Orthotopic Liver Transplant: Hepatic Artery and Portal Vein Complications - SD

Introduction

I am Dr. Mindy Harrow from the Albert Einstein Medical Center in Philadelphia, Pennsylvania.

My talk will be on orthotopic liver transplant complications involving the hepatic artery and portal vein.

Surgical Technique

The surgical technique for an orthotopic liver transplant involves an end to end arterial anastomosis between the donor and the recipient hepatic arteries.

There can be a variety of variations of anatomy, which will require the surgeon to hook up the arteries in different ways.

And if in the unusual circumstance there is insufficient inflow from the native hepatic artery, they may use some type of interposition graft to anastomosis to the infrarenal aorta, which is helpful to know when you're scanning.

The other anastomosis will be the portal vein anastomosis, typically an end to end anastomosis between donor and recipient portal veins.

Similarly, there can be problems if the patient has preoperative chronic portal vein thrombosis, and then portal inflow can be reestablished again with an interposition graft from an iliac vein to one of the other venous sites postoperatively.

Sonographic Evaluation

A variety of things are evaluated in the sonogram, but we're going to concentrate on those problems related to the hepatic artery and the portal vein supplemented by gray scale imaging that can be useful in this regard.

There are many interesting things that happen hemodynamically in the immediate postoperative period, the perioperative period, and then eventually thereafter.

And these interesting complicated things are sometimes unique to this situation of a transplant and can be actually very interesting and intriguing to figure out.

And when things happen abruptly, one can make significant impact in the patient's outcome by doing adequate and sonography.

Normal Waveforms

To understand abnormal, obviously you have to know normal and the normal waveforms from hepatic artery and the portal vein after a transplant should be the same as they would be if you didn't have a transplant.

The hepatic artery has a nice prompt upstroke, it has continuous flow during diastole.

One should measure the angle corrected velocities and the resistive index, the main portal vein and the intrahepatic right and left branches should be interrogated as well as other aspects of the portal system if possible.

And you typically get a nice low level continuous flow with respiratory variability and one can do an angle corrected velocity as well in the immediate perioperative time period.

Perioperative Variations Within Normal Range

There are a variety of things that can change a bit in both the hepatic artery and the portal vein, which are considered within the range of normal.

They may be followed and typically resolve, but one wouldn't want to act on it or return a patient to surgery or other procedures.

So it's important to know about them.

These involve a high resistance waveform in the hepatic artery and a variety of portal vein issues, which we'll look at.

This patient on day one had a perfectly normal looking hepatic artery and power doppler next to the portal vein, but if you notice, the waveform had lack of flow during diastole, so the resistive index is essentially one.

Nonetheless, the peak systolic velocity was still reasonably adequate.

It was easy enough to find the artery, and most importantly, clinically the patient was doing well.

We followed carefully, and sure enough, by day three, the waveforms had normalized.

The resistive index was now 0.64.

Since there was flow during diastole, this early high resistance pattern is well known and it can be low diastolic flow or even no diastolic flow.

The resistive index is zero, and it's actually not uncommon.

It is not even necessarily associated with a poor outcome or decreased survival, but probably has more to do with wherever and however long it took to get the liver available.

Longer ischemia times older donors typically are associated with an elevated resistive index.

If there is concern, instead of doing other imaging studies, one can actually administer a vasodilator, which typically will increase the amount of diastolic flow.

And on the portal venous side, there are a variety of findings that can show up that might initially be of concern.

One of these is a pattern in which it looks as though there is a stenosis in the portal vein.

Oftentimes there can be a relative mismatch in size between the donor and the recipient portal vein.

And at the level of the anastomosis, which you can see here through this area, the velocities will increase and there will be more turbulent flow as you can see in the corresponding duplex doppler.

In the area of maximal narrowing, the velocities considerably elevated with a lot of turbulent flow, proximal to the narrowing.

The flow was normal.

Again, one can just watch this assuming that the patient is doing well and there are no other signs of increasing portal hypertension such as ascites.

And this patient actually was seen quite a bit later, had been doing well, and one year later, this relative mismatch has essentially resolved and there's perfectly normal flow without any elevated venous waveform.

Another unusual pattern that one can see in the portal vein is this one.

It is a very pretty one.

It's red, blue, red, blue, and a sort of candy cane circling thing.

And that is due to helic flow in the portal vein.

It doesn't mean that the flow is going towards the liver and then reversing and towards the liver, but rather a spiral of flow.

And as the spiral comes towards or away from the transducer, the corresponding color changes from red to blue to red.

And so you get this kind of pattern totally fine, totally normal, typically will resolve eventually, and it can be seen in many patients.

Also probably has something to do with a mismatch in size between the donor and the recipient portal vein, the last pattern is really much less common, but you might see occasionally, and that is portal venous gas immediately after a liver transplant.

You can notice here in the portal vein these tiny little bright echoes, sort of these little parallel things.

And if it was a realtime image, you would see them moving with the blood out into the periphery of the liver.

And a pulse stopper will show these very typical spikes superimposed on the underlying normal portal venous waveform.

And it sounds funny in the pulse doppler, again, an uncommon but likely insignificant finding, which can just be followed.

True Complications

Now let's turn to some true complications and the majority of these have to do with a hepatic artery and we will be discussing thrombosis at various times stenosis, and then a variety of other less common, but nonetheless important for the patient findings.

Hepatic Artery Thrombosis

Hepatic artery thrombosis has a significant incidence.

It's higher in children who have a smaller hepatic artery but has a low, relatively low level in adults.

Clinically it can have very non-specific findings as many of these vascular abnormalities can risk factors.

There's a whole host of them, as you can see here.

But for the artery in thrombosis, there can be significant complications.

The biliary tree within the liver has its blood supply solely from the hepatic artery.

So that is actually what suffers when there is a hepatic artery thrombosis or even a significant ischemia and a variety of issues can then ensue.

And of course there can be gray scale findings in the biliary tree that might help you in the diagnosis of a hepatic artery ischemic event.

How do we deal with this?

The screening test for all of these is duplex doppler color combination duplex doppler, and it all depends on when you're looking as far as the sensitivity in the immediate postoperative and perioperative period.

You should really achieve a near a hundred percent sensitivity for hepatic artery thrombosis later on as we'll discuss as time goes on, the sensitivity actually decreases.

CTA and MRA can both be used as a follow-up and conventional catheter angiography as necessary.

Microbubble contrast agents clearly could improve our sensitivity in these cases to make sense of hepatic artery thrombosis.

It is convenient to divide it into both early and late stages.

Early thrombosis can be considered to be within a week or some number of days extra after surgery, and this is usually considered a surgical complication and will require surgical revascularization.

One can't do stents and a variety of other procedures on an anastomosis that is that fresh.

These clinically will typically present with symptoms.

Late thrombosis after a week and thereafter can be due to a variety of factors.

And as time goes out after the transplant, we become less and less sensitive to hepatic artery thrombosis.

So again, in the early time period, immediately after surgery, lack of flow is the best indicator of what's happening, and ultrasound is a very sensitive modality.

Later thrombosis as I'll show you why ultrasound becomes less sensitive.

So let's look at some examples.

This was a routine postoperative study done eight hours after the transplant.

And I should preface this to say that within 12 to 16, 15, 16 hours, we typically do the first bedside ultrasound in all of these patients, even if there are no symptoms.

So there we were routinely nothing had thought to be unusual about the surgery or the patient's course To this point, we imaged at the port of Heus and there's normal flow in the portal vein.

And even with the gain all the way yanked up, we could not really find any flow at all in the artery.

Interestingly enough, at this time, and it was only eight hours after the surgery, we were able to perceive that there were odd sorts of heterogeneity in the left lobe of the liver only with areas of slightly lower and slightly increased echogenicity.

So we had apps and flow already.

There was some sort of abnormality of the left lobe.

This patient had a immediate angiography one, one injection.

You can see the catheter here in the celiac axis.

Splenic artery, no flow in the hepatic artery, and the patient went back to the operating room and had an emergency thrombectomy.

Immediately after the thrombectomy, the waveform normalizes had a good resistive index of 0.64.

Nonetheless, in that short period of time, the left lobe of the liver had become ischemic.

And these are subsequent cts over the intervening weeks.

And you can see that this area was ischemic and it eventually involuted, but the rest of the liver maintained itself.

And on in this regard, patient had other issues.

The patient's right hepatic artery maintained enough liver.

There can be a situation in which you might happen to image the hepatic artery when something is happening to it, but it hasn't gone down completely.

I believe this is really relatively rare.

Most of the time in the immediate postoperative period, there is either is an artery or there isn't an artery.

But once in a blue moon, this might happen.

As in this case, this was the first study.

Nothing was deemed unusual clinically, it was very, very difficult for us to find any flow in color or power doppler.

We had our settings quite sensitive, and when we finally found some, we got these tiny little high resistance blips.

Now you might say that this overlaps with the high resistance pattern that I showed you before, but in this case it was very, very difficult to even find the flow.

And then they became concerned about the patient.

We don't have contrast bubble agents to give, but the patient went back to the operating room and they found that there was a kinking of the hepatic artery.

It had seemed straight initially, then they sort of put everything in, closed the patient and there was a kinking.

They took out a piece of the artery and straightened it out.

The immediate waveform on the table before they closed the patient reverted to normal with a resistive index of 0.7 later in the day.

However, we did another ultrasound, and initially this seemed alarming and that the waveforms were still very, very high resistance, but the velocities were considerably improved and the pa and we could easily find the vessel in color flow.

And the patient seemed fine.

And so we waited and eventually this reverted to a normal waveform several days later, and things seemed fine.

Then two weeks later, the patient returned for a routine follow-up and we had an interesting sequence of waveform in the left lobe of the liver.

We always image the main hepatic artery and the intra hepatic left and right branches.

The main was fine.

The left hepatic artery had wonderful waveform and a relatively normal resistive index, but we had a great deal of difficulty seeing and finding the flow in the right lobe.

And what we got was just very low velocity and very abnormal high resistance wave forms.

The patient went on to have an angiogram, and as you can see, here's the main hepatic artery and here's the left branch.

And the right branch had a pattern of dilatation narrowing, dilatation narrowing.

And we believe that this happened during the thrombectomy that was done at the time of the surgery, that there was an injury, if you will, and probably some dissection of the blood vessel.

But nonetheless, the gray scale images were fine and the patient was left alone at this point.

Moving on, there's a patient who on post-op day one had a perfectly normal hepatic artery weaving in and out of the plane of view here next to the portal vein.

And on post-op day 15, no hepatic artery to be seen.

Next to the portal vein, this patient had a CTA, this is the aorta celiac axis, the splenic artery, and there was an abrupt cutoff of the hepatic artery.

So this was a hepatic artery thrombosis slightly later than the other cases.

On post-op day 15 when we did the imaging, the left lobe of the liver looked somewhat heterogeneous and 10 days later, these large low density collections appeared within the left lobe of the liver.

And these are myelomas.

This is the effect on the biliary tree of the ischemia due to the fact that the hepatic artery had gone down and for whatever reason was only affecting the left lobe of the liver.

The patient went on to require drainage of this bile, which became infected.

You can see here an injection of the bile duct and these casts filling the biliary tree and the catheter was placed, but eventually with antibiotics and drainage.

This did resolve as they often do.

Interestingly enough, however, several months later after the myeloma was cleared, we noted that the patient had developed diffuse portal vein thrombosis, though he was doing much better.

You can see the varice increasing in size.

The patient also was having sonography during this period and it was clear to us that the portal vein had thrombosis.

But very interestingly, all of a sudden with this patient who had a known hepatic artery thrombosis, we were able to visualize numerous small arterial vessels all over the port of hepititis and obtain relatively normal looking waveforms.

Normal resistive index, a slight delay in upstroke.

So this is the interval development of arterial collaterals.

And the patient had an angiogram and it confirmed here's the superior mesenteric artery injection early on and later during the phase, you can see numerous collateral vessels that had developed that were now supplying the liver.

This is the late portal phase and you can see splenic vein and superior mesenteric vein and the main portal vein was down.

So this patient amazingly had no main hepatic artery and actually no portal vein, but had collateralized the arteries.

And this phenomenon of collateralization is actually might seem counterintuitive in patients who have only one artery going to the liver, the one that was reconstructed at the time of surgery.

But we have come to understand that collaterals can develop from a variety of places and as you see, can give wave forms that aren't that dissimilar from a normal hepatic artery.

So to reiterate, here's a another patient.

This one is two years out from transplant and had some rising liver function studies.

The finding that was most easily appreciated on sonography were some dilated ducts in the left lobe of the liver.

These are the waveforms we obtained from what looked like the hepatic artery at the port of Hetus.

And in this case, they are abnormal, though they are present, the resistive index is low and there is a delay in systolic upstroke.

This is this typical parvis tus waveform.

And not surprisingly, this patient also had hepatic artery thrombosis.

This is the injection, the catheter here.

This is the hepatic artery with an abrupt cutoff.

You can see the supply of the spleen, and when we injected again the superior mesenteric artery, all of these collateral vessels going to the liver can be seen.

So this is late hepatic arterial thrombosis with arterial collaterals.

And this phenomenon is the reason that we are less sensitive for hepatic artery thrombosis.

Later after surgery, patients can develop collaterals, they can develop them relatively rapidly.

Sometimes the wave forms look odd.

Sometimes they do not look that odd and all things being equal, we might not know that the hepatic artery had actually thrombose.

Nonetheless, when this happens, some of the patients will develop ischemia to the biliary tree, get biliary dilatation and sludge, as did our patient.

And this is the cholangiogram.

And you can see this very dilated, narrow, dilated beaded appearance of the bile ducts, reminiscent of primary sing cholangitis as they're filled with casts.

And this may require a variety of drainages and catheters.

If you can get the patient through any biliary sepsis, they may actually do quite well.

So the phenomenon of late hepatic arterial thrombosis and arterial collaterals is well known in the literature.

Hence as I said, the reason that we are less sensitive for thrombosis as time goes on.

People have called this pattern the collateral transformation of the hepatic artery akin to what can happen to the portal vein.

And one should have a suspicion for this if all of a sudden biliary complications occur, myelomas, sludge, et cetera.

Probably this phenomenon happens more than is recognized because some patients probably are asymptomatic if their biliary tree does not become obstructed and go on and live well with this phenomenon.

We've studied and followed a long series of patients and we have patients who have been alive 8, 9, 10 years with known thrombosis and seem to live with it and do fine.

Other Hepatic Artery Issues

Now let's turn to some of the more sub selective specialized issues that can happen.

This was a patient who had also very dissimilar waveforms coming from the left and right lobe of the liver.

It's a little bit similar to the other case I showed you.

The left hepatic artery waveform had a resistive index of point A three a little bit high, and the right hepatic artery had almost very little difference between systole and dia.

It almost looked venous with a calculated resistive index of 0.36.

So what was going on here?

This patient had an angiogram and you can see here's the catheter and the main hepatic artery.

This is the gastroduodenal artery.

And this patient had a complete thrombosis of the right hepatic artery.

The left int hepatic artery was supplying the right lobe via collaterals.

And when patients vessels are re collateralized, they typically demonstrate a parvis tus waveform, which is what was happening.

There was nothing we could do at this point about the right hepatic artery thrombosis.

So it was decided to thrombo to coil embolize the gastroduodenal artery so that it would not steal flow and that did improve the flow into the left lobe, which then ultimately improves the flow to the right lobe.

The phenomenon of an initial low hepatic artery resistive index after surgery has been noted in the literature and this is felt to be a more significant finding than the elevated resistive index.

It may be a predictor for subsequent events of a variety of kinds and if especially when this is persistently low, if it's a transient finding, not as much of a problem.

So a low resistive index in the perioperative period should cause one to carefully follow the patient.

Low resistive indices can also be seen if there is a hepatic artery Stu stenosis more proximally and stenosis of the hepatic artery is the second most common arterial complication.

It typically occurs at the anastomosis for a variety of reasons.

It can go on to thrombosis and it can in and of itself cause a significant enough ischemia to the liver to have biliary complications.

If one is able to actually interrogate the stenotic and sight and visualize it, it will typically demonstrate what stenosis elsewhere would of an elevated velocities and a distal par t waveform.

The problem exists with the fact that the hepatic artery often innately or is particularly when it a transplant has been performed, is rather redundant and tortuous.

This makes it difficult to actually find stenosis.

In addition, there are often calculated high velocities and this just due to the wave to the vessel curling around.

So we often rely in this case more on the downstream parvis tardis waveform When a the vessel is very di very tortuous or b, we can't even visualize the entire hepatic artery.

A relatively low grade stenosis may not even cause any significant doppler finding, and that is why if the clinical suspicion for a hepatic artery problem is high due to findings of ischemia, et cetera, and we can't figure it out on ultrasound, one should go on to C-T-A-M-R-A or catheter angiography.

So here's a typical example in which we could actually find the stenotic segment of the artery.

This is the hepatic artery and actually what you're seeing mostly here is a color brewery due to the high velocities and significant turbulence at the site of the stenosis.

And at the actual s stenotic site, we were finding peak systolic velocities that were quite high of 4 29 centimeters per second with a quite low resistive index posts stenotic.

Further distally at the port of Hetus, we had a parvis TARDIS waveform.

And mind you, this might be the only place that we could visualize in some patients and if so, if you see a waveform like that, it could be due to a more proximal stenosis.

This patient went on to have a catheter angiogram and you can see the stenotic segment right here.

This is after an angioplasty and stent was placed and the post stent ultrasound does show some improvement in the waveform.

The velocities were still high but had significantly decreased.

The resistive index was still relatively low.

Often there is some edema and some residual stenosis that needs to be followed.

This patient had rising liver function studies clinically.

The waveform more proximally was relatively normal and distally wasn't that abnormal, but there was a slight delay in upstroke.

And the resistive index though in the normal range was lower.

So that was all that we saw.

We couldn't see the entire artery, but the liver gray scale showed areas that were hypodense and little areas of biliary dilatation concerning for ischemic changes.

So this patient, since we couldn't see the entire artery, went on to have A-A-C-T-A and these are the reconstruction from that study.

You can clearly see that there is a stenotic segment and post stenotic dilatation.

As a result, the patient went on to have a catheter angiogram.

Here's the catheter and you can see an interesting phenomenon.

The large vessel that you see to the left of the image is the splenic artery.

And that filled briskly and rapidly and the hepatic artery filled more slowly.

As time went on in during the injection, we were able to find this area of stenosis that corresponded to the CTA, but you'll notice still that there was quite exuberant flow into a large splenic artery and delayed slower filling into the hepatic system.

And late images showed sort of persistent dwelling flow on the hepatic arterial side instead of the splenic side.

So this patient not only had some degree of stenosis, but also had the phenomenon of a splenic artery steel and the combination of those were decreasing the inflow on the arterial side to the liver.

Therefore, this patient went on to a combination of intervention, had an angioplasty of the hepatic artery and also a splenic artery embolization.

This resulted in normalization of the waveforms and a lower resistive index.

And over time, the spleen actually decreased in size.

This is a relatively uncommon phenomenon in which there is shunting of arterial flow to the splenic artery.

In these cases, often you'll see higher velocities in the portal vein at the liver because flow is going to the spleen and then back into the portal system.

And so there may be an increased hepatic arterial resistive index and high portal vein velocities.

Here's another patient.

The immediate postoperative waveforms in hepatic artery were fairly normal with a low normal resistive index.

And six months later the resistive index had decreased.

We went looking around and you scan a variety of vessels and we noted that at the relative origin of the celiac axis there was an area of narrowing.

And at this location we found relatively high velocities.

This patient had a CTA and demonstrates almost a web-like phenomenon at the origin of the celiac axis in the axial image.

And the reconstruction shows a kind of narrowing right here at the origin of the celiac axis and post stenotic dilatation.

This is the median arcuate ligament syndrome, a piece of the diaphragm that kind of curves under where the celiac axis arises.

Many people have this, but in most of the time it's of no clinical significance due to the rich collateral blood supply to the liver from other vessels.

But in a liver transplant patient in whom the only inflow is in through the hepatic artery, such a ligament syndrome can become clinically significant.

There are other causes for hep celiac artery stenosis as well, including atherosclerosis.

But it is one of the extra things to think about when the inflow to the hepatic artery appears abnormal.

But there is no stenotic segment in the hepatic artery itself.

This patient had an episode of septicemia fairly shortly after surgery and the hepatic artery had a focal area of dilatation in color.

And you can see that in this area of dilatation there's flow towards the transducer and away in a red and blue pattern of the so-called yin yang sign that is typical of a pseudo aneurysm, which is what the patient had as you can see on this angiogram.

And that was treated with a catheter with angioplasty.

Several days later, a gray scale ultrasound, a new cystic lesion popped up in the liver and it's always important to turn on color again.

This one had a red blue pattern as well.

And this was the second pseudo aneurysm.

Since this one was in the liver, it was treated with a thrombin injection.

So hepatic artery pseudo aneurysm is another complication, relatively rare, usually related to some sort of septic episode.

Always remember when new cystic lesions appear to turn on the color and see if it is a pseudo aneurysm.

Portal Vein Complications

And turning now to portal vein complications.

These are relatively rare.

There are not as many kinds of complications as in the hepatic artery, but there can be thrombosis and there can be stenosis.

Whether or not these come from faulty technique or other issues aren't clear.

And then one has to always be careful if there are large varis left behind that there can be a steel effect.

It's important to know if there has been any variation in the standard anastomotic procedure in the operating room so that you can appropriately evaluate.

And clearly if the portal vein narrows down or then becomes thrombosis, the patient may develop acute findings of portal hypertension.

So let's look at some examples.

This was a perioperative portal vein thrombosis within a few a day or two of surgery, we're imaging the patient, the portal vein has become slightly small in size with low level echoes.

The artery has become larger and has elevated velocities as happens as the flow increases through the artery because the portal vein has thrombosis.

This was the CT scan confirming the thrombosis.

The patient went back to the operating room after thrombectomy flow had normalized and one day after thrombectomy the patient was fine.

This patient you can see in the portal vein developed a significant area of narrowing with post stenotic dilatation and elevated velocities.

And this was far enough away from the surgery and the patient was symptomatic that we believe this was not just a donor mismatch.

You can see the lower velocity pres stenotic and quite high velocities at the level of the stenosis.

This patient had an angioplasty done of that segment.

This is the area of narrowing and this is post angioplasty.

So portal vein stenosis can occur.

It is sometimes difficult to know if it is a donor mismatch or if it is truly a hemodynamically significant stenosis.

There's not a huge amount in the literature about it, but basically people think if there's a three to one or four to one increase in velocity, this may be quite significant.

But you need to usually have clinical findings of that support this to even do something about it.

And balloon angioplasty and stent placement, if it's at far enough out from the surgery would be the appropriate treatment.

This patient came five months after a transplant and the portal vein had essentially very little flow in it, very, very low velocities and low level echoes.

You can see this is a direct injection and there was thrombus in the portal vein that had developed.

This was acute thrombosis and the patient also, if you notice, had developed these large coronary varis.

This patient went on to have embolization of the varis and then an angioplasty of the portal vein and that reestablished the flow afterwards.

And this case is was a very interesting one.

We actually published as a case report.

The patient on day one had relatively poor but nonetheless flow in the correct direction.

In the main portal vein, we watched carefully and on day two it was very difficult to see the portal vein.

There was very low level two and fro flow and so the patient returned to the operating room and to have a revision and thrombectomy.

However, after revising the anastomosis, the surgeons felt that the flow was still quite poor and asked us to scan in the operating room where indeed it was very difficult to pick up any flow.

The flow remained quite low level and the anastomosis had been completely redone and there was no obvious cause for this problem at the level of the port of heus.

So we went looking around and happened to notice that on the left side of the patient, which hadn't been opened up.

And so we were scanning through the omentum.

There was a very, very large spleen, renal varix with blood being shunted towards this varix.

And the patient then went on to have ligation of the varix after which time portal flow was reestablished.

This is the patient's preoperative ct and you can see these very, very large spleen renal varis that had begun to steal flow from the main portal vein and needed to be ligated.

Suggestions and Tips

So suggestions and tips for dealing with the hepatic artery and portal vein after liver transplant in the preoperative period, CTA should be performed to look for all kinds of variations in vascular anatomy.

Any varis, figure out the size of things you need to talk to the surgeons and find out what they did and if there was any unusual anatomy.

And then when doing imaging afterwards, don't just look at the liver, you need to look at the spleen.

Look for varis, look for various kinds of stenosis such as the celiac axis and remember that complications can occur in combinations.

Use the gray scale imaging to look for signs of ischemia to the biliary tree because that may be a tip off that there really is an ischemic problem going on with the hepatic artery.

In general, I have always a high index of suspicion for any complication and if the patient seems to be having problems that you can't figure out.

Ultrasound is not the be all and end all, and the patient should go on to whatever procedures might help in that regard.