Management of "unfavourable" carcinoma of unknown primary site: synthesis of recent literature.

Carcinomas of unknown primary (CUP) approximately represent 2-3% of all adult cancers. Various clinicopathological subsets of CUP have been identified, which may be treated with tailored approaches. Nevertheless, 80% of CUP do not fall into these subsets. Even when at least 4 prognostic models have been developed and validated in independent patient cohorts, there is no consensus or reliable guidance for estimating the prognosis of these "unfavourable" CUP. Consequently, targeting patients who benefit from palliative chemotherapy is difficult. Thirty-eight phase II trials were published between 1997 and 2011; a systematic analysis of these trials did not allow the recommendation of any of the tested regimens as a standard of care. Currently, there is only one published phase III clinical trial (Paclitaxel/carboplatin/etoposide versus gemcitabine/irinotecan); without significant difference between both regimens. Thus, with the promise of molecular profiling, we are waiting for a large collaborative clinical trial that validates the concept of targeted treatment in this population of patients with "unfavourable" CUP.