Comorbidity predicts poor prognosis in nasopharyngeal carcinoma: development and validation of a predictive score model

Radiother Oncol. 2015 Feb;114(2):249-56. doi: 10.1016/j.radonc.2014.12.002. Epub 2015 Jan 21.

Abstract

Background and purpose: The impact of comorbidity on prognosis in nasopharyngeal carcinoma (NPC) is poorly characterized.

Material and methods: Using the Adult Comorbidity Evaluation-27 (ACE-27) system, we assessed the prognostic value of comorbidity and developed, validated and confirmed a predictive score model in a training set (n=658), internal validation set (n=658) and independent set (n=652) using area under the receiver operating curve analysis.

Results: Comorbidity was present in 40.4% of 1968 patients (mild, 30.1%; moderate, 9.1%; severe, 1.2%). Compared to an ACE-27 score ⩽1, patients with an ACE-27 score >1 in the training set had shorter overall survival (OS) and disease-free survival (DFS) (both P<0.001), similar results were obtained in the other sets (P<0.05). In multivariate analysis, ACE-27 score was a significant independent prognostic factor for OS and DFS. The combined risk score model including ACE-27 had superior prognostic value to TNM stage alone in the internal validation set (0.70 vs. 0.66; P=0.02), independent set (0.73 vs. 0.67; P=0.002) and all patients (0.71 vs. 0.67; P<0.001).

Conclusions: Comorbidity significantly affects prognosis, especially in stages II and III, and should be incorporated into the TNM staging system for NPC. Assessment of comorbidity may improve outcome prediction and help tailor individualized treatment.

Keywords: Comorbidity; Multi-cancer centre; Nasopharyngeal carcinoma; Prognosis; Survival.

Publication types

  • Research Support, Non-U.S. Gov't

MeSH terms

  • Adult
  • Aged
  • Carcinoma
  • Comorbidity
  • Disease-Free Survival
  • Female
  • Humans
  • Male
  • Middle Aged
  • Models, Statistical
  • Nasopharyngeal Carcinoma
  • Nasopharyngeal Neoplasms / diagnosis*
  • Nasopharyngeal Neoplasms / pathology*
  • Neoplasm Staging
  • Prognosis
  • Reproducibility of Results
  • Risk Factors