Serum Fusion Transcripts to Assess the Risk of Hepatocellular Carcinoma and the Impact of Cancer Treatment through Machine Learning.

Journal: The American journal of pathology
Published Date:

Abstract

Hepatocellular carcinoma (HCC) is one of the most fatal malignancies. Early diagnosis of HCC is crucial in reducing the risk for mortality. This study analyzed a panel of nine fusion transcripts in serum samples from 61 patients with HCC and 75 patients with non-HCC conditions, using TaqMan real-time quantitative RT-PCR. Seven of the nine fusions frequently detected in patients with HCC included: MAN2A1-FER (100%), SLC45A2-AMACR (62.3%), ZMPSTE24-ZMYM4 (62.3%), PTEN-NOLC1 (57.4%), CCNH-C5orf30 (55.7%), STAMBPL1-FAS (26.2%), and PCMTD1-SNTG1 (16.4%). Machine-learning models were constructed based on serum fusion-gene levels to predict HCC in the training cohort, using the leave-one-out cross-validation approach. One machine-learning model, called the four fusion genes logistic regression model (MAN2A1-FER≤40, CCNH-C5orf30≤38, SLC45A2-AMACR≤41, and PTEN-NOLC1≤40), produced accuracies of 91.5% and 83.3% in the training and testing cohorts, respectively. When serum α-fetal protein level was incorporated into the machine-learning model, a two fusion gene (MAN2A1-FER≤40, CCNH-C5orf30≤38) + α-fetal protein logistic regression model was found to generate an accuracy of 94.8% in the training cohort. The same model resulted in 95% accuracy in both the testing and combined cohorts. Cancer treatment was associated with reduced levels of most of the serum fusion transcripts. Serum fusion-gene machine-learning models may serve as important tools in screening for HCC and in monitoring the impact of HCC treatment.

Authors

  • Yan-Ping Yu
    Department of Pathology, University of Pittsburgh School of Medicine, Pittsburgh, Pennsylvania; High Throughput Genome Center, University of Pittsburgh School of Medicine, Pittsburgh, Pennsylvania; Pittsburgh Liver Research Center, University of Pittsburgh School of Medicine, Pittsburgh, Pennsylvania.
  • Silvia Liu
    Department of Pathology, University of Pittsburgh School of Medicine, Pittsburgh, Pennsylvania; High Throughput Genome Center, University of Pittsburgh School of Medicine, Pittsburgh, Pennsylvania; Pittsburgh Liver Research Center, University of Pittsburgh School of Medicine, Pittsburgh, Pennsylvania.
  • David Geller
    Pittsburgh Liver Research Center, University of Pittsburgh School of Medicine, Pittsburgh, Pennsylvania; Department of Surgery, University of Pittsburgh School of Medicine, Pittsburgh, Pennsylvania.
  • Jian-Hua Luo
    Department of Pathology, University of Pittsburgh School of Medicine, Pittsburgh, Pennsylvania; High Throughput Genome Center, University of Pittsburgh School of Medicine, Pittsburgh, Pennsylvania; Pittsburgh Liver Research Center, University of Pittsburgh School of Medicine, Pittsburgh, Pennsylvania. Electronic address: luoj@upmc.edu.