Recent Improvements in the Sonographic Detection of Ovarian Cancer

Introduction and Presentation Overview

Greetings.

I'm Dr. Arthur Fleischer from Vanderbilt University Medical Center, Nashville, Tennessee, also known as Music City, USA.

My presentation is going to be on improved techniques for sonographic detection of ovarian cancer.

This presentation will cover the recent improvements in the sonographic detection of ovarian cancer.

I have no conflicts of interest.

Basically, I hope to present the principles and concepts regarding improvements in early detection of ovarian cancer with sonography.

Discuss the use of 2D morphology, transvaginal color doppler sonography 3D morphology and color doppler, as well as contrasted transvaginal sonography for early detection of ovarian cancer.

My presentation will cover microbubble ultrasound contrast, which can be used off-label for general purposes, but is currently not FDA approved.

This is a rather scholarly and important topic in this, disease, which is a so-called silent killer.

These microbubbles are the newest application of contrast sonography and have been used at a few academic medical centers throughout the world.

I appreciate the help of Dr. Barrow, RAF, and Phillips. In some of the images that I will present.

Dedication to Gilda Radner

I wanna dedicate this talk to Gilda Radner, who was a comedian, most notably in Saturday Night Live.

As you can see here, she did a cassette recording called, it's Always something describing her own personal experience with this deadly killer with ovarian cancer.

She, in fact, she had some risk factors that were not very clear, but two aunts with breast cancer.

And unfortunately, when she was being treated for infertility, they found that she indeed had ovarian cancer.

And unfortunately, she died, in May of 1989 and recorded this, account of it, only a month before her demise and her husband, gene Wilder, put funds together to start a, a screening program at Cedars Sinai in Los Angeles, which was a great, program and examined approximately 1200 women with transvaginal ultrasound and ca 1 25.

Interestingly enough, several ovarian cancers were found in asymptomatic women, but most of those were of the disseminated, type.

Anyhow, there are, due to this impetus, many cities have what's called gildo clubs, which are wonderful, ways for families of patients that are affected with ovarian cancer and other cancers to get together and to, become familiar with each other.

Ovarian Cancer: The Silent Killer

Well, ovarian cancer is a silent killer.

Approximately three quarters of patients that have ovarian cancer, first time present with stage three or stage four disease that is disease beyond the pelvis, or with distant metastases.

And as you can see, the five year survival for those patients is not good, 25% and 5%.

However, the five year survival of ovarian cancer in stage one, which is confined to the ovary, is 95% with pelvic extension is 80%.

So it's been a dream of many to shift the detection of ovarian cancer to its earliest stage, and therefore improve the longevity of patients affected with ovarian cancer.

So the thought is that with earlier detection moving, patients at the first time with di disseminated disease from, a terrible 25% to a better prognostic stage, that is stages one and two could decrease mortality by at least 50%.

Now, recently, there has been shown a very strong association of patients with ovarian cancer that have the BRCA breast cancer, antigen mutation.

The lifetime risk for BRCA one patients is 46%, which is very, very high.

And for BRCA two is 20%.

Now, it's also been shown that these patients that are BRCA positive have an ovarian cancer that is actually different from the standard ovarian cancer that arises from the epithelium.

This is called type type one, type two ovarian cancers actually arise from the tubal intra epithelial cells.

And these tubal intra intra epithelial carcinoma is seen in, many patients with BRCA one and two, and also have peritoneal disease in over half of these patients.

So the hope is that by screening for ovarian cancer, we can decrease the mortality and improve the five year survival of this terrible killer disease.

I should say that the incidence is 33 per hundred thousand in the us.

Screening Programs for Early Detection

Now, there have been four screening programs that have addressed the ability of sonography to detect early stage disease.

The one in the us, the PCLO, prostate, colon, lung, ovarian, project has enrolled over 78,000 subjects.

But it was found that the ovarian cancers that were found were not particularly early stage disease.

The largest study so far is from the uk where they've had over 200,000 women, and they have clearly found that, ovarian cancer screening has increased the detection of early stage disease.

And this is an ongoing project that the results will be finalized in the next few years.

There was a very large Japanese program, which is also ongoing of 41,000 patients, and they clearly showed that there was an earlier stage of detection in patients that were screened.

And finally, at the University of Kentucky, they have examined over 41,000 women and clearly showed that the overall longevity of or five year survival in the screen group is much better than in the control group.

So this data is very hardening that in fact, early cancer screening can improve the longevity of patients.

Types of Ovarian Cancer

Now, when we think about how sonography can detect ovarian cancer, we have to have the two state two types of ovarian cancer in mind.

The type one, as I mentioned, arises from the ovarian surface epithelium versus the type two, which tends to be very aggressive, actually arises from the tubal epithelium.

And this is a major challenge to the detection of early cancer.

In patients that are BRCA positive, there is a potential for labeled microbubbles, which could, actually both diagnose and treat in, in the future of this disease so-called theos sons.

So remember, the difference between type one and type two ovarian cancer, type one is relatively slow growing, developed stepwise and is associated with with A-B-R-A-S or A-K-R-A-S gene mutation that has a relatively long five year survival.

Type two, on the other hand, is rapidly growing.

High grade at presentation is related to P 53 or tumor suppression gene malfunction and has a short five year survival.

And this is a diagram showing the two different cancers and where they arise in the type one ovarian cancer, which arises from the tube from the epithelium of the ovary itself.

You can see that it's related, to genetic changes in the epithelium, in the oma epithelium that covers the ovary as opposed to the type two ovarian cancer, which arises from the tubal epithelium.

And these cells undergo a malignant change as you can, as is depicted here, developing tubal intra epithelial carcinoma.

And it's important for us when we talk about ovarian cancer, to account for these two types of ovarian cancer, that we have described.

Improvements in Sonographic Detection

Now, the improvement in sonography has been in many areas, firstly in color doppler sonography, where we're looking at vessel morph mess vessels.

We're looking at the flow in those vessels, and this is related to the detection of tumor neo angiogenesis.

We have to realize that with doppler, there can be normal physiologic findings that can mimic, angiogenesis.

Now we've also gone into 3D morphology and contrast, which I think are important breakthroughs in the early detection of ovarian cancer.

Color Doppler Sonography

Now starting with color doppler sonography, this has been around since the late eighties and basically we're detecting abnormal areas based on the detection of neovascular.

And over the years color doppler was first described as being extremely specific, but now, now over the years, through the various phases of any scientific evaluation of a, of a new clinical application, we can see in this graph that in the early days it's great and in, after what I call the era of inappropriate enthusiasm, we have first papers we have back to question mark applications.

And then hopefully finally we have that it has an important clinical application.

And clearly sonography has an important role in the early detection of ovarian cancer.

And the flip side of that is, has a very important role in confirming that a lesion is benign.

And I'm not going to address this, but it's very important of course, in torsion evaluation.

So with color doppler, what we're looking at phase shifts of the blood flow inside the structure we're interested in, of course, it's very dependent on that doppler angle.

And in fact, if you're not at 30 to 60 degrees to the flowing structure, you may not get an appropriate signal.

Now, the doppler that we evaluate with the spectrum is related to several things, the downstream resistance.

And we know in tumors there's very high downstream resistance because they lack a lymphatic system to, for the return of blood.

And we also know that the velocities that we detect relate to the vascular tone of the vessel, whether it's vaso constricted or vasodilated.

And we also know that it's related to the pressure inside the structure we're looking at.

Now, the vessels in, tumors lack a muscular media, and one would think that you see a complete different waveform, and you do, basically it's related to increased diastolic flow.

As we can see in this waveform, there's increased diastolic flow relative to systolic velocities.

But we know that there's an in-between state with a dilated normal vessel that gives a similar waveform.

Now with a muscular media and in constriction, the normal vessel with a muscular media will show a high resistance waveform.

Now, we can quantitate this by several ways.

We can do a resistive index, which is systolic minus diastolic over systolic.

We can take the pulsatility index, which is systolic minus diastolic over mean.

We can do a S 2D ratio like we do in the umbilical artery doppler, or we can look in the area under the curve to the power of that signal.

And when we do use power doppler, we can actually see smaller vessels that have slower flow.

And so, there is a major application improvement in using power doppler.

However, the disadvantages, it's very susceptible to tissue motion.

There's typically no directional information, and, there's limited temporal resolution, which is improved of course, by the use of contrasted microbubbles.

Now, here is the first example of velocity based, color doppler on the left, in this multiloculated hemorrhagic cyst versus the power doppler, which show vessels inside the wall, probably the areas of, of the site of hemorrhage.

This is a nice example of hemorrhagic corporal ludia, where we can see the vessel wall having, these, vessels versus with the velocity base, we barely get a signal in this hemorrhagic luteal cyst, which we can see on h and e stain on the left and on CD four staining on the right that we are vi are visualizing the rim vessels in this hemorrhagic lesion.

Now, also, we can look or get an idea of vessel branching patterns.

This, of course, is a microscopic picture, used in a computer reconstruction of the normal ovarian, vascularity shown by the group from Israel Schoenfeld many years ago.

And as, as you can see, the vessels have a very orderly branching pattern going from larger vessels to smaller vessels.

And histologically, these normal vessels will have this kind of rim of muscular media.

Now, compare that if you would to this, which is a tumor neovascular, which has clustered vessels, and the vessels themselves don't have the muscular mania.

So the question was, could doppler be sensitive enough to detect these changes in vessel morphology?

As I mentioned, angiogenesis as a normal process in corporal luteal formation and placental formation and wound healing.

But all of these are regulated, processes in tumors.

They are not regulated.

This is a picture of a corpus lutetium.

Now, angiogenesis or tumor angiogenesis was first described by Judah Folkman.

And basically what he pointed out is that for a tumor to enlarge from a few millimeters to several centimeters, it has to induce a blood supply usually off the host system.

And the blood vessels that are, developed are very tenuous.

This is a picture of a breast cancer electron micrograph of all of these vessels surrounding the tumor.

And as you can see, the vessels that form are very abnormal.

And in fact, at early stages of tumor development, there's focal areas of ischemia, inside the tumor itself.

So depending on when we actually see the tumor, we can get a different doppler appearance of, these ovarian cancers.

Now, color doppler is a basic way of looking at the vessel, network, but also looking at the physiology involved.

And as you can see on these diagrams with color doppler, we get the image and we get the doppler, spectrum, but we have a lot more information.

Now, when we look at the 3D evaluation of tumors, we can detect the tumor network, the relative vascularity of the tumor.

And these improvements have come about with 3D harmonics and contrast.

Now, back to physiology.

We know that during follicular development, we get development in the corpus lium of a vascular arcade, which, is extremely vascular as opposed to focal areas where there's no follicular development.

The ovarian vessels are tortuous and have very high resistance flow.

So during the examination, we try to look at the entire ovary looking for the intra ovarian vessels, as well as the larger feeding vessels to the ovary.

And as you know, the ovary has a dual blood supply.

It has a it vessel coming from the in fibular pelvic ligament, and it also has a vessel coming from the uterine artery 'cause the so-called adnexal, branch.

And we also have vessels which crossed with the tubal vessels.

Now, this is an image showing what color doppler of the mature follicle looks like.

And why is this important?

Because we need to differentiate tumor vascularity from physiologic changes.

And as we can see in this colored Doppler image, there's a focal area of relative ischemia where ovulation will occur, and this is a trans illumination of that ovary.

Now, here in a corpus lutetium, we have low impedance flow, and if we didn't have this color Doppler image, I think it would be difficult to pick this out, as in fact, a corpus lutetium.

But the typical low impedance flow of a functioning corpus lutetium is shown here.

And with power, we can see even, even finer vessels, surrounding the corpus lutetium.

So this is a picture of a so-called ring of fire.

The corpus lium, in fact, is a very endocrinological active structure.

And we can see on this doppler image, the vascularity, created by these vessels.

So in the late eighties, early nineties, it was proposed that the Doppler assessment of vessels around the ovary and in the ovary, could differentiate benign from malignant.

However, in our experience, we found this to be about 90% overall, accuracy that depending on which vessels you are imaging, you may get in fact different, doppler, spectral waveform.

So we expanded not only looking at the RI or the pi, they're looking at the vessel, distribution and the overall, velocity when we could obtain it, when we could see the doppler, angle.

And clearly tumors had central vessels compared to benign lesions.

The velocities were not statistically different, but clearly there was a trend toward lower pi in malignancies.

And the fact that there was a diastolic notch, usually indicated a branch point in a normal vessel.

So putting all this together, trying to come up with the absolute truth, whatever that is, we kind of, went overboard, I think, in trying to come up with all these parameters.

And we must remember that morphology is important.

Morphology is the study of shapes and benign lesions are thin walled septate, spider web-like internal interfaces and suspicious lesions are ones with papillary expressives, which can be detected very nicely with 3D and 2D for that matter, thick walled lesions and lesions with solid mineral nodules.

So if we look at this netter drawing of a typical S cyst adenocarcinoma, we can see that this, this, tumor has papillary expressives both projecting into the tumor as well as outside of the capsule.

And in fact, it's where these, papillary expressive, are, is where the neoplasia occurs.

Now, here's a patient that was studied for early pregnancy, yet in scanning her ovary, we can see that there's a focal, papillary expressions here.

So, the first thing is to realize that this could be other things, a clot, for example, or a benign lesions, and ask her to, come back.

This was the initial scan.

Eight weeks later, the lesion still is present, and in fact, it has no blood flow in the center of the lesion.

This turned out to be a borderline ovarian cancer.

Now, borderline ovarian cancers look like malignancies under the microscope or histologically, but are benign in their clinical, course.

Now, I think with 3D we can detect this difference in morphology.

3D can be done with a transvaginal probe.

And as we can see in this image from Dr. RAF's work, if we have a wall that has thin and regular, this is a very good sign of benignity.

This is a four DA 3D in, in, real time of a, of a ovarian, lesion sent to me by my friends in China.

And we can see that the walls are very thin, very regular, and, we can see this beautifully now with, 3D and some of our, so-called matrix array probes.

This is just a nice example of a cystadenoma, showing the thin septations.

Now with 3D, we can differentiate, these thin fibrin strands in this hemorrhagic ovarian cyst.

Now, hemorrhagics ovarian cyst can actually look morphologically very, very complex in that the septations can be thick, but the big difference is that there's no blood flow inside of these, septations where there's organized clot.

This is a 3D showing irregular septations.

I'm not far from Kentucky, and those caves where there's s stalagmites and s stalactites, this is kind of, the ultrasound equivalent of a sta s stalagmite.

Anyway, here we have a thick and irregular septation on 3D in an ovarian cancer.

This one from Dr. Braf shows a, dermoid cyst that has a very complex internal structure, but we, what we really, can detect on 3D that needs to be in the back of our mind, are these papillary expressives shown beautifully in this image from Dr. Ben RAF's work.

So with that in mind, with morphologic assessment in mind, we try to improve with color doppler, the detection of ovarian cancer.

And we do this by looking at the blood flow in and around, the tumor.

So tumors typically have peripheral, I'm sorry, central flow versus peripheral flow in benign lesions and have lower impedance in malignancies and have clustered vessels as opposed to regularly spaced and branching vessels.

This is work from Timberman, and the IOTA group, international Ovarian Tumor Analysis Group.

And basically they use color doppler in a scale, but benign lesions have no color.

Doppler are un ocular.

They have thin, solid components up to seven millimeters.

And, versus malignancies, which have irregular areas, ascites, papillary expressions, irregular multiloculated tumors and have high content of color.

When these parameters are used, the so-called rules, they achieved 93% sensitivity, 90% specificity in over 1200 lesions.

Here are some examples of where I think color doppler is helpful.

This was a patient many years ago now where, we could see in her ovary this abnormal, area.

And I also detected that this tumor was involved the fallopian tube, and this was the first case that I've seen that was of a so-called type two ovarian cancer with low impedance flow.

Here's the, picture at surgery.

The hemostat is on the tube.

And, and in the low power micrograph, we can see all these very abnormal vessels, in this, stage two ovarian cancer.

This was a patient that was, being evaluated for follicular monitoring.

And you can see she has a mature follicle here and one here.

But in this area, we have a large papillary expressions with a central feeding vessel with low impedance and histologically.

This is her, low power micrograph showing tumor in the epithelium and some central vessels within these papillary expressives.

Now this patient has a focal echogenic area in the wall of this cyst, and you can see it's of irregular thickness.

And clearly this was suspicious just based on morphology.

Now looking at the wall, we saw what looked like a venous signal.

As you can see here, there's no muscular media.

And, what was also very concerning in this patient was the cluster of abnormal vessels.

And this was another one of the cancers that arise, from the tube, the type two ovarian cancers.

This is a patient with an echogenic lesion about two centimeters in size, with a abnormal vessel with low impedance.

That was a clear cell carcinoma.

This is a patient that has a papillary, expressions with blood flow, and this was a serous, cyst adenocarcinoma shown here on histology.

It's important to differentiate, hemorrhagic lesions, such as the one shown here that has no blood flow in the center of this lesion.

And this is just due to hemorrhage inside of the, lesion.

Hemorrhagic lesions have a, lace like appearance, and these are two examples of hemorrhagic al ludia that have, on the left here some mild irregularity of the wall, but thin no vascular signal inside of them, even though they look morphologically, su suspect.

And as we know, we can see peritoneal cysts.

These are cysts usually after hysterectomy or other pelvic surgery, which are adhesions surrounded by fluid.

So our initial work in the late nineties showed that we had a pretty good predictive value with color Doppler 83% positive predictive value, a 98% negative predictive value.

And so we could improve the triage of patients and her certainly improve the odds ratio that is the patient, that go to surgery, the ones that are suspected of having, cancer.

And there were many, many other studies, all around the world.

But clearly one could show an improved specificity of, using transvaginal color and improved positive predictive value, using this technique.

So, is color doppler clinically useful?

The answer clearly is yes.

A lesion with blood flow in a, in a young patient, could be physiologic as opposed to in a post-menopausal patient.

They should not have lesions with blood flow.

So it's imp it's accuracy is improved in the post-menopausal group.

And we look for, clusters of vessels, a as I've shown you that have low impedance and high velocity.

Can 3D help?

This certainly 3D can quantify the overall blood flow, and there is a potential for targeted ultrasound using labeled microbubbles.

So back to understanding what we're looking for, we're looking for abnormal, network of vessels that are clustered, such as this.

Now on this 3D planar image, we have a patient that has two lesions, an ovarian cancer here, and an ovarian cyst here.

And we can clearly see on this color Doppler 3D image, abnormal vessels in the ovarian cancer shown in this, set of images.

This is a patient with a papillary expressions with abnormal flow inside of it on 3D.

This is a patient that has an, focal mural nodule with a large feeding vessel shown on the color Doppler, right here, which when we, when we show this on the video, you can see the vessel, quite clearly.

Now, metastatic lesions to the ovary tend to have a central feeding vessel.

And, this is a 3D volume that I'm showing, on this video, and we'll see that there's an abnormal group of vessels supplying this tumor in this patient with an ovarian, metastatic lesion.

Here is the central vessel on 3D.

I hope you can appreciate this group of vessels, as opposed to vessels which are peripheral.

So this is a central vessel in a, metastatic lesion, to the ovary.

And here is the, 3D slice showing this.

This is the vessel, map, which you can see the central abnormality.

Now, here's a patient with a tube ovarian abscess.

I still believe that there's an overlap with vessel morphology.

This patient, had a large multiloculated lesion that clearly had the similar morphology to an ovarian cancer.

And you can see morphologically the vessels coming off the pelvic sidewall, simulate that the vessels in a tumor.

With this 3D volume set, we can do a, bloodless surgery on ultrasound, cutting through the tumor, taking that area off, and, tilting the tumor down to see the multiple LOEs inside of this, tube, tube ovarian abscess.

And so 3D is very helpful, not only in morphology, but looking at the vessel network in, lesions such as this.

Okay, so these are more, images of ovarian tumors with 3D color doppler showing, central vessels in these lesions.

In this lesion we have gray scale thickened abnormal wall inside of these, lesions with very abnormal vessels.

3D Morphology and Vascularity

So let's go back to the idea of angiogenesis and tumor angiogenesis in particular.

This is a process at the capillary, level.

And tumor profusion varies depending on the story of ischemia, central ischemia.

And what we really want to see is the tumor, I'm sorry, is the flow inside of capillaries.

Now, to do this, we need these microbubbles, and here is a, schematic drawing of a small capillary and red blood cells.

And these microbubbles are about a third the size of an RBC.

Now, we don't need a lot of these to get a doppler signal using harmonics.

And, here's a nice picture showing some of the profusion in, in vivo microscopy.

And, So, this is kind of, a, voyage through the body here.

I guess you can see the red blood cells in these small vessels, coursing through the, the mesentery.

And this is microscopy, but we're down at the level now of the capillaries.

And, that's what the, microbubbles are depicting.

Now we started doing the, microbubble work many years ago.

Dr. Peter Burns and Stephanie Wilson in Toronto were of course pioneers, to use the microbubbles primarily in the liver.

And we thought we could use this to detect ovarian tumors.

Now, we have used the mic, the microbubble, called Definity.

It comes in a, in a vial shown here.

We put in, some saline.

It gets it all shook up, in 45 seconds, becomes a suspension.

This milky, contrast.

We, put this into a tuberculin syringe, which holds one cc.

We have intra, venous access through an antecubital vein, and we inject very tiny doses of this, followed by 10 ccs of, sterile saline.

And what we see here is a typical, series of images.

This is the fundamental image of this patient with ovarian cancer.

The red shows the area of interest, and this is the harmonic image, and this is the time intensity curve.

And as you can see, we can, we can quantitate the time to peak, which is shown here.

We can quantitate the area under the curve, which is the overall vascularity.

And we can detect the washout, which, is very important because as opposed to, some MRI of breast cancers, the washout in ovarian cancer is actually increased as opposed to decrease.

And, here is a, picture, microscopic picture, of course of a tumor.

We can also do microvascular imaging showing these very abnormal vessels.

Now, tumors are approximately 5%, made up of vessels.

The vessels are very abnormal, and this is a microscopic picture, on scanning electron microscopy of, normal on the left with arteries going to arterials, going to capillaries, going to venal as opposed to tumor.

Vascularity on the right, where there's a mishmash of vessels, vessels that are big and small and blind ending.

And as you can imagine, if you're a red blood cell, you're going to take a lot more time getting through the very abnormal, collection of vessels on the right, which is a tumor than on the left, which is a normal vessel.

Now tumor angiogenesis begins by erosion of the basement membrane and, the vessels escape through the, the endothelial gap, the fibroblasts, and begin to form vessels.

And basically it's very important to differentiate the microbubbles, which are intravascular versus gadolinium and other Mr contrasts that leak through these endothelial cell gaps into the interstitium.

So we have an intravascular agent, which is very, potentially helpful not only in diagnosis, but also in therapy.

If we can put a, target on the outer part of the bubble, which is a lipid, coating, we could potentially have it go specifically to tumor and then potentially have either drug on the, on the outer shell or in the, in the bubble itself and break the bubble.

So this has tremendous application for targeted drug therapy and also for gene therapy.

This is a microbubble, which, is exploding in front of the eyes of the, microscope here showing you can oscillate the microbubble and break it up.

Now, this is some work that my colleague, Dr. Lich and I, were doing and continue to do, and finding the labeled microbubbles that go to the so-called VEGF vasogenic, endothelial growth factor sites in a vessel, and potentially, control the angiogenesis of tumors.

And we can do this by the strep Aden, molecular systems shown here.

And in one study from, Dr. Bro's group, in a hen model, they've shown that targeted microbubbles may in fact, improve early detection, and we're working on this, in our facility as well.

Microbubble Contrast-Enhanced Sonography

Okay, so firstly I wanna go back to, the contrast so-called enhancement kinetics.

It was shown early on that tumors have a somewhat different enhancement kinetic.

In fact, the images, shown here from orin's work from Finland showed that malignancies have higher intensities, have longer washout times than benign lesions.

This was also shown from the French group, mar mare and, his associates In that, contrast, images show a difference in the contrast kinetics in malignancies as shown in red and in benign lesions.

Just to refresh, what we're looking at is the washout phase, and the area under the curve.

Now we can also show this in parametric display and we can show the peak enhancement in these lesions.

The wash in rate, the time to peak and the area under the curve.

This was the study from Europe of 72 patients, and they showed that contrast was almost as good as pat pattern recognition.

And their main problem was differentiation of benign, I'm sorry, from borderline, from malignant lesions.

Well, both are ovarian cancers.

And I think when we go back to what we're trying to accomplish that is detection of tumors, we can see that color doppler is still very, very helpful.

This is our study, which we, published many years ago now, where we first did a transvaginal found abnormal morphology, only in the vascular areas.

We, we, took a three DA three minute cine loop and quantitated the wash in peak enhancement wash out and the area under the curve.

This is a, example of such, a patient.

We can see the fundamental image here of this cyst, the microbubbles coming in.

This is the harmonic image and the so-called buildup image.

Here we can see that there's no particular area of increased activity in the wall.

When we look at the time intensity curve, you can see that there's a very rapid wash in and a very rapid washout in this benign lesion.

And this is the 3D, doppler of that lesion, which was a per ovarian cyst.

This, on the other hand, is a patient that has clearly an abnormal area here.

If you watch this region, it washes in rather quickly, but also washes out equally quickly.

So this is a cystadenoma, a benign ovarian epithelial tumor.

As shown on this, time intensity curve, there's very rapid washout of this benign lesion.

And here's the parametric image of that lesion.

Now contrast that to this patient.

Here's an ovary that's normal in size about a centimeter cystic area, and watch on the harmonic image the buildup of the ovarian, blood flow.

As you can see, they're completely different.

There's increased flow, which is maintained, throughout the venous cycle in, in fact.

And here's the time intensity curve showing very long washout.

In fact, at the end of three minutes, the lesion had not totally washed out.

Now this is the other ovary in the same patient.

We can see this ovary is enlarged and there's a peripheral area of enhancement.

And this lesion is, an stage two ovarian cancer, shown very nicely with the contrast.

Now again, very long washout in this adenocarcinoma, and this is the growth specimen showing the tumor on the right with a cystic area.

And this was the tumor on the left.

This is a patient with a very complex appearing, lesion.

Echogenic material is floating throughout this lesion, and on the mr there's even in fat suppression, it lights up, so it's not a dermoid.

And on the contrast enhancement focus on this nodule right here, we can see a buildup of contrast inside of a nodule in this borderline, mucinous cyst adenocarcinoma.

And, echogenic, the echogenic material was just debris floating, inside of this gi type of borderline tumor.

When we look at the time intensity curve, we again see long washout, as shown here, and also on the parametric image.

Contrast that again to a solid lesion here with very quick washout being a fibroma.

And here's the parametric image of that fibroma.

Now, our, study had about 50 cases with 12 borderline or malignant tumors.

When we look at the time to peak, there was no difference.

When we look at the peak enhancement, there was clearly a big difference in benign versus malignant.

When we look at the washout, there was complete difference in the two sets, and the area of the curve were very significantly different.

And so we built a receiver operating curve with these parameters showing that the washout was the, and the area under the curve were extremely, helpful and specific for the ovarian cancer diagnosis.

Of course, we need larger series and different types of ovarian cancer To confirm this initial observation, if we look at just the number of vessels with color doppler, we can see in this lesion that was scanned by my chief sonographer at the time, Donna Keppel, we can see lots of vascularity.

Even before we turned the color on, we could see lots of speckles and that was a peritoneal tumor with, lots of vessels.

We can quantitate this using vascularity index, which is a number of pixels with color over the total flow index, which is power weighted pixel, or the vascular vascularity flow index.

We could even if we are so prone to, to look at the fractal dimensions of these vascular networks and also comment on the vessel branching and the caliber.

This was a study that we did in at Vanderbilt, just looking at differences in overall vascularity and benign versus malignant.

And, the volumetric, assessment of these vessels is very, very helpful in differentiating the different, vessel branching and distribution types.

The penetrating vessels versus the vessels that remain in the periphery of the lesion.

This is kind of a combination of both.

This shows a feeding vessel inside of a tumor.

This is a 3D showing a smooth wall in a benign lesion.

And with color doppler, we can see lots of blood flow, but it's regularly spaced.

And this is the power doppler of that same lesion.

And we can contrast that with the 3D vascular set shown here in the tumor with vessels that are big versus small versus blind ending.

And, this is the 3D volume set of these vessels and in malignancy.

So, by quantitating these values, alazar in Spain showed that we could reduce the number of false positives, using these parameters.

Conclusion and Acknowledgments

Now, I'd like to emphasize as we finish the, the role of, of, contrast seems to improve not only sensitivity and specificity, but also detection.

So, I wanna leave you with these overall, ideas.

When we look for ovarian cancer, the best, and most specific finding is papillary expressions.

Now, most of these lesions will be borderline cancers, but they're still important to detect.

Irregular or focal focally, thickened walls or septation is a sign of ovarian cancer.

Obviously, ascites and peritoneal or omental mass masses are found in malignancies.

On contrast, what we see is greater peak enhancement, longer washout time, greater vascular volume.

And uh, with quantitative 3D we can see, we can quantitate these values.

I want to thank my, very capable associates, Dr. Lehe in particular for his, very long and continued help with early detection of ovarian cancer, as well as the, clinical colleagues, shown here.

Thank you.