Salivary Glands - HD

Introduction

Thank you for your attention this afternoon,
and I will be speaking on salivary gland ultrasound.
I have no disclosures.
I'll begin with this somewhat problematic case.
This was a 48-year-old healthy woman who came in
for a lump under her chin,
which ended up being a very unexciting lymph node,
but had this incidental finding in her right parotid,
just under a centimeter.
Looks almost cystic, very well circumscribed,
but we put color on. It had internal blood flow,
and therefore it was presumed to be solid.
It was biopsied,
and very unfortunately for her,
this was adenoid cystic carcinoma, which as we'll describe later has a fairly bad prognosis.
And this highlights some of the challenges of salivary gland ultrasound.

Anatomy Review

Before I get into more of the sonographic findings,
I'm gonna review a little bit of anatomy.
Parotid gland anatomy is tough
because there's a lot of stuff that is near it
and runs through it, particularly the facial nerve,
which goes through the parotid gland
and divides the gland anatomically
and surgically into superficial and deep portions.
We usually cannot see the facial nerve,
but we can often see the retro mandibular vein,
and that's a good imaging marker for the facial nerve
because the nerve is just lateral
to the retro mandibular vein.
In most cases, the
external carotid artery goes into the gland,
terminates in the gland, giving its final branches,
including the internal maxillary.
The distal ICA is not in the gland,
but it is in the neighborhood.
It's just deep to it. The parotid gland is drained
by Stensen duct, which comes anteriorly,
to drain into the side of the mouth.
And most parotid glands contain lymph nodes,
on average, about 10 lymph nodes per gland,
and almost all of which are in the superficial portion.
And here's just an example
of a normal intraparotid lymph node.

Some pictures to illustrate this anatomy on CT.
Here is the retro mandibular vein on a
corresponding ultrasound.
This is the parotid tissue.
Here is the mandible, the masseter,
and the retro mandibular vein.
The external carotid artery
can often be seen right next
to the retro mandibular vein here in this Doppler image.
This is the retro mandibular vein heading inferiorly,
and just deep to it, the distal aspect
of the external carotid artery in the neighborhood.
Just deep to that, the internal carotid artery,
the internal jugular vein,
and actually fairly close in some patients,
the vertebral artery, which
can actually come into play with some biopsies.
I just have to watch for that. The facial nerve,
the dreaded thing for our biopsies, we can't
usually see it with ultrasound occasionally,
if MR works exceedingly well, we can see it,
just
below the stylomastoid foramen.
And here is where it is coming into the parotid.
And this is the rough course that it takes.
Of course, this is a three dimensional structure, so it is
arborizing, the pes anserinus, right, the pes as it goes
through the parotid towards the masseter muscle.
So it goes up to down quite a bit.
But this is the rough course
that it takes the plane of the vein so superficial,
this superficial parotid deep to this, the deep parotid.

Now, I said, we can't see this on ultrasound.
That's most of the time. This is an unfortunate case
of a relatively young man who had a horribly aggressive
squamous cell carcinoma of the midface that presented
with neurological symptoms.
And all of this linear hypo echogenicity is branches of the facial nerve,
coming anteriorly towards the masseter.
And the retro mandibular vein can also be used to sort out, if you're not sure if something superficial
or deep, this pleomorphic adenoma is superficial
to the retro mandibular vein, therefore it's superficial.
And another example of a pleomorphic adenoma,
this time deep to the retromandibular vein.
So in the deep portion.

Submandibular Gland Anatomy

Submandibular gland anatomy is a little bit simpler.
It is the submandibular gland, as you know, is in the,
level one B submandibular fossa.
And it is between the mandible and the mylohyoid muscle.
It is drained by Wharton's duct, which heads posteriorly,
and then laterally to drain right next
to the sublingual gland.
This is anti-gravity drainage, assuming that you're upright,
which is important probably in stone formation.
There are a lot of lymph nodes near the submandibular gland,
but usually not in the submandibular gland.

Inflammation

One of the things we see with imaging
that I'm not gonna cover in great detail,
but we certainly see is inflammation.
And that can be divided into acute or chronic.
Acute can often be viral.
It can be bilateral, can bacterial certainly occurs,
can be unilateral, such as in this case,
where you see the left parotid is expanded, hypervascular,
heterogeneous as compared with the right parotid.
If this gets really bad, it can
end up resulting in an abscess such
as this intraparotid abscess.
And with abscess in almost any part, superficial,
a little bit of gentle pressure
and release of pressure can help to confirm
that there's mobile debris inside of it.

Chronic inflammation can be obstructive,
fairly commonly due to stones
or non-obstructive due to a lot of causes,
including autoimmune disease.

Sialolithiasis

Stone disease.
The vast majority
of this occurs in the submandibular gland or in Wharton's duct.
One of the reasons is likely the anti-gravity drainage
of Wharton's duct, less common in the parotid.
These stones are often not seen on conventional
radiographs can usually be seen on CT.
It is important to determine if it's in the gland
or at the margin, or just outside the gland.
And sometimes,
stones in Wharton's duct can be confused
for material debris in the floor of the mouth.
Here is just an image scanning the stone here.
I thought it was probably a stone,
but I did what's been suggested
and just put a glove finger into the mouth to just confirm
that that is a stone with gentle manipulation.
And actually, I got the stone right out,
which wasn't what I was expecting.
The patient was extremely happy.
And here's what it looked like.
After the stone is removed,
the submandibular gland was still inflamed,
with some dilated ducts,
but the patient actually felt better almost immediately.
He was actually referred for a biopsy
of the submandibular mass, which obviously we didn't do.

Sjogren's Syndrome

Sjogren's syndrome.
Most common autoimmune condition you're going to see,
in the salivary glands involves the major
and the minor salivary glands.
It can be primary or secondary.
These are the autoantibodies that are often seen as well
as rheumatoid factor, and you get a heterogeneous
sonographic appearance of the salivary glands,
with areas of hypoechoic foci that can be rounded
with some intermixed hyperechoic linear bands.
To some extent, it kind of looks like Hashimoto's
of the thyroid, although it may not be nearly as diffuse,
it may be a little bit more patchy.
And they're actually, if you are in a situation
where you have a lot of rheumatology there also,
there actually are some staging schemes
that have been proposed
and are in clinical use for both diagnosis
and monitoring of therapy in Sjogren's syndrome.

One thing important to know about Sjogren's syndrome is
that these patients do have a risk
of developing lymphoma.
Often a MALT type lymphoma, five to 10% of patients
with Sjogren's may develop lymphoma within 15 years of diagnosis.
Here's a case of bilateral lymphoma.
So any confluent, large mass in the setting of Sjogren's,
should be biopsied with testing
as appropriate for lymphoma.

Cysts and HIV-Related Lesions

You can get simple cysts in the salivary glands,
particularly the parotid,
and you're never gonna biopsy something
like this in the thyroid, in the parotid glands, not uncommonly.
We will aspirate these just to make sure
that it is a simple cyst.
If there's no solid component,
you can completely aspirate it,
and there are no worrisome cells in the aspirate.
You can say this is a benign lesion,
even if the cytopathology is technically non-diagnostic,
but the radiology report saying it was completely
aspirated with no solid component is necessary for that.

In HIV disease.
Occasionally you can see this entity,
multiple benign lymphoepithelial lesions,
within the parotid glands.

Salivary Gland Pathology

Salivary gland pathology, several
good factoids here.
70% of things that clinically present
as masses in the parotid are neoplasms.
The risk of malignancy
for a salivary neoplasm is inversely proportional
to the size of the gland.
20 to 25% of parotid neoplasms are malignant.
50% of submandibular gland neoplasms are malignant,
and 75%
of the rare sublingual gland neoplasms are malignant.
70% of salivary malignancies,
however, occur in the parotid, and that is
because parotid tumors are way more common than tumors
in the smaller salivary glands, 80 90%
of parotid tumors arise in the superficial lobe.
The WHO classification includes,
and I'm not kidding here, 28 different histologic types
of primary salivary malignancy in the salivary glands,
it's basically four, maybe five,
and many of the salivary primary tumors have low,
intermediate and high grade variants.
It has been described as the most heterogeneous of any group
of cancers in the human body,
which makes a real pathologic challenge, particularly
with fine needle aspiration.
And there is overall five-year survival
with primary salivary cancer.
There also are at least a dozen primary benign lesions
of the salivary glands.
I'm going to just go through the most common primary salivary benign and malignant lesions.

Pleomorphic Adenoma

The most common benign lesion,
actually the most common lesion in general,
in the salivary glands is the pleomorphic adenoma.
It is usually solitary.
80% occur in superficial parotid,
10% in the deep parotid,
and then the rest in the other salivary glands.
It is composed of epithelial myoepithelial
and mesenchymal tissue.
Thus, the pleomorphic adenoma
typically hypoechoic solid mass with lobulated borders.
That's kind of classic, may have calcifications,
occasionally, some cystic change, often not that vascular.
What is important to know, though,
is this is potentially a pre-malignant lesion,
with reported up to 5% chance of going, converting over
to the dark side and becoming a carcinoma ex pleomorphic,
which has a fairly grim 40% five year survival.
And here is an example of a carcinoma ex pleomorphic.
It is solid and invasive, some calcification, a little bit
of internal cystic change.
It's pretty bizarre on CT
and a lot of heterogeneous low signal on T2,
MR concerning, but certainly not specific.

Warthin's Tumor

The other common benign tumor is a Warthin's tumor.
The old term for that is cystadenolymphoma,
which I think has largely been abandoned
because it sounds like it's a malignancy and it's not.
It's benign. These almost always occur in the parotid.
So I do harass my residents when they include it in a
differential for a submandibular mass,
because it probably shouldn't be in that differential.
It also has an extremely strong association with smoking.
Does not have to be active smoking,
but usually occurs in people who either are
or were committed smokers.
It contains epithelial and lymphatic tissue.
It may have cystic components.
They tend to be fairly vascular, but not always,
and this is a benign lesion.
Malignant transformation is rare if it ever occurs.
When I pulled our pathologist at OU, they feel
that this does not have a malignant potential.

Mucoepidermoid Carcinoma

The most common malignancy in the salivary glands is the
mucoepidermoid carcinoma.
It is composed of mucus secreting,
intermediate, and epidermoid,
or squamous cells, thus the mucoepidermoid carcinoma.
This has low intermediate
and high grade histology variance,
which have very discordant
clinical behavior from a fairly indolent, 75
to 89% five year survival
to a less than 50% five year survival for the high grade.
They may have cystic components
such as this one and this one.
The high grade tumors often are very ill-defined
infiltrative, but the low grade in small tumors can be very
well-defined and honestly overlap with benign lesions.

Adenoid Cystic Carcinoma

The second most common malignancy
of the salivary glands is the adenoid cystic carcinoma.
It is considered an intermediate grade tumor.
It is indolent, yet often highly fatal.
That may seem confusing what that means,
it is a slow moving tumor, but it is a problem.
And unfortunately this case in that young woman,
hopefully she will be one of the few that is cured.
But it has a 35 to 70% five year survival,
but only a 10 to 20% 10 year disease free survival,
and no better than a 50% 20 year survival.
Late distant metastases
and local recurrence with perineural invasion,
unfortunately is common,
has a variable appearance on imaging.
It's been reported in the literature as having a lot
of color Doppler signal.
Honestly, I have not found that to be consistent in the cases I have seen.

Acinic Cell Carcinoma

Potentially the third most common salivary malignancy,
primary salivary malignancy is an acinic cell carcinoma.
It's a low grade malignancy.
It's actually previously was regarded as a benign lesion.
It has an unpredictable behavior,
but overall, a very good five year survival.

Metastases to the Salivary Glands

The other thing to consider is that metastases
to the parotid are fairly common,
because there are lymph nodes in the parotid,
and those parotid lymph nodes are the primary drainage
for a lot of the skin of the midface,
the ear, a lot of the scalp.
And so skin cancer is metastatic to the parotid,
at least in our head and neck practice at OU are really common.
Almost every week at tumor board,
we discuss at least one.
And here are some examples.
Squamous cell carcinoma, basal cell carcinoma, melanoma,
and Merkel cell carcinoma.
So all of the major skin cancers
that are seen predominantly in an elderly population.
You can also occasionally get distant metastases
to the parotid.
Here is an example of a very hypervascular solid lesion,
which was a distant met from renal cell carcinoma.
This was interesting to biopsy.
And then you can see it also from lung cancer,
such as this case, breast cancer, and others.

Can Ultrasound Determine if a Parotid Mass is Benign or Malignant?

A very important question in my opinion.
Can ultrasound determine if a parotid mass
is benign or malignant?
You've heard earlier today some really great talks
about risk stratifying thyroid nodules.
Does any of that apply to the salivary glands, specifically
to the parotid since that's
where most of these tumors occur?
Well, before I answer
the question, I'm gonna give you a try.
So I'm gonna show you a number of cases.
One is benign, one is malignant.
Nobody's really guessing.
Well, this is a pleomorphic adenoma.
This is a high grade adenoid cystic
carcinoma, very bad lesion.
The only difference, really,
and it's subtle, is that this one is wrapping around
the retro mandibular vein.
But otherwise, I think these look about identical.

Okay, so, that was hard. How about another one?
So this is mucoepidermoid carcinoma.
This is a pleomorphic adenoma. Well, how about this one?
Which lesion is malignant?
I think they look pretty similar,
except this one is more vascular and it's benign.
It's a Warthin's tumor. This was a high grade muco
epidermoid carcinoma.
And these two, they look, I think, almost identical.
They're honestly different patients.
This one's a little bit more blood flow.
It was a Warthin's tumor.
This was metastatic squamous cell carcinoma
and in an intraparotid lymph node.

Okay, one more. Some fairly circumscribed,
well-defined hypoechoic solid lesions.
This one may have a touch of calcification,
and it's a high grade salivary ductal carcinoma.
A very aggressive tumor,
that has some similarities with breast cancer.
This was a benign pleomorphic adenoma.

Okay, so I think you might be able to guess
where I stand on this question based on the
cases I've shown you.
The biggest article in the literature to
actually address this question was from 1989.
And they found 272 of 302 cases.
Parotid lesions were correctly identified as
malignant or benign.
But there's a couple of big buts about this.
One is that unsharp border was the only sign of malignancy,
and in this case, 28% of malignancies had a sharp border,
and they were called benign.
I don't think this is good,
because in my opinion,
it may be those benign looking malignancies
that have the greatest chance for cure of
what can be a very aggressive lesion.
And the other thing is,
if you call everything in the parotid benign,
you're gonna be right about 80% of the time.
More recent articles, including at least one that
looked at contrast ultrasound, have mostly concluded
that ultrasound cannot correctly classify lesions,
but they're limited by very small numbers
of malignant lesions,
because most parotid tumors are benign.

So what are malignant features?
An irregular or infiltrative margin
and encasement of the retro mandibular vein
or external carotid artery, local adenopathy,
which is usually level two or level three, but that's rare.
And encasement of the vein is rare as well.
Rapid growth, if you get that clinical history,
and here, this is probably the most important one,
and possibly a SAM question.
Neural symptoms, if somebody has pain
or even worse, a slowly developing ipsilateral facial droop,
that's really worrisome
because benign things usually don't take out the nerves.
Unfortunately, many malignancies have
none of these features.

Pitfalls in Parotid Mass Imaging

I'm gonna describe just a few other pitfalls
and parotid mass imaging.
One of them is that the parotid is a wonderful,
is kinda like the kidney produces wonderful opportunities
to make stuff up that does not exist.
And here is an example.
This kinda looks like a real lesion.
And as I've described, you know, we having to biopsy this.
However, when you really looked at it
and turned in an oblique plane, you saw this was continuous
with the adjacent fat as was shown on
real time cine imaging.
And it actually, these, the CT isn't great
because the artifact, but it's this little bit
of interdigitated adipose tissue,
'cause the parotid can kind
of have little fingers to it.
And so this is a common pitfall.
It's easy to fall into this.
We probably have this issue maybe
once a week in our lab.
We do a lot of head and neck ultrasound.
But it's a common issue.

Another really problematic issue is not finding stuff deep
enough in the parotid.
This was a relatively young woman, had an MR
for a headache, didn't show anything in the brain,
but showed this bright T2 lesion
of the deep margin of the parotid.
Probably a pleomorphic adenoma,
but appropriate for biopsy.
She came for biopsy, and the initial sonographer
images, there's no lesion.
She found this benign looking lymph node, which corresponds
to this, but was not looking deep enough.
So using a lower frequency, looking deep enough knowing
where to look, we see this lesion, which was
biopsied and proven to be pleomorphic adenoma.

The last thing is failure to visualize the deep aspect
of a deep mass.
So this looks like a circumscribed lobulated,
solid hypoechoic mass.
Not specific, but good for a pleomorphic adenoma.
And it looks like you're seeing the whole thing.
I wanna thank Dr.
Tefy for this amazingly beautiful image of an iceberg.
And you know, here you're not seeing the whole thing.
Well, let's look at this on MR.
This is what we saw with ultrasound.
This is just a little extra bit of it
that might affect patient management.
And actually it was subtle on the PET CT.
We were really worried about a carcinoma
ex pleomorphic,
and the biopsy was suspicious as well,
when I performed it.
But fortunately,
for the patient ended up being just a very
cellular pleomorphic adenoma.
At excision. We're gonna watch it really closely
because we're still concerned about it.

So the moral of that story is that deep,
lobe parotid lesions, as much
as I love ultrasound, they need MR.
Or CT because you can miss something.
It's difficult to see everything,
and you don't want the patient
to be managed inappropriately.

Biopsy of Salivary Masses

So lastly, I'm just gonna briefly cover biopsy.
So why should we biopsy salivary masses?
Why not just cut 'em all out?
That's a great question.
And I've kind of talked
to some surgeons about this at times.
And one of the easy answers is that some
of the masses are not treated surgically, like Warthin's,
lymphoma, malignancies that are not resectable,
non neoplastic lesions and some metastases.
But even if surgical excision is planned
pre-op diagnosis surgeons will argue is still helpful.
Does this patient need surgery in one
or two weeks because they got a cancer?
Or can we do this over Christmas break when it's convenient for the patient
because it's a pleomorphic adenoma.
And it allows for more accurate counseling
of the patient regarding the prognosis
and the likelihood of facial nerve injury or sacrifice.
A generation ago surgical incisional biopsy was performed
that is now considered contraindicated not
to be performed in the surgery community due
to both the risk of nerve injury
and the high risk of tumor seeding.

So if we do a biopsy, do we FNA, do we core? Do we do both?
There are a lot of opinions.
This has been studied actually quite a bit.
And FNA has a long track record, great safety record.
There's a lot of evidence for core biopsy,
good safety record, and almost always has been shown
to work better than fine needle aspiration.
And some people have advocated that you shouldn't FNA,
these, you should just core them.
That is actually not our approach at OU.
Our approach is based on the fact
that we have some really good cytopathology support,
and we will almost always begin with fine needle aspiration.
After a little bit of lidocaine, we generally try
to traverse as little salivary tissue as possible.
And then based on a real time assessment
by the cytopathologist, in conjunction
with the radiologist being familiar
with the imaging in the history, we will decide,
do we need to core it?
Do we need to collect for flow cytometry cultures, et cetera.
If we core it, we use 18 gauge semi-automatic adjustable
throw side cut needles.
Here are a couple of manufacturers that we use.
We do not make an incision.
Almost always, it's single needle technique,
very rarely coaxial.
And what we do is we get in the lesion
and then we manually deploy the whole cutting area,
make sure it is completely in the lesion
before firing the needle and cutting.
And here's a movie of doing it.
You can see getting in the lesion deploying. It's good.
So go ahead and cut.
And these were both pleomorphic adenomas.
If it's a mixed cystic solid, just like in the thyroid,
you have to target the solid component.
I did an FNA first
and the cytopathologist said, you got a few atypical cells.
This is not gonna be diagnostic.
And so took a core biopsy of it,
diagnosed the mucoepidermoid carcinoma.

When we reviewed our experience
with this FNA first core if needed,
that we recently published, over a three year period,
135 procedures, we only cored about 25% of patients.
And we had a 93% diagnostic pathology.
And as of yet, have not identified any missed cancers in
that group and had no significant complications.
But again, I would say that I am blessed to work
with some excellent cytopathologists,
and this would only work if you have complete buy-in
of excellent cytopathologists.

Deep Lobe Parotid Lesions

Briefly, deep lobe parotid lesions.
These can result in a little bit of adrenergic
and sphincter hyperactivity.
When they come for biopsy,
fortunately, they're uncommon.
I think these I like for them to go
to our multidisciplinary head and neck tumor board
before I biopsy them if possible.
And they need MR
or CT imaging to make sure that we're not missing something
that is, we're not getting fooled.
But there are three approaches to deep lobe parotid.
If you have to biopsy them, one of them is to just say,
whatever, I'm just gonna go right after it.
Go right through the parotid knowing that,
unless you're really posterior,
you will cross the plane of the nerve.
Even if you do that, the safety profile is very good.
You're gonna make sure if you have to core it,
your core is completely in the mass
not getting parotid on the way in.
And the thought is, if there is nerve in the mass,
the nerve is hosed anyway.
Basically is the thinking there.
I don't like doing this if I can avoid it.
But sometimes it's unavoidable.
If it is a very posterior
or inferior lesion,
you can do a posterior inferior approach.
This is, you have to turn the patient.
This is way posterior inferior going through the
very origin of the sternocleidomastoid.
And you can often get into something in this location
without traversing any parotid
and not crossing the plane of the nerve.
I will say though, the vertebral artery is
not far away from this.
So if I'm doing this approach, I find it
and make sure I'm not gonna stick it.

The final approach for a more anterior superior lesion
is, again, I love ultrasound,
but I will occasionally switch over to a CT biopsy
with a transfacial approach.
This is, this looks crazy,
but I just snow them with plenty of conscious sedation.
This is actually very well tolerated.
Just using a skinny needle, FNA.

Take Home Points

Take home points,
sialolithiasis occurs mainly in the submandibular gland.
Lymph nodes are a normal finding in the parotid.
Evaluate them just like you would any other cervical lymph
node masses with an infiltrative border are frequently
but not always malignant.
Some infectious
and inflammatory conditions can look infiltrative.
There is a substantial overlap in the ultrasound appearance
of benign and malignant salivary masses.
Masses in the deep lobe of the parotid need MR.
Or CT imaging and ultrasound guided core biopsy
of salivary masses is safe
and improves diagnostic yield and sensitivity.
Although if you have great cytopathology support,
it is reasonable to begin with FNA.

So thank you for your attention.