Abstract
Introduction
In the United States, most patients with large bowel adenomas judged not resectable via standard colonoscopic means are referred to surgeons who perform segmental colorectal resection. Whereas bowel resection is prudent for high-grade dysplastic polyps, which contain invasive cancers in 32% to 41% of cases, it is not clear that a formal resection is needed for plain adenomas.1-3 If it were possible to remove these polyps without colectomy, it would be preferable because it would avoid resection-associated morbidity and mortality.
During the past decade, 2 alternative methods— one developed by surgeons and the other by gastroenterologists—have been used for these polyps. Segmental colectomy has been avoided by performing colonoscopy together with laparoscopy in the operating room. The polyp is laparoscopically manipulated, so that it can be grasped and removed via the colonoscope with a standard snare. After removal, the bowel wall integrity is assessed laparoscopically and reinforced, if necessary. If colonoscopic removal is not feasible, then a laparoscopic partial circumference “wedge” resection of the polyp is attempted. A standard segmental bowel resection is the last resort.4-8
The alternative approach is to remove the polyp colonoscopically using endoscopic submucosal dissection (ESD) and endoscopic mucosal resection (EMR) methods. Both ESD and EMR require a “saline lift” that is established by injecting saline or other solution into the submucosal layer of the bowel wall, which increases this layer’s thickness and “lifts” the polyp off the muscularis propria. The normal mucosa surrounding the polyp is then scored colonoscopically with a needle knife or other electrocautery instrument. An attempt is then made to colonoscopically expose and divide the submucosal attachments of the polyp. Ideally, the polyp is removed en bloc; however, if need be, it is removed piecemeal. ESD/EMR were developed by the Japanese for gastric neoplasms. 9 ESD/EMR have also been used for rectal, sigmoid, descending, and some transverse colon lesions. Thus far, these methods have not often been used in the right colon, which is the most common location of these polyps.
ESD/EMR methods are attractive for large polyps because, if successful, an abdominal surgical procedure is avoided. In the United States, ESD/EMR are performed by some gastroenterologists for select large-bowel polyps. It is logical for colorectal surgeons to learn ESD/EMR methods because most of these lesions are referred to them. Several years ago, in an effort to avoid colectomy, gain experience with ESD/EMR, and to determine if these methods are appropriate for right-sided polyps, the authors began performing ESD/EMR in conjunction with laparoscopy for all non-high-grade dysplastic benign polyps sent for bowel resection. Unlike the combined laparoscopic/endoscopic method described above, in this instance, the ESD/EMR was done without laparoscopic assistance. However, after polypectomy, the bowel wall was assessed laparoscopically. If ESD/EMR methods failed or were not feasible, a wedge partial bowel wall resection or standard segmental resection was performed. This is a retrospective review of the results of the first 26 cases. The long-range goal is to refine the selection process as well as ESD/EMR methods, such that large sessile adenomas could be removed without an accompanying laparoscopy.
Methods
This is a retrospective review of one surgeon’s experience with ESD/EMR and wedge resection (WR) methods for benign colorectal neoplasms; surgery was performed at either New York Presbyterian Hospital (Columbia University Campus) or St Luke’s Roosevelt Hospital in New York City. Because these methods are not experimental and have been shown to be effective, the cases in question were not done as part of an institutional review board (IRB)-sanctioned study. The majority of the data for each patient was prospectively collected and maintained in an IRB-approved prospective database maintained by the colon and rectal surgery service at each institution.
Adult patients with a biopsy-proven benign neoplasm, judged not removable via a colonoscope, and referred to the senior author for surgical evaluation were candidates for this procedure. The ESD/EMR and wedge partial circumference bowel wall resection methods as well as the standard segmental bowel resection, were each presented and explained. ESD/EMR was attempted only on patients who consented to a combined laparoscopic and colonoscopic procedure, wherein the polyp would be removed via 1 of the 3 methods. It was understood that the procedure of choice was ESD/EMR followed by wedge resection; the last option was segmental bowel resection. It was made clear that frozen section pathological analysis would be carried out immediately for polyps removed via ESD/EMR or WR and that if invasive cancer was found, a segmental bowel resection would be performed immediately. It was also understood that there was a small chance that a segmental bowel resection might need to be performed at a later date if the final pathological analysis revealed invasive cancer.
Patients with polyps containing invasive carcinoma, high-grade dysplasia, or carcinoma in situ were not eligible; the same was true for patients with familial adenomatous polyposis or other polyposis syndromes. Immunosuppressed patients were also excluded, as were ASA-4 patients and those with a history of a recent myocardial infarction.
Study data were obtained from office and hospital charts as well as from operative and pathology reports. Details concerning the preoperative, intraoperative, and postoperative colonoscopic exam(s), including polyp location and prior polypectomy attempts, were recorded and noted.
Short-term outcome, including length of stay (LOS), time until flatus and bowel movement (BM), complications, and reoperations, were noted. For successful ESD/EMR and WR patients, follow-up colonoscopy data, including biopsy results, were collected. Finally, operative, pathological, and outcome data were collected for patients who underwent a bowel resection at a second operation.
Intraoperative Procedures
Colonoscopy
Consenting patients were placed in the modified lithotomy position after general anesthesia was established. All had an intraoperative colonoscopy with CO2 insufflation of the bowel. In most cases, several laparoscopic ports were placed prior to colonoscopy and a brief abdominal inspection carried out. In some patients with distal left-sided polyps, the procedure was begun with colonoscopy. A double-channel colonoscope was inserted transanally, and an examination to the cecum was carried out, except for those with rectal or sigmoid polyps. Next, a saline and epinephrine solution (1:40 000 concentration) was injected into the submucosal layer beneath the index polyp using a sclerotherapy needle. The goal was to elevate or “lift” the polyp off the bowel wall. Based on the results of the saline lift test and the polyp’s size, position, and character, it was decided if ESD/EMR would be attempted. Also, if not already present, India ink tattoos were placed along the polyp’s borders.
Endoscopic submucosal dissection
After saline lift, a needle knife (Boston Scientific) was used to score the mucosa 0.5 to 1 cm away from the polyp edge. Next, using a dissection cap, gravity, snare, or biopsy forceps (inserted via the second channel to push or lift the polyp), the submucosal attachments beneath the polyp were exposed and divided using the needle knife. The saline lift was reestablished as needed during the procedure. Finally, the polypectomy was completed by dividing the final submucosal attachments with a needle knife or standard snare.
Endoscopic mucosal resection
EMR is defined as the use of a snare in conjunction with a saline lift to remove a polyp in piecemeal fashion (with or without a needle knife). EMR was used when en bloc resection was judged not feasible.
Wedge partial circumference resection
WR was considered for polyps that lifted but were judged not appropriate for ESD/EMR and for polyps where ESD/EMR had failed. Factors taken into consideration were polyp size and colon diameter. The proximity of the polyp to the ileocecal valve was noted. The tattoos alerted the surgeon as to the polyp’s location. WR was attempted with a 12-mm reticulating laparoscopic linear GIA stapler; 60-mm cartridges were most often used. The colonoscope was used to observe and assist the positioning of the stapler. Multiple staple cartridges were sometimes used. For polyps close to the ileocecal valve, the colonoscope was inserted into the terminal ileum prior to stapler placement to protect the valve. Next, the colonoscope was withdrawn into the colon and the stapler’s position assessed. The stapler position was adjusted until there was no polyp visible. For proximal right colon polyps, the ileocecal valve was reintubated after WR to ensure that the lumen was open.
Laparoscopic inspection after ESD/EMR and WR
Except for extraperitoneal rectal polyps and for some rectosigmoid polyps, laparoscopy was performed. If not already present, ports were placed and the abdomen insufflated. The segment in question and tattoo were located and the serosal surface examined. Next, the abdomen was filled with saline and the polyp-bearing segment submerged. The colon was insufflated via the colonoscope and a leak test performed. If the bowel wall was judged to be injured or weakened, then the colon wall was suture imbricated or a WR carried out, after which a visual inspection and leak test were repeated.
Segmental colorectal resection
A standard oncologic segmental bowel resection was done for polyps judged inappropriate for ESD/EMR or WR and for ESD/EMR failures. The blood vessels supplying the colon segment were divided close to their origin and the lymph node basin included in the resection.
Intraoperative pathological evaluation
All ESD/EMR polyp specimens were immediately retrieved and removed from the colon after harvest and sent for frozen section analysis. WR specimens were likewise removed from the abdomen and sent for immediate frozen section. The wedge specimens were carefully opened along the staple line and unfolded, after which the circumferential margin was assessed grossly and via frozen section. The pathologists performed multiple sections on all ESD/EMR and wedge specimens and looked carefully for high-grade dysplasia and invasive cancer. Segmental colorectal resection specimens were opened by pathology and inspected. The polyp size and location was noted; however, frozen sections were not obtained in these cases.
Follow-up Colonoscopy
All patients who successfully underwent either an ESD/EMR or WR were requested to return for a follow-up colonoscopy or sigmoidoscopy 3 to 4 months after surgery. The polyp site was identified via the tattoo(s), and multiple biopsies were obtained of the polypectomy scar (ESD/EMR) or the staple line (WR). If the pathology results confirmed a complete resection (no adenoma found), then the patient was requested to return for a second follow-up colonoscopy/sigmoidoscopy in 6 months, at which time biopsies of the scar were again taken. The third exam was scheduled 1 year after the second provided the biopsy results showed no neoplasm. Multiple follow-up exams were carried out because microscopic foci of residual polyp could be missed on the first or second examinations; with time, residual disease will become apparent.
If obvious residual adenoma was found at the first follow-up exam, then an attempt was made to remove the remaining polyp using a biopsy forceps or standard monopolar diathermy snare after submucosal injection of saline in the area. Additional colonoscopy/sigmoidoscopy exams were done at 3- to 4-month intervals until the polypectomy base biopsies revealed no adenoma, at which point the patient was asked to return for their next exam 6 months later.
Patients who underwent a segmental resection were requested to return for a full colonoscopy 1 year after their surgery.
Statistical Methods
The main outcome measures were presented as median with 95% confidence interval. A Mann-Whitney U test was used to compare the nonparametric results. P values of less than .05 were considered to indicate significant differences between groups. All statistical tests were performed using GraphPad Prism (version 4.0, GraphPad Software, USA).
Results
A total of 26 patients with adenomas went to surgery having consented to polyp removal via ESD/EMR, wedge, or segmental resection (mean age = 60.7 years; 38.5% male, 61.5% female). The most common polyp locations were the right colon (57.7%), rectum (26.9%), and sigmoid colon (7.7%; Table 1). Successful ESD/EMR was carried out in 13 patients (50%), whereas a WR was carried out in 4 (15%), and a standard oncologic segmental resection was done in 9 (35%) patients (mean lymph node harvest, 13.7; Table 2). The median LOS was as follows: ESD/EMR, 3 days; wedge, 5 days; resection, 5 days. The ESD/EMR LOS was significantly shorter than that of the resection group. The median times to first flatus and BM were significantly shorter for the ESD/EMR group when compared with that for the resection group (Table 3). There were no major complications (leaks, abscesses, reoperations, and so on), and there was no mortality (Table 4); 2 ESD/EMR patients and 1 WR patient had a prolonged ileus (defined in these patients as first BM or flatus occurring on or after postoperative day 3).
Polyp Characteristics.
Operative Characteristics.
Abbreviations: ESD, endoscopic submucosal dissection; EMR, endoscopic mucosal resection; SD, standard deviation.
One patient had 2 procedures performed: index polyp removed via ESD/EMR, resection for cancer in different area. Excluded in surgery time calculation.
Postoperative Recovery Data. a
Abbreviations: BM, bowel movement; ESD, endoscopic submucosal dissection; EMR, endoscopic mucosal resection.
In days, median (95% confidence interval).
ESD/EMR: n = 12; 1patient who had both cecal ESD/EMR and sigmoid resection for cancer was excluded from analysis.
ESD/EMR: n = 11; 2 exclusions: same patient as above, and 1 patient discharged the same day.
Mann-Whitney U: ESD/EMR versus resection; significance P < .05.
Complications.
Abbreviations: ESD, endoscopic submucosal dissection; EMR, endoscopic mucosal resection.
Final pathology revealed 18 adenomas, 3 high-grade dysplastic adenomas, 3 invasive cancers, and 2 rectal carcinoids. Complete resection was confirmed for all en bloc ESD/EMR, WR, and segmental colectomy patients. Small foci of invasive carcinoma were found in the polyps of 3 ESD/EMR patients with large bulky pedunculated/sessile polyps whose frozen-section analyses had shown adenoma only. In these 3 cases, to expose the base and permit ESD, the pedunculated parts of these polyps were first removed; the cancer foci were found in the superficial parts. Despite the documented complete removal of these 3 lesions via ESD/EMR, these 3 patients underwent subsequent segmental colon resection without incident; in all 3, pathology revealed no cancer in the bowel wall or in the lymph nodes.
Regarding the successful ESD/EMR cases, a needle knife was used in 61%, whereas a snare was used in almost all cases. Of the polyps, 31% were removed en bloc, and 69% were removed piecemeal (Table 5). In 5 out of 9 segmental resection and 3 out of 4 WR patients, ESD/EMR was first attempted. With regard to right-sided lesions, ESD was attempted in 10/15 (66.7%) of these polyps and was successful in 3 cases. Of the remaining patients, 3 had a WR and 4 a segmental colectomy.
Colonoscopic Polypectomy Methods in Successful Cases (n = 13).
Abbreviations: ESD, endoscopic submucosal dissection; EMR, endoscopic mucosal resection.
Follow-up colonoscopy was done in 9 of the 13 ESD/EMR patients; in 7 patients, biopsies of the scar confirmed complete removal. In 2 patients, residual adenoma was found and removed endoscopically (Figure 1). At this time, follow-up exams and biopsies were done in 3 of the 4 WR patients and confirmed complete resection (Tables 6 and 7).
Postoperative follow-up of ESD/EMR patients.
Final Pathology Results.
Abbreviations: ESD, endoscopic submucosal dissection; EMR, endoscopic mucosal resection.
Follow-up Data for 3 Cancer Patients.
Abbreviations: F/U, follow-up; LN, lymph node.
Patient underwent colon resection at an outside institution; only final path results available.
Discussion
In the setting of a solitary large or difficult benign non-high-grade dysplastic polyp in patients without Hereditary nonpolyposis colorectal cancer (HNPCC) or Familial adenomatous polyposis (FAP), it is logical to avoid colectomy if the lesion can be fully and safely removed by other means. In this series, using the combined laparoscopic/colonoscopic approach as well as ESD/EMR and WR methods, it was possible to avoid bowel resection in 65.4% of patients. The LOS and ileus resolution time were shorter for the ESD/EMR group when compared with the times for the resection group. Of note, most of the successful endoscopic polypectomies were piecemeal resections. Ideally, polyps would be removed en bloc with a margin of normal mucosa, so that the margins can be assessed. 10 However, for the few polyps that are bulky, pedunculated, and sessile, piecemeal removal may be necessary to expose the base. Regardless, provided that the polyp is fully removed and follow-up evaluation and biopsy of the polyp site performed 3 to 4 months later, piecemeal removal is preferable to segmental colectomy. With greater experience and better ESD tools, the number of completed en bloc ESDs should increase.
The finding of 3 invasive cancers despite frozen-section analysis is worrisome; however, no cancer was found when segmental resection was later carried out in all 3 of these patients. Of note, all 3 were very large pedunculated/sessile polyps; perhaps, these polyps are best treated with segmental resection because their large mass increases the chances of finding a cancer. However, it is not clear that a small focus of cancer centimeters away from the base carries the same risk for lymph node involvement as in a sessile polyp. In a sessile polyp, the cancer foci are close to the muscularis mucosa–associated lymphatic network, whereas in the pedunculated/sessile polyp, the cancer was well away from the bowel wall. There are no published data regarding this subset of polyps. Clearly, more data are needed.
It is interesting to note that none of the patients who had colectomy because of “failure to lift” had an invasive cancer. The saline lift test is not reliable if a prior polypectomy attempt has been made. More reliable assessment methods are needed; endoscopic ultrasound could be used. The present study also verifies that right-sided polyps are a challenge; contributing factors are distance from the anus and the thin bowel wall. Based on our limited experience, it appears that WR is reasonable mainly for select cecal and proximal right-colon polyps. This series attempted ESD/EMR on select right-sided sessile polyps, which have been largely avoided thus far, and was successful in only 20% of cases. It is hoped that with better tools and greater experience, the completion rate for right-sided polyps will increase.
The operative setting and the inclusion of laparoscopy allowed for immediate WR or segmental resection in patients in whom ESD/EMR methods failed. This method is attractive to patients because their polyp will be removed by the least invasive method possible during a single procedure. WR is feasible in the cecum and proximal right colon if the polyp is far enough away from the ileocecal valve. It is logical for endoscopically proficient surgeons to learn ESD/EMR because the difficult polyps are usually sent to surgeons. Tools for ESD/EMR for colonic polyps are in the early stages of development. Currently, there are needle knifes and plastic “dissection” caps. Ideally, what is needed is an endoscopic method of lifting the edge of the scored polyp without looking away, so that the submucosal attachments can be placed on traction and cut. Another challenge is maintaining the mucosal lift; better solutions are needed.
Conclusion
ESD, EMR, and wedge polypectomy during combined laparoscopic/endoscopic procedures appear to be safe and effective for a subset of patients. ESD/EMR methods are associated with a shorter LOS and ileus resolution time when compared with resection. Patients who undergo ESD/EMR or WR require short-interval colonoscopy and biopsy to assess for residual or recurrent polyps. ESD/EMR tools and techniques are evolving; larger studies are needed to better assess these methods.
Footnotes
Authors’ Note
This was a podium presentation at the American Society of Colorectal Surgeons Annual Meeting; May 14-18, 2011; Vancouver, British Columbia, Canada. The authors Joon Ho Jang, MD, Daniel Kirchoff, MD, Kevin Holzman, MD, Koji Park, MD, Michael Grieco, MD, Vesna Cekic, MSc, Samer Naffouje, MD, Jon Kluft, BS, have no conflicts of interest or financial ties to disclose. Author Richard L. Whelan, MD, is a consultant for Olympus Corporation, Ethicon Inc, and Atrium Medical Corporation. Applied Medical, WL Gore & Associates Inc, and Atrium Medical Corporation have other research enrollment affiliations with Dr Whelan.
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.