Abstract: |
BACKGROUND: Conditional survival can provide valuable information for both patients and healthcare providers about the changing prognosis in surviving patients over time. OBJECTIVE: This study estimated conditional survival for patients with anal cancer in the United States through analysis of a national population-based cancer registry. DESIGN: Log-rank test identified significant covariates of cause-specific survival (defined as time from diagnosis until death from anal cancer). Significant covariates were considered in the multivariable regression of cause-specific survival using Cox proportional hazards models. SETTINGS: Covariates included cancer stage and demographic variables. PATIENTS: Patients in Surveillance, Epidemiology, and End Results regions diagnosed with anal squamous cell carcinoma as their first and only cancer diagnosis from 1988 to 2012 were selected from this database, and 5145 patients were included in the retrospective cohort study. MAIN OUTCOME MEASURE: Five-year conditional survival stratified by each variable in the final Cox models was measured RESULTS: : The final multivariable models of overall and cause-specific survivals included stage, grade, sex, age, race, and relationship status. Over the first 6 years after diagnosis, conditional survival of distant stage increased from 37% to 89%, whereas regional stage increased from 65% to 93% and localized stage increased from 84% to 96%. The other variables had increasing prognosis as well, but the subgroups increased at a more similar rate over time. LIMITATIONS: The data source used does not include information on chemotherapy treatment, patient comorbidities, or socioeconomic status. CONCLUSIONS: Conditional survival showed improvement over time. Patients with advanced stage had the greatest improvement in conditional survival. This is the first study to provide specific conditional survival probabilities for patients with anal cancer. |