Carcinoma of unknown primary site: development in a single institution of a prognostic model based on clinical and...

To identify clinical and biologic variables with significant impact on survival in patients with carcinomas of an unknown primary site (CUP) and to develop a simple prognostic model.

In this retrospective study, univariate and multivariate prognostic factors analyses were conducted in a population of 100 patients with CUP. Patients with features requiring well defined treatments had previously been excluded.

Overall survival (OS) was significantly related to the following pretreatment adverse prognostic clinical factors: a poor performance status (2 or 3), weight loss more than 10% in the last six months, the presence of liver metastases and more than two metastatic sites. Two biological parameters predicted a significantly shorter survival: elevated serum levels of alkaline phosphatase and of lactate dehydrogenase. In the multivariate analysis, only two independent adverse prognostic parameters were retained: a poor performance status and the presence of liver metastases. We developed a prognostic model for OS based on the following subgroups: good prognosis (PS 0 or 1 and absence of liver metastases), intermediate prognosis (PS> or =2 or presence of liver metastases) and poor prognosis (PS> or =2 or presence of liver metastases). Median OS for the three groups was 10.8, 4 and 1.9 months respectively, p<0.0001.

A simple prognostic model using performance status and presence of liver metastases was developed. It allowed the assignment of patients into three subgroups with different outcomes. Treatment strategies could be adapted for each subgroup. We think that this prognostic model could be useful and should be validated in other patient series.