Predictive performance of m TNM staging system
The differences in prognostic prediction between the eighth edition of the TNM staging system and the mTNM classification system were compared directly. The -2 log likelihood of the mTNM stage was 4688.361, which was less than the value of the TNM staging system (4694.876), which indicates a better performance in predicting survival (Table V ). The ROC curves showed that the AUC of the mTNM stage was significantly greater than that of the TNM stage (AUC: 0.646 vs 0.587%, P=0.0007) (Figure 6 ).
Discussion
Both tumor burden and malignant degree can reflect biological characteristics of tumor, and they are also associated with the survival of patients with malignancy. The AJCC TNM classification which reflects tumor burden was the most widely used staging system for malignancy, however, it was a macroscopic classification mainly based on anatomy and did not incorporate the factors reflecting microscopic tumor burden which could not be detected by imageology[14], neither the factors reflecting malignant degree of tumor. As a result, there were always biases of accuracy when the TNM stage was used to predict prognosis[8,9]. In the present study, we supposed to incorporate preoperative serum CA19-9, which could compensate the deficiency of TNM stage in evaluating microscopical tumor burden, and histology grate, which could make up for the defect of TNM stage in reflecting malignant degree, into the TNM staging system to reduce the biases.
In this study, patients with preopertive serum CA19-9 level >112 U/ml demonstrated a significantly lower OS rate than those with CA19-9 ≤112U/ml (5-year OS: 15.9% vs. 22.6%, P<0.001). Patients with high grade also had a significant decreased OS than those with low grade (5-year OS: 15.7% vs. 23.5%, P<0.001). In the multivariate analysis, both preoperative serum CA19-9 and histology grade were identified as independent prognostic factors for OS. Our proposed biological risk of PDAC was established according to the preoperative serum CA19-9 and histology grade. All patients were categorized into three groups according to the biological risk of PDAC: low-risk, middle-risk and high-risk group. Significant differences in OS were observed among the three groups. After incorporating the biological risk into the eighth edition of TNM staging system, mTNM staging was established. It was confirmed that the mTNM staging system was more accurate in predicting postoperative prognosis of PDAC patients than the eighth edition of the AJCC TNM staging system.
CA19-9 was the most widely used serum tumor marker. Kang et al reported that both pretreatment and posttreatment CA19-9 levels or their changes after treatment had good prognostic value in determining the survival of pancreatic cancer patients[17]. At present, there was no recognized cut-off value of CA19-9 to predict the prognosis of PDAC [18,19,20,21], as the level of CA19-9 fluctuated very widely, in our study, the CA19-9 level ranged from 0.6 to 178850 U/ml, besides, the proportion of negative Lewis-antigen patients varied across countries, ethnicities and studies (6% in white population, 22% in black population, 5-10% in yellow population)[22,23,24,25]. Even so, multiple researches have confirmed the advantage of CA19-9 over other serum tumor markers in predicting prognosis of PDAC[26,27,28], which was consistent with our results.
Histology grade was defined as the degree of differentiation of the tumor. Differentiation refers to the morphologic and functional resemblance between a tumor cell and a normal cell of the same tissue. Malignant neoplasms usually evolve from low grade to high grade. In the process of malignancy, the higher the tumor grade, the more aggressive, and the greater the tumor burden[29,30]. Many studies have demonstrated that histology grade was significantly associated with survival of patients with PDAC and could be used to modify the TNM staging system[31,32], so was the case in our study.
To the best of our knowledge, this study is the first to categorize patients with PDAC according to the biological risk based on preoperative CA19-9 and histology grade, and assess the impact of tumor biology on the survival. Our results reveled that patients with low risk had a significant higher OS than those with middle risk, and patients with middle risk also demonstrated a better prognosis than those with high risk. The biological risk was confirmed to be an independent prognostic factor in the multivariate analysis. In our opinion, the combination of preoperative serum CA19-9 and histology grade contains unique prognostic information which represents the microscopic tumor burden and malignant degree, which is not included in the TNM staging system.
There are several limitations to the study. First, it was a retrospective study conducted at a single center. Second, patients were included from 2011 to 2020, therefore, current neoadjuvant therapy were not used routinely. In addition, we could not exclude patients with negative Lewis antigen, which may have an uncertain impact on accuracy. As the constraints listed above may cause bias in the results, data from other centers is required for further validation of our findings.