Science Spotlight

Understanding the tumor microenvironment for a better fight

from the Pollack lab (Clinical Research Division)
The picture represents an electron microscope picture of a soft tissue sarcoma.
Electron microscope picture of a soft tissue sarcoma. Released from the NIH/NCI.

Soft tissue sarcomas (STS) develop in fat, muscle or fibrous tissues and represent less than 1% of adult cancers. The average survival of patients in advanced stage soft tissue sarcoma is between 12 and 18 months. Several types of sarcoma exist characterized as simple or complex types. Simple sarcomas are characterized by specific dysregulated signaling pathways supporting the oncogenesis and a limited number of mutations, while complex tumors have no identified oncogenic driver in spite of numerous mutated genes.

The variety of STS tumor types represent a challenge for treatments. For this reason, the characterization of the diverse tumors and tumor environments that could be relevant for immunotherapy and more specifically checkpoint inhibitors appears critical. Checkpoint inhibitors, generally consisting of antibodies, target proteins such as CTLA-4 or PD-L1 at the surface of cancer cells, or their receptors (e.g. PD-1) at the T cells surface. Interaction between PD-1 and its ligand PD-L1 inhibit T cell activation and repress anti-tumor immunity. By preventing this from happening, checkpoint inhibitors are a promising strategy to overcome tumor generated immunosuppression. The efficacy of checkpoint inhibitors could be increased by a better understanding of the tumor and surrounding immune cells or tumor microenvironment.

A recent study published in Cancer Journal from the Pollack lab (Clinical Research Division) investigated the most frequent types of STS and characterized the tumor microenvironment surrounding these tumors with the aim to better adapt immunotherapy treatment per tumor type and individual.

To this end, the authors analyzed 81 patient samples representing five STS tumor types such as liposarcoma  (n=27) (including well differentiated/differentiated or WD/DD) and myeloid/round cell liposarcoma (MRCL) in addition to synovial sarcoma (SS, n=15), undifferentiated pleomorphic sarcomas (UPS, n=20) and leiomyosarcoma (LMS, n=19). Both UPS and LMS are complex tumor types while WD/DD and MRCL belong to simple class of STS. Using these tissues, the authors analyzed tumor phenotype and T cell infiltration/clonality by immunohistochemistry (IHC) and T cell receptor V beta (TCR-Vβ) sequencing, in addition to gene expression quantification in the tissues.

The expression of a set of 760 genes was analyzed. Almost half (48%) were expressed at significantly different levels between at least two different tumor types. Among these, genes with functions related to the antigen presentation were significantly higher in the UPS and LMS subtypes. Antigen presentation is critical for CD8+ T lymphocytes attraction and recognition of the tumor cells. Genes such as CD3D or CD8a, markers of T lymphocytes infiltration, interleukin 27 (CD127 or IL-7R), expressed on activated and antigen experienced T cells, were also highly expressed in UPS and LMS.  In addition, the LMS type was characterized by high gene expression related to tumor-associated macrophages presence in the tumor microenvironment. UPS tumor types were also characterized by higher expression of both PD-1 and PD-L1, confirmed by RNA quantification and IHC. Hence, both UPS and LMS, which belong to complex tumor types, are characterized by higher T cell immune response gene signature as well as higher PD-1 and PD-L1 checkpoint inhibitors expression. This suggests that tumors with inflammatory phenotypes might be better candidates for checkpoint inhibitors immunotherapy targeting.

Individual T cells can be identified through sequencing of the variable Vβ region of the TCR (T cell receptor). Such analysis provides information regarding the diversity and consequently propensity, of the T cell response and infiltration. Again, the complex UPS and LMS subtypes presented a higher diversity with more individual clones recruited at the tumor sites. Both PD-1/PD-L1 expression and histological subtypes correlated with the T cell infiltration. Cytokines secretion associated with T cell infiltration could explain the higher PD-1/PD-L1 expression.

These data suggest a clear correlation between tumor inflammation, T cell infiltration and checkpoint inhibitors expression, dependent on the tumor subtype. This study is the first to provide such analysis of the tumor microenvironment per individual subtypes and will allow for selection of optimal treatment according to the tumor type and phenotype. For instance, the UPS complex tumor type that comprise more mutations, and is characteristic of inflammation, represent a relevant candidate for checkpoint immunotherapy treatment, opposed to the WD/DD subtype presenting few cases of PD-L1 overexpression in spite of a high T cell infiltration. SS and MRCL subtypes had low T cell infiltration associated with low PD-1/PD-L1 expression making them uncertain candidates for checkpoint inhibitors targeting immunotherapy.


This study was funded by the Sarcoma Alliance for Research through Collaboration (SARC), the Sarcoma Foundation for America, the Gilman Sarcoma Foundation, the National Institute of Health.


SM Pollack, Q He, J Yearley, R Emerson, M Vignali, Y Zhang, M Redman, KK Baker, S. Cooper, B. Donahue, ET Loggers, LD Cranmer, MB Spraker, YD Seo, VG Pillarisetty, RW Ricciotti, BL Hoch, TK McClanahan, E Murphy, WM Blumenschein, SM Townson, S Benzeno, SR Riddell, RL Jones. T-Cell Infiltration and Clonality Correlate With Programmed Cell Death Protein 1 and Programmed Death-Ligand 1 Expression in Patients With Soft Tissue Sarcomas. 2017. Cancer. (epub ahead of print).