A panel discussion on the subject of MRI Contrast Agents, moderated by Lawrence R. Muroff, MD, FACR.
Lawrence R. Muroff
, MD Clinical Professor of Radiology University of Florida, and
University of South Florida Colleges of Medicine Tampa, Florida
, MD Director, Residency Programs Clinical Associate Professor
University of Missouri, Kansas City St. Luke's Hospital Kansas
, MD Neurovascular Radiologist HealthSouth Doctor's Hospital Coral
Steven B. Halls
, MD, FRCPC Director, Department of Oncologic Imaging Cross Cancer
Institute Edmonton, Alberta, Canada
, MD Professor of Radiology Director of Clinical and Educational MR
University of Pittsburgh Medical Center Pittsburgh,
Gerald M. Pohost
, MD Professor of Cardiovascular Medicine University of Alabama at
Birmingham Birmingham, Alabama
Val M. Runge
, MD Rosenbaum Professor of Diagnostic Radiology University of
Kentucky Lexington, Kentucky
Lawrence N. Tanenbaum
, MD Assistant Professor, Neuroscience Seton Hall University
Section Chief--Neuroradiology, MRI & CT
New Jersey Neuroscience Institute J.F.K. Medical Center--Edison
Imaging Associates Edison, New Jersey
Dr. Kanal talked about the gadolinium-based contrast agents being
similar in some ways and dissimilar in others. Do these factors
affect how you look at the agents?
Many of us are in situations where the contrast agents may not be
selected by the radiologist per se but I think it's valuable to
have the information as a radiologist to make the distinction. If
you have the choice of what contrast agents to use, safety, which
seems to be quite similar for all agents, is extremely important.
As Dr. Kanal pointed out and Dr. Tanenbaum mentioned in his
presentation, radiologists should be educated regarding things like
injection rates and viscosity and osmolality.
You bring out a point that some of us don't have the choice of
contrast agents. We noted when we polled our the registrants at
Educational Symposia, Inc., meetings 3 years ago that about 70% of
physicians were able to designate their agents. Today, less than
half of them can. What's the experience on the panel? Can all of
you pick the agents you want to use?
My experience is similar to that of Dr. Delgado in that it is a
purchasing manager who will choose a vendor. Unless you raise
strong objections and make your voice heard--and you can only do
that once or twice on a couple issues per year--you are going to
get what they decide, usually based on price.
I think we all have a similar experience that when it comes time
to evaluate the vendors, you'll get a box of free samples from each
and you try them all out dutifully and, honestly, I don't think you
notice any significant difference in the clinical context. That
being said, there are some things that do sway you one way or the
other, and I think that's been brought out in the
Dr. DiPrima, how about you? How are things done in your place? Do
you get to pick what you want to use?
I think that it is less and less common that the radiologist is
solely able to choose their own contrast agent, but your voice is
heard. If you argue and there is not a large price difference, you
are often able to at least sway the decision. At my institution, we
tend to use a lot of higher-dose administration, particularly on
older patients with cancer, and so the osmolality is a powerful
argument. Additionally, administrators don't want to be caught on
the wrong side of a lawsuit any more than we do. So with some
success we've been able to help guide the decisions, as long as
there is not a large difference in cost. It's important to
remember, however, that we don't buy just one contrast agent.
Computed tomography (CT) considerations, iodinated contrast, and so
forth also enter into the decision-making process.
Do the rest of the panelists choose your own agents? Dr. Tanenbaum,
do you pick what you use at your hospital?
We are fortunate to actually make the choice of contrast agent at
our institution, although sometimes it's a bit of a negotiation,
and price is always a consideration. I think there are a number of
compelling reasons, especially with the advanced applications, to
favor one agent over the other. At my institution, we are fortunate
enough to have the ability to do that: to choose or force a change
at the institutions with which we are affiliated.
Dr. Pohost brought up the concern about extravasation and viscosity
in terms of cardiovascular applications. You do a lot of
power-injecting. Is this part of your decision-making process?
Frankly, the differentiating issues--ionicity and osmolality--are
critical issues in most of the advanced applications. Aside from
the times when we actually put a butterfly in and hand inject, a
significant percentage of our contrast administrations are power
injected at higher and higher rates. Cardiac cases are done at
least 5 mL/sec. And you heard Dr. Pohost advocate warming the agent
and going as fast as you can, limited only by the pressure.
Localized pain will kill your study. So it's critically important
that you minimize all those factors and, frankly, those are the
critical considerations. That is why we are so comfortable with
ProHance in this clinical setting.
Dr. Runge, do you have any other considerations in selecting an
I'd like to address the issue of extravasation. We recently did a
mouse study looking at extravasation. We took the four
gadolinium-based contrast agents approved in the United States and
did a rather large animal study. It was a blinded study from two
perspectives. The individual giving the animal the extravasated
dose was blinded to the agent used and also, of course, the
pathologist that graded the response was blinded to the agent used.
In this study, to no surprise, when we looked at necrosis and the
other histologic changes that occurred with the gadolinium
chelates, (and of course we were looking at extravasation, not
intravenous injection) the results were statistically significant.
The amount of necrosis and other histological changes correlated
closely with osmolality. So the high-osmolar agent, Magnevist,
caused the greatest damage, including necrosis and other histologic
changes. The two low-osmolar agents, ProHance and Omniscan, caused
the least. The newest agent on the market, which is intermediate in
osmolality, OptiMARK, caused an intermediate response. These three
groups were statistically differentiable.
I think one of the things that many radiologists have not looked
at very intensely is specifically what amount of damage is caused
in extravasation with the gadolinium chelates. In this study, we
used saline as a control and we did a comparison with one of the
old ionic X-ray agents. It turns out that the response to
Magnevist, in terms of extravasation and damage, was not
statistically differentiable from the old ionic X-ray agent. The
only reason I think that we haven't been exposed to this problem to
a greater extent in the last 10 years is that only recently have we
been giving large doses and using power injectors and high
In this study, the reaction with ProHance was not statistically
different from saline. But again, that is based on the osmolality
of the agents. I think that also reflects another thing that is in
the package insert. In many of the countries, including Japan, the
United States, and some European countries, the Magnevist package
insert states and adds precautions about extravasation due to cases
of necrosis. In the U.S. package insert, it says that one has to be
careful in regard to this adverse reaction. Every time you give an
intravenous injection, you need to make sure you look at the site.
In fact, the package insert for Magnevist actually now warns about
things like limb amputation that might occur because of a terrible
reaction with extravasation. Therefore, I think with the high doses
that we're now giving and the use of power injectors, we need to be
very careful. I would advocate that anyone in that instance be
looking at the low-osmolar agents.
: Dr. DiPrima, what's the mix in your practice of MR? How does it
shake out in terms of neurologic, body, and musculoskeletal
I think that our experience probably parallels most surveys I've
seen in the literature. We do probably about 60% neuro-imaging,
including spinal images. The rest is predominately musculoskeletal,
with an increasing but small number of body applications, including
specifically: breast, which does require contrast; liver, which
requires contrast; and MRCP hydrography of the pancreas.
When you say a small percentage of others including breast, how
many breast patients are you imaging in a month, let's say?
I would estimate about 10 a month on average.
And what is your overall percentage of MR contrast use in your
practice, per se, is it in that 20% ballpark?
No, it's substantially higher. We have a high mix of oncology
patients so we're sitting somewhere in the 50% to 55% range for
I'd actually be interested in polling the panel to find that out as
well. Everyone's quoting 20% to 30%, but we have not been able to
get below 40%. I'm interested in hearing what might be the cause of
that, and to see if that's actually representative of this panel as
I can tell you that among the people we've polled, the difference
seems to be neuro-imaging. I also think that entrepreneurial
centers tend to use contrast slightly less than other sites. In
addition, I think that we're going to see cardiovascular and
peripheral vascular imaging increasing the percentage of contrast
use. That's on the other side of things, but maybe we can start and
take the whole panel.
I'm doing mostly body imaging, so I'm not involved directly with
the neurologic applications. On the body application side, I tend
to use contrast frequently, but I think it's being used more and
more particularly with some of the applications that Dr. Tanenbaum
spoke about, peripheral vascular studies and cardiac MR.
Dr. Tanenbaum, you were talking about the rate of use at your
practice being 20%.
Contrast use at our practice is about 20%, and, frankly, that
reflects a predominantly neurology bias. We are probably 70% to 75%
neurology, head and spine. The 20% number reflects a strict
criteria for the utilization of contrast that coincides with a
traditional list of indications. We don't give contrast for
headache, we don't give it for new-onset seizures, or for seizures,
or just because we are very worried about a particular case. We use
a strict list of indications, and that puts us at very low
utilization. We use it rarely in the spine, for tumor, and for only
10% of postoperative cases. We perform a decent amount of body
work, and with high-performance gradient systems, contrast became
routine in circumstances where before it was never used. Of course
we use contrast in the advanced applications, but our numbers are
in the low end of the range because we use these agents under
strict circumstances, and avoid use in the more "fuzzy"
It was interesting that Dr. Muroff mentioned that
entrepreneurial centers use contrast less. He probably means,
euphemistically, centers that run without a radiologist on site. On
the other hand, there are entrepreneurial centers that will use
contrast almost 100% of the time in neurology areas because there
is a CPT code directly assignable to the post-contrast study, so it
tends to shift the utilization in those circumstances as well.
That's an interesting point, the way you segmented the
entrepreneurial centers, and I think that's very important. What
are your views about having a physician on site for contrast and
injection? Again, I can tell you in our polling, 75% or more of the
people believe very strongly that there should be a physician on
site. I know that there is a very small, but very vocal minority of
people who believe that ought not to be the case. Dr. Kanal, why
don't you kick off this part of the discussion?
Let me start it out by saying that this is, as you mentioned, quite
controversial. It seems to me that if we're going to be
administering drugs, we will, of course, be responsible for the
reactions, but I don't want to be on the record saying I believe
there has to be a physician on site. It's really my opinion that if
we're going to be administering these agents, then someone who is
in a position professionally to be able to identify, recognize,
stabilize, and appropriately respond to any emergent reaction or
adverse events that might occur should be on site, trained
appropriately, and be able to demonstrate and document that they
followed the protocols that this site has set. If I have a
physician assistant there, if I have a nurse practitioner there, if
I have anyone who is adequately trained in monitoring for adverse
events, has a minimum of training in basic life support and CPR,
knows exactly what the policy is for contacting professional
assistance should it be required, and having them available in a
timely fashion, I think that would probably be sufficient. The
exact definition of "a timely fashion" is going to be
From a financial point of view, it's a considerable burden to
have a licensed MD or DO on site for every MR center in the
country. On the other hand, financial considerations may not be the
ones that will attract the jury's attentions, should there be, God
forbid, a life-threatening reaction, or a severe adverse reaction
that was not adequately handled in a timely fashion. So most sites
are essentially taking a gamble. They're saying that they have
trained personnel and if it's a severe life-threatening reaction,
just have them dial 911. As long as the patient prospectively
understands that his or her chances of having a severe reaction are
extremely small, but if one occurs there is no one on site to cover
it: if the patient agrees to that, then perhaps you have a
relationship that can be documented prospectively as being
reasonable and understood. If the patient is not aware that these
drugs have a very small rate of life-threatening adverse events and
there is no one there to handle it, I'm not quite sure what defense
the site is going to be able to offer.
One of the purposes of this conference is to talk about myth versus
reality. We're talking about adverse events, and I think one of the
things that comes up in discussion often is nausea, as one of the
adverse events that people see with all MR contrast agents. I was
wondering if some of the panel can talk about their experience with
one agent versus another to see if there are any differences in
either number or types of adverse events.
Anecdotally, the two agents I've used are Magnevist and, more
recently, ProHance. With regard to nausea, for some reason, I don't
think our incidence is as high as some of the studies that were
quoted earlier in the discussion. I wonder sometimes if that's an
issue of prepping the patient properly. We fully tell the patients
that they will feel some warmth in their arm and probably some
nausea. We also tell them that it is unlikely that they will vomit,
and for whatever reason, we just don't see it that often.
I would say that probably most of us rely heavily on our
technologists who feed us this kind of information. They are the
ones who have direct contact with the patients. We are very
fortunate at our institution to have some wonderful technologists,
and the information given to me is that there are very few
incidences of small adverse reactions, those being hives usually
and nausea to some degree. I think most of the physicians here feel
pretty comfortable giving ProHance. At our institution we have used
Magnevist, ProHance, and Omniscan. Our experience demonstrates
there are no differences in severity and/or frequency of adverse
events between any of the gadolinium chelates.
To the point that, if a physician is not on site, a simple phone
call alerting a physician that contrast is being administered is
sufficient, and the likelihood of anything happening is very, very
Since we're not in a clinical trial setting, we're really only
aware of some of the reactions. I think the technologists, to a
certain extent, are really only going to comment on the cases that
are true vomiting or near emesis. Mild nausea is going to escape
our notice, unless we go out of our way to question the patient. I
think that is why we get the impression that, colloquially, the
drug is like water; we don't see many adverse events.
I made sort of a flip comment that we've seen more reactions
with hand injections than we have with power injections, but,
frankly, that is a true statement. We've had maybe one or two
reactions in thousands of power injections, yet I can recall cases
in which I've gone in and hand injected somebody and seen reactions
with all the agents we've used. So, I think it is a very, very
small number of patients that will actually come to our notice as
having acute adverse events, and that's why the numbers are so low,
and why it seems even lower with power injectors.
I'd like to address that issue of power injection because I think
that it is true for the low osmolar non-ionic agents. We noticed
early on, when just Magnevist was available, that when we gave a
bolus our nausea rate went way up. And so I would caution people in
this day when we commonly use power injectors, that you need to go
with a low osmolar agent.
Let me just say this for the record, that my experience with power
injections is 100% with ProHance, so my comments reflect only that
I would like to add, if I may, I think your experience is very
real, and I'd like to present another potential explanation. We
have the same experience that you have; with power injectors we
find less of a "reaction rate." I don't believe that this is real.
I believe that when we hand injected, you're standing there, by
definition, and the patient will say something more readily than
when it's a remote power injector and there's no one in the room.
You just pick up less reactions.
There are differences in levels of nausea. The ones that we find
out about are the ones that we consider disruptive nausea. The
physicians in these retrospective surveys, I think, will tend to
pick up what we refer to as disruptive adverse events, or the
adverse events that tend to come to your attention whether you like
it or not. A lot of the patients, for all we know, are having
nausea, and, if it was a small amount, until you do a prospective
study and ask them, they just wait for it to pass and don't say
anything about it. Therefore, it never gets reported to anyone as
an adverse event if you are not disrupted from your typical
Dr. Pohost, in your cardiac patients you said you've used both
Magnevist and ProHance. Have you noted any differences in nausea
rates with those patients?
I haven't noticed any difference between the two, but again we have
noticed that when you deliver the agent to the patient by power
injection, the rate seems to decrease. The problem with the
perfusion studies is that nausea may be caused by the stress agent.
We know that those two agents produce nausea in and of themselves
to a much greater extent than contrast agents.
I would just like to comment that we actually do have someone
standing by the patient when we power inject. It's not quite the
same experience as injecting someone anticubitally with them
sitting next to you out of the magnet. The magnet is generally
chirping and the patient is in the machine and there's a lot of
things going on, but we are standing by doing the best job we can
to monitor the site. I'm not sure, however, how effective we are at
monitoring for extravasation with the patient in the bore. We make
a reasonable attempt to be there at the patient's side when we
I just want to make a comment with regard to the cardiac patient.
We require that a physician be there anytime we place a cardiac
patient in the magnet, and that's for obvious reasons. So there is
a physician there all the time. It's not a law but it's our own
internal laboratory regulation. We have had some patients who have
had pulmonary edema episodes: things that you wouldn't want to have
happen without a physician around.
We have had a very good experience. We switched from Magnevist to
ProHance in 1993, and since then we have done over 5,000
injections. Our technologists log every reaction in a log book and
we recently ran through that log book and discovered our rate was
1.3% for reactions of some kind. That's lower than some of the
numbers we've heard in the presentations. Of that, 80% were either
nausea or vomiting, and smaller percentages were minor things like
headache and unsteadiness.
Do you sense there is a difference from one agent to another, or
have you only had experience with ProHance? How does your
experience relate from one agent to another?
I'm just going to stick to saying I think it was equivalent to what
it was before we switched to ProHance. But in my mind, it
illustrates that the very act of asking the patient if they're
having a reaction or standing by them influences the outcome and
our numbers reflect where we don't ask. We just make note of what
the patients tell us and it's been a good experience.
In the 1970s, Dr. Anthony Lalli published a landmark study on
contrast media and adverse reactions. He suggested that it was
possible to talk a patient into a serious contrast reaction. I
really believe he is right. Depending on the type of questions you
ask and the approach to the patients, you can have a significant
impact on adverse event rates.
Dr. Halls, looking at the data here today, you note a 1.3% reaction
rate in your retrospective review. Since you did not question the
patients, that means these are the reactions that were
spontaneously reported to you. And then the 5,280 Magnevist
patients we studied had a 1.3% adverse reaction rate. It's amazing
that your study and ours were that identical, considering that
neither of us knew the other one was doing it at different
Our patients are cancer patients, and a lot of them come back for
repeat imaging, and if you don't pick up reactions the first time,
the patient will often mention to the technologist the next time
that they had a delayed onset of hives. This way we still pick up
ones we would normally miss.
When the purchase choice comes up, when you've got two agents
that are exactly the same price, it is very nice to be able to say,
we have had 5 or 10 years of experience with this particular agent.
Anecdotally, about 6 months ago, one of our technologists reported
that we had four reactions in 1 month. We've never had that before;
it was always one or two or three or zero. So we said bring the
numbers. We put them in the computer, did a chi-squared test on it,
and, of course, it wasn't statistically significant. Then I wagered
to them that during the next month there would be less than four
reactions. I guaranteed it, and, of course, it was less than four.
It just proves that adverse events are random and several reactions
in a short period of time are generally not significant.
: Dr. Pohost, talking about safety: your cardiac patients are a
special subset of patients that have so many superimposed problems
going on. You talked about your concerns with injections rates.
What about your concerns--or do you have any concerns--about
osmolality and osmotic load in these patients, one agent versus
We have great concerns because the possibility of these heart
failure patients that we're studying for perfusion may have
episodes of pulmonary edema or heart failure; exacerbations of
heart failure are a real potential threat. The same thing is true
of the iodinated contrast agents. We've had a few patients develop
pulmonary edema in the magnet and you have to treat them rapidly.
So it is something that is realistic and something that should go
through your mind when you are trying to decide between a high- or
Dr. Kanal talked about whether or not there is supervision
available on-site. Are there any risk factors that alert you to the
fact that a particular patient might be more or less prone to have
This is a very good question and often is asked nationally at
meetings. I think you have to be very careful. Patients with asthma
have a higher susceptibility to having an adverse reaction, as do
patients with allergies to multiple medications or multiple
different substances. If you look at the package inserts in
general, those two conditions are mentioned. The third type of
patient with increased susceptibility is one that is often not
well-appreciated in the United States: those with a previous
reaction to an iodinated agent. Such a prior reaction does put you
in a higher risk category. So asthma, multiple allergies, and a
prior allergic reaction to iodinated contrast media are all
If I may add to that also, the allergy to a prior iodinated
contrast administration just about doubles your incidence to about
6%. But an allergy to a prior administration of a gadolinium agent
increases to over 20% the potential to react to a future one. In
fact, at UPMC, our standard of care and policy is that if anyone
has had an allergic reaction to a prior MR contrast
administration--and we've used them all now so it doesn't make a
difference to which one--then, when they come back, they receive a
different agent, not the one they initially received.
There are stories about patients who have had adverse reactions
to Drug X, and every single time they got Drug X, there was a
problem. Then we gave them Drug Y with no problems whatsoever and
we stuck with Drug Y since then. You can plug in anything you want
for X and Y, because they have had a problem with this one, not
that one, or that one and not this one. And it is that way across
I'd just like to ask a question. With MR contrast agents, when
would you pretreat? What criteria would you use?
Does anybody on the panel pre-treat patients? Let's start with that
question. Do you pre-treat any of your gadolinium patients whom you
are concerned may be predisposed to a reaction?
The only pre-treatment we'll do is if someone comes in with a
history of a previous severe reaction, and the severe reaction is
not to iodine. We've had patients anaphylactoid to iodine that I
have not pre-treated when I give them any of the gadolinium-based
agents. However, if they have had a severe, potentially
life-threatening reaction to a gadolinium-based agent, and, for
whatever reason we feel it's still indicated to give them another
such agent this time, we'll switch to a different one and we will
pretreat. But I will make an observation. I don't know if this is
universally reflective of the reality of MR care in this country,
but we do have our patients monitored on a pulse oximeter. We
monitor everyone who is at increased risk for a potential
significant reaction, even if it's just due to an iodinated
reaction in the past. We monitor all patients who have had previous
allergies, a history of allergy
or allergic respiratory phenomena, asthma, or an iodinated
Let's deal with other issues of myth versus reality. Are there
patients who ought not to get gadolinium? Dr. Halls, in your
practice are there patients to whom you avoid giving
The answer is no. We use it when it's called for and it would take
a very severe set of circumstances to persuade us not to use
I agree with Dr. Halls. The only indication probably would be, as
Dr. Kanal pointed out, a severe reaction to gadolinium. But with
patients with decreased renal function and cardiac patients, we
don't hesitate at all.
You go to a meeting and the first question and the last is always
what about patients with severe renal failure. How do you deal with
those patient, Dr. DiPrima?
In general, we administer a normal dosage to such patients when
indicated. Our philosophy is not to give contrast unless it is
indicated. But once that decision has been made, even somebody who
is in renal failure gets a standard 0.1 mmol/kg dosage. We don't
decrease our dosage.
I'd like to just clarify one point. It has never been shown with
intravenous injection of the approved doses of gadolinium chelates
that there is any deleterious effect upon the kidney. The only
concern in renal failure would be with ultimate excretion of the
gadolinium chelate and the possible issue of free gadolinium. So,
when we are injecting a gadolinium chelate, we are, of course, not
causing any renal damage.
It should also be pointed out that, as you brought up so nicely the
concept of myth versus reality, the package inserts on all of these
agents have sickle cell as an area that one should be aware of a
potential problem. I hope that anyone who reads this article can
contact me if they have any idea why that has made it into the
package insert of these agents. All four of the companies have been
directed to do so, to my knowledge, by the FDA. It is true that
sickled-cells do orient relative to the magnetic fields. The forces
with which they orient are exquisitely small, to the point where
they are felt to be less than normal physiologic forces on these
cells as they go through the normal capillary flow. Nevertheless,
any potential "potentiation" by these contrast agents makes no
sense to me whatsoever. There are time-varying concerns, and I have
no idea why that it is in the package inserts. We have most
definitely administered these agents to patients with sickle-cell
traits, sickle-cell disease, and have used it diagnostically when
we felt it was clinically indicated without hesitation.
I'd like to go into another area: high-dose uses or altered doses.
Perhaps we can even get into low dose as well. But in high dose,
I'd like to deal with two different issues, the first being the
central nervous system (CNS) indications and then the other being
the off-label indications. First let's deal with CNS. What
percentage of times do you use greater than 0.1 mmol/kg for CNS
patients in your practice other than angiography perhaps?
When we presented the data for our study, we looked at the dose for
that study as well as for
that initial study. We found that in the two studies, approximately
5% of patients received non-standard dose. Of that 5% about 50%
received a low dose and 50% a high dose. So I would guess that not
counting MR angiographic procedures for the body, approximately
2.5% of our patients who received contrast got high dose.
I think our experience will be a little higher than Dr. Kanal's. We
do a lot of imaging before radiosurgery, and metastatic disease has
become the number one indication for radiosurgery these days. You
know that the detection of additional enhancing lesions is
important. So we've become pretty frequent users of high doses of
contrast agents prior to radiosurgery; not so much in a primary
brain neoplasm, but in the scenario of intracranial metastatic
We've dabbled in high-dose usage, but I think the reason that we're
not doing it as much anymore has to do with the fact that we just
keep getting busier and busier. You have to interact with the
technologist to make the decision to give more, and so much of our
practice is driven by protocols now. Especially in neurologic
imaging, we don't do it much anymore. The few times that we do give
extra doses is when we're doing breast MRI and we're localizing and
I'm afraid the scans will have a gadolinium washout.
We use gadolinium for our hepatic studies, some of our pelvic
studies, and enhancement studies of the uterus and pelvis. When
speaking to Dr. Tanenbaum, I told him that we routinely use high
dose--double or even triple initially, although that has come down
to some degree. I'd like to hear his comment regarding the fact
that he is able to do it on standard dose.
How about contrast use for MRA?
My feelings on the appropriate dose for MR angiography are kind of
symptomatic of my feelings any time we get a contrast agent dose
recommendation handed down to us in clinical practice. The
recommendations always come in at the highest dose and I
immediately go to the lowest dose and then try and figure out if
that will work as well. We figured out that with surface coils and
small body parts as we see in the neurology area that 0.05 mmol/kg
works very well, and that's our standard dose. Half dose is our
routine dose for neuroangiography. For single-station body imaging,
we've been using standard dose and, frankly, I think that less than
standard dose would work as well, but we're at standard dose. For
aortic or renal imaging, we use 0.1 mmol/kg and the results are
spectacular. We using very high resolution now and results are
Can that be attributed to your instrument a little bit?
Technically yes. I'm talking about a high-performance magnet and as
opposed to a system with which you want to keep the blood pool
opacified longer, but first pass, first or second generation
high-performance gradients. When we do a run-off and we want to
chase the agent "down to the legs," we will go to double dose. Now,
is that because I know that double dose is better than single dose?
No, and, frankly, I have a lot of experience at single dose doing
run-offs. How do I get that? Well we practice with our
technologists. We train our technologists by chasing renal studies
down to the ankles. So I know how good single-dose studies look.
But because my patient is there for that purpose, I want to give it
the best shot and therefore we're using double dose. I understand
that many institutions are using triple dose. It's nice to know you
have that safety profile that allows you to do that, but until I
prove to myself that triple is genuinely different and better,
we're staying at double. Frankly, there are some issues of venous
opacification that may be better at double dose than they are at
triple dose. That is our experience after 5 years of doing a decent
volume of cases.
I'd like to suggest that Dr. DiPrima's comment is right on the
money. It is exquisitely dependent on the hardware and software
capabilities of your imaging system. If your system can generate
the images it requires for the spatial and temporal resolution that
you're asking for during first pass of a single-dose study, then
there is no additional benefit--none--of double dose. Again, let me
say that if you have the caveat that you are able in the temporal
resolution of the first pass of the volume given in single dose,
then someone would need to explain what the potential benefit of
increasing dose would be. As you increase dose, you're not changing
the relaxivity of the agent. It's still the same as it had been
before. So the impact on the blood is going to be the same. The
changes are going to occur when you have a somewhat slower system
and the study takes longer to do, so if you just waited for first
pass in the space of 7 or 8 seconds, it's already bolusing through
and now you're already getting the tail end of that bolus. It's
spreading out. Your relaxivity is not as great. Your T1 times are
starting to lengthen out. Now you need to give a double dose just
to physically give you a bigger bolus to be imaging during passage
of a longer bolus.
What the industry will be heading toward is equipment that is
faster that can get better temporal resolution at the same time as
getting the same or better spatial resolution. As we head in that
direction, there will be much more of a drive for the
pharmaceutical firms that are manufacturing these agents to provide
new agents entirely--agents that have higher relaxivities, that'll
allow us even more imaging. It is especially true for perfusion
imaging today that we accept what they have given us. It's far from
ideal. We would far prefer to have an agent today that would be an
even higher R1 and especially R2 value than what is available
What about the whole concept when we're dealing with myth versus
reality of lower field strengths versus higher field strengths and
the impact that field strength has on contrast. Dr. Kanal, can you
sort the myth from the reality?
I'm afraid this does have reality there. It depends on the
application. For example, if we're talking about perfusion-weighted
imaging, then these are essentially used as susceptibility agents.
And at that point, the amount of signal lost that you'll see on the
image at a given imaging sequence is very definitely going to be
dependent upon the field strength. The higher the field strength,
the more sensitive you will be to a first pass if you're doing
susceptibility imaging. So, for perfusion imaging, I would
recommend going to the higher field strengths, and by higher, as
high as your system will allow you to go. A 3 Tesla would be a lot
nicer than a 1.5, for example.
For some of the other sequences, the physics being what they
are, the relaxivities are not constant. They change as you change
field strength, and what ends up happening is that the T1
shortening is not the same at a 0.5 Tesla as it is at a 3.0 Tesla.
But in reality, what we end up suffering with more than anything
else is a signal-to-noise trade-off. So as we're going to higher
field strength, the bottom line is we are going to see more
signal-to-noise overall in the image, and therefore the contrast
enhancement should be that much more evident.
Earlier we talked about using lower than the 0.1 mmol/kg in
selected patients and about pituitary microadenomas and acoustic
schwan-omas. Are other people on the panel using less than 0.1
mmol/kg and, if so, are you using it for different or other than
those two indications?
We use the standard dose.
We do indication-related dosing across the board, which means we
agree with everything that's been said to this point. We do use
half dose if we are following a known meningioma and it works very
well. Even perfusion imaging with half dose with meningiomas can be
But Dr. Tanenbaum, I know you work with both 1.5 tesla units and
open units. Do you alter your dosing based on the field strength?
Will you use less than 0.1 mmol/kg if you're in an open magnet
I'm glad you asked that question, because I wanted to add something
to what Dr. Kanal said about low field. We have no problem creating
signal-to-noise at low field. It just takes longer and we do it
usually at the cost of lower spatial resolution, although not
markedly lower. We really do think you can see the difference in
the relaxivity in the contrast agents at 0.2 mmol/kg when compared
to what we're used to in most of our practice at 1.5 Tesla.
We do everything we can to maximize enhancement. We'll use the
various forms of fat suppression that are now available. We'll use
magnetization transfer contrast, or background suppression to
maximize contrast. But all of that said, with those devices at our
disposal, we don't use half dose contrast at low field. We stay at
the standard dose and, for the traditional circumstances, we might
go to a high dose. Pituitary imaging is one of those circumstances
where you're dealing with a dynamic effect at high field so you
don't want the lesion to enhance that much, so you want to use a
half dose. At low field, it's more of a steady state exam, and,
frankly, I don't know that you really get an enormous amount of
benefit going to half dose. So it's full dose and up at low field
and the full range at high field for us.
You mentioned magnetization transfer, and I know again we talked
about polling that has taken place over the last 3 years. We've
asked our registrants about their use of magnetization transfer.
Less than 50% are using magnetization transfer in conjunction with
single dose or standard dose contrast. Do you use it routinely in
That's a complicated question to answer. We do use it selectively.
Our rule-out metastatic disease surveys will have a single plane
done with magnetization transfer. Suppressing the background does
improve the conspicuity of your contrast enhancement. It's not a
triple dose worth, but it is more than you would get without
magnetization transfer. There are some caveats. People have
reported punctate lesions, punctate bright spots on magnetization
transfer scans, which have been referred to as pseudo-lesions in
the literature, that you'll see in the pre-contrast scan. We
double-check against that in theory by not using magnetization
transfer for the alternate planes, although, in 13 years, I have
never seen a "pseudo-lesion" that I couldn't document with a
different technique on another plane.
At low field, however, it's a very nice device. I do a 3-D
magnetization transfer, gradient echo as my sole post-contrast
sequence and display that in 3-D and often oblique planes. That is
a very nice way to maximize enhancement at low field and still keep
your time efficiency in hand. If your degrading echo is not
adequate for contrast enhancement, even on high field, add a
magnetization transfer pulse and it's a very viable technique for
Are any of the panel members using magnetization transfer with
single-dose contrast on a routine basis, as opposed to not using
magnetization transfer pulse with your contrast?
We use it as part of our brain metastases protocol. One of our
three post-gadolinium sequences has magnetization transfer. I
wouldn't use it on all three because you'll lose the gray
matter/white matter differentiation, which is nice, but it's
helpful to use it.
We use at least
one sequence routinely. When you administer contrast, your intent
is to detect abnormal contrast enhancement. Anything I can do to
increase my sensitivity, I'm going to go ahead and do. I don't
think the time penalty is substantial at all. For one sequence,
typically our axial sequence, it's just not an issue with time.
I'd just like to point out that the potential benefit of
magnetization transfer is applicable regardless of the dose that is
applied. If you're doing half dose, standard dose, double, triple,
it doesn't make a difference what dose you're applying; the
background suppression is applicable across the board and it should
serve to increase your conspicuity regardless of which dose you
Bill Bradley made an excellent point at the last meeting I heard
him lecture. He said he doesn't use it because in a very
competitive, private practice environment, the pictures are not as
good with a magnetization transfer pulse, and, basically, no matter
how smart we can be, we're ultimately judged by our pictures in a
clinical, competitive setting. That is a very interesting argument
and may delineate the differences between private practice and
I would say that I agree with that. The pictures are ugly, the
contrast is suppressed, and you lose a little signal-to-noise. You
can, on rare occasion, make a lesion harder to see. But when
sensitivity is the critical issue, and you know that's the critical
issue, I can make a pretty coronal image with a T1 FLAIR sequence
if I like. On the other hand, when I'm following up my
neuro-oncology cases, and they really want to see the morphology
and the contrast between normal and abnormal tissue, we don't use
magnetization transfer or fast spin echo techniques, which have a
lot of magnetization transfer. We'll use some kind of really
spectacular-looking T1 FLAIR, and that's the marketing issue that
clinical practices have to consider as well.
Again, this is something that I see less in academic practices and
far more in clinical practices, that is the need or the desire to
market your images to your competitive groups of physicians so that
they send patients to you. Are there any other differences between
academics and community practice in terms of contrast use that we
I would say that, responding to what people have said in terms of
their percentage of contrast use, our percentage is more than 50%
and that probably reflects the fact that we're getting a fairly
sick patient population in an academic center.
Dr. Halls, you're our lone representative from Canada. Is there
anything in terms of practice patterns or reimbursement issues that
impact contrast use in Canada differently than the United
In the private practice setting, it's still fee-for-service and
sort of government-funded and that has some minor influence on how
you plan your examinations. It may influence, in some cases,
whether you do a pre-contrast examination and then bring the
patient back later, or do it all in one setting. In the
contract-for-service setting where I work at a cancer hospital,
there's no influence at all. You want to get the patient in and out
as quickly as possible and do whatever you need to do and you don't
worry about it. There are a few private practices that are purely
private practice settings where the patient has to pay and
government doesn't pay, but even then there is hardly any influence
on the cost of the procedure, actually.
Our time is drawing to an end. I just want to thank everybody for
your participation. I think it's been a very productive and
informative session, and I hope all of you enjoyed it as much as I
enjoyed moderating and participating with you. Thank you.