moderately enlarged" before starting treatment. This is ignoring the best evidence we have and relying (as always) on robust opinion and belief. As my mentor Mark Rishniw says, belief belongs in church. I know some people like Mark Kittleson are firmly in the "wait to treat" camp but I think that way lies danger - how big does the LA have to be before you start treatment? Will some dogs miss out on treatment that they need?

I think its pretty cool how much debate EPIC has created though, especially as it has us all thinking hard about how we measure, which is a great thing.

I also totally agree about diagnosis of CHF. It is definitely a spectrum, not an all-or-nothing condition, and diagnosing CHF is probably the most difficult routine thing I do in practice. I had a case last week with elevated pressures on echo, normal radiographs, high normal SRR, but I ended up trialling frusemide and the dog is much better. So who knows! I agree taking a multimodal approach is always the best way - for those of us lucky enough to have the time to do it.

Dave

RE: M Mode query

AlanH — Sun, 19 May 2019 19:07:14 GMT

Hi all,

Thanks for clearing that one up , seriously though appreciate all the comments - very helpful! Echo certainly seems to be something that can create as many questions as it can answer and I definately agree that taking into account the clinical picture is required rather than producing endless measurements.

The EPIC discussion is useful as I have looked at several dogs recently with this exact question - can I justify starting them on pimobendan with a LA/Ao of 1.6 and slightly enlarged LVIDd - so far I have put them all onto medication, I have only measured the LA using LA/Ao and in cats by measuring the width on a right parasternal 4-chamber view - should I look to routinely do LA volume measurements as well ?? If so are people using the area-length biplane measurement or simpsons method?? The paper quoted by Roger (https://www.ncbi.nlm.nih.gov/pubmed/25056960) certainly suggests I should be looking at volume estimation rather than relying on single dimension. We have literally just got a Vivid IQ which looks amazing! and Im wondering wether I can do this calculation on this - shall check 2moro!

Dave - would be interested in the intra-observer variability testing you mentioned, could well be an eye opener ;)

Cheers

Alan

RE: M Mode query

Roger Wilkinson — Fri, 17 May 2019 20:47:24 GMT

I should have specifically mentioned Dave's Springer paper -it was a revelation (at least to me). I ran back to the practice and anxiously reviewed all the Springers I'd echo-ed in recent times.

There are some outstanding interesting issues as far as I'm concerned with echocardiography. I realise this is a slight expansion from the original question into cardiology in general but it's interesting stuff...

Essentially, on the basis of a one-off echo in a given dog it may be impossible to distinguish between B1 and B2. As Dave says, if dilation is marked then exact measurements don't matter. If the matter is more borderline the trouble is, making more accurate measurements doesn't solve the problem (that we still don't know whether this heart has got bigger or not).

That's the root of the biggest issue with EPIC: that we don't really know how many of the dogs might actually have been in B1. A tighter, more unequivocal set of inclusion criteria would have increased our confidence in the outcome. As it is, the possibility exists that there could have been (unintentional) bias in the B1/B2 composition of the treatment/placebo groups. I feel that we're left with a scenario where we are often choosing when to start pimobendan on a slightly ad hoc basis. There's a general feeling that the later the B2 the stronger the case for pimo but it's hard to draw a line.

To be even more controversial I'm not even sure that we can tell for sure whether some dogs are in CHF or not. It's always painted as a relatively all-or-nothing phenomenon. However, I suspect that the development of pulmonary oedema may be more gradual than conventionally presented. There are dogs out there with localised lung changes consistent with pul oedema on lung ultrasound. Respiratory rate is obviously useful as an indicator but there are lots of potentially confounding factors. Radiography in human medicine is relatively insensitive in the diagnosis of pulmonary oedema (something like 40-60% in published papers). Lung ultrasound is probably more sensitive. CT probably more sensitive again. Although some echocardiographic paremeters are strongly associated with CHF, none of them are foolproof in borderline cases.

I used to feel pretty confident writing echo reports. It was all nice and simple. Now I find myself getting more and more indecisive! At a practical level I find myself needing more and more case-specific history, physical exam, echo, lung ultrasound etc etc to try to come to a rounded decision on what to do.

Probably none of that is a surprise to anybody. I just feel maybe available sources tend to simplify a bit.

Roger

RE: M Mode query

Dave Dickson — Thu, 16 May 2019 11:02:32 GMT

Hi everyone

Some interesting replies and I agree a bit with what everyone has said. Forgive me, but measurement of echo is one of my research interests so I'm going to go into more detail than most people probably want!

Firstly, as everyone knows, echo (like all tests) has inherent inaccuracies. What it gives us is an estimate of cardiac size - but it is still just an estimate. As tests go, it's a fantastic one, but it isn't perfect. What makes people really good at echo is knowing when to measure and when not to measure. Alan - I see the same thing as you: in some patients, no matter how well aligned you try to be, there is sometimes asynchronous contraction of the LV. In some dogs, as Sarah rightly points out, it is due to the 'twist' of the left ventricle. I'll see if I can add an echo example of this. In those cases, I do my 'best guess' of the measurement. It isn't perfect, but it's good enough for me. Alan you are entirely correct, the cursor should be placed on the tissue-blood interface on the peak inward incursion of the septum. Sometimes I'll fudge the free wall cursor to match what I think is better, rather than going exactly on the tissue-blood interface of the free wall.

I've published some normal reference values in Springers (I have a second paper coming soon, with a larger sample size and including radiographs and proBNP) and yes, we definitely see a subset of Spingers with big hearts that don't contract very well. We are getting better at differentiating this normal population (which I think are almost certainly athletic remodelling) from disease but as Roger said, it can be hard as we also see some myocardial disease in Springers - though much less than people think.

When it come to measurements I'm with Roger - know which ones are useful, when to use them and when not to use them and you'll be fine. If the LA is huge, who cares if the LA:Ao is 2.5 or 2.7? Same fo the LV - if it's massive, who really cares? However, when the fine judgements need to be made (like B1 vs B2 dogs), measurements are the difference between treatment and no treatment. We have to use certain measurements (like LA:Ao and LVIDd in EPIC) because they are what determine the treatment decisions. So we need to be able to use these in clinical practice, despite their flaws. Also when measurements really become important is when doing serial examinations, to follow dogs over time, or compare treatment effect. Without measuring, it is almost impossible to make proper clinical decisions in specific cases.

Ditching short-axis views altogether is interesting! I know some cardiologists prefer long-axis to short, but the wealth of evidence in vet cardiology shows that short axis views are reliable and repeatable and most of the reference values come from short-axis measurements. They aren't perfect, but if you make measurements you need a reference to compare to. M-mode gives us a way to look at the moving heart with a high frame rate and some specific things are best looked at with M-mode. I agree with David Mills that it is a simplistic way of looking at a complex structure, but it still serves a purpose and if done carefully it is a valuable tool (at least I find it to be). I have done my own measurement repeatability studies, to see how repeatable I am. This is a good exercise for everyone to do - so you know how good/bad you are at a test. I'm very good at volumes, pretty good at the left atrium and aorta, pretty good at M-mode and terrible at the right heart!

If anyone is interested I can give you a method for doing your own mini intraobserver measurement variability study. It's one way of doing some internal validation and checking that you are doing a good enough job with measuring the heart. Of course, I realise most people aren't as geeky as me about echo.

Not sure if that helped or not Alan, but at least it showed you not to ask echo questions unless you want an essay in response! Can't help myself.

Dave

RE: M Mode query

David Mills — Thu, 16 May 2019 00:56:50 GMT

M mode is fraught with errors, and whilst it produces a nice picture, I haven't used it for years. Think of the physics of its production, one line through a complex, impressive, beguiling organ.

For the same reason I don't use short axis views.

Just dispense with it altogether would be my advice.

RE: M Mode query

Roger Wilkinson — Wed, 15 May 2019 22:56:59 GMT

If I were honest, Alan, I would suspect you are a bit oblique across the ventricle rather than truly perpendicular to its long axis. But it's not just you! In some dogs it's not easy at all to get nice transverse plane views due to chest conformation. I find that several of the 'standard' echo views are actually quite difficult to get spot on -which is something that you don't really see represented in a lot of texts. Reproducibility is often an issue. For example transverse LA:Ao doesn't correlate very well with LA volume -especially with more severe enlargement:

https://www.ncbi.nlm.nih.gov/pubmed/25056960

When I started doing referral echos I used to always save one image with the LA looking big and another with it looking normal: so I could decide later on which I wanted to use ;)

That's not necessarily to say that it's not worth measuring lots of parameters but, ultimately, a lot of the time you're almost as well judging systolic function subjectively.

Blinking Springer Spaniels are very difficult. There are a lot of perfectly healthy, athletic springers out there with FS and EF way below what you'd expect in other breeds....and there are also Springers with primary myocardial failure who develop CHF as a result.

hope that's some help

Roger

RE: M Mode query

Sarah Keir — Wed, 15 May 2019 20:59:58 GMT

I am not a cardiologist but... the ventricles and septum do not contract in unison, it is more of a spiral contraction which is better appreciated on functional MRI. I do a best guess as to maximum contraction of both septum and free wall in M mode when I know I have position as optimal as I can. Now for a cardiologist....

As an aside, my perfectly healthy springer spaniel had a resting FS of 12%. She did not have DCM and lived to nearly 16. Did freak me out when I used her for practice but was lovely as her resting HR was incredibly slow too so the old machine could cope with frame rate!