Thyroid-Specific T Cells in Patients With Differentiated Thyroid Cancer: Implications for Immune-Based Therapies?

Abstract

This editorial reviews the recent paper by Ehlers et al and provides a perspective on thyroid cancer immunology.

Inflammation has long been associated with autoimmune thyroid disease and thyroid cancers, raising seminal questions about the role of the immune system in the pathogenesis and potential therapy of advanced thyroid cancers (1). Lymph node (LN) metastasis, although common in patients with differentiated thyroid cancer (DTC), rarely leads to distant metastasis and can persist for years without evident progression. Slow progression of DTC may be explained, in part, by the indolent nature of the tumor; however, the immune system may be responsible for limiting tumor growth and metastases. In support of this theory, CD8+ T cells are commonly found infiltrating primary DTC and metastases (2, 3). Despite these advances in our knowledge of DTC and the immune response, little is known about the epitope-specific responses in DTC.

In this issue of JCEM, Ehlers et al. (4) investigate whether T cells specific for thyroglobulin (Tg) and thyroperoxidase (TPO) are evident in patients with papillary thyroid cancer (PTC), and whether this immune response correlates with clinical outcome. Using major histocompatibility complex I tetramers loaded with Tg- and TPO-derived peptides, the authors identified Tg- and TPO-specific T cells in the peripheral blood of patients with PTC. The authors suggest the frequencies of these cells were, on average, elevated over that seen in normal control patients and similar to that observed in patients with autoimmune thyroiditis (AIT).

One important caveat in this study is that the authors did not directly account for evidence of AIT in the primary PTC tumor. AIT was defined by the presence of autoantibodies (TPO >105 mIU/mL; Tg >120 mIU/mL) and optional hypoechogenicity in the thyroid sonography, which is not unreasonable. However, histological assessment of lymphocytic aggregates in the normal thyroid is necessary to verify whether Tg- and TPO-specific T cells are present in patients with PTC in the absence of underlying autoimmune disease (3).

Baseline measurements of Tg- and TPO-specific T-cell frequencies in peripheral blood were low: <0.6% of CD8 T cells. The authors chose to verify the presence of these populations by stimulating T cells overnight with Tg and TPO peptides. In a subset of patients with PTC, Tg- and TPO-specific T cells were elevated to 2% to 10% of CD8⁺ T cells, which was increased over that achieved by most samples with a nonspecific ovalbumin peptide (∼1%). The authors did not include normal or AIT controls in their analysis of tetramer-positive T cells after stimulation, which may have provided a reference for interpreting the stimulation seen in some patients with PTC. Importantly, a significant number of patients with PTC did not show evidence of Tg- and TPO-specific T cells either at baseline or after stimulation, suggesting that Tg and TPO may not be universal targets for the immune response to PTC.

One would expect that tumor-specific T cells would be found in association with the tumor. The authors do show evidence of tetramer binding to CD8⁺ T cells in tumor-involved LNs. However, they note in the Figure 4 legend that these Tg- and TPO-specific T cells were not evident within or in proximity to the metastatic tumor. The authors did not report evidence of Tg- and TPO-specific T cells in primary tumors. The lack of evidence for Tg- or TPO-specific T cells infiltrating or surrounding the tumor may suggest that Tg and TPO are not targeted by the antitumor immune response. The potential of Tg and TPO as targetable tumor antigens would be reinforced by evidence of Tg and TPO expression and peptide presentation by the tumor in patients that have generated T cells specific for Tg and TPO.

Ehlers and colleagues (4) also investigated the contribution of HLA class II haplotype in disease severity. The percentage of patients with distant metastases was lower in patients expressing HLA DQB1*03 (2.5%) compared with those who did not express HLA DQB1*03 (15%). In contrast, metastatic spread was more common in patients who expressed DRB1*03 and DQB1*02. Of note, Tg- and TPO-specific tetramer-positive CD8⁺ T cells were largely restricted to patients expressing DQB1*03⁺. If DQB1*03 is contributing to a productive immune response involved in controlling metastatic spread, it is likely that a professional antigen-presenting cell, perhaps a TPO- or Tg-specific B cell, is responsible for presentation of peptides to both a DQB1*03-restricted CD4⁺ T cell and antigen-specific CD8⁺ T cells. This hypothesis would be strengthened by further investigation into whether DQB1*03 is superior in presenting Tg- and TPO-derived peptides and whether DQB1*03 tetramer reactive CD4⁺ T cells are also enriched in these patients.

The potential of Tg as a tumor antigen is currently under investigation in an ongoing clinical trial [ClinicalTrials.gov no. NCT02390739 (registered 17 March 2015)]. In this study, autologous T cells are transduced with a murine T-cell receptor specific for human Tg in patients with Tg-expressing advanced DTC. Recombinant human interleukin-2 (aldesleukin) will be given within 24 hours of the cell infusion and continued every 8 hours for 3 days. Based on the current study by Ehlers and colleagues (4), one may hypothesize that expression of the HLA DQB1*03 allele may predict better responsiveness to the autologous T-cell therapy.

Clinical trials are now investigating the efficacy of immune checkpoint blockades (i.e., anti-PD-1 and anti-CTLA-4) in patients with thyroid cancer. These therapies enhance existing antitumor immunity through a general blockade of inhibitory signals that negatively regulate the immune response and maintain self-tolerance. Preliminary reports from a recent phase 1 trial for pembrolizumab (anti-PD-1) in advanced solid tumors [ClinicalTrials.gov no. NCT02054806 (registered 3 February 2014); MK-3475-028/KEYNOTE-28] revealed that two of 22 (9.1%) patients with DTC achieved a partial response to anti-PD-1 as a single agent (5). A phase 2 trial investigating combination therapy with pembrolizumab plus lenvatinib in patients with advanced DTC is now in development in collaboration with the International Thyroid Oncology Group [ClinicalTrials.gov no. NCT02973997 (registered 23 November 2016)]. In addition, a current trial testing the efficacy of stereotactic body radiation therapy of lung or liver metastases followed by ipilimumab treatment has now added a thyroid cancer expansion cohort [ClinicalTrials.gov no. NCT02239900 (registered 11 September 2014)].

Ehlers and colleagues’ findings (4) may encourage the analysis of HLA class II haplotype and/or peripheral TPO or Tg epitope-specific T cells as biomarkers of responsiveness to therapy for the current and upcoming immune-based therapy studies in patients with advanced DTC.