LAUSR

BACKGROUND Patients with cancer of unknown primary site are divided into two distinct groups, favourable and unfavourable subsets. For the unfavourable subset, empiric treatment or site-specific treatment is recommended, but limited knowledge exists about the efficacy of site-specific treatment compared with empiric treatment in clinical practice. METHODS In this multicentre retrospective study, we reviewed the medical records of patients with cancer of unknown primary site treated with chemotherapy (or chemoradiotherapy) as first-line treatment from eight institutions during 2006-18. We investigated the workup modality and categorized the patients into favourable and unfavourable subsets, which were further divided into site-specific and empiric treatment groups. Site-specific treatment is defined as a standard chemotherapy for an estimated primary site. We examined the efficacy in the favourable and unfavourable subsets and performed multivariable analysis for estimating the overall survival in the unfavourable subset. RESULTS Of 177 patients with cancer of unknown primary site, 33 and 144 were categorized into favourable and unfavourable subsets, respectively. In the unfavourable subset, 84 patients (58.3%) received empiric therapy, and 60 patients (41.7%) received site-specific treatment. Median overall survival was 10.0 and 10.1 months in site-specific and empiric treatment groups, respectively, with no significant difference (hazard ratio 1.01, 95% confidence interval 0.70-1.45, P = 0.95). Multivariable analysis revealed performance status, number of metastatic sites and hypoalbuminaemia as independent prognostic factors for overall survival in the unfavourable subset. CONCLUSIONS Overall survival in site-specific and empiric treatment groups was similar in the unfavourable cancer of unknown primary site subset in this study. Further research is needed to prolong overall survival in patients in the unfavourable cancer of unknown primary site subset.