Evolution in the diagnosis and treatment of carcinoma of unknown primary: a multicenter Canadian analysis.

Background: Guidelines for the management of patients with cancer of unknown primary (CUP), who have metastatic disease without an identified primary tumor site, have evolved. We sought to describe the diagnostic work-up and outcomes of patients with CUP in Canada over the last decade. We also sought to identify factors associated with improved prognosis in CUP, including primary tumor site identification, identification of "favorable subtypes," and concordance with published guidelines.

Methods: With ethics board approval, patients with histologically confirmed CUP between 2012 and 2021 in 3 Canadian cancer centers were reviewed and clinicopathological variables retrospectively collected. The primary endpoint was to describe significant trends in CUP diagnosis and management over the decade using linear regression models. Univariable (UVA) and multivariable (MVA) logistic regression analyses identified variables correlated with primary site identification and overall survival (OS). Kaplan-Meier curves with the log-rank test were used to compare OS outcomes.

Results: In total, 907 patients were included, with a median follow-up of 5.1 months. There was an increase in both 5-year survival and identification of primary tumors over the decade. Diagnostic tests including next-generation sequencing were independently associated with primary site identification on UVA. However, primary site identification was not found to be predictive of survival; instead, patients with "favorable subtypes" of CUP had significantly longer OS.

Conclusions: Survival in patients with CUP in Canada has been increasing over the last decade. Identifying the primary site does not influence survival, and efforts should be focused on discovering novel "favorable subtypes" which have superior outcomes.