Gender-specific AI model advances understanding of glioblastoma progression and treatment response

Gender-specific AI model advances understanding of glioblastoma progression and treatment response

Distinguishing on current imaging between disease progression and pseudoprogression in patients with glioblastoma is one of the most difficult clinical issues, according to Manmeet Ahluwalia, MD and Pallavi Tiwari, PhD.

Despite advances in the management of primary glioblastoma—chemotherapy, radiation therapy, and surgery—the disease continues to have a poor prognosis, with patients typically surviving only 15 to 18 months. Clinicians face many challenges when caring for the 15,000 Americans diagnosed with glioblastoma each year.

One of the most difficult clinical issues is distinguishing on current imaging between disease progression and pseudo-progression, particularly after treatment with temozolomide (Temodar), due to the inflammatory changes that occur in the brain after treatment with chemotherapy and radiotherapy. The benign treatment-related radiological effect and the “true” tumor recurrence are similar clinically and radiographically. Pseudo-progression occurs in about 40% of patients, most of them harboring O6-methylguanine-DNA methyl-transferase methylated tumors. The result is that a large number of patients with benign changes related to chemoradiotherapy often undergo additional imaging, and sometimes unnecessary and invasive surgery or intracranial biopsies.

Research developments are proving valuable in distinguishing the effects of radiation from tumor recurrence. Building on an earlier study by our group that suggested that sex differences should be considered as an influence on prognostic outcomes1 and a more recent study to identify signaling pathways that govern the biology and processing of sex-specific tumors2,, we are moving forward in taking the next steps to generate what we believe to be powerful and more accurate tools in glioblastoma patient risk stratification for personalized decision-making.

By exploring and extracting computational features that are not visually apparent to radiologists or clinicians, it is possible to map and distinguish between radiation-related changes and tumor progression. Funded by an NIH R01 grant (1R01CA264017-O1A1), our group is working on the development of an image-based recidivism risk classifier (IRRisC), which will use advanced radiomics and machine learning approaches to distinguish between effects radiation and tumor recurrence. , for men and women. Uniquely, unlike “black box” machine learning and deep learning approaches that have already been explored in the literature, IRRisC will take advantage of “handcrafted” image features that capture the heterogeneity of the underlying disease existing in GBM tumors, via measures of local gradient entropy (degree of disorder associated with aggressiveness), as well as a new class of biophysical strain and surface topology attributes―characteristics which capture the tumor microenvironment on routine MRI sequences (Gd-T1w, T2w, FLAIR).

So far, our results on the preliminary analysis have been very promising with nearly 90% accuracy in distinguishing histologically proven radiological necrosis from tumor recurrence, across a cohort of multi-institutional datasets. Our goal is to demonstrate the value of IRRisC as a decision aid (to complement neuroradiologists in decision making) on ​​a much larger cohort of histologically confirmed cases of radionecrosis and tumor recurrence via our multi-collaborative -institutional involving Case Western, Miami Cancer Institute, University of Wwashington-Madison, Northwestern University and Cleveland Clinic. The collaboration will also include GE Research, through which we will extend these diagnostic tools to be deployed on a cloud platform. The cloud platform will provide IRRisC with global reach and accessibility, both in clinics and hospitals in the United States and internationally. Additionally, the research has implications for immunotherapeutic therapies for glioblastoma, which can also cause inflammation in the brain that is not actually indicative of aggressive tumor growth.

As we lead the charge of developing these improved diagnostic tools in collaboration with GE Research, we hope this five-year initiative will allow us to move into clinical trials and ultimately improve the quality of life of patients with GBM.

REFERENCES:

1. Gender-specific probabilistic atlases of glioblastoma reveal the impact of tumor location on progression-free survival. https://doi.org/10.1093/neuonc/noz175.333. Posted November 11, 2019.

2. Sexually dimorphic radiogenomic models identify distinct imaging and biological pathways that are prognostic for overall survival in glioblastoma. https://doi.org/10.1093/neuonc/noaa231. Posted February 25, 2021.

#Genderspecific #model #advances #understanding #glioblastoma #progression #treatment #response

Leave a Comment

Your email address will not be published.