Thyroid Cancer and Autoimmunity

Christopher Alan Muir, MBBS, FRACP

The University of Sydney, Australia

Abstract

Background: Thyroid toxicity is common following immune checkpoint inhibitor (ICI) treatment. Published studies estimate the incidence at 10-20%, although rates vary widely between different ICIs. The etiology of ICI-associated thyroid immune related adverse events (irAEs) is unknown & not all patients develop a classic thyroiditis-like presentation of transient hyperthyroidism followed by a hypothyroid phase. Only small observational cohorts have been reported & the clinical & biochemical features of thyroid irAEs have not been well characterized. The current study aimed to describe thyroid irAEs in a large cohort of patients with melanoma. Methods: We reviewed outcomes in a prospective cohort of adult patients undergoing ICI treatment for advanced melanoma. Thyroid function was measured at baseline & at regular intervals during treatment. Thyroid irAEs were defined as new biochemical thyroid dysfunction developing over the course of routine follow-up. Results: Thyroid irAEs occurred in 518 of 1246 (42%) patients. Median follow-up was 11.3 months. Multiple patterns of thyroid-irAEs were observed, such as hyperthyroidism (subclinical or overt) in 31%, hypothyroidism in 8%, & euthyroid hyperthyroxinemia, hypothyroxinemia & isolated low FT3 syndrome each in 1% of participants. Thyroid irAEs were more frequent following combination (CTLA-4 + PD-1) ICI treatment (56%) than following PD-1 (38%) or CTLA-4 (25%) based monotherapies (p=0.001). The severity of thyroid irAEs differed by ICI, with higher rates of overt (vs. subclinical) thyroid dysfunction following combination ICI treatment (47%) relative to PD-1 (37%) & CTLA-4 (19%) monotherapies (p=0.001). Younger age (OR 0.88 per 10-yrs; 95% CI 0.81-0.96), female sex (OR 1.62; 95% CI 1.27-2.08) & combination ICI-treatment (vs. CTLA-4, OR 3.76, 95% CI 2.49-5.75; vs. PD-1, OR 1.90, 95% CI 1.45-2.49) were associated with higher odds of thyroid irAE. Time to onset of thyroid dysfunction was shorter in patients with overt hyperthyroidism relative to other types of thyroid irAE (log rank p=0.001). Overt hyperthyroidism was associated with increased irAEs in other organ systems (colitis, hepatitis, etc), increased irAE severity & increased multi-system irAEs than euthyroid patients or patients with other subtypes of thyroid irAE (p=0.003). Overt hyperthyroidism was also associated with improved progression free survival (HR 0.57; 95% CI 0.39-0.84; p=0.005) & overall survival (HR 0.68; 95% CI 0.49-0.94; p=0.02). No benefit to cancer survival was observed with other thyroid irAE subtypes. Conclusions: Thyroid irAEs were common. Combination ICI treatment resulted in more frequent, more severe, & earlier onset thyroid irAEs. Of thyroid irAE subtypes, overt hyperthyroidism was uniquely associated with increased immune responsiveness, as evidenced by higher incidence of extra-thyroidal irAEs & improvements in cancer survival.

Sophie Leboulleux, MD, PhD

Gustave Roussy, France

Abstract

Background: The benefits of post-operative radioactive iodine (RAI) administration have not been demonstrated in patients with low risk differentiated thyroid cancer (DTC). The objective of this randomized phase III trial is to assess in low risk DTC patients the non-inferiority of a follow-up strategy as compared to a systematic adjuvant post-operative RAI administration. Methods: ESTIMABL2 is a French multicentric randomized phase III trial in patients with low-risk DTC treated with total thyroidectomy with or without prophylactic neck lymph node dissection (pT1am N0 or Nx with a sum of the diameters of tumor lesions ≥ 10mm, pT1b N0 or Nx). Two to five months after surgery, in the absence of suspicious lateral neck lymph node on ultrasonography (US), patients were randomized either to the follow-up group (FU, no RAI administration) or to the ablation group and received post-operative RAI (1.1 GBq following rhTSH stimulation). Yearly controls under levothyroxine treatment consisted in thyroglobulin (Tg) and Tg antibodies (TgAb) determinations and neck-US. The primary objective was to assess at 3 years after randomization the non-inferiority of the proportion of patients without tumor-related event in the FU group as compared to the ablation group. Non-inferiority is demonstrated if the rate of patients without event at 3 years does not differ by more than ΔL=-5%. A tumor-related event was defined by the occurrence of subsequent treatment (RAI administration or surgery) for abnormal RAI uptake on the post-therapeutic WBS or by elevated Tg or TgAb levels and/or abnormal neck US during controls. Tg levels on levothyroxine treatment were considered elevated if > 2ng/mL in the FU group and > 1ng/mL in the ablation group. TgAb were considered elevated if > the upper limit range with an increase above 50% on 2 consecutive determinations performed 6 months apart. Results: 776 low-risk DTC patients were included between 2013 and 2017 in 35 French centers within the TUTHYREF network; 83% females, mean age: 52 years, papillary TC: 96%, pT1bNx: 43.6%, pT1bN0: 37.5%, pT1amNx: 12.6%, pT1amN0: 6.3%. Among the 729 patients evaluable at 3 years after randomization, tumor-related events occurred in 18/367 patients (4.9% IC95%=[2.9 ; 7.6]) in the FU group and in 15/362 patients (4.1% IC95%=[2.3 ; 6.7]) in the ablation group. Thus, 95.1% of patients in the FU group had no event at 3 years and this percentage is not inferior from the 95.9% of patients observed in the ablation group (difference = -0.8% [95% CI:-3.3%; 1.8%]. The number of subsequent surgery and/or RAI administration was 6 (1.6% IC95%=[0.6 ; 3.5]) in the FU group and 9 (2.5% IC95%=[1.1 ; 4.7]) in the ablation group. Conclusion: this phase III trial demonstrates the non-inferiority of a follow-up strategy compared to a systematic adjuvant post-operative administration of RAI (1.1GBq following rhTSH) in low risk DTC patients (PHRC 2012; NCT01837745).

Christine Do Cao, MD

CHRU, Hôpital Claude Huriez, France

Abstract

Background: Two-thirds of patients with metastatic differentiated thyroid cancer (DTC) become refractory to radioactive iodine (RAIR). The inhibition of the MAP-kinase pathway that is activated in case of BRAFV600E mutation might increase RAI incorporation into metastatic foci and reverse the RAI refractoriness.MERAIODE is a prospective multicentric open-label phase II trial, using a one-stage Fleming design, evaluating the efficacy and tolerance of trametinib (a MEK inhibitor) and dabrafenib (a BRAF inhibitor) treatment followed by the administration of RAI in metastatic RAIR DTC patients. Methods: Patients with BRAFV600E mutated RAIR metastatic DTC with RECIST progression within 18 months prior to enrollment and no lesion > 3 cm were included. A baseline rhTSH-stimulated diagnostic whole body scan (dc WBS) was performed prior to treatment initiation. Patients were treated with dabrafenib (150 mg bid) and trametinib (2 mg per day) for 42 days. At day 28, a second rhTSH-stimulated dc WBS was performed. After 35 days, a therapeutic activity of RAI (5.5 GBq) was administered. Primary endpoint was objective response rate (ORR) at 6 months according to RECIST v1.1 (central review). Patients: Among the 24 patients (mean age 67 years, 15 females) with a BRAFV600E mutated RAI refractory papillary DTC included between March 2018 and January 2020 in 8 French centers from the TUTHYREF netwok, 24 patients were treated and 21 patients were evaluable for the principal outcome at 6 months. Results: Abnormal RAI uptake was present in only 1 of the 21 patients (5%; 95%CI 0-24%) on a RAI diagnostic whole body scan (dc-WBS) performed prior to treatment initiation, in 11 patients, 11/17 (65%; 95%CI 38-86) on a dc-WBS performed 4 weeks after dabrafenib-trametinib initiation and in 20/21 (95%; 95%CI 76-100) on the post-therapeutic WBS performed after 5.5 GBq of RAI. The RECIST 6-months tumor response (central review) was partial response (PR) in 38% (95%CI 18-61), stable disease (SD) in 52% (95% CI 30-74) and progressive disease (PD) in 10% (95% CI 1-30). The median change in the sum of target lesions was -22% (range: -79 to +46) at 6 months after baseline. The 6-month fluorodesoxyglucose metabolic PET response was PR in 11/17 (65% 95%CI 38-86), SD in 4/17 (23%) (95% CI 7-50) and PD in 2/17 (12%; 95% CI 1-36). Among the 15 patients without Tg antibodies, 7 (47%) patients had a decrease of serum thyroglobulin level on T4 treatment by more than 50%All patients experienced at least one grade 1-2 adverse event, mainly asthenia, nausea, fever, diarrhea and cutaneous eruption. Nine grade 3 toxicities occurred in 6 treated patients. No grade 4-5 adverse event occurred?Conclusion:The association of dabrafenib and trametinib in BRAFV600E mutated patients is effective for restoring RAI uptake and is followed by a tumor control in 90% of patients and by tumor response in 38% with limited adverse events. (PHRC 2015, NCT 03244956)

Cheuk Wun Li, PhD

Albert Einstein College of Medicine, NY, United States

Abstract

Introduction: There is a strong genetic association between autoimmune thyroiditis (AITD) and Type 1 diabetes (T1D), and the co-occurrence of AITD and T1D in the same individual is referred to as Autoimmune Polyglandular Syndrome type 3 variant (APS3v). We previously discovered a unique amino acid signature of the HLA-DR pocket (designated APS3v HLA-DR pocket) that predisposes to APS3v, and we found that both thyroid (Tg.1571, TPO.758) and islet (GAD.492) peptides can bind to this flexible APS3v HLA-DR pocket. The goal of the present study is to identify small molecules that can block the presentation of these thyroid and islet peptides by the APS3v HLA-DR pocket as a potential therapeutic approach for APS3v.Methods: We screened a panel of small molecules using an in vitro screening assay we developed using Baculovirus-generated recombinant APS3v-DR. To validate “hit” compounds discovered by the in vitro assay in a mouse model we established a novel mouse model of APS3v. The APS3v mouse model was established by immunizing humanized NOD-DR3 mice with Tg.1571, TPO.758 and GAD.492 combined, together with adjuvant. We then validated hit small molecules for their effectiveness in blocking stimulation of T-cell responses in the APS3v mouse model. Results: Our screen identified 4 small molecules (S5, S9, S15, S53) that showed more than 50% inhibition in our in vitro assay. We were also able to establish a novel APS3v mouse model. Immunized mice showed an increase in cytokine production in response to all 3 thyroidal and islet peptides. The immunized mice also developed antibody responses against each peptide. When we tested the 4 “hit” compounds, only Cepharanthine (S53) which we previously identified as a small molecule blocking AITD immune responses, could block all 3 thyroidal/islet peptides in the APS3v mouse model. Treating with Cepharanthine prior to inducing APS3v in the mice also reduced T-cell proliferation in the immunized mice.Conclusions: We established a novel humanized APS3v mouse model by co-immunizing 3 thyroidal and islet peptides. We also identified Cepharanthine as a small molecule that can block thyroid and islet specific T-cell responses in this model. Cepharanthine can be further tested in APS3v patients carrying the specific APS3v-DR pocket as a potential therapeutic treatment.

Eyal Robenshtok, MD

Rabin Medical Center, Israel

Abstract