CD33+/p-STAT1+ double-positive cell as a prognostic factor for stage IIIa gastric cancer
- First Online:
- Cite this article as:
- Dong, J., Li, J., Liu, SM. et al. Med Oncol (2013) 30: 442. doi:10.1007/s12032-012-0442-2
- 783 Downloads
Tumor-infiltrating immune cells are associated with tumor prognosis, although the type of immune cells responsible for local immune escape is still unknown. This study examined the relationship between gastric cancer survival and the density of immune cells, including CD8+ T cells, CD20+ B cells, and CD33+/p-STAT1+ cells, which represent myeloid-derived suppressor cells, to evaluate the role of immune cells in the progression of gastric cancer. One hundred pathologically confirmed specimens were obtained from stage IIIa gastric cancers between 2003 and 2006 at Sun Yat-sen University Cancer Center, China. The density of tumor-infiltrating immune cells in tumor tissue was examined using immunohistochemical analysis. Clinicopathologic parameters and the survival rate were analyzed in relation to the density of immune cells. A high density of CD8+ T cells and CD20+ B cells was associated with a good clinical outcome, but a high density of CD33+/p-STAT1+ cells was associated with a poor clinical outcome. Most importantly, the density of CD33+/p-STAT1+ cells was an independent prognostic factor and inversely related to the infiltration of CD8+ T cells. Although the infiltration of CD8+ T cells and CD20+ B cells is involved in the progression of gastric cancer, these data suggest that CD33+/p-STAT1+ cells play a central role in the regulation of the local immune response, suggesting that CD33+/p-STAT1+ cells might be therapeutic targets in gastric cancer.
KeywordsGastric cancer CD33 STAT1 T cell B cell Regulation
Gastric cancer is one of the most aggressive diseases worldwide, particularly in Asian countries, such as China . In addition, gastric cancer is the second leading cause of cancer-related death each year due to the resistance and intolerance of this cancer to cytotoxic therapy [2, 3, 4]. New therapeutic strategies are needed to treat gastric cancer. Recently, immunotherapy has received increased attention because antibodies targeting immune checkpoint molecules and therapeutic vaccines have been identified [5, 6, 7, 8]; therefore, identifying patients who are suitable for immune therapy is critical.
The density of tumor-infiltrating immune cells is associated with the prognosis of multiple types of tumors [9, 10, 11, 12, 13]. However, tumor-infiltrating immune cells consist of both tumor-rejecting cells and tumor-promoting cells. Therefore, a detailed analysis of tumor-infiltrating immune cells would lead to more effective prognostic indicators, and their potential use as biomarkers for immunotherapy could be determined [14, 15].
Immune cells infiltrates are different among tumor types and from patient to patient. Tumor-infiltrating lymphocytes (TILs) are considered to be a manifestation of the host immune reaction to cancer cells . Large numbers of T and B lymphocytes are associated with a good clinical outcome in many different tumor types [13, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26].
In addition to tumor-rejecting immune cells, immune suppressor cells play a pivotal role in tumor progression. Recently, several studies have demonstrated the importance of myeloid-derived suppressor cells (MDSCs) in tumor-associated immune suppression [27, 28]. MDSCs are recognized as a heterogeneous population of myeloid cells that consist of immature myeloid cells (IMCs) and myeloid cells at early stages of differentiation [27, 29, 30]. The role of the MDSCs that infiltrate into solid tumor tissue is still unclear. An obstacle to assaying the role of MDSCs in human tissue is the lack of defined markers for MDSCs. Among the markers reported, CD33 and CD11b are the basic markers of myeloid-derived suppressor cells. Although they are feasible to assay in blood cells, staining cells in tissue for multiple markers is very difficult, especially for markers with the same subcellular localization. To develop a set of markers that are suitable for the detection of infiltrative MDSCs, we reviewed the mechanisms of activation and expansion of MDSCs. The expansion and activation of MDSCs are influenced by several different factors, including VEGF, GM-CSG, SCF, TGF-β, and MMP9, among others [31, 32, 33, 34, 35, 36, 37, 38, 39]. These factors trigger the activation of several different signaling pathways in MDSCs that promote JAK-STAT signaling, including STAT1, STAT3, and STAT6 [28, 40]. To investigate this role, CD33 was used as a marker of immature myeloid cells, and p-STAT1 was used as a marker of IMCs activated by immune regulatory cytokines [40, 41]. Therefore, CD33+/p-STAT1+ cells were proposed to be a subset of MDSCs in gastric cancer tissues. This study examined the relationship between the survival rate of gastric cancer and the density of immune cells, including TILs and CD33+/p-STAT1+ cells, which represented MDSCs, to evaluate the role of the immune response in gastric cancer progression.
Materials and methods
Formalin-fixed, paraffin-embedded tissues from 100 gastric cancer patients were used. Gastric cancer biopsy specimens were collected from stage IIIa (2009 UICC staging system) gastric cancer patients between 2003 and 2006 at the Sun Yat-sen University Cancer Center. All of the patients underwent radical resection, and none of the patients had chemotherapy or radiotherapy before sample collection. Patients received 5-FU-based adjuvant chemotherapy post-operatively for 6 months. If recurrence or metastasis occurred, 5-FU-based chemotherapy was given according to the NCCN guidelines. This study was conducted in accordance with the Declaration of Helsinki, and all patients signed a consent form approved by the Research Ethics Committee of the Sun Yat-sen University Cancer Center.
Immunohistochemistry and scoring systems
Paraffin-embedded tissues were sectioned continuously at a thickness of 4 μm and heated for 1 h at 65 °C. Briefly, the sections were deparaffinized using xylenes and rehydrated with a graded alcohol series and distilled water. The sections were immersed in an EDTA antigen retrieval buffer (pH 8.0), subjected to high pressure for 3 min for antigen retrieval and allowed to cool to room temperature. After blocking with sheep serum, the sections were incubated overnight at 4 °C with mouse monoclonal antibodies against human CD8, CD20, and GrB (Zymed, San Diego, CA, USA), all of which were diluted 1:400. Following incubation with secondary antibodies, the sections were developed using diaminobenzidine tetrahydrochloride (DAB) and counterstained with hematoxylin.
The co-expression of CD33 and p-STAT1 was detected with sequential immunohistochemical staining using the EnVison™ G/2 Doublestain System (Dako Cytomation, Glostrup, Denmark) according to the manufacturer’s instructions. Endogenous peroxidases and alkaline phosphatase enzymes were blocked with the dual endogenous enzyme-blocking reagent provided in the kit. The sections were treated with normal goat serum for 20 min to reduce nonspecific binding and incubated overnight at 4 °C with rabbit polyclonal anti-CD33 antibody (1:100; Protein Tech Group, Chicago, IL, USA) and rabbit monoclonal anti-p-STAT1 (1:400; Cell Signaling, Boston, MA, USA). For the color reaction, diaminobenzidine (brown) and permanent red (red) were must used. As a negative control, the antibodies were replaced with phosphate-buffered saline (PBS).
The density of immune cells within the tumor specimens was scored as we previously reported . Briefly, the number of cells and the cell size were counted in at least 10 different fields of each section. The size of each high-powered field (400×) was approximately 300 μm × 300 μm, and the cells were counted in the intratumoral compartment. The areas of highest density were chosen, and necrotic areas were avoided. Two observers counted the cells at the same time in the same field using a multiple-lens microscope. The median value was used to distinguish the different groups of immunohistochemical variables in the results.
All statistical analyses were performed with the SPSS 16.0 statistical software package. The median value was used to differentiate high- and low-density groups for each immunohistochemical variable. The correlation between the density of immune cells and patient characteristics and the correlation between the density of TILs and CD33+/p-STAT1+ cells were analyzed using correlation test. Kaplan–Meier curves were used to estimate the distribution of variables in relation to survival, which were compared using the log-rank test. Univariate and multivariate analyses were based on the Cox proportional hazards regression model. Overall survival (OS) was defined as death from any cause, and disease-free survival (DFS) was defined as the time prior to relapse of the primary tumor. p < 0.05 was considered to be statistically significant.
Among the 100 patients, there were 71 men and 29 women, and the median age was 59.5 years, with a range of 29–82 years. Fifty-five patients (55 %) were Borrmann classification II, 38 were Borrmann classification III, and 7 cases were Borrmann classification I or IV. All of the patients presented lymph node metastasis before treatment, with 36 patients at the N1 stage, 34 patients at the N2 stage, and 30 patients at the N3 stage. During follow-up, 54 died, and 44 presented with disease progression.
CD8, CD20, GrB, and CD33+/p-STAT1+ cells in gastric cancer tissues
To characterize MDSC infiltration in gastric cancer tissue, we defined MDSCs as CD33+/p-STAT1+ cells. These CD33+/p-STAT1+ double-positive cells have been found mostly in the stroma among gastric cancer (Fig. 1g, h), whereas they had not yet been observed in the normal mucosa adjacent to gastric cancer and normal mucosa far from the gastric cancer (data not shown). The CD33 immunostaining demonstrated cytomembrane staining, and p-STAT1 demonstrated nucleus staining. Figure 1g, h shows the CD33/p-STAT1 double-positive cells in a subset of cells around the tumor nests.
Relationship between the density of CD33+/p-STAT1+ cells and the density of CD8+ and CD20+ lymphocytes
Relationship between CD33+/p-STAT1+ cells and the density of TILs
Density of CD8+ T cells
Density of CD20+ B cells
Relationship between the density of CD8, CD20, and CD33+/p-STAT1+ cells and clinicopathologic characteristics
Correlation between clinicopathologic features and immune cell density
CD8+ T lymphocytes
CD20+ B lymphocytes
(n = 38)
(n = 62)
(n = 43)
(n = 57)
(n = 43)
(n = 57)
Tumor size (cm3)
Association of the density of immune cells with patient survival
Univariate analysis of factors associated with OS and DFS
OS (n = 100)
DFS (n = 100)
HR (95 % CI)
HR (95 % CI)
Size of tumor
CD8+ T cells
CD20+ B cells
Multivariate Cox proportional hazard analysis
Multivariate analysis of factors associated with OS and DFS
OS (n = 100)
DFS (n = 100)
HR (95 % CI)
HR (95 % CI)
1.515 (0.857, 2.677)
1.603 (0.903, 2.846)
CD8+ T cells
0.595 (0.321, 1.101)
0.571 (0.309, 1.054)
CD20+ B cells
0.555 (0.312, 0.986)
0.632 (0.356, 1.121)
4.674 (2.525, 8.652)
4.670 (2.537, 8.596)
To enhance our understanding of the contribution of the local immune response to the progression of gastric cancer, this study characterized CD8+ T lymphocytes, CD20+ B lymphocytes, and CD33+/p-STAT1+ cells in the intratumoral region. The results showed that the density of CD8+ T cells, CD20+ B cells, and CD33+/p-STAT1+ cells in tumors was a useful criterion for the prediction of survival in advanced gastric cancer patients. Most importantly, Cox multivariate analysis demonstrated that the density of CD33+/p-STAT1+ cells within the cancer tissue was an independent prognostic factor. These data indicated that CD33+/p-STAT1+ cells play an important role in the progression of gastric cancer.
Several recent studies have analyzed the important role of tumor-infiltrating cytotoxic CD8+ T and B lymphocytes in the antitumor immune response [42, 43, 44, 45, 46, 47, 48, 49]. The effects of CD8+ T cells in cancer cell nests might be related to the effector function of killer T cells, and the role of B lymphocytes as part of the adaptive humoral immune response has been associated with improved survival in several types of cancer [25, 26, 50].
Although some studies have indicated that immune surveillance and immune escape exist in gastric cancer [51, 52, 53], the type of immune cells responsible for local immune escape is still unknown. Experimental data indicated that MDSCs suppress the immune response, in contrast to CD8+ T cells and CD20+ B cells. The main obstacle is the markers for MDSCs in the human tissue. As mentioned above, besides the basic markers of human MDSCs, the signaling pathways associate with the MDSCs expansion and activation are also very important. STAT1 is one of the pathways in MDSCs activation; Gabrilovich demonstrated that STAT1 activation in tumor-associated macrophages is responsible for the upregulation of inducible nitric oxide synthase (iNOS) and arginase 1 activity in these cells, which results in T-cell suppression . STAT1-deficient MDSCs are unable to inhibit T-cell activation due to their inability to upregulate iNOS and arginase 1 activity [54, 55]. Therefore, CD33 staining on the cell membrane was used as a marker of myeloid cells in this study, and p-STAT1 nuclear staining was used to represent a transcription factor activated by pro-inflammatory cytokines. For this reason, CD33+/p-STAT1+ cells are defined as a specific type of MDSC in gastric cancer tissues, although further functional analysis is required. Using this set of MDSC markers, it was possible to examine the infiltration of MDSCs into cancer tissues.
To confirm the role of CD33+/p-STAT1+ cells, which represented MDSCs in gastric cancer tissue, this study detected the density of CD8+ T cells, CD20+ B cells, and CD33+/p-STAT1+ immune cells infiltrating tumor tissue. A higher density of CD33+/p-STAT1+ cells was inversely related to infiltrating CD8+ T lymphocytes but not B cells, supporting that CD33+/p-STAT1+ cell might be a subset of MDSCs in gastric tumor tissue. Although Kaplan–Meier analysis and log-rank tests revealed that the all of the immune cells were significantly associated with OS and DFS duration, Cox multivariate analysis demonstrated that only the density of CD33+/p-STAT1+ cells within the cancer tissue was a prognostic factor for OS and DFS, independently, indicating that CD33+/p-STAT1+ cells play a key role in the regulation of the local immune response.
Overall, this study demonstrated that the density of CD33+/p-STAT1+ cells is an independent prognostic factor for patients with stage IIIa gastric cancer, indicating that CD33+/p-STAT1+ cells play a central role in the local immune response in gastric cancer. Therefore, CD33+/p-STAT1+ cells might be a useful therapeutic target.
This work was supported by the grants from the National Nature Science Foundation of China (30972882, 81272341) and the Key Project of Science and Technology of Guangdong Province (2008B030301079).
Conflict of Interest
Open AccessThis article is distributed under the terms of the Creative Commons Attribution License which permits any use, distribution, and reproduction in any medium, provided the original author(s) and the source are credited.