Search

In this study we use artificial intelligence (AI) to categorise endometrial biopsy whole slide images (WSI) from digital pathology as either "malignant", "other or benign" or "insufficient". An endometrial biopsy is a key step in diagnosis of endometrial cancer, biopsies are viewed and diagnosed by pathologists. Pathology is increasingly digitised, with slides viewed as images on screens rather than through the lens of a microscope. The availability of these images is driving automation via the application of AI. A model that classifies slides in the manner proposed would allow prioritisation of these slides for pathologist review and hence reduce time to diagnosis for patients with cancer. Previous studies using AI on endometrial biopsies have examined slightly different tasks, for example using images alongside genomic data to differentiate between cancer subtypes. We took 2909 slides with "malignant" and "other or benign" areas annotated by pathologists. A fully supervised convolutional neural network (CNN) model was trained to calculate the probability of a patch from the slide being "malignant" or "other or benign". Heatmaps of all the patches on each slide were then produced to show malignant areas. These heatmaps were used to train a slide classification model to give the final slide categorisation as either "malignant", "other or benign" or "insufficient". The final model was able to accurately classify 90% of all slides correctly and 97% of slides in the malignant class; this accuracy is good enough to allow prioritisation of pathologists' workload.

Abstract The tumour microenvironment is an important factor for colorectal cancer prognosis, affecting the patient's immune response. Immune checkpoints, which regulate the immune functions of lymphocytes, may provide prognostic power. This study aimed to investigate the prognostic value of the immune checkpoints TIM‐3, LAG‐3 and PD‐1 in patients with stage I–III colorectal cancer. Immunohistochemistry was employed to detect TIM‐3, LAG‐3, PD‐1 and PD‐L1 in 773 patients with stage I–III colorectal cancer. Immune checkpoint protein expression was assessed in tumour cells using the weighted histoscore, and in immune cells within the stroma using point counting. Scores were analysed for associations with survival and clinical factors. High tumoural LAG‐3 (hazard ratio [HR] 1.45 95% confidence interval [CI] 1.00–2.09, p = 0.049) and PD‐1 (HR 1.34 95% CI 1.00–1.78, p = 0.047) associated with poor survival, whereas high TIM‐3 (HR 0.60 95% CI 0.42–0.84, p = 0.003), LAG‐3 (HR 0.58 95% CI 0.40–0.87, p = 0.006) and PD‐1 (HR 0.65 95% CI 0.49–0.86, p = 0.002) on immune cells within the stroma associated with improved survival, while PD‐L1 in the tumour (p = 0.487) or the immune cells within the stroma (p = 0.298) was not associated with survival. Furthermore, immune cell LAG‐3 was independently associated with survival (p = 0.017). Checkpoint expression scores on stromal immune cells were combined into a Combined Immune Checkpoint Stromal Score (CICSS), where CICSS 3 denoted all high, CICSS 2 denoted any two high, and CICSS 1 denoted other combinations. CICSS 3 was associated with improved patient survival (HR 0.57 95% CI 0.42–0.78, p = 0.001). The results suggest that individual and combined high expression of TIM‐3, LAG‐3, and PD‐1 on stromal immune cells are associated with better colorectal cancer prognosis, suggesting there is added value to investigating multiple immune checkpoints simultaneously.