A practical nomogram and risk stratification system predicting the cancer‐specific survival for patients with early hepatocellular carcinoma

Bing Yan, Bingbing Su, Dousheng Bai, Jianjun Qian, Chi Zhang, Shengjie Jin, Guo‐Qing JiangCancer 2020 · DOI

TL;DR

This paper introduces a practical nomogram and risk stratification system to predict cancer-specific survival (CSS) in early-stage hepatocellular carcinoma (HCC) patients. By integrating key clinical and demographic factors, this tool aims to improve prognostic evaluation and guide personalized treatment strategies for clinicians.

What Problem Does This Paper Address?

Hepatocellular carcinoma (HCC) is a significant global health concern, ranking as the sixth most common cancer and fourth leading cause of cancer-related mortality worldwide. While many HCC patients are diagnosed at advanced stages with poor prognoses, an increasing number are being identified at earlier stages due to improved diagnostic techniques and screening. However, accurately predicting the survival of these early-stage HCC patients remains a challenge. Existing staging systems, like the AJCC TNM system, are sometimes insufficient for precise survival prediction. Therefore, there's a need for more accurate and clinically applicable tools to guide treatment decisions and patient management for this specific population.

Methods

The researchers developed a nomogram, a graphical calculating device, to predict the probability of cancer-specific survival (CSS) for patients with early HCC. They utilized a retrospective approach, extracting data from the Surveillance, Epidemiology, and End Results (SEER) database.

The core methodology involved:

  • Data Extraction and Cohort Splitting: Data from early HCC patients diagnosed between 2010 and 2015 was extracted from the SEER database. This dataset was then randomly split into a training cohort (70%) and a validation cohort (30%) to build and test the nomogram, respectively.
  • Factor Identification: Univariate and multivariate analyses were performed on the training cohort to identify significant prognostic factors for CSS. Factors with a p-value less than 0.05 were considered significant.
  • Nomogram Construction: A nomogram was constructed using the identified significant prognostic factors. This visual tool assigns points to each factor, which are then summed to estimate the probability of survival.
  • Performance Evaluation: The nomogram's performance was rigorously evaluated using several metrics:
  • Concordance Index (C-index): Measures the model's ability to discriminate between patients who will survive longer and those who will not.
  • Calibration Plots: Assess how well the predicted probabilities from the nomogram align with the observed survival rates.
  • Decision Curve Analysis (DCA): Evaluates the clinical utility and net benefit of the nomogram across various threshold probabilities.
  • Receiver Operating Characteristic (ROC) Curves: Used to assess the sensitivity and specificity of the nomogram.
  • Risk Stratification: Based on the total score generated by the nomogram, patients were categorized into three risk groups: low, intermediate, and high risk.
  • Survival Analysis: Kaplan-Meier curves and log-rank tests were employed to compare the CSS among these risk groups.

Datasets

The study utilized data from the Surveillance, Epidemiology, and End Results (SEER) database (1973-2015). Specifically, data for patients diagnosed with HCC between 2010 and 2015 was extracted. The dataset comprised 3620 eligible patients with early HCC, which was then divided into a training cohort (n=2536) and a validation cohort (n=1084). No new benchmarks were introduced; the study leverages existing, comprehensive population-based data.

Key Insights and Ideas

  • Beyond TNM: The study highlights that while the AJCC TNM staging system is a standard, it may not be sufficient for precise survival prediction in early HCC. The developed nomogram, incorporating a broader set of factors, demonstrates superior predictive performance.
  • Multifactorial Prognosis: Cancer-specific survival in early HCC is influenced by a complex interplay of factors, including not only tumor characteristics (grade, stage, size) but also patient demographics (age, race, marital status) and treatment received (surgery, chemotherapy).
  • Clinical Utility of Nomograms: Nomograms offer an intuitive and visually accessible way to integrate multiple prognostic factors, making complex statistical predictions more understandable and actionable for clinicians.
  • Risk Stratification for Targeted Management: The ability to stratify patients into distinct risk groups (low, intermediate, high) allows for more personalized treatment planning and closer monitoring of high-risk individuals.

Key Results

The developed nomogram demonstrated good predictive performance:

  • C-indexes: 0.755 in the training cohort and 0.737 in the validation cohort, indicating a good ability to discriminate between survival outcomes.
  • Calibration: Calibration plots showed good consistency between predicted and observed 3- and 5-year CSS.
  • Clinical Utility: DCA curves indicated excellent clinical utility for the nomogram.
  • AUCs: 3- and 5-year AUCs for ROC curves were around 0.78 in the training cohort and 0.77 in the validation cohort.
  • Risk Stratification Impact: Kaplan-Meier analysis revealed significant differences in CSS across the low, intermediate, and high-risk groups, with 3-year CSS rates of 81.2%, 45.6%, and 24.0% respectively, and 5-year CSS rates of 71.3%, 32.1%, and 14.4% respectively in the overall cohort.

Limitations and Future Work

The authors acknowledge several limitations:

  • Retrospective Design: The study is retrospective, which inherently has limitations in establishing causality.
  • SEER Database Reliance: The nomogram is based on data from the SEER database, which may not be fully representative of global patient populations.
  • Missing Data: The SEER database lacks certain important information, such as the specific cause of HCC, and some blood and inflammatory indicators, which could have further refined the model.
  • Internal Validation: While internal validation was performed, external validation on independent datasets is crucial to confirm the nomogram's generalizability.

Future work should focus on external validation of the nomogram and risk stratification system. Further research could also explore the differences in prognostic factors between early and advanced HCC, and investigate the impact of specific treatment modalities in more detail.

Summary updated

Abstract

We developed a practical nomogram in early HCC patients for predicting the CSS, and a risk stratification system follow arisen, which provided an applicable tool for clinical management.
BibTeX
@inproceedings{yan2020practical,
  author = {Bing Yan and Bingbing Su and Dousheng Bai and Jianjun Qian and Chi Zhang and Shengjie Jin and Guo‐Qing Jiang},
  title = {{A practical nomogram and risk stratification system predicting the cancer‐specific survival for patients with early hepatocellular carcinoma}},
  booktitle = {Cancer},
  year = {2020},
  doi = {10.1002/cam4.3613},
}
Figure