Relationship between the presence of liver metastases with histological grading , depth of invasion and nodal involvement in sporadic adenocarcinoma of the large intestine

Introduction: Large intestine adenocarcinoma (LIA) is the most common cancer of the gastrointestinal tract, and corresponds to the fifth most common malignancy in Brazil. The main prognostic factors related to LIA are depth of tumor invasion and perivisceral lymph nodes status. Objective: To estimate the relationship between pathological findings and the presence of liver metastases (LM) in LIA cases. Method: We evaluated 51 cases of LIA, previously submitted to surgical resection, in order to determine the following variables: topography, tumor size, macroscopic appearance, degree of differentiation, depth of invasion, nodal status, and presence of LM. Results: The average age was 64.8 years, with predominance of men (n = 26/51.0%) and lesions in the sigmoid colon (n = 18/35.3%). The main general characteristics of the sample were ulcerative-vegetative lesions (n = 20/39.2%), no annular tumors (n = 3/64.7%), moderately differentiated tumor (n = 44/86.3%), absence of mucinous areas (n = 40/78.4%), and mesocolon invasion (n = 29/56.9%). LM were found in 14 cases (27.5%), and is associated with presence of nodal metastases (p = 0.005). Tumor size (p = 0.72), macroscopic appearance (p = 0.362), histological grade (p = 0.147), and depth of invasion (p = 0.195) showed no association with LM presence. Conclusion: LIA has a wide anatomical and pathological heterogeneity. In this study, the presence of LM associated with LIA was related to perivisceral lymph nodes status, with no relation to tumor size, degree of differentiation, and depth of invasion, which suggests that identifying neoplastic angiolymphatic invasion is a possible predictor of liver involvement.


intRoDuCtion
The primary adenocarcinoma of the large intestine is among the ten most common types of cancer worldwide, and it is also the most common form of malignancy of the gastrointestinal tract.It corresponds to the third leading cause of death in women and the fifth among men in Brazil, although these rates are higher in North America and Europe (1)(2)(3) .A progressively higher incidence over the last 15 years have been observed; however, the distribution of colorectal cancer in relation to gender and age has remained constant (1)(2)(3)(4) .The risk for developing this cancer increases after the age of 40 for both men and women, and this risk doubles with each subsequent decade.Family history is a risk factor for colorectal carcinoma, especially in patients younger than 50 years, the prevalence of adenomas of the colon in relatives of patients with colorectal carcinoma is 39% (1)(2)(3)(4) .
Colorectal adenocarcinoma may present polypoid, vegetative, ulcerative, ulcerative-vegetative, and infiltrative macroscopic forms.Histologically, it can be classified into grades based on tubular architectural pattern of neoplastic cells (2,3,5) .Although poorly differentiated and mucinous neoplasms are associated with lower disease-free survival rates, the two most important prognostic factors are the depth of invasion and the presence of metastases in regional lymph nodes.Besides these latter, the metastases may involve the lungs and bones, but as a consequence of the venous drainage of the colon, the liver is the most common location for metastatic lesions (2,3,5) .About 25% of patients with colorectal cancer may present liver metastasis at the time of primary tumor diagnosis, and 50% of patients may show liver metastases during the course of the disease (2,3,5) .
In this study, the authors assessed the relationship between liver metastases and anatomopathological findings of colorectal adenocarcinoma, as topography, tumor size, macroscopic appearance, degree of differentiation, presence of mucinous areas, depth of invasion, and in periviscerais lymph nodes metastasis.

MEthoD
This is a cross-sectional retrospective study involving patients with large intestine adenocarcinoma (LIA), who had undergone previous surgical resection and staging.Initially, 51 different LIA cases were selected for the study, previously analyzed in the Pathology Laboratory of the Grupo Hospitalar Conceição (GHC) in Porto Alegre (RS).Data collection was carried out between March 2014 and October 2014, and was approved by the Ethics Research Committee (ERC) of GHC, project nº 14-001.We included in the sample only those cases histologically classified as primary large intestine adenocarcinoma according to World Health Organization criteria.All other histological types of primary benign or malignant neoplasm of the large intestine and cases of metastasis or secondary involvement of the organ were excluded.
The demographic characteristics (age and sex) were collected from the medical records.Each case was previously fixed in 10% formalin, stained with hematoxylin-eosin (HE), and the following anatomopathological data were reviewed and described: 1) anatomical position: cecum, ascending colon, sigmoid colon, transverse colon, and rectum; 2) tumor size: is in centimeters, in the longer axis of the lesion; 3) macroscopic appearance: infiltrative, ulcerative, ulcerative-infiltrative, ulcerative-vegetative and vegetative; 4) presence of annular or no annular lesion; 5) degree of differentiation: well-differentiated, moderatelydifferentiated and poorly-differentiate; 6) presence of mucous producing area; 7) depth of invasion: mucosa, submucosa, muscle, perivisceral tissue, serous and adjacent organ invasion; 8) surgical margins status: free or positive; 9) presence of lymph nodes metastases.
The evaluated cases were also grouped by TNM staging system; T: tumor, N: lymph node, M: metastasis.Data were recorded and analyzed in a Microsoft Excel 2007 spreadsheet.Quantitative variables were expressed as mean and standard deviation (DP) or median and interquartile range.Categorical variables were described by absolute and relative frequencies.To compare means between groups we used the Student-t test; in case of asymmetrical distribution, we used the Mann-Whitney test; Pearson chi-squared tests or Fisher's exact test were used for the difference between proportions; in case of significance, the adjusted residual test was applied.The significance level was 5% (p ≤ 0.05), and analyzes were performed in SPSS software version 21.0.

RESuLtS
The group of patients analyzed had a mean age of 64.8 years, predominantly male.The prevalence of liver metastasis was 27.5% (confidence interval [CI] 95%: 15.2%-39.8%).Tables 1 and 2 show the results obtained by the research.
The number of lymph node metastasis (p = 0.005) and group N of TNM system (p = 0.015) were associated with the presence of liver metastases.The pathological variables, such as anatomical position (p = 0.242), tumor size (p = 0.722), macroscopic appearance (p = 0.362), degree of differentiation (p = 0.147), presence of mucous producing area (p = 0.251), depth of invasion (p = 0.195), and positive resection margins (p = 0.478), had no significant association with the presence of liver metastasis.

DiSCuSSion
Colorectal carcinoma is among the ten main types cancers in Brazil, affecting both sexes in similar prevalence.The age profile of colorectal carcinoma has dramatic increase in incidence from 40 years for women and 50 years for men.It is estimated that, in 2014, in Brazil, 32,600 cases of colorectal carcinoma were found, 15,070 men and 17,530 women (6) .In this study, data of age group was equivalent to those from the literature.The study population consisted of an age average of 64.8 years, ranging 12.2 years, in 26 male (51%) and 25 female (49%).Oliveira et al. analyzed 129 cases of colorectal carcinoma and determined that the mean age of patients was 56.9 years, ranging 25-87 years, of which 51.2% were female patients (51.2%, p > 0.05) (7) .Torres Neto et al. evaluated 355 cases of colorectal cancer, which was more frequent in women, the age ranged from 5-98 years, mean age of 60.2 years (8) .
Colorectal adenocarcinoma may grade in well-differentiated, moderately-differentiated, poorly-differentiated and undifferentiated lesions.The infiltrating edge of adenocarcinomas may have rounded or spiculated tumor growth and, in its progression, the adenocarcinoma gradually invades the intestine layers.During this mural progression of the cancer, there may be angiolymphatic invasion in cancer and potential spread and tumor implantation in regional lymph nodes and distant organs (1,2,4,9,10) .In this study, the anatomopathological variables, such as anatomical position (p = 0.242), tumor size (p = 0.722), macroscopic appearance (p = 0.362), the presence or absence of annular lesion (p = 0.204), degree of differentiation (p = 0.147), presence of mucous producing area (p = 0.251), depth of invasion (p = 0.195), and presence of positive surgical margins (p = 0.478), were not associated with the presence of liver metastases.Madeira et al. analyzed 50 LIA cases and reported that there was no association between the presence of nodal metastases and age, sex, microscopic edge, presence of mucous producing areas, degree of differentiation, and tumor size (2) .Fujimoto et al. reported that the presence of liver metastases on 122 cases with clinical five years follow-up, was significantly associated with tumor size, degree of differentiation, depth of invasion, presence of neoplastic angiolymphatic invasion and presence of metastases in regional lymph (11) .Kitajima et al. investigated the depth of submucosa invasion in 865 cases of invasive colorectal carcinoma and concluded that nodal metastasis rates are associated with depth of submucosa lesion (12) .Inoue et al. compared the depth of submucosa lesion in 118 cases of colorectal carcinomas sessile and superficial type, which were resected endoscopically, and total depth was significantly higher in sessile type (13) .
The liver is the most common site of distant metastases of LIA, found in up to 25% patients at diagnosis (1-3, 5-7, 11) .The prevalence of liver metastases in our study was 27.5%.The average age of the group with liver metastasis was significantly higher than the group without metastasis.Fernandes et al. analyzed 749 LIA cases and determined that the main sites of metastases were liver (64.51%), peritoneum (19.35%), and lungs (9.67%).Among metastases associated with colorectal cancer, the liver was also the most affected organ (54.11%) (14) .Murad et al. reported that in a sample of 49 patients with liver metastases from colorectal adenocarcinoma, 31 cases (63.2%) had already had hepatic involvement in the initial surgical procedure or presence of metastasis within six months of initial diagnosis.In 18 cases (36.8%), the appearance of metastasis occurred six months after colorectal surgery (9) .Ambiru et al. determined that the survival of 168 patients with liver metastases associated with colorectal carcinoma was 42% in three years and 26% in five years.The main prognostic factors included nodal status of the initial resection of the primary tumor, surgical margin status, number of liver metastases, and adjuvant chemotherapy (15) .Arru et al., among the 297 cases of resected liver metastases associated with colorectal carcinoma, determined that the time period of disease-free survival was associated with the degree of differentiation, serum level of carcinoembryonic antigen, tumor size greater than 5 cm and time period of metastasis detection after initial surgical treatment (16) .Ohji et al. selected 120 cases of advanced colorectal carcinoma, including liver metastases in 60 cases and determined that the presence of nodal metastasis, and the immunohistochemical expression of neprilysin (CD10) and vascular endothelial growth factor (VEGF) are associated with the development of metastatic lesions (17) .Ueno et al. reported that the degree of differentiation was associated with groups T and N of the TNM system, the serum levels of carcinoembryonic antigen, the presence of liver, lung and peritoneal metastases, the number of liver metastases and survival time after hepatectomy (18) .Lin et al. determined that in cases of colorectal adenocarcinoma, the presence of tumor deposits in perivisceral tissues was associated with liver and nodal metastasis, perineural invasion and disease-free interval after resection of the primary tumor (19) .Adachi et al. reported that the presence of liver metastases in adenocarcinoma colorectal cancer was associated with primary tumors with size equal or greater than 6 cm, presence of serosal invasion, neoplastic angiolymphatic invasion, and metastasis to regional lymph nodes (20) .
In our study, the presence of regional lymph node metastasis (p = 0.005) was associated with the presence of liver metastasis.Kuhen et al. analyzed 132 patients undergoing surgical treatment
for colorectal cancer resection, and the presence of nodal metastases was associated with the degree of differentiation (21) .Fonseca et al. assessed 521 surgical specimens of patients operated for colorectal cancer and the average of lymph nodes affected by metastases corresponds to 2.57 ± 5.34 lymph nodes (22) , while Santo et al. found an average of 3.4 ± 3.3 metastatic lymph node from 51 specimens examined (42.8%/n = 119) (23) .Homma et al. analyzed 65 patients with colorectal adenocarcinoma invading the submucosa or the muscle propria, and concluded that the resection of the affected regional lymph nodes by metastases reduced the local recurrence rate of these carcinomas (24) .In another study Homma et al. determined that sex, histologic grade, neoplastic angiolymphatic invasion or nodal status were not significantly associated with tumor recurrence (25) .Ahmadi et al. analyzed 824 patients with colorectal carcinoma and found that the right colon tumors are more often associated with nodal metastases, and were also associated with age (26) .Toh et al. investigated 207 cases of adenocarcinoma colorectal and demonstrated that the radial extent of the tumor and submucosal invasion area can predict the regional nodal status (27) .

ConCLuSion
LIA has good anatomopathological heterogeneity, and the depth of invasion and lymph node status are considered important prognostic factors.In this study, the presence of liver metastases determined by LIA was associated with the status of perivisceral lymph nodes, with no relation to tumor size, degree of differentiation, and depth of invasion, suggesting that identifying neoplastic angiolymphatic invasion by microscopic examination is a potential predictor of liver involvement.