Learn how advances in biomedicine hold the potential to revolutionize drug development, drug treatments, and disease prevention: where are we now, and what does the future hold? This course will present short primers in genetics and mechanisms underlying variability in drug responses. A series of case studies will be used to illustrate principles of how genetics are being brought to bear on refining diagnoses and on personalizing treatment in rare and common diseases. The ethical and operational issues around how to implement large scale genomic sequencing in clinical practice will be addressed. After completing this course, learners will understand 1. The ways in which genetic variants can contribute to human disease susceptibility 2. How to choose among drug therapies based on genetic factors 3. That the functional consequences of the vast majority of genetic variants discovered by modern sequencing are unknown. This course is targeted primarily at physicians 5+ years out of training. Other healthcare providers, medical/health sciences students, and members of the public may also be interested. Course launches January 15, 2016. * The information presented in “Case Studies in Personalized Medicine” is offered for educational and informational purposes only, and should not be construed as personal medical advice. If you have questions or concerns about a medical matter, please consult your doctor or other professional healthcare provider.
A Bleeding Complication 5 Days After Starting Warfarin
Loading...
Из курса от партнера Vanderbilt University
Case Studies in Personalized Medicine
162 оценки
Explore Related Videos
At Coursera, you will find the best lectures in the world. Here are some of our personalized recommendations for you
Из урока
UNIT 3: CASE STUDIES IN PERSONALIZED MEDICINE, PART 1
In Module 5 we will begin to discuss specific cases as these apply to personalized medicine. We will first look very closely at a case of familial hypercholesterolemia as we investigate how we use genomic medicine to move from a rare disease to a common medication, using genomics to find new drug targets, and a discussion of the side effects of statin therapy. In Module 6 we will look at a collection of "high risk pharmacogenetics"cases that illustrate adverse reactions due to drug metabolism and variable drug responses.
Познакомьтесь с преподавателями
Dan Roden, M.D.Professor of Medicine and Pharmacology
Assistant Vice Chancellor for Personalized Medicine
[MUSIC]
This module continues our exploration of the sources of variability
after administration of a drug, looking at the genetics in particular.
And the example for today is a bleeding complication after starting Warfarin.
As you'll see, Warfarin is a drug where there's a complication because there's
more than one gene that plays an important role in variability in its action.
So, the case is a 76 year old woman who is noted at examination to have incidental
atrial fibrillation.
She's started on a standard dose of Warfarin, five milligrams a day.
Has other medications,
none of which are very likely to interact with Warfarin or cause problems.
Five days later she presents with severe back pain,
and the question is what should you be worried about?
Now, in an elderly patient there's all kinds of things you could worry about,
including a compression fracture.
But having just started Warfarin,
one of the things you should be aware of is that there's a potential for
hemorrhage, and in this particular case, a retroperitoneal hemorrhage.
This is a CT scan of the abdomen in a patient,
who has a very large retroperitoneal hemorrhage, as shown by the arrows.
So, why does this happen to some individuals
relatively early during Warfarin therapy?
Now, Warfarin is a drug that has a relatively narrow therapeutic index, and
what I wanna do on this slide is to illustrate to
you what I mean by a narrow therapeutic index.
Two dose response groups are shown here.
One shows the cumulative incidents of a therapeutic action as a function of
increasing dose or plasma concentration on the x axis,
and the cumulative incidence on the y axis.
That's in blue.
And the other curve shows the incidence of adverse events as a function of
increasing dose or concentration.
You can see these curves are widely separated, that the drug that we're
interested in here, has a high likelihood of producing a therapeutic effect.
And actually not all patients get adverse events even when you push
the concentrations as high as you'd like them to.
This is a drug that has a very, very wide therapeutic index.
Here's an example of a drug with a narrow therapeutic index,
the curves are much closer together.
Everybody gets side effects if you increase the dose high and
not everyone will get a therapeutic action.
And what's most important is at concentrations, or at doses
that produce efficacy in some patients there will be adverse events in others.
So, that's a drug with a narrow therapeutic window, or
a narrow therapeutic index and Warfarin is an excellent example of that.
Little too much drug, adverse event.
Little too little drug, no therapeutic effect.
So, it turns out that bleeding, or complications due to blood thinning drugs,
Warfarin, Clopidogrel, and aspirin, are the commonest causes for
admission to hospital.
This is a list of admissions to hospital in two general hospitals in Great Britain.
The most common category of drugs that provoke admission,
are non-steroidal, anti-inflammatory drugs.
They include aspirin in that group.
The second most common are diuretics.
Many, many diuretics producing adverse events.
But Warfarin stands by itself as a drug, alone in third place, and
when you add up the Clopidogrel complications, and
the aspirin complications, bleeding due to the these drugs,
is one of the commonest cause of admission to hospital because of side effects.
So, it becomes important to understand mechanisms underlying
variability in response to any anti-coagulant drug that we use.
Warfarin it turns out,
is administered as that is it's an equal part of S-wafarin and R-warfarin.
S-warfarin is the active in R-warfarin is less active.
It's not devoid of activity, but it's less active.
S-warfarin undergoes bioinactivation through metabolism at the seven hydroxy
position and the enzyme that accomplishes that bioinactivation is cyp2c9.
So, another player in the Cytochrome P450, or CYP super family.
There are two functionally important and common snips in the CYP2C9 gene.
Both of them are non synonymous.
They change in amino acid and they change function.
The *2 Allele is relatively common.
12% minor Allele frequency in Caucasians.
And it has reduced function.
The star 3 Allele, a little less common, 5%, but it's a loss of function Allele.
So, here is a study from the late 1990s that was one of the first to explore
the relationship between variation in a CYP, CYP2C9 in this case,
and outcome of treatment with a drug that was a CYP2C9 substrate Warfarin.
So, the first panel that I'm showing you shows the frequency of
different genotypes in patients who are receiving normal doses, 5 mg a day, or
so is a normal dose, and patients in whom the dose was adjusted downwards because of
monitoring the INR, which is the standard way of adjusting Warfarin dose.
So, and patients who end up on a maintenance dose of Warfarin that
is less than 3 mg a day, the frequency of varying Alleles, *2,
*2, *2, *3, *2, *1, *3, *3, is much higher than
in the reference population where most patients are *1 ,*1.
Some of them are heterozygous, and very few are homozygotes or
compound heterozygotes.
That's one observation.
The other observation, which is really interesting and
actually has not yet been perfectly explained.
Is that patients who ended up
on less than 3 mg a day tended to have more bleeding events.
It's as though they were on 3 mg a day,
but some days they would need 2 mg a day or some days they would need 4 mg a day.
That's what we think is going on, but whatever the reason, patients who end up
on a lower dose of Warfarin tend to have a higher incidents of bleeding long term.
And that maybe because of the genetics or
because of factors that have not yet been clearly understood.
So, if you were to draw a pathway for Warfarin disposition
at the beginning of the 21st century, this is what it would look like.
You would say S-warfarin, the active metabolized by CYP2C9 and are warfarin
less active metabolized by something else to produce inactive metabolites.
Now in 2004, investigators in Germany looked at a family,
or a series of families with an unusual pharmocogentic trait and
that is complete absence of response to Warfarin.
What these families have is you can give them hundreds of milligrams of Warfarin
and nothing happens to their INR.
And they discovered that mutations in a gene they call V-K-O-R-C-1 or
VKORC1, are the cause of Warfarin resistance.
And it turns out that Warfarin, a drug that has been used since the 1950s as
anti-coagulant and parathentically as a rat poison, it's pharmaceutical target,
the protein with which it interacts to produce its effects,
was actually unknown until this study came out.
It turns out the protein encoded by VKORC1, which is also called VKORC1,
is in fact the protein with which Warfarin interacts to effect vitamin K synthesis,
and that's how Warfarin acts as an anti-coagulant.
There are variations in the promoter region of the VKORC1 gene.
That's the region that controls how much VKORC1 the liver makes.
And there are variations that are classified by haplotype.
This is a simple example haplotype A and haplotype B.
You can see there are sequences across the nucleotide sequence as shown here
that are different, [COUGH] but they attract together.
Haplotype A is associated with much less liver production messenger RNA for
VKORC1 and haplotype B.
So the prediction would be that, that would result in less protein.
And therefore, patients who carry haplotype A would require less warfarin
for inhibition of vitamin K dependent synthesis of clotting factors.
So if you draw the pathway now, you would draw it this way, but
VKORC1 modulates the cycling of vitamin K and the fashion shown on this slide and
there are two major genes that effect warfarin response, CYP2CP and VKORC1.
VKORC1, the variants are not in the coding region,
except in those rare patients who have the warfarin resistances trait,
but they are in the promoter region.
So that's the trait and then there are multiple other genes that
control different enzymes that are responsible for cycling.
I'm not gonna talk very much about them,
I'll just point out that one on the left called CYP4F2 that's responsible for
vitamin K metabolism itself and we'll come back to that one in a second for a second.
One of the things that people who administer warfarin around the world have
known for a long time is that different ancestries require different maintenance
doses of warfarin.
So the idea is your start warfarin you measure the RNR over time and
you adjust the dose for the RNR's between two and three and
everybody ends up on a different, stable warfarin dose.
Caucasians, the average dose is five milligrams a day, average.
Some people need less, some people.
Among African subjects, it's about six milligrams and among Asian subjects,
it's about three or four milligrams.
Those don't sound like big differences, but of course,
a little bit of overdose with warfarin and
you will have a bleeding complication like the patient I presented initially.
And too little and you'll have a thrombosis complication,
because of failure of drug efficacy.
We at Vanderbilt along with many, many other investigators around the world were
part of something called International Warfarin Pharmocogenomics Consortium
that investigated the genetic basis for variability of warfarin response and
one of the the first studies we did was to look at the relationship of CYP2C9 and
VKORC1 genotypes and steady state dose.
So, it turns that the genotypes vary by ethnicity and
the VKORC1 genotype that predicts a lower dose requirement is
much more common in Asian subjects than it is in African subjects.
They're labeled as AA and AG here and not haplotype A and haplotype B,
but the AA is the haplotype A that I show you on the previous slide.
The GG is the haplotype B.
You can see that the frequency of those star two and star three alleles are much
higher in the European population than they are in the Asian or
the African populations, we'll come back to that in a moment.
So one of the things that the Warfarim Pharmacogenomics Consortium did is showed
in this slide and it's a little complicated, so
let me walk you through it.
The idea is that you can use clinical factors that are known to affect warfarin
concentration.
Clinical factors such as age, sex, presence of interacting drugs, such as
certain statins or amiodarone and predict what the study state warfarin dose is.
That's called a clinical dosing algorithm.
You can use a clinical dosing algorithm along with the genetic information that
you know to create a pharmacogenetic dosing algorithm or
you can use what's called a fixed does algorithm.
So what we did is a simulation, basically, consortium that we said.
Suppose we compare those three approaches to prescribing morphine does,
how do well do they do in a very large population of subjects in whom we
know what the dose should be?
So among patients who end up needing five milligrams a day,
a fixed dose regimen of five milligrams a day is perfect.
Would that we knew that those patients would need five milligrams a day, but
that's perfect and the clinical algorithms and
the pharmacogenetic algorithms perform just as well.
The y-axis on this slide shows the proportion of subjects that
are close to their predicted dose by the algorithm.
So as long as you're in the middle,
as long as you're average five milligrams a day is just fine.
Now it turns out that there is 20 to 30% of the population that's below average,
it turns out and 20 to 30% that are above average.
So in those patients, five milligrams a day is not gonna cut it,
because they will never be close to the actual steady state dose.
The clinical algorithms do pretty well, that's the blue bars and
the pharmacogenetic algorithms at the extremes.
Do a little bit better than the clinical algorithms.
So this is becomes important when I talk about the randomized clinical
trials looking at warfarin dosing.
So genetics seems to add a little bit.
Not surprisingly, it adds a little bit of the extremes.
It doesn't help in the average subject and if you knew your subject was average,
you wouldn't need genetics.
But of course, we don't need it until we have it and
we know that the subjects are average or not.
People have done GWAS's looking at the relationship
between common genetic variance and steady state warfarin dose.
This is a GWAS in a large group of Caucasian subjects.
The top panel shows the raw results and
shows that there's a very strong signal at or near the CYP2C9 locus,
not surprisingly and at the VKORC1 locus, not surprising either.
If you condition the analysis, you do the statistics and say, let me make CYP2C9 and
VKORC1 at covariant, and see if there's any other signals.
The other signal that emerges is shown on the bottom side and
that's the CYP4F2 variant.
So it turns out that there is a variation in that CYP.
[COUGH] So it turns out that there is a variation that CYP that is responsible for
vitamin K metabolism itself that contributes to variability in
warfarin dose, but not to the extent that CYP2C9 and VKORC1 do.
Now the other problem is that of ancestry.
So, I showed this slide before and I showed you the incidents or
the frequency of the star two and star three oils is much higher in European
populations than it is in African populations or Asian populations.
The table at the bottom shows you the oil frequencies across Europeans and
Africans at our own hospital when we look not only at star two and star three, but
a couple of other variants that are known to be common in African subjects,
star 6, star 8 and star11.
And of course, it turns out that those are hardly detectable at all in European
populations but quite prevalent In the African population.
So any algorithm, any approach that we're gonna use to try to personalize
warfarin care has to take into account the fact that there are variant
alleles that are ancestry specific.
There aren't the Caucasians and
Asian subjects tend to look the same in this regard.
The star two and the star three are the ones that are frequent in Asian
populations, but they're obviously as you can see not as frequent in Europeans.
We've gone on to do a genome wide association study as part of
the International Warfarin Consortium in African american subjects and
this is the raw result and
that signal that you see is VKORC1 signal not a surprise, it should be there.
So again, you do the statistics and say that's the covariant.
Let me see if there are other statistic signals and there is a signal.
That signal's pretty close to CYP2C9, but it's not CYP2C9.
We don't know what that signal is, it's real and it's not a variant that
predicts outcome in other subjects in subjects of other ancestries.
So this is an approach using genome-wide association to To uncover
the mechanism whereby certain ancestries have different dose requirements for
this commonly used anticoagulant.
The other problem with Warfarin is that there are occasional patients who take
Warfarin and don't have much of a therapeutic response at normal doses.
This is a slide showing the relationship between Warfarin plasma concentration and
Warfarin effect in a group of subjects who received very high doses of Warfarin,
generally without much anticoagulant affect untill high doses were achieved.
Now the most common explanation, not surprisingly is that the patient's were
probably not taking the drug, so their plasma concentrations were quite low.
But there are patients who take the drug.
Their concentrations are within the therapeutic range.
Those are the triangles at the top of the graph, and
yet their anticoagulant effect is pretty minimal.
So these are patients with rare non-synonymous variance in VKORC1.
These are patients who have sort of a variant on the Warfarin resistance
syndrome that originally identified VKORC1, but
they obviously have a response.
They don't have total lack of response to Warfarin, but they have a coded region
variant that makes them less likely to respond to Warfarin.
So you would say, well, if I have a patient who needs ten or
12 milligrams a day.
That's an unusually high dose, and I guess they have one of these.
But, if you run an anti-coagulation clinic,
you have to be aware of the ancestry of your subjects.
So, there's one of the variants here is D36Y in VKORC1.
D36Y is rare, except In subjects of Ashkenazi origin,
and there the allele is quite frequent.
It's about 5% of the population.
So if I ran an anti coagulant clinic in an area where most of my patients where
Ashkenazi Jews, I think I would insist on genotyping this particular
variant of VKORC1 as part of my ability to predict dose requirement or not.
But in other populations I don't have to worry about that.
I might have to worry about other rare variants.
So the last part of the Warfarin story is shown here, and
I'm just gonna talk at this slide rather than explain much of what it is.
This slide basically shows that there were three large trials that examined
[COUGH] the influence of using pharmacogenetics to guide Warfarin therapy
during initiation of treatment.
Each of the trials were designed a little bit differently.
The nuances might or might not be important but the idea was pick
either a clinical algorithm or a pharmaco genetic algorithm.
In one study, they picked a fixed dose algorithm versus a pharmaco genetic
algorithm, and then randomly assigned patients to be managed through one
algorithm or another and then asked the question how many days,
over the course of the first 30 days, or 90 days of therapy,
were patients in the therapeutic range for their anti-coagulation?
And the answer is, that genetics contributed very, very,
little to getting people into the therapeutic range.
So one argument is the genetics,
while they appear to contribute mightily To the overall dose of warfarin,
studies using GWAS and other approaches have estimated that 40 to 50%
of the variability in Warfarin dose requirement is genetic.
That genetics doesn't translate into time and
therapeutic range over the course of the first month or three months of therapy.
The other argument is that asking the question about time and
therapeutic range over a month or three months is the wrong question.
What we really want to know, are there subjects that are gonna bleed early or
subjects who are gonna take two or three weeks to get anti coagulated?
And there's not enough bleeding episodes in any of these trials to answer
the question of bleeding susceptibility, so that's left open.
But right now the position of genetics with respect to Warfarin is a little bit
uncertain.
There's no question there's a huge genetic effect,
but how that translates into clinical medicine is not so clear.
Of course, the other part that affects Warfarin dosing is
that there's a whole series of new oral anticoagulant drugs that are available,
which in large randomized clinical trials appear to do as well as Warfarin,
sometimes better, sometimes equivalent.
And the argument is, well Warfarin is a difficult to use drug.
You have to do all this dosing adjustment.
You have to think about the genetics.
The newer anticoagulants, a single dose fits all.
My own bias is that it's unlikely that the new drugs must be really good,
because if you could dose adjust the new drugs because they
are anti-coagulants with a narrow therapeutic range,
you could dose adjust them somehow they would do even better.
So the fact that a fixed dose of a new drug beats dose adjusted Warfarin in
general suggests that these drugs really are very, very effective,
but they're new, and again, expensive.
In certain situations, Warfarin is still the drug of choice, and
we don't actually know what the long, long term outcomes with these medicines may be.
So stayed tuned.
But Warfarin, if nothing else, has been a great example of a drug whose
complicated pharmacogenetics involving not variation in one gene but
common variation in two genes effects mightily the last dose,
the final steady state dose the people end up taking, and may effect response.
So again just to summarize one more time.
Warfarin is an example of a single pathway to drug elimination.
In this case,
CYP2C9 with the extra added feature that it also has variability of its target.
And so drugs that inhibit CYP2C9, there are a couple or genetic
variants that decrease CYP2C9 activity will result in increased Warfarin effect.
[MUSIC]
[MUSIC]
Ознакомьтесь с нашим каталогом
Присоединяйтесь бесплатно и получайте персонализированные рекомендации, обновления и предложения.
Coursera делает лучшее в мире образование доступным каждому, предлагая онлайн-курсы от ведущих университетов и организаций.
© Coursera Inc., 2019 Все права защищены.
