Casey M. Cosgrove

Training, Validating, and Testing Machine Learning Prediction Models for Endometrial Cancer Recurrence

PURPOSE Endometrial cancer (EC) is the most common gynecologic cancer in the United States with rising incidence and mortality. Despite optimal treatment, 15%-20% of all patients will recur. To better select patients for adjuvant therapy, it is important to accurately predict patients at risk for recurrence. Our objective was to train, validate, and test models of EC recurrence using lasso regression and other machine learning (ML) and deep learning (DL) analytics in a large, comprehensive data set. METHODS Data from patients with EC were downloaded from the Oncology Research Information Exchange Network database and stratified into low risk, The International Federation of Gynecology and Obstetrics (FIGO) grade 1 and 2, stage I (N = 329); high risk, or FIGO grade 3 or stages II, III, IV (N = 324); and nonendometrioid histology (N = 239) groups. Clinical, pathologic, genomic, and genetic data were used for the analysis. Genomic data included microRNA, long noncoding RNA, isoforms, and pseudogene expressions. Genetic variation included single-nucleotide variation (SNV) and copy-number variation (CNV). In the discovery phase, we selected variables informative for recurrence ( P < .05), using univariate analyses of variance. Then, we trained, validated, and tested multivariate models using selected variables and lasso regression, MATLAB (ML), and TensorFlow (DL). RESULTS Recurrence clinic models for low-risk, high-risk, and high-risk nonendometrioid histology had AUCs of 56%, 70%, and 65%, respectively. For training, we selected models with AUC >80%: five for the low-risk group, 20 models for the high-risk group, and 20 for the nonendometrioid group. The two best low-risk models included clinical data and CNVs. For the high-risk group, three of the five best-performing models included pseudogene expression. For the nonendometrioid group, pseudogene expression and SNV were overrepresented in the best models. CONCLUSION Prediction models of EC recurrence built with ML and DL analytics had better performance than models with clinical and pathologic data alone. Prospective validation is required to determine clinical utility.

A working group report from the 2024 National Cancer Institute / Gynecologic Cancer Steering Committee endometrial cancer clinical trials planning meeting: refining the approach to endometrial cancer in the immunotherapy era

Abstract Endometrial cancer is now the leading cause of gynecologic cancer death in the United States. Recognizing the urgent need to improve outcomes for patients diagnosed with endometrial cancer, the National Cancer Institute Gynecologic Cancer Steering Committee convened a clinical trials planning meeting, Refining the Approach to Endometrial Cancer in the Immunotherapy Era, on January 8 and 9, 2024. Multidisciplinary experts were charged with addressing critical challenges to optimize treatment of endometrial cancer in the new immunotherapy landscape. As part of the clinical trials planning meeting, working groups were assembled to address several important aspects of clinical trial design. Working group 1 focused on translational science and was tasked with reviewing the scientific literature for data on validated discriminants of response to immunotherapy to inform trial concept development by the therapy-focused groups. The working group established that molecular subtyping of endometrial cancer is now the standard approach for classifying endometrial tumors. Molecular subtyping for prognostic and predictive applications should be considered when assessing biomarkers as well as therapeutic targets. Additionally, strategies to improve immune response like incorporation of radiation as well as therapy sequencing considerations should continue to be explored. A major key observation from working group 1 was lack of validated discriminants for immunotherapy response beyond mismatch repair status, and tumor mutational burden and exploration of additional discriminants of response and resistance will be critical with the increasing use of immunotherapy in endometrial cancer.

Bridging the Gap from Bench to Bedside: A Call for In Vivo Preclinical Models to Advance Endometrial Cancer and Cervical Cancer Immuno-oncology Research

Abstract Advanced-stage endometrial and cervical cancers are associated with poor outcomes despite contemporary advances in surgical techniques and therapeutics. Recent clinical trial results have led to a shift in the treatment paradigm for both malignancies, in which immunotherapy is now incorporated as the standard of care up front for most patients with advanced endometrial and cervical cancers as the standard of care. Impressive response rates have been observed, but unfortunately, a subset of patients do not benefit from immunotherapy, and survival remains poor. Continued preclinical research and clinical trial development are crucial for our understanding of resistance mechanisms to immunotherapy and maximization of therapeutic efficacy. In this setting, syngeneic models are preferred over xenograft models as they allow for the evaluation of the tumor–immune interaction in an immunocompetent host, most closely mimicking the tumor–immune interaction in patients with cancer. Unfortunately, significant disparities exist about syngeneic models in gynecologic malignancy, in which queries from multiple large bioscience companies confirm no commercial availability of endometrial or cervical cancer syngeneic cell lines. Published data exist about the recent development of several endometrial and cervical cancer syngeneic cell lines, warranting further investigation. Closing the disparity gap for preclinical models in endometrial and cervical cancers will support physician scientists, basic and translational researchers, and clinical trialists who are dedicated to improving outcomes for our patients with advanced disease and poor prognosis.