A clinical review of the investigation and management of carcinoma of unknown primary in a single cancer network

Clin Oncol (R Coll Radiol). 2007 Feb;19(1):87-95. doi: 10.1016/j.clon.2006.09.009.


Aims: Carcinoma of unknown primary (CUP) is a common encounter in oncological practice and represents 2.0-6.0% of all invasive malignancies. Evidence to support particular therapeutic strategies in this patient population is scarce, and empirical therapies are frequently derived from research on patients where the primary tumour site is known.

Materials and methods: This retrospective study reviewed the management of all patients recorded to have a diagnosis of CUP in a single cancer centre over a period of 12 months. Health records were reviewed documenting the CUP subtype, the investigations carried out both in the referring cancer unit and subsequently at the cancer centre and the recommended treatment (type and regimen), together with survival. The outcomes were examined in respect to a number of prognostic factors. Statistical tests were considered significant at P < 0.05.

Results: One hundred and sixty-six patients were recorded to have a diagnosis of CUP, representing 3.7% of all referrals to the cancer centre. The median age of patients was 68 years (range 32-94 years), and 52.0% were women. The three most common CUP subgroups were CUP-liver/multiple sites (25.0%), CUP-bone (21.0%) and CUP-brain (16.0%). The remaining subgroups occurred at frequencies of less than 10% each. Histological confirmation was only obtained in 55.0% of cases. Even within a single subtype, 41 patients with CUP-liver/multiple sites underwent a total of 19 different investigations before any treatment being given. Forty-seven (28.0%) patients received radiotherapy, 30 (18.0%) received chemotherapy and 58 (35.0%) received supportive care alone. Nine different 5-fluorouracil-containing regimens were used in 11 patients treated with chemotherapy for CUP-liver. The overall median survival for all patients was 4.0 months. Survival was better in patients with a good performance status (0-1) and absent liver metastases (median survival 15.0 months; 95% confidence interval 8.0-22.1) and those who received chemotherapy (median survival 13.0 months; 95% confidence interval 7.4-18.6). Multivariate analysis confirmed female gender (P = 0.006), a good performance status (0/1) (P < 0.0001) and absent liver metastases (P = 0.002) as favourable prognostic indicators.

Conclusions: The appropriate management of patients with CUP is unclear and this study revealed a high degree of variation in clinical practice. This area is in urgent need of clinical research to ensure that the treatment of CUP is evidence based. Until such time, clinical recommendations are suggested for the investigation and treatment of such patients. Therapeutic progress will be facilitated by designating a clinical lead for CUP in each clinical network.

MeSH terms

  • Adenocarcinoma / mortality
  • Adenocarcinoma / secondary*
  • Adenocarcinoma / therapy
  • Adult
  • Aged
  • Aged, 80 and over
  • Bone Neoplasms / mortality
  • Bone Neoplasms / secondary*
  • Bone Neoplasms / therapy
  • Brain Neoplasms / mortality
  • Brain Neoplasms / secondary*
  • Brain Neoplasms / therapy
  • Female
  • Humans
  • Kaplan-Meier Estimate
  • Liver Neoplasms / mortality
  • Liver Neoplasms / secondary*
  • Liver Neoplasms / therapy
  • Male
  • Middle Aged
  • Neoplasms, Unknown Primary / mortality
  • Neoplasms, Unknown Primary / therapy*
  • Outcome Assessment, Health Care
  • Prognosis
  • Proportional Hazards Models
  • Retrospective Studies
  • Survival Analysis