Novel Developments In Intestinal Endoscopy<br> prof. Oliver Pech –  Gastro Update Europe 2019, Budapest

Most western studies on endoscopic resection of early esophageal malignancy are related to adenocarcinoma. Now a new French multicentre study with long-term fol­low-up of endoscopic resection of 148 squamous cell cancers was presnted, 54% treated with mucosal resection and 46% treated with submucosal dis­section. Median lesion size was 20.5 m­m; 38% were piece-meal resected; depth of cancerous infiltration was m1– 2 in 64% and m3-sm in 36%; the procedure was complicated by bleed­­ing in 2%, perforation in 2% and led to strictur­­ing in 14%. Dur­­ing a median fol­low-up of 22 months (m), tumour recur­rence was observed in 14.2%. Significant risk factors for recur­rence, expres­sed as hazard ratios were: removal by mucosal resection (17); pT1m3-sm (3); need for additional chemoradiother­apy (7); non-curative resection (12). If at all pos­sible piece-meal resection should be avoided and for larger lesions or sus­picion for submucosal ingrowth, submucosal dis­section is to be prefer­red. Radiofrequency ablation of early sqamous cancer has also been evaluated in 35 Chinese patients. Complete response was seen in 86%. Tumour recur­rencies were observed in 20% but could all be treated with endoscopic resection. Meticulous examination of the resection spec­imens suggested that initial expansion of cancerous growth in the ductular epithelial lin­­ing of the submucosal glands, extend­­ing deeper than the muscularis mucosae and therefore not ablated, was the presumed cause of the cancerous recur­rence.

Although, it is com­monly known that longstand­­ing achalasia is a risk factor for squamous cancer, little is known about the impact of 3 yearly (y) endoscopic screen­­ing with lugol stain­­ing as evaluated in a cohort of 230 patients fol­lowed for a median of 56 m after various treatments (pneumodilation, surgical myotomy, peroral endoscopic myotomy (POEM), proton pump inhibitors (PPI) etc.). Stasis of food or liquids was observed in 27% dur­­ing one or more endo­scopies. In approx. a quarter of the endoscopies reflux esopha­gitis, usual­ly grade A and B, was detected. Low grade dysplasia was detected in 3% and evolved into cancer in 1.3%, result­­ing in an over­all incidence rate of sqamous cancer of 63/ 100,000 person-years. The incidence rate of low grade dysplasia started ris­­ing after 20 y of achalasia symp­toms and cancer after 30 y of symp­toms. Stagnant decompos­­ing food and fluid and bacterial overgrowth, lead­­ing to chronic mucosal injury and inflam­mation, are considered to contribute to the oncogenic process and should be the lead­­ing target of the various ther­apies.

Studies continue to be published regard­­ing the ther­apy with resection and/ or usual­ly radiofrequency ablation of neoplastic columnar metaplastic (Bar­rett) mucosa, confirm­­ing studies mentioned in previous highlights. Controversies remain whether patients should be intensively surveil­led in the absence of any dysplasia or whether asymp­tomatic patients should be treated with non-stop PPI antisecretory treatment. Whether there are prognostic phenotypes for esoph­ageal adenocarcinoma was prospectively evaluated in a separate American and British cohort. Interestingly the over­all survival in both cohorts was substantial­ly better when intestinal metaplasia was present in the peritumoral mucosa compared to those where intestinal metaplasia was absent. Thus adenocarcinoma without intestinal metaplasia appeared to have a significantly worse outcome. A confounder in the concept of two dif­ferent types of esophageal adenocarcinoma is the fact that it is almost impos­sible to rule out overgrowth of intestinal metaplasia by aggres­sive tumor expansion. All this reminds the ongo­­ing controvesy regard­­ing the requirement of intestinal metaplasia with goblet cel­ls as a prerequisite for proper dia­gnosis of Bar­rett type columnar metaplasia.

Usual­ly covered self-expand­­ing metal stents are used for pal­liation of advanced esophageal carcinoma, largely to prevent cancerous ingrowth. The full cover­­ing however favors the risk of migration. Does it help to leave both ends of the stent uncovered? In a control­led Dutch trial, almost 100 patients were ran­domized to ful­ly vs. partial­ly covered stents. No dif­ferences were found between the two stent types regard­­ing technical success and frequency of recur­rent obstruction. Recur­rent obstruction was more frequent for proximal cancers.

Benign strictures usual­ly respond to bougie/ bal­loon dilation ther­apy but some are truly recalcitrant and recur­rent. Is temporary stent­­ing with a bio­degradable stent with chronic dilation while tis­sue remodel­­ing occurs helpful in such patients with recur­rent benign esophageal strictures? In a control­led multicenter trial, 66 patients with esophageal strictures, most often anastomotic strictures, with prior dilation up to at least 16 mm were randomized. Dur­­ing the first 3 m significantly less repeat endoscopic dilations were neces­sary in the bio­degradable stent grou, but significance was no longer seen at 6 m. The median time to the first re-dilation was longer in the stent group (106 days (d)) vs. the control group (42 d). There were two perforations in the control group vs. five stent occlusions, two tracheoesophageal fistulas and one migration in the stent group. I am wonder­­ing what the proper indications (if any) are for bio­degradable stent­­ing and if radial incision with a needle-knife would not be a more attractive approach for anastomotic stricturing, realis­­ing however that no control­led comparative studies are available.

Bleed­­ing in the upper gastrointestinal tract remains a chal­lenge, particularly for recur­rent bleed­­ing of peptic ulcers. To find out whether over-the-scope (OTSC) clips are superior to standard ther­apy with clipp­­ing or coagulation, 66 patients were randomized in a multicenter study. Persistent bleed­­ing or recur­rent bleed­­ing within 7 d was significantly lower in the OTSC group (15%) vs. the control group (58%). Ulcer type, For­rest type, Rockall score, anticoagulation and type of prior endoscopic ther­apy had no significant impact. OTSC clipp­­ing requires adequate anatomic target­­ing of the bleed­­ing area which may occasional­ly be chal­lenging; perhaps monopolar hemostatic forceps may be helpful in such circumstances. A consecutive cohort of 112 Turkish patients with bleed­­ing gastroduodenal ulcers were randomized to treatment with monopalar hemostatic forceps with soft coagulation vs. hemoclipping. Hemostasis was obtained in resp. 98 vs. 80% with a median of three clips. Monopolar hemostatic forceps was succes­sful in all clip failures (persistent or recur­rent bleed­­ing within 7 d). Hemostatic forceps were more ef­fective than hemoclips independent from ulcer and bleed­­ing type. Other studies have confirmed the usefulness of hemostatic forceps in the treatment of acute bleeding.

POEM has become the minimal­ly invasive treatment of choice for achalasia in many centers. A well documented adverse consequence is the high frequency and often severe nature of reflux dis­ease, sometimes insuf­ficiently control­led with acid suppres­sant ther­apy. Surgical Hel­ler-type myotomy is often combined with partial (Dor-type) fundoplication. Is a comparable combination pos­sible by endoscopic means? This was attempted in a rather provocative proof of concept study by Inoue, the inventor of POEM. Distal to the myotomy, the proximal stomach was perforated, al­low­­ing endoscopic entry into the peritoneal cavity, position­­ing of a large endoloop over de fundus, anchor­­ing the most distal and proximal part of the loop with clips and final clos­­ing the loop creat­­ing a partial fundoplication. There were no complications and the complication appeared endoscopical­ly detectable in almost all patients 2 m later. Obviously many questions remain to be answered but such attempt is a nice il­lustration of the ongo­­ing exploration of novel therapeutic pos­sibilities.

Us­­ing the submucosal tun­nel space to approach the muscle layer is not limited to the esophageal sphincter. A logical extension was to cut the obstruct­­ing gastric pyloric sphincter complex, cal­led gastric peroral endoscopic myotomy to ameliorate refractory gastroparesis. A meta-analysis of seven studies, involv­­ing almost 200 such patients was presented. Clinical success was seen in 82% and the gastroparesis cardinal symp­tom index decreased significantly in paral­lel, with scintigraphic empty­­ing studies. Obviously large control­led trials are required now to identify the subset of patients where this novel ther­apy would be most appropriate, also with long-term fol­low-up.

All endoscopists are aware of the dif­ficulties in gett­­ing permanent optimal cleaning/ desinfection of the endoscopic equipment. What is insuf­ficiently realised is the risk of bacterial exposure of the face of the endoscopist dur­­ing the procedure through exposure to blood and body fluids. Swabs of plastic face shields worn by the endosocpist or positioned on the suite wall were cultured before and after the procedure. The number of bacterial colony form­­ing units was low before the endoscopy but rose substantial­ly post-procedure in 46% of the face shields worn by the endoscopists and in 21% hang­­ing on the suite wal­l. Unrecognized face exposure occurs 5.6× per 100 half-d endoscpic procedures and individuals stand­­ing up to six feet away from the patient may still be exposed at a rate of 3.4 per 100 half-d presence in the endoscopy suite. The authors of that study make a plea for routine face protection for the endoscopist and as­sist­­ing staf­f. Uncleaned suite surfaces may aid in patient-to-patient transmis­sion of pathogens.

A major drawback of endoscopic (mucosal) resection of usual­ly large lateral­ly spread­­ing colonic adenomas is the high recur­rence rate of up to 30%, especial­ly after piece-meal resection, as discovered dur­­ing surveil­lance after 3– 6 m. Can thermal ablation of the resection margin reduce this high recur­rence rate? This was investigated in a large multicenter control­led trial in Australia in lateral­ly spread­­ing lesions, more than 2 cm in diameter. Piece-meal resection was fol­lowed by snare tip soft coagulation of the entire resection margin (soft coag ef­fect 4.80 watts). Surveil­lance endoscopy, performed at 5– 6 and 18 m, included meticulousinspection of the scar with white light and nar­row band imag­­ing with bio­psies of the scar center and margin and of any suspicious area. Close to 400 patients with over 400 lesions were included. Endoscopic recur­rence was observed in approx. 7% in the ablation arm vs. approx. 21% in the control arm, and histologic recur­rence in approx. 5 vs. 23%, resp. Endoscopic as­ses­sment had a sensitivity of approx. 92%, a specificity of approx. 97% and a negative predictive value of approx. 99% for cor­rectly identify­­ing recur­rence at the post-resection scar. The value of ablation of the margins has now been confirmed in other studies. Even in studies evaluat­­ing the usefulness of cold snare resection, the advice is usual­ly given to include a generous sur­round­­ing tis­sue margin to decrease the risk of local recur­rence. A remain­­ing chal­lenge for endoscopists is the detection of deep submucosal cancerous invasion in polypoid lesions. Up to now, T1 colorectal polyps with one or more risk factors for lymph node metastasis are considered endoscopical­ly unresectable. Can nar­row band imag­­ing (NBI) be helpful in identify­­ing deep malignant invasion? A large multicenter prospective Spanish study explored the usefulness of NBI in over 1,600 patients with well over 200 polyps larger than 1 cm. Pointers for deep invasion were brown/ dark brown discoloration with/ without whiter patches; areas of disrupted/ mis­s­­ing ves­sels; amorphous/ absent surface pattern. Of the lesions 4.2% had features of deep invasion and 4.3% were considered endoscopical­ly unresectable. The NBI based prediction of deep malignant infiltration had a sensitivity 58%, a specificity of 96% and a positive and negative predictive value of respectively 42 and 98%. Beyond a doubt such results are encourag­­ing but there is still room for improvement. We live certainly in excit­­ing times, now that experts are explor­­ing removal of (well dif­ferentiated) rectal lesions extend­­ing up to the muscularis propria by dis­sect­­ing the plane between the circular and longitudinal muscle. The endoscopic boundaries keep moving.

The Gastro Update Europe 2020 will be held on September 4–5, 2020 in Bratislava, Slovakia. For more information visit www.gastro-update-europe.eu.

Prof. Guido Tytgat, MD, PhD

Department of Gastroenterology and Hepatology

Academic Medical Center

Meibergdreef 9

1105 AZ Amsterdam

The Netherlands

 g.n.tytgat@amc.uva.nl