A neotype nomogram based on accessible clinical indicators to predict the overall survival of esophageal cancer patients underwent radical resection

Xinye Li, Jinming Xu, Linhai Zhu, Sijia Yang, Yu Li, Wang Lv, Jian HuResearch Square 2020 · DOI

TL;DR

Researchers have developed a "neotype" nomogram—a visual predictive model—that uses six easily accessible clinical indicators to forecast the overall survival of esophageal cancer patients following radical surgery. This model significantly outperforms the traditional AJCC staging system, providing a more nuanced tool for personalized prognosis and treatment guidance.

What Problem Does This Paper Address?

Esophageal cancer (EC) remains one of the most lethal malignancies worldwide, particularly in China. Currently, the American Joint Committee on Cancer (AJCC) TNM staging system is the "gold standard" for prognosis. However, it relies heavily on anatomical data (tumor size, lymph node involvement, and metastasis) and often fails to account for the biological diversity of individual patients.

Clinicians frequently observe that patients with the same TNM stage can have vastly different outcomes. This gap suggests that other factors—such as a patient’s nutritional status or systemic inflammation—play a critical role in survival. The authors sought to bridge this gap by creating a comprehensive, easy-to-use nomogram that integrates traditional staging with routine laboratory markers.

Methods

The researchers conducted a retrospective study of 311 EC patients who underwent radical esophagectomy. They split the data into a primary cohort (215 patients) to build the model and a validation cohort (96 patients) to test it.

To identify the most impactful variables, they used Cox proportional hazards regression analysis. Unlike standard models that might only look at one or two factors, this study evaluated a wide array of demographic and clinicopathological data. The final nomogram was constructed using six independent prognostic factors. The model's performance was then rigorously tested using:

  • C-index (Concordance Index): To measure predictive accuracy.
  • Calibration Curves: To see how well predicted survival matched actual survival.
  • ROC Curves (Receiver Operating Characteristic): To compare the nomogram’s efficacy against traditional AJCC staging.

Datasets

The study utilized a dataset of 311 patients treated at The First Affiliated Hospital of Zhejiang University School of Medicine between 2008 and 2013.

  • Inclusion criteria: Confirmed EC diagnosis, no prior anti-tumor therapy, and complete clinical data.
  • Exclusion criteria: Non-squamous/adenocarcinoma types, incomplete surgical resection (non-R0), or secondary cancers.

While the dataset is robust for a single-center study, the authors acknowledge it does not represent a new public benchmark but rather a specific clinical population.

Key Insights and Ideas

  • The "Whole Patient" Approach: The study moves beyond anatomy by including the Albumin to Globulin Ratio (AGR) and the Prognostic Nutritional Index (PNI). These markers reflect the patient's nutritional and immunological "reserve," which are often overlooked in traditional staging.
  • Ratios Matter More Than Counts: One of the strongest predictors in the model was the Positive Lymph Node Ratio (PRL). The authors argue that PRL is a superior indicator compared to simple N-staging because it accounts for the thoroughness of the surgical dissection.
  • Practicality is King: All six factors (Age, Gender, AGR, PRL, N stage, PNI) are "accessible clinical indicators." They can be derived from routine pre-operative blood tests and standard pathology reports, making the nomogram immediately useful in a standard hospital setting without requiring expensive genomic sequencing.

Key Results

The model demonstrated impressive predictive power:

  • Predictive Accuracy: The nomogram achieved a C-index of 0.75 in the primary cohort and 0.70 in the validation cohort.
  • Superiority over TNM: In the primary cohort, the Area Under the Curve (AUC) for the nomogram was 0.801, significantly higher than the AUC for T stage (0.629) or N stage (0.693) alone.
  • Risk Stratification: The researchers used the nomogram scores to divide patients into three risk groups (low, intermediate, and high). The survival differences were stark: the high-risk group had a median overall survival of just 18 months, while the low-risk group had not yet reached its median survival point during the follow-up period.

Limitations and Future Work

Despite the promising results, the authors note several hurdles:

  • Retrospective Bias: As a retrospective study, it is subject to inherent selection biases.
  • Single-Center Scope: The data comes from a single hospital in China, where squamous cell carcinoma is the dominant type. Results might differ in Western populations where adenocarcinoma is more prevalent.
  • Need for Prospective Validation: The authors emphasize that while the model is statistically sound, it requires validation through large-sample, multi-center randomized controlled trials before it can be widely adopted in clinical guidelines.

Summary updated

Abstract

<title>Abstract</title> <bold>Background: </bold>Esophageal cancer (EC) is a malignant tumor with dreadful mortality, nomogram is a prognosis tool of great significance in therapeutic guidance and assessment. We aimed to establish a newly-built nomogram for OS prediction of EC patients with radical esophagectomy.<bold>Methods: </bold>A total of 311 EC patients underwent radical esophagectomy were retrospectively investigated with their survival and demographic and clinicopathological data. Patients were randomly divided into the primary and validation cohorts. The establishment of nomogram was based on Cox hazard regression analysis in primary cohort, the calibration curves and Harrell’s concordance index were performed to verify the predictive accuracy while ROC curves was adopted to reflect the efficacy of nomogram. Kaplan–Meier curves showed the clinical significance of risk classification system and Pearson correlation test was utilized to show the correlation between risk classification system and TNM staging.<bold>Results: </bold>The median OS and 5-year survival rate are 44 months and 29.8% in primary cohort. In validation cohort, they are 52 months and 27.1%, respectively. In Cox hazard regression analysis, we extracted six independent prognostic factors—age, gender, AGR, PRL, N stage, PNI—to establish the nomogram. The C-index of nomogram is 0.75 in primary cohort and 0.70 in validation cohort. Calibration curves indicated high consistency between accurate and predicted OS in both primary and validation cohorts. ROC curves showed a better efficacy of nomogram compared with AJCC T and N stage. The area under curve (AUC) of primary cohort is 0.801 and 0.727 in validation cohort. Patients in primary cohort were divided into three risk groups according to the nomogram score, the median OS between each group was significantly different. Analogical results were obtained in validation cohort. Furthermore, the risk classification system was strongly correlated to AJCC TNM staging system in total cohort (r<sup>2</sup>=0.647, P&lt;0.001), and it also demonstrated a better OS prediction efficacy (AUC=0.742).<bold>Conclusions: </bold>We established a neotype nomogram and a relevant risk classification system with verified accuracy and efficacy in OS prediction of EC patients after radical esophagectomy. They may provide feasible value in prognosis assessment and treatment guidance prospectively.
BibTeX
@inproceedings{li2020neotype,
  author = {Xinye Li and Jinming Xu and Linhai Zhu and Sijia Yang and Yu Li and Wang Lv and Jian Hu},
  title = {{A neotype nomogram based on accessible clinical indicators to predict the overall survival of esophageal cancer patients underwent radical resection}},
  booktitle = {Research Square},
  year = {2020},
  doi = {10.21203/rs.3.rs-87555/v1},
}
Figure