Characteristics of Colorectal Cancer and Adenoma Observed in Surveillance Colonoscopy after Endoscopic Resection of All Neoplasms

Masafumi Tabuchi Nakameguro Digestive Disease Clinic, Department of Surgery University of Tokyo School of Medicine

tabuchi@mrg.biglobe.ne.jp

Aim: Little is exactly known about the speed of growing and invading of colorectal cancer in early stage and recurrence rate of tumors. We must elucidate this problem, to make the optimal guideline for surveillance colonoscopy after colonic tumor resection.

Method:

From 1988 to 2004, consecutive total colonoscopies were designed to make colon free of cancer and adenoma, (namely clean colon). All cancer and adenoma was entirely removed in each colonoscopy. Even less than 5mm tumors were the object of resection. To make the data precisely, for detecting tumor, more than 20 minutes were spent and dye-splay method was used in every colonoscopy. All suspicious lesion more than 1.5mm in size was tried to be undergone the magnified examination. To distinguish the tumor, pit-pattern diagnosis criteria was applied with high resolution magnified chromo-endoscope.

Tumors found after this clean colon procedure are recognized to be representative of newly borne tumors in earlier stage. We compared the data of surveillance tumors to initial ones to elucidate the characteristics of colonic tumors in earlier time.

In this study, initial tumors are defined that were detected and removed with endoscope within one year from the first colonoscopy. Surveillance tumors are defined that were detected and removed by endoscope after twice procedure of clean colon and more than one year past from the first colonoscopy.

Objects:

The initial ones were 8479 cases and mean age was 54.3±11.1 and gender ratio was 33.9%.

The surveillance ones were 13101 colonoscopies, average of interval was 412±253 day and mean age was 59.8±9.8 and gender ratio was 25.7%. The time of observation was 14783 person-years.

Results:

In the initial data, there were

100 advanced cancers (C1, A16, T12, D12, S24, R35: right colon/total colon=rr=29/100=29%),

100 sub-mucosal cancers (including

25 pedanculated (average of size: 16.9±5.2mm)(rr=2/25=8%), 42 sessile (14.8±7.8mm) (rr=8/42=19%) and

21 depressed (13.9±6.7mm)(rr=10/21=48%)),

824 intra-mucosal cancers (including

297 pedanculated (14.6±6.3mm)(rr=80/297=27%),

351 sessile (11.3±6.7mm)(rr=107/351=30.5%) and

45 depressed (12.3±9.4mm)(rr=29/45=64%)),

812 adenomas with high grade dysplasia (including

178 pedanculated (11.3±3.9mm)(rr=64/178=36%),

488 sessile (7.4±3.9mm)(rr=205/488=42%) and

39 depressed (12.3±9.4mm)(rr=23/39=59%)),

23046 adenomas with low grade dysplasia (including

520 pedanculated (8.9±4.1mm)(rr=243/520=47%),

11777 sessile (4.0±2.1mm)(rr=7041/11777=60%) and

1469 depressed (3.6±2.2mm)(rr=897/1469=61%)),

219 villous tumors (14.1±9.3mm),

57 serrated adenomas (9.6±7.1mm).

Initial Data

Adv ca	100
sm ca	25	41	13	21
m ca	297	351	131	45
HGD	178	488	107	38
LGD	520	11777	9280	1469
	pedanculated	sessile	others	Depressed

Case1, KI Age of 61 male sm1 depressed cancer, 4mm in size, was detected in the 3^rd colonoscopy, 1 month after the last colonoscopy and 2 month .after first colonoscopy

In surveillance data, there were

4 advanced carcinoma consisting of

1 newly detected advanced cancer,

1 direct invasion from pancreas cancer and

2 recurrece from the first advanced colonic cancers ,

3 sub-mucosal cancers (including

0 pedanculated,

1 sessile (20mm,rr=1/1) and

2 depressed (9.5±9.1mm rr=1/2),

18 intra-mucosal cancers (including

2 pedanculated (5.0mm±0mm rr=1/2),

7 sessile (5.8±2.7mm rr=3/7=42%) and

5 depressed (7.4±3.9mm rr=3/5=60%)),

32 adenomas with high grade dysplasia (including

1 pedanculated (10mm, rr=1/1) and

6 sessile (3.2±1.7mm rr=4/6=67%) and

7 depressed (4.0±3.3mm rr=4/7=57%),

10698 adenomas with low grade dysplasia (including

11 pedanculated (5.6±2.5mm rr=8/11=73%),

3982 sessile (2.8±1.2mm rr=2745/3982=69%,) and

797 depressed (3.1±1.6mm rr=547/797=69%),

7 villous tumor (6.5±7.0mm),

12 serrated adenoma (5.3±6.0mm).

Surveillance Data

Adv ca	1
sm ca	0	1	0	2
m ca	2	7	4	5
HGD	1	6	18	7
LGD	11	3982	5908	797
	pedanculated	sessile	others	Depressed

In surveillance, advanced cancers, pedanculated tumors and villous tumors were small in number. This means that it takes more than one year for them to develop their own stage. On the contrary, the rate of depressed mucosal and sub-mucosal cancers was much larger than that of other polypoid type of early stage of cancers. So, depressed tumors were thought to develop more rapidly than other polypoid type of tumors.

Case2, YK Age of 83 female sm2 depressed cancer, 3mm in size, was detected in the 13^th surveillance colonoscopy, 7 months after the last surveillance colonoscopy and 4 year .after first colonoscopy

Considerations:

Initial data is thought to be in proportion with prevalence data. And surveillance data is thought to be in proportion with incidence data. There is a relationship between prevalence, incidence and duration that prevalence equals to incidence multiplied by duration. So the rate of surveillance data divided by initial data is thought to be in proportion with a reciprocal value of duration to stay in each category; namely the rate of surveillance data divided by initial data is in proportion with the speed of changing such as promotion, demotion or disappearance. In the phase of mucosal and sub-mucosal cancer, it is hard to imagine that tumors disappear or demote. So in the phase of early stage of cancer, the rate of surveillance data divided by initial data is thought to be in proportion with the speed of promotion.

S/I ∝ Speed of Changing

Adv ca	1.0％
sm ca	0.0%	2.4%	0.0%	9.5%
m ca	0.7%	2.0%	3.1%	11.1%
HGD	0.6%	1.2%	16.8%	18.4%
LGD	2.1%	33.8%	63.7%	54.3%
	pedanculated	sessile	others	Depressed

From the data of this study, the value was 7/66=10.6% for depressed early stage of cancer and 8/392=2.04% for sessile early stage of cancer. This consideration induces the hypothesis that depressed type of early cancers invade 5 times faster than sessile type of early cancers. From the reported study of size and invasion in my data and other Japanese groups, the size, where half of early cancers invade into sub-mucosal layer, is about 7mm for the depressed and about 13mm for the polypoid. This fact is also consistent with the hypothesis.

Conclusions:

It is estimated that it takes more than one year to form the pedanculated shape, villous tumor and advanced cancer. In the surveillance colonoscopy, the most dangerous lesion was thought to be depressed cancers because they were estimated to invade 5 times faster than sessile type of cancers.