Clinicopathological features and survival outcomes of patients with different distant and lymph node metastasis in pancreatic cancer
摘要
Pancreatic cancer is a highly lethal malignancy, usually diagnosed when metastasis is present. Detailed analysis of distant (DM) and lymph node (LN) metastasis patterns is crucial for prognosis. This study evaluated metastasis-specific prognosis in pancreatic cancer patients from 2010 to 2016 and developed a validated nomogram for survival prediction.
MethodsData of patients with metastatic pancreatic cancer who received chemotherapy or radiotherapy were retrieved from the Surveillance, Epidemiology, and End Results (SEER) database. Pearson’s chi-square test was utilized to analyze clinicopathological variables. Kaplan–Meier survival analysis and log-rank tests were employed to evaluate overall survival and prognostic outcomes. A nomogram for prognosis prediction was constructed using univariate and multivariate analyses, combined with the Cox proportional hazards regression model.
ResultsDistant metastasis (DM) most commonly occurred in the liver (44.8%), followed by the bone (9.5%). Approximately 63.2% of patients had single-site metastasis, whereas 26.0% had metastases in two distant organs. Regional lymph node (LN) involvement was observed in 14.3% of pancreatic cancer patients, among whom only 3.7% concurrently had DM, with the liver being the most frequently involved organ, in which the liver was the most common organ. At the end of the follow-up period 152 patients in 1805 were alive. For pancreatic carcinoma patients with DM the median survival decreased to 4.0 months, while with LN that was 31.05 months. The nomogram was proven to be acceptable consistency in train and validation sets, and the c-index was 0.763.
ConclusionWe conducted a comparative analysis of clinicopathological characteristics and clinical outcomes among cases of metastatic pancreatic cancer. Furthermore, we developed a prognostic nomogram that demonstrated excellent performance in both the training and validation cohorts.