Endoscopic diagnosis of submucosal infiltration in colorectal cancer and treatment
The recent state in Japan

Nobuhiko Okawa

Division of Gastroenterology - Department of Internal Medicine
Showa University Fujigaoka Hospital - Yokohama, Japan

This time, as I have an opportunity to speak about colorectal cancer with submucosal invasion, I want to explain the present state in our country showing many endoscopic pictures.

We define the colorectal cancer whose invasion reaches the submucosal layer but does not reach the proper muscular layer as submucosal cancer (sm cancer). It dose not matter whether metastasis exists or not.

Acording to statistics, sm cancers account for 4-10% of all colorectal cancers which are detected in endoscopic examination. About 70-79% of all sm cancers exists in sigmoid colon and rectum, and the frequency by location well correlates with that of advanced cancers.

We sometimes find sm cancers of superficial type which are smaller than 10mm in size. So it is necessary to use endoscopes of good performance and to carry out enough lavage of the lumen to recognize the presence of these lesions. It is also important to carefully observe slightly uneven mucosa, change in color, small bleeding spot, loss of visible vascular pattern, slight deformity of the lumen arc etc., and dye spraying method (e.g. Indigo Carmine) will be a good help.

Macroscopically, sm cancer is classified into protruded type (Type I) and superficial type (Type II) in general. Type I can be further classified into pedunculated type (Ip), subpedunculated type (Isp) and sessile type (Is), and Type II into superficial elevated type (IIa, IIa+IIc), superficial flat type (IIb) and superficial depressed type (IIc, IIc+IIa). Some of the superficial elevated type lesions that are remarkably flat and extend along the lumen are called nodular aggregated tumor or laterally spreading tumor.

Frequency of each type is quite different among hospitals, i.e., 37-82% in whole protruded type and 18-47% in whole superficial type. Such discrepancy arises because the criterion of type Is, IIa and IIa+IIc differs in each hospital. In any case, at least 20% of all sm cancers is superficial type and perhaps their origin would not be a protruded adenoma.

In case of classifying the grade of submucosal (sm) invasion in colorectal sm cancer, the classification by Kudo is most frequently used because of its simplicity and convenience. This classification can be made by dividing the submucosal layer into three levels equally, i.e., sm1, sm2 and sm3 in order from the muscularis mucosae.

This classification has two problems as follows. In the case of endoscopically resected specimen, the lack of proper muscular layer makes it difficult to divide the submucosal layer into three equal parts accurately. And judgement of the degree of sm invasion can be influenced by the size and the macroscopic type (pedunculated type or sessile type etc.) of the lesion, even if the amount of sm invasion is the same. To solve these problems, we usually modify the classification by Kudo and judge the cancer invasion which slightly exeeds beyond the muscularis mucosae(e.g., 200-300µm or several glands) as sm1(or mild sm invasion) and the other sm invasion as sm2 and sm3(or massive sm invasion together) in the case of endoscopically resected specimen.

Clinically it is impotant to differentiate from sm1 cancer which might be treated with endoscopic removal because of its very low lymph node metastatic ratio to sm2-3 cancer which needs surgical resection.

At present, it is generaly considered that mucosal and sm1 cancer can not be differentiated by the endoscopic features. On the other hand it is almost possible to differentiate from m and sm1 cancer to sm2 and sm3 cancer(moderately to massively submucosal invasive cancer). It is reported that a skillful colonoscopist can diagnose with precision 70-80% of cancers with massive sm invasion.

During the lecture, I will explain practical diagnosis of sm cancer showing endoscopic pictures, but here I only list macroscopic features which were often observed in sm2 and sm3 cancer. They are rigidity of the lumen arc, SMT like marginal swelling, fold convergence, deep central depression, uneven base of depression, abnormal protrusion, surface erosion, impression of tenseness, tumor redness, luck of luster, amorphous findings of surface stracture, circumferential white spot etc..

These are not definite findings for diagnosis, but only show the possibility of sm2 and sm3 cancer. We also take into consideration size and macroscopic type of cancers etc. in practical diagnosis.

The following is the outline of treatment plan for colorectal sm cancer. In case of the lesion which can be highly suspicious of sm2 and sm3 cancer, surgical operation with lymph node dissection is chosen. If the lesion does not show endoscopic features which can not be highly suspicious of sm2 or sm3 cancer, endoscopic treatment is performed at first.

After endoscopic resection we should consider histological risk factors which correlate with higher incidence of lymph node metastasis. They are as follows; 1) moderate to massive sm cancer invasion, 2) lymphatic and venous vessel permeation, 3) moderately to poorly differentiated adenocarcinoma (histological type in the main body of the tumor), 4) moderate or poor tumor differentiation in the submucosal invasive front, 5) tumor budding(ahead of the submucosal invasive front ).

If there are no such risk factors as mentioned above, treatment can be finished with endoscopic resection only. In other case, additional surgical bowel resection with lymph node dissection is needed.

Almost all reports mention that the rate of lymph node metastasis of whole colorectal sm cancers which were resected by surgical operation is about 10% (3-5% in case of sm1). Moreover, since minutely sm invasive cancers, which were endoscopically resected and were not evaluated whether lymph node metastasis existed or not, are not included in these statistics, I suppose that the true rate of lymph node metastasis of sm cancer might be smaller. So this fact signifies that surgical operation was not necessary in about 90% of the cases. Of course, there are some difficult cases which can not be resected endoscopically for technical reasons. However, it is true that we have performed surgical operation too much as a result.

There is no agreement of opinions yet, though many studies have been done so far in order to extend the indication for endoscopic treatment of sm cancer. However, some evaluations of sm cancers without lymph node metastasis indicate that it is possible to consider sm cancers whose vertical depth of sm invasion is less than 1000µm as sm1 cancer.

Because tumors over 20mm in size are often required piecemeal resection and it is not easy to judge the degree of sm invasion of such specimen properly, indication of endoscopic treatment for large size lesions should be considered carefully.

It is reported that 5-10% of sm cancers recurred after endoscopic treatment and almost all of recurrent cancers were found at the same site as the initial lesion. Therefore, follow-up colonoscopy is necessary for long term. Surveillance should be continued at least for two years but there is such a case that a recurrence tumor was found six years later.

In case of treatment of most superficial type and some of sessile type lesions, endoscopic mucosal resection (EMR) is often performed in order to make easy to snare the lesion and resect them safely.