Cancer Colon Precoces 0033 1344234

A large-scale multicenter study of long-term outcomesafter endoscopic resection for submucosal invasivecolorectal cancer

Authors Yusuke Yoda1, Hiroaki Ikematsu1, Takahisa Matsuda2, Yuichiro Yamaguchi3, Kinichi Hotta4, Nozomu Kobayashi5,Takahiro Fujii6, Yasuhiro Oono1, Taku Sakamoto2, Takeshi Nakajima2, Madoka Takao3, Tomoaki Shinohara4,Takahiro Fujimori7, Kazuhiro Kaneko1, Yutaka Saito2

Institutions Institutions are listed at the end of article.

submitted 16. August 2012accepted after revision6. May 2013

BibliographyDOI http://dx.doi.org/10.1055/s-0033-1344234Published online: 5.8.2013Endoscopy 2013; 45: 718–724© Georg Thieme Verlag KGStuttgart · New YorkISSN 0013-726X

Corresponding authorH. Ikematsu, MDDepartment ofGastroenterology and GIOncologyNational Cancer Center HospitalEastKashiwanoha 6-5-1KashiwaChiba 277-8577JapanFax: [email protected]

Original article718

Introduction!

Submucosal invasive colorectal cancer (SM-CRC)has been observed in approximately 0.2%–2% ofcolorectal polyps removed endoscopically [1,2].Lymph node metastases are found in 10%–15%of patients with SM-CRC [3–5]. While the likeli-hood of lymph node metastases from SM-CRCs islow, surgical resection with the removal of regio-nal lymph nodes is the standard treatment forthis disease [6]. Several histological features areassociated with lymph node metastasis and localcancer recurrence, including poorly differentiatedhistological features, vascular or lymphatic inva-sion, and incomplete resection [7,8]. Further-more, the rate of lymph node metastasis is signif-

icantly associated with the depth of tumor inva-sion into the submucosa [9].Consensus guidelines over the past 10 years havesuggested that endoscopic resection alone is ade-quate for treatment of low risk SM-CRCs, and thatsubsequent curative surgical resection should beperformed for high risk SM-CRCs [10,11]. Accord-ing to the Paris classification and the Japanese So-ciety for Cancer of the Colon and Rectum (JSCCR)guidelines, SM-CRCs were classified as having alow risk for lymph node metastasis and local re-currence when the following features were pres-ent: (i) negative vertical margin; (ii) well or mod-erately differentiated adenocarcinoma; (iii) noevidence of vascular or lymphatic invasion; and(iv) invasion depth of less than 1000µm [10,11].

Yoda Yasuke et al. Outcomes of endoscopic resection for SM-CRC… Endoscopy 2013; 45: 718–724

Background and study aims: Patients with sub-mucosal invasive colorectal cancer (SM-CRC)treated with endoscopic resection who are at lowrisk of lymph node metastasis and local recur-rence may be followed upwith observation alone,while additional surgery is recommended forthose with high risk features. However, the long-term outcomes that these strategies offer are stillunclear. The objective of our study was to retro-spectively evaluate the long-term outcomes of pa-tients with SM-CRCmanaged with endoscopic re-section.Patients and methods: We retrospectively ana-lyzed all patients with SM-CRC treated by endo-scopic resection at six institutions between 2000and 2007. SM-CRCs with (i) negative vertical mar-gin, (ii) well or moderately differentiated adeno-carcinoma, (iii) absence of lymphovascular inva-sion, and (iv) invasion depth<1000µmwere clas-sified as low risk. Patients with SM-CRCs withoutthese characteristics were classified as high risk.Outcomes were assessed by 5-year recurrence-free survival (RFS) and recurrence rate.

Results: During the study period, 428 patientswith SM-CRC (low risk, 126; high risk, 302) whounderwent endoscopic resection as their firsttreatment were enrolled (median follow-up 61months). Among the 120 patients with low riskfeatures treated by endoscopic resection alone,the 5-year RFS and recurrence rates were 98%and 0.8%, respectively. Of the 302 patients withhigh risk features, 196 underwent additional sur-gery and 106 were managed with endoscopic re-section alone. For those who underwent addi-tional surgery, the 5-year RFS and recurrencerates were 97% and 3.6%, respectively. Amongthe 106 patients managed with endoscopic re-section alone, RFS and recurrence rates were 89%(P<0.05 vs. low risk patients treated by endo-scopic resection alone) and 6.6% (P<0.05), respec-tively.Conclusions: Endoscopic resection alone is ade-quate for the management of patients with SM-CRC and low risk features. However, in those pa-tients with SM-CRC and high risk features, sur-gery should be considered in addition to endo-scopic resection.

Thi

s do

cum

ent w

as d

ownl

oade

d fo

r pe

rson

al u

se o

nly.

Una

utho

rized

dis

trib

utio

n is

str

ictly

pro

hibi

ted.

SM-CRCs without these characteristics were considered to carrya high risk.Management of SM-CRC must strike a balance between curingthe patient andminimizing the morbidity andmortality associat-ed with the treatment. However, the long-term outcomes for pa-tients with low risk SM-CRC undergoing endoscopic resectionalone are still unknown. Therefore, this study aimed to retrospec-tively evaluate the long-term outcomes of patients with SM-CRCtreated with endoscopic resection as their first treatment.

Patients and methods!

PatientsWe reviewed the clinical records of patients with SM-CRC at sixinstitutions from January 2000 to December 2007. The subjectswere recruited from patients who had undergone endoscopic re-section as first-line treatment for SM-CRC, which included endo-scopic mucosal resection (EMR), endoscopic submucosal dissec-tion (ESD), transanal endoscopic microsurgery, and local resec-tion.The protocol of this retrospective study was approved by the in-stitutional review board of each hospital. Before all procedures,comprehensive informed consent for treatment and clinical datacollection had been obtained from the patients; however, writteninformed consent was not obtained for this study itself. The studywas performed in accordance with the ethical principles thathave their origin in the Declaration of Helsinki.The exclusion criteria were as follows: (i) presence of familialadenomatous polyposis (FAP), hereditary nonpolyposis colorec-tal cancer (HNPCC), or inflammatory bowel disease (IBD); (ii)presence of active malignant disease in any other organs; (iii)presence of synchronous or metachronous advanced colorectalcancer; (iv) age>85 years; and (v) follow-up period of<1 year.

ProcedureAll patients were prepared for colonoscopy with 2–3L of pol-yethylene glycol–electrolyte solution. To avoid bowel move-ments, patients with no contraindications received intravenousscopolamine butylbromide (10mg) or glucagon (0.5mg) beforethe examination. Midazolam (0.03mg/kg) and/or pethidinehydrochloride (35mg) was used for conscious sedation onlywhen a patient complained of discomfort or pain.All examinations were performed using magnifying colono-scopes (CF-Q240ZI, PCF-Q240ZI, CF-200Z, and CF-H260AZI;Olympus Optical Co., Tokyo, Japan) that enhance the image up to75–100 times using a one-touch operation power system. Endo-scopically resectable tumors were identified by endoscopistsusing chromoendoscopy and magnification techniques. Lesionswere evaluated under magnifying chromoendoscopy with 0.4%indigo carmine and 0.05% crystal violet and assessedwith Kudo’sclassification and by clinical classification [12,13]. If lesions wereidentified as neoplastic noninvasive lesions (adenomas, mucosalcarcinoma, and slightly submucosal invasive carcinoma), endo-scopic resection was performed. If lesions were determined toextend deep into the submucosa, patients were referred for sur-gery. When patients rejected surgical resection or requestedendoscopic resection, endoscopic resection was performed asthe first-line treatment.

HistologyAll resected specimens were retrieved, immediately fixed in 10%buffered formalin solution, and examined histologically after he-matoxylin and eosin staining. The specimenwas cut into longitu-dinal slices measuring 2mm in width. Sections were evaluatedmicroscopically for resected margin status and tumor character-istics, including size, histological type, depth of invasion, vascularinvasion, and lymphatic invasion.Histological diagnosis was determined and measurements of tu-mor invasion depth were obtained according to the guidelines ofthe JSCCR [11]. When it was possible to identify the muscularismucosae, the depth of submucosal invasion was taken to be thedistance between the deeper edge of the muscularis mucosaeand the deepest invasion. When it was not possible to identifythe muscularis mucosae, the depth of invasion was the distancebetween the surface of the tumor and the deepest invasion. Inpedunculated tumors with disrupted muscularis mucosae, thedepthwas the distance between the deepest invasion and a refer-ence line defined as the boundary between the tumor head andthe pedicle.

Indications for treatmentIn line with the JSCCR guidelines, patients with SM-CRCs with (i)negative vertical margin, (ii) well or moderately differentiatedadenocarcinoma, (iii) no evidence of vascular or lymphatic inva-sion, and (iv) an invasion depth of less than 1000µmwere classi-fied as being at low risk for lymph node metastasis and local re-currence [11]. For lesions identified as being low risk SM-CRCs,endoscopic resection alone was indicated as a minimally invasivetreatment for these patients. Subsequent curative surgical resec-tion was indicated for patients with high risk SM-CRC.

Follow-upPhysical examinations, chest radiography, computed tomog-raphy (CT) of the abdomen and pelvis, and blood tests includingcarcinoembryonic antigen (CEA) level were performed post-operatively, every 6 months for the first 3 years, and every 12months thereafter. A total colonoscopy was performed everyyear.Recurrence was confirmed on the basis of imaging. Local recur-rence was defined as recurrence within the surgical field for co-lon cancer or within the pelvis for rectal cancer.

StatisticsClinical outcomes were assessed by 5-year recurrence-free survi-val (RFS), 5-year overall survival, and recurrence rates. RFS wasdefined as freedom from confirmed recurrence or death fromthe cancer; overall survival as death from any cause. Overall sur-vival and RFSwere calculated by the Kaplan–Meier method, andthe differences were compared using the log rank test with theSPSS package (SPSS 11.0 for Windows; SPSS Inc., Chicago, Illinois,USA).

Results!

A total of 487 patients received treatment for SM-CRC at our sixinstitutions, and 428 patients (266 men [62%]; median age 64,range 22–80) were enrolled for this study. The median follow-up period for all 428 patients was 60.5 months (range 12–75months). For 357 patients (83%) the follow-up period was more


Original article 719

Thi

s do

cum

ent w

as d

ownl

oade

d fo

r pe

rson

al u

se o

nly.

Una

utho

rized

dis

trib

utio

n is

str

ictly

pro

hibi

ted.

than 3 years; however, 151 patients (35%) were lost to follow-upwithin 5 years in this retrospective cohort study.An overview of the histological risk and treatments performedfor the enrolled patients is shown in●" Fig.1. According to thecriteria, 59 patients were excluded from this study, including 47patients whose follow-up period was less than 1 year. There were126 patients (29%) with low risk SM-CRC and 302 (71%) withhigh risk disease.Of the 126 patients with low risk disease, additional surgical re-section with lymph node dissection was performed for only sixpatients. For the rest, endoscopic resection alone was performed.

Of the 302 patients with high risk disease, 196 underwent addi-tional surgical resection; 106 patients did not undergo additionalsurgery because they refused or had significant comorbidities.With regard to the complications of treatment, there were nocases of perforation at endoscopic resection requiring emergencysurgery and no mortality from the additional surgery.The characteristics of each subgroup, according to their risk andtreatment, are shown in●" Table1. The median tumor size was15mm (range 3–100mm). The tumor was located in the rightcolon in 128 patients (30%), left colon in 199 patients (46%), andrectum in 101 patients (24%). There were no significant differen-ces in sex, age, tumor size, or location between the subgroups.All of the patients with positive vertical margins at endoscopicresection underwent additional surgery. Therefore, there wereno patients with positive margins treated by endoscopic resec-tion alone. Lymph node metastases were detected in 12% of thehigh risk patients who underwent additional surgery after endo-scopic resection.Among the 120 low risk patients treated by endoscopic resectionalone, there was one case of recurrence (recurrence rate 0.8%; 95% confidence interval 0.2%–4.5%). Distant metastases of the lungand bone were detected 23 months after endoscopic resection;the patient consequently received palliative treatment and diedof CRC. Three other patients died of other diseases. Lymph nodemetastasis and recurrence were not found in the six low risk pa-tients who underwent surgery after endoscopic resection (●" Ta-ble2). Among the low risk patients treated by endoscopic resec-tion alone, the RFS and overall survival rates were 98% and 94%,respectively.Among the high risk patients treated by endoscopic resectionalone, tumor recurrence was detected in 6.6% (7/106; P<0.05compared with the low risk patients who underwent endoscopicresection alone). Of the seven patients with recurrence, two pa-tients had distant metastasis and five patients had local recur-rence. Six of these patients underwent salvage surgery, and ofthese, three patients developed distant metastasis during fol-

▪ Follow-up period of <1 year (n = 47)▪ Synchronous or metachronous advanced colorectal cancer (n = 8)▪ FAP, HNPCC or IBD (n = 3)▪ Active malignant disease in any organs

Histologically proven submucosal invasive colorectal cancerafter endoscopic resection (n = 487)

Submucosal invasive colorectal cancer enrolled in this study

Low risk (n = 126)

Endoscopic resection alone

(n = 120)

Endoscopic resection alone

(n = 106)

Endoscopic resection and

additional surgery(n = 6)

Endoscopic resection and

additional surgery

(n = 196)

High risk (n = 302)

Fig.1 Flow diagram of patients entered into the study divided by patho-logical risk and treatment (FAP, familial adenomatous polyposis; HNPCC,hereditary nonpolyposis colorectal cancer; IBD, inflammatory bowel dis-ease).

Table 1 Clinical and pathological characteristics of the patients with low risk and high risk submucosal invasive colorectal cancer (SM-CRC) treated by endo-scopic resection with and without additional surgery.

Low Risk High Risk Total

(n=428)Without surgery

(n=120)

With additional surgery

(n=6)

Without surgery

(n=106)


(n=196)

Median age(range), years

65(37–80)

72(52–75)

72(32–80)

61(22–79)

64(22–80)

Sex, n

Male 85 4 71 106 266

Female 35 2 35 90 162

Location, n

Right colon 48 1 22 57 128

Left colon 56 4 47 92 199

Rectum 16 1 37 47 101

Median tumor size(range), mm

15(4–100)

15(8–40)

15(6–74)

16(3–100)

15(3–100)

Positive vertical margin atendoscopic resection, n 0 0 0 44 44

Lymphatic invasion, n 0 0 13 48 61

Venous invasion, n 0 0 8 43 51

Invasion depth > 1000 μm, n 0 0 96 178 274

Lymph node metastasis afteradditional surgery, n (%) – 0 – 23 (12%)


Original article720

Thi

s do

cum

ent w

as d

ownl

oade

d fo

r pe

rson

al u

se o

nly.

Una

utho

rized

dis

trib

utio

n is

str

ictly

pro

hibi

ted.

low-up, with two dying of CRC. Among the high risk patientstreated by endoscopic resection alone, two patients died of otherdiseases.For the high risk patients who underwent additional surgeryafter endoscopic resection, the recurrence rate was 3.6% (7/196;P=0.26 compared with the high risk patients who underwentendoscopic resection alone). A single patient had local recur-rence and six had distant recurrence. Four of these seven patientsunderwent salvage surgery, and three developed distant metas-tasis. These three patients died of recurrence, along with anotherthree patients who died of other diseases.Regarding the long-term outcome of the high risk group, the 5-year RFS was higher in patients who underwent surgery afterendoscopic resection than in those patients treated by endo-scopic resection alone (97% vs. 89%; P=0.13;●" Fig.2). In thehigh risk group, the 5-year overall survival rates were similar forpatients treated by endoscopic resection alone and for those whounderwent endoscopic resection with subsequent surgery. Timeto recurrencewasmostly within 60months, although cases of re-currence were reported more than 5 years after treatment. Theclinical and pathological characteristics of patients with recur-rence are summarized in●" Table3.

Discussion!

This is the largest retrospective, multicenter cohort study of long-term outcomes for patients with SM-CRC to date. Several pre-vious studies have reported pathological and prognostic factorsfor predicting lymph node metastasis in SM-CRC patients treatedwith colorectal resection and lymph node dissection [7–9]. Re-cently, other studies have reported the long-term outcomes ofpatients with SM-CRC treated with endoscopic resection [14–16]. Di Gregorio et al. looked at 105 patients with malignantpolyps over 5 years and reviewed past studies [15]. They report-ed an unfavorable clinical outcome was significantly more preva-lent in the high risk group compared with the low risk group.The present study is particularly important because consecutivecases of SM-CRC treated with endoscopic resection were ana-lyzed in both low risk and high risk groups over 5 years. Long-term outcomes of endoscopic resection alone were good in pa-tients with low risk SM-CRC (5-year RFS 98%). This outcomeshowed that endoscopic resection alone was adequate for thetreatment of low risk SM-CRC. In patients with high risk SM-CRC, 5-year RFS was slightly higher in those who underwent

endoscopic resection and additional surgery compared withthose not undergoing surgery (97% vs. 89%; P=0.13).In the present study, low risk SM-CRCs were defined as having anegative vertical margin, well or moderately differentiated ade-

1 2 63 74

P = 0.13

0

ER + additional surgeryER alone

5Time, yearsa

b

Recu

rren

ce-fr

ee s

urvi

val r

ate

100 %

90 %

80 %

70 %

60 %

50 %

40 %

1 2 63 74

P = 0.52

0

ER + additional surgeryER alone

5Time, years

Ove

rall

surv

ival

rate

100 %

90 %

80 %

70 %

60 %

50 %

40 %

Fig.2 Kaplan–Meier curves for the 302 high risk patients who were treat-ed with or without additional surgery showing: a recurrence-free survival;b overall survival (ER, endoscopic resection).

Table 2 Comparison of recurrence and survival rates in patients with low risk and high risk submucosal invasive colorectal cancer (SM-CRC) treated by endo-scopic resection with and without additional surgery.

Low risk SM-CRC High risk SM-CRC

Without surgery

(n=120)


(n=6)

Without surgery

(n=106)


(n=196)

Recurrence, n (%) 1 (0.8%) 0 7 (6.6%)1 7 (3.6%)2

Local recurrence 0 0 5 1

Distant metastasis 1 0 2 6

5-year recurrence-free survival, % 98% 100% 89%3 97%4

5-year overall survival, % 94% 100% 98% 99%

1 P<0.05 (vs. low risk patients treated by endoscopic resection alone).2 P=0.26 (vs. endoscopic resection alone in high risk patients).3 P<0.05 (vs. low risk patients treated by endoscopic resection alone; log rank test).4 P=0.13 (vs. endoscopic resection alone in high risk patients; log rank test).



Thi

s do

cum

ent w

as d

ownl

oade

d fo

r pe

rson

al u

se o

nly.

Una

utho

rized

dis

trib

utio

n is

str

ictly

pro

hibi

ted.

nocarcinoma, no evidence of vascular or lymphatic invasion, anda depth of invasion less than 1000 μm, according to JSCCR guide-lines. In patients whose SM-CRC fulfilled four of these low riskpathological factors, long-term outcomes of endoscopic resectionalone were sufficiently good. However, it is notable that therewas one case of recurrence among the low risk SM-CRC patients.For this case, the original pathology specimen and additional tis-sue slices were re-examined, and lymphovascular invasion wasfound in one of the additional slices. The case highlights that con-sideration should be given to the risk of recurrence even if the le-sion is a low risk SM-CRC and care should be taken with thepathological diagnosis of SM-CRCs. Moreover, this case emphasi-zes the importance of taking adequate levels through the speci-men and of review by expert gastrointestinal pathologists.Kudo first classified submucosal invasion as SM1 (upper third ofthe submucosa), SM2 (middle third of the submucosa), and SM3(lower third of the submucosa) [17]. Since then, Kikuchi et al.have reported lymph node metastasis in 0%, 10%, and 25% of182 patients with SM1, SM2, and SM3 invasive colorectal cancer,respectively [18].Conventional measurement of submucosal invasion using theSM1–SM3 grading was originally devised for examination of sur-gical specimens where the full thickness of the colonic wall wasavailable to the pathologist. Kitajima et al. [19] defined the super-ficial aspect of SM-CRCs and measured submucosal depth from

this baseline to the deepest point of invasion. They reported thatthe rate of lymph node metastasis was 0% if the submucosaldepth was<1000 μm for nonpedunculated tumors. This methodof measuring invasion depth is highly applicable to endoscopicresection specimens because the muscularis propria is not in-volved.Seitz et al. [20] classified SM-CRCs as low risk if a complete resec-tion was performed, there was no vascular invasion, and the le-sion was histologically well or moderately differentiated; theydid not observe recurrence after endoscopic resection of lowrisk SM-CRCs. In our study, however, three patients with recur-rence did not exhibit the classical risk factors, with the exceptionof invasion depth, which was greater than 1000µm. This showedthe “1000-µm rule” for submucosal invasion is a valuable factorfor predicting recurrence.Additional surgery is recommended for patients with high riskSM-CRC. Few studies have assessed the beneficial effect of sur-gery after endoscopic resection on oncological outcome in SM-CRC. Choi et al. [14] reported no cases of recurrence in 30 patientswith high risk SM-CRC whowere treated by endoscopic resectionwith subsequent surgery; three cases of recurrence occurred in20 patients who did not receive subsequent surgical treatment.In the present study, 196 of 302 patients with high risk SM-CRC(65%) underwent additional surgery. The 5-year RFS for patientswho underwent endoscopic resection and additional surgery

Table 3 Clinical and pathological characteristics of the 15 patients with submucosal invasive colorectal cancer (SM-CRC) who developed recurrence afterinitial treatment.

Sex;

age,

years

Location Size,

mm

Margin

status

at endo-

scopic

resection

Differen-

tiation

Lympha-

tic inva-

sion

Venous

inva-

sion

Depth

of inva-

sion

Treatment Time to

recur-

rence,

months

Location of

recurrence

Treat-

ment

after re-

currence

Male;65

Rectum 12 – Good – – Shallow Endoscopicresection alone

23 Distant Palliation

Female;55

Sigmoidcolon

15 – Moderate – – Deep Endoscopicresection alone

36 Local Salvagesurgery

Female;50

Rectum 8 – Good + – Deep Endoscopicresection alone


Male;80

Rectum 20 – Good – + Deep Endoscopicresection alone


Male;58

Rectum 45 – Good – – Deep Endoscopicresection alone


Female;55



Male;65


54 Distant Salvagesurgery

Female;69

Rectum 20 – Moderate – – Deep Endoscopicresection alone


Male;59

Rectum 10 + Good – – Shallow Endoscopic andsurgical resection


Male;63

Sigmoidcolon

50 + Good + + Deep Endoscopic andsurgical resection


Male;63

Rectum 20 – Moderate – – Deep Endoscopic andsurgical resection


Male;57

Sigmoidcolon

9 – Moderate + + Deep Endoscopic andsurgical resection


Male;66

Ascend-ing colon

10 + Moderate – – Deep Endoscopic andsurgical resection


Female;63

Sigmoidcolon

20 – Good – – Deep Endoscopic andsurgical resection

42 Local Palliation

Female;44

Rectum 25 + Good + – Deep Endoscopic andsurgical resection



Original article722

Thi

s do

cum

ent w

as d

ownl

oade

d fo

r pe

rson

al u

se o

nly.

Una

utho

rized

dis

trib

utio

n is

str

ictly

pro

hibi

ted.

was 97%, which was slightly higher than that seen in patientswho underwent endoscopic resection alone. Although this out-come was not significantly different, it suggests that patientswith high risk SM-CRCmay benefit from surgery after endoscopicresection.For high risk SM-CRC, it is still unknown whether endoscopic re-section before surgery influences lymph node metastasis, recur-rence, or long-term survival. Further studies are required to clar-ify these clinical questions. If endoscopic resection before surgeryincreases the incidence of lymph node or distant metastasis, itwould be crucial to diagnose high risk SM-CRCs with magnifyingchromoendoscopy or narrow band imaging (NBI), which arehighly effective for predicting the depth of invasion of colorectalneoplasms [13, 21].Among patients with high risk SM-CRC who were treated withendoscopic resection alone, tumor recurrence was detected inseven patients. Of these seven patients, five had local recurrenceonly, and they all underwent salvage surgery to treat the recur-rent cancer. At the time of analysis, two patients had had a fur-ther episode of recurrence or had died of the disease after thissalvage surgery.Weiser et al. [22] retrospectively reviewed 50 patients who un-derwent salvage surgery after transanal excision for submucosalinvasive cancer and muscularis propria invasive cancer. The ma-jority of the patients (49/50) had successful salvage surgery, and47 patients had pathologically complete resection. Disease-specific survival was 53%, which was lower than would havebeen expected for an early stage cancer. When high risk SM-CRCpatients are treated with endoscopic resection alone, the risk oflocal recurrence and the modest rate of cure with salvage surgeryshould be considered.Recurrence was detected in 15 patients in the present study:eight recurrences were detected before 3 years, five within 3–5years, and two were detected more than 5 years after the treat-ment. To detect recurrence early, several surveillance guidelineshave been proposed [23, 24]. Our results suggest that a follow-upprogram of at least 5 years is necessary for the early detection ofrecurrence. In this study, most of the local recurrences were ob-served in rectal cancer. Further analysis of the differences in long-term outcomes between colon and rectal cancer is required.There are some limitations to this study. First, this study is a ret-rospective analysis of the clinical records, andwe did not re-eval-uate all of the pathological features. Second, tumor budding,which has also been referred to as a risk factor for lymph nodemetastasis, was not evaluated in this study [25, 26]. Furthermore,with regard to the follow-up program after endoscopic resection,there are some limitations and a risk of bias. There were 71 pa-tients (16.6%) who were lost in follow-up within 3 years and, asthis was a retrospective study, some patients did not have ade-quate regular follow-up examinations.In conclusion, long-term outcomes were good in patients withlow risk SM-CRC treated with endoscopic resection alone. Endo-scopic resection alone was considered an adequate treatment forthis group, particularly in the interests of minimizing the mor-bidity and mortality associated with treatment. In patients withhigh risk SM-CRC, the higher rate of recurrence suggests that sur-gery is required in addition to endoscopic resection.

Competing interests: None

Institutions1 Department of Gastrointestinal Oncology & Endoscopy, National CancerCenter Hospital East, Kashiwa, Japan

2 Endoscopy Division, National Cancer Center Hospital, Tokyo, Japan3 Division of Endoscopy, Shizuoka Cancer Center, Shizuoka, Japan4 Department of Gastroenterology, Saku Central Hospital, Nagano, Japan5 Department of Diagnostic Imaging, Tochigi Cancer Center, Tochigi, Japan6 TF Clinic, Tokyo, Japan7 Department of Surgical and Molecular Pathology, Dokkyo University Schoolof Medicine, Tochigi, Japan

References1 Coverlizza S, Risio M, Ferrari A et al. Colorectal adenomas containing in-

vasive carcinoma. Pathologic assessment of lymph node metastaticpotential. Cancer 1989; 64: 1937–1947

2 Morson BC, Whiteway JE, Jones EA et al. Histopathology and prognosisof malignant colorectal polyps treated by endoscopic polypectomy.Gut 1984; 25: 437–444

3 Wilcox GM, Anderson PB, Colacchio TA. Early invasive carcinoma in co-lonic polyps. A review of the literature with emphasis on the assess-ment of the risk of metastasis. Cancer 1986; 57: 160–171

4 Cooper HS. Surgical pathology of endoscopically removed malignantpolyps of the colon and rectum. Am J Surg Pathol 1983; 7: 613–623

5 Tanaka S, Haruma K, Teixeira CR et al. Endoscopic treatment of submu-cosal invasive colorectal carcinoma with special reference to risk fac-tors for lymph node metastasis. J Gastroenterol 1995; 30: 710–717

6 Colacchio TA, Forde KA, Scantlebury VP. Endoscopic polypectomy: in-adequate treatment for invasive colorectal carcinoma. Ann Surg 1981;194: 704–707

7 Nascimbeni R, Burgart LJ, Nivatvongs S et al. Risk of lymph node metas-tasis in T1 carcinoma of the colon and rectum. Dis Colon Rectum 2002;45: 200–206

8 Muller S, Chesner IM, Egan MJ et al. Significance of venous and lympha-tic invasion in malignant polyps of the colon and rectum. Gut 1989;30: 1385–1391

9 Ueno H,Mochizuki H, Hashiguchi Y et al. Risk factors for an adverse out-come in early invasive colorectal carcinoma. Gastroenterology 2004;127: 385–394

10 The Paris endoscopic classification of superficial neoplastic lesions:esophagus, stomach, and colon: November 30 to December 1, 2002.Gastrointest Endosc 2003; 58: 3–S43

11 Watanabe T, Itabashi M, Shimada Y et al. Japanese Society for Cancer ofthe Colon and Rectum (JSCCR) guidelines 2010 for the treatment ofcolorectal cancer. Int J Clin Oncol 2012; 17: 1–29

12 Kudo S, Tamura S, Nakajima T et al. Diagnosis of colorectal tumorouslesions by magnifying endoscopy. Gastrointest Endosc 1996; 44: 8–14

13 Matsuda T, Fujii T, Saito Y et al. Efficacy of the invasive/non-invasivepattern by magnifying chromoendoscopy to estimate the depth of in-vasion of early colorectal neoplasms. Am J Gastroenterol 2008; 103:2700–2706

14 Choi DH, Sohn DK, Chang HJ et al. Indications for subsequent surgeryafter endoscopic resection of submucosally invasive colorectal carci-nomas: a prospective cohort study. Dis Colon Rectum 2009; 52: 438–445

15 Bories E, Pesenti C, Monges G et al. Endoscopic mucosal resection foradvanced sessile adenoma and early-stage colorectal carcinoma.Endoscopy 2006; 38: 231–235

16 Di Gregorio C, Bonetti LR, de Gaetani C et al. Clinical outcome of low-and high-risk malignant colorectal polyps: results of a population-based study andmeta-analysis of the available literature. Intern EmergMed 2012: [Epub ahead of print]. DOI: DOI 10.1007/s11739-012-0772-2

17 Kudo S. Endoscopic mucosal resection of flat and depressed types ofearly colorectal cancer. Endoscopy 1993; 25: 455–461

18 Kikuchi R, Takano M, Takagi K et al. Management of early invasive colo-rectal cancer. Risk of recurrence and clinical guidelines. Dis Colon Rec-tum 1995; 38: 1286–1295

19 Kitajima K, Fujimori T, Fujii S et al. Correlations between lymph nodemetastasis and depth of submucosal invasion in submucosal invasivecolorectal carcinoma: a Japanese collaborative study. J Gastroenterol2004; 39: 534–543

20 Seitz U, Bohnacker S, Seewald S et al. Is endoscopic polypectomy anadequate therapy for malignant colorectal adenomas? Presentation of114 patients and review of the literature Dis Colon Rectum 2004; 47:1789–1796



Thi

s do

cum

ent w

as d

ownl

oade

d fo

r pe

rson

al u

se o

nly.

Una

utho

rized

dis

trib

utio

n is

str

ictly

pro

hibi

ted.

21 Ikematsu H,Matsuda T, Emura F et al. Efficacy of capillary pattern typeIIIA/IIIB by magnifying narrow band imaging for estimating depth ofinvasion of early colorectal neoplasms. BMC Gastroenterol 2010; 10:33

22 Weiser MR, Landmann RG,WongWD et al. Surgical salvage of recurrentrectal cancer after transanal excision. Dis Colon Rectum 2005; 48:1169–1175

23 Waye JD, Braunfeld S. Surveillance intervals after colonoscopic poly-pectomy. Endoscopy 1982; 14: 79–81

24 Patchett SE, Mulcahy HE, O’Donoghue DP. Colonoscopic surveillanceafter curative resection for colorectal cancer. Br J Surg 1993; 80:1330–1332

25 Fujimori T, Fujii S, Saito N et al. Pathological diagnosis of early colorec-tal carcinoma and its clinical implications. Digestion 2009; 79: 40–51

26 Shimomura T, Ishiguro S, Konishi H et al. New indication for endoscopictreatment of colorectal carcinomawith submucosal invasion. J Gastro-enterol Hepatol 2004; 19: 48–55


Original article724

Thi

s do

cum

ent w

as d

ownl

oade

d fo

r pe

rson

al u

se o

nly.

Una

utho

rized

dis

trib

utio

n is

str

ictly

pro

hibi

ted.

Documents

Cancer Colon Precoces 0033 1344234