Sixty-eight pathologically-confirmed specimens were obtained from patients with stage IIIB (T3N1M0; AJCC, 2002) colon cancer between January 1999 and May 2002 at the Cancer Center of Sun Yat-Sen University, Guangzhou, China (Table 1). All the patients underwent radical resection and 5-FU-based adjuvant chemotherapy after operation for 6 mo. The patients were evaluated every 3 mo during the first year, every 6 mo in the second year, and once every year thereafter for a total of 5 years. If a recurrence or a metastasis occurred, 5-FU-based chemotherapy was given according to the national comprehensive cancer network (NCCN) guidelines. No patients received preoperative blood transfusion or non-steroidal anti-inflammatory drugs. Overall survival was defined as the time from surgery to death. Data analysis was done on the last known day when the patient was alive.

Formalin-fixed, paraffin-embedded tissue was cut into 4-μm thick sections. The size of each tissue section was about 1 cm × 1 cm. Then, the sections were dewaxed, rehydrated, and blocked with hydrogen peroxide. Antigens were retrieved in 10 mmol/L citrate buffer (pH 6.0) for 10 min and cooled to room temperature. After blocked with sheep serum, the sections were incubated overnight at 4°C with either rabbit polyclonal antibody against human calreticulin at a dilution of 1:2000 (Abcam, Cambridge, MA, USA) or mouse monoclonal antibody against human CD3 and CD45RO (Zymed, San Diego, CA, USA), both of which were diluted to 1:100. Subsequently, biotinylated secondary antibodies and streptavidin-biotinylated horseradish peroxidase complexes were used. The sections were developed with diaminobenzidine tetrahydrochloride (DAB) and counterstained with hematoxylin. Negative controls in which primary antibody was replaced with a phosphate buffered solution (PBS), were employed.

Infiltration of lymphocytes in the tumor was scored with Hussein’s method[38] and expression of calreticulin in colon cancer was interpreted via immunoreactivity using the 0-4 semi-quantitative system derived from Remmele and Stegner for both the intensity of staining and the percentage of positive cells (labeling frequency percentage)[39]. The cells were counted in at least 10 different fields for each section, and the size of each high-power field (× 400) was about 300 μm × 300 μm. The cells were counted in tumor stroma. The highest infiltration areas of lymphocytes were chosen. Necrotic areas were avoided. Two observers counted the cells at the same time and in the same field under a multiple-lens microscope. The results were expressed as mean ± SE. Cytoplasm staining was divided into no staining/background of negative controls (score = 0), weak staining above background (score = 1), moderate staining (score = 2), and intense staining (score = 3). The labeling frequency was scored as 0 (≤ 1%), 1 (1%-24%), 2 (25%-49%), 3 (50%-74%), and 4 (≥ 75%), respectively. The product index was obtained by multiplying the intensity and percentage and scored as (-), (+), (++), and (+++), which indicate the cross-indices of 0-2, 3-5, 6-8, and 9-12, respectively. (-) was defined as no or negative expression, and (+)-(+++) as positive expression. Each section was independently scored by two pathologists. If an inconsistency occurred, a third pathologist was consulted to achieve a consensus.

Correlation between calreticulin expression, or infiltration of lymphocytes and parameters of patients was analyzed by χ² test or Fisher’s exact test. Factors, including gender and age of the patients, pathologic grade, tumor site, infiltration of CD3+ cells and CD45RO+ cells, and calreticulin expression level in colon cancer, were assessed by univariate and multivariate analysis to determine their influence on the overall survival rate of patients. Kaplan-Meier curve and log-rank test were used to estimate the distribution of variables in relation to survival. Cox regression model was used to correlate the assigned variables with the overall survival rate. All statistical analyses were carried out using SPSS 13.0 software (SPSS Inc., Chicago, IL, USA). P < 0.05 was considered statistically significant.

CRT was stained in cytoplasm rather than in nuclei of cancer cells and normal epithelium. The expression level of CRT was lower in colon cancer than in its adjacent normal epithelium. Of the stained samples, 61.8% (42/68) were positive for CRT in cancer nest (Figure 1A and B, Table 2). CD3+ and CD45RO+ cells were observed in all samples from tumor stroma and its adjacent normal mucosa. Positively stained antigens were found on cell membrane (Figure 1C-F).

Open in New Tab   Full Size Figure   Download Figure

Figure 1 Expression level of CRT in cancer nest (A) and in its adjacent normal epithelium (B) [× 100 in A, × 200 in B, normal epithelium (red arrow), atypical hyperplasia (black arrow) and tumor tissue (blue arrow), respectively], high infiltration of CD3+ cells (C) and CD45RO+ cells (E) in colon cancer stroma (× 400), low infiltration of CD3+ cells (D) and CD45RO+ cells (F) in colon cancer stroma (× 400).

The cut-off value for infiltration of lymphocytes in colon cancer was 75%. Infiltration of 18 CD3+ cells and 26 CD45RO+ cells was observed in per high-power field, and was thus recorded as high and low infiltration. Log-rank test was used to analyze the relation between expression of CRT and infiltration of CD3+ and CD45RO+ cells in colon cancer, showing that positive expression of CRT (+-+++) was associated with high infiltration of CD45RO+ cells (P = 0.010, Table 2). No correlation was observed between expression of CRT and infiltration of CD3+cells or other parameters of the patients, such as age, gender, tumor site.

By the end of a 5-year follow-up period, 52 patients were alive with a 5-year survival rate of 76.5%. Kaplan-Meier survival analysis indicated that positive expression of CRT was associated with a higher 5-year survival rate. The 5-year survival rate of patients with positive and negative expression of CRT was 85.5% (36/42) and 61.5% (16/26), respectively (P = 0.022, Table 3). However, the survival curves for patients with positive and negative expression of CRT were crossed at 18 mo (Figure 2). In addition, high infiltration of CD3+ and CD45RO+ cells in colon cancer was associated with a higher 5-year survival rate (P = 0.000, Figure 3).

Open in New Tab   Full Size Figure   Download Figure

Figure 2 Relation between CRT expression and 5-year survival rate.

Open in New Tab   Full Size Figure   Download Figure

Figure 3 Correlation between high infiltration of CD3+ cells (A) and CD45RO+ cells (B) in colon cancer and 5-year survival rate of patients.

Cox regression model revealed that neither expression of CRT nor infiltration of CD3+ cells or CD45RO+ cells in colon cancer had an independent prognostic value (Table 4).

There is evidence that colorectal cancer is immunogenic. The mechanisms by which colorectal cancer cells induce an immunologic response remain unknown. A subset of damage-associated molecular patterns (DAMPs) has been recently found to be involved in the induction of immune response. Calreticulin is one of the most important DAMPs. Whether calreticulin is associated with the immunogenicity of colon cancer remains controversial.

Calreticulin is a multifunctional chaperone protein which mainly locates in the lumen of endoplasmic reticulum (ER). When exposed on the cell surface, calreticulin works as an “eat me” signal for dendritic cells and macrophages, initiating an immune response. Whether the expression of calreticulin in colon cancer is involved in induction of immune response is a hot topic on the role of calreticulin in pathogenesis of colon cancer.

Studies indicate that the expression of calreticulin is decreased in non-mucinous colonic adenocarcinoma while is increased in poorly -differentiated colon cancer and advanced tumor. Other studies have shown that one of the calreticulin fragments inhibits angiogenesis and growth of colon cancer cells. However, the correlation between expression of calreticulin and survival of colon cancer patients remains unknown. In this study, the expression of calreticulin in stage IIIB colon cancer was associated with the lower infiltration of CD45RO+ lymphocytes which was associated with a shorter 5-year survival time, suggesting that reduced expression of CRT may serve as a mechanism underlying immune escape in colon cancer.

In this study, reduced expression of endogenous “danger signals”, such as calreticulin, was found to be a new mechanism underlying immune escape in colon cancer, which is important for a better understanding of the immunogenicity of colon cancer, thus providing a new immunotherapy for colon cancer.

Calreticulin: A highly conserved 46 kDa Ca²⁺-binding protein, which is mainly located in the lumen of ER and has various versatile functions, like chaperone activity, regulation of Ca²⁺ homeostasis, and adhesion signaling. When cells were treated with anthracyclines, oxaliplatin, radiation and hypoxia, calreticulin exposure works as an “eat me” signal for dendritic cells and macrophages, initiating an immune response.

This is a well-conducted study on the correlation between calreticulin expression and infiltration of lymphocytes in colon cancer and their impact on the 5-year survival time of patients with stage IIIB colonic cancer.