Neoadjuvant immune checkpoint blockade (ICB) is a promising treatment for melanoma and other cancer types, and has recently been shown to provide a modest survival benefit for patients with recurrent glioblastoma. To improve the treatment efficacy, researchers are looking for vulnerabilities in surgically removed glioblastoma tissues, but this has been difficult due to the vast differences within the tumor and between patients.

To address this challenge, researchers at Institute for Systems Biology (ISB) and their collaborators developed a new way to study tumors. The method builds mathematical models using machine learning-based image analysis and multiplex spatial protein profiling of microscopic compartments in the tumor.

The team used the approach to analyze and compare tumor tissues collected from 13 patients with recurrent glioblastoma and 23 patients with high-risk melanoma, with both sets of patients treated with neoadjuvant ICB. Using melanoma to guide the interpretation of glioblastoma analyses, they identified the proteins that correlate with tumor-killing T cells, tumor growth, and immune cell-cell interactions.

"This work reveals similarities shared between glioblastoma and melanoma, immunosuppressive factors that are unique to the glioblastoma microenvironment, and potential co-targets for enhancing the efficacy of neoadjuvant immune checkpoint blockade," said Dr. Yue Lu, co-lead author of the paper describing the research.

"This framework can be used to uncover pathophysiological and molecular features that determine the effectiveness of immunotherapies," added Dr. Alphonsus Ng, co-lead author of the paper.

The work was published today in Nature Communications, and is a collaborative project by ISB, UCLA and MD Anderson. Brain cancer represents one of the most challenging settings for achieving immunotherapy success. The fruitful collaboration between scientists and clinicians provides a tremendous opportunity for improving patient care and achieving an understanding of cancer immunotherapy at the deepest levels.

"We believe that the integrated biological, clinical and methodological insights derived from comparing two classes of tumors widely seen as at the opposite ends of the spectrum with respect to immunotherapy treatments should be of interest to broad scientific and clinical audiences," said ISB President Dr. Jim Heath, corresponding author of the paper.

The response of patients with recurrent glioblastoma multiforme to neoadjuvant immune checkpoint blockade has been challenging to interpret due to the inter-patient and intra-tumor heterogeneity. We report on a comparative analysis of tumor tissues collected from patients with recurrent glioblastoma and high-risk melanoma, both treated with neoadjuvant checkpoint blockade. We develop a framework that uses multiplex spatial protein profiling, machine learning-based image analysis, and data-driven computational models to investigate the pathophysiological and molecular factors within the tumor microenvironment that influence treatment response. Using melanoma to guide the interpretation of glioblastoma analyses, we interrogate the protein expression in microscopic compartments of tumors, and determine the correlates of cytotoxic CD8+ T cells, tumor growth, treatment response, and immune cell-cell interaction. This work reveals similarities shared between glioblastoma and melanoma, immunosuppressive factors that are unique to the glioblastoma microenvironment, and potential co-targets for enhancing the efficacy of neoadjuvant immune checkpoint blockade.

For mRNA quantification data, expression values were normalized using positive and negative controls and housekeeping genes and analyzed using the nSolver analysis software 4.0 (NanoString). Digital counts from barcodes corresponding to proteins were processed in three steps using Microsoft Excel (Redmond, WA). First, raw counts were normalized with ERCC spike-in controls to account for batch and system variation. Second, the normalized counts were subtracted with the appropriate IgG isotype control counts from each ROI to control for nonspecific antibodies; resulting counts that fall below zero were set to zero. Third, the resulting counts were normalized by the ultraviolet-light mask area to yield count density. For inter- and intra-tissue protein expression comparisons, an average count density was first calculated within each tissue compartment (i.e., immune-rich or immune-poor).

For more details go through: Archives in Cancer Research.

Journal accepts the article submission directly from the authors. If anyone has manuscripts or word documents, submit to cancer@imedpub.com or online portal.

Media Contact:

Allison Grey      

Managing Editor

Archives in Cancer Research