Immunosurveillance in esophageal carcinoma: The decisive impact of regulatory T cells

ABSTRACT

Neoadjuvant radiochemotherapy of esophageal carcinomas causes a variable degree of depletion of tumor-infiltrating FOXP3⁺ regulatory T cells (Tregs). The frequency of local Tregs in the operative specimens negatively correlates with the pathological response and overall patient survival. These results underscore the importance of immunosurveillance in determining the fate of patients with esophageal cancers treated with radiochemotherapy.

Abbreviations

CTL

CD8⁺ cytotoxic T lymphocytes

ICD

immunogenic cell death

P2RX7

purinergic receptor P2X, ligand-gated ion channel 7

PD-1

programmed death 1

TLR4

toll like receptor 4

Treg

FOXP3⁺ regulatory T cell

TRG

tumor regression grade

Patients with advanced esophageal cancers are treated by preoperative chemoradiation therapy consisting in the combination of local radiotherapy and systemic chemotherapy, generally with either cisplatin or oxaliplatin. Radiotherapy is known to induce immunogenic cell death (ICD), and so does oxaliplatin (although cisplatin exemplifies a class of agents that is not able to induce ICD).^1,2 Driven by the growing awareness that several major cancer types (such as breast and colorectal cancer) are under immunosurveillance (meaning that the immune contexture largely determines patient prognosis) and that their treatment with ICD inducers is strongly influenced by changes in the local immune infiltrate,³ we investigated the anticancer immune response in esophageal cancer patients receiving neoadjuvant radiochemotherapy. For this, we obtained specimens from close-to-200 esophageal cancer patients and used immunohistochemical methods to assess the immune infiltrate in the operative specimens that is constituted either by the residual tumor or by cicatricial tissue (in the advent of a complete pathological response to the treatment). Importantly, the level of the histological response to preoperative chemoradiation therapy, as measured according to tumor regression grade (TRG) system, correlated with the density of FOXP3⁺ regulatory T cells (Tregs), meaning that the more pronounced was the therapeutic response, the lower was the frequency of Tregs invading the tumor or its remnants. More impressively, the rarefaction of local Tregs also correlated with overall and cancer-specific survival of the patients.⁴ In other words, the therapeutic response led to, or was favored by, the depletion of Tregs from the tumor, underscoring a strong link between the efficacy of radiochemotherapy and local immunosurveillance.

The aforementioned data support the notion that the immunosurveillance system plays a major role in disease outcome in esophageal cancer. The positive prognostic—and perhaps predictive—impact of low Treg infiltration into the tumor is very similar to that observed in breast cancer.⁵ Locally advanced mammary carcinomas treated with anthracycline-based neoadjuvant chemotherapy may manifest the depletion of tumor-infiltrating Tregs (which can be detected already after one single cycle of chemotherapy), which in turn correlates with the complete pathological response detectable after six cycles of chemotherapy.⁵

Nonetheless, there are also notable differences in the prognostic impact of the lymphocyte infiltrate and other immune parameters on esophageal and breast cancer. For example, in breast cancer successful neoadjuvant chemotherapy causes an increase in tumor-infiltrating CD8⁺ cytotoxic T cells (CTL). Hence, a favorable change in the CTL/Treg ratio predicts a pathological complete response.⁵ Such a positive correlation between CTL infiltration and treatment outcome is not found in esophageal carcinoma, and there is actually a tendency for a negative correlation,⁴ meaning that the CTL/Treg ratio cannot be used as a prognostic parameter. The absent (and perhaps paradoxical) role of CTL in immunosurveillance against esophageal cancer cannot be explained by an increased expression of the exhaustion marker programmed death 1 (PD-1) (as this has been found in renal cancer, in which local CTL have a negative prognostic impact).⁶ Indeed immunohistochemical staining of adjacent sections revealed that PD-1 is mostly expressed by other cell types than CD8⁺ CTL infiltrating the esophageal cancer post-radiochemotherapy (Fig. 1). This result appears intriguing in view of the possibility of using immune checkpoint-blocking antibodies that target either PD-1 or PD-ligand 1 l(PD-L1) for the treatment of relapsed esophageal carcinoma.

Figure 1.

Immunosurveillance in neoadjuvant radiochemotherapy of esophageal cancer. (A) Schematic relationship between treatment, Treg infiltration and therapeutic outcome. (B), (C) Representative immunohistochemical detection of CD8⁺ and PD-1 from two different patient specimens. Operative specimens were recovered after completion of radiochemotherapy.

Breast cancer patients bearing one or two alleles of a loss-of-function mutation in the genes coding for Toll-like receptor 4 (TLR4) or the purinergic receptor P2X, ligand-gated ion channel, 7 (P2RX7) exhibit a reduced therapeutic response to adjuvant chemotherapy.³ However, in esophageal cancer patients that were heterozygous or homozygous for the TLR4 mutation are characterized an improved therapeutic outcome.⁴ The biological mechanisms of this difference are unclear, although it is tempting to invoke the pleomorphic function of TLR4 in inflammatory processes (that favor esophageal carcinogenesis), metastatic spreading,⁷ as well as in anticancer immune responses triggered by ICD,³ that may explain these contradictory findings. Loss-of-function mutations of P2RX7 had no significant impact on the survival of esophageal cancer patients,⁴ yet did increase the local density of Tregs, underlining the importance of purinergic signaling for controlling Treg infiltration.⁸

The aforementioned findings underscore the general importance of immunosurveillance in multiple distinct cancer types⁹ including esophageal carcinomas.⁴ They also emphasize the importance of studying local immune parameters with great detail to unravel unexpected and hitherto unexplainable differences in the immunosurveillance system of distinct cancer types.

Disclosure of potential conflicts of interest

No potential conflicts of interest were disclosed.