MD Anderson Research Highlights for November 3, 2022

Featuring a new target to improve immunotherapy responses, tools for predicting treatment response and risk genes, and combination therapy to improve ALL outcomes

The University of Texas MD Anderson Cancer Center’s Research Highlights provides a glimpse into recent basic, translational and clinical cancer research from MD Anderson experts. Current advances include new insights into a known target gene that may improve immunotherapy responses, a novel 3D organoid model to predict treatment response, combination treatment with blinatumomab and chemotherapy for acute lymphocytic leukemia, the benefits of using genomic sequencing to match treatment for patients with ultra-rare sarcomas, a new statistical method to better identify genes linked with specific diseases, a standardized approach to include patient-reported outcomes for trials involving adolescents and young adults, and the identification of two new subtypes of ovarian cancer that may guide treatments.

New role for an established target gene could improve immunotherapy responses
Immune checkpoint inhibitors (ICIs) can result in lasting responses for many patients but some eventually develop resistance, making combination therapy necessary to improve efficacy. To uncover the underlying mechanisms of treatment resistance and to identify new therapeutic strategies, researchers led by Zhicheng Zhou, Ph.D., Mei-Ju May Chen, Ph.D., Yikai Luo,  and Han Liang, Ph.D., performed an integrated analysis of several immune-oncology targets in patients with melanoma who received anti-PD-1 treatment. The SIRPA gene is a known inhibitory immune regulator in macrophages, but the researchers discovered a new role for tumor-intrinsic SIRPA in melanoma cells. Higher SIRPA expression in tumor cells was associated with better patient responses to ICIs. Lab models without SIRPA showed no response to ICIs, while SIRPA overexpression significantly improved treatment response, highlighting the distinct roles of SIRPA in tumor cells and the tumor microenvironment. The findings suggest potential opportunities to improve immunotherapy responses through targeting SIRPA in a more specific way. Learn more in Cancer Cell

Novel patient-derived organoid model platform helps predict drug treatment response in pancreatic cancer
Pancreatic cancer has unpredictable responses to chemotherapy, so personalized treatment options are needed to improve outcomes. However, tumor tissue from patients is very limited, making it difficult to study and test patient tumors. Unlike current lab models that do not completely mimic human tumors, patient-derived organoid (PDO) models are 3D structures representative of pancreatic tumors and require very little patient tumor tissue to generate personalized, patient-matched drug sensitivity profiles. To test and quantify very small numbers of PDOs for sensitivity to various drugs, Michael Kim, M.D., and colleagues developed an ex vivo organoid-based platform with image-based quantification of cell death. This approach allows researchers to test PDO responses to individual drug treatments in a high-throughput, more efficient manner. This model also identified a novel predictive biomarker (α-SMA/CK-19 ratios) associated with patient outcomes, highlighting its potential to aid in therapy selection. Learn more in The Journal of Clinical Investigation.

Front-line chemo with blinatumomab improves long-term survival for patients with ALL
Studies have shown minimal residual disease (MRD) negativity is associated with superior outcomes in patients with acute lymphocytic leukemia (ALL). Blinatumomab, a bispecific CD3/CD19 T-cell-engaging monoclonal antibody, has demonstrated higher response rates and superior survival rates than chemotherapy. Additionally, it is very effective at producing MRD negativity in patients with residual disease after chemotherapy. Researchers led by Elias Jabbour, M.D., and Hagop Kantarjian, M.D., evaluated the use of blinatumomab in combination with intensive chemotherapy in adults with newly diagnosed Ph-negative B-cell ALL. The single-arm, Phase II study enrolled 38 patients. After 37 months, 30 patients (79%) survived and 28 (74%) were in continuous first response. These durable responses, including in patients with high-risk cytomolecular features, warrant a randomized study of this treatment regardless of MRD status. Learn more in Lancet Haematology.

Novel integrated statistical method improves predictive value in identifying risk genes
Genome-wide association studies (GWASs) are helpful in highlighting genes linked with specific traits and diseases, but their use is limited to genes with higher expression patterns. This underscores a need for an alternate method to capture moderate to low-expression genes that play important roles in determining traits. Researchers led by Chong Wu, Ph.D., developed a novel method to increase the statistical power and predictive value of GWASs. This model, the Summary-level Unified Method for Modeling Integrated Transcriptome (SUMMIT), allows for more accurate identification and prediction of risk genes with low expression heritability compared to current available methods. Using SUMMIT, they identified 11 genes associated with COVID-19 severity, showcasing its predictive value. Learn more in Nature Communications.

Genomic sequencing for ultra-rare sarcomas in Phase I trials helps identify patients likely to benefit
Recently published guidelines for recognizing ultra-rare sarcomas (URSs) defined them as sarcomas with an incidence of less than 1 per 1,000,000. Despite the rarity of each type, when combined, URSs make up 20% of all bone and soft tissue sarcomas. In a review of records from MD Anderson patient on Phase I trials over eight years, Vivek Subbiah, M.D., and colleagues identified that 106 of 587 (18.1%) sarcomas would have been classified as URSs. Of these, 33 patients were on molecularly matched treatments, meaning that genomic sequencing was done to help match the patients to the trial. The response rate among the treatment-matched patients (24%) was roughly triple that of unmatched patients (8.2%), and the clinical benefit rate increased from 26% to 36.4%. Though this data is limited, it points to the need for genomic sequencing in URSs to better match patients with potentially beneficial Phase I trials. Learn more in Clinical Cancer Research.  

Researchers identify patient-reported outcomes critical to adolescent and young adult cancer clinical trials
Adolescents and young adults face unique challenges with a cancer diagnosis, and understanding the impact on their quality of life and survival is critical. Non-consensus and limited data collection of patient-reported outcomes (PROs) in adolescent and young adult (AYA) clinical trials has led to slow progress in improving patient outcomes. Michael Roth, M.D., assembled the National Cancer Institute’s Clinical Trials Network AYA PRO Task Force to fill these gaps. The task force developed recommendations for a standardized approach to assessing health-related quality of life and treatment-related toxicity in AYA cancer clinical trials using previously validated measures. Eight core principles guided the development of the AYA PRO battery, including its relevance to AYAs, availability in multiple languages, and applicability to different cancer types and treatments. This consensus from the Task force already has increased inclusion of PROs in AYA trials, highlighting the importance of quality of life during and after cancer treatment. Learn more in the Journal of the National Cancer Institute.

Study identifies two subtypes of high-grade serous ovarian cancers
Despite looking similar under a microscope, high-grade serous ovarian cancers (HGSOCs) exhibit noticeable differences. In a study led by Anil Sood, M.D., and Katherine Foster, researchers developed a classification system for HGSOCs and determined two specific subtypes with distinct clinical and molecular features. The study included 112 women with advanced-stage ovarian cancer who received laparoscopic review of disease burden before treatment. Patients with type 2 tumors were more likely to have a lower predictive index value score and therefore to receive primary tumor reductive surgery. Analysis of tumor samples found that the type 1 tumors were enriched in PI3K/AKT/mTOR and Hedgehog signaling, and type 2 tumors had enrichment of pathways related to MYC signaling and cell-cycle progression. The findings may have implications for directing patients to surgery or chemotherapy, for predicting outcomes and for developing personalized therapeutic approaches. Read more in JAMA Network Open.

In case you missed it
Read below to catch up on recent MD Anderson press releases.

Organoid cells are enumerated by digital software using 3D confocal images of an organoid generated from a pancreatic cancer patient. Image courtesy of Michael Kim, Ph.D.