POEM Contraindications and Pitfalls

Fig. 7.1
(a, b) Demonstration of luminal narrowing secondary to peptic strictures after Heller myotomy. (c, d) Demonstration of luminal narrowing secondary to achalasic sphincter


Fig. 7.2
Patient status after prior remote Heller with good initial response now referred for possible POEM for recurrent dysphagia and poor emptying on barium. HRM shows minimal residual LES pressure (mean LES pressure of 9.6, residual pressure of −1.2!) and common cavity between the stomach and esophagus. This patient would not benefit from POEM and should be considered for possible esophagectomy


  • 3.

    Achalasia in the setting of prior bariatric surgery. Patients that have undergone bariatric surgery such as gastric bypass or sleeve gastrectomy may suffer from idiopathic achalasia or achalasia secondary to the bariatric surgery itself [13]. POEM has been anecdotally reported to be efficacious in patients that have had prior bariatric surgery. However, based on our preliminary unpublished experience, some caution is indicated. The excellent efficacy of POEM in disrupting the LES (which underlies its dramatic and durable relief of dysphagia and, on the downside, clinically relevant GERD in approximately 30–40% of patients) may be a cause of concern in patients with prior bariatric surgery. POEM may significantly increase the severity of GERD in sleeve patients and may facilitate regurgitation from the surgically restricted small-capacity gastric pouch in bypass patients and the high-pressure narrow stomach in sleeve gastrectomy patients. In these patients, severe GERD or regurgitation can diminish any post-POEM quality-of-life improvement from dysphagia relief. Effective management of such symptoms can be difficult since surgical revision or antireflux procedures are limited in these patients.


      Procedural Pitfalls

      Preprocedure Preparation

      POEM is unique in that it is a mediastinal surgical procedure that can be performed in endoscopy units where procedure protocols may be somewhat less stringent than in formal operating rooms, as they are geared toward traditional endoscopic procedures in which reaction time delays, omissions, or other errors are much less likely to result in mission-critical disruptions compared to surgery. To minimize serious and potentially life-threatening events, it is important to replicate operating room protocols including detailed equipment checklists that ensure that all devices that may be needed, particularly ones that may be needed infrequently (such as overtubes, stents, and specialized clips or sutures) or emergently (such as Veress needles or angiocaths for decompression), are readily available. Detailed “time-out” protocols are essential, including, for example, confirmation that air insufflation has been turned off and appropriate antibiotics have been administered. It is instructive to look at air insufflation as an example of a potentially catastrophic event that may result from a minor oversight that would be of little import in most other endoscopic procedures. The high frequency of adverse events and severe adverse events resulting from use of air rather than CO2 was amply illustrated in a study by a group that intentionally utilized air in their first 119 POEMs [14]. Endoscopy consoles at the present time still have air as the default insufflation setting with add-on appended equipment required to use carbon dioxide. To avoid air insufflation one needs to ensure prior to the procedure that the unit’s default air insufflation is switched off. Including the “air switch-off” step in the standard procedure “time-out” minimizes the risk of inadvertent air insufflation which can occur even in expert centers. Inadvertent air insufflation was reported as the cause of the single occurrence of pneumothorax requiring chest drainage in a recent study by Inoue et al. following a series of 500 POEM cases [15]. This is not a “learning curve” related event and is most likely to occur when the procedure becomes more routine, and vigilance by the operator and support team decreases. We recommend including an “air-off/CO2 on” check to the preprocedure “time-out”, as we have done at our institution, and positioning an angiocath and betadine wipes at a standard location within immediate reach of the operator to minimize any delay in emergent venting of capnothorax or capnoperitoneum.

      Other routine preprocedure setup tasks should be included in the preprocedure checklist. For example, routine taping of the distal cap attachment with a water-resistant tape can avoid dislodgment of the cap in the submucosal tunnel, which can result in a quite cumbersome and time-consuming extraction of the dislodged cap [16, 17].

      It is also important to have a highly trained dedicated team. Since achalasia is a rare disorder and POEM is performed in small numbers in most centers, errors can result without a dedicated team. The anesthesiology team needs to be prepared for circumstances that may result in severe morbidity. For example they need to anticipate the presence of massive amounts of food debris in patients with advanced or end-stage achalasia and severely dilated esophagus and preemptively apply cricoid pressure and rapid sequence intubation. At our center, on some occasions, we employ additional maneuvers such as semi-erect intubation in certain severe end-stage patients with the history of aspiration episodes during prior endoscopies. An endoscopy team that is unfamiliar with POEM may also fail to correctly interpret signs of pneumothorax or pneumoperitoneum. Delay in diagnosis of such conditions may result in cardiac arrest, whereas prompt recognition allows correction by desufflation or venting with an angiocath, thus minimizing morbidity. Anesthesiologists familiar with traditional endoscopic procedures performed under general anesthesia but unfamiliar with POEM need to recognize signs of emerging tension pneumothorax or pneumoperitoneum (e.g. difficulty in ventilating the patient and high airway pressures) versus endotracheal tube displacement by the endoscope, bronchospasm, or inadequate paralysis. The latter sort of differential diagnosis would be appropriate for a traditional endoscopic procedure performed under general anesthesia such as endoscopic retrograde cholangiopancreatography, but in the case of POEM, it could result in delay in diagnosis of tension pneumothorax or pneumoperitoneum. If having a dedicated POEM-operative team is not feasible, it is incumbent upon the surgeon or gastroenterologist to discuss with the anesthesiologist the potential POEM anesthesia pitfalls prior to the procedure.

      Tunnel Initiation and Orientation

      On insertion of the endoscope, one may encounter a situation where excessive loss of insufflation from the upper esophageal sphincter is encountered with the resultant inability to properly distend the esophageal lumen (or submucosal tunnel lumen later in the procedure). In such cases, insertion of a short esophageal overtube with an air-tight silastic ring around the shaft of the endoscope may be helpful (Guardus Overtube, US Endoscopy, Mentor OH).

      After removal of any debris from the esophagus, we recommend irrigation with at least 500–1000 cm3 of saline based on studies regarding NOTES indicating significant reduction in bacterial colonies after copious irrigation with similar reductions, whether or not a disinfectant was included in the irrigant [18].

      During this step, a common mistake involves aggressively and repeatedly inserting the endoscope through the GEJ, which in patients with an extremely tight LES results in mucosal tears compromising the mucosal flap which serves as the essential barrier that prevents leaks in POEM.

      Careful measurement of the location of the GEJ from the incisors is required to determine the proximal and distal extents of the tunnel and myotomy. A common pitfall here involves overly rigid adherence to standard recommendations such as initiating the tunnel at a certain fixed distance proximal to the LES to the GEJ. Recent data suggest that a standard surgical myotomy of at least 8 cm may be longer than necessary in the esophageal body for nonspastic achalasia patients (type 1 and 2) [19]. Employing this approach in patients with advanced disease and dilated esophagus with mild (S1) or severe (S2) sigmoidization is likely excessive since these patients have a very short obstructing segment consisting of the LES only. Extension of the myotomy proximal to the LES on the expansive esophageal body may be of no benefit and even predispose the patient to diverticulum formation in the area of weakened muscle. Furthermore, in these advanced stage patients, severe angulation and lumen-indenting folds in the dilated distal esophageal body can make POEM technically challenging unless tunneling is initiated close to the LES distal to the meandering expansive lumen. With proper technique, POEM can provide substantial symptomatic improvement even in patients with sigmoid esophagus [20, 21]. Other scenarios that may complicate POEM can also be alleviated by judicious selection of the initiation site. Orientation that would require traversing areas of ulceration, diverticula, severe angulations, or a prior Heller myotomy should be avoided. It should also be noted that initiating the tunnel in an area that may make tunnel initiation and, importantly, tunnel closure difficult should be avoided even if this requires selecting a more proximal site by creating a longer submucosal tunnel than required for the planned myotomy length. This approach allows one to avoid areas with scarring and scant submucosa from recurrent ulcerations due to food stasis, areas in which a sigmoid esophagus “dives posteriorly” (making contact with the endoscope for a posterior POEM tenuous) or “ascends anteriorly” (causing the endoscope to be perpendicular to the wall or nearly retroflexed rather than in the optimal tangential position). Selecting an initiation site that is more proximal, away from areas where chronic food stasis may have caused the mucosa and submucosa to be thickened, may also facilitate closure as reviewed below in the “Tunnel Closure” section of this chapter.

      Although a specific discussion regarding anterior vs posterior orientation is offered below in the “Submucosal Tunnel” section, we should note here that there is no consensus regarding the optimal orientation among expert centers with some favoring the anterior approach popularized by Inoue and some the posterior approach favored by the group in Shanghai and our group [1] (Fig. 7.3).


      Fig. 7.3
      POEM orientation among pioneering centers polled in the International POEM Survey (IPOEMS), including all centers having performed >30 POEMs at that time. Only two centers favored a posterior orientation (Mineola, Shanghai) at that time. (Figure from Stavropoulos SN, Modayil RJ, Friedel D, Savides T. The International Per Oral Endoscopic Myotomy Survey (IPOEMS): a snapshot of the global POEM experience. Surg Endosc. 2013 Sep;27(9):3322-38. doi: 10.1007/s00464-013-2913-8). Permission obtained

      Initial Submucosal Injection

      In achalasia patients, injection into the submucosa may be difficult due to alterations in the thickness of the layers of the esophageal wall. For example, in patients with long-standing achalasia, the entire wall of the esophagus may be severely thickened including the mucosa which may result in inadvertent injection of the deep mucosa superficial to the muscularis mucosae. In this case, attempts to establish a submucosal tunnel will be in vain unless the operator appreciates that what he/she considers to be muscularis propria is in fact a hypertrophic muscularis mucosae and proceeds to incise it (Fig. 7.4).


      Fig. 7.4
      Patient with long-standing achalasia with thick muscularis mucosae (full arrow) that can be confused for muscularis propria. Incision of this thickened muscularis mucosae reveals the submucosal space (dashed arrow)

      In severely malnourished patients in whom the submucosal layer may be very thin and in some early nonspastic achalasia patients with thin esophageal wall layers, the operator may inadvertently inject deep to the submucosa into the muscularis propria, adventitia, or mediastinal pleura. This may be recognized by appreciating that the resultant bleb is flatter than usual and has a pale white coloration with very little blue hue seen due to lack of transmission of the color of the injectate through the thickened overlying layers (Fig. 7.5). If not recognized, this deep bleb can result in layer confusion since the injected areolar tissue of the adventitia and pleura can mimic the submucosa. This can induce even experienced operators to incise through the muscularis propria and start tunneling in the adventitia or pleura plane deep to the muscularis propria with high attendant risks to injury to adjacent organs [22]. Once this is recognized, correction would necessitate closure of the full thickness perforation leading to this deep mediastinal tunnel with a secure modality such as suturing [22].


      Fig. 7.5
      Initial submucosal injection. (a) Shows a flat pale mount with very little blue hue suggesting that an inadvertent deeper injection into the muscularis propria or beyond has been performed rather than the desired submucosal injection (demonstrated in (b) as a markedly raised translucent bleb)

      Regarding the injectate used for submucosal injection, unlike endoscopic submucosal dissection, most operators avoid epinephrine due to the risk of necrosis of the devascularized mucosal flap. Such severe necrosis has been reported by one group [3].

      Initial Mucosal Incision

      Optimal incision is important in order to facilitate tunnel entry and facilitate secure closure at the conclusion of the procedure.

      To avoid oozing from the edges of the incision from sizable mucosal and submucosal veins present in the midesophagus, we recommend selecting a site with the lowest density of such visible vessels and using a current with a significant coagulation component (e.g. dry-cut current in the ERBE VIO generator) for the initial incision.

      As noted above, site selection and orientation should also take into account esophageal morphology in that area. Extensive nodularity from chronic food stasis likely represents cycles of ulcerations and healing that may make establishing a submucosal tunnel difficult. Even mild angulations of the esophagus may make tunnel entry and closure technically difficult.

      Generally, it is accepted that a longitudinal incision allows easier closure with endoscopic clips than a transverse incision. Our group has used endoscopic suturing for closure in the last 250 POEMs and we prefer a transverse incision to avoid luminal narrowing. It should be noted, however, that even for closure with endoscopic clips, at least one group has advocated initially a transverse incision [23] and more recently an “inverted-T” incision [24]. Their argument, as we understand it, is that a transverse incision allows easier entry into the tunnel and also allows better escape of CO2 from the tunnel, thus potentially decreasing insufflation-related adverse events. One would think, however, that this might also result in poor tunnel distension and decreased visibility. Furthermore, the closure issue remains since placement of clips along a transverse incision in the esophagus is more challenging.

      The initial submucosal dissection at the entry site should be made close to the muscularis propria in order to avoid denuding the underside of the mucosal flap of submucosa, resulting in a structurally weakened flap at the tunnel opening that can tear during endoscope manipulations within the tunnel. Such tearing results in a much larger opening with devitalized torn edges that may be hard to approximate securely at the time of tunnel closure.

      Initial Entry into the Submucosal Space and Tunnel Initiation

      Operators early in their learning curve may have some difficulty in achieving initial entry into the submucosal space. As noted in the “Initial Incision” section, it is helpful to select a propitious entry site based on flat favorable morphology, lack of visible vascularity, and lack of submucosal scarring. Methods that may assist in submucosal entry include use of an oblique transparent distal cap attachment as initially used by Inoue or a tapered distal cap attachment (e.g. ST Hood; Fujifilm, Tokyo, Japan). In our first few POEMs in 2009–2010, we employed balloon dilation to establish the submucosal tunnel [25]. This technique greatly facilitates entry into the submucosal space and also carries the risk of balloon catheter perforation of the muscularis propria or mucosa during the blunt insertion prior to inflation [26].

      For posterior POEM, which is our favored orientation currently, entry into the tunnel may be hampered by the much lower maximum down-angulation versus up-angulation capacity of gastroscopes (e.g. 90° vs 220° for Olympus GIF-HQ190 gastroscope). Therefore, we have developed and taught a technique that facilitates posterior entry which consists of reversing the orientation of the endoscope during entry by torqueing 180° while simultaneously using irrigation to retract the mucosal flap (demonstrated in Video 7.1).

      Submucosal Tunnel Dissection

      Submucosal tunnel dissection is usually the most time-consuming and challenging portion of POEM (e.g. mean duration of 44 min for submucosal access and tunnel creation vs. 25 min for the myotomy in a recent US study) [27]. Less intraprocedural bleeding and faster procedure durations have been reported with the use of the multifunctional ERBE hybrid T-type knife (ERBE, Tubingen, Germany) which allows injection and dissection by the same device compared to the triangular tip knife (TT knife, Olympus America, Center Valley PA) [28, 29]. However, neither effect appeared to have a significant impact on clinical outcomes, and, therefore, use of the hybrid knife is not a substitute for careful, precise, deliberate dissection which is the best strategy for preventing errors such as accidental mucosal injuries and excessive bleeding. Novice operators often attempt to use blunt dissection by forceful endoscope insertion to achieve faster or easier tunnel dissection. Although this technique is often successful in the less vascular and softer porcine submucosa used in preclinical training, in humans it can have the following undesirable consequences: (1) Multiple small submucosal veins which would normally be obliterated by the electrosurgical dissection of the submucosa without any specific hemostatic maneuvers required can be avulsed via the technique of mechanical blunt dissection which then necessitates time-consuming coagulation of multifocal oozing. (2) Unrecognized buckling of the endoscope at the tunnel insertion site which may result in tearing of the opening may in turn make closure more challenging (Fig. 7.6) (3) “Muscle splitting” especially in the area of a tight LES, an important pitfall of submucosal dissection, is discussed in detail below.


      Fig. 7.6
      Tearing of the tunnel opening caused by aggressive maneuvering of the endoscope during submucosal tunnel dissection. White arrows indicate the cautery changes that mark the original distal extent of the tunnel opening. The blue arrow demonstrates the distal extent of the now much larger opening after tearing occurred (note the absence of cautery along the tear confirming that this extension was caused by mechanical tissue tearing rather than electrical energy)

      Although submucosal tunnel dissection in the esophageal body is usually straightforward, occasionally certain challenging scenarios and pitfalls can occur. One such scenario is that of thin, absent, or fibrotic submucosa thwarting the endoscopist’s attempts to create a submucosal tunnel. Aborted POEMs due to this phenomenon have been reported anecdotally even by expert operators. However, the best described series of such cases comes from a group in Rome, Italy [30]. This group reported a 6% early termination rate on their first 100 POEMs, with all five cases halted due to this phenomenon. We submit here an excerpt from their report as it describes this pitfall of submucosal dissection. They state that “In 5 cases, the procedure failed because of inadequate lifting of the mucosa and the impossibility to proceed with submucosal dissection. Two patients had received radiation therapy for breast cancer. The esophageal wall appeared very thin, sclerotic, and any attempt at submucosal injection of glycerol solution ultimately failed, more likely because of severe submucosal fibrosis after radiation. The other 3 patients had no peculiar clinical history. Nevertheless, in these patients, the mucosal lifting was incomplete and submucosal dissection impossible: 2 of these patients had a very dilated and tortuous esophagus, which additionally complicated submucosal lifting and dissection. Any attempt at dissection resulted only in a laceration of the mucosa, which was repaired with clips.” We focus the reader’s attention on the fact that in three of the patients described, there was no clearly identifiable cause for this “absent submucosa” phenomenon which occurred in 3% of the patients in this Italian series. Based on observations from our series of over 300 POEMs [31] with no aborted POEMs, this phenomenon can usually be overcome with the maneuvers described below. It is encountered in patients with long-standing disease and severe food stasis likely resulting in pervasive inflammation and cycles of mucosal injury and healing that cause widespread submucosal sclerosis. This is most prevalent along the posterior wall of the esophagus or in patients with severe malnutrition in whom the absent submucosa is probably a sign of a severe prolonged catabolic state. We suggest the following maneuvers to overcome this challenging phenomenon: (1) Abandon the original site where tunnel initiation attempts have failed and reattempt at a new tunnel orientation (e.g. move from the posterior wall to a lateral wall) and/or new location (more distally or sometimes more proximally to the initial site, attempting to target an area with the least amount of mucosal nodularity, thickening, or other surface abnormalities). (2) Use of the I-type Hybrid knife (ERBE, Tubingen, Germany). The I-type Hybrid knife, which we have used in our last 280 POEMs, delivers a saline injection at pressures of up to 1400 PSI, which is powerful enough to dissect tissue via a tiny 0.12 μm port at the tip of this straight knife. In our experience, this can often achieve enough injection to delineate a submucosal dissection plane even in cases of very minimal fibrotic submucosa. Needless to say, even though these maneuvers may make a seemingly impossible dissection feasible, it would still remain an expert-level, slow, meticulous dissection requiring as much patience as skill.

      Once the submucosal tunneling is initiated, a common pitfall involves spiraling of the tunnel dissection. Spiraling occurs due to preferential dissection on one flank of the tunnel more than the other and usually results in progressive clockwise rotation of the orientation of the tunnel. In patients with a relatively straight esophagus, it can be recognized by the operator as a progressive change in the angle between the long axis of the tunnel and the circular muscle fibers from a 90° angle to a more oblique angle [32]. Potential problems due to spiraling include the following: (1) Spiraling of the myotomy which results in a less powerful disruption of the ability of the circular muscle to achieve lumen-effacing contractions (2) Moving from an anterior POEM (2 o’ clock orientation) or a posterior POEM (5 o’ clock orientation) to a greater curvature-oriented POEM at a 7 o’ clock position. A greater curvature POEM is much more challenging as it involves dissection across the angle of His [33]. One simple methodology first proposed by our group to avoid tunnel spiraling involves placing a marker on the shaft of the endoscope that indicates the torque rotation of the endoscope that maintains the desired orientation within the tunnel which is illustrated in our open-access narrated POEM technique video [34]. Advanced sigmoidization constitutes one of the most challenging and time-consuming POEM clinical scenarios [20, 21, 35]. The main challenge in these patients consists of completing a properly oriented submucosal tunnel. Proper orientation can be so challenging in these scenarios that experienced practitioners in India have described inadvertent retroflexion of the endoscope during tunneling, with the tunnel making a U-turn just prior to the GE junction and leading back to the esophageal body. This was recognized and corrected by the use of fluoroscopy which this group has used and now advocates in difficult sigmoid patients to help maintain orientation [36]. We have found our endoscope shaft torque marker method to be adequate in these patients, but we feel that it is important to be knowledgeable about the full armamentarium of useful adjunctive techniques such as fluoroscopy that can help avoid POEM pitfalls, particularly early in one’s experience and in exceptionally challenging cases.

      Submucosal tunnel dissection in the area of the GE junction can be challenging due to two potential reasons: a very tight LES or fibrosis from a variety of causes such as prior biopsies (frequently performed by referring physicians to exclude neoplasia or eosinophilic esophagitis), reflux or stasis erosions and ulcerations, prior Botox injections, or prior surgery including Heller myotomy. Fibrosis encountered as the tunnel approaches the GEJ is best approached via a detour, whereby the direction of the tunnel is deviated to the left or right of the fibrotic area depending on which side is most convenient and provides the best submucosal expansion [37]. A very tight LES can present a formidable obstacle to tunnel extension into the cardia and may also complicate the myotomy portion of the procedure. The Chinese group from Harbin has proposed a POEM technique, whereby myotomy is performed without prior separate submucosal tunnel dissection achieved by injecting the submucosa and then cutting the muscle by dissecting it off the injected submucosa as the endoscope advances in a proximal to distal direction [38, 39]. This technique may be of value in the hands of experienced operators. In the hands of less-experienced operators, it may result in “layer confusion” with resultant “splitting” of the esophageal muscle, thus leaving an unrecognized, and thus uncut, portion of the LES on the underside of the mucosal flap. This allows us to segue into a discussion of muscle “splitting,” an important pitfall of submucosal tunnel dissection particularly in the area of a thick, tight LES. LES splitting is one of the two main technical causes of POEM clinical failures, with the other being inadequate myotomy extension onto the cardia to be addressed below. Muscle splitting is mainly an early learning curve pitfall which occurs as the novice operator, duly concerned about causing an inadvertent mucosal injury injects and dissects ever closer to the muscular layer, especially within the tight quarters of a high-pressure LES zone or in areas of scant fibrotic submucosa in the esophageal body. Splitting of the muscle may be initiated by excessive forward mechanical force with the endoscope in and ill-advised attempt to add blunt dissection with the endoscope to electrosurgical dissection in areas where the latter appears risky (such as segments with minimal submucosal expansion or a tight GE junction). Injection near the split muscle fibers can expand the fascia between circular muscle bundles, thus leaving some bundles attached to the mucosa camouflaged by injected fascia that mimics injected submucosa. Recognition of this pitfall of POEM can be difficult. It may be suspected in the following circumstances: (1) Once the myotomy is completed, the exposed cut edges of the LES are not significantly thicker than the cut edges of the muscle of the gastric cardia as is the norm. (2) Apparent premature entry into the peritoneal cavity, as heralded by exposure of omental fat while there is still a substantial high-pressure narrowing of the GE junction as assessed by intraluminal endoscope insertion. (3) Substantial residual narrowing and resistance to endoscope insertion at completion of the myotomy which can also be confirmed by functional luminal assessment using the EndoFLIP device (Crospon, Dublin, Ireland) [40]. Recovery from this pitfall is simple in theory but may require some experience in practice. The operator needs to “back-track” along the tunnel to the point where the muscle split originated. This can usually be determined by noting that the cut edges of the muscle appear thinner than expected and usually occurs in an area of difficult tunnel dissection. At that point, careful dissection of the underside of the mucosal flap is performed using ample submucosal injection, pure cutting current, and if necessary a specialized knife such as the hook knife (Olympus America, Center Valley, PA) to avoid injury to the mucosa. This delicate dissection exposes the true submucosal plane and isolates the split muscle bundles that remained attached to the mucosa. Figure 7.7 illustrates a case of muscle splitting in the distal esophageal body just proximal to the GEJ which was recognized and treated as discussed above. The technique for isolating and incising the missed muscle fibers in this case is demonstrated in the first half of Video 7.2. After correction of this pitfall, the second half of the video illustrates resumption and completion of the anterior POEM myotomy including full-thickness muscle dissection along the high-risk location posterior to the pericardial sac (discussed below).


      Fig. 7.7
      Anterior POEM with inadvertent muscle splitting in the distal esophagus during submucosal tunnel dissection. (a) Demonstrates circular muscle fibers at 7 o’ clock position (where normally the mucosa and submucosa forming the roof of the tunnel should be seen in an anterior POEM) in addition to the 2 o’ clock position (which is the expected location of the circular muscle fibers in an anterior POEM). This can also be seen in (b) (i.e. circular muscle fibers at both 2 o’ clock and 7 o’ clock positions) as the endoscope is withdrawn in an attempt to identify the area of the dissection where the inadvertent muscle splitting commenced. (c) Illustrates recovery from this pitfall as the split fibers have been incised and the proper dissection plane in the submucosa has been re-established

      A number of POEM submucosal dissection pitfalls relate to bleeding. Acute bleeding can be classified as minor––usually resulting from inadvertently or inexpertly divided veins and very small caliber submillimeter arteries in the esophageal body––or major, usually resulting from accidental injury or inadequate coagulation of large arteries in the 1–2 mm range which are generally encountered in the cardia and represent branches of the left gastric artery that penetrate through the muscularis propria and arborize to supply the overlying mucosa and submucosa. The best approach to intraprocedural bleeding is prevention by identification of vessels and pre-emptive coagulation. Submucosal dissection, particularly in the area of the cardia, should be performed with short superficial swipes of the knife that ensure that the bundle of submucosal fibers being cut does not harbor undetected vessels. For this reason, it is important to avoid injection solutions that are too dark due to excessive blue dye and may prevent visualization of submucosal vessels. Treatment of bleeding is inferior to prevention for a number of reasons: (1) Even if successfully controlled, bleeding episodes can result in significant prolongation of the procedure time, since identification of the bleeding vessels and effective treatment can be quite time-consuming. (2) Copious bleeding can stain the submucosa red, which can make submucosal tunnel dissection substantially harder since the usually pink/tan underside of the mucosa and submucosal vessels do not appear distinct from one another (Fig. 7.8). (3) Multiple poorly targeted coagulation efforts resulting from the suboptimal visibility conditions of an active bleed can result in mucosal thermal injury or deep injury to the muscle and adjacent structures or at a minimum heavily coagulated, contracted, or even charred tissue. This hinders progress since the submucosa needs to be carefully dissected before a clean submucosal dissection plane can be re-established.


      Fig. 7.8
      Clot and extensive red staining of the submucosa after a large arterial bleed that required several minutes to identify and control. Red staining of the submucosa hinders proper identification of the submucosal dissection plane and identification of vessels within the submucosa that should be avoided or pre-emptively coagulated, thus predisposing to further intraprocedural bleeds until a clean unstained submucosal plane can be recovered distally as the tunnel is extended

      The endoscopist needs to distinguish arteries from veins since even small arteries generally require more coagulation treatment with graspers rather than simply using the tip of the endo-knife (which conversely is adequate for all but the largest veins when properly applied). Figure 7.9 illustrates the differences in the appearance of veins (generally larger, more cylindrical, more compressible with a deeper red color than arteries) and arteries (smaller, flatter, paler, sometimes with detectable pulsations, and with well delineated pale white borders representing their thicker muscular wall). Proper coagulation technique using the tip of the knife (Fig. 7.10) involves first heating the vessel indirectly by addressing the submucosa surrounding it and only proceeding with the division of the vessel once it has been desiccated and its lumen obviously obliterated. This avoids the potential for electrosurgical energy applied directly to the vessel, resulting in the division of the vessel prior to luminal sealing. A coagulation grasper should be used rather than the tip of the knife in the case of arteries (including the large arteries in the cardia), where the rapid luminal flow results in a powerful heat sink effect that can only be overcome by using a grasper to coapt the walls and disrupt blood flow prior to coagulation. A coagulation grasper should also be used for vessels under tension being stretched between their origin at the muscle layer and their insertion in the mucosa by the presence of the endoscope and insufflation within the emerging submucosal tunnel. In such vessels under tension, attempted coagulation with the tip of the knife may result in tearing of the vessel as soon as the structural integrity of its wall is weakened but prior to effective sealing of the lumen of the vessel. In fact, veins under such tension are fragile enough that injudicious use of suction via the endoscope can injure them and cause bleeding emphasizing the importance of gentle suction within the tunnel (unlike the customary use of suction in traditional luminal endoscopy [41]. Proper pre-emptive coagulation technique with the coagulation grasper (Fig. 7.11, Video 7.3) involves skeletonizing the entire circumference of a large vessel or multiple vessels within a vascular bundle. They can be grasped followed by extensive coagulation using a coagulation current algorithm that minimizes spread, and avoiding mucosal injury. The coagulation should be continued until impedance sharply rises and energy delivery sharply drops indicating tissue desiccation. Only then can the vessel be safely divided to proceed with tunnel dissection.


      Fig. 7.9
      (a) Illustration of a penetrating vein, larger, cylindrical bulging, soft, compressible with deeper red color. (b) Illustration of a penetrating artery which is smaller, flatter, firmer, often with visible pulsatile flow and paler red color often with pale white borders, annotated with white arrows here (an appearance caused by the thick muscular wall)


      Fig. 7.10
      Proper coagulation technique using the tip of the knife. (a) The submucosa surrounding the vessel is injected (b) A small incision is made with the knife in the submucosa next to the vessel (c) Electrosurgical energy is delivered to the vessel initially indirectly by targeting the submucosa surrounding it (d) Direct energy to the vessel to effect division of the vessel is only applied once the vessel has been desiccated and its lumen obliterated


      Fig. 7.11
      Proper technique for coagulation of larger vessels or vascular bundles. (a) A vascular bundle containing at least three penetrating vessels has been “skeletonized” using careful dissection with the knife (blunt dissection using the tip of the forceps can also be used to skeletonize vessels as shown in the accompanying video). (b) The isolated vascular bundle is grasped and coagulated with a coagulating forceps

      Massive pulsatile bleeding from an accidentally divided penetrating branch of the left gastric artery can result in substantial hemorrhage. Such events require immediate intervention since the tunnel may rapidly fill with blood, eliminating visibility and thus the ability to identify and effectively treat the source of the bleeding (Fig. 7.12). A coagulation grasper should be immediately available. We use the hot biopsy forceps rather than the Olympus coag-graspers for this type of predicament since the thinner caliber and larger jaws of the hot forceps allow for better suction of blood and irrigation of fluid and less need for precision placement of the jaws compared to other devices. Tamponade of the bleeding by exerting pressure with the tip of the endoscope should be applied before proceeding with definitive coagulation. Once the grasper is applied, irrigation should be employed to clean the site and ensure that the flow of blood has been arrested. If this is not the case, suction should be applied to remove fluid mixed with blood that may obscure visualization. Suction should be combined with insufflation and avoidance of excessive irrigation while the grasper is readjusted. Another useful technique involves placing a sticker on the shaft of the coagulation grasper that marks the proper length of insertion to the tip of the endoscope, which allows much faster insertion of the grasper since the operator is less concerned about overshooting with the grasper insertion and causing further injury. We also note that there may be a higher density of large cardia vessels when POEM is performed anteriorly (2 o’ clock orientation) rather than posteriorly (6 o’ clock orientation) which may shift the desired route for POEM to a posterior course [41].
    1. Sep 30, 2017 | Posted by in GASTROENTEROLOGY | Comments Off on POEM Contraindications and Pitfalls

      Full access? Get Clinical Tree

      Get Clinical Tree app for offline access