13 Q&A with the Faculty

Q&A Session: Strategies for Rectal Air Artifact in Diffusion Assessments

We've set up a microphone here,
and we'd like to ask you if you have questions
to step up to the microphone.
If you are shy, you can write something out
and we will read the questions out for you.

Question for the whole panel.
Air artifact on the diffusion assessments a problem,
and we've certainly had trouble with that in our practice.
Do you have any general suggestions for strategies
to eliminate rectal air before the exam?
We've asked our radiation oncology planners
and they give gas x believe it or not, things like that.
Have you had any luck with any other strategies?

Okay, Andy, you sort
of addressed that before.
Maybe you can repeat
and then we can quickly go through the panel
and see what the suggestions they have.
Because for most of us, this is a problem.
I wish one of us did have some magic
solution that would solve it.
You've heard this suggestion so far this morning.
I'm sure there'll be more coming up.
Instructing the patient to evacuate,
there was a couple mentions of trying
to suck it out.
A rectal catheter, perhaps having the
a technologist can check the initial T two eight image
and see if there's a lot of gas in the rectum
or if it's suspended and make a decision if something
else is needed.
Positioning in your experience help at all?

I'm sorry, say that again? Positioning of the patient.
Yes. This question versus Positioning.
Could you lie the patient on their stomach
and have the gas go somewhere else?
I've not tried that. Second,
Just curious. So let's just go quickly
down the line here.
What do you say?

I think if they know ahead of time that the
they need to evacuate prior, then
that helps them plan a little bit.
So we have our when they called a schedule,
we make certain that they're aware of that.
I really don't have any much more stuff than the fact
that the I do have a prep the night before the morning of that was,
and I use coil as well.
Yep. And that helps.

Yeah. It was an odd discussion to have at lunch,
but we often use endorectal coils with flu art,
and of course do not have a problem in that respect.
And Jeff suggested what they do made a lot
of sense, which was,

Well, thank you, Peter.
I was actually gonna talk about this in another talk,
but by the way,
this isn't the problem in general,
if you're using an endo rectal coil,
when you don't use an endo rectal coil, right?
It's a problem. So one solution is if you see it just,
you see air in the rectum,
take the patient out and stick in an endo rectal coil.
What we do in our practice is we
wanna do these without an endorectal coil.
We talk to the patients beforehand.
We tell them we're gonna try
to do it without an endorectal coil,
but there's a 10 to 20% chance
that it's not gonna come out well.
And we're gonna have to put the endo rectal coil in
and there for the most part.
Okay. With that. The downside to that is
that it takes a lot of time.
You gotta monitor the exam
and determine whether there's air in the rectum.
And the first thing we do, if we see air in the rectum,
we have somebody monitoring this, we're trying
to train our techs to monitor it.
But we have fellows
or generally check this.
If we see air in the rectum, we do
what you already heard about, which we call a ectomy.
We put a Foley catheter, which costs money by the way.
They're not cheap.
But this seems to work the best.
And we first we ask the patient to expel the gas.
We might tell them to try,
does he have to go to the bathroom?
It doesn't work so well. So we try to suck the air out.
I'd say with mixed results,
sometimes it helps a lot.
Most of the time it helps a little.
Sometimes it doesn't help at all.
And so we end up putting the endorectal coil in.
This is not a good solution, honestly.
So there's probably better ways.
So we've talked about, but we haven't tried just looking
for novel ways of doing this.
'cause it really is a problem without an endorectal coil.

Yeah. In the community, people don't want things placed
There. Yeah, don't,
well, we don't want, again,
this is a serious problem if you're not using
an endorectal coil.
We've talked about putting them on their stomach,
but for them to lie there for 45 minutes
or an hour on their stomach, not terrific.
And I don't know that it would make a
difference, but we haven't tried.
We've talked about putting ultrasound gel in their
rectum, Kashas tried that.
We've done it and doesn't really help a whole lot.
We do have the patient, we do do a prep on the patient
now before at home.
They get a fleet sentiment.
Now that seems to have helped,
but it doesn't help when they're there.
We've talked about infusing barium into their
rectum and then like plugging it up
and seeing if the that would help,
but I doubt that's going to really fly
for 45 minutes or something like that.
So, to be honest with you, I don't think any of us, we
we actually have discussed this amongst ourselves,
including at lunch or over the phone and in conferences.
And we don't have a great solution now,
but I bet there's probably some real simple way
of solving this that we haven't thought of.
Thank you. Okay, thank you.

Assessing Lesions in the Transition Zone

On one of the slides you showed in the transitional zone, on the
Transition zone, so Transition zone,
I have to tell you, this is one of these pet peeve things.
When we had our steering committee together,
almost everybody was calling a transitional zone,
like the junction zones, not just me zone in the uterus,
but it's the transition zone.
Okay. So the gentleman you mentioned the first
that the well, you said,
you said you showed the third time
where there is a lesion on the transition
in the transition zone.
Can we go back to that slide?
I just have a specific question on it
because you mentioned that there is a lesion there and
and it was something on the other side. I wanna know how

Yeah, there was something on the other side.
I didn't Absolutely. Okay.
I didn't wanna make Want to every
Lesion. Well, I want know,
how did you decide,
one side or the other?
Not in the not in the peripheral zone.
No, no, in reality.
Again, I would've done a pi rads assessment on that other thing on the other side,
but I left it out in this case
because it's one of those things I wasn't sure how
to do it, to be honest with you.
And I didn't want to confuse you.
I just wanted to go over some things
that are relatively straightforward. Okay. So, yeah,
Absolutely. Well, I just wanted to
know how you decided and,
and so we can learn from you how you decide,
one or not the other. Okay.
So that, the thing on the other side was very well
defined, exactly was discrete
and round those things in the transition zone
tend not to be cancers.
Whereas the thing that we described in the transition zone
and did a pi eds assessment on was amorphous not, didn't have nice circumscribed margins.
And that's what cancers look like in the transition zone.
Thank you.

PSA Density, Volume Studies, and PI-RADS Cancer Percentages

Hi, my name's Scott Slavi. I'm a urologist.
I'd like to make one comment and then have one question.
The comment is, we like to have PSA densities
and do volume studies, and that really helps.
So if it's a really big prostate
and the you check the PSA with the density,
it really eliminates that as a as an issue.
The question is what,
and I you may have mentioned it earlier,
what's the percentage of cancers
for PI RAD five pi RAD four? Do we have those numbers?
Right? So again, the question is
what are the percentages assigned
to those PI rads one through five?
And as I mentioned in my talk, we don't know
because it hasn't been standardized yet.
So people have been using different scoring systems.
So now that we have it standardized,
and hopefully everybody will start using it we'll at some
point, and the not too distant future be able to tell you,
we hope that a RADS four means that there's a
80 whatever, a 75% chance
of cancer of clinically significant cancer.
And what percentage of patients have
that is gonna depend on the patient
population you're studying.
And we do prostate MR
for a whole variety of different reasons.
So it would be silly to think
that patients in whom we're using it
for surveillance are gonna have the same percentage
as somebody in who had a positive biopsy.
And we're doing it for localization, for example.

Yeah. Our urologists actually published a series
of targeted biopsies for both active surveillance
and for prior negative biopsies,
and found a difference in the positivity rate for
the what we call the assessment category, fours
and fives, just not the new pi rads categories.
But because those populations were different,
they got different yields.
And we did an informal study.
It hasn't been published of
the PI Rads version two in our setting.
The sensitivity was actually quite good at
in the mid eighties, 85%.
The down observation was
that there was inter reader observer variability that
is pretty substantial. Yeah.
And we can expect that because it's new
and it's gonna take time for us
to diminish the integrative variability,
but we've had the same experience with birads
that everybody was reading it all over the place
and now it's reasonably reproducible.
So this will take, this will take time,
but we couldn't possibly get there without standardization.
There is a paper coming out from in Europe
where they did a meta-analysis of
looking at all the literature published on using ac using
PY Rads version one or some semblance of it.
And not unexpectedly expectedly,
they were all over the place
because the instructions on how to use
that were really not very good.
And and and,
but clearly it depended on what the indication
for the MRI was.
But we're hoping to have, if we can end up
with a sensitivity
for clinically significant cancer of 80% in that range
or higher and a
and a specificity in that range, boy,
that would be a big step forward compared to what
we've been doing up to now.
Other questions?

PI-RADS Assessment for Seminal Vesicle Lesions

Yes, I have, I think recently what would be a unique case,
and I don't recall if the patient had had a
positive biopsy in the past or not,
but when I was reviewing the MRI,
it appeared the only significant lesion was in
the seminal vesicle.
And I didn't ha, it didn't look like in the prostate
specifically that it moved into a semi vesicle.
That was pretty much all I had.
I'm not sure how I assessed that with py rads. Yeah,
That, that's a good question.
Anybody wanna address that?
We've all had those kinds of cases.
Well, that, that is the seminal vesicles are one of the
location categories.
Beyond that it doesn't count
as either peripheral or transition.
So you can at least say where it is
and say that it looks abnormal,
but we don't have a way to rank the suspicion level.
How large was this lesion?
Off the top of my head, I think it was about
a centimeter, centimeter and a half. Oh, okay. So it's
Fairly good size.
Anybody else wanna comment on it?

Well, it's an unusual situation.
Primary carcinoma in the semi vesicle is diminishingly rare.
So you'd have to be presuming that there's an occult
prostate cancer somewhere that you're not seeing,
which is possible about one
or 2% in our series were sparse aggressive cancers.
So I think our approach would be to biopsy that lesion
under ultrasound guidance
and see whether it was, I'm assuming
that it was a bonafide lesion and not like a stone
or and see if it is cancer
and if it is cancer, then you know, so
You would just obviously not be able
to assess the true PY rads
and then no, just go into
your verbiage and recommend a biopsy.
Yeah. Yeah.
And again, I think what we're gonna see is
PY RADS is not gonna cover every single
situation that occurs.
We've all found things that we look at it
and go, I'm not sure really how this fits in,
but it will cover the vast majority of them.
Okay.

B Values for ADC Map Calculation

Are you all using three B values
to calculate the a DC map or is too sufficient?
The question was for the A DC map, you need
three B values or two sufficient.
So two is sufficient.
I'm actually aware of one article that actually
compared performance with two versus more B values
and showed no difference in diagnostic performance.
I mean, there could be value of doing a larger number
of devalues, for instance, doing
some more sophisticated modeling.
But just for standard
or for instance, just to do
what we've been seeing in this course so far.
And for the pirates approach, I mean, two be values is
fine.
I don't feel
that a third one is necessary, right?
It's just a calculation
and the calculation just plugs in.
Two numbers,
in my sense we have a certain amount
of time in the protocol that I'll devote to diffusion,
and that time we can get total number
of iterations which can be applied across different B values
or averages per B value.
So this is how much time we spent on diffusion.
If I'm gonna get more B values, that'll be less time
or less averages per B value,
including less averages on the highest B value.
And I really want that maximum B value image, have a lot
of averages and be have highest NR and be have high quality.
So on the one hand, even though we have less fewer B
values, there are now more averages per B value in,
not in the prostate, but in other organs.
There's been montecarlo simulations performed
to identify how to best distribute all the acquisitions we
obtained between B values and averages per B value.
And it's been suggested that it could actually be beneficial
to fewer B values with more averages per B value
to have each value v value be more precise.
So there's a trade off there.
If you're saying do more B values,
if this is within a fixed total scan time,
then you're having less averages per B value.
So we I go with the two B values just right.

ADC Value Standardization and Thresholds

What another important point that we
haven't discussed at length is, the reason why, so you heard people talk about a
DC values and how we use those in our practices,
but we can't tell you what a DC value
to use in your practice
because it is dependent on
the B values that you use.
It's dependent on the scanner you use.
It's just not reproducible
across institutions at this point in time
and across scanners.
So in the pi rad steering committee,
when we ask people what a DC value they're using to sort of,
as a threshold for
clinically significant cancer versus not, we got a range
of seven 50 to 900 and,
and we're using actually a thousand now
because we wanna sort of play a safe in a sense

Yeah. To keep as your maximum view as the B value goes up,
the a DC value in the same ROI will decrease.
So if one center's using a maximum value of
a thousand and other centers directly acquiring a 1400
and using that to compute the a DC map,
then all the a DC values will be decreased.
So you'll need a lower threshold.
Right? So the point is
that I think PI rads didn't standardize everything,
and hopefully there'll be more research over time
and over time everybody will use the same devalues
and our A DC values will be comparable and we can and,
and we can start talking to each other.
Yeah. But right now it's still very difficult.

I Mean, I think that's tricky. Some
vendors, yeah.
Systems will struggle more at very high B value.
So cer or even with at a single vendor, certain platform.
So some,
and depending on your hardware,
you might have a more difficult time getting
a certain B value at somebody else
that their center might Yeah.
Be able to do more readily. So it's not Daniel
You wanted to say something.
Oh, and you get a cleaner a DC if you
if your minimum B value is at least a hundred,
because you'll have less of the perfusion component,
on the other hand, you'll get a lower calculated A DC.
So if you're trying to compare versus those ADCs
with a B zero, you'll have to
to take into account that offset.
Yeah, I'd say most people have just taken
what the vendors gave us and so on.
Our scan is a B 5,400 and 800.
And frankly it's as good as anything.
We don't know enough
to say we want to change that
or screw around with it. Go ahead.

Endorectal Coil Usage in Prostate MRI Protocols

So this question is Dr.
Rosencrantz, your images were beautiful in your
DDWI talk.
They looked like they were all taken without
an endorectal coil. Is that pretty much,

That was correct.
They were all our protocol is three T without an endo rectal coil. That's
Correct. Gives me hope. Okay.
Second, second question.
I'll tell you also, there's a clear trend,
for better or worse, 'cause it might be for worse
to not use an endorectal coil for obvious reasons.
And certainly in Europe,
by far people are not using endorectal coils.
I'm gonna show you some stuff
in another talk about sort of
what we're doing amongst ourselves. That's

Actually interesting. There was actually a
an article,
this just came out just a few months ago in JCR that
it was a survey of medical centers in the US
and amongst just their protocol they used
for prostate MRI and it wasn't the largest sample size,
but nonetheless the most commonly reported protocol
by a very small margin.
Currently it was three T without an endorectal coil
followed by a three T with an endorectal coil with one
Point, right. But in academic departments
it was mostly
with an endo recal coil at three T.
Yeah. So that's, go ahead. Sorry.

Post-Prostatectomy MRI for Recurrent Lesions and Fusion Biopsy

The second question is
are you gonna be talking at some point during this
course post prostatectomy MR for identification
of recurrent lesions and then fusion biopsy?
We've done a couple of those, believe it or not,
and we're thrashing.
So any help with that?

I Know. Is anybody gonna talk about that in
their

Yes, the answer is yes.
Great. Thank you.

ADC Cutoff for Malignancy at 3T without Endorectal Coil

Sorry.
If you use a three t system, know in recile call,
which cutoff you have for a DC number to call it malignancy,
is it 800 or eight between eight to 900, or is it what
Daniel, you wanna comment?
We actually, so when we switched from our old protocol
to our current protocol, we had a number
of strata that we use.
So we actually stratified using a K series of
whole mount correlations where it used
to be if it was above 1200, that was completely normal,
and our low cutoff was 600.
So it's below 600. We would consider that highly suspicious.
When we switched to our new protocol, we redid
that analysis and had to adjust,
basically add 200 to each strata.
So we have a we were using a non-standard
assessment category,
but above 1400 is basically normal, 12 to 1400,
low suspicion, a thousand to 1200, intermediate, 800
to 1200 high
and below 800, very high with the sense
that if there's no signal, if it's like a very low a DC
because it's calcium, then you discount that.
But that only works on our scanner platform
with our protocol.
When we look at our GE scanner
that we basically don't use anymore,
we could not apply that.
And even when we look at our 1.5 Tesla scanner,
we're not certain if we can still use those same values.
So I think the important thing is
that you use one protocol ideally on all
of the scanners, if you have more than one
and correlate with outcomes, see
what the pathologist is actually saying
and set your own strata.

Yeah. And we're gonna talk a little bit more about
this tomorrow when we talk about sort
of the business of prostate MRI
and how important it is to do these correlation type things

Use of Glucagon in 3T Prostate MRI Protocols

Regarding glucagon, everybody, I mean, three T system,
do you still glucagon?
I mean, they're pretty fast glucagon you use, I mean,
routinely only every patient get glucagon when you,

So glucagon is a tricky issue.
I think we're all using glucagon.
I don't know that any of us knows that it actually helps.
And 'cause I don't know
that anybody's actually done that experiment.
We think it helps, but
I don't use it. Yeah.
And I have tried
to use it several times and,
and we got very good anti peristalsis
of the small bowel, but
I don't think it works that well in the rectum,
So. Right. And the
studies in Canada
and Europe, they're using Buscopan, which is a
better agent, frankly,
that we just don't have in the us.
So I would say, in the Pi rads document,
all we did is say that some people recommend
or people use this, but we don't say you have to use it
or even suggest you have to use it.
I think we don't have data
to support things one way or the other at this point.
Good Study to do. Did you do the study yet?
But we Did everything else,
No. We had used it
at one point, similar to other comments,
we'd stop using it.
I don't think we've observed any
appreciable difference or drop off in poverty.

Yeah, I think the other problem with glucagon is that,
when you no matter how you give it,
it's just unreliable.
You don't know when it's gonna kick in
and how long it's gonna last
unless you give it intravenously in a pretty,
as a two milligrams, it
will actually start working pretty much immediately
and last a fairly long time.
But a lot of times we are doing the scan
by the time they get the patient in
and everything, if you give it,
IM the glucagon's already one off,
or they do the key studies before the glucagon is effective.
So this has to be studied systematically.
And I can't, I don't feel strongly one way
or the other personally at this point.
Okay. No more questions. We, I've been instructed. Done.