Data presented at the annual meeting of the American Association for Cancer Research (AACR), held April 14-19, 2023 at the Orange County Convention Center in Orlando, Florida, shows that a novel, investigational deep learning model requiring one histopathologic slide accurately predicted the risk of distant recurrence in patients with endometrial cancer.
Endometrial cancer is the most common type of uterine cancer.
“While patients with early-stage uterine cancer have an approximately 95% five-year survival rate*, those who develop a distant recurrence have a very poor survival outcome,” noted Sarah Fremond, MSc, a Ph.D. candidate in computational pathology and deep learning for endometrial cancer at Leiden University Medical Center (LUMC) in the Netherlands.
Distant recurrence is associated with a 5-year overall survival of 10-20%. This risk may be reduced by adjuvant systemic therapy. Hence, correctly identifying patients at high risk and low risk of distant recurrence is crucial for personalized adjuvant treatment recommendations and for reducing unnecessary morbidity from toxic treatments, Fremond explained.
Current methods of risk stratification, which rely on pathologists assessing histopathologic images, are limited by significant inter-observer variability between pathologists. While molecular testing of tumor tissue is of additional value, this method comes with high costs and the need for complex infrastructure.
Fremond added that the increasing number of prognostic variables has made it difficult to combine the relevant factors into a single risk score.
Deep learning models
To overcome these challenges, Fremond and colleagues examined the potential of deep learning models (a form of artificial intelligence) which can predict a patients’ risk of distant recurrence by identifying relevant visual features using digitized histopathologic images and whole slide images (WSIs) at different resolutions without prior assumptions.
The study was a collaborative effort by the AIRMEC Consortium, which involved the Department of Pathology at Leiden University Medical Center, the TransPORTEC Consortium, and the Department of Pathology and Molecular Pathology at the University of Zurich.
“Deep learning is a powerful computer-aided predictive technology that has entered the field of pathology because it can be trained to read complex visual information from tumor slides after digitization,” Fremond said.
“In this study, we aimed to evaluate whether a deep learning model could be trained to predict risk of distant recurrence in patients with endometrial cancer using routine histopathological slides as a cost- effective input.”
To develop the model, Fremond and colleagues utilized long-term follow-up data from patients enrolled in the PORTEC-1/-2/-3 randomized clinical trials and patients in three separate clinical cohorts, amounting to 1,761 patients with endometrial cancer who had not received prior adjuvant chemotherapy.
The researchers randomly sampled 20% as a held-out internal test set (N=353 with 62 events; 8.45 year median follow-up) and used one representative histopathological slide image of the tumor to perform a 5-fold cross-validation to train and optimize the model (N=1408).
“This means that the model was exposed many times to the histopathological image and to the information regarding the time to distant recurrence in each patient until the model started to recognize visual features that were predictive of distant recurrence,” Fremond explained.
The model’s performance of correctly ranking patients by predicted risk scores and true time to DR was measured with the concordance-index demonstrating a concordance-index of 0.764 [95%CI 0.754-0.773] on 5-fold cross validation and 0.757 on the test set, as compared to 0.704 [95%CI 0.662-0.746] with a Cox’ Proportional Hazards (CPH) model fitted on histopathological variables (histotype, grade, lymphovascular space invasion, stage).
To assess its performance and generalizability, the resulting model was then tested on the previously unseen dataset of 353 patients whose data were not used to train the model. The model accurately identified 89 of these patients as having a low risk of distant recurrence, 175 an intermediate risk, and 89 a high risk.
Consistent with patient outcome
These predictions were consistent with the patients’ outcomes: 3.37% of patients (n=3 categorized as low-risk experienced a distant recurrence, as compared with 15.43% (n=27) and 36% (n=32) of those categorized as intermediate-risk and high-risk, respectively.
Fremond noted that the results outperformed pathologist-identified features, such as tumor type, grade, and molecular class, typically used to assign risk groups.
“We are currently working on improving performance by integrating clinical variables that cannot be read in the histopathologic slides,” she added.
“This deep learning model is capable of predicting the risk of distant recurrence for patients with endometrial cancer using one digitized histopathological slide as a cost-effective input,” said Fremond.
“Although additional external validation is needed, the performance of this model serves as an important proof of concept that deep learning models have the potential to optimize clinical care for patients with endometrial cancer,” she said.
Limitations of the study include its retrospective design and the lack of large-scale external validation to assess the generalizability of the model across different populations.
Additionally, it is currently unclear how differences in tissue scanning and slide preparation protocols might impact the performance of the model.
The study was supported by the Hanarth Foundation.
Note: * Statistic from the the National Cancer Institute’s Surveillance, Epidemiology, and End Results Program (SEER).
Post Operative Radiation Therapy in Endometrial Carcinoma 1 (PORTEC-1) trial
External-Beam Radiation Therapy Compared With Vaginal Brachytherapy After Surgery for Stage I Endometrial Cancer (PORTEC-2) – NCT00376844
Randomized Trial of Radiation Therapy With or Without Chemotherapy for Endometrial Cancer (PORTEC-3) – NCT00411138
Fremond S, Andani S, Barkey Wolf J, Ørtoft G, Høgdall E, Dijkstra J, Jobsen JJ, Jürgenliemk-Schulz IM, Lutgens LCHW, Powell ME, Singh N, et al. Deep learning risk prediction model of distant recurrence from H&E endometrial cancer slides. In: Proceedings of the 114th Annual Meeting of the American Association for Cancer Research; 2023 April 14-19; Orlando, FL. Philadelphia (PA): AACR; 2023. Abstract nr 5695.
 Cancer Stat Facts: Uterine Cancer (NCI: SEER). Online. Last accessed on April 18, 2023.
Featured image by Tianyi Ma on Unsplash