Zeng J, Zhu L, Zhou G, Pan F, Yang Y. Prognostic models based on lymph node density for primary gastrointestinal melanoma: a SEER population-based analysis.
BMJ Open 2023;
13:e073335. [PMID:
37798018 PMCID:
PMC10565139 DOI:
10.1136/bmjopen-2023-073335]
[Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Track Full Text] [Journal Information] [Submit a Manuscript] [Subscribe] [Scholar Register] [Received: 03/06/2023] [Accepted: 09/11/2023] [Indexed: 10/07/2023] Open
Abstract
OBJECTIVE
This study aimed to construct prognostic models to predict the overall survival (OS) and cancer-specific survival (CSS) of patients with primary gastrointestinal melanoma (PGIM).
DESIGN
An observational and retrospective study.
SETTING
Data were obtained from the Surveillance, Epidemiology and End Results (SEER) programme database, encompassing a broad geographical and demographic spectrum of patients across the USA.
PARTICIPANTS
A total of 991 patients diagnosed with PGIM were included in this study.
METHODS
A total of 991 patients with PGIM were selected from the SEER database. They were further divided into a training cohort and a validation cohort. Independent prognostic factors were identified by Cox regression analysis. Two prognostic models were constructed based on the results of multivariable Cox regression analysis. The concordance index (C-index) and area under the time-dependent receiver operating characteristic curve (time-dependent AUC) were used to evaluate the discriminative ability. Calibration curves were plotted to evaluate the agreement between the probability as predicted by the models and the actual probability. Risk stratification was developed given the model.
RESULTS
By the multivariable Cox regression analysis, we identified four independent risk factors (age, stage, lymph node density and surgery) for OS, and three independent risk factors (stage, lymph node density and surgery) for CSS, which were used to construct prognostic models. C-index, time-dependent AUC, calibration curves and Kaplan-Meier curves of risk stratification indicated that these two models had good discriminative ability, predictive ability as well as clinical value.
CONCLUSIONS
The prognostic models of OS and CSS had satisfactory accuracy and were of clinical value in evaluating the prognosis of patients with PGIM.
Collapse