Estimates of Conditional Survival in Gastric Cancer Reveal a Reduction of Racial Disparities with Long-Term Follow-Up

Introduction In prior analyses, conditional survival (CS) estimates for gastric cancer have weighed clinical and pathologic factors to predict prognosis at time intervals after surgery. Since racial disparities in gastric cancer outcomes were not considered, our objective was to determine whether race influences CS estimates. Methods Data from the Surveillance, Epidemiology, and End Results cancer registry were used to identify gastric adenocarcinoma patients who underwent curative surgical intervention between 1988 and 2005. Five-year relative conditional survival (RCS) was computed for patients who survived at least 1 to 5 years after surgery. RCS was calculated by assessing observed and expected survival in an age- and race-matched standard population. Results were compared across time and racial groups (white, black, and Asian) using z test statistics. Results Of 14,067 patients, 63.8 % were white, 15.5 % black, and 20.7 % Asian. Racial disparities among groups were observed with improved survival of Asians at time point zero and improved RCS at 1 year. At 5 years following curative surgery, each racial group had increased RCS and the greatest magnitude of relative increase was observed in white and black patients (121 and 118 %, respectively). Comparison of RCS at the 5-year time point revealed a reduction of racial disparities in survival among the three groups. Conclusions Our investigation shows that racial disparities in gastric cancer outcomes are pronounced at the time of curative surgery but diminish after years of survival, thus suggesting race has less influence over outcomes the longer patients survive. The reasons for reduction of racial disparities remain unclear and warrant future study.