Clinical Challenges in Hepatobiliary Surgery: Pancreatic Anastomoses in Whipples

Greetings, everyone. Welcome to another HPV episode on Behind the Knife. This is your HPV team coming to you from Stanford University in California. I'm Anish, a PGY3 general surgery resident here at Stanford, and I'm excited to be joined by some great mentors and Dr. Monica Dua, the associate PD of our HPV fellowship, and Dr. John Harrison, our second year HPV fellow and soon to be HPV faculty at MGH. Today we'll be discussing the pancreatic anastomosis during pancreatoduodenectomies, or the Whipple procedure. Specifically, we'll discuss the standard reconstructive options, as well as alternative approaches when patients have unfavorable and or aberrant anatomy or physiology. A lot of the concepts that we'll discuss in today's episodes can be found in a recent invited commentary in surgery that was written by the HV faculty here at Stanford and the HV faculty at MGH. And we'll include that article in the show notes. So Dr. Dua, could you just kind of talk to us about the standard

Bloomgard style anastomosis that most people are familiar with and most residents should be familiar with? Thank you. I think it's really an important topic because the pancreatic adenostomy is a really critical part of the reconstruction. And it's really important to pay attention to technique. And even though you have a favorite type of anastomosis, it's good to have different options available. So for me, I preferably do the Blumgaard anastomosis. In this particular anastomosis, it involves There's three silk sutures which are trans pancreatic, and they basically approximate the parenchyma to the dejunum using seromuscular U stitches. And then once these are placed, we do a duct to mucosa anastomosis using 6 0 PDS suture. Okay, awesome. And John, sometimes when we're on rounds, you talk to me about this end to side dunking technique. Can you describe that for our listeners? Right,

so this whole paper was inspired by the fact that there are many ways to do this in Asthmosis. And so, in Fellowship, Dr. Dua and Dr. Visar really taught me Bloomgarth style. But in Residency and the other half of the authors on the paper, they really use a more Duncan style anasthmosis. And It achieves the same goal, which is a two layered anastomosis, but the stitches are just put in in a different way, and it causes the jejunum to sort of roll over, just like a Bloomgart does, but it's done with a back row of silks, and then a ducto mucosa anastomosis with Vicryl, and then it has an anterior layer, again, of silk. And those trans pancreatic sutures that Dr. Dua talked about are on either side. Okay, perfect. And actually, John, in this paper, there's a nice phrase that you used, the anastomosis, the Achilles heel of the whip operation. Can you describe what you mean by that? Yeah, so this is a phrase I've heard throughout training. It's used in meetings, it's used in papers. It sort of gets to the fact that this operation gets

done at high volume centers a lot of the time and people have mastered it and yet when this operation fails, it's because this anastomosis leaks. And that is it's sort of weak point, the bile duct does fine, the GJ does fine, we've gotten really really good at the dissection and even pushed the limits of dissection with more vascular techniques, but still this pancreas anastomosis, even when you felt like you have so much experience with this and really perfected it, it can still come back and get you, and so that's what that means with the Achilles heel. Yeah, definitely sounds like the key portion of the operation. So, Dr. Dua, when you Put together the anastomosis. What goes through your mind as you make the decision about How are you going to do it? What you're going to do? What are the key factors that you look at? I think it's important to recognize that some of these pancreatic jejunostomies will leak, right? That's just a common portion. So even though it is the Achilles heel, we have to acknowledge that certain glands Small duck size, very

soft texture glands, sometimes glands that have experienced a lot of pancreatitis. Those are all more prone to leak and so I think it's important to have that discussion with the patient that even if you are at a very high volume center with very very good technique that still up to 20 percent of these can leak. I think the key part is really recognizing that and doing the best we can from a technical standpoint to minimize that. So the things, like I mentioned, that I look at for factors are duct size, so very small ducts, soft glands. Sometimes I also pay attention to how much atrophy they have had as a result of the tumor how much gland will be left over after a Whipple. So many patients after the resection may not have all of their pancreas that extends to the splenic hilum, and they just have a very, very small amount. So those are all different factors that I think about when thinking about which anastomosis to use or some of the other

techniques that we'll discuss. That's a great summary on the important things for everyone to consider. So, let's say that you do have a patient who has like a very, very small duct. It doesn't lend itself to these standard approaches for reconstruction. What are some alternative approaches, John, that you might consider in a patient who has a small duct? Right, so, like Dr. Du was alluding to, if you're doing your anastomosis, you're planning to do your anastomosis, but then you really can't find a pancreatic duct, or you have a very atrophic pancreatic gland, meaning that it's small, the patient has already been on pancreatic enzyme replacement therapy, it's not unreasonable in that situation to just ligate the duct or staple off the remnant gland and not perform an anastomosis. And you still leave a little bit of parenchyma, so hopefully they have some endocrine function preserved. But. You're not really relying on that pancreas to drain and produce any sort of digestive enzymes since it hasn't been for a long time. And that was sort of what actually prompted this paper was in residency a couple times

we had encountered this and we sort of wondered about what to do because, you know, certainly we didn't want to leave a duct undrained. But we looked through the literature and we had actually done these ligations and it works pretty well in this sort of situation. I think that's a really good option, you know, we always try our best if we can even find a very, very small duct to do some kind of an osmosis or put a stent in, but it's good to have these other options. Another one that John mentioned in the paper was bringing the pancreatic stump to the stomach, so that's called a pancreatic gastrostomy. So sometimes if you know the duct is there, but you can't exactly find it to do an actual connection to it, you can bring the whole face of the pancreas into the stomach. And he wrote about the technique, but essentially it does involve making an anterior gastrotomy and a posterior gastrotomy because you're essentially pulling that cut edge into the stomach to just allow it to drain into the stomach. And so what about if you have a duct that has a very soft

texture, a texture that doesn't lend itself well to an anastomosis or kind of might make you lose sleep at night if you try to anastomosis it the normal way? What are some options for that? Yeah, I mean sometimes you just can't get around the fact that the pancreatic parenchyma is really soft and that's these sort of cases where you're anticipating that there's going to be a leak even if everything goes technically very well. So, you know, it's not necessarily the duct around pancreas texture is what's going to drive your decision making about the type of anastomosis. It's really that duct caliber. So if you're not able to do a duct mucosa anastomosis, then you're thinking about a pancreatic gastrostomy or if you have familiarity with an invagination technique that's sort of achieving the same thing where You're just trusting that wherever the duct, as minuscule as it is, is going to drain from the face and go right into a lumen. And you're not dealing with a new connection that needs to epithelialize because you can't put stitches into a duct that's, you know, sub millimeter, intrivial in size. So it's less about the texture

of the gland, but more about the duct caliber. So basically what I get out of that is this, the duct is almost like your rate limiting step. That duct caliber and that duct size is more important than the actual texture of the pancreas. But what about if someone has an atrophic pancreas? How does that change your decision making in terms of your anastomosis and are there other options for those kinds of patients? Sure. I think that atrophy of the gland can happen, you know, they have a significant blockage of their duct with the tumor and sometimes these patients are going through a lot of neoadjuvant chemotherapy so over time that, you know, pancreas can atrophy. I still think that if there is a certain amount of atrophic pancreas at the end, you know, A lot of times those ducts are actually bigger because they, you know, have been blocked. So I will try and do an anastomosis if I can, even to a small segment of pancreas, because I think it can contribute to providing some endocrine, exocrine function. But if I absolutely cannot see the duct at all, like if it's the same

thing that you mentioned, very small duct and a very, very small remnant because of the atrophy, then in those cases you can just So, a lot of times we use the same sort of staple load that we would when we were doing distal pancreatectomies it's a black reinforced staple load, and really you're just closing that cut edge that you would normally do in an osteomosis too, but you're essentially just sealing it off at that point. And it will continue to atrophy and rarely cause as much sequelae after that. What about a bailout situation where you feel like you can't perform a safe anastomosis, whether it's because maybe the patient's not doing well on the table, you need to get out of there, or things are just not looking good. What are some of your options then to complete the operation there? Right, so, you know, this is a rare situation. I'd say this happens very infrequently, maybe a few times in, you know, a 5 year stretch, but, you know, the tumor might be more

involved than you think it is You're trying to do your dissection, there's bleeding, and you, you just really can't safely do a reconstruction because the patient, like he says, is hemodynamically unstable and getting sicker. And so, one technique is to actually place an externalized drain in the pancreatic duct lumen and have that mature out through the skin as a controlled fistula. And so in that situation, you're trusting that, you know, most of the exocrine fluid that the pancreas is making is going to be diverted. And this patient is going to be on enzyme replacement therapy, but at the same time, it's not, you're not putting that patient at risk for an intradermal leak, which could cause sepsis and they could get even sicker from, so you just use a really small silastic catheter. And just place it in the duct, try to secure it some way to the duct parenchyma, and then you just bring one end out through the skin, and then that just glues the pancreatic face up to the abdominal wall, and hopefully that will contract down and become atrophic and they won't have so much

output every day, but that is one way to get out of DODGE if you need to quickly. Dr. Du, is there another technique that you maybe use? I think for me, my typical bailouts would be I think you, I mean, you do what you're most familiar with, right? So I think, in general, I don't tend to use as much external drainage. I don't, I think that it is a good a good solution to have, just in case, for sure. But I tend to use the pancreatic gastrostomy more as a bailout because I, I feel like at least the whole face and the duct that's somewhere in there will be drained, you know, back into the GI tract. But I, I think it's important to discuss stents like we were because the external drainage is certainly an option that's used. Internal drainage is something that we typically use for small ducts and, you know, soft pancreas texture. I would say that it's not necessarily used to, you know, prevent the leak. If you're going to leak, it's going to happen. But I think in those high risk gland scenarios, it certainly will help direct flow, right? So, like, it, it helps direct flow. Kind of provide the path of least resistance

and direct flow into the bowel more and those stents can be placed at the time of the anastomosis sort of before you tie the anterior row of your duct to mucosa you put one of these Five French stents that goes partly into the duct and then kind of has a pigtail into the lumen of the bowel and then you finish your anastomosis on top of that. And those tend to stay for a couple weeks. Sometimes they may slide out and usually it will just pass. But I think in the initial healing period sometimes that can help minimize the sequelae of if you do have a leak. Yeah, you can use the externalized drainage in that same way too, and that was a technique I used a lot in residency, where we actually had a trans anastomotic stent that was then pa it was sewn with a little purse string as it came out through the bowel near the staple line of our sort of candy cane for our, our biliopancreatic limb. And then that is matured through the

skin in the same way you would do in this sort of damage control situation. But unlike the trans anastomotic stent that Dr. Doolittle was talking about, which Generally passed on their own and, and patients don't even know that they've fallen out. This is something you have to pull out eventually, but you can't pull it out too quickly because you need a little epithelialized tract so that everything will collapse down and they won't have a pancreatic fistula. Understood. Yeah, it's an interesting thought. I did have just one question about reading the article. John, Dr. Dewey, you guys mentioned the use of a total pancreatectomy. What situation would you use that for, do that in? Yeah, actually, so Dr. Gadon, who's one of the authors on the paper, brought this up There is a series that's published, I believe, out, it's out of Europe, I think it's out of Italy, where they were doing a lot of vascular reconstructions and in, especially with arterial reconstructions, if you have a pancreatic fistula or a leak in that situation, it can be catastrophic because you can get a pseudoaneurysm or blowout. And so they wrote about their experience actually doing total pancreatectomies in the setting of these

really extensive revascularization procedures. And so that's to minimize the risk of a leak and a pseudoaneurysm from happening. But that's, that's the instance that that sort of comes up. Makes sense. Yeah, I would agree with that. It's not something that we standardly use but I think that it, the risk of when you have pancreatic enzymatic juice near, Graphs or reconstructed arteries can really be very, very high risk situations. So in those particular cases, especially if you're dealing with a very atrophic remnant, then it's one way to sort of take that out of the picture. All right, well I think we hit on all like the major points of the papers. Anything else you guys want to add? I octreotide, we didn't discuss that mitigation strategies, yeah. That's a great topic, good mitigation strategies that We should be aware of trying and reduce the risk of a fistula or a leak post-op. Yeah, so there, I mean, a lot of groups have tried to do randomized control studies to

look at this. You know, Dr. Peter Allen at Duke has looked at Pide and octreotide. Leah Co at University Hospitals in, in Ohio has looked at antibiotics. And then in Norway they actually looked at steroids and giving preoperative steroids. And so all of these things have been studied. Carlos Fernandez, who is an author on this paper, will tell you that externalized drainage is the one and only thing that reduces leak. And he believes that very, very strongly, and I think there's great evidence to support that. But, Arc'teryx has what we use here. Yeah, I mean, I think we use it, like I said, it's not going to necessarily prevent your leak, right? It's about the technique and the gland texture and the duct size, but it certainly can minimize the leak complications or sequelae. So it's just like Ducts as, I mean duct stents as well. So we tend to use octreotide 200 micrograms given at the time of the anastomosis and then we'll continue it In an IV fashion every eight hours for five days. So

as they're recovering in the hospital We continue to do this unless of course, you know, they leave prior to that time period but I think it in general You know may help minimize secretions flowing through the anastomosis or a decreased blood pressure The volume of the leak, for sure. I, I think a side effect that we have to be, you know, aware of is that it can cause the patient's nausea sometimes. So, in the first couple of days, sometimes they can feel a little nauseated with it. One, one other thing is that we really use drains liberally with PJs. So, in all of these anastomoses, most of the time, unless it's a really slam dunk, hard pancreas with a giant duct, We're leaving at least one drain and some centers will leave two drains and that's routine practice and whatever is routine practice is what you should go with but it's very important to have a drain around if you have any concern about synastomosis because you want to control the fistula. Yeah, that's a good point. We didn't even mention that. So this is, we're talking about the

surgical external drain, so nothing trans and asthmatic or anything, just your surgical drain that comes out and, and I agree, I think For the most part, the gastrojejunostomy, even the hepaticojejunostomy, those are pretty resilient anastomoses with good blood supply, but you know, with any concern, then we definitely put a drain near the pancreaticojejunostomy, and that can be managed even in a very short fashion, you know, for a lower risk pancreatic anastomosis, we tend to check the amylase level in it pretty quick, even as early as day three. And if it's low or we felt like the anastomosis was very secure, then we're also very quick to take them out so that they don't cause, you know, further complications because certainly having a drain in for a long period of time increases the rate of also fistula or other problems, you know, migration of the drain, pseudoaneurysms, things like that. So, I think it's important to have it. And then also be timely in the way that you use it. That's a great point. As someone who you're very concerned about, it's better to leave the

drain in rather than to be cavalier. Well, I think that this has been a great conversation, and we just want to leave our listeners with some key takeaways. The first is that technical failure of the PJ anastomosis is the Achilles heel of the Whipple procedure, as it can be a significant source of perioperative morbidity and mortality, which has been well documented in the literature. End to side dunking, or Bloomgaard style duct to mucosa anastomotic techniques, are the standard reconstructive approaches during the Whipple, but as we've mentioned throughout this episode, you should do what you are most comfortable slash familiar with. There are several key characteristics to consider when deciding on the approach for the anastomosis that may preclude these standard approaches and require one of the alternative approaches we've discussed. These factors to consider include a small pancreatic duct size. Atrophy of the pancreatic gland and or a small remnant pancreas, unfavorable pancreatic texture, and some patient factors. And sometimes you may not even

be able to perform an anastomosis. And as we've discussed today, it's important to have a bailout method in the back of your mind. And lastly, as Dr. Dua mentioned, up to 20 percent of patients can have an anastomotic leak. And there are some mitigation strategies like steroids, external drainage, or octreotide. And also, if you have any concern about your anastomosis, it's best to leave a drain in place. Once again, thanks for tuning in and listening, and as always, dominate the day.