Today, several considerations about the resection of pedunculated lesions and closure after resection on the ASGE SUTAB tip of the week. Last week, we discussed evidence showing that for non-pedunculated lesions, the best evidence for prophylactic clip closure is for lesions 20 millimeters or larger and proximal to the splenic flexure removed by traditional inject and resect EMR using electrocautery. What about closure after resection of pedunculated lesions? The Multi-Society Task Force recommends that prophylactic treatment to prevent delayed bleeding using a detachable loop or clips is recommended for pedunculated lesions with a head 20 millimeters or a stalk 5 millimeters or larger. Here's a list of things to consider in the resection of large pedunculated lesions. First of all, you want to strive very hard to get it out on block, that is, cut it through the stalk in one piece and never piecemeal it. That's essential for proper pathologic assessment if cancer is present. The pathologist wants to bivalve the lesion down through the head and stalk, so one piece resection. Second, choose a snare of proper size, something big enough to get over the head, but often these lesions are in the sigmoid, which is narrow, not so big that it's unwieldy in a narrow space. Third, are there measures that will facilitate snaring, such as changing the position of the patient or floating the lesion up in a water column? For cautery, forced coagulation current is best, preferably with a microprocessor-controlled current. And for our discussion, something that I think is often not considered is, what if there's an unexpected cancer at pathology? How can I resect this lesion to optimize the chance to minimize the risk that the patient is going to need adjuvant surgical resection? And does that approach affect the choice of a detachable snare before resection versus clip closure after resection? For flat lateral spreading lesions like the one on the left, endoscopic assessment is pretty darn accurate in predicting the presence of cancer, and it's unusual to get an unexpected report of cancer from the pathologist. That's not as true when there is a sessile 1S or in a pedunculated lesion because these are bulkier, and it's harder to predict what's going on in the middle by assessment of the surface of the lesion. We get more unexpected reports of cancer, and that's what leads to my own preference to resect pedunculated lesions closer to the colon wall than to the polyp head. The argument for moving down the stalk toward the colon wall to perform the transection where the snare is, where the red arrow is, is that if any cancer is present, we've moved farther away from it. If instead we put a detachable loop or snare where the red arrow is, that's going to force us to go back up the stalk toward the yellow arrow and closer to any cancer present to perform the transection. So in a malignant pedunculated polyp, there are three histologic features that are critical in determining whether or not surgery is advised. One is the degree of tumor differentiation. Poor differentiation is bad, but it only affects about 10% of the polyps. Lymphovascular invasion is bad, but it only affects about 20-30% of the polyps. For the rest of the polyps, the critical factor is going to be the distance between the tumor and the resection line, and in fact, that's something that's going to be measured in every single polyp. One objection to putting the endo loop on first is that when we put it on, it's always going to force us to move up higher on the stalk, and it's going to shorten the distance between the tumor and where we end up cutting the polyp. So my own preference is to just cut the polyp first low on the stalk so that we can maximize that distance, and then after the resection, close the site with clips. So here's endo loop being cinched up against the stalk of a pedunculated polyp near the colon wall. I think these can be elegant devices. I like to see them placed. I enjoy placing them at times, although they are floppy and sometimes they're quite difficult to place. The problem here is that if there's cancer in the head of this polyp, then endo loop is going to force us up the stalk toward the head to perform the sneer transection. We have to stay several millimeters away from endo loop to avoid cutting it, and also just so that endo loop will stay on the stalk after the transection is completed. So if cancer is present, it's that movement up the stalk that creates the potential for a suboptimal outcome. So here's an overtly malignant sigmoid pedunculated polyp. You can see all the ulceration on the surface. This is obviously cancer. It's okay to remove this endoscopically because it's pedunculated. If it were sessile or flat, it would not be okay to remove it endoscopically, but we can remove this. Many times there's still going to be unfavorable histological criteria, but I wouldn't put a loop on this first because I want to stay as far away from the cancer as possible. So I'd prefer to sneer through the stalk, remove it on block directly, and then close the stalk with clips. So here's another overtly malignant pedunculated polyp, and you can see the ulceration on the surface of this, the very irregular surface telling us that this is going to be cancer, and it has a short stalk on it. I think it would really create problems to put a detachable snare on it. So I think the best approach to this lesion has a thick stalk. Closure for delayed hemorrhage is indicated, but we're going to proceed with snaring. We're going to do it close to the colon wall with the strategy of trying to maximize the distance between the cancer and the resection line because we know that that's going to be something that's going to be evaluated in the histologic resection. So we've performed the on block resection, and now for prevention of delayed hemorrhage, we can proceed with clip closure of this defect. So here's another pedunculated polyp in the sigmoid colon. It's lying in an attitude that doesn't make snaring easy, so we're filling the sigmoid colon with water, which makes it float up into the water column, makes snaring easier. And this is the more typical situation where there's no overt evidence of cancer. But again, we get a fair number of unexpected results from the pathologist of cancer in these lesions. So there's a rationale for approaching this lesion as if it might have cancer, which I think means avoiding the use of the detachable snare, going directly to snare resection close to the colon wall. And then after we've achieved this on block resection with a maximal length of stalk, then we'll use clips to close the site to achieve the goal of reducing the risk of delayed hemorrhage. Now, admittedly, this is a more expensive approach than the use of the detachable snares. But I think the potential for an improved oncologic outcome justifies this approach. And I think in general, it's also more efficient. It's just faster than placement of the detachable snares. What about prophylactic clip closure after removal of smaller lesions in patients on anticoagulants or non-aspirin antiplatelet agents? Currently, there are insufficient data to answer this question, and our clinical judgment must guide us. Next week, endoscopic tricks on achieving full clip closure on the ASGE SUTAB tip of the week.

resection

pedunculated lesions

closure techniques

snare size

oncologic outcomes