Integrative deep learning analysis improves colon adenocarcinoma patient stratification at risk for mortality.

BACKGROUND: Colorectal cancers are the fourth most diagnosed cancer and the second leading cancer in number of deaths. Many clinical variables, pathological features, and genomic signatures are associated with patient risk, but reliable patient stratification in the clinic remains a challenging task. Here we assess how image, clinical, and genomic features can be combined to predict risk. METHODS: We developed and evaluated integrative deep learning models combining formalin-fixed, paraffin-embedded (FFPE) whole slide images (WSIs), clinical variables, and mutation signatures to stratify colon adenocarcinoma (COAD) patients based on their risk of mortality. Our models were trained using a dataset of 108 patients from The Cancer Genome Atlas (TCGA), and were externally validated on newly generated dataset from Wayne State University (WSU) of 123 COAD patients and rectal adenocarcinoma (READ) patients in TCGA (N = 52). FINDINGS: We first observe that deep learning models trained on FFPE WSIs of TCGA-COAD separate high-risk (OS < 3 years, N = 38) and low-risk (OS > 5 years, N = 25) patients (AUC = 0.81 ± 0.08, 5 year survival p < 0.0001, 5 year relative risk = 1.83 ± 0.04) though such models are less effective at predicting overall survival (OS) for moderate-risk (3 years < OS < 5 years, N = 45) patients (5 year survival p-value = 0.5, 5 year relative risk = 1.05 ± 0.09). We find that our integrative models combining WSIs, clinical variables, and mutation signatures can improve patient stratification for moderate-risk patients (5 year survival p < 0.0001, 5 year relative risk = 1.87 ± 0.07). Our integrative model combining image and clinical variables is also effective on an independent pathology dataset (WSU-COAD, N = 123) generated by our team (5 year survival p < 0.0001, 5 year relative risk = 1.52 ± 0.08), and the TCGA-READ data (5 year survival p < 0.0001, 5 year relative risk = 1.18 ± 0.17). Our multicenter integrative image and clinical model trained on combined TCGA-COAD and WSU-COAD is effective in predicting risk on TCGA-READ (5 year survival p < 0.0001, 5 year relative risk = 1.82 ± 0.13) data. Pathologist review of image-based heatmaps suggests that nuclear size pleomorphism, intense cellularity, and abnormal structures are associated with high-risk, while low-risk regions have more regular and small cells. Quantitative analysis shows high cellularity, high ratios of tumor cells, large tumor nuclei, and low immune infiltration are indicators of high-risk tiles. INTERPRETATION: The improved stratification of colorectal cancer patients from our computational methods can be beneficial for treatment plans and enrollment of patients in clinical trials. FUNDING: This study was supported by the National Cancer Institutes (Grant No. R01CA230031 and P30CA034196). The funders had no roles in study design, data collection and analysis or preparation of the manuscript.

Clinicopathologic Characteristics and Outcome Descriptors of Metaplastic Breast Carcinoma.

CONTEXT.­: Metaplastic breast carcinoma is an aggressive form of breast cancer that accounts for 0.5% to 3% of all breast cancers. OBJECTIVE.­: To study the clinicopathologic characteristics and outcomes of this rare disease. DESIGN.­: Retrospective study of patients with a diagnosis of metaplastic breast carcinoma between 2000 and 2019. Hematoxylin-eosin-stained slides were reviewed and additional clinical data were obtained from electronic medical records. Univariable and multivariable Cox proportional hazard regression analyses were used to determine associations between overall survival and several clinicopathologic variables. RESULTS.­: Of the 125 patients with metaplastic breast carcinoma identified, only patients with high-grade disease (N = 115) were included in the data analysis. A total of 38 participants (33%) were white, 66 (57%) were African American, and 11 (10%) belonged to other ethnicities. The median age at diagnosis was 57 years. The median tumor size was 3 cm. Heterologous histology was seen in 30% of cases. Multivariable analyses showed that patients with a larger tumor size had worse overall survival (hazard ratio [HR], 1.25; 95% CI, 1.10-1.44; P < .001). Distant metastatic disease was also associated with worse overall survival on multivariable analysis (HR, 10.27; 95% CI, 2.03-55.54; P = .005). In addition to treatment with either partial or complete mastectomies, 84 patients (73%) received chemotherapy. Multivariable analyses showed that chemotherapy had no effect on overall survival (HR, 0.53; 95% CI, 0.09-6.05; P = .55). CONCLUSIONS.­: A larger tumor size and distant metastatic disease are associated with worse overall survival in patients with metaplastic breast carcinoma. Additional studies are needed to further characterize our findings.

Impact of positive cytology in uterine serous carcinoma: A reassessment.

OBJECTIVES: The aim of this study was to evaluate the prognostic value of peritoneal cytology status among other clinicopathological parameters in uterine serous carcinoma (USC). METHODS: A retrospective study of 148 patients diagnosed with uterine serous carcinoma from 1997 to 2016 at two academic medical centers in the Detroit metropolitan area was done. A central gynecologic pathologist reviewed all available slides and confirmed the histologic diagnosis of each case of USC. We assessed the prognostic impact of various clinicopathological parameters on overall survival (OS) and endometrial cancer-specific survival (ECSS). Those parameters included race, body mass index (BMI), stage at diagnosis, tumor size, lymphovascular invasion (LVSI), peritoneal cytology status, receipt of adjuvant treatment, and comorbidity count using the Charlson Comorbidity Index (CCI). We used Cox proportional hazards models and 95% confidence intervals for statistical analysis. RESULTS: Positive peritoneal cytology had a statistically significant effect on OS (HR: 2.09, 95% CI: [1.19, 3.68]) and on ECSS (HR: 2.02, 95% CI: [1.06 - 3.82]). LVSI had a statistically significant effect on both OS (HR: 2.27, 95% CI: [1.14, 4.53]) and ECSS (HR: 3.45, 95% CI: [1.49, 7.99]). Black or African American (AA) race was also found to have a significant effect on both OS (HR: 1.92, 95% CI: [1.07, 3.47]) and ECSS (HR: 2.01, 95% CI: [1.02, 3.98]). Other factors including BMI and tumor size > 1 cm did not show a statistically significant impact on OS or ECSS. CONCLUSIONS: Peritoneal washings with positive cytology and LVSI are important prognostic tools that may have a significant impact on overall survival in USC and can be used as independent negative prognosticators to help guide adjuvant treatment.