Aunt Minnie’s Ovary: Patterns, Pearls, and Pitfalls in Adnexal Imaging - HD

Introduction

Hi, my name is Tre Patel.

I work at the Mayo Clinic in Arizona

and my talk today is Ant mini's ovary patterns, pearls

and pitfalls in ovarian imaging.

I have no disclosures to make on this lecture

and my objective today is to outline an algorithm

for thinking about an adnexal mass when we

see one on ultrasound.

To review the sonographic appearance of a variety of types

of physiologic and pathologic adnexal cysts,

describe criteria that will allow you

to make a confident diagnosis and discuss if

and when a mass that you identify

on ultrasound needs follow up or surgical intervention.

Categorization of Adnexal Disease

Now this is a categorization of adnexal disease

from the perspective of a sonologist.

You can see that these

categories are rather idiosyncratic

because there are some very specific categories like

hemorrhagic ovarian cyst

and some pretty broad categories like solid mass.

But the reason why I like this categorization is

that if we can put an adnexal mass

that we identify on ultrasound into one of these categories,

then we know exactly what to do with that in terms

of any follow-up imaging or surgical intervention,

because that's the basic approach that the sonologist has

to an adnexal mass.

What you want to do is ask yourself, am I reasonably confident

that I can put this mass

that I'm looking at into one of those categories?

Do I know what this is?

If not, is there a subsequent test

that might even include a follow-up ultrasound

that won't harm the patient

and that has a reasonable chance

of showing characteristic findings to either avoid surgery

or to improve surgical planning?

Pattern Recognition vs. Mathematical Models

Now, clearly pattern recognition works.

When we look at the adnexa,

there have been many scoring systems

that have been developed to try to distinguish

between benign and malignant adnexal masses.

And the most extensively tested

formula is this prediction model that's been put out

by the IOTA, the International Ovarian Tumor Analysis Group.

It's a mathematical model,

and several studies have shown

that this mathematical model works almost as well

as pattern recognition by experts, but not quite.

So I'll submit to you that the choice is yours.

You can either learn to become an expert in pattern

recognition when you look at these adnexal masses on

ultrasound, or you can get out your calculator.

Now this is a silly slide,

but this is actually the mathematical model

that explains how you might be able to analyze

an ovarian mass

that you identify on ultrasound based on a variety

of different criteria.

Clearly, we're not going to be doing that at the

view box on a daily basis,

but it's important for us to understand

how we fit into the picture as imagers.

The management of adnexal masses is highly influenced

by the sonographic impression

that you will provide in your report.

We want to strive to avoid unnecessary intervention

because remember, the vast majority of the masses you see

in the adnexa of a woman are going

to be non neoplastic ovarian cysts,

and most of these are going to spontaneously resolve.

Algorithm for Adnexal Masses

This is an algorithm that we'll be working through to try

to help us in formulating recommendations for follow up,

for not having any follow up or for surgical intervention.

And the first step in this algorithm is

to decide whether the mass you're looking at is a simple

cyst, an anechoic cyst.

What do we mean by that? We mean a cyst

that lacks internal septations that has no nodules,

it has no wall irregularity and it's oval or round in shape.

If you can make the observations that lead you to confidence

that you're dealing with a simple cyst,

then the diagnosis has been established

as either a non neoplastic cyst or a benign neoplasm.

Malignancy is practically excluded when you

see a simple cyst.

Now it's important for you to have faith in that assessment.

The risk of malignancy in anechoic cysts is practically zero.

In a study of 3000 anechoic cysts done in over 15,000

post-menopausal women,

and these weren't just small cysts, these were cysts

that measured up to 10 centimeters in size, the risk

of malignancy was less than 0.1%,

and that's with a 95% confidence interval.

But what about those rare malignancies?

Well, it turns out that in this series,

those rare malignancies were always borderline malignancies.

They were always over five centimeters in size

and they always had identified a change on a follow-up

sonogram, meaning that septations

or tiny wall nodules associated

with the cyst were identified on the follow-up study.

It's not clear whether those are new

or whether they were simply missed on the first study.

So if the answer to the first question is

that you are in fact dealing with a simple cyst,

the next thing that we want to try to assess mentally

is whether this is a non neoplastic cyst or not.

Predicting non neoplastic cysts is important

because it allows us to ignore cysts

that are almost certainly self-limiting

or clinically unimportant or to follow

or remove cysts that we think are not non neoplastic cysts

because they have the potential to cause harm

that harm being mainly that they'll enlarge in size

and potentially become an emergency if they tors.

So the four factors that we will look at to try

to predict whether a cyst is non neoplastic

or not is to look at stability if we're able to location

size and the patient's hormonal or clinical status.

So let's assess those one by one.

Stability

Now, stability is probably the easiest to understand.

Clearly if a cyst is decreasing in size

between two exams that's

a non neoplastic cyst, there are no instances

that have been shown

where a neoplasm actually got smaller on a follow-up study.

And so if you can show that the cyst is decreasing in size,

you don't need to treat that cyst.

It's expected to resolve should

not be a problem for the patient.

If a cyst doesn't change in size, it's also likely

to be a non-neoplastic cyst.

Assuming that you've had a long enough interval

of observation, clearly a cyst that's been there

for 10 years and not changed in size is not

a neoplastic cyst.

The question is how long do you have to wait

to assess the absence of change in size?

And we'll discuss that in a bit.

If a cyst is increasing in size,

I think at this point you have to conclude

that it's probably a neoplasm

because non neoplastic cysts typically do not progressively

increase in size over time.

Surgical removal therefore would probably be important

to perform if the patient is relatively young

because the cyst will just increase in size

until it gets quite large.

And finally, if you can show

that the cyst is rapidly developing, meaning

that it wasn't there three weeks ago

or wasn't there three months ago

and now is there, it's almost certainly going

to be a non neoplastic cyst

because neoplasms do not grow like that

where they weren't there three months ago

and now exist at least not benign neoplasms,

which is your differential diagnosis

for a simple cyst in the ovary.

And so if it's rapidly developing, you could ignore

that cyst if the symptoms allow you to ignore it.

Location

Well, let's turn our attention to location then.

If the cyst can be shown to be next to the ovary,

it's quite unlikely to be a neoplastic cyst.

Now there are case reports of para ovarian cystadenomas

and even para ovarian cystadenocarcinomas.

But a good rule of thumb is

that if you can see a separate ovary, the cyst is not going

to be a neoplasm paraovarian location also means

that it's less likely to

therefore require surgical treatment.

Size

And finally, size. And this is what most people want

to pay a lot of attention to, unfortunately,

there really are no absolute thresholds that allow us

to distinguish a non neoplastic cyst from a benign neoplasm.

There's no number below which everything is a non neoplastic

cyst and above which everything is a neoplasm.

But there are some numbers that can be helpful in patients

who are still in pre before menopause.

If a simple cyst is less than

or equal to five centimeters, it's very likely

to be a non neoplastic cyst.

On the other hand, if it's over seven centimeters,

that's not very likely to be a non neoplastic cyst and

therefore it's more likely to be a neoplasm.

In postmenopausal patients, it somewhat depends on

how many years they are away from

their final menstrual period.

But one can confidently state that if a cyst is less than

or equal to one to three centimeters in size,

it's more likely to be a non neoplastic cyst.

This is a simple cyst of course.

One thing we all should be on the same page on is

understanding how we talk about adnexal cysts.

How big is this cyst

where we have four different measurements, two taken

in the longitudinal plane

and two taken in the transverse plane?

And it turns out that the Society of Radiologists

in Ultrasound had a consensus conference,

and this is back in 2009

and concluded that when we talk about adnexal

and ovarian cysts,

we should probably be talking about the largest measurement,

at least as it refers to simple cysts

and as it refers to things that are round or oval in shape.

And so this cyst is a 4.4 centimeter cyst

because the largest measurement is 4.4 centimeters.

Hormonal and Clinical Status

The hormonal and clinical status

of the patient is quite important to consider.

A non neoplastic cyst is most likely if

the patient is menstruating.

But remember that a cyst

that's simple in a postmenopausal woman is

oftentimes non neoplastic.

In fact, about 50%

of these simple cysts in postmenopausal patients will

resolve on their own.

Now that resolution process typically will take

about one to two years.

Resolution is also much more common if the patient is within

the first decade from her final menstrual period.

So because of that, in our laboratory at Mayo Clinic,

we always will have an assessment of

and a recording of when the patient's

last menstrual period is.

And we want to understand that in the context

of when we're doing the study.

We also have benefit of using the electronic medical record

at our facility and that helps us

because the clinical status is just

as important to consider.

And I won't go through all of these notes,

but this 21-year-old woman with the cyst that I'll show you

had pain that started at 10 o'clock in the morning

and it hadn't got better.

So she was concerned, but now the pain is dull in lessening.

So it was a relatively acute onset of pain that's resolving.

And so that is a very common scenario

for a non neoplastic cyst

and that will factor into our assessment.

Follow-Up for Simple Cysts

So in this algorithm, if we've decided

that the cyst is un simple

and we've decided that it's unlikely

to be a non neoplastic cyst because it's increasing in size

and follow up or because it's over seven centimeters in

size, we'll take action on that

and I'll be talking about

that a little bit later in this lecture.

But in many cases, the cyst will be one

that we very likely will predict

that it's a non neoplastic cyst

and one that we don't even want follow up on.

For example, a patient who has a cyst that didn't have

that same cyst three months ago,

that's clearly a non neoplastic cyst.

One can conclude in the asymptomatic patient

with these suspected non neoplastic cysts

that you can say something to the effect that in the absence

of relevant clinical signs

and symptoms, follow up ultrasound is very unlikely

to be necessary in other patients

where we're not sure whether we're dealing

with a non-neoplastic cyst,

but we think we probably are, we'll probably want

to get a follow-up ultrasound

to help us make that determination.

So the important thing to try to do is distinguish

when we're going to take this approach

where we're not going to get follow up

and when we're going to take this approach

where we are going to get follow up.

Can we agree on some size threshold

to help make that distinction?

Well, the consensus conference did come

to a consensus on this issue.

For simple cysts, the experts in attendance for

that conference agreed

that really follow up was not indicated when an asymptomatic

woman had a simple cyst and she was premenopausal

and that cyst measured less than

or equal to five centimeters in maximum diameter.

There was a little bit less consensus with respect

to the size cutoff in postmenopausal patients.

All of the participants understood

that three centimeters could be used

by reasonable practitioners of imaging and

basically say that a postmenopausal patient

with a simple cyst less than three centimeters might not

need follow up, but realized

that some people might want to follow up cysts

between one to three centimeters.

And so basically advocated that the threshold between one

and three centimeters could reasonably vary

between practices based on how

important it was to that practice.

Not to mistake a benign neoplasm from a non neoplastic cyst

that's asymptomatic.

So again, the SRU criteria for ignoring a simple cyst,

asymptomatic patients less than

or equal to five centimeters.

If you're premenopausal and less than

or equal to one to three centimeters, you get

to choose if it's postmenopausal.

Now, my criteria are slightly different.

What I would advocate is that many of the patients

that we're imaging actually do have some vague symptoms.

So if the symptoms are vague

and then go away, I would consider that

to be an asymptomatic patient.

And so they don't have to be asymptomatic at the get go

If it's less than or equal to five centimeters, I agree,

a premenopausal patient would not need to be imaged again.

And in our practice we use two centimeters as the threshold.

So a postmenopausal patient with the simple cyst

that we well evaluated

that's less than two centimeters would not need

to have follow up imaging.

And finally, I've added a new criteria.

If you've seen the cyst as a new finding, so

that you in other words, saw the ovary

as a normal structure within the last six months on an

ultrasound, and now you see a cyst that's simple,

that cyst does not need follow up

because that by cyst almost by definition is going

to be a physiologic non neoplastic cyst.

Now it's important to understand why

you're following the cyst.

This is oftentimes confused by people.

You're not following the cyst

to determine if the cyst is malignant or not.

You've already made that determination.

Remember, by calling it a simple cyst,

you've practically excluded malignancy from the diagnosis,

especially if the cyst is less than five centimeters.

No, the reason you're following the cyst is

to determine if the cyst is more likely

to be a benign cystic neoplasm or a non neoplastic cyst.

So with that in mind, if you are presented with this set

of images and this clinical scenario,

a 55-year-old woman who's asymptomatic

and has a four centimeter cyst,

when would you perform the follow-up study?

There can be a number of potential choices

that you would make, but in my estimation,

the proper choice is six to 12 months, not three months

or six weeks as many people reflexively would do

for a premenopausal patient.

And the reason for that is

that if this is a non neoplastic cyst

as we predict it probably might be,

versus a benign neoplasm, your distinction

between those two is going to rely on time

to decide whether the cyst is stable

or decreasing in size,

in which case it's a non neoplastic cyst

or increasing in size, in which case you can predict

that it's going to be a benign neoplasm.

And so if you take a look at that, the expected size

of a four centimeter cyst over time,

if it's a moderately enlarging benign neoplasm

with a doubling time of let's say 180 days in six weeks,

if you image that patient again,

it would only be 4.2 centimeters

and you will not be able to tell

that it's actually changed in size.

In fact, you're very likely to make the

erroneous assumption that it hasn't changed in size

when in fact it is slightly increasing in size.

And so you need to give yourself enough time to be able

to distinguish whether the cyst

has actually changed in size.

And this is even more true, the smaller the cyst becomes.

If two centimeter mass over time, if it's doubling in size,

every 180 days would only be 2.1 centimeters in size.

If you looked at it in six months, you will not be able

to make that diagnosis of enlargement.

Whereas if you look at it in six months, it'll be two

and a half centimeters in size

and will clearly look different.

So if it's a simple cyst and you're going to follow the cyst

because it falls into your follow up criteria, you need

to pick a reasonable interval for that follow up the Society

of Radiologists in Ultrasound as a simple rule stated

that if it's a premenopausal patient

and if it's an over five centimeter simple cyst

and you're going to follow it,

get the follow up in 12 months.

Again, remember these are asymptomatic patients.

If it's a postmenopausal asymptomatic patient

and it's above your threshold, which is between one

and three centimeters, again,

they said get a 12 month follow up

and keep getting that 12 month follow up

until you're fairly confident

that either the cyst has resolved or is resolving

or isn't changing in size,

there really weren't any recommendations on when

to stop the follow up.

Again, I've modified this a bit for my practice,

which I think is reasonable.

What I do is in the premenopausal patient, if I'm going

to follow the cyst,

I'll get the first follow up in three months, not

because I know that I'll be able

to make a change in assessment of the size

of the cyst if it's enlarging.

On the other hand, most of those cysts between five

to seven centimeters that are simple are in fact going

to be still non neoplastic physiologic cysts

and they clearly will demonstrate resolution

or change in size for the smaller within three months.

If it's a postmenopausal patient

and the cyst is over three centimeters in size,

I'll get the first follow up in six months

because that gives me an opportunity to make an assessment

as to whether the cyst is increasing in size

and gives me another chance to look at the margins

of the cyst to make sure that

I really believe that it's simple.

On the other hand, if it's a smaller cyst between two

to three centimeters in size

and postmenopausal, I do do a 12 month follow up

and I'll cut these intervals in half.

If there's anything about the cyst

that makes me feel a little queasy about it,

we'll get a follow up sonogram yearly

and then we'll double the interval of the follow up

as long as it's stable.

So in other words, if the cyst is stable

after the first year,

we'll get the next follow up in two years.

If then it's stable at two years,

we'll get the next follow up in four years

and so on and so on.

So again, the SRU consensus criteria, no follow up needed.

If it's less than or equal

to five centimeters in a premenopausal patient less than

or equal to one to three centimeters,

you pick the threshold in a postmenopausal patient,

get a follow-up ultrasound.

If it's over five centimeters or above that threshold,

and still less than seven centimeters,

and if it's over seven centimeters,

you might consider a surgical consultation.

If Not a Simple Cyst: Physiologic Processes

Now, back to the algorithm.

Let's say you've decided that the cyst

or mass you're looking at is not a simple cyst.

The next question then is to say, could the findings

that are making you decide

that this is not a simple cyst be due

to a physiologic process?

Well, what are those physiologic processes

that can cause concern but really should not?

Well, one of them is if you see two

simple cysts next to each other.

So adjoining cysts can simulate an intervening septation,

the corpus luteum, which is sometimes cystic,

and sometimes involuting can have wall irregularity

or even look like a small solid mass.

And finally, hemorrhage can have a confusing appearance

that can simulate septations and nodules.

These are three physiologic events that can cause us

to worry about a cyst.

Here are some examples of adjoining cysts.

In both of these cases, this could reasonably be interpreted

by a imager as a possible septation,

although I think there are clues in both cases that suggest

that these are not septations,

but rather an intervening area of parenchyma

between two cysts.

Here's another case of two simple cysts actually separated

by ovarian parenchyma.

Now I know that they're simple cysts

because I know the outcome,

but I think it would be a reasonable conclusion for somebody

to make to say that this is not a simple cyst,

that this might in fact be a septated cyst.

And so you would not be able to say that

that's a simple cyst.

Note that the presence

of blood flow within this pseudo septation

does not really allow you to be more worried about this

as a neoplastic septation

because trapped ovarian parenchyma between two simple cysts,

as in this case will have blood flow.

Here's the other spectrum of change

where a simple cyst that is physiologic in both

of these cases actually has wall irregularity

because of evolution of the cyst over time or

because of wall based clot.

And in fact,

corpus luteums oftentimes look like small solid masses.

They don't look cystic at all. So here,

if you were really to look at this

and categorize it,

you might be worried if you didn't have a lot of experience

that this was a small solid mass within the ovary,

but in fact, this is the corpus luteum as manifest

by this ring of flow around the cyst.

Hemorrhage can be very confusing to the uninitiated,

but you want to look for fibrin strands,

but it can look like septations,

it can look like wall nodules.

Now notice that this wall clot

actually has a different echogenicity than the wall itself.

So that's one clue that you're dealing with a clot

and not a wall nodule,

because if this was a wall nodule growing from the wall,

it should have the same echogenicity as the wall itself.

The other clue is the shape of that nodule.

So the key ultrasound features

of a hemorrhagic ovarian cyst are

that there are internal echoes, which can be clumped

as we see in these instances,

oftentimes

with a lace like or fishnet appearance.

And so these are not septations,

these are fibrin strands within a sea of clot,

and the retracting clot sign is very useful

for you to look for.

So again, as I mentioned

before, this wall based clot isn't a mural nodule

because it doesn't have the same echogenicity as the wall,

but also has these concave margins

or straight margins that allow you to be confident

that you're dealing with clot.

Now, if you suspect clot on the wall of the cyst

as a possible cause for the irregularity of that cyst,

you want to test that hypothesis by applying color doppler

because clot really should not have blood

flow as we see in these cases.

Another useful feature to suspect a physiologic process

is this circumferential flow.

So if I see ring flow around a heterogeneous cystic area,

I'm not worried as much about this being possible

wall irregularity

because it has the characteristic signature on doppler

of a corpus luteum.

So if the cyst

or mass that we're seeing isn't a simple cyst,

but we decide that the findings that we're seeing

are reasonable for a physiologic process, then we're back

to this side of the equation where we want

to decide whether we're going to follow this at all

or not.

Cysts that are due to hemorrhage do resolve over time.

So that's one thing in our favor.

If we think it's a physiological process, it should change.

This is a landmark article that showed

that hemorrhagic ovarian cysts when scanned on a weekly basis

for three months, 30%

of them would resolve within two weeks.

And a hundred percent

of them in this series resolved in eight weeks.

Now, there are anecdotal cases both reported in the

literature and in my own experience

that take longer than eight weeks.

So although eight weeks is a very good timeframe

to expect all hemorrhagic cysts to have resolved,

sometimes they take a little longer.

There's really no data to suggest that the size

of a hemorrhagic cyst really matters

in its eventual resolution.

So based on this evidence,

the SRU consensus conference decided

that if you had an asymptomatic patient

with a hemorrhagic ovarian cyst based on the sonographic

appearance, you really didn't need follow up at all if it

was less than or equal to five centimeters in size,

and you could be confident

that it was a hemorrhagic ovarian cyst.

In other words, it had the classic features

of a hemorrhagic ovarian cyst, those being the presence

of fibrin strands, the presence of retracting clot,

the appropriate history and circumferential flow.

You should get a short interval follow-up in six to 10 weeks

when it was over five centimeters

or if the findings were suggestive but not classic.

Specific Diagnoses

So if we get to this stage

and we decide that no, the findings are not reasonably due

to a physiologic process

or, they might have been due to a physiologic process,

but because we followed it and it hasn't changed or

because it's over a certain size,

we can't be reasonably certain

that this is a physiologic process, then we want to go

to the next step in the algorithm

and say, are there findings to suggest a specific diagnosis?

So this is our list of diagnoses,

and if we're looking at specific diagnoses, yes,

we can make specific diagnoses for a number of these.

Endometrioma

So let's first talk about the endometrioma.

And the advantage of being able to make a specific diagnosis

with some confidence on ultrasound is

that then the patient can be treated

and managed as appropriate for that expected diagnosis.

And surgery is not always what is needed, especially

for endometriomas.

Endometriomas have been evaluated

for many years on ultrasound.

There's lots of literature in this regard,

but recognized the vast majority of studies

that looked at the characteristics of endometriomas

really didn't test those characteristics against the

competing diagnosis of acute hemorrhagic ovarian cyst.

And you need to recognize

that acute hemorrhagic ovarian cyst can mimic endometriomas

in premenopausal women.

Now, endometriomas can be found in pre

and postmenopausal women.

And the other key concept here in endometriosis

and endometriomas is that it's been clearly shown now

that patients who desire fertility, who want to undergo IVF

have reduced ovarian responsiveness

if you have a surgical intervention for their endometrioma.

So nowadays, if an infertility patient has an asymptomatic

or relatively asymptomatic endometrioma, it's best thought

to leave it alone, not to take that out.

It's a don't touch lesion

because the patient will have better ovarian responsiveness

to IVF if you don't do the cystectomy.

So this patient who has two masses, one in each ovary

has two endometriomas,

and they're both showing the classic appearance

of diffuse ground glass echoes or low level echoes.

Notice how the echogenicity

of those echoes is different within each endometrioma.

Here's another patient that has an endometrioma

with diffuse low level echoes.

However, this patient who has two small masses

with diffuse echoes does not have an endometrioma.

In fact, this is one of the pitfalls is not understanding

that acute hemorrhagic ovarian cyst can mimic this appear.

And in fact, this patient had

two acute hemorrhagic ovarian cysts,

which resolved on follow up.

So to help distinguish acute hemorrhagic ovarian cyst from

endometriomas, one can observe

and look for these hyperechoic wall foci,

these bright dots on the wall.

Now what are these bright dots?

Well, it's thought that what these are,

are cholesterol deposits from the lysed red blood cells.

Remember that red blood cell membranes

have cholesterol in them.

All cell membranes have cholesterol when they lyse,

as they do in an endometrioma

and fall to the edge of the endometrioma, there are

phagocytes in the ovarian parenchyma that will chew up

that cellular debris.

The cholesterol is indigestible,

it gets stuck in those phagocytes

and then forms these echogenic foci,

which resemble the cholesterol com,

the cholesterol deposits that one would see

with adenomyosis in the gallbladder.

In fact, that's what we think these things are,

are small cholesterol deposits on the wall

of a chronic cyst.

Now, this doesn't happen only in endometriomas,

but the good news is, is that these things do not happen

in acute hemorrhagic ovarian cysts.

So if you use a diagnostic criteria of the presence

of low level echoes, the absence of neoplastic features

and hyperechoic wall foci, you're only going to capture

about 30 to 40% of endometriomas

because most endometriomas don't have these hyperechoic foci.

But you will not mistake any acute hemorrhagic ovarian cysts

as endometriomas.

The likelihood ratio for endometrioma

with this diagnostic criterion set

is 32, which is quite high.

This is what you would get from a series,

and this is actual data if you use the criteria

of low level echoes with no neoplastic features

and not considering whether

or not they had hyperechoic wall foci.

So yes, most endometriomas,

in fact probably all endometriomas,

have low level echoes with no neoplastic.

I mean no low level echoes, sorry, not absence

of neoplastic features, but low level echoes.

But unfortunately many acute hemorrhagic cysts also have

diffuse low level echoes.

Now, I want to be sure that you don't mistake

the bright foci

that I showed you on those other cases with something like this.

This yes has foci of increased echogenicity,

but they're not clearly on the wall.

They're on the edge of this mass.

And although there are low level echoes in this mass,

this proved to be a borderline cyst adenocarcinoma.

And these are small wall nodules.

These two cases actually were not cancers,

these are endometriomas,

but you have to worry a bit about the possibility

of neoplasm based on the

ultrasound imaging without doppler consideration.

So here, this focal wall thickening should be viewed

with some concern for the possibility of neoplasm.

Now, the fact that we didn't see any blood flow in here may

predict that this isn't a neoplasm,

but clearly this hasn't been fully tested.

And when we start implementing contrast enhancement

or we use MRI, we would have more confidence

that this doesn't have any blood flow in it.

In fact is just clot adherent on the wall

of this endometrioma, whereas this nodule in this mass

has blood flow, but actually this also proved

to be an endometrioma

and this mass, which we don't see much blood flow in,

if any blood flow in this mass,

but it has convex margins, can't be assumed

to be an endometrioma.

And in fact, this was another borderline cystic cancer.

There is a small risk

of future cancer in endometriomas if you're watching these

endometriomas, if you're not taking them out,

there's a 1% risk

of cancer developing in an endometrioma over time.

Typically, these masses are over six centimeters in size

and in fact most of them are quite large over

nine centimeters in size.

And typically the patient is older.

But this is why the SRU recommended

that we follow endometriomas if we don't take them out.

So the recommendations that one might put together are

that you would get short interval follow up prior

to surgical intervention.

If you're going to intervene on

what you thought was an endometrioma,

you probably want to get a short interval follow up prior

to surgically intervening if you don't see

that hyperechoic wall foci.

So if you see low level echoes

but you don't see those hyperechoic wall foci, you need

to exclude the possibility

of acute hemorrhagic ovarian cyst.

And typically there's no clinical urgency to intervene in.

So, wait a few weeks and get the follow-up sonogram.

In those cases, if you're not going to do surgery,

if the patient isn't going to have surgery

because they're asymptomatic,

and you suspect an endometrioma,

a yearly follow up is probably indicated

because you'd like

to increase your confidence in the diagnosis

by showing

that this is not increasing in size

and also monitor for that small

but definite risk

of malignancy developing within the endometrioma.

So based on that, the SRU recommended

that the initial follow-up

for suspected endometriomas be at six to 12 weeks to ensure

that the mass is not an acute hemorrhagic ovarian cyst.

Assuming that you don't see those hyper wall foci,

which could probably obviate the need for the six

to 12 week follow up,

and then a yearly follow up if it's not surgically removed,

to ensure that you're not mistaking a benign neoplasm

as an endometrioma

or to be sure that there's not a malignancy that develops.

Cystic Teratomas (Dermoids)

Now let's turn our attention to cystic teratomas.

Again, if we can make a specific diagnosis,

we could manage these as expected for cystic teratomas.

Most of these will go to surgery,

but if the patient wasn't a good surgical candidate,

you could certainly elect not

to take out an asymptomatic teratoma.

There are certain features

that have been well characterized

and well associated with dermoids,

and these include acoustic impedance as we see here,

regional or diffuse bright echoes as we see here,

hyperechoic lines

and dots, the so-called dermoid mesh, which we see here.

And in this sagittal transvaginal image,

the fluid fluid level,

which has also been associated in the

literature with dermoids.

Now, one pitfall when you're evaluating dermoids

or when you're evaluating the pelvis

and the dermoid is there is you may go completely

unrecognized by the sonographer.

So you need to be somewhat aware of the possibility

of these dot dash appearance mimicking bowel.

So this longitudinal image of the uterus

did not recognize the fact

that there was this large mass in the cul-de-sac.

And in fact, on review of the images, we saw

that there was this area here.

Could this be bowel possibly,

but it had enough of the features of dermoid that we wanted

to go back in and make sure.

And in fact, this turned out

to be quite a large dermoid sitting in the cul-de-sac.

Here's another case of a dermoid

that has a very heterogeneous appearance.

Now, focal acoustic impedance is

an important thing to look for.

And so sometimes to help you find

that you want to look past the finding

and see if there is a swath of decreased echogenicity

that you can see in the tissues behind the mass

dermoids can be found in pre and postmenopausal women.

It's the most common ovarian neoplasm,

and it's the one that's most subject

to observer variability.

This is a very interesting article that showed

that even the same observer

looking at a dermoid at two points in time may come

to different conclusions about the mass

here in this particular study where a hundred

and I believe 10 different types of lesions were evaluated

by reviewers at six month intervals,

meaning that

the same set of images were looked at six months later,

the observers did not agree with themselves for

how they would diagnose the mass when it was a dermoid

as much as you might expect

and inter observer agreement is even worse in the

diagnosis, experts do

pretty well in evaluating dermoids.

So these are two early articles back in 99

and 95 using the criteria.

And these are both from Europe using the criteria

of an echogenic ball.

And this is the exact terminology

that was used in these papers in the corner of cyst

or filling up the whole mass

or the presence of echogenic lines and dots

or the presence of shadowing.

These investigators showed that the likelihood ratio

using these criteria was quite high

for making the diagnosis of dermoid.

Yes, there were some false positives,

but this is a pretty good likelihood ratio

for the diagnosis.

But it's important to remember that all of those studies

were performed by the physicians interpreting the

examination even in Europe

where it is more common for patients to be scanned

by physicians and then to be interpreted, those scans

to be interpreted by the same physicians

who scan the patients.

If you just look at images,

the experts don't do as well.

And here the likelihood ratio for the diagnosis

of dermoid varied between seven to 12 using those criteria

even in the most expert ologists

that were looking at the cases.

Well, how do you make mistakes when you're looking

and thinking that you're dealing with a dermoid?

And here's an example from the paper

that Gary Ero wrote in 2009 that gives us an insight into

how mistakes are made.

All of the operators in Garros paper showed

or suggested that this was going to be a dermoid

because of the increased echogenicity in what was suspected

to be this part of the mass.

And I think that you can make a mistake when you diagnose a

dermoid but relying too much on the

echogenicity of the mass.

So we looked at a different pattern approach

and said that if you saw multiple patterns of dermoids,

you were in far stronger ground for making that diagnosis.

So what do we mean by that?

Well, clearly increased echogenicity can be

a feature of dermoids.

When we looked at this of the 74 dermoids in the series,

we looked at a good 60%.

Almost 60% of them had increased echogenicity.

So this transabdominal in this endovaginal image

of two different dermoids show a mass

that has increased echogenicity

and here increased echogenicity, which is defined basically

as echogenicity in the mass that matches that

of the surrounding fat.

But if that's the only feature of the dermoid,

you can make a mistake.

And in fact, this is a dermoid,

but this is not a dermoid, this is an endometrioma.

So not everything with bright echoes is a dermoid hemorrhage

can also give you bright echoes.

What you really want to look

for is multiple features of dermoid.

So, as we showed in that earlier image reproduced here,

if we can see more than one feature as occurs in 75%

of dermoids, then we'll be far more confident

that we're dealing with the dermoid.

In fact, in this paper that we wrote many years ago,

when any combination of two features

or more we're seeing there were no false positives in our

series, giving us a likelihood ratio for the diagnosis

of dermoid that exceeds 132 using the

criteria I have described.

So you want to look for multiple features of dermoid, look

for acoustic impedance, look for increased echogenicity,

look for the dermoid mesh.

It turns out that the fluid, fluid level

by itself is not a very good feature,

but in combination with those other features can be helpful.

And if you see more than one feature of a dermoid,

then you can be very confident that you're dealing

with a cystic teratoma.

You also do need to check for internal blood flow

because if it's present,

the diagnosis is not a benign cystic teratoma.

You would have to consider something else such

as a fatty exophytic, uterine mass, like a lipoleiomyoma

or leiomyoma or potentially a malignant teratoma

or struma ovarii.

And in these instances, if you can't make one

of these diagnoses, but you see blood flow within

what looks like a dermoid, you can't call that a dermoid

and you should consider an MRI.

So for example, in this case,

this looks like a dermoid based on the images.

This was an adnexal mass in the right adnexa,

it has increased echogenicity in one area.

It definitely has acoustic impedance, but

before we called it a dermoid, we put color flow on it

and in fact, this has blood flow in it,

so it can't be a cystic teratoma.

This could have been a number of things

and it turned out to be an exophytic lipoma.

Now, the growth rate of dermoids is slow.

One study suggests

that it can differ based on whether you're pre

or postmenopausal patients,

but the vast majority

of those dermoids grew only a few millimeters a year.

The other thing to remember is that dermoids can

transform into malignancies,

especially when the dermoid is large.

This doesn't occur very often. The rate is around 1%.

If you average three studies,

these dermoids were all over 10 centimeters in size in one

study, and the mean was 14 centimeters in

size in another study.

So, these are not usually small masses

based on this evidence.

The SRU recommended that if the ultrasound was suggestive

of a dermoid,

but not pathognomonic, you should probably confirm that

with an MRI or an imaging.

If it is a dermoid in your mind on the ultrasound,

you might consider an initial follow-up in six

to 12 months just to be sure

that you haven't made a diagnostic error after that.

A yearly follow-up might be indicated if you don't take out

the dermoid just to monitor the growth over time

because there is a small risk of malignant transformation.

Hydrosalpinx and Peritoneal Inclusion Cysts

Okay, let's turn our attention then to hydrosalpinx

and peritoneal inclusion cyst.

Two things that occur outside the ovary.

If we can find specific diagnoses

or specifically findings that suggest these diagnoses,

then we may not need to do any surgery at all.

There are some features that suggest hydrosalpinx,

a mass that's cystic

and tubular that's separate from the ovary,

with incomplete septations or short linear projections.

There has been a picture

that's been called beads on a string, which I'll show you,

which are small round projections

representing the endometrial folds along the wall

of the fallopian tube.

And the waist sign is another important feature

that can suggest that you're dealing with a hydrosalpinx.

So here's a hydrosalpinx that has a number of these features.

It has an incomplete septation as we see here.

It has the waist sign, which is

where you have an indentation along the wall of the mass

and you go completely 90 degrees opposite to that

and you see another indentation.

So this is a hydrosalpinx.

Now, by itself, some

of these findings are more useful than others.

So the likelihood ratio for incomplete septation is only two

because some cystadenomas will have incomplete

septations, as in these two cases.

Similarly, the presence of small linear projections

or small round projections as we see here by itself

only has a pretty low likelihood ratio,

whereas tubular shape

and the waist sign are much more useful

as isolated findings.

But the real advantage comes when you

combine observations.

So if it's a tubular cystic collection

and has an incomplete septation,

it's about 7.4,

in terms of the likelihood ratio.

But more importantly, if it's tubular in shape

and has the waist sign,

it was always a hydrosalpinx in our experience.

So you really want to look for multiple features.

So if you're presented a case like this

where you see some incomplete septations,

but you're not seeing a tubular shape,

you're not really seeing other features that might suspect,

you'd think that this is a hydrosalpinx.

So that there is a little bit of a waist sign here

as we see here, the best thing to do is

to watch a ciné clip or watch the sonographer scan

because that can be very useful in showing you

that this is a folded tube that is a hydrosalpinx,

and if it looks like a hydrosalpinx

and you're confident of that,

you could ignore it if the patient is asymptomatic.

Now, peritoneal inclusion cyst can be very confusing,

especially to those who haven't seen a lot of these.

But once you get the hang of it, these can be very,

not easy to diagnose,

but can be very nice to diagnose

because what happens is you take the situation where the

patient is suspected

to have a malignancy based on the appearance

and with your experience you're able to say, no,

this isn't a malignancy.

We're just dealing with a peritoneal inclusion cyst.

So what are we looking for? Well,

peritoneal inclusion cysts typically have angular margins,

especially as they abut what looks like to be solid tissue.

Then if you look more carefully at the solid tissue, look

for the presence of follicles as we see here,

and the presence of strands

that come off from the tissue at right angles,

the spider webbing of the peritoneal inclusion cyst.

Here's another example of a cyst with, at first glance,

you might have thought a solid area,

but in fact this angular margin, okay, is this solid?

Well, it has small follicles.

These features should make us think

of a peritoneal inclusion cyst.

And if it looks like a peritoneal inclusion cyst

and you're confident of that, you can then ignore it.

Multilocular Cystic Masses and Neoplasms

Turning our attention to masses

that have other features.

So here's a cystic mass that doesn't have angular margins,

doesn't have the strands, the solid area

or the, there's not really much of a solid area,

but there's certainly septations here.

This has no features of hemorrhage.

When we encounter a mass like this, a multilocular

cystic mass with septations, albeit rather uniform

and thin, we need to think about a neoplasm.

And especially in the setting

where there are septal nodules and wall nodules.

Clearly here on these two different examples, this has

to be viewed with a very high suspicion for malignancy.

Again, this is that model to predict malignancy

and it's a little bit silly

to actually even show you these numbers.

It's crazy, right?

But I'm showing you two mainly to show you the inputs

to the data because if we understand the inputs to the data,

we can see how important ultrasound is in

making this diagnosis.

The full model shows you

that there are only four clinical parameters

that have any influence on whether a mass is going

to be a cancer or not.

The other eight parameters are all imaging parameters,

the diameter of the lesion, the presence

of ascites blood flow in the solid area,

whether it's purely solid, whether

the solid area has a sizable component, the irregularity

of the cyst wall, if there's acoustic shadowing,

that is actually a negative predictor

for malignancy and the color score.

And even in the simple model, if you even simplify it

and use the simple model mathematical model,

which we're not doing, of course you can see

that the clinical parameter, the only clinical parameter

that really makes a difference is how old a patient is.

Everything else is an imaging feature.

So it's not appropriate for you as an imager to say,

clinical correlation necessary to review whether this mass

that I'm seeing is a malignancy or not.

That's not a very useful thing

because you can make the clinical correlation

because the patient age is the really,

the only thing that matters.

CA 125 doesn't help you. It's not of additional value.

In fact, expert subjective impression performs about as well

as the math model, and that's important to realize.

So when you're trying to make that subjective impression,

certainly take into consideration the patient's age.

Look for ascites

and look to see whether the mass has irregular walls,

thick septations, papillary projections, solid areas,

the degree of vascularity.

These are the things that will allow you

to predict malignancy.

And if it has those things, you want to make the diagnosis

of malignancy and send the patient to surgical evaluation.

Conclusion

So in conclusion then this is

the full spectrum of our algorithm.

We're going

to first see whether the mass is a simple cyst or not.

If it is, we know what to do.

If it's not, we know how to analyze it

to see whether it might be due to physiology

that it is causing us to be confused.

If that's not the case,

there are specific things we want to look for.

We will then consider the possibility of MRI

to help us in the diagnosis if we haven't found specific

things that allow us to make the diagnosis.

And in the end, some patients will have

to have a surgical diagnostic procedure performed,

but the vast majority are going to be patients where you know

what to do based on your sonographic evaluation.