Pembrolizumab Fails to Yield Responses in Recurrent Medullary Thyroid Cancer

Standard-of-care (SOC) pembrolizumab (Keytruda) did not generate therapeutic responses in patients with recurrent or metastatic medullary thyroid cancer, regardless of prior immunotherapy treatment status, according to findings from a phase 2 trial (NCT03072160), which were published in Oncologist.¹

No evidence of therapeutic response was observed among patients who had previously received the experimental immunotherapy carcinoembryonic agent GI-6207 (n = 13) or those without prior GI-6207 exposure (n = 4). Additionally, no patients on the trial (n = 17) had confirmed 50% declines in either calcitonin levels or serum carcinoembryonic antigen (CEA) levels. Moreover, investigators observed no radiographic responses and no evidence of the immune effect of pembrolizumab in peripheral blood mononuclear cell (PBMC) analyses.

In the prior immunotherapy cohort, the median progression-free survival (PFS) was 210 days, and the 2-year overall survival (OS) rate was 100%. In the cohort of patients with no prior immunotherapy, the median PFS was 55 days, and the 2-year OS rate was 50%.

Due to the lack of responses among patients who had received prior immunotherapy, the study was closed early.

“This study did not show any evidence that sequencing immunotherapy with immune checkpoint inhibitors [ICIs] improves clinical efficacy of ICIs in tumor types where they do not have independent activity,” lead study author Jaydira Del Rivero, MD, and coauthors wrote in the paper.

Del Rivero is a physician scientist in the Developmental Therapeutics Branch of the Center for Cancer Research at the National Cancer Institute in Bethesda, Maryland.

What was the rationale for investigating pembrolizumab in GI-6207–exposed thyroid cancer?

Patients with medullary thyroid cancer are not known to have therapeutic responses to ICIs. However, the trial investigators hypothesized that sequencing antigen-directed immunotherapy with ICIs may prime tumors, including medullary thyroid tumors, for immune response to ICIs.

Notably, the phase 2 QUILT-3.006 trial (NCT01856920) previously investigated the efficacy of GI-6207, a yeast-based, CEA-targeted immunotherapy agent, in patients with recurrent medullary thyroid cancer.² The trial was supported by preclinical and early-phase data indicating a strong CEA immune response and antitumor activity associated with GI-6207.

What was the design of the phase 2 trial evaluating pembrolizumab in medullary thyroid cancer following GI-6207 treatment?

The current phase 2 trial included 2 cohorts of patients with stage IV medullary thyroid cancer.¹ One cohort consisted of patients who had previously received GI-6207 as part of QUILT-3.006, and the other cohort enrolled patients with no treatment history with GI-6207. All patients received SOC intravenous pembrolizumab at 200 mg every 3 weeks for a maximum of 2 years. Patients also underwent imaging every 3 months during the trial. Additionally, investigators evaluated PBMCs for immune responses every 3 months.

The primary end point was a 50% or greater decrease in calcitonin levels or a partial or complete response on imaging. Secondary end points included the effect of previous immunotherapy on study treatment, immune responses, changes in calcitonin and CEA kinetics, and safety.

What were the characteristics of the patients enrolled in the trial examining ICI sequencing in pretreated medullary thyroid cancer?

The patients with prior GI-6207 exposure had a median age of 52.4 years, most were female (n = 10), and all had an ECOG performance status of 0 or 1. The median CEA level was 125 ng/mL (range, 3.7-1650), and the median calcitonin level was 509 pg/mL (range, 65-35,607).

Among the patients without prior GI-6207 treatment, the median age was 58.8 years, most were female (n = 3), and all had an ECOG performance status of 0 or 1. The median CEA level was 207 ng/mL (range, 153-943), and the median calcitonin level was 60,000 pg/mL (range, 2672-103,312).

What additional findings were observed with pembrolizumab in ICI-pretreated medullary thyroid cancer?

An immune analysis in evaluable patients (n = 16) showed that on days 1 vs 16, the respective median percentages of PBMCs in the following classic subsets were:

• CD4-positive cells: 34.73% vs 33.95% (P = .926)
• CD8-positive cells: 13.70% vs 12.50% (P = .445)
• T-regulatory cells: 0.70% vs 0.77% (P = .210)
• Natural killer (NK) cells: 10.47% vs 10.26% (P = .780)
• NK T cells: 3.20% vs 3.42% (P = .838)
• B cells: 7.32% vs 8.35% (P = .780)
• Classical dendritic cells: 0.38% vs 0.55% (P = .210)
• Plasmacytoid dendritic cells: 0.15% vs 0.18% (P = .564)
• Myeloid-derived suppressor cells: 10.13% vs 11.93% (P = .196)
• Monocytes: 11.58% vs 11.43% (P = .381)

Regarding safety, the investigators observed no new safety signals with pembrolizumab. Adverse effects (AEs) of grade 2 or higher that were attributed to pembrolizumab included acute kidney injury (grade 2, n = 0; grade 3, n = 1), increased alanine aminotransferase levels (n = 1; n = 0), decreased neutrophil counts (n = 1; n = 0), hearing impairment (n = 1; n = 0), headache (n = 0; n = 1), fatigue (n = 2; n = 0), and rash (n = 3; n = 1). Notably, no grade 4 or 5 AEs were reported.

“The lack of therapeutic response of this sequencing strategy is consistent with findings in other cancers like prostate cancer, where ICIs are not effective beyond a minor subset of patients, such as those with high mutation burden or microsatellite instability,” the authors concluded. “Future research strategies may focus on novel immunotherapy strategies (e.g. bispecific [antibodies]) as opposed to sequencing of or combinations with ICIs.”

References

1. Del Rivero J, Donahue RN, Marte JL, et al. Pembrolizumab in recurrent or metastatic medullary thyroid cancer. Oncologist. Published online December 27, 2025. doi:10.1093/oncolo/oyaf348
2. QUILT-3.006 for recurrent medullary thyroid cancer. ClinicalTrials.gov. Updated July 9, 2024. Accessed January 13, 2026. https://clinicaltrials.gov/study/NCT01856920