Abstract
Background: Macroscopic serosal classification of gastric cancer has been reported in previous studies, but rarely reported about it of colorectal cancer. The purpose of this study was to propose a macroscopic serosal classification of colorectal cancer and to investigate clinical significance of this classification. Materials and methods: Morphologic features of colorectal cancer were analyzed according to the macroscopic serosal appearance and clinicopathologic characteristics of these patients were retrospectively reviewed. Microscopic serosal structure was compared between different types under light microscope and transmission electron microscope. Results: Macroscopic serosal classification was divided into normal type, reactive type, nodular type and colloid type according to the macroscopic serosal appearance and microscopic structure. There were significant differences in tumor size, tumor gross type, histological type, histological grade, tumor necrosis, pT stage, number of nodes metastasis, lymph node metastasis ratio, pN stage, M stage and peritoneal metastasis between patients with different serosal types. Univariate analysis of prognosis revealed macroscopic serosal classification as one of factors significantly correlated with patient survival. However, multivariate analysis only revealed TNM stage significantly correlated with patient survival, while macroscopic serosal classification did not, maybe due to insufficient samples. Conclusions: Macroscopic serosal classification of colorectal cancer is preliminarily defined and divided into four types. Different macroscopic serosal types indicate different clinicopathologic features and correlate with prognosis of patients with colorectal cancer, but still cannot be proven as an independent factor.
Keywords: Colorectal cancer, macroscopic serosal classification, clinicopathologic characteristics, prognosis
Introduction
Macroscopic serosal classification (MSC) of gastric cancer was defined firstly in 1986 by professors of our department, based on the observation of more than 200 patients before 1986 and had reached agreement among all the surgeons and pathologists in First Affiliated Hospital of China Medical University in that year. But this study was published in Chinese journal and had not be known worldwide [1]. MSC of gastric cancer was classified as normal type, reactive type, nodular type, tendonoid type and color-diffused type, playing a critical role in predicting peritoneal micrometastasis. Compared with other methods, MSC was more sensitive than peritoneal cytological examination of peritoneal wash samples and more convenient than quantitative detection of the expression level of CEA mRNA which need a complex and time-consuming procedure. Further analysis also revealed that MSC was an independent predictor for both peritoneal recurrence and prognosis [2]. Besides peritoneal metastasis, lymph node metastasis also had been confirmed to have some relationships with serosa invading and MSC of gastric cancer [3,4].
Colorectum as a distal part of digestive system has some similarities in the structure and biological behavior with stomach. But very few studies focus on analyzing clinicopathologic features and prognosis according to serosal invasion of colorectal cancer, more rare for classifying macroscopic serosal type. It may because that approximately half side of colorectum was wrapped by mesentery or in retroperitoneum, so near a half of colorectal cancer couldn’t be viewed in the side of serosa. However, for those patients that lesions invade serosal side, macroscopic serosal appearance will be observed firstly by surgeons when they perform the operation. If macroscopic serosal type correlates with clinicopathologic characteristics and can predict the prognosis of patients with colorectal cancer, it can determine the extent of rational resection and application of other anti-cancer treatments during surgery. In the present study, we propose a macroscopic serosal classification (MSC) of colorectal cancer according to morphology and structure of serosa, then hypothesize different macroscopic serosal types have different clinicopathologic characteristics and prognosis. Based on light microscope (LM) and transmission electron microscope (TEM), we compared microscopic serosal structure and components between different macroscopic serosal types. Furthermore, we investigated clinicopathologic characteristics and prognosis through a retrospective analysis of 213 patients with colorectal cancer.
Materials and methods
Patients with colorectal cancer underwent surgery at department of colorectal surgery, Liaoning Cancer Hospital & Institute during December 2010 to September 2011 were entered into this study. A total of 213 cases was suitable to analyze, 129 were men and 84 were women, with an age range from 28 to 86 years old (median 60). The change of bowel habits and bloody stool were the original symptom. Preoperative colonoscopy was taken to confirm tumor site and pathological diagnosis by biopsy. Application of computed tomography, ultrasound and magnetic resonance or other imaging techniques was to explore invasive depth of tumors, lymph node and distant metastasis. Histological staging was in accordance with the latest 7th edition of the UICC TNM system (T is primary tumor; N is regional lymph nodes; M is distant metastasis), with a total of four stages [5]. I to IV stage were 26 cases (12.2%), 69 cases (32.4%), 83 cases (39.0%) and 35 cases (16.4%), respectively. Among patients of IV stage, there were 13 cases of peritoneal metastasis, accounting for 6.1% of the total. All patients were firstly treated with curative resection for primary tumor, followed with appropriate chemotherapy or radiation regimen according to TNM stage [6]. Resections were deemed curative when no gross residual disease was evident at the time of operation, with tumor-free margins on histological examination [7].
Patient selection criteria: (1) primary colorectal cancer, with or without metastasis; (2) lesions were located on the serosal side; (3) curative resection for primary tumor, regardless of metastasis; (4) without preoperative therapy; (5) patients undergoing elective surgery. The following criteria were excluded: (1) tumor directly invaded or was adherent to other organs or structures (T4b); (2) palliative, debulking surgery for primary tumor; (3) with other malignant tumors synchronously; (4) history of abdominal operation.
The type of serosa was prospectively determined according to their macroscopic appearance by three surgeons, who had experiences in more than fifty surgeries of colorectal cancer, immediately after the abdominal cavity being opened. If there was co-existence serosa type in one lesion, only the type that occupied the maximal area was recorded for analyzing. The same time we obtained a piece of serosa with 1.0-cm diameter and 0.2-cm thickness and divided it into three parts after rinsed in physiological 0.9% NaCl solution. One part fixed in 2.5% glutaraldehyde buffered with 0.1 M phosphate (pH 7.4) for transmission electron microscopy and the second part fixed in 10% formalin for light microscopy. The third one was stored in -80°C ultra-low temperature freezer (MDF-382E, SANYO, Japan), in order to extract proteins and mRNAs from serosal tissues for further study in the future.
Microscopic serosal appearance of 10% formalin-fixed tissues were observed with LM by experienced pathologist using hematoxylin-eosin staining as previously described [8]. Longitudinal sections of lesions after fixation, staining and a series of steps, were also observed with LM by the same pathologist. TEM sample preparation: the specimens fixed in 2.5% glutaraldehyde buffered, after fixation for 2-3 hours in 1% OsO4, were rinsed with 0.1 M phosphate and dehydrated three times in an ascending ethanol series (50, 70, 90, and 100%), then embedded by Epon 812. The semi-thin sections (about 50 nm thickness) obtained by the slicer (PowerTome-XL, RMC company, US), double stained with 3% lead citrate and uranyl acetate, were visualized with TEM (H-7650, Hitachi, Japan) in the accelerating voltage at 80 kV.
Clinicopathologic features such as sex, age, tumor size, tumor site, tumor gross type, histological type, histological grade, tumor necrosis, pT stage, number of nodes retrieved, number of nodes metastasis, lymph node metastasis ratio, pN stage, M stage and peritoneal metastasis were compared between patients with different serosal appearance. Univariate and multivariate analysis were applied for all 213 patients that were followed up completely, to identify the significant factors correlated with survival.
Follow-up of survival for the entire study population was obtained in household registration system at the Public Security Bureau, so the follow-up rate was 100%. Overall survival was defined as the time from surgery to the last follow-up (October 18, 2014) or patient death. Mean and median follow-up periods were 36.7 months and 40.7 months (range: 0.5-51.1 months), respectively.
This study was approved by the Research Ethics Committee of Liaoning Cancer Hospital & Institute (Shenyang, China). Written informed consent was obtained from all patients.
Statistical analysis
All the statistical analyses and graphics were performed with the SPSS for windows 19.0 statistical package (SPSS Inc., Chicago, IL). The mean of continuous data was expressed as x̅ ± s and categorical data was expressed as rate or percentage. In univariate analysis, two-tailed chi-square test for categorical variables and two-tailed t-test for continuous variables (Kruskal-Wallis test for heterogeneity of variance) were used for statistical comparisons. Overall survival rates were determined using the Kaplan-Meier estimator, an event being defined as death from a cancer-related cause. The log-rank test was used to identify differences between the survival curves of different patients groups. Multivariate Cox regression was used to identify independent factors correlated with prognosis. For all analyses, only P values < 0.05 were considered significant.
Results
Determination of macroscopic serosal classification
Referring to the articles about macroscopic serosal classification of gastric cancer written by Chen [1] and Sun [2], and summarizing the characteristics of hundreds cases of colorectal cancer serosal change, we preliminary divided macroscopic serosal appearance into four types: normal type, reactive type, nodular type and colloid type. Normal type: the serosal surface, which color and shape were completely consistent with normal serosal tissues, was smooth, soft touch and pressed without depression (Figure 1A). Reactive type: the serosal surface was red or white changes, with obscure boundary, touched smoothly or slightly rough (Figure 1B). Nodular type: there were some protruding nodules scattered or accumulated together on the surface of serosa, with hard and rough texture (Figure 1C). There were also some serosa sunken changes, due to serosa contracture. Colloid type: obvious colloid change was viewed on the serosal surface (area > 50% lesion), with or without tumor nodules or necrosis (Figure 1D).
Figure 1.
Macroscopic view of serosal classification. A. Normal type: normal serosal surface; B. Reactive type: red serosal changes, with obscure boundary; C. Nodular type: obvious tumor nodules protruding the surface of serosa; D. Colloid type: colloid changes predominantly, accompanied with massive small peritumoral nodules.
Observation under LM and TEM
The microscopic serosal structure and characteristic components of different serosal types were observed using LM and TEM. What we found with LM was as follows. Normal type: observed in longitudinal sections, the morphology and structure of the serosal surface were no change and tumor cells didn’t penetrate the serosa. The outermost layer of serosa was monolayer mesothelial cells, between it and muscularis propria was full of collagen fibers. In cross sections, only mesothelial cells, fibroblasts and collagen fibers could be observed. Reactive type: there were plenty of inflammatory cells (mainly lymphocytes) accumulated among collagen fibers in serosa and tumor cells didn’t penetrate the serosa or penetration was invisible. Nodular type: the whole serosal layer was permeated with glandular cancer nests, with or without tumor penetration of serosal surface. Colloid type: normal serosal structure outside of muscularis propria disappeared, replaced by mucin pool in extracellular space, or irregular close signet ring cells, or some necrosis tissues and cell fragments. Figure 2 shows some representative sections observed with LM.
Figure 2.
Microscopic serosal appearances and structure observed with LM (representative sections, not for all). A. Normal type viewed in longitudinal section, tumor nests invaded into serosa but the surface of serosa was unchanged (H.E. ×100); B. Normal type viewed in cross section of serosa, the main structure of serosa included mesothelial cells and fibroblasts, accompanied with numerous collagenous fibers (H.E. ×400); C. Reactive type viewed in longitudinal section, a tumor nest invaded muscularis propria and a large number of inflammatory cells infiltrated subserosa (H.E. ×100); D. Reactive type viewed in cross section of serosa, a large number of inflammatory cells (lymphocytes) were among the collagenous fibers (H.E. ×400); E. Nodular type viewed in longitudinal section, approximately a dozen nests infiltrated serosa but not penetrated (H.E. ×100); F. Nodular type viewed in cross section of serosa, the serosal surface was covered with glandular tumor nests (H.E. ×400); G. Colloid type viewed in longitudinal section, the structure of serosa outside of muscularis propria was not clear and full of signet ring cells (H.E. ×200); H. Colloid type viewed in cross section of serosa, the serosal surface was covered with signet ring cells (H.E. ×400).
Characteristics of different serosal types observed with TEM are detailed described in Figure 3. Consistent with the results observed with LM, the characteristic cells of normal type were mesothelial cells, with the diameter of 5~10 μm of oval shape. On cell surface, numerous microvilli were noted. In reactive type, the dense small lymphocytes (diameter < 5 μm) between mesothelial cells and collagen fibers played the crucial structural role differing from other three types. Huge special-shaped tumor cells consisted of the main structure of nodular type. And in colloid type, large amount of intracellular or extracellular mucinous granules were characteristic components which widely existed in signet ring cell cancer or mucinous adenocarcinoma.
Figure 3.
Microscopic serosal structure and characteristic cells observed with TEM. A. Normal type: typical mesothelial cell was observed in serosa, with numerous microvilli in its surface. Nucleus was in center with a prominent nucleolus and stained nuclear membrane. Abundant cytoplasm contained a wealth of endoplasmic reticulum and mitochondria. Collagen fibers existed outside of cell. B. Reactive type: lymphocytes were visible among mesothelial cells and fibroblasts in serosa, containing a large nucleus and less cytoplasm. Nucleus accounted for most of the cell, accompanied with heterochromatin that obviously accumulated in perinuclear. C. Nodular type: three intact tumor cells were visible, with large nucleus, less cytoplasm and stained chromatin. D. Colloid type: a signet ring cell penetrating muscularis propria reached serosa. It contained intracytoplasmic mucinous granule, pushing its nucleus to one side.
Clinicopathologic characteristics
Clinicopathologic characteristics of patients with different serosal types are listed in Table 1. As shown, there were significant differences in tumor size, tumor gross type, histological type, histological grade, tumor necrosis, depth of invasion (T stage), number of nodes metastasis, lymph node metastasis ratio, pN stage, M stage and peritoneal metastasis between patients with different types of serosa.
Table 1.
Comparison of clinicopathologic features between patients with different types of serosa
| Normal type | Reactive type | Nodular type | Colloid type | Statistics | P | |
|---|---|---|---|---|---|---|
| Sex | ||||||
| Male | 27 | 42 | 35 | 25 | χ2 = 3.068 | 0.381 |
| Female | 15 | 20 | 30 | 19 | ||
| Age (yrs) | 57.81±10.77 | 62.40±11.26 | 60.12±11.02 | 59.73±12.49 | F = 1.350 | 0.259 |
| Tumor size (cm) | 3.64±1.41 | 4.65±1.66 | 5.05±1.69 | 6.02±2.72 | F = 11.787 | 0.000 |
| Tumor site | ||||||
| Colon | 19 | 21 | 31 | 26 | χ2 = 6.785 | 0.079 |
| Rectum | 23 | 41 | 34 | 18 | ||
| Tumor gross type | ||||||
| Protruded | 26 | 8 | 15 | 3 | χ2 = 76.599 | 0.000 |
| Ulceration | 16 | 54 | 30 | 30 | ||
| Infiltrative | 0 | 0 | 20 | 11 | ||
| Histological type | ||||||
| Non-MUC | 36 | 55 | 53 | 9 | χ2 = 73.156 | 0.000 |
| MUC | 6 | 7 | 12 | 35 | ||
| Histological grade | ||||||
| Well | 24 | 10 | 10 | 2 | χ2 = 71.556 | 0.000 |
| Moderately | 16 | 52 | 49 | 27 | ||
| Poorly | 2 | 0 | 6 | 15 | ||
| Tumor necrosis | ||||||
| Yes | 6 | 43 | 38 | 18 | χ2 = 33.999 | 0.000 |
| No | 36 | 19 | 27 | 26 | ||
| UICC T stage | ||||||
| T1/T2 | 26 | 5 | 0 | 0 | χ2 = 116.158 | 0.000 |
| T3 | 16 | 47 | 38 | 22 | ||
| T4 | 0 | 10 | 27 | 22 | ||
| Number of nodes retrieved | 20.00±17.10 | 19.58±8.56 | 22.26±8.56 | 24.64±10.61 | F = 2.127 | 0.098 |
| Number of nodes metastasis | 1.62±4.10 | 2.16±4.08 | 5.86±6.94 | 9.18±10.12 | χ2 = 42.420* | 0.000 |
| LNM ratio (%) | 8.15±19.06 | 12.05±21.05 | 24.74±27.91 | 37.31±33.90 | χ2 = 38.163* | 0.000 |
| UICC N stage | ||||||
| N0 | 32 | 35 | 23 | 8 | χ2 = 47.222 | 0.000 |
| N1 | 6 | 18 | 11 | 12 | ||
| N2 | 4 | 9 | 31 | 24 | ||
| UICC M stage | ||||||
| M0 | 42 | 53 | 52 | 31 | χ2 = 14.537 | 0.002 |
| M1 | 0 | 9 | 13 | 13 | ||
| Peritoneal metastasis | ||||||
| Yes | 0 | 0 | 3 | 10 | χ2 = 28.229 | 0.000 |
| No | 42 | 62 | 62 | 34 |
Kruskal-Wallis test of mean ± SD.
MUC: Mucinous adenocarcinoma; LNM: Lymph node metastasis.
Concerning the tumor site between four types of MSC, there were no significant differences. But comparison of tumor site between non-colloid type and colloid type, we could revealed that colorectal cancer with colloid type was more likely to occur in the proximal intestine (P = 0.043) (Table 2).
Table 2.
Comparison of tumor site between non-colloid type and colloid type
| Tumor Site | Serosal type | OR (95% CI) | χ2 | P value | |
|---|---|---|---|---|---|
|
| |||||
| Non-Colloid Type | Colloid Type | ||||
| Colon | 71 | 26 | 0.502 (0.256-0.984) | 4.106 | 0.043 |
| Rectum | 98 | 18 | |||
Considering the morphology, structure, and clinicopathologic characteristics between MSC, it was more easily to summarize that the malignant degree of colorectal cancer with nodular and colloid types was higher than other two types.
So, we singled out these two types of cancers for further analysis. Comparing the depth of invasion and lymph node metastasis (LNM) between two types, no significant differences were noted (P > 0.05), but the p value of LNM was very close to significance (P = 0.051) (Table 3). The ratio of peritoneal metastasis between two types was different significantly (P = 0.003) (Table 4), and we just compared the tumors that penetrated the surface of serosa (pT4). Because from the previous studies, tumor cells penetrating the serosa had the probability to cause peritoneal metastasis and seldom exfoliation cancer cells passed through other access such as the surface of lymph node [9].
Table 3.
Comparison of UICC T stage and LNM between nodular type and colloid type
| Serosal type | OR (95% CI) | χ2 | P value | ||
|---|---|---|---|---|---|
|
| |||||
| Nodular Type | Colloid Type | ||||
| UICC T stage | |||||
| T3 | 38 | 22 | 1.407 (0.652-3.039) | 0.759 | 0.384 |
| T4 | 27 | 22 | |||
| LNM | |||||
| No | 23 | 8 | 2.464 (0.983-6.180) | 0.384 | 0.051 |
| Yes | 42 | 36 | |||
LNM: Lymph node metastasis.
Table 4.
Tumor penetrating serosal surface (T4) and peritoneal metastasis
| Normal type | Reactive type | Nodular type | Colloid type | χ2 | P | |
|---|---|---|---|---|---|---|
| Number of T4 | 0 | 10 | 27 | 22 | —— | —— |
| Number of PM | 0 | 0 | 3 | 10 | —— | —— |
| Ratio of PM (%) | 0 | 0 | 11.11 | 45.45 | 11.726 | 0.003 |
PM: Peritoneal metastasis.
Prognosis of patients
The 1- and 3-year survival rates of the entire patients followed up were 92.0% and 77.3%, respectively. The overall average survival time was estimated to be 43.3 months, while the median survival time cannot be estimated. Univariate analysis of prognostic factors identified tumor gross type, histological type, histological grade, pT stage, pN stage, M stage, TNM stage and MSC were significantly correlated with survival (Table 5). Putting all above significant variables selected by univariate analysis into Cox regression model, and after interaction of T, N and M stage with TNM stage, we drew a conclusion by multivariate analysis that TNM stage was the unique independent prognostic factor (Table 6).
Table 5.
Univariate analysis of prognostic factors for 213 patients
| Factors | Number | Survival rate (%) | χ2 | P | |
|---|---|---|---|---|---|
|
| |||||
| 1-year | 3-year | ||||
| Sex | |||||
| Male | 129 | 94.6 | 80.6 | 1.741 | 0.187 |
| Female | 84 | 88.1 | 72.6 | ||
| Age (year) | |||||
| ≤ 60 | 113 | 93.8 | 77.9 | 0.143 | 0.705 |
| > 60 | 100 | 90.0 | 76.9 | ||
| Tumor size (cm) | |||||
| ≤ 5 | 146 | 93.2 | 76.7 | 0.143 | 0.705 |
| > 5 | 67 | 89.4 | 75.5 | ||
| Tumor site | |||||
| Colon | 97 | 86.6 | 73.2 | 2.357 | 0.125 |
| Rectum | 116 | 96.6 | 81.0 | ||
| Tumor gross type | |||||
| Protruded | 52 | 98.1 | 96.2 | 28.500 | 0.000 |
| Ulceration | 129 | 93.0 | 77.3 | ||
| Infiltrative | 32 | 78.1 | 46.1 | ||
| Histological type | |||||
| Non-MUC | 160 | 94.4 | 80.6 | 4.772 | 0.029 |
| MUC | 53 | 84.9 | 67.8 | ||
| Histological grade | |||||
| Well | 46 | 95.7 | 91.3 | 7.025 | 0.030 |
| Moderately | 144 | 93.1 | 75.7 | ||
| Poorly | 23 | 78.3 | 60.3 | ||
| Tumor necrosis | |||||
| Yes | 78 | 90.4 | 78.5 | 0.393 | 0.531 |
| No | 135 | 94.9 | 75.6 | ||
| UICC T stage | |||||
| T1/T2 | 31 | 96.8 | 93.5 | 7.208 | 0.027 |
| T3 | 123 | 91.9 | 78.0 | ||
| T4 | 59 | 89.8 | 67.6 | ||
| UICC N stage | |||||
| N0 | 98 | 96.9 | 92.9 | 32.090 | 0.000 |
| N1 | 47 | 95.7 | 76.6 | ||
| N2 | 68 | 82.4 | 55.5 | ||
| UICC M stage | |||||
| M0 | 178 | 94.4 | 84.3 | 42.273 | 0.000 |
| M1 | 35 | 80.0 | 42.5 | ||
| TNM stage | |||||
| I | 26 | 96.2 | 96.2 | 49.463 | 0.000 |
| II | 69 | 97.1 | 91.3 | ||
| III | 83 | 91.6 | 74.7 | ||
| IV | 35 | 80.0 | 42.5 | ||
| MSC | |||||
| Normal type | 42 | 95.2 | 88.1 | 10.968 | 0.012 |
| Reactive type | 62 | 95.2 | 82.3 | ||
| Nodular type | 65 | 92.2 | 76.6 | ||
| Colloid type | 44 | 86.4 | 65.1 | ||
MSC: Macroscopic serosal classification.
Table 6.
Multivariate analysis of prognostic factors for 213 patients
| Variable | B | SE | Wald | df | Sig. | Exp (B) | 95.0% CI for Exp (B) | |
|---|---|---|---|---|---|---|---|---|
|
| ||||||||
| Lower | Upper | |||||||
| Tumor gross type | 0.357 | 0.271 | 1.739 | 1 | 0.187 | 1.429 | 0.841 | 2.429 |
| Histological type | 0.510 | 0.294 | 3.015 | 1 | 0.082 | 1.665 | 0.936 | 2.960 |
| Histological grade | 0.168 | 0.289 | 0.338 | 1 | 0.561 | 1.183 | 0.671 | 2.086 |
| TNM*T interaction | -0.097 | 0.086 | 1.281 | 1 | 0.258 | 0.907 | 0.767 | 1.074 |
| TNM*N interaction | 0.135 | 0.074 | 3.350 | 1 | 0.067 | 1.144 | 0.990 | 1.322 |
| TNM*M interaction | 0.086 | 0.142 | 0.370 | 1 | 0.543 | 1.090 | 0.826 | 1.439 |
| TNM stage | 0.689 | 0.275 | 6.289 | 1 | 0.012 | 1.991 | 1.162 | 3.411 |
| MSC | -0.050 | 0.203 | 0.061 | 1 | 0.805 | 0.951 | 0.639 | 1.416 |
Comparing survival rates according to MSC, survival curves indicated that colloid type was the worst, followed by nodular type, reactive type and normal type (Figure 4). Survival rate of different TNM stage was also compared by log-rank test described by Figure 5.
Figure 4.
Survival curves of patients after operation according to macroscopic serosal classification (n = 213, P = 0.012).
Figure 5.
Survival curves of patients after operation according to TNM stage (n=213, P=0.000).
Discussion
Typically, macroscopic classification of colorectal cancer has been divided into three types according to its gross appearance, including protruded, ulceration and infiltrative, which was concluded from the morphologic observation of mucosal surface. This macroscopic classification has been indicated to have some certain correlation with the patient prognosis that prognosis of patients with ulceration type is in an intermediate position between the best of protruded type and the worst of infiltrative type [10,11]. Morphological studies seem to be superficial, but in fact it is closely related to the biological behavior, differentiation, growth pattern, and many other features of tumor cells [12]. For surgeons, the first sight when entering abdominal cavity at the beginning of surgery is serosal appearance instead of mucosa, then to conduct the appropriate surgery and other clinical decision making partly according to serosal changes, especially more significant to those patients without distant metastasis. So, serosal morphological characteristics of colorectal cancer are particularly important. The present study, which is firstly reported worldwide, merely focused on the macroscopic serosal classification (MSC) and investigated clinicopathologic characteristics and prognosis of different MSC.
Through summarizing macroscopic serosal appearances of colorectal cancer and observing microscopic characteristics with LM and TEM, MSC was preliminarily classified as normal type, reactive type, nodular type and colloid type. From the results, there was no statistical significance in tumor site between patients with different types. But compared colloid type with other three types, it was more frequently found in the proximal colon, mainly due to the serosal histological components of colloid type which composed by mucinous adenocarcinoma, signet ring cell cancer and a small part of necrosis. Previous studies comparing clinicopathologic characteristics of mucinous carcinomas (MCs) with non-mucinous carcinomas (NMCs) of the colon and rectum, indicated that MCs occurred in the right colon significantly more frequently than did non-mucinous carcinomas (NMCs) [13,14]. Similar study by Hugen et al. reported that patients with signet ring cell cancer presented more frequently in the proximal colon (57.7 vs. 32.0%) than patients with adenocarcinoma cancer [15]. However, tumor location of signet ring cell cancer remains controversial. Song et al. revealed signet ring cell tumors were more commonly found in the rectum than mucinous and non-mucinous adenocarcinoma [16].
Regarding clinicopathologic characteristics of MSC, it is very easy to discriminate that nodular type and colloid type are more malignant than normal type and reactive type, according to the aspects in the depth of invasion, lymph node and distant metastasis. Nodular type and colloid type invade deeper than other two types, but without significant difference within themselves. Lymph node metastasis compared between nodular type and colloid type, there is a worse trend emerged for colloid type, but it has not result in a statistically significant difference. Previous studies revealed that mucinous and signet ring cell tumors presented in a later stage more often than non-mucinous adenocarcinoma, including lymph node involvement [14,16]. Maybe the differences of LNM between nodular type and colloid type will be statistical significance with samples increasing in the current study. Distant metastasis between nodular type and colloid type has no significant differences, excluding the peritoneal metastasis which has a significant higher proportion in colloid type than that in nodular type. Free cancer cells penetrating serosa may be the crucial step and main reason for the development of peritoneal metastasis. In accordance with it by histological examination, the depth of invasion should equal or exceed pT4 besides changes in macroscopic serosal morphology. So we further compared the patients with pT4 of different serosal types, peritoneal metastasis rate of colloid type was still significantly higher than that of other types. In other words, free cancer cells of colloid type have stronger ability to exfoliate from the lesions where tumors penetrate the serosa and to result in metastatic carcinogenesis in peritoneal cavity, than those of nodular type and reactive type. Analyzing from the view of histological components, colloid type is composed by mucinous carcinoma and signet ring cell cancer, which more frequently had peritoneal metastases compared with classic adenocarcinoma cancer [17]. Therefore, one of meaningful contents about application of MSC is to make intraoperative decision regarding proper treatment strategy. For patients with colloid type of colorectal cancer, besides conventional surgery including lymph nodes dissection, it is necessary to use intraperitoneal chemotherapy or other methods for preventing peritoneal metastasis.
Our further analysis confirmed that MSC, one of predictors for prognosis of patients with colorectal cancer, will become a positive complement of traditional prognostic factors. Clinicopathologic characteristics and histological components of different serosal types decide different prognosis, such as signet ring cell and mucinous components in colloid type differ from classical adenocarcinomas in nodular type [18,19]. But in the present study, MSC is still not an independent predictor according to statistical results, maybe due to the insufficient sample. While TNM stage, the unique independent prognostic factor for patients with colorectal cancer, is always in an irreplaceable position for prognostic evaluation.
Of course, MSC has some disadvantages. For example, a small part of patients with pT3 or pT4 can’t be accurate judgment quickly by surgeons when entering abdominal cavity through serosal changes in both nodular type and colloid type, even in a few reactive type. Then the method of serosal cytologic smears could be applied to confirm the depth of invasion as an effective supplement of MSC [20]. Serosal cytologic smears could estimate whether tumor cells easily exfoliate from the serosal surface into peritoneal cavity. Panarelli et al. concluded that cytologic smears improve detection of peritoneal penetration among pT3 tumors compared with histology alone. Tumors close (≤ 1 mm) to a fibroinflammatory tissue reaction on the serosa are likely associated with peritoneal involvement by cancer. Peritumoral abscesses that communicate with the serosa and hemorrhage or fibrin on the serosa also predict cancer involvement of the peritoneum [21]. According to macroscopic appearance of abscesses, which is often associated with necrosis, we put it into colloid type. But without being classified about serosal hemorrhage and fibrin, further studies need to be carried out to make MSC more refined and detailed.
The present study provided the morphological difference of serosal appearance from macroscopic observation to light microscopy, then indicated the components and structural differences from light microscopy to transmission electron microscopy, eventually summarized the clinical significance of MSC through retrospective analysis of clinicopathologic materials. Looking forward for further study, in addition to making definition and classification of MSC detailed, we would analyze the molecular changes and compositions in different serosal type from histological level, in order to provide partial supports for molecular serosal classification of colorectal cancer.
In conclusion, we found MSC of colorectal cancer, which was preliminarily defined and classified as normal type, reactive type, nodular type and colloid type. Patients with different serosal types have different clinicopathologic features and prognosis. The information about serosal type and invasion, which can be known at the first time by surgeons in the beginning of the surgery, aids them in making proper intraoperative decisions including surgical performance and other anti-cancer treatments.
Disclosure of conflict of interest
None.
References
- 1.Chen JQ QC, Shan JX, Wang SB, Zhang WF, Wang MX, Zhang YC, Chen X, Liu RD. The serosal types of gastric cancer and its biological significance. J Chin Med Univ (Engl) 1986;1:38–43. [Google Scholar]
- 2.Sun Z, Xu YY, Wang ZN, Zhu Z, Zhang H, Huang BJ, Xu Y, Chen JQ, Xu HM. Macroscopic serosal classification predicts peritoneal recurrence for patients with gastric cancer underwent potentially curative surgery. Ann Surg Oncol. 2010;18:1068–1080. doi: 10.1245/s10434-010-1449-6. [DOI] [PubMed] [Google Scholar]
- 3.Zhu HT, Zhao YL, Wu YF. [Relationship between serosal invasion types and lymph node metastasis after total gastrectomy in gastric cancer and its significance in selection of rational dissection] . Zhonghua Zhong Liu Za Zhi. 2009;31:474–477. [PubMed] [Google Scholar]
- 4.Jiang CG, Wang ZN, Sun Z, Liu FN, Yu M, Xu HM. Clinicopathologic characteristics and prognosis of gastric cancer invading the subserosa. J Surg Oncol. 2010;102:737–741. doi: 10.1002/jso.21678. [DOI] [PubMed] [Google Scholar]
- 5.Akkoca AN, Yanik S, Ozdemir ZT, Cihan FG, Sayar S, Cincin TG, Cam A, Ozer C. TNM and Modified Dukes staging along with the demographic characteristics of patients with colorectal carcinoma. Int J Clin Exp Med. 2014;7:2828–2835. [PMC free article] [PubMed] [Google Scholar]
- 6.Benson AB 3rd, Venook AP, Bekaii-Saab T, Chan E, Chen YJ, Cooper HS, Engstrom PF, Enzinger PC, Fenton MJ, Fuchs CS, Grem JL, Hunt S, Kamel A, Leong LA, Lin E, Messersmith W, Mulcahy MF, Murphy JD, Nurkin S, Rohren E, Ryan DP, Saltz L, Sharma S, Shibata D, Skibber JM, Sofocleous CT, Stoffel EM, Stotsky-Himelfarb E, Willett CG, Gregory KM, Freedman-Cass DA. Colon cancer, version 3.2014. J Natl Compr Canc Netw. 2014;12:1028–1059. doi: 10.6004/jnccn.2014.0099. [DOI] [PubMed] [Google Scholar]
- 7.Jiang CG, Wang ZN, Sun Z, Liu FN, Yu M, Xu HM. Clinicopathologic characteristics and prognosis of signet ring cell carcinoma of the stomach: results from a Chinese mono-institutional study. J Surg Oncol. 2011;103:700–703. doi: 10.1002/jso.21878. [DOI] [PubMed] [Google Scholar]
- 8.Elias JM. Chemistry of H and E staining. Am J Med Technol. 1974;40:513–514. [PubMed] [Google Scholar]
- 9.Roviello F, Marrelli D, de Manzoni G, Morgagni P, Di Leo A, Saragoni L, De Stefano A. Prospective study of peritoneal recurrence after curative surgery for gastric cancer. Br J Surg. 2003;90:1113–1119. doi: 10.1002/bjs.4164. [DOI] [PubMed] [Google Scholar]
- 10.Li M, Li JY, Zhao AL, Gu J. Colorectal cancer or colon and rectal cancer? Clinicopathological comparison between colonic and rectal carcinomas. Oncology. 2007;73:52–57. doi: 10.1159/000120628. [DOI] [PubMed] [Google Scholar]
- 11.Vasile L, Olaru A, Munteanu M, Plesea IE, Surlin V, Tudorascu C. Prognosis of colorectal cancer: clinical, pathological and therapeutic correlation. Rom J Morphol Embryol. 2012;53:383–391. [PubMed] [Google Scholar]
- 12.Cheah PY, Kirzin S, Marisa L, Guimbaud R, De Reynies A, Legrain M, Laurent-Puig P, Cordelier P, Pradère B, Bonnet D, Meggetto F, Portier G, Brousset P, Selves J. Sporadic Early-Onset Colorectal Cancer Is a Specific Sub-Type of Cancer: A Morphological, Molecular and Genetics Study. PLoS One. 2014;9:e103159. doi: 10.1371/journal.pone.0103159. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Leopoldo S, Lorena B, Cinzia A, Gabriella DC, Angela Luciana B, Renato C, Antonio M, Carlo S, Cristina P, Stefano C, Maurizio T, Luigi R, Cesare B. Two Subtypes of Mucinous Adenocarcinoma of The Colorectum: Clinicopathological and Genetic Features. Ann Surg Oncol. 2008;15:1429–1439. doi: 10.1245/s10434-007-9757-1. [DOI] [PubMed] [Google Scholar]
- 14.Nozoe T, Anai H, Nasu S, Sugimachi K. Clinicopathological characteristics of mucinous carcinoma of the colon and rectum. J Surg Oncol. 2000;75:103–107. doi: 10.1002/1096-9098(200010)75:2<103::aid-jso6>3.0.co;2-c. [DOI] [PubMed] [Google Scholar]
- 15.Hugen N, Verhoeven RH, Lemmens VE, van Aart CJ, Elferink MA, Radema SA, Nagtegaal ID, de Wilt JH. Colorectal signet-ring cell carcinoma: benefit from adjuvant chemotherapy but a poor prognostic factor. Int J Cancer. 2015;136:333–339. doi: 10.1002/ijc.28981. [DOI] [PubMed] [Google Scholar]
- 16.Song W, Wu SJ, He YL, Cai SR, Zhang CH, Zhang XH, Zhan WH. Clinicopathologic features and survival of patients with colorectal mucinous, signet-ring cell or non-mucinous adenocarcinoma: experience at an institution in southern China. Chin Med J (Engl) 2009;122:1486–1491. [PubMed] [Google Scholar]
- 17.Hugen N, van de Velde CJ, de Wilt JH, Nagtegaal ID. Metastatic pattern in colorectal cancer is strongly influenced by histological subtype. Ann Oncol. 2014;25:651–657. doi: 10.1093/annonc/mdt591. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Nitsche U, Zimmermann A, Späth C, Müller T, Maak M, Schuster T, Slotta-Huspenina J, Käser SA, Michalski CW, Janssen KP, Friess H, Rosenberg R, Bader FG. Mucinous and Signet-Ring Cell Colorectal Cancers Differ from Classical Adenocarcinomas in Tumor Biology and Prognosis. Ann Surg. 2013;258:775–783. doi: 10.1097/SLA.0b013e3182a69f7e. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Inamura K, Yamauchi M, Nishihara R, Kim SA, Mima K, Sukawa Y, Li T, Yasunari M, Zhang X, Wu K, Meyerhardt JA, Fuchs CS, Harris CC, Qian ZR, Ogino S. Prognostic Significance and Molecular Features of Signet-Ring Cell and Mucinous Components in Colorectal Carcinoma. Ann Surg Oncol. 2015;22:1226–1235. doi: 10.1245/s10434-014-4159-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Zeng Z, Cohen AM, Hajdu S, Sternberg SS, Sigurdson ER, Enker W. Serosal cytologic study to determine free mesothelial penetration of intraperitoneal colon cancer. Cancer. 1992;70:737–740. doi: 10.1002/1097-0142(19920815)70:4<737::aid-cncr2820700404>3.0.co;2-s. [DOI] [PubMed] [Google Scholar]
- 21.Panarelli NC, Schreiner AM, Brandt SM, Shepherd NA, Yantiss RK. Histologic features and cytologic techniques that aid pathologic stage assessment of colonic adenocarcinoma. Am J Surg Pathol. 2013;37:1252–1258. doi: 10.1097/PAS.0b013e3182960e7c. [DOI] [PubMed] [Google Scholar]