Imaging of Placenta Accreta - HD

Introduction

Hello, my name is Anne Kennedy.

I'm a professor of radiology at the University of Utah Hospitals and Clinics and I'm going to share with you some of our experiences in imaging placenta accreta.

Today we're gonna talk about imaging of placenta accreta and I would like to acknowledge my co-author Dr. Bowman, who is a maternal fetal medicine fellow, currently working with us at the University of Utah.

And he has provided the intraoperative photos that illustrate why we see what we see.

Objectives

The objectives of this talk are to review the classic imaging findings of placenta accreta spectrum to think about how we may be seduced by findings and make pitfalls in interpretation.

And although the focus of the talk is on ultrasound, I will show some MRI images for correlation and discuss the role of how MRI may contribute to the diagnosis.

Nomenclature and Pathology

If you think about the nomenclature of placenta accreta, we talk about accreta, increta and percreta and these are terms that are defined by the pathologists as they look at samples under the microscope.

In placenta accreta, the villi of the placenta adhere to the myometrium.

In increta they actually invade the myometrium and in percreta placental villi are seen through the myometrium and beyond that out into the uterine serosa or adjacent structures.

And obviously the imaging findings are not going to be as discreetly divided as the pathology findings.

Therefore, it's thought better to use the term morbidly adherent placenta as an umbrella term for all of these.

The amount of placenta that's involved can also vary from being all of the placental surface to a small part or to a very focal amount.

And while again it's useful to think about what actually goes on in reality in the clinical situation, any morbidly adherent placenta can cause catastrophic bleeding for the mother at the time of delivery.

Clinical Picture

So this is a clinical picture.

This is a cesarean hysterectomy.

This is the maternal head end.

This is the foot end, this is the incision through which the baby was delivered and the uterus has been sewn closed.

And this is the lower uterine segment and you can see this prominent bulge with extremely large vessels seen on the uterine serosa.

And what happens is you have lack of the formation of a normal decidua basalis, therefore the placenta can extend into the myometrium.

And then at the time of delivery, the placenta cannot separate normally.

And because at term 25% of your cardiac output goes to the uterus, there is a lot of blood volume going there and there can be catastrophic hemorrhage with maternal demise.

This is a sample obtained in the what they call the back table in the operating room.

And this is after cesarean hysterectomy.

These are actually Dr. Bowman's fingers and you can see that here.

This is the uterine serosa and lower uterine segment that's being retracted back toward the cervix, which is down here.

This is the fundal end of the uterus way over here.

And then this is actually placental tissue.

So you can see that you have placental villi going through adherent to the uterine serosa and obviously this placenta is not going to fall away and be delivered like a normal one, which is why this patient was scheduled to have a cesarean hysterectomy.

Issues with Placenta Accreta

So what are the issues with placenta accreta?

Obviously we have morbidity and mortality both for mom and baby oftentimes because this is associated with placenta previa, there's bleeding which results in a premature delivery for baby and the hemorrhage can result in disseminated intravascular coagulation.

Just the sheer volume of blood loss can cause systemic shock and there are many, many documented cases of maternal death.

Gold Standard for Diagnosis

When we think about conditions, we always refer to what is the gold standard for diagnosis.

And in our own experience, it was interesting that we had some cases read by pathologists who did not have particular expertise in placental pathology.

And when these were reviewed by an expert placental pathologist, there were cases of placenta accreta and increta that had been missed by a pathologist without specific training.

Management

And then you have to think about how you're going to manage these patients.

And in my own institution we counsel patients with strong suspicion for accreta that they should have an elective cesarean hysterectomy which provides a controlled delivery.

The baby is delivered by the obstetric department and then the hysterectomy is performed by one of the obstetric members in association with one of the gynecologic oncologic surgeons.

Risk Factors

What are the risk factors for placenta accreta?

The single biggest risk factor is previous instrumentation and in the modern world, a lot of the time, that is a previous cesarean section and the number in brackets refers to the odds ratio for having accreta.

Many patients who have infertility issues may end up with having a myomectomy on the rationale that a fibroid can prevent normal pregnancy implantation.

The fibroid is removed, the patient successfully becomes pregnant, but then there's the concern that the site of the fibroid will be an area of denuded decidua and accreta can occur at that site.

Also, many patients will have had a D&C at some point in the past for a previous miscarriage.

Placenta previa in and of itself is associated with the accreta spectrum and if you have a combination of placenta previa and a prior cesarean section, the numbers increase or the risk increases.

Another risk factor is thought to be advanced maternal age, but it is confusing as part of the impact of advanced maternal age may in fact be multiparity.

And so the relative risk for that is 3.2, but bear in mind that it may not be the mother's age.

It may be more the fact that she has had more than one pregnancy.

Imaging Evaluation

So when we see a patient with a possible placenta previa, placenta accreta, what are we looking for?

All obstetric studies include documentation of the location of the placenta and this is a transvaginal ultrasound.

This is a small transducer footprint.

This is the bladder, this is myometrium, this is placenta and you can see that there is placental tissue abutting the posterior bladder wall.

And here's the same case with color doppler applied and you can see that there is some flow in the wall but some interruptions or gaps in the myometrial blood flow.

And these are some of the signs that are reported.

People refer to loss of the retroplacental hypoechoic zone, which is also called loss of the retroplacental clear space.

So here is the clear space and as you come down here, you no longer see that clear space.

The myometrium itself may be thinned, you may have an irregular bladder interface.

If you have direct invasion of placenta into the bladder, you may see placenta extending beyond the uterine serosa at any site.

And as I mentioned, you can have abnormal color doppler findings on MRI.

We see the same stuff but it's called different things.

So we refer to dark intraplacental bands and this is a T2 weighted image which is the mainstay of anatomic imaging with MRI and the placenta usually comes out in this medium shade of gray.

So when you see these dark areas, these are concerning for abnormal vascularity.

You also get diffusely heterogeneous signal of the placenta.

You can have disruption of the myometrium and I'm going to showcase that, demonstrate that better later.

But you can see here this is the myometrium and it gets very thin here over the dome of the bladder, a bulge in the uterine contour or tenting of the bladder have been associated.

And obviously direct invasion of adjacent structures is a sign of placenta percreta.

Color Doppler Findings

Abnormalities on color doppler finding have been described relatively recently and this is a reference that interested readers can consult.

But one of the things is that you can get disruption of the normal flow in the myometrium and that you may also see vessels bridging from the placenta to the adjacent tissues at site of interface disruption.

Now I would caution you that any single image may look as though there is disruption in blood flow and you have to evaluate this in real time here again as an example where on gray scale you can see that the placental tissue seems to be traversing into the myometrium and here is a bridging vessel right along that site of disruption.

And here's another area where we have a potential area of disruption with blood vessels going from the placenta into the myometrium.

The thing is we don't often look at blood flow of the placenta in patients in whom we are not worried about accreta spectrum.

So here's an example of an anterior placenta in a primigravida, no prior uterine instrumentation, lovely homogeneous gray scale signal of the placenta.

Here is anterior myometrium, here is the abdominal wall.

Here's the same patient with color doppler applied to document the site of the cord insertion into the placenta and we just happen to see the subplacental flow on this picture and you can see that it is homogeneously filled in with no gaps.

Here's an example of a patient who had three prior cesarean sections and a complete placenta previa.

You'll notice the prominent bulge in the lower uterine segment.

The fact that the placenta is inhomogeneous and has these multiple vessels, the bladder interface looks irregular and on color doppler there seems to be a lot of flow and even when the gain is adjusted down you can see that there are areas interruption apparently of the blood flow and areas of abnormal vessels extending from the placenta into the bladder wall.

So putting those side by side.

Here's gray scale of a normal anterior placenta.

Here's gray scale of placenta that we believe is morbidly adherent.

Here's the color with the normal subplacental flow.

Here's color with some interruptions and some branching vessels.

Here's a patient with a complete previa and a history of three prior cesarean sections.

Fortunately, most of the placental implantation is posterior in this case, but you can see that we have homogeneous appearance to the placenta.

We don't appear to have any irregularity of the bladder wall.

Contrast that with this case in which the patient had had one prior cesarean section.

And look at how thick and inhomogeneous this placenta is with all these big vessels and a lack of normal architectural features.

No definite decidua line where you can say the placenta ends.

There's myometrium, there's bladder.

We have this just inhomogeneous mass of placental tissue.

MRI Correlation

On MRI.

We don't do MR for placental location, but we do do MR for fetal anomalies.

And this is a fetus with a diaphragmatic hernia.

This is a posterior placenta.

You can see it's homogeneously gray and this darker gray tissue is the posterior myometrium and there's a clear delineation between the placenta and the myometrium.

Contrast this with the study that we did to evaluate morbidly adherent placenta.

And yes, the placenta is present, it looks thicker, it looks more inhomogeneous.

This is myometrial signal.

This is myometrial signal.

And in this area with the red arrow, this is the anterior abdominal wall musculature of mother, there's the myometrial signal ending and there's a gap right here where placenta has invaded through.

Potential Pitfalls

So what are the potential pitfalls that you can run into?

I did mention earlier that we don't usually look at the placental implantation site and do color doppler and evaluate blood flow because we're not concerned about it.

So very few of us have a good visual database of what's normal.

We also have the ability to press quite hard with the transducer.

The harder you press, oftentimes the better picture you get.

But if you press hard, you will obliterate low pressure vascular spaces.

There are some normal structures to be aware of in pregnancy and some physiologic changes and you also need to think about your angle of insonation and your gain settings and of course observer bias.

So here's some pictures taken with different transducers.

This is a vector four and this is a curved six.

Now on this study, this is a patient who has a placenta previa but no prior instrumentation.

And you can see this is the bladder serosa and this is the myometrium, but I don't see a subplacental hypoechoic zone here and it could be that we just have excess transducer pressure but I don't see it.

We weren't concerned in this case, we didn't mention it and the patient delivered a term, a normal delivery.

Over here on the curved six you have improved resolution and here you can see myometrium and now you see placenta and you don't see myometrium deep to it.

You can also see that the texture of this placenta is abnormal with big vascular spaces coming over here.

Now using a nine megahertz linear transducer, which obviously you could only do in a slim patient, you have exquisite detail.

This is placental tissue and you can see these little pseudopods as it were of placental tissue extending out into the subplacental hypoechoic zone, which is this vascular space.

And again an area here represented tissue is extending posteriorly.

The gold standard in my mind is transvaginal ultrasound and this is an eight megahertz image.

Now transvaginal ultrasound is also gonna be dependent on your choice of megahertz and sometimes you have to turn down the megahertz to get adequate depth penetration, but because you're close to the area of interest, you tend to get really exquisite images.

So this is a vaginal ultrasound, this is cervix tissue, this is the bladder neck, this is the bladder serosa and bladder mucosa.

This is amniotic fluid presenting part of the fetus.

And this is very abnormal placenta with big vascular spaces and you can see that it is extending through the myometrium posteriorly and coming close to the bladder.

But that there is intervening myometrium at this point.

So you can really see the anatomy in detail when you use the higher frequency transducers.

Here's another transvaginal image and this is the bladder again, this is the subplacental hypoechoic zone and you can see placental tissue moving up into that area.

In this, in the same case in a slightly different scan plane.

This is the bladder wall and you see how it's become very irregular and you have placental tissue touching the bladder mucosa.

And here's color doppler of the same one showing a lot of vascularity in the area.

So there may not be disruption in flow but there is a lot of placental flow extremely close to the bladder mucosa.

Normal Anatomy and Physiologic Changes

What about normal anatomy?

This is an image of a patient who is pregnant with twins.

First pregnancy, this is the placenta.

There are huge vessels here, but this is normal myometrial flow.

And remember that at term 25% of your cardiac output goes to the uterus and there will be a lot of blood flow surrounding the placenta, which is a highly vascular organ designed for gas exchange for the fetus.

This is another patient in which we were just doing an AFI and if you look, this is a transabdominal midline sagittal image and there seems to be no bladder mucosa here.

But look at the angle of insonation.

There is no 90 degree reflection here.

So you're not going to get a good interface for bladder if it exists there.

So you have to be careful with how you are scanning and what inferences you make based on things that you see on an image that was obtained as part of a routine study.

The uterus is a very vascular organ.

Here's a transabdominal image.

This is a nice long closed cervix and look at all of these vessels in the myometrium.

Now in this case the placenta is posterior and clear of the cervix.

So we're really not worried about accreta.

But if those vessels were there and a patient happened to have an anterior placenta, people can get very agitated and concerned that they have direct invasion from the placenta to the myometrium when in fact this is normal myometrial vasculature.

And this is an example in the same case of the uterine artery with really low flow in myometrial vessels.

Another example, here's placenta.

Here's big vessels in the myometrium.

Here's slow venous flow but the vessels are nowhere in relation to the placenta.

This is a busy slide and it's a poster that we presented some years ago because it's a pitfall that I think people need to be aware of that you can have varices in the bladder of a pregnant woman that reflect the congestion and increased intravascular volume in pregnancy and do not in fact relate to placental blood flow at all.

And this patient was actually counseled that she probably did have a placenta percreta because she was a primigravida with a placenta previa who presented with bleeding at 24 weeks.

We followed her, she settled and we were able to identify later that in fact the fetal head presenting was abutting the cervix.

There was no placenta anywhere near the uterine wall but there were varices in the bladder.

So be aware of that as a pitfall.

I'm gonna show you some more.

Here is a case and where's the retroplacental clear zone up here?

Probably obliterated by transducer pressure.

You can clearly see there's placenta.

Here's myometrium, but then there's an irregular bladder interface and there's blood flow in the bladder wall.

Well these are bladder varices.

This is normal myometrium and normal placenta.

This is a patient who had a prior cesarean section.

Here is placenta also don't see the subplacental hypoechoic zone and I see these huge vessels in the bladder and I see a dehiscence of the cesarean section scar, but the placenta which is further superior in the uterus is not invading.

These vessels are in the bladder and are nothing at all to do with the morbidly adherent placenta spectrum.

Scan Plane and Gain Settings

What about your scan plane?

What I like to do in cases where we are concerned, so there is a placenta previa or a low lying placenta in a patient with a history of a prior cesarean section is to evaluate it in real time and to actually walk the placental insertion site and the site of the prior hysterotomy in sagittal sections from right to left so that you always are aware of your orientation and you always know where the bladder is, where the myometrium should be or is and where the placenta is.

Be careful also with your gain settings and color doppler.

This is a case I showed you earlier and it looks like there's all sorts of flow here.

Now remember again, fetal movement will cause flash artifact in the amniotic fluid and you can look like you have invading vessels going from the placenta to the bladder just by turning the gain up.

You must adjust the gain so that you don't have bleed.

And this is the same patient with a better gain setting.

And now you can see that there is flow along the myometrium with some interruptions and an abnormal vessel here going from the placenta out to the bladder wall.

Observer Bias

Another thing to always consider in medicine is observer bias.

And we recently completed a study which will be presented at the Society for Maternal Fetal Medicine by Dr. Bowman, in which we blinded three radiologists and three maternal fetal medicine specialists and gave them still images from multiple patients who were being evaluated for possible placenta accreta.

And what we found was it is extremely difficult to do this in a blinded manner.

I'm sure that poor Dr. Bowman would've liked to slap me many times because I wrote at the bottom of the sheet, I can't tell without going in the room.

And I am actually R2 in this study and you can see my sensitivity is pretty pathetic really.

It's hardly better than a coin toss.

I'm pretty good when I said it was a placenta accreta that I was right 90% of the time.

But the end game of the study was that ultrasound for placenta accreta may not be as sensitive as previously described.

And the things that we found most useful were the placental lacunae, the loss of the retroplacental clear space and abnormal doppler findings.

Diagnosis in Early Pregnancy

When can you make the diagnosis?

People refer to this and think about it as more second and third trimester condition, but it has to start somewhere.

And so there's a body of literature now suggesting that cesarean scar pregnancy or implantation of the pregnancy into a cesarean scar pregnancy is the first trimester manifestation of the placenta accreta spectrum.

And again, this is a recent review of the literature that the interested reader can consult.

This is an example of a first trimester cesarean scar implantation.

So transvaginal image, here's the cervix, the uterine fundus is off of the picture.

This is the lower uterine segment, this is the bladder and this is the gestational sac.

And if you look, you can see that it's definitely eccentric with respect to the cavity.

But when you magnify up you can see that this is the chorionic tissue around the gestational sac.

Here's the endocervical canal, this is the bladder wall.

So the epicenter of this pregnancy is in fact in the prior scar and not in normal myometrium.

Here's another example.

This is a retroverted uterus.

So we have cervix and again, here's gestational sac sitting right at the top of the cervix.

The uterine body and fundus are off the picture down here and there's a live embryo in this case.

And again, note the proximity to the bladder angle.

Treatment of Cesarean Scar Pregnancy

The best way to treat these pregnancies is to do an injection of methotrexate.

And this is often most easily accomplished using the vaginal transducer in the needle guide.

And here's just an example and yet another case cervix endocervical canal sac implanted in the region of the cesarean scar, yolk sac embryo.

Here's the needle guide and here's the needle injecting methotrexate into the region of the embryo to effectively terminate that pregnancy and allow resolution of this sac so that it cannot progress.

Here's another example that is taken images before and after injection of methotrexate.

And here's the sac in the scar.

Here's the involuting sac following methotrexate injection.

And you can see there's some free fluid in the pelvis as a result of the procedure.

There was actually some in the patient's pelvis even beforehand.

This is the external os, internal os bladder neck epicenter of your sac in the scar tissue.

Why It Matters: Example of Uterine Rupture

You may say, okay, you've kind of shown several cases of this.

Now why does it matter?

And here's an example of why this is a case shared with me by my colleague Dr. Winter.

And it's an old case but very instructive.

This is a patient who presented at 13 weeks pregnancy with abdominal pain transabdominal ultrasound, showed a uterus which was empty but a large gestational sac type appearance here in the region of the lower uterine segment and cervix.

And this was part of a fetus and MRI was done.

And you can see that this is a T1 post contrast and there's a lot of enhancing tissue here, which means vascularity.

That's the gestational sac here on T2 is the gestational sac as well.

But note all the gray signal from myometrium stops up here and the cervix is down here.

So the assumption was that this patient had a cesarean scar implantation with a placenta percreta.

And at surgery, this is a diagram drawn by the surgeon, there was almost complete separation of the uterus from the cervix.

The fetus was dead.

Here's the uterus, here's the cervix and the surgical specimen.

And this was a uterine rupture secondary to a placenta percreta at 13 weeks.

Thankfully the mom had a very good outcome apart from the fact that she had a hysterectomy.

So remind yourself always when you see a first trimester pregnancy that seems to be a low implantation check if the patient has a history of prior cesarean section and make sure that the pregnancy is not implanted at the scar site.

It's relatively unusual for uterine rupture to occur in this condition, but more often it progresses and the placenta continues to grow.

And we go back to what we showed earlier, which are these examples of abnormal thick placenta with these huge dilated vessels sometimes known as tornado vessels.

This is a vector four.

This is getting out the curved six and walking the placental implantation site and you can see how the placenta extends right up to the uterine serosa in these pseudopods of tissue reaching the serosa ready to extend across and possibly even become placenta percreta.

And here are the abnormal vessels along the bladder wall.

Role of MRI in Evaluation

So is MRI necessary in evaluation of placenta percreta?

I will in the interest of true disclosure tell you I do both ultrasound and MRI and I find MRIs extremely difficult to interpret, but I do see that they occasionally have a role to play.

And what I've done here is I've put a sagittal MR and a sagittal ultrasound side by side.

So B is bladder, P is placenta and you can see here it's just the lower uterine segment.

This is all placental tissue.

There is subplacental hypoechoic zone, there is a normal bladder.

There's the bladder wall here, this is the bladder.

You can possibly better perceive the tenting of the bladder angle, but that could happen as a result of scarring from a prior cesarean and it doesn't necessarily imply that there is a percreta.

And here's placenta with normal placental vessels on the surface.

Here's the maternal colon.

Maternal spine is back here.

This is part of the presenting fetal head and multiple loops of cord and fetal extremity.

The thing the MRI is better at than ultrasound sometimes is creating the big picture.

So these are sagittal images, this is maternal bladder.

You can see flow voids.

Here is vessels in the bladder wall.

You can see that the dome of the bladder is tented at the site of the prior scar.

And here there's an area where the placental gray signal tissue seems to go straight through into the dome of the bladder or over the dome of the bladder.

And we have an interruption in the normal myometrial signal at that site.

So this is a way where you can show a small focus of percreta and because MRI is a volumetric acquisition, you get the whole organ with ultrasound, you have to keep your eyes peeled as you walk the scar in real time.

I think it's possible to do this but many, many people are reading ultrasound nowadays where they have a series of images obtained by a sonographer.

And if there's nothing else you take from this talk, it is the fact that you have to interrogate the scar in detail in real time with the highest resolution transducer you can find.

Here's another example on MRI.

And this was something that I did for our surgeons.

So this is a sagittal T1 side.

This is the sacral ala and that is the greater trochanter of the maternal hip.

And you can see all of this vascular stuff going out to the pelvic side wall.

And that's the far right, this is the far left in the midline, which is maternal lumbar spine and sacrum.

We have a cesarean which is evidence of prior instrumentation in the uterus.

We have normal placenta and myometrium here.

But watch what happens as we come down, there's a defect in the myometrium with what appears to be placental tissue breaching it.

And this is maternal colon lying perilously close to what appears to be placental tissue.

This was an axial image and you can see this is the uterus, this is the psoas muscle, psoas muscle sacrum.

Here's fat behind the uterus in front of the psoas muscle on the left.

Here's the psoas muscle on the right and we have what appears to be placental vascularity breaching through the myometrium sitting right on top of the psoas muscle.

Also note that this is the right common iliac artery flow void and where the internal and external iliac artery branches should have been visible.

They were lost in placental vascularity.

So I was extremely concerned about this patient and contacted the surgeons and made this PowerPoint for them.

And I heard that things went fine.

I looked up the operative report and it said, no evidence of placenta accreta but lots of neovascularity.

And I was astonished.

I thought, look, this is the cervix, this is the endocervical canal and the entire pelvis is full of big abnormal vessels.

They reach all the way to the urethra, they reach to the rectum.

I don't know what this is, if it isn't placenta accreta percreta.

Well I then read the path report, which came out several days later and there was indeed a placenta percreta with a five centimeter mass of placental tissue through a defect in the uterus and extensive accreta and increta elsewhere.

So this is another issue that you may have to deal with.

If you want to know why we see what we see, you have to read op reports and read path reports and sometimes it's quite disturbing to have a patient in whom you were convinced there was an abnormality and then have the operative reports say everything was fine.

And I think what happens is what the surgeons refer to as neovascularity is in fact vascularized placental tissue invading the structures in the pelvis.

Conclusion

So is MRI necessary?

Bottom line is I think as a radiologist we should use all the tools in our armamentarium and sometimes MRI can be helpful.

It gives you an ability to show the lay of the land and that's important if it changes management.

It's also being published in the literature that MRI is better for evaluation of the posterior myometrium or lateral extension in patients in whom there's been previous interventions such as a myomectomy.

So it may be that you're better able to image areas that you can't reach with an ultrasound transducer in the third trimester of pregnancy.

I like to do ultrasound first and use MRI as a problem solving tool, but I don't think that we should say that one or the other is the only thing that we can do.

And be aware that there are plenty options for mistakes and pitfalls and be very careful and follow all your cases and see what the pathology shows.