Early detection of pancreatic cancer by comprehensive serum miRNA sequencing with automated machine learning

Published in Cancer and Cell & Molecular Biology
Early detection of pancreatic cancer by comprehensive serum miRNA sequencing with automated machine learning
Like

Share this post

Choose a social network to share with, or copy the URL to share elsewhere

This is a representation of how your post may appear on social media. The actual post will vary between social networks

Pancreatic cancer is often diagnosed at advanced stages, and early-stage diagnosis of pancreatic cancer is difficult because of nonspecific symptoms and lack of available biomarkers. We performed comprehensive serum miRNA sequencing of 212 pancreatic cancer patient samples from 14 hospitals and 213 non-cancerous healthy control samples. We randomly classified the pancreatic cancer and control samples into the two cohorts: a training cohort (N = 185) and a validation cohort (N = 240). We created ensemble models that combined automated machine learning with 100 highly expressed miRNAs and their combination with CA19-9 and validated the performance of the models in the independent validation cohort. The diagnostic model with the combination of the 100 highly expressed miRNAs and CA19-9 could discriminate pancreatic cancer from non-cancer healthy control with high accuracy (area under the curve (AUC), 0.99; sensitivity, 90%; specificity, 98%). We validated high diagnostic accuracy in an independent asymptomatic early-stage (stage 0-I) pancreatic cancer cohort (AUC:0.97; sensitivity, 67%; specificity, 98%). We demonstrate that the 100 highly expressed miRNAs and their combination with CA19-9 could be biomarkers for the specific and early detection of pancreatic cancer.

Our ultimate objective is to spread this early detection technology for pancreatic cancer around the world as a next-generation sequencing (NGS) testing.

 

However, in general, several challenges remain for stable and sustainable operations of NGS measurement systems. For example, miRNA profiles tend to fluctuate depending on the facility where the sample is collected. One of the reasons is the differences in hemolysis levels1. Additionally, batch-to-batch difference in NGS measurements also tends to affect miRNA profiles because library preparation for NGS measurements requires multiple reaction steps1. These differences in miRNA profiles could lead to low accuracy or false determination in the discrimination results. In order to advance practical application, many challenges remain to be resolved, including those mentioned above. In this study, ARKRAY, Inc. participated in sample collection and quality control of NGS measurement, and with the tremendous cooperation of the doctors at each facility, we achieved stable sample collection and data acquisition. In order to make NGS testing more widely used in the world, we believe that further improvements such as higher discrimination accuracy, lower testing costs, and enhanced sustainability are required.

 

As one of the means to achieve these, ARKRAY, Inc. is currently developing its in-house library preparation kit. We have evaluated the pancreatic cancer discrimination performance of this in-house kit under development and report the results here. Validation data were obtained using this in-house kit and the Ion GeneStudio S5 system and evaluated with the pancreatic cancer discrimination model constructed in this study. As summarized in the figures and table below, the results confirmed that the discrimination accuracy was comparable to that shown in this study.

 

Figure 1. ROC curve for the performance of (A) miRNA model and (B) miRNA+CA19-9 model in the validation cohort using the in-house library preparation kit.

 

Table 1. The performance of miRNA model and miRNA+A19-9 model to discriminate pancreatic cancer patients from healthy controls in the validation cohort using commercial or in-house library preparation kit.

 

We will continue research and development to further improve the performance for practical application and to implement NGS testing for early detection of pancreatic cancer in society.

 

1 Suzuki K, Igata H, Abe M, Yamamoto Y. small RNA based cancer classification project. Multiple cancer type classification by small RNA expression profiles with plasma samples from multiple facilities. Cancer Sci. 2022; 113: 2144–2166.

Please sign in or register for FREE

If you are a registered user on Research Communities by Springer Nature, please sign in

Follow the Topic

Pancreatic Cancer
Life Sciences > Biological Sciences > Cancer Biology > Cancers > Gastrointestinal Cancer > Pancreatic Cancer
miRNAs
Life Sciences > Biological Sciences > Molecular Biology > Non-coding RNAs > miRNAs

Related Collections

With collections, you can get published faster and increase your visibility.

Digital Imaging

BJC’s Digital Imaging series is open to receiving submissions assessing:
  • State-of-the-art in digital imaging technology and computational output;
  • Radiomics;
  • Algorithms and approaches designed to predict therapeutic response and enable “personalized medicine”;
  • Interrogation of the immune microenvironment and implications for immunotherapy selection;
  • Challenges of implementation in clinical practice;
  • Controversial applications of digital imaging

Publishing Model: Hybrid

Deadline: Ongoing