Identification of death risk factors of primary melanoma based on the competing risk model

Bull Cancer. 2021 Oct;108(10):915-930. doi: 10.1016/j.bulcan.2021.04.020. Epub 2021 Jul 30.

Abstract

Objective: To preliminarily explore death risk factors in primary melanoma patients.

Method: Competing risk model analysis was used using a large sample public cohort and Cox proportional hazard model was compared.

Result: In the competing risk model analysis, age, gender, ethnicity, stage, site, TMN stage and metastases were the independent risk factors of single primary melanoma (SPM) death. T stage had a particularly important impact on SPM death. T2 stage had a 3.212 times greater risk of interest event than T1 stage [hazard ratio (HR)=3.212, 95%CI: 2.994-3.446], T3 stage was 5.747 times greater than that T1 stage (HR=5.747, 95%CI: 5.337-6.187) and T4 stage had a 7.086 times than T1 stage (HR=7.086, 95%CI: 6.514-7.708). Gender, ethnicity, stage, site, T stage and brain and liver metastases were the independent risk factors of multiple primary melanoma (MPM) death. When some groups had a very high death rate or the reference group had a very low death rate in competing events, the results of Cox proportional hazard model may not be as accurate as the results obtained by fine-Gray regression model.

Conclusion: Early diagnosis and therapy, and prevention of tumor progression and metastases of primary melanoma patients are important measures to improve its prognosis and survival.

Keywords: Competing risk model; Primary melanoma; Risk factors for death.

MeSH terms

  • Adult
  • Age Factors
  • Aged
  • Bone Neoplasms / mortality
  • Bone Neoplasms / secondary
  • Brain Neoplasms / mortality
  • Brain Neoplasms / secondary
  • Cohort Studies
  • Confidence Intervals
  • Female
  • Humans
  • Liver Neoplasms / mortality
  • Liver Neoplasms / secondary
  • Lung Neoplasms / mortality
  • Lung Neoplasms / secondary
  • Male
  • Melanoma / ethnology
  • Melanoma / mortality*
  • Melanoma / pathology
  • Melanoma / secondary
  • Middle Aged
  • Neoplasm Staging
  • Proportional Hazards Models
  • Regression Analysis
  • Risk Factors
  • Sex Factors