Epidemiological Analysis and Prognosis of Conjunctival Cancer in the Past Twenty Years: A Population-Based Retrospective Study Using SEER Data

Abstract

Epidemiological studies of malignant primary conjunctival tumors are rare. We extracted data pertaining to primary site-labeled conjunctival cancer patients present within the Surveillance, Epidemiology, and End Results (SEER) database from 1992 to 2001 and from 2002 to 2011. The Kaplan-Meier approach was used for comparisons of overall survival (OS) between patients, while OS-related risk factors were identified via a Cox proportional hazards regression approach. We then constructed a nomogram that could be used to predict the 3- and 5-year OS, with the accuracy of this predictive model based on receiver operating characteristic (ROC) curve. We observed a significant reduction in age-adjusted incidence of conjunctival cancer in the 50-69-year-old age group of the 2002-2011 cohort relative to the 1992-2001 cohort (APC, P<0.05). There were no significant differences in OS between the 1992-2001 and 2002-2011 conjunctival cancer patient cohorts. Being ≥30 years old (P<0.05), male (P<0.001), single (P<0.05), divorced (P<0.001), or widowed (P<0.001) were all associated with an increased OS-related risk of primary conjunctival cancer (1992-2011). Our nomogram was able to accurately predict 3- and 5-year OS in conjunctival cancer patients. In verification mode, the 3-year area under the curve (AUC) was 0.697 and the 5-year AUC was 0.752. We found that age, sex, and marital status were all associated with primary conjunctival cancer survival. Our results further suggest that conjunctival cancer incidence and survival rates have been relatively stable over the last two decades, and using these data, we were able to generate a satisfactory risk prediction model for this disease.