A collaborative workflow between pathologists and deep learning for the evaluation of tumour cellularity in lung adenocarcinomaAims: The reporting of tumour cellularity in cancer samples has become a mandatory task for pathologists. However, the estimation of tumour cellularity is often inaccurate. Therefore, we propose a collaborative workflow between pathologists and artificial intelligence (AI) models to evaluate tumour cellularity in lung cancer samples and propose a protocol to apply it to routine practice. Methods and results: We developed a quantitative model of lung adenocarcinoma that was validated and tested on 50 cases, and a collaborative workflow where pathologists could access the AI results and adjust their original tumour cellularity scores (adjusted-score) that we tested on 151 cases. The adjusted-score was validated by comparing them with a ground truth established by manual annotation of haematoxylin and eosin slides with reference to immunostains with thyroid transcription factor-1 and napsin A. For training, validation, testing the AI and testing the collaborative workflow, we used 40, 10, 50 and 151 whole slide images of lung adenocarcinoma, respectively. The sensitivity and specificity of tumour segmentation were 97 and 87%, respectively, and the accuracy of nuclei recognition was 99%. One pathologist's visually estimated scores were compared to the adjusted-score, and the pathologist's scores were altered in 87% of cases. Comparison with the ground truth revealed that the adjusted-score was more precise than the pathologists' scores (P < 0.05). Conclusion:We proposed a collaborative workflow between AI and pathologists as a model to improve daily practice and enhance the prediction of tumour cellularity for genetic tests.

More than half of patients with solid tumor malignancies undergo treatment with radiotherapy (RT). In addition to causing direct tumor cell death, RT results in release of tumor antigen and damage associated molecular patterns that elicit a CD8 T cell and IFN-g dependent anti-tumor immune response. To further bolster this response, RT has been combined with immune checkpoint inhibitors such as anti-PD-1 and anti-CTLA-4. Despite a modest increase in survival when these therapies are used alone or together in conjunction with RT, many patients fail to respond to treatment, suggesting other mechanisms of immune suppression exist in the tumor microenvironment (TME). Using syngeneic B16 F10 melanoma and C38 colorectal adenocarcinoma models, we have studied the role of the inhibitory receptor NKG2A. We have observed that NKG2A is expressed only on tumor infiltrating lymphocytes (TILs) and expression of this receptor does not change with RT. Furthermore, RT increases intratumoral expression of the ligand for NKG2A, Qa-1b, as does IFN-g stimulation in vitro. Blockade of NKG2A alone through use of B16 F10 cells lacking Qa-1b expression or blocking antibodies did not significantly increase survival of mice treated with RT. Further analysis revealed that among CD8 TILs, only a minority of cells express NKG2A alone, with most TILs co-expressing NKG2A and PD-1 or PD-1 alone. Thus, we combined anti-NKG2A and anti-PD-1 blockade and observed increased survival in mice treated with RT, whereas either therapy alone was ineffective. These results suggest that NKG2A blockade could be combined with RT and existing immunotherapies clinically to improve patient response. Supported by awards T32AI007285 from the NIAID and R01CA028332 from the NCI.

Owing to the high demand for molecular testing, the reporting of tumor cellularity in cancer samples has become a mandatory task for pathologists. However, the pathological estimation of tumor cellularity is often inaccurate. We developed a collaborative workflow between pathologists and artificial intelligence (AI) models to evaluate tumor cellularity in lung cancer samples and prospectively applied it to routine practice. We also developed a quantitative model that we validated and tested on retrospectively analyzed cases and ran the model prospectively in a collaborative workflow where pathologists could access the AI results and apply adjustments (Adjusted-Score). The Adjusted-Scores were validated by comparing them with the ground truth established by manual annotation of hematoxylin-eosin slides with reference to immunostains with thyroid transcription factor-1 and napsin A. For training, validation, retrospective testing, and prospective application of the model, we used 40, 10, 50, and 151 whole slide images, respectively. The sensitivity and specificity of tumor segmentation were 97% and 87%, and the accuracy of nuclei recognition was 99%. Pathologists altered the initial scores in 87% of the cases after referring to the AI results and found that the scores became more precise after collaborating with AI. For validation of Adjusted-Score, we found the Adjusted-Score was significantly closer to the ground truth than non-AI-aided estimates (p<0.05). Thus, an AI-based model was successfully implemented into the routine practice of pathological investigations. The proposed model for tumor cell counting efficiently supported the pathologists to improve the prediction of tumor cellularity for genetic tests.

No abstract

No abstract