Cancer of unknown primary: time trends in incidence, United States

Abstract

Purpose: To describe the epidemiological features and trends of cancer of unknown primary (CUP) in a large and diverse US population.

Methods: The Surveillance Epidemiology and End Results registry was used to examine incidence rates, adjusted to the World Segi 1960 population, by demographic and tumor characteristics among patients diagnosed with CUP between 1973 and 2010. Annual percent changes in incidence rates were estimated using Joinpoint regression.

Results: The incidence rate of pathologically investigated CUP was 4.1 per 100,000 and is consistent with reports from other countries. In the USA, CUP incidence rates have been decreasing since the early 1980s, 3.6 % per year in the last two decades. The USA experienced decreases earlier than other countries. US males and African Americans had the highest rates of CUP. The rates of non-microscopically confirmed CUP have dropped 2.6 % per year since 1973, but 24 % of CUP patients do not receive microscopic confirmation and 21 % of those with microscopically investigated cancer receive a vague histology (i.e., epithelial) diagnosis. Twenty percent of patients with pathological investigation receive radiation. Patients were twice as likely to be diagnosed with a non-pathologically investigated CUP if they were living in areas with the lowest income quartile relative to areas with the highest income quartile.

Conclusion: Although the incidence of CUP is decreasing, we document CUP that may be due to insufficient diagnostic inquiry. Questions raised by the findings in this data provide hypotheses for further epidemiological and biological studies in the elucidation of CUP incidence and treatment.