Prognostic Significance of CD3⁺ Tumor-infiltrating Lymphocytes in Patients with Endometrial Carcinoma

Discussion

The present study adds to the growing body of evidence indicating that the immune system plays an important role in the progression and prognosis of endometrial carcinoma. A significant trend of increasing CD68⁺ TAM and decreasing CD3⁺ TIL counts associated with higher stage or grade was recorded. Compared to patients with early tumors, stromal CD8⁺ TIL counts were increased relative to intraepithelial CD8⁺ TIL and stromal CD3⁺ TIL counts in patients with advanced disease. Moreover, a significant difference was observed in patients with high-risk and low-risk early-stage tumors in stromal CD3⁺ TIL and CD20⁺ TIL counts. Most importantly, although none of the TIL populations studied was associated with prognosis when the whole cohort was examined, among patients with early-stage, high-risk tumors, intraepithelial CD3⁺ TIL counts below median were predictive of poor prognosis. The Kaplan–Meier overall survival curves started to diverge at three years after diagnosis, and later, eight to nine years after diagnosis, both curves seemed to reach a plateau, with the difference in long-term survival rates of patients with high versus low intraepithelial CD3⁺ TIL counts approaching almost 40%. In multivariate analysis of patients with early tumors, high intraepithelial CD3⁺ TIL counts and low intraepithelial CD8⁺ TIL counts were significant prognostic biomarkers that surpassed the prognostic significance of the risk classification itself, which was not a significant predictor of outcome in multivariate analysis. Interestingly, high intraepithelial CD3⁺ and CD8⁺ TIL counts had opposing effects on prognosis. In the sub-group of patients with high-risk early tumors, CD3⁺ TIL counts were also an independent prognostic factor.

Endometrial cancer usually manifests early by vaginal bleeding, resulting in timely diagnosis and curative surgical therapy. In contrast to other gynecological malignancies, e.g. epithelial ovarian carcinoma, most patients with endometrial carcinoma are cured by surgery-alone or in combination with external-beam radiation with or without vaginal brachytherapy. Consequently, compared to epithelial ovarian carcinoma, patients with advanced/metastatic disease are rare such that metastatic endometrial carcinoma represents an almost ‘orphan’ disease, and the information on the systemic treatments and utilization of biomarkers in endometrial cancer is limited (12). There is an unmet medical need for new treatments and biomarkers that would allow the prediction of prognosis and response to therapy. Currently, tumor stage, depth of myometrial invasion, and tumor grade are commonly used to predict clinical outcome and to plan treatment modalities in patients with endometrial carcinoma (13). Precise histological assessment of myometrial invasion and cervical involvement is crucial in the staging of early disease. Patients with more than 50% myometrial invasion are at increased risk for extrauterine metastasis and may require more aggressive surgical staging. Vascular and lymphatic vessel invasion are encountered in approximately one-fifth of cases of endometrial carcinoma and correlate significantly with the extension of the primary tumor, depth of myometrial invasion and histological grade. The presence of vascular invasion has been shown to be a powerful independent indicator for increased risk of recurrence and diminished survival in patients with clinical stage I endometrial carcinoma. The grading system relies firstly on the architectural pattern of the glands and secondarily on the cytological and nuclear appearances, but grading itself, as well as the subtyping of high-grade endometrial carcinoma, was found to be frequently subjective and poorly-reproducible. Among parameters that can be assessed using immunohistochemistry, estrogen receptors and progesterone receptors expression play a major role. Approximately 60-70% of endometrial carcinomas express both receptors, which seem to characterize clinically less aggressive tumors, with a better chance of response to endocrine therapy. Data have also been reported on the significance of other molecular alterations, such as p53 mutation which is more frequently detected in serous endometrial tumors, microsatellite instability, mutation of phosphatase and tensin homologue on chromosome ten, β-catenin, or Kirsten retrovirus-associated DNA sequences genes that characterize the pathogenesis of endometrial tumors (13). Potential prognostic significance has also been attributed to markers of proliferation of tumor cells, such as Ki-67.

Considerable progress has been made in the past few years in understanding the role of the immune system and host–tumor interactions in tumor progression (14). The presence of TILs within the tumor microenvironment is considered to represent a biomarker of the host immune response against the tumor. T-Cells and macrophages constitute major components of tumor-infiltrating leukocytes in epithelial malignancies. Depending on the function and phenotype, T-cells may either act as effectors in tumor cell destruction, or induce immune tolerance. Macrophages may also play a more complex role in tumor progression and metastasis. Depending on the context of the tumor microenvironment, macrophages may represent potent effectors of the host antitumor response and a crucial component of the stroma contributing to tumor cell proliferation and angiogenesis (15-17). In different reports examining patients with endometrial carcinoma, the presence of TILs was associated with microsatellite instability (11), expression of B7-H7, a membrane protein that negatively regulates T-cell response (10), and the expression of indoleamine-2,3-dioxygenase, the enzyme responsible for tryptophan depletion (6). The reported results on the prognostic significance of the presence of TILs in patients with endometrial carcinoma are inconsistent, or even conflicting because of differences in size of the cohorts examined, number of events or differences in methodology, including differences in lymphocyte populations analyzed or methods for the detection of lymphocytic infiltration. In an early report, the presence of higher numbers of intraepithelial CD8⁺ TILs at the invasive border was an independent predictor of improved overall survival (7). The association of higher CD8⁺ TIL counts with prolonged time to recurrence and improved overall survival was also reported in a large cohort of 368 patients with endometrial carcinoma (8). A positive association of overall survival with TILs that exhibit CD45RO⁺ (memory) T-cell phenotype has also been noted (8). In some studies, the prognostic significance of forkhead box P3 (FOXP3)⁺CD4⁺ TILs was examined (5, 8, 9), and an association of high FOXP3⁺CD4⁺ TIL counts with disease-free survival (5) and overall survival (8) has been reported in patients with endometrial carcinoma. The lack of prognostic significance of FOXP3⁺CD4⁺ TILs reported in another study may be attributed to an insufficient number of events (9). The presence of FOXP3⁺CD4⁺ TILs was also demonstrated to correlate with clinical and pathological characteristics associated with high risk, including vascular density, stage, differentiation, and lymphovascular space invasion (5, 9). In a study that demonstrated the association between indoleamine-2,3-dioxygenase expression and tumor lymphocyte infiltration, high stromal CD3⁺ TIL counts were reported to independently predict better relapse-free survival (6). Less is known about the significance of TAMs in patients with endometrial carcinoma. In one study, CD68⁺ TAM counts at the invasive margin were associated with tumor stage, grade, myometrial invasion, lymph node metastasis and vascular space invasion (18). Intratumoral CD68⁺ TAM counts were positively associated with Ki-67 expression and microvascular density. High CD68⁺ TAM counts at the invasive margin were associated with poor prognosis in univariate, but not multivariate analyses (18).

• Download figure
• Open in new tab
• Download powerpoint

Figure 1.

Intraepithelial and stromal tumor-infiltrating lymphocytes. Immunohistochemically-stained cluster of differentiation (CD3)⁺ lymphocytes in endometrial carcinoma are shown. Both stromal and intraepithelial CD3⁺ lymphocytes are evident (original magnification, ×200).

• Download figure
• Open in new tab
• Download powerpoint

Figure 2.

Survival of patients with early-stage, high-risk endometrial carcinoma according the intraepithelial cluster of differentiation (CD3)⁺ tumor-infiltrating lymphocyte (TIL) counts. The Kaplan–Meier survival curves of patients with intraepithelial CD3⁺ TIL counts <17/mm² and ≥17/mm² in patients with early-stage, high-risk tumors are shown (median survival of 52.7 months versus not reached, log-rank test, p=0.01).

Adjuvant radiation prevents only locoregional relapse, and the benefit of adjuvant chemotherapy in endometrial carcinoma is controversial. In fact in most prospective trials, the benefit of adjuvant chemotherapy was not evident. The lack of demonstrable benefit of adjuvant systemic treatment could be due to the fact that subsequent distant metastases occur infrequently in patients with early-stage disease. A prognostic biomarker would first allow identification of a high-risk patient population to be enrolled into prospective studies of adjuvant systemic therapy and, subsequently, serve as a biomarker for treatment decisions. Future systemic therapies may include not only cytotoxic chemotherapy regimens, but also targeted-therapy. Conceptually, targeted treatments aim at one or more hallmarks of cancer. Hallmarks of cancer are defined as pathological features that enable tumor growth and escape from control, and one of the eight hallmarks currently recognized is the escape of tumor from destruction by the immune system (14). CD3⁺ TIL counts could serve as a biomarker for targeted therapies that aim at manipulating the immune system, e.g. ipilimumab or nivolumab.

In the present study, no prognostic significance of intraepithelial CD3⁺ TIL counts was detected when the whole cohort of patients was examined; an association with prognosis was found only in patients with early-stage tumors. However, patients with advanced endometrial cancer represent a subgroup of patients that is different from a point of view of prognosis and biology. The outcome of patients with advanced endometrial carcinoma is generally poor, and other factors, including therapy, age or comorbidity, may modify the effect of the immune response on patient prognosis. Moreover, patients with advanced disease constituted only 20% of patients in the present study. Despite the fact that the prognostic significance of CD3⁺ TIL counts was not evident in the whole cohort, CD3⁺ TIL counts were independent predictors of prognosis in patients with early-stage tumors, i.e. patients for whom the decision on the choice of therapy is the most urgent and controversial. Of note, patients with early-stage tumors in the present cohort were almost equally split between the low- and high-risk groups. In fact, the identification of patients with high risk is most crucial in patients with early disease, some of whom would have excellent prognosis without any adjuvant treatment, while others will eventually experience relapse.

As in other reports (2, 3), the presence of CD3⁺ TILs in the present study was more relevant in the tumor epithelium compared to the tumor stroma. In general, lymphocytic infiltration is more abundant in the stroma compared to the tumor epithelium, but intraepithelial lymphocytes are more likely to be associated with the immune response to the tumor than lymphocytes encountered in the tumor stroma. The finding of prognostic significance for the intraepithelial CD3⁺ TIL count is in contrast with a report demonstrating prognostic significance of stromal CD3⁺ TIL counts that, however, analyzed progression-free survival based on a smaller number of events (6). Interestingly, high CD8⁺ TIL counts were associated with adverse prognosis in the present cohort, in contrast with studies reporting positive prognostic significance of high CD8⁺ TIL counts in patients with endometrial carcinoma (7, 8).

The current study has obvious limitations that are inherent to its retrospective nature. Because of the retrospective nature of the study, overall survival rather than relapse-free survival or progression-free survival was analyzed in the present cohort. For obvious reasons, overall survival represents the most reliable end-point in a study that attempts to identify prognostic biomarkers. From the current perspective, surgical management of the patients in the present cohort may appear suboptimal. On the other hand, the strong association of intraepithelial CD3⁺ TIL counts with overall survival observed in the present study indicates the activity of the immune system determines the natural history of this cancer.

In conclusion, the present data demonstrate that in patients with endometrial carcinoma, a decrease of intraepithelial CD3⁺ TIL count is associated with advancing stage and risk group. The intraepithelial CD3⁺ TIL count is also an independent predictor of survival. Future studies should address the role of the intraepithelial CD3⁺ TIL count as a prognostic and predictive biomarker in adjuvant therapy of endometrial carcinoma.