Dataset Information

Immune cell topography predicts response to PD-1 blockade in cutaneous T cell lymphoma

ABSTRACT: Anti-PD-1 immunotherapies have transformed cancer treatment, but the determinants of clinical response are largely unknown. We performed CODEX multiplexed tissue imaging and RNA sequencing on 70 tumor regions of advanced cutaneous T cell lymphoma (CTCL) from 14 patients enrolled in a clinical trial of pembrolizumab therapy. Clinical response was not associated with the frequency of tumor-infiltrating T cell subsets, but rather with striking differences in the spatial organization and functional immune state of the tumor microenvironment (TME). After treatment, pembrolizumab responders had a localized enrichment of tumor and CD4+ T cells, which coincided with immune activation and cytotoxic PD-1+ CD4+ T cells. In contrast, non-responders had a localized enrichment of Tregs pre- and post-treatment, consistent with a persistently immunosuppressed TME and exhausted PD-1+ CD4+ T cells. Integrating these findings by computing the physical distances between PD-1+ CD4+ T cells, tumor cells, and Tregs revealed a spatial biomarker predictive of pembrolizumab response. Finally, the chemokine CXCL13 was upregulated in tumor cells in responders post-treatment, suggesting that chemoattraction of PD-1+ CD4+ T cells towards tumor cells facilitates a positive outcome. Together, these data show that T cell topography reflects the balance of effector and suppressive activity within the TME and predicts clinical response to PD-1 blockade in CTCL.

ORGANISM(S): Homo sapiens

PROVIDER: GSE162137 | GEO | 2021/09/17

REPOSITORIES: GEO

ACCESS DATA

Similar Datasets

Project description:Hepatocellular carcinoma (HCC) has dismal treatment responses to systemic therapies and is caused by both, viral and non-viral etiologies, including non-alcoholic steatohepatitis (NASH) 1–5. NASH - triggered by high caloric intake and sedentary lifestyle - has become an important driver for HCC development. Immunotherapy has been recently approved for HCC but stratification of responders versus non-responders has remained an unmet need 5–8. Here, we found a progressive accumulation of non-classical activated CD8+PD-1+ T-cells in livers of NASH-affected patients and mice. PD-L1/PD-1-targeted immunotherapy of NASH-induced HCC administered either at cancer-initiation or after cancer-establishment lacked beneficial effects in mice. On the contrary, PD-1-targeted immunotherapy drove hepatic necro-inflammation and increased liver cancer incidence, tumor number, and nodule size. Anti-CD8 or anti-CD8/anti-NK1.1 treatment suppressed liver cancer development, implicating CD8+ T-cells as liver cancer drivers in the context of NASH. In line, PD-1-/- mice challenged with a NASH-inducing diet displayed early-onset of liver cancer. Mechanistically, PD-1-targeted immunotherapy triggered necro-inflammation and a pro-tumorigenic environment in the context of NASH, increasing the abundance of hepatic TOX+CXCR6+ expressing CD8+PD-1+ TNF+ T-cells. Anti-CD8/anti-PD-1 or anti-TNF/anti-PD-1, but not anti-CD4/anti-PD-1 treatment reverted anti-PD1 treatment-related increase of liver inflammation, NASH severity and tumorigenesis. Gene expression profile and increased abundance of murine hepatic CD8+PD-1+ T-cells were corroborated in human CD8+PD-1+ T cells derived from NASH patients. In a meta-analysis of clinical trials testing PD-1-targeting immune checkpoint inhibitors alone (i.e. pembrolizumab) or in combination in 1656 patients with advanced HCC, immunotherapy was not favorable over to control treatment in non-viral- when compared to viral-related HCC. Similarly, in two clinical cohorts tested, patients with NASH-driven HCC had a significantly worse overall survival with PD-1-targeted immunotherapy than HCC patients with other etiologies. Our data identify non-viral etiologies, particularly NASH, as potentially non-responsive in the context of HCC immunotherapy, providing a strong rational basis for patient stratification.

Dataset Information

Immune cell topography predicts response to PD-1 blockade in cutaneous T cell lymphoma

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure