Deep learning predicts postsurgical recurrence of hepatocellular carcinoma from digital histopathologic images

Rikiya Yamashita,Atif Saleem,Daniel L Rubin,Jeanne Shen,Jin Long

doi:10.1038/s41598-021-81506-y

Abstract

Recurrence risk stratification of patients undergoing primary surgical resection for hepatocellular carcinoma (HCC) is an area of active investigation, and several staging systems have been proposed to optimize treatment strategies. However, as many as 70% of patients still experience tumor recurrence at 5 years post-surgery. We developed and validated a deep learning-based system (HCC-SurvNet) that provides risk scores for disease recurrence after primary resection, directly from hematoxylin and eosin-stained digital whole-slide images of formalin-fixed, paraffin embedded liver resections. Our model achieved concordance indices of 0.724 and 0.683 on the internal and external test cohorts, respectively, exceeding the performance of the standard Tumor-Node-Metastasis classification system. The model’s risk score stratified patients into low- and high-risk subgroups with statistically significant differences in their survival distributions, and was an independent risk factor for post-surgical recurrence in both test cohorts. Our results suggest that deep learning-based models can provide recurrence risk scores which may augment current patient stratification methods and help refine the clinical management of patients undergoing primary surgical resection for HCC.

Highlights

Recurrence risk stratification of patients undergoing primary surgical resection for hepatocellular carcinoma (HCC) is an area of active investigation, and several staging systems have been proposed to optimize treatment strategies
To stratify patients according to their expected outcome in order to optimize treatment strategies, several staging systems, such as the American Joint Committee on Cancer (AJCC)/International Union against Cancer (UICC) Tumor-Node-Metastasis (TNM)[3] and the Barcelona Clinic Liver Cancer (BCLC) systems[4], have been proposed and validated
We developed and independently validated a deep convolutional neural network for predicting risk scores for the recurrence-free interval (RFI) after curative-intent surgical resection for HCC, directly from digital whole-slide images (WSI) of hematoxylin and eosin (H&E)-stained, formalin-fixed, paraffin embedded (FFPE) primary liver resections

Summary

Introduction

Recurrence risk stratification of patients undergoing primary surgical resection for hepatocellular carcinoma (HCC) is an area of active investigation, and several staging systems have been proposed to optimize treatment strategies. We developed and validated a deep learning-based system (HCC-SurvNet) that provides risk scores for disease recurrence after primary resection, directly from hematoxylin and eosin-stained digital whole-slide images of formalin-fixed, paraffin embedded liver resections. Our results suggest that deep learning-based models can provide recurrence risk scores which may augment current patient stratification methods and help refine the clinical management of patients undergoing primary surgical resection for HCC. Mobadersany et al.[14] and Zhu et al.[15] applied convolutional neural networks, a type of deep learning network, to predict patient survival directly from histopathologic images of brain and lung cancers, respectively. We developed and independently validated a deep convolutional neural network for predicting risk scores for the recurrence-free interval (RFI) after curative-intent surgical resection for HCC, directly from digital whole-slide images (WSI) of hematoxylin and eosin (H&E)-stained, formalin-fixed, paraffin embedded (FFPE) primary liver resections. We present a fully automated approach to HCC recurrence risk prognostication on histopathologic images, which can be adopted for use in clinical settings to refine treatment and follow-up plans

Methods

Results

Conclusion