Patient Selection for a TIPS Procedure

[Michael Barraza]
I'd like to start maybe by having you each tell us what your TIPS practice looks like. Overall, I'm mostly treating patients with acute or recent variceal bleeding. Tell us what you're mainly doing these for and where you're getting most of your patients.

[Peter Bream]
We're a transplant center at UNC, so we have a large liver failure population, a large cirrhosis population. We work very closely with the hepatologists and the transplant surgeons here. That's where most of our patients come from. We do a fair amount of TIPS for both of the main indications, for ascites and for variceal bleeding—mainly esophageal variceal bleeding that's refractory to GI management, but also for gastric varices as well, although it's not as robust. We usually end up doing an antegrade obliteration of the varices, along with the TIPS, if we're dealing with gastric varices that do not have a splenorenal shunt. In my previous practice, we rarely ever did an emergent TIPS in the middle of the night… We had a pretty strong former colleague that felt like the mortality on those were so high that you needed to get the patient stabilized, maybe put in a Minnesota tube, or a Blakemore tube, transfuse them, do the full work up with the Echo, looking at the right heart, and making sure that the TIPS is going to be safe, rather than just barrel in there with somebody that's bleeding out in the middle of the night.

[Peder Horner]
We're a private practice. My scenario is more community-based hospital, so we got some big ones and some smaller ones. It really depends on where we are, and the indications, because some of the more refractory ascites cases I can do at one hospital where's there a big liver population of non-alcoholic steatohepatitis, cirrhosis, or your cryptogenics. Then there are peripheral folks in the smaller hospital, sometimes very high acuity, high alcoholic livers with variceal hemorrhages and what not. We try not to TIPS in the middle of the night either. I don't think anyone likes to do that, but I think it does come up sometimes when all other situations and options have been exhausted.

[Peter Bream]
BRTOs have become a huge part of our practice and cases where we might have done TIPS in the past, we're now doing BRTOs.

[Peder Horner]
My colleagues and I have switched to almost like a BRTO first, in approach to hemorrhage. If they've got gastric varices, if you've got a big shunt, we'll BRTO first. We don't even talk about TIPS. I really think that approach has been very helpful in our practice, and it's been pretty cool getting to share that with some of the GI docs. Some of them are, again, smaller community hospitals and they're not aware of the other options out there besides TIPS.

[Michael Barraza]
Are you guys approaching BRTO like the Japanese do, or are you doing it strictly for gastric varices in the setting of a splenorenal shunt?

[Peter Bream]
I would say mainly for us, if they have a splenorenal shunt, we're considering a sclerosing type of procedure to start with. There’s two camps. There's the western camp of diversion or decompression, and then you have the eastern camp, mainly the Japanese and the Asian, that have the sclerosant, or the obliteration side. As I've thought about this over the years, I start to think of these gastric varices being very similar to an arteriovenous malformation, where if you do not get rid of the actual varix—let’s say you just coil a couple of the inflows—it will recruit other vessels. There are so many vessels in this area that could be potentially recruited, and if you can't get directly into the varix and at least reflux sclerosant up into the varix, then you may really be treating it now, but you're not treating it down the road.

[Michael Barraza]
One last question about BRTO, are you guys strictly doing them for gastric varices or are you occasionally doing them for patients with esophageal varices?

[Peder Horner]
Mine are usually for gastric varices. I just don't see the splenorenal shunt in the esophageal camp, I really don't.

[Peter Bream]
I've done it before for small bowel varices. You have another route where it's going up through the duodenum and actually I’ve done more of an antegrade approach to those, trying to sclerose the varix that may be in the duodenum. I've only had to do that a couple of times, though. That's kind of unusual.

[Peder Horner]
I’ve had a couple patients over the last year where they've had stomal varices. They were very high functioning people, did not want any kind of risk of hepatic encephalopathy, so we actually did transhepatic portal vein access. Then, we went down to the varices and sclerosed them just like a BRTO. Those have worked really, really well so far. I'm out about a year now.

[Peter Bream]
Whether we stick them directly or go transhepatic…I can tell you that my practices have varied on that. When I first started doing those we were just coiling, and then as I got more facile with Sotradecol, I was doing more obliteration. I can't tell you which is better, but I can tell you that just blocking the portal venous flow and getting that main channel that's coming from the portal vein is dramatic for these patients.

[Michael Barraza]
Regarding alternative and adjunctive techniques that you use during a TIPS, when, if ever, are you guys embolizing varices in addition to placing the TIPS?

[Peder Horner]
I've always embolized varices if I see them. That's just how I was trained. Someone told me recently that you can't bill for embolization, and so, I don't really care, but if I see big varices, I usually still embolize them with coils.

[Peter Bream]
We would say do a hand injection of 10 ccs of contrast with the catheter in the splenic vein, and if you did not see any varices, you wouldn't embolize them. If you did a power injection, you're going to reflux up anyway, and you get false positive. We felt like that was enough to prove they were not filling anymore, but if you did see them filling, you would embolize them at that time. I've kind of used that. That's not very scientific, but it works.