Strokes/INR/Mechanical valves

[perimeno]

"Recently, a question was discussed on this website concerning INR and the
likelihood of strokes following aortic valve replacements. In one reply,
reference was made to research which appeared to imply that after 10 years
upwards of 25 per cent of people with the St Jude mechanical valve had
suffered strokes -- despite the benefit of Warfarin. Is the inference drawn
correct? What is the long term likelihood of strokes for people with SORIN
or mechanical valves other than St Jude?"

Thanks,

Peter

 

[Valve-Viking]

Hi Peter,

I did not see these posts, but my spontaneous thought is the age of an MVR-patient has to be considered when you start comparing. I mean you could say there is a great risk of dying within 10 years when you replace your valve at the age of 85 too......by mere age the risk of a stroke is probably quite big even without any MVR too. There´s a lot of factors to consider and I think
it´s hard to compare different valves in general ways due to it as well.



/
Martin

 

[liamty]

I'll have to look back, I too missed those posts. Steve is 23 years post avr op, and the concern of stroke seems no more of a threat now than it was back then. We were mindful then and still are. It is worth checking though!!

 

[tobagotwo]

Peter, I have been doing searches, and can't find the post that those numbers originally came from. This is of interest, as it seems out of kilter to much of the other information I have seen. It would be good to know whether there are any limitations to this study that might tend to produce the results that it apparently did.

Is there any possibility that you know the source for this?

Thank you,

 

[Gemma]

Peter,
When I read your question it reminded me of a post by Al Lodwick a few weeks ago about stroke risks with various valve types:

"The Annals of Thoracic Surgery (2004;78:77-83) has an article about the risks of stroke after valve replacement. The study was done by Ruel et al in Canada. They studied 3,189 adults with various types of valves for a total of 20,096 patient-years. This is a big study over a long period of time so the results should be pretty valid. Here are the stroke rates and some other bits of information.

Site Type Stroke Rate
Aortic Bioprosthetic 1.3% per year
Aortic Mechanical 1.4% per year
Mitral Bioprosthetic 1.3% per year
Mitral Mechanical 2.3% per year

Those who had two valves were not at greater risk for stroke.

Taking aspirin or dipyridamole in addition to warfarin did not lower the stroke risk.

It does not appear that they were able to include information about whether the people kept their INR in range or not. Presumably this is everybody lumped together. Some would have tight control of the INRs, some probably skipped considerable doses of warfarin and there were even times when people were off warfarin. This is a look at real-people in the real-world averages."

I can't recall seeing the 25% figure - but hope this post by Al helps. Guess it was done over about 7 years, judging from the patients/patient years figures.
Gemma.

 

[tobagotwo]

That is one of the studies that the 25+% information is out-of-kilter with...

I'm hoping there's a clue in the source which would help us to define the large variance in results.

Best wishes,

 

[bvdr]

Well, if the risk of mechanical mitral valves is 2.3% per year, then is the risk projected over a ten year period of time 23%? It has been years since I took statistics and I really am not sure of this.

 

[perimeno]

strokes

strokes

Hi All,

Thank you for the replies you sent my brother, Peter. I was not able to call him tonight as it is 2 am in England now. But I think this is the post I copied from this site and sent to him a while back, and is what he was referring to. I'll check with him tomorrow and also send him your comments.

Thanks again

Peri

khan2 11-11-2003 02:54 AM

--------------------------------------------------------------------------------

I'd definitely agree that the risk of stroke (embolism) is generally higher in patients receiving a mitral prosthetic valve than an aortic prosthetic valve. However, in an article I published on our first 1,000 St. Jude valves in 1994, we did not see a significant difference:

"With 4328 patient-years of follow-up, 83% of the mitral group, 76% of the aortic group, and 77% of the double valve group were free of thromboembolism at 10 years after operation,"

The St. Jude Medical valve. Experience with 1,000 cases. Khan S, Chaux A, Matloff J, Blanche C, DeRobertis M, Kass R, Tsai TP, Trento A, Nessim S, Gray R, et al. J Thorac Cardiovasc Surg. 1994 Dec;108(6):1010-9; Division of Cardiothoracic Surgery, Cedars-Sinai Medical Center, Los Angeles, CA 90048.

The general recommendation with the St. Jude valve is to run the INR 2.5-3.5 with aortic St. Jude valves being kept at a target of 2.5 and mitral St. Jude valves at a target between 3.0 and 3.5. Since there is some natural variation in the INR, this should hopefully result in the aortic patients running an INR between 2.0 and 3.0 (with an average of 2.5).

 

[tobagotwo]

Thank you for supplying that, Perimeno. It helps me understand.

The current St. Jude site talks about their "gold standard" (Regent) valve design having not changed in 18 years. The study was from 1994, a decade ago. That means that any St. Jude valves over 10 years old at the time of the study had to be an older model, not the current, Regent design.

It was heart care twenty years ago, as well - different drugs, different priorities at the cardiologist's.

It was also before home testing, which improves INR compliance considerably.

As the average person in the study had owned his or her valve for only 4.3 years, the valves over 10 years old had to be in a minority.

I would look to a more current study, such as in Al's post, for a more accurate view of the current equipment from St. Jude.

Betty, the Jurassic Cowboy did some interesting statistical math in one of his recent threads, which was in response to that exact question: http://www.valvereplacement.com/forums/showthread.php?t=8420&page=2&pp=30 Basically, I believe he said no, that would be an overestimation. I agree with him in principle, but my statistics gets fuzzy on and off, as it frequently argues vehemently with my other logical functions.

Perhaps someone out there can collar their college-attending, business-major offspring, and prod them to do the statistical math for us lazy oldsters.

Best wishes,

 

[Guest]

If the risk is 2.5% per year, then that would mean that it would be about 40 years before all of the group had strokes. Maybe not exactly but close enough. When you weigh that against your risk of living trouble-free had you not gotten the valve it does not look too bad.

It also assumes that everyone has the older model valves. The new ones may be better. You have to remember Billy's wife (in Northern Ireland) who was given a new St. Jude model that was not so good. If they have a life expectancy of about 20 years, that also means that you will be getting a (hopefully) less risky model before your 40 years is up. Also the anticoagulation business should be much improved in another few years.

So even if you had someone run the statistical analysis there are still too many variables to come up with a valid conclusion. Just play with the hand you have been dealt and don't worry about the odds of getting a royal flush or getting royally flushed.

 

[PapaHappyStar]

I looked up that paper, the abstract with a result is available here:

http://www.sciencedirect.com/scienc...d=864464&md5=74663328e77123720336df368e70f09f

It was mentioned before that this is dependent on whether you have independent events ...
if the probability of the event happening in any year does not depend on the outcomes of previous years, then you can use a binomial type probabilty ( i.e. like a coin toss -- two outcomes -- stroke, no stroke ).

so calculating from the numbers given, assuming independent probability ( and validity of this approach ):

valve type		3y	5y	10y	15y	20y	30y	40y
Aortic Bioprosthetic 	3.9 	6.4 	12.3 	17.9 	23.1 	32.5 	40.8
Aortic Mechanical 	4.2 	6.9 	13.2 	19.1 	24.6 	34.5 	43.2
Mitral Bioprosthetic 	3.9 	6.4 	12.3 	17.9 	23.1 	32.5 	40.8 
Mitral Mechanical 	6.8 	11.0 	20.8 	29.5 	37.3 	50.3 	60.6

since they mention that they see a 20% rate after 15 years this calculation seems reasonable upto 15 years, i.e. not much difference between mechanical and bioprosthetic valves in the aortic position. They dont mention anything about a control sample ( and I havent included statistical errors ). There are standard disclaimers about method and sampling at the end etc.

if anyone is interested I have a full copy of the paper in pdf format...

 

[PapaHappyStar]

There is a commentary to this paper which raises the point ( for aortic valve ):

The group of patients recieving bioprostheses would be older than the group with mechanical prosthesis, so the incidence of stroke in the bio-prosthesis patients would be higher due to aging related causes not coupled to their valve prosthesis. This, in the opinion of the commentator, is indistiguishable from valve related stroke. I did see a age realted correction factor in their analysis but am not familiar with their technique ( they seem to use a linear fit to the data, to come up with the independent probability estimate and correct their data for independent risk factors for stroke ). This would be a systematic error overestimating the incidence of stroke in the bio-prosthesis group.​

I think this is a valid critique -- I dont see a mention of subtracting a age-related background rate for the incidence of stroke.

Here is a link to the citation:

http://www.sciencedirect.com/scienc...d=864464&md5=8af488bba7fa11eb3b3e7780ce3100d1

I have the full text for this too...

 

[catwoman]

Do the statistics also include vr patients with a-fib or other problems that might produce clots, thus strokes?

 

[PapaHappyStar]

Do the statistics also include vr patients with a-fib or other problems that might produce clots, thus strokes?

Here is what they have to say:

Atrial fibrillation

Chronic atrial fibrillation was an independent risk factor for embolic stroke in patients with aortic prostheses, with a relative risk of 2.2 (Table 3). In patients with mitral prostheses, atrial fibrillation lost statistical significance as the stronger effects of advanced left ventricular dysfunction (see below) on embolic stroke risk concomitantly entered in the model ( Table 4).

I will e-mail you the full text of the paper ( pdf ) so you can see this in context.

Regards,
Burair

 

[JimChicago]

I would wonder too what their criteria for a stroke was? Did it take into account transient events that left no damage or just include strokes that left paralysis or other problems?

Most of the other studies I've heard of just showed liklihood of adverse events including bleeding, stokes, etc not just limited to just strokes like this statistic in this thread.

 

[PapaHappyStar]

Here is their definition:

Definition of stroke

The occurrence of stroke was defined, in accordance with the Guidelines for Reporting Morbidity and Mortality after Cardiac Valvular Operations, as the presence of a neurologic deficit lasting more than 3 weeks [1]. Patients diagnosed with a stroke after 1986 had computerized tomography (CT) of the head. Strokes were characterized either as embolic or as an intracranial bleeding event based on the primary mechanism and CT appearance of the lesion. The diagnosis was confirmed by a neurologist or internist in all cases.

 

[PapaHappyStar]

Here is the entire section on the risks associated with higher incidences of stroke.

Risk factors for embolic stroke
Gender

Female gender was an independent risk factor for embolic stroke in patients with an aortic or a mitral prosthesis (Table 3 and Table 4). Women with prosthetic valves had approximately 1.7 times the embolic stroke risk of men, despite adjustment for other risk factors.

Table 3. Multivariate Predictors of Late Embolic Stroke?Aortic Prostheses
Full Size Table

Nonsignificant variables also included in the model were: history of cerebrovascular accident, left ventricular grade 3 or 4 at the time of surgical referral, diabetes mellitus, primary indication of aortic insufficiency versus stenosis, redo status, mechanical aortic prosthesis, currently available versus discontinued aortic prosthesis, manufacturer aortic prosthesis size, warfarin anticoagulation of an aortic bioprosthesis (within subset of aortic bioprostheses), and aspirin use.

CI = confidence interval.

Table 4. Multivariate Predictors of Late Embolic Stroke? Mitral Prostheses
Full Size Table

Nonsignificant variables also included in the model were: history of cerebrovascular accident, atrial fibrillation, diabetes mellitus, primary indication of mitral insufficiency versus stenosis, redo status, concomitant tricuspid valve procedure, number of bypass grafts performed, currently available versus discontinued mitral prosthesis, manufacturer mitral prosthesis size, tilting-disc versus bileaflet mitral valve (within subset of contemporary mitral mechanical prostheses), warfarin anticoagulation of a mitral bioprosthesis (within subset of mitral bioprostheses), and aspirin use.

CI = confidence interval.

Age

Age more than 75 years was an independent risk factor for embolic stroke regardless of the site of implant or type of prosthesis. The embolic stroke risk was 1.9 times higher in elderly patients with aortic prostheses, regardless of other risk factors and valve type, and 3.1 times higher for elderly patients with mitral prostheses.

Atrial fibrillation

Chronic atrial fibrillation was an independent risk factor for embolic stroke in patients with aortic prostheses, with a relative risk of 2.2 (Table 3). In patients with mitral prostheses, atrial fibrillation lost statistical significance as the stronger effects of advanced left ventricular dysfunction (see below) on embolic stroke risk concomitantly entered in the model ( Table 4).

Coronary artery disease

Coronary disease was a significant risk factor for embolic stroke in patients with an aortic prosthesis, where a 24% increase in relative risk per additional graft required at operation was found (Table 3). This relationship was not significant in patients with mitral prostheses.

Left ventricular function

Depressed left ventricular function at the time of surgical referral (defined as grade 3 or 4; Table 4) was independently associated with an increased risk of embolic stroke in patients with mitral prostheses. Patients with grade 3 or 4 left ventricular dysfunction preoperatively who received a mitral prosthesis had 3.1 times the relative risk of embolic stroke of patients with a mitral prosthesis and normal or mildly depressed left ventricular function (grade 1 or 2), despite adjustment for other risk factors. Separate analyses similarly revealed that a 61% linearized augmentation in embolic stroke risk was observed for each unit increase in preoperative left ventricular grade ( Fig 2).


Enlarge Image (6K)

Fig 2. Risk-adjusted effect of increased preoperative left ventricular grade on late postoperative embolic stroke in patients who underwent mitral valve replacement. With each unit increase in preoperative LV grade, a linearized stroke risk augmentation of 61% was observed. (LV = left ventricular.)

Smoking status

Both current and past smoking were strong independent risk factors for embolic stroke after heart valve replacement. Current smoking was independently associated with a relative risk of stroke ranging from 2.3 to 3.0, regardless of the site of implantation or type of prosthesis. These effects were additive and independent to those of previous smoking, which was also significantly associated with an increased stroke risk in both implant positions.

Type of prosthesis

Patients with mechanical mitral valves had 1.24 the risk of embolic stroke of patients with bioprosthetic mitral valves after adjustment for other risk factors (Table 4). This relationship was not observed in the aortic position, where the embolic stroke risk was not significantly different between mechanical and bioprosthetic valve patients.

Within the subset of aortic mechanical prostheses, contemporary tilting-disc aortic prostheses (ie, Medtronic-Hall) were associated with a 1.7 times higher stroke risk than contemporary aortic bileaflet valves (ie, St. Jude and Carbomedics). This relationship was not observed with mitral mechanical valves.
Antiplatelet and anticoagulation therapy

There was no difference in embolic stroke rates between patients with bioprosthetic valves who had been anticoagulated for a period of 3 months after operation versus those who had not. The adjunct of aspirin or dipyridamole had no significant effect on the incidence of embolic stroke in patients with mechanical valves, regardless of implant site or valve model.

 

[tobagotwo]

General thoughts about studies

General thoughts about studies

This is a group of general thoughts relating to studies, and how we may choose to relate to them:

While it is important to address the realities of valve disease, it can help to bear in mind that not all studies relate to all individuals, and that numbers can be accurate without having real meaning. We do realize that statistically, we are not actually "good as new," but we are not all that far off. I believe it much more useful to look at the relative risks indicated by these studies, rather than treat the numbers as an accurate forecast of events. By their nature, the numbers will change with changes to the hardware, peripheral treatments of related health issues, and other environmental factors.

It should be a thought in mind that various health-related groups obtain their funding by scaring the pants off the public and producing OMyGosh numbers to ballyhoo their causes. Frequently, their numbers are accurate, but don't actually portray what they appear to. Heart attacks and strokes are the number one killer of people in the US. True. And many of those people are in their 80s or 90s (have to be - look at the average US life span). But those numbers are aimed at people in their 40s, who have disposable incomes to donate.

It's not that it is not important, but the implied link to the average 42-year-old is a very weak one, indeed. Try to look at studies insofar as they reasonably relate to your own, personal circumstance.

Another common statistical misdirection is to say that your risk "doubles" when you (insert "bad" thing here). Well, if your risk was .00001%, then your risk is still negligible. But it sure got your attention. The doubling or tripling of your apparent risk of something is absolutely meaningless without knowledge of your original, specific risk (by age, etc.).

Most studies don't intend to misdirect. As well, they don't generally pretend to provide perfect and complete answers to a proposition. They usually are aimed at adding another, usable portion to a larger, composite knowledge.

We frequently only look at a summary of their results, denying ourselves the context of the numbers given, and thus some of their meaning. As such, it makes sense to temper our reactions as much as possible, in relation to what we really know about the genesis of the numeric outcomes.

It is interesting to know that 52% of the people who have had procedure X feel "crappy" three weeks later. However, it is more useful to know how you feel.

Best wishes,

 

[PapaHappyStar]

right but reading opinion is also performing a study albeit with larger error bars -- one ought not to discount information, most of it is comon-sense .... i.e. dont smoke if you have a prosthetic valve.

not paying attention to opinion or statistical study borders on solipsism, i would say study everything and pay attention ( i.e. look before you leap -- and once you've leapt look some more before stepping forth )

 

[Guest]

Studies give a reasonable assessment of what is going on and never apply to any one individual. When you were getting ready for surgery the question was, "How long will this valve last?" or "How long will I live if I don't do this?". The studies gave an average. It was good to use in making a decision but not very useful in planning future events.

If the doc said that the valve would last 20 years, you didn't schedule another surgery on that date 20 years in advance, nor reserve space at the funeral home in 180 days if you were given 6 months to live. You used them to guide you in making a "global" decision.

The other one that gets me is when the news reporter says that you have 30 times less risk... Shouldn't it be 1/30th of the risk ... but most people hate dealing with fractions.