NEW YORK – In an effort to better understand why cervical squamous cell carcinoma (CSCC) often shows only a limited response to immune checkpoint blockade immunotherapy, a team from China and the US has applied a multiomic approach to evaluate the gene expression and cellular heterogeneity features found in the common cervical cancer.
"Cellular heterogeneity and plasticity are fundamental in tumor progression and tumor immune microenvironment (TIME) regulation," the authors explained in a study published in Nature Medicine on Monday, adding that "much remains to be determined regarding [cellular states'] functional impact on tumor pathophysiology, interaction with the TIME, and patient outcomes, as well as the mechanisms underlying their formation and plasticity."
With that in mind, the investigators used single-cell RNA sequencing to profile nearly 163,900 individual cells from 14 treatment-naïve tumor samples, including half a dozen CSCC tumors with immune infiltrated phenotypes and another eight CSCC tumors marked by immune-excluded or immune desert features, as well as three normal cervical samples from cancer-free individuals.
With expression patterns found in the cells, they identified 14 clusters of major cell types and 50 minor subtype clusters, while highlighting eight "meta-programs" (MPs) representing within-tumor heterogeneity and cell state diversity.
"To better characterize the heterogeneity and plasticity of malignant cells in CSCC, we classified each tumor cell into distinct MP-related subtypes based on the expression score of their corresponding signature genes," the authors explained.
By bringing in mass spectrometry-based spatial proteomic data for 60 tumor regions of interest and Stereo-seq-based spatial transcriptomic profiles for 15 CSCC samples — together with bulk transcriptomic data for 94 other CSCC samples assessed at the same centers and 245 samples previously profiled for the Cancer Genome Atlas project — the investigators went on to consider meta-program-based cell state relationships to the tumor immune microenvironment.
In particular, they pointed to three MPs with distinct TIME features that appeared to track with different squamous differentiation stages: the epithelial-cytokeratin-like MP6 program, the epithelial-immune (Epi-Imm) program MP7, and the epithelial senescence (Epi-Sen) program MP8.
Along with more detailed analyses of the immune cell, CSCC MP state, and epigenetic interactions and dynamics, the team turned to patient organoids and models to explore the consequences of inhibiting or removing a fatty acid binding protein-coding gene called FABP5 that appeared to have a prominent role in the epithelial-cytokeratin MP6 state tumors, which tended to contain features linked to advanced CSCC stage and poor outcomes.
From their findings, the authors suggested that "[b]idirectional interactions between epithelial-cytokeratin malignant cells and immunosuppressive cancer-associated fibroblasts form an immune exclusionary microenvironment through transforming growth factor beta pathway signaling mediated by FABP5."
On the other hand, the team saw signs that the MP7 Epi-Imm program tended to coincide with enhanced response to immune checkpoint immunotherapy. When they did bulk RNA-seq on samples from 21 participants in a clinical trial of neoadjuvant platinum-based chemotherapy followed by the anti-PD-1 treatment camrelizumab in locally advanced cervical cancer, for example, the authors saw potential ties between MP6/7 state transitions and immunotherapy response — a pattern similar to that reported in head and neck squamous cell carcinoma (HNSCC) in the past.
Nevertheless, they noted that "further investigations are needed to generalize our findings to other squamous malignancies and determine whether the observed heterogeneity is dependent on [human papillomavirus] infection, which is common in CSCC and HNSCC."