V.I. Nesterova, T.G. Kuznetsova,
V.I. Metlushko, N.L. Ogluzdina
Introduction
Colonoscopy is successfully used today to diagnose colon new growths. Based on a
number of indications endoscopic investigation provides reliable information about the
colonic growth surface in order to correctly classify its pattern and take a sample for
morphological identification. Yet, colonoscopy does not give an idea of the kind of
internal structure the new growth has, nor does it allow an assess the depth of the
invasion of the colon wall by a maligant tumor, determine its proliferation to adjacent
organs or metastases to regional lymph nodes. Besides, colonoscopy does not provide
information about extra intestinal new growths unless they have already permeated the
intestinal wall.
The NLS-investigation of the colon using a 4.9 GHz high frequency nonlinear sensor can help
clear up all of these issues.
The NLS-investigation provides information on intestinal wall layers and the adrectal
cellular tissue.
This research aimed to define the potentials of the NLS-method in a more specific
diagnostic of straight-and segmented intestine tumors.
The matter and investigation methods
87 patients were examined in whom 91 new growths were investigated using NLS
techniques. The examinees included 41 men and 46 women aged from 31 to 83 with
most of them (82%) aged 50 and over. All the patients affected by colon new growths
were subject to surgical treatment depending on the patient, size and location of the
growth. In 23 cases endoscopic polypectomy was performed. In 61 cases a resection was
undertaken on different parts of the colon and in 3 patients trans-anal endo-microsurgery
was performed. All of the NLS investigation results were verified by a
pathomorphological examination of macro preparations according to which the colonic new
growths were represented by simple tumors in 30 cases and by glandular cancers
with different degrees of differentiation in 61 cases.
The stages of the malignant process were defined according to TNM classification
adopted by the International Anticancer Association in 1997 (the 5th revision). Phase T1
was diagnosed in 13 patients (21%), phase T2 in 26 patients (43%), phase T3 in 17
patients (28%) and phase T4 in 5 patients (8%).
According to a patho-morphological examination, metastases into regional lymph nodes
were detected in 11 of 61 cases.
14All the patients underwent NLS-investigation and ultrasound colonoscopy to diagnose and
localize new growths, define their size, growth patterns and approximate
morphological characteristics together with ultrasound scanning of the abdominal cavity
and small pelvic organs to assess the condition of the organs adjacent to the colon and
diagnose distant metastases.
The NLS-investigation used a 4.9 GHz nonlinear sensor. The endoscopic
ultrasonography made use of the endoscopy ultrasonographic system UM-20 complete with
the ultrasonic colonoscope CF-UM20 (Olympus, Japan). The echographica of the abdominal
cavity made use of the diagnostic unit SSD-630 (Aloka, Japan) and Logiq-700 (General
Electric, USA).
Discussion of results
We know from experience that every NLS-investigation should be preceded by
diagnostic colonoscopy, which evaluates anatomic characteristics of the colon and
defines the number, localization and macroscopic characteristics of the new growth, and by
ultrasound scanning of the abdominal cavity as well. A thorough trans-abdominal
ultrasound scanning is required to assess the condition of the organs adjacent to the colon
and diagnose remote metastases.
A comparison of the NLS results with those of pathomorpholocal investigations was
made in order to define the potential of the NLS-method in differential diagnosis of
benign and malignant colonic new growths.
The results of the NLS-investigation coincided with the pathomorphological investigation
in 87 of 91 cases. Most of the errors occurred in diagnosing colon adenomas. In 6 of 31
cases the patient was suspected of having cancer. The analysis of the observations noted
that difficulties in diagnosis were related to the deformation of intestinal wall layers due
to the pressure of a nodal villous tumor rather than to a genuine invasion. Two falsenegative results were obtained in the case of malignant adenoma and cancer diacrises.
Thus, the accuracy of the NLS method in differential diagnosis of malignant and benign
colon tumors amounted to 81.3% of sensitivity to 79.8%, while the specificity was 76.4%.
The method of treatment to be chosen for patients affected by colon cancer depends on
the tumor process phase. A comparison was made to the patho-morphological
investigation data in 61 cases in order to assess the diagnostic efficiency of the
NLSmethod in classifying the colonic cancer phase.
The correct definition of the phase of tumor process was possible in 68.4% of the
observations. The best results were obtained in defining phases T3 and T4, where the
diagnostic accuracy was 78.2% and 81.2% respectively. It should be noted that most of
the errors occurred in determining phases T1 and T2, where the data of NLS and patho-
15morphological investigations coincided only in 54.2% and 47.4% of the observations
respectively.
In diagnosing phase T1 mistakes were made in 4 cases with 3 of the errors toward
overstating the phase; in one case signs of intestinal wall invasion were not found and the
tumor was taken for adenoma. In the analysis of phase T2 diagnostic errors in overstated
phases were noted in 7 of 9 cases; an under-statement of phase of the tumor process
occurred in one case and yet in one case no evidence of invasion proved to be found. The
analysis of the post surgical morphological conclusions revealed that in 6 of 7 false
positive results the patho-morphological investigation of a macro preparation detected a
deeper infiltration into the intestinal wall. However, according to microscopic
examination, the infiltration was of inflammatory rather than of a tumor kind. It should
also be noted that in all of the cases it had to do with an infiltrative tumor process in the
inferior ampullar section of the straight intestine free of serous membrane while the
inflammatory infiltration area was located in adrectal cellular tissue.
To understand better where there is imprecision in diagnosis, the efficiency of the
NLSmethod was analyzed in terms of the size, localization and form of germination of
neoplasms. The best results were obtained in diagnosing new growths sized under 2 cm and
over 5 cm.
Epithelial tumors over 5 cm in size is represented by phases T3 and T4 in 12 of 17 cases.
It has to be noted for large neoplasms the data of NLS assay did not coincide with pathomorphologic data only in phase T2, where the process phase was overestimated because
of the presence of inflammatory infiltration in deeper layers, than the layers where the
tumor invasion occurred. Thus, at neoplasms larger than 5 cm in size the diagnosis of the
invasion degree of the intestinal wall correlates in 78.2% of observations. High
correlation was also obtained for depth of tumor invasion by neoplasms sized up to 2 cm.
Most of them are represented by a tumor in phases T1 and T2. The results of ultrasonic
colonoscopy coincided with those of patho-morphologic conclusions in 76.7% of the
observations. It should also be noted, that tumors sized up to 2 cm are most convenient
for examination since they have the least number of artefacts.
In this study the greatest group was the tumors sized from 2 to 5 cm, where the results
proved to be lower, than in two first groups. The NLS data and those of the pathomorphologic essays coincided in 66.7% of cases. An appreciable error band (60%)
occurred in phase T2, where the intestinal wall invasion depth was overestimated in all
observations.
The great value has the fact, that according to pathomorphologic essay, in 5 of 6 cases of
hyper-diagnosis apart from the tumor infiltration an expressed inflammation was detected in
deeper layers of the intestinal wall. The relatively low accuracy of diagnosed depth of the
intestinal wall invasion by a tumor sized from 2 to 5 cm is due to the fact that 24 of 30
observations of this group corresponded to phases T2 and T3. A differential diagnosis of the
tumor infiltration depth in these phases is complex.
16At the next research stage we made comparative analysis of the effect of the form of
growth of the neoplasm for accuracy of defining the phase of tumor invasion in the
intestinal wall. All neoplasms were classified into three groups. In function of the shape of
the tumor growth: polypiform, saucer-shaped and infiltrative.
The highest results were obtained when diagnosing the phase of the saucer-shaped
growth cancer process where the accuracy of defining the tumor invasion in the intestinal
wall was 78.3%.
It seems however possible to fully estimate the accuracy of the NLS method in defining
the depth of a tumor invasion at neoplasms with saucer-shaped growth because of its
dismal occurrence among other forms of patients surveyed by us.
The polypiform of the growth was noted in 30 neoplasms. The growths had a distinct
interface with unaltered sections of the intestinal wall and did not block the intestine
lumen by more than half, which created favorable conditions for the survey. The
accuracy of NLS method in defining the depth of tumor invasion in the intestinal wall
was as high as 65%. It has to be noted, that half of all cases divergent with the
pathomorphologic conclusions is due to the overestimated depth of tumor infiltration at
defining the phase T2, which is connected with the presence of a perifocal inflammation.
This fact suggests difficulties in defining the phase of a cancer process in cases where the
tumor invasion is compounded by the inflammatory component penetrating deeper layers
of the intestinal wall and beyond its limits.
The neoplasms with an infiltrative growth shape have proved to be most difficult in
defining the degree of the tumor invasion into the intestinal wall. In this this group the
results of NLS method and those of the pathomorphologic essays coincided only in
49.8% of observations. It was due to the fact that these neoplasms, as a rule, had a large
size and occupied more than a half of the intestine wall circle.
In the next investigation phase we estimated the accuracy of the NLS method in defining
the degree of the intestinal wall invasion depending on the tumor location in the colon.
In 40 cases the tumor was localized in the rectum and in 21 cases in the segmented
intestine. The accuracy of diagnosing the phase of the tumor process in the colonic
intestine is significantly higher, than at finding the tumor invasion depth with the
neoplasms located in the rectum and amounts to 71 and 62.5% respectively. This high
result can be most likely explained by the fact that this department of colon contains a
serous membrane, which distinctly separates the muscular layer from the abenteric organs
and tissues. Also it is noted, that the serous membrane of the intestine is less subject to
penetration of the inflammatory infiltration, than the para-rectal cellular tissue. The
majority discrepancy relates to over-estimated depth of the invasion at defining Phase T2.
These researchers have noted that the accuracy of diagnosing the phase of a tumor
process was higher in colonic intestine, than in rectum. The greatest number of
17discrepancies occurs in Phase T2, which is conditioned by the presence of abscesses,
inflammatory infiltration or radial therapy in the neoplasm area.
Damaged regional lymph glands are an important prognostic factor in diagnosing rectum
cancer. To define the capabilities of the method in diagnosing metastases in regional lymph
glands, the results of the NLS method were compared with those of the pathomorphologic
essay. In the latter the malignant damage to the regional lymph glands was detected in 11
observations from 22 cases.
The analysis of the derived data proved that the NLS essay had correctly defined the
pattern of damage to the lymph glands in 63.6% of cases.
The metastatic pattern of damage to the lymph nodes was defined in 74.8% of cases, and
in inflammatory changes the results of the ultrasonic colonoscopy and those of the pathomorphologic essay coincided only in 45.5% of observations. In 6 from 11 of cases the
presence of metastasizes in lymph nodes was assumed (false-positive result). Such
mistakes can be attributed to oneologic vigilance of the researcher and complexity of
differential diagnosis of inflammatory and metastatically-altered lymph glands.
Conclusions
1. NLS diagnosis is a highly efficient method of diagnosing the neoplasms of the colon,
allowing to diagnose neoplasms and regional lymph glands.
2. The NLS method enables to detect the colon adenoma and cancer by the presence or
absence of the tumor invasions in the intestinal wall.
3. The diagnostic efficiency of NLS method in defining the phase of tumor process in the
rectum is lower, than in segmented intestine.
4. The diagnostic accuracy of the cancer phase in colon depends as much on the size as
on the anatomic shape of the tumor growth. The best results were obtained at
defining depth of invasion of the intestinal walls by a tumor sized under 2 cm and
over 5 cm.