A long-term follow-up study on the prognosis of endoscopic submucosal dissection for colorectal laterally spreading tumors

Accepted Manuscript A long-term follow-up study on the prognosis of endoscopic submucosal dissection for colorectal laterally spreading tumors Zhi-Jie Cong, M.D., Liang-Hao Hu, M.D., Jun-Tao Ji, M.D., Jun-Jie Xing, M.D., YongQi Shan, M.D., Zhao-Shen Li, M.D., En-Da Yu, M.D. PII:

S0016-5107(15)02815-1

DOI:

10.1016/j.gie.2015.08.043

Reference:

YMGE 9528

To appear in:

Gastrointestinal Endoscopy

Received Date: 13 January 2015 Accepted Date: 19 August 2015

Please cite this article as: Cong Z-J, Hu L-H, Ji J-T, Xing J-J, Shan Y-Q, Li Z-S, Yu E-D, A long-term follow-up study on the prognosis of endoscopic submucosal dissection for colorectal laterally spreading tumors, Gastrointestinal Endoscopy (2015), doi: 10.1016/j.gie.2015.08.043. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT Title Page

A long-term follow-up study on the prognosis of endoscopic submucosal dissection for colorectal laterally spreading tumors Zhi-Jie Cong1,2*, M.D., Liang-Hao Hu3*, M.D., Jun-Tao Ji3*, M.D., Jun-Jie Xing2,

RI PT

M.D., Yong-Qi Shan2, M.D., Zhao-Shen Li3, M.D., En-Da Yu2, M.D. 1. Department of Colorectal Surgery, Ren Ji Hospital, School of Medicine, Shanghai Jiao Tong University, Shanghai, China.

SC

2. Department of Colorectal Surgery, Changhai Hospital, Second Military Medical University, Shanghai, China.

University, Shanghai, China.

M AN U

3. Digestive Endoscopy Center, Changhai Hospital, Second Military Medical

* These authors contributed equally to this paper.

Corresponding author

EP

En-Da Yu, M.D.

TE D

Running title: Follow-up on ESD for LST

Department of Colorectal Surgery

AC C

Changhai Hospital

The Second Military Medical University 168 Changhai Road

Shanghai 200433, China Tel: +86-21-31161344 Fax: +86-21-31161365 E-mail: [email protected]

ACCEPTED MANUSCRIPT ABSTRACT Background Colorectal laterally spreading tumors (LSTs) are divided into homogenous (LST-G-H), nodular mixed (LST-G-M), flat elevated (LST-NG-F), and pseudodepressed

RI PT

(LST-NG-PD) subtypes. We hypothesized that based on rates of advanced histology, the recurrence rates of the LST-NG-PD and LST-G-M group may be higher than those of the other subgroups.

SC

Methods

Endoscopic submucosal dissection (ESD) was performed in 156 patients with 177

specific subtype were investigated. Results

M AN U

LSTs. The clinicopathological features and long-term prognosis of ESD according to

LSTs were most commonly found in the rectum, and the highest percentage of rectal

TE D

lesions was observed in the LST-G-M group (71.1% vs. overall 55.4%, P = 0.032). The LST-G-M lesions were larger (60 ± 22 mm vs. 40 ± 33 mm, P = 0.034) than the LST-G-H lesions. The LST-G-M group also demonstrated more high-grade

EP

intraepithelial neoplasias (32.2% vs. 10.8%, P = 0.003) and submucosal carcinomas

AC C

(13.6% vs. 1.5%, P = 0.010) compared with the LST-G-H group. The LST-NG-PD group exhibited the highest incidence of submucosally invasive cancer (16.7%). The overall perforation rate was 2.3%. The perforation in the LST-NG group was higher than that in the LST-G group (5.7% vs. 0.8%, P = 0.047). All recurrences (7.7%) were found by colonoscopy without any detection of cancers and no difference was found among the subtypes. Conclusions No significant differences were observed among subgroups with a 44.4 ± 16.3 months

ACCEPTED MANUSCRIPT follow-up. Considering that all recurrences were discovered through colonoscopy, and most can be cured by repeated ESD, the LSTs of all subgroups require more intensive follow-up compared with smaller adenomatous lesions. Key Words: laterally spreading tumors; outcome; colon; rectum; endoscopic

RI PT

submucosal dissection

Introduction

SC

A colorectal laterally spreading tumor (LST) is a flat and broad-based lesion, with 1 cm or greater in diameter that extends laterally and circumferentially along the

M AN U

colorectal wall rather than perpendicular to it [1], which should receive considerable attention because of its high malignant potential [2–5]. Endoscopic submucosal dissection (ESD) is the standard treatment for LSTs [6, 7]. In 2008, Kudo et al. divided LSTs into 4 subtypes based on different surface morphologies [8].

TE D

Considerable medical literature shows that the risk of containing advanced histology significantly increases

in some LST subgroups.

For example,

malignant

transformation and premalignant lesion (HGIN/CIS) were reported frequently in the

EP

LST-NG-PD and LST-G-M groups [9–13]. The proportion of submucosa-massive

AC C

(SM-m) lesions also reportedly increases in the LST-NG-PD group [13, 14]. We hypothesized that based on rates of advanced histology, the recurrence rates of the LST-NG-PD or LST-G-M group may be higher than those of the other subgroups. In this study, we performed a retrospective analysis of the outcome and a prospective analysis of the long-term prognosis of LST patients with ESD and assessed the clinico-pathological features of the different subtypes of LST. Materials and methods General information

ACCEPTED MANUSCRIPT A retrospective analysis was carried out on LST cases with ESD in our hospital from January 2003 to December 2007. During this initial period, no international accepted guidelines were available for the treatment of colorectal LSTs or large polyps. Polypoid and nonpolypoid lesions were discovered by colonoscopy. Hot

RI PT

biopsy, polypectomy, endoscopic mucosal resection (EMR), or piecemeal EMR was employed for lesions smaller than 30 mm with no routine staining. Lesions larger than 30 mm were identified by ESD after staining and observation by a senior endoscopist

SC

(Dr. Yu). Surgery was carried out instead of ESD when one of the following situations occurred: (1) non-lifting sign after submucosal injection; (2) an invasive type V pit

M AN U

pattern; and (3) biopsy-confirmed local canceration [15]. LSTs are defined as lesions greater than 10 mm in diameter. However, all cases included in this study were lesions larger than 30 mm because smaller LSTs were not stained and treated with ESD in our hospital at that time.

TE D

Classification of LSTs

LSTs are divided into 2 subtypes according to endoscopic features: LST-Gs with nodules or granules distributed evenly or not on the surface of the lesion and

EP

LST-NGs with smooth surface and no nodules or granules. In 2008, Kudo et al. [8]

AC C

further divided the former subtype into homogeneous (LST-G-H) and nodular mixed (LST-G-M), and the latter subtype was divided into flat-elevated (LST-NG-F) and pseudo-depressed (LST-NG-PD) (Figure 1). In the present study, 2 experienced endoscopic doctors blindly reviewed the endoscopic images of LST cases. The LSTs were stained, received ESD treatment, and were divided into four subtypes according to Kudo’s standard. Consensus was reached by discussion when disagreement arose. ESD operation All patients received full communication and signed written informed consent

ACCEPTED MANUSCRIPT before operation. A needle knife (KD-10Q; Olympus) was used in the procedure only during the initial period of ESD in our hospital. A hook knife and an insulated-tip (IT) knife were gradually adopted in incision and separation, and a transparent hood (ST hood) was started to be placed on the top of the scope in 2005. We initially chose to

RI PT

snare the lesion after circumferential mucosal incision with a needle knife in some patients, which is now called as simplified or hybrid ESD (S/H-ESD) [16], as an introductory step to full ESD. Indigo carmine (0.4%) or methylene blue (0.5%) was

SC

used for the pre-operation staining. The surface structures and the type of the openings of colonic crypts were observed before the lesion margin was determined by

M AN U

common electronic colonoscopy or magnifying colonoscopy. Repeated submucosal injections (100 mL 0.9% saline solution containing 0.4% indigo carmine and 0.0001% epinephrine) were carried out to uplift the lesion with a injection needle (NM-4L-1, Olympus, Japan). A high-frequency generator (ICC-200; ERBE, Tübingen, Germany)

TE D

was used for all ESD procedures (endocut E3, 45-60 W for marginal incision; forced coagulation E1, 40W for submucosal dissection). An incision was made along the margin of the lesion by an IT knife assisted by a needle knife before the submucosal

EP

separation. ESDs were performed without intravenous sedation or analgesia. The

AC C

lesion was pulled open gravitationally by changing the patients’ body position without the assistance of an ST hood in some of the cases. Careful hemostasis was observed to visible vessels at the base of the mucosal defect, and hemoclips were used when bleeding or perforation occurred during the operation. Histopathological assessment Excised specimens were paraffin embedded and sectioned perpendicularly with an interval of 2 mm. Pathological diagnosis was made after hematoxylin and eosin staining and microscopic observation. Assessment criteria included lesion size,

ACCEPTED MANUSCRIPT invasive depth, presence of fibrotic scar, lymphatic and vascular involvement, lateral and basal margins with residual adenoma or tumor tissue. En bloc resection is defined as a tumor removed in a single piece. En bloc R0 resection is histologically defined as a tumor removed as a single piece without adenomatous tissue or residual carcinoma

RI PT

at the lateral or deep margins. High-grade intraepithelial neoplasia (HGIN)/carcinoma in situ (CIS) is defined as a lesion with the morphological characteristics of high-grade dysplasia or adenocarcinoma that is confined to the glandular epithelium

SC

or invades the lamina propria without submucosal invasion [17]. Submucosal carcinoma is defined as carcinoma that invades through the muscularis mucosa into

M AN U

the submucosa [17]. Follow-up

This study was approved by the Committee on Ethics of Biomedicine Research of the Second Military Medical University, Shanghai, China. All the patients signed the

TE D

informed consent form of follow-up. All LST cases with ESD were requested to undergo follow-up colonoscopy regularly and sign the informed consent form of follow-up. The local recurrence rate, incidence of new primary tumors/polyps in other

EP

areas, and overall/ disease-specific survival were assessed and analyzed prospectively.

AC C

Follow-up colonoscopies were recommended every 6 months for the first year after ESD to assess the ulcer healing and the growth or residual of tumor. Then colonoscopy was taken yearly to discover local recurrence and new primary tumors/polyps in other areas until 3 years after ESD. After that, the surveillance interval could be extended to 2 to 3 years if no polyps were found in other areas. The follow-up period was counted from the ESD operation to the last colonoscopy. If tumor was discovered at the previous ESD site or adjacent to ESD scar within 1 to 2 mm, local recurrence or residual neoplastic disease was highly suspected. An

ACCEPTED MANUSCRIPT extended ESD could be carried out after biopsy. If histology confirmed invasive disease involving the submucosa, surgical resection was recommended. Submucosal carcinoma patients with surgical resection were followed up according to American National Comprehensive Cancer Network (NCCN) guidelines for colorectal cancer.

RI PT

Statistical analysis

SPSS 14.0 (SPSS Inc, Chicago, Ill, USA) was used for the statistical analysis. Mean value and standard deviation with range were measured for the continuous

SC

variables, whereas frequency and percentage were calculated for the categorical variables. The differences between the 2 groups were analyzed by an independent

M AN U

Student t test or Mann–Whitney U test. The Pearson chi-square test and Fisher exact test were performed for categorical variables. A P<0.05 was considered statistically significant. Results

TE D

Overall characteristics of LST cases Overall characteristics

A total of 31,503 endoscopic examinations and 7,133 endoscopic polypectomies

EP

were operated in our hospital from January 2003 to December 2007. The 156 patients

AC C

with 177 LSTs had lesions larger than 30 mm and were treated with ESD, accounting for 2.2% of the total polypectomies. In addition to the aforementioned cases, 8 ESD procedures were aborted because of the difficulty encountered during the process, and these patients were converted to surgical operations. Seventeen patients refused ESD and chose surgical operation. Demographic data and clinico-pathological features are shown in Table 1. The average age of the patients was 62.9 ± 11.2 years (range 34-85 years), with a male to female ratio of 1.26:1 (87/69). The mean of lesion diameter was 52 ± 26 mm (30-140 mm). The most common location was the rectum (98, 55.4%),

ACCEPTED MANUSCRIPT followed by the ascending colon (30, 16.9%) and the transverse colon (29, 16.4%). The most common histological type was tubular adenoma with 51 cases (28.8%), and the second most common type was tubulovillous adenoma with 41 cases (23.2%). Submucosal carcinoma was found in 13 cases (7.3%) (Table 1).

RI PT

Short-term outcome of ESD

The average procedure duration was 64.3 ± 45.9 min (range 10–294 min). A total of 147 (83.1%) of 177 lesions received en bloc resection, of which 3 were diagnosed

SC

with basal margin involvement by histological evaluation, resulting in the en bloc R0 resection rate down to 81.4% (144). A total of 4 (2.3%) perforations occurred,

M AN U

including 1 in the transverse colon and 3 in the rectum. Three of these perforations occurred during the procedure, and the patients recovered after being treated with endoscopic clipping and antibiotics. The other perforation with mild symptoms occurred 24 hours after ESD, and the patient recovered with long period of fasting

TE D

and antibiotic treatment instead of surgical intervention. Six delayed bleedings (3.4%) were observed, including 1 in the ascending colon and 5 in the rectum. These bleedings occurred in the first 3 days after the ESD procedure. One of the 6 patients

EP

stopped bleeding after conservative treatment, whereas the other patients underwent

AC C

endoscopic clipping and reached successful hemostasis. The points of bleeding were observed from the resection bed with vessels on the surface. No blood transfusion was required.

Long-term outcome of LSTs with or without submucosal carcinoma Among the 13 cases of submucosal carcinoma, 8 had no lateral and basal margin involvement; 3 had basal margin involvement as previously mentioned; and 2 cases, who received piecemeal resection instead of en bloc resection, could not be assessed. Nine of the 13 cases received additional surgical resection (including 3 cases with

ACCEPTED MANUSCRIPT basal margin involvement and 2 cases with piecemeal resection). No lymph node metastasis or recurrence occurred after additional surgical resection. Four rectal LST cases with en bloc R0 resection (3 of 4 with basal margin involvement) chose radiotherapy with close follow-up because of old age or heart and lung problems.

RI PT

Moreover, 1 of the 4 cases received additional surgical resection for recurrence at 18 months after ESD, and no recurrence was observed in the other 3 cases.

In addition to the 13 cases of submucosal carcinoma included in the colorectal

SC

cancer follow-up program according to NCCN guidelines, the other 133 cases with 152 lesions received follow-up colonoscopies for an average duration of 44.4 ± 16.3

M AN U

months (range 18–87 months), and 10 were lost to follow-up with a follow-up rate of 93.0% (133/143) (Figure 2). A total of 127 cases (127/133, 95.5%) were followed up for more than 24 months. Local recurrence occurred in 11 cases at 7.7 ± 10.6 months (range 2–20 months) during the follow-up period. All 11 cases presented with no

TE D

discomfort and were discovered by colonoscopy. In all recurrence, 8 occurred within 12 months after the procedure, and 10 (90.9%) underwent 1–2 repeated ESD successfully without recurrence after the procedure (Table 1). One case received

EP

surgical resection for recurrence after second ESD. Slight scar stenosis was observed

AC C

in only 1 patient (accounting for 0.7% of all cases), who did not need balloon dilatation. Polyps in other areas of the colon were discovered during the follow-up period in 26 cases (26/133, 19.5%), and these polyps were endoscopically treated with ligation or cauterization. The average time interval from ESD operation to the discovery of polyps in other areas was 23.5 ± 17.6 months (range 2–59 months). No LST was observed in other areas during the follow-up period. Survival The 3-year overall/disease-specific survival was 100%/92.2%, and the

ACCEPTED MANUSCRIPT corresponding 5-year values were 95.5%/92.2% (Figure 3). One patient died of cerebral tumor, and 2 patients died of cardiovascular incident. None of the patients

Characteristics of each subtype of LSTs Demographic data

RI PT

died of colorectal cancer.

The distribution of subtypes was as follows: 124 LST-Gs including 65 cases of

SC

homogenous type (LST-G-H, 36.7%) and 59 cases of nodular mixed type (LST-G-M, 23.2%); and 53 LST-NGs including 41 cases of flat-elevated type (LST-NG-F, 23.2%)

M AN U

and 12 cases of pseudo-depressed type (LST-NG-PD, 6.8%). No statistically significant differences were found in lesion size (mean 50mm vs. 50mm) or location between LST-G and LST-NG groups, but the mean diameter of LST-G-Ms was significantly larger than LST-G-Hs (60 mm vs. 40 mm, P=0.034). In all pathological

TE D

types, the most common lesion location was the rectum, which accounted for 71.1% (42) of all LST-G-Ms. This result is higher than the corresponding value of the LST-G-H group (46.2% [30], P=0.004). The lesion distribution showed no significant

EP

difference between the LST-NG-F and LST-NG-PD groups (Table 2).

AC C

Histological assessment

Pathological types showed no significant difference between the LST-G and

LST-NG groups, but further analysis among subtypes found that tubular adenomas were more common in the LST-G-H group (35.4%) than in the LST-G-M group (16.9%, P=0.020). HGIN (premalignant) and submucosal carcinomas (malignant) accounted for 32.2% and 13.6% of LST-G-M, respectively. Both were significantly higher than the corresponding percentages in LST-G-H (10.8% and 1.5%). Moreover, no significant difference was observed in the pathological types between the

ACCEPTED MANUSCRIPT LST-NG-F and LST-NG-PD groups, whereas the sum of the percentage of HGIN and submucosal carcinomas was significantly higher in LST-NG-PD than in LST-NG-F (41.7% vs. 14.6%, P=0.042). Outcome of LST subtype

RI PT

En bloc resection rate and en bloc R0 resection rate showed no significant difference among types or subtypes. The incidence of postoperative bleeding did not also show any significant difference among types and subtypes. The rate of

SC

perforation for the LST-NG group was significantly higher than that for the LST-G group (5.7% vs. 0.8%, P=0.047). However, no statistical difference of perforation

M AN U

existed among the four subgroups (P>0.05). No significant difference was also found in mean procedure duration and recurrence rate among types or subtypes. Discussion

This study retrospectively analyzed a large number of LST cases with ESD from

TE D

a tertiary referral hospital during the initial period when ESD was introduced intoChina. The rates of en bloc resection and en bloc R0 resection were 83.1% (147/177) and 81.4% (144/177), respectively. The long-term outcomes were

EP

prospectively analyzed during the average follow-up period of 44.4 ± 16.3 months.

AC C

Local recurrence occurred in 11 cases at 7.7 ± 10.6 months after ESD and all recurrences were found by colonoscopy with no canceration. Most of the cases with recurrence received radical resection by ESD, and no colorectal tumor-related deaths occurred.

Rectum is the most common site among all the LST subgroups and the highest percentage of rectum was observed in the LST-G-M group (71.1%, P=0.032). The lesions of the LST-G-M group were larger (60mm vs. 40mm, P=0.034) than those of the LST-G-H group. In reference to pathological distribution, LST-G-Hs were more

ACCEPTED MANUSCRIPT often diagnosed as tubular adenomas with low malignant potential, whereas LST-G-Ms had a higher percentage of both HGIN and submucosal carcinomas than LST-G-Hs. Although former studies showed a relatively low malignant potential in LST-Gs [18, 19], LST with large nodules is considered an independent risk criterion

RI PT

to submucosal carcinomas because tumorigenesis usually occurs in cells with the largest LST nodule [18]. With accordance to former reports [18, 19], the present study showed a high malignant potential in LST-NG-PDs. The LST-NG-PD group with a

cancer of 16.7% among the four subtypes.

SC

small sample size in the study had the highest incidence of submucosal-invasive

M AN U

Given the differences among the subtypes of LSTs, it is necessary to subdivide LSTs into four subtypes rather than 2 types (i.e., LST-G and LST-NG). Further investigation of sub-differentiated LST lesions and comparative study on subtypes will help devise more targeted and effective treatment for each subtype of LST. For

TE D

instance, given the high malignant potential in LST-G-M and LST-NG-PD, en bloc resection rate should be pursued as high as possible in ESD with these 2 subtypes, which will contribute to accurate pathological diagnosis in determining whether

EP

additional surgical resection is needed. Although malignant transformation and

AC C

premalignant lesion (HGIN/CIS) were frequent in the LST-NG-PD and LST-G-M groups, no significant differences in recurrence rates were observed among all the subgroups. Huang et al. [20] reported similar results, indicating that although the rates of malignant lesions were significantly higher in the LST-NG type than those in the LST-G type (27% vs. 5.6%, P=0.011), no significant difference was found in recurrence between the 2 types (15.4% vs. 7.1%, P = 0.419), with a follow-up of 7.8 ± 5.8 months (range 3–26 months). Bleeding is more commonly encountered in rectal lesions and the incidence of

ACCEPTED MANUSCRIPT bleeding showed no significant difference between the LST-G and LST-NG groupsor among the four subtypes. The overall perforation rate was low (2.3%), but we found a higher rate in the LST-NG group (5.7%) than in the LST-G group (0.8%, P=0.047). As perforation often occurs in LST-NG lesions, ESD operations should be conducted

RI PT

with more care in this type. For example, before performing ESD, it is important to observe the lesion color, presence of depression, surface unevenness and fold convergence under chromoendoscopy (with indigo carmine, methylene blue, etc) or

SC

image-enhancement technology (narrow-band imaging [NBI]) in order to identify a flat or pseudodepressed lesion. Repeated submucosal inject of a lifting solution as

M AN U

necessary to form a safe plane for dissection during the whole ESD procedure. The main solution used was normal saline solution during the initial period of ESD in our hospital, but now we suggest longer-lasting solutions such as glycerol or hyaluronic acid. Care should also be taken with the cautery setting to prevent perforation and

TE D

bleeding. Dry cut current can be opted for densely fiborsed area, and spray coagulation current can be used for hard-to-reach areas. More coagulation current should be delivered while dissecting heavily vascularized surgical plane. Excessive

EP

coagulation may increase the risk for delayed perforation.

In addition, compared

AC C

with the wide variety of knives used today, needle knife was the only alternative in the early years; this knife causes more intraprocedure bleeding compared with other knives and increases the risk of perforation [21–23]. So the needle knife should be used under the direct vision of the surgeons during the procedure. Without an ST hood, the area of the separated lesions may block the vision, which could cause serious difficulty in continuing the operation. As a result, intravenous sedation and anesthesia/analgesia were seldom used in ESD performance in our hospital，because patients were frequently requested to change positions, using gravitation to pull open

ACCEPTED MANUSCRIPT the lesion for better vision. But when good vision couldn’t be gotten even after changing the patients’ positions, the patients had to choose piecemeal resection instead of en bloc resection in order to prevent perforation, leading to a lower en bloc resection rate. Moreover, analgesia and loss of pain sensation also reduce the

RI PT

opportunity for the endoscopist to discover perforation during the procedure, for patients being unable to provide immediate and crucial feedback on pain when perforation occurs [24]. Last but not least, one should observe carefully for the sign of

SC

subtle perforation after ESD, such as dark brown muscle fiber in the resection bed. We think if we can do a gentle, accurate, and careful ESD procedure according to the

M AN U

attentions mentioned above, perforation can be avoid.

The limitations of this study are as follows. First, it is a single-center retrospective study. Second, given the lack of awareness about classification of LSTs in the initial periods of ESD, information on 1 to 2 cm LST lesions was not included

TE D

because only lesions larger than 3 cm were stained, observed, and treated with ESD. Even without the large number of cases in each subtype, the research is still valuable because the number and size of LST lesions included in the study were large. The

EP

third limitation is that all cases were admitted in early periods of ESD when the en

AC C

bloc resection rate was low, which may fail to reveal the realities of the ideal prognosis of ESD. Recent cases have been followed up in our hospital. In conclusion, LSTs require closer follow-up than ordinary polyps. The relatively

long-term outcomes may be excellent when all LST recurrences are discovered by colonoscopy without any detection of cancers and mostly cured by repeated ESD. Although malignant transformation and premalignant lesion (HGIN/CIS) were frequent in the LST-NG-PD and LST-G-M groups, no significant difference was observed in the recurrence rate among all the subgroups with a follow-up time of

ACCEPTED MANUSCRIPT 44.4 ± 16.3 months. We recommend that all LST lesions require post-ESD intensive surveillance, the suggested frenquency of which is semiannually for the first year and then yearly. References

RI PT

1. Kudo S. Endoscopic mucosal resection of flat and depressed types of early colorectal cancer. Endoscopy 1993;25(7):455-461.

2. Imai Y, Terai T,Miwa H, OhnoY, Ogihara T, Sato N. Marginal irregularity of flat

SC

elevated type of colorectal tumor as a marker of malignant potential. Gastrointest Endosc 1998;48(3):263-266.

M AN U

3. Matsui T, Yao T, Iwashita A Natural history of early colorectal cancer. World J Surg 2000;24(9):1022-1028.

4. Matsuda T, Saito Y, Hotta K, Sano Y, Fujii T. Prevalence and clinicopathological features of nonpolypoid colorectal neoplasms: should we pay more attention to

TE D

identifying flat and depressed lesions? Dig Endosc 2010;22(Suppl 1):S57-S62. 5. Kudo S, Kashida H, Tamura T, Kogure E, Imai Y, Yamano H, Hart AR. Colonoscopic diagnosis and management of nonpolypoid early colorectal cancer.

EP

World J Surg 2000;24(9):1081-1090.

AC C

6. Matsuda T, Gotoda T, Saito Y, Nakajima T, Conio M. Our perspective on endoscopic

resection

for

colorectal

neoplasms.

Gastroenterol

Clin

Biol

2010;34(6-7):367-370.

7. Tamegai Y, Saito Y, Masaki N, Hinohara C, Oshima T, Kogure E, Liu Y, Uemura N, Saito K. Endoscopic submucosal dissection: a safe technique for colorectal tumors. Endoscopy 2007;39(5):418-422. 8. Kudo S, Lambert R, Allen JI, Fujii H, Fujii T, Kashida H, Matsuda T, Mori M, Saito H, Shimoda T, Tanaka S, Watanabe H, Sung JJ, Feld AD, Inadomi JM, O'Brien

ACCEPTED MANUSCRIPT MJ, Lieberman DA, Ransohoff DF, Soetikno RM, Triadafilopoulos G, Zauber A, Teixeira CR, Rey JF, Jaramillo E, Rubio CA, Van Gossum A, Jung M, Vieth M, Jass JR, Hurlstone PD. Nonpolypoid neoplastic lesions of the colorectal mucosa. Gastrointest Endosc 2008;68(4 Suppl):S3-S47.

RI PT

9. Rembacken BJ, Fujii T, Cairns A, Dixon MF, Yoshida S, Chalmers DM, Axon AT. Flat and depressed colonic neoplasms: a prospective study of 1,000 colonoscopies in the UK. Lancet 2000;355():1211-1214.

SC

10. Tanaka S, Haruma K, Oka S, Takahashi R, Kunihiro M, Kitadai Y, Yoshihara M, Shimamoto F, Chayama K. Clinicopathologic features and endoscopic treatment of

M AN U

superficially spreading colorectal neoplasms larger than 20 mm. Gastrointest Endosc 2001;54(1):62-66.

11. Kim KO, Jang BI, Jang WJ, Lee SH. Laterally spreading tumors of the colorectum: clinicopathologic features and malignant potential by macroscopic morphology. Int J

TE D

Colorectal Dis 2013;28(12):1661-1666.

12. Rotondano G, Bianco MA, Buffoli F, Gizzi G, Tessari F, Cipolletta L. The Cooperative Italian FLIN Study Group: prevalence and clinico-pathological features

EP

of colorectal laterally spreading tumors. Endoscopy 2011;43(10):856-861.

AC C

13. Kim BC, Chang HJ, Han KS, Sohn DK, Hong CW, Park JW, Park SC, Choi HS, Oh JH. Clinicopathological differences of laterally spreading tumors of the colorectum according to gross appearance. Endoscopy 2011;43(2):100-107. 14. Horiuchi Y, Chino A, Matsuo Y, Kishihara T, Uragami N, Fujimoto Y, Ueno M, Tamegai Y, Hoshino E, Igarashi M. Diagnosis of laterally spreading tumors (LST) in the rectum and selection of treatment: characteristics of each of the subclassifications of LST in the rectum. Dig Endosc 2013;25(6):608-614. 15. Ishiguro A, Uno Y, Ishiguro Y, Munakata A, Morita T. Correlation of lifting versus

ACCEPTED MANUSCRIPT non-lifting and microscopic depth of invasion in early colorectal cancer. Gastrointest Endosc 1999;50(3):329-333. 16. Toyonaga T, Man-I M, Morita Y, Azuma T. Endoscopic submucosal dissection (ESD)

versus

simplified/hybrid

ESD.

Gastrointest

Clin

N

Am

RI PT

2014;24(2):191-199.

Endosc

17. Dixon MF. Gastrointestinal epithelial neoplasia: Vienna revisited. Gut 2002;51(1):130-131.

SC

18. Uraoka T, Saito Y, Matsuda T, Ikehara H, Gotoda T, Saito D, Fujii T. Endoscopic indications for endoscopic mucosal resection of laterally spreading tumours in the

M AN U

colorectum. Gut 2006;55(11):1592-1597.

19. Fujishiro M, Yahagi N, Kakushima N, Kodashima S, Muraki Y, Ono S, Yamamichi N, Tateishi A, Oka M, Ogura K, Kawabe T, Ichinose M, Omata M. Outcomes of endoscopic submucosal dissection for colorectal epithelial neoplasms in

TE D

200 consecutive cases. Clin Gastroenterol Hepatol 2007;5(6):678-683. 20. Huang Y, Liu S, Gong W, Zhi F, Pan D, Jiang B. Clinicopathologic features and endoscopic mucosal resection of laterally spreading tumors: experience from China.

EP

Int J Colorectal Dis 2009;24(12):1441-1450.

AC C

21. Fukami N, Ryu CB, Said S, Weber Z, Chen YK. Prospective, randomized study of conventional versus hybridknife endoscopic submucosal dissection methods for the esophagus: an animal study. Gastrointest Endosc 2011;73(6): 1246-1253. 22. Yamamoto H, Yahagi N, Oyama T. Mucosectomy in the colon with endoscopic submucosal dissection. Endoscopy 2005;37(8):764-768. 23. Liu Feng, Li Zhao-shen. Currect status of instrument development in endoscopic submucosal dissection. Zhonghua Xiao Hua Nei Jing Za Zhi 2012;29(12):661-664. 24. Shi X, Shan Y, Yu E, Fu C, Meng R, Zhang W, Wang H, Liu L, Hao L, Wang H,

ACCEPTED MANUSCRIPT Lin M, Xu H, Xu X, Gong H, Lou Z, He H, Xing J, Gao X, Cai B. Lower rate of colonoscopic perforation: 110,785 patients of colonoscopy performed by colorectal

AC C

EP

TE D

M AN U

SC

RI PT

surgeons in a large teaching hospital in China. Surg Endosc 2014;28(8):2309-2316.

ACCEPTED MANUSCRIPT

Figure Legends Figure 1: Subtypes of laterally spreading tumors.

RI PT

Figure 2: Patient’s clinical course after endoscopic submucosal dissection for laterally spreading tumors.

AC C

EP

TE D

M AN U

submucosal dissection for laterally spreading tumors.

SC

Figure 3: Kaplan–Meier estimates of survival in patients who underwent endoscopic

ACCEPTED MANUSCRIPT

Tables Table 1 Demographic data, clinico-pathologic features, and outcome of ESD treatment for LSTs. No.

Age (yrs); mean (range)

62.9 (34-85)

Sex (male/female)

92/73 (1.26:1)

Size (mm); mean (range)

52 (30-140)

LST type and subtype, n (%) 124 (70.1%)

Granular

65 (36.7%)

Nodular mixed

59 (33.3%)

SC

Homogeneous

Nongranular

53 (29.9%)

Pseudodepressed

Cecum

12 (6.8%)

3 (1.7%)

Ascending colon Transverse colon Descending colon

TE D

Sigmoid colon Rectum

41 (23.2%)

M AN U

Flat-elevated

Tumor location, n (%)

RI PT

Category

30 (16.9%) 29 (16.4%) 4 (2.3%)

13 (7.3%)

98 (55.4%)

Histological classification, n (%)

51 (28.8%)

Tubulovillous adenoma

41 (23.2%)

Villous adenoma

32 (18.1%)

Serrated adenoma

7 (4.0%)

HGIN/CIS

33 (18.6%)

Submucosal carcinoma

13 (7.3%)

EP

Tubular adenoma

AC C

Procedure duration (min); mean (range)

64.3 (10-294)

Outcomes of ESD En bloc resection; n (%)

147 (83.1%)

En bloc R0 resection; n (%)

144 (81.4%)

Postoperative bleeding; n (%; 95 % CI)

6 (3.4%; 95% CI, 0.7–6.1)

Perforation; n (%; 95 % CI)

4 (2.3%; 95% CI, 0.1–4.5)

Recurrence; n (%; 95 % CI)

11/142 (7.7%; 95% CI, 3.3–12.0)

ACCEPTED MANUSCRIPT

Granular LSTs All

Nongranular LSTs

Homo-

Nodular-

geneous

mixed

124

65

59

50 (30-140)

40(30-130)

60 (30-140)

72 (58.1%)

30(46.2%)

42 (71.1%)

9 (7.3%)

6(9.2%)

Transverse colon

24 (19.4%)

16(24.6%)

Cecum/ascending colon

19 (15.3%)

13(20.0%)

Tubular adenoma

33 (26.6%)

Tubulovillous adenoma Villous adenoma

Tumor size (mm); mean (range)

HGIN+Ca Outcomes of ESD, n (%)

depressed

P value

12 50 (30-100)

0.108

0.083

0.004

26(49.1%)

19 (46.3%)

7 (58.3%)

0.558

0.270

3 (5.1%)

0.374

8(15.1%)

5 (12.2%)

3 (25.0%)

0.276

0.105

8 (13.6%)

0.120

5(9.4%)

4 (9.8%)

1 (8.3%)

0.882

0.102

6 (10.2%)

0.129

14(26.4%)

13 (31.7%)

1 (8.3%)

0.106

0.083

23(35.4%)

10 (16.9%)

0.020

18(34.0%)

16 (39.0%)

2 (16.7%)

0.194

0.323

30 (24.2%)

18(27.7%)

12 (20.3%)

0.340

11(20.8%)

8 (19.5%)

3 (25.0%)

0.680

0.619

21 (16.9%)

12(18.5%)

9 (15.3%)

0.634

11(20.8%)

9 (22.0%)

2 (16.7%)

0.691

0.545

5 (4.0%)

4(6.2%)

1 (1.7%)

0.207

2(3.8%)

2 (4.9%)

0

0.935

0.936

26 (21.0%)

7(10.8%)

19 (32.2%)

0.003

7(13.2%)

4 (9.8%)

3 (25.0%)

0.170

0.225

1(1.5%)

8 (13.6%)

0.010

4(7.5%)

2 (4.9%)

2 (16.7%)

0.174

0.946

8(12.3%)

27 (45.8%)

<0.001

11(20.8%)

6 (14.6%)

5 (41.7%)

0.042

0.299

9 (7.3%) 35 (28.2%)

TE D

EP

Submucosal carcinoma

elevated

P2

41

AC C

HGIN/CIS

Pseudo-

40 (30-80)

Histological classification, n (%)

Serrated/ hyperplastic polyps

Flat-

53

0.034

M AN U

Descending/Sigmoid colon

All

50 (30-100)

Tumor location, n (%) Rectum

P1

SC

Number of lesions

RI PT

Table 2 Clinicopathological features and outcome of ESD treatment for specific subtypes of LSTs.

En bloc resection

102 (82.3%)

50(76.9%)

52 (88.1%)

0.103

45(84.9%)

36 (87.8%)

9 (75.0%)

0.276

0.667

En bloc R0 resection

100 (80.6%)

50(76.9%)

50 (84.7%)

0.271

44(83.0%)

35 (85.4%)

9 (75.0%)

0.400

0.710

Postoperative bleeding

5 (4.0%)

1(1.5%)

4 (6.8%)

0.138

1(1.9%)

1 (2.4%)

0

0.585

0.470

Perforation

1 (0.8%)

1(1.5%)

0

0.961

3(5.7%)

2 (4.9%)

1 (8.3%)

0.649

0.047

ACCEPTED MANUSCRIPT

6 (4.8%)

2 (3.1%)

4 (6.8%)

0.337

8/102 (7.8%)

5/57 (8.8%)

3/45 (6.7%)

0.695

Procedure duration (min); mean (range)

68.2 (11-294)

74.2 (11-294)

58 (15-170)

0.118

3/40 (7.5%) 48.1 (10-180)

AC C

EP

TE D

M AN U

SC

Recurrence

4 (7.5%)

3 (7.3%) 2/31 (6.5%)

RI PT

Bleeding + perforation

54.9 (16,175)

1 (8.3%)

0.907

0.236

1/9 (11.1%)

0.801

0.945

42.5 (10-180)

0.468

0.163

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT Author contributions

Zhi-Jie Cong, M.D., Liang-Hao Hu, M.D., Jun-Tao Ji, M.D. and Zhao-Shen Li, M.D.

published.

RI PT

analyzed the data, wrote and revised the draft, and approved the manuscript to be

Jun-Jie Xing and Yong-Qi Shan conducted the endoscopic interventions and provided clinical data, and approved the manuscript to be published.

SC

En-Da Yu, M.D. planned the study and revised the draft, and approved the manuscript

AC C

EP

TE D

M AN U

to be published.

ACCEPTED MANUSCRIPT Acronyms Colorectal Laterally Spreading Tumor –LST Homogenous Colorectal Laterally Spreading Tumor --LST-G-H

RI PT

Nodular Mixed Colorectal Laterally Spreading Tumor --LST-G-M Flat Elevated Colorectal Laterally Spreading Tumor --LST-NG-F

Endoscopic Submucosal Dissection --ESD

M AN U

Endoscopic Mucosal Resection --EMR High-grade Intraepithelial Neoplasia –HGIN

AC C

EP

TE D

Carcinoma in Situ -- CIS

SC

Pseudodepressed Colorectal Laterally Spreading Tumor –LST-NG-PD