An Analysis of Prognostic Biomarkers of the Systemic Inflammatory Response in Cancers of Unknown Primary


Session type:

Rebekah Patton1,Rebekah Patton1,Gillian Knowles1,Rachel Haigh1,Lucy Dobbs2,Barry Laird1,Colin Barrie1,Sally Clive1
1NHS Lothian, Edinburgh, UK,2University of Edinburgh, Edinburgh, UK



Cancers of unknown primary (CUP) represent between 3-5% of all malignancies worldwide. In favourable clinico-pathological subsets median survival is 24 months but can be as little as 1-3 months in ‘poor prognosis’ patients. To rationalise investigations and treatment appropriately it is essential to provide accurate prognostic data. Biomarkers of the systemic inflammatory response (albumin and CRP combined in the modified Glasgow prognostic score (mGPS)) predict survival in cancers with an established site. We sought to establish their prognostic utility in CUP.


A prospective data collection was undertaken of patients with provisional or confirmed CUP (using NICE guideline definitions) referred to the CUP service at the Edinburgh Cancer Centre between 2010-2019. Patients with favourable clinico-pathological subgroups were excluded. mGPS at the time of clinical diagnosis was recorded. Survival, defined as the date of biopsy until death, or censorship if alive, was calculated.


Data were available for 247 patients. Median survival was 2.2 months. 87% of patients had an elevated CRP (>10mg/l) at diagnosis. mGPS stratified median survival from 11.7 months with mGPS 0 to 1.6 with mGPS 2 (p<0.001). 77% of patients with mGPS 0 were alive at three months compared to only 30% of those with mGPS 2.


mGPS, a biomarker of the systemic inflammatory response, is a strong prognostic factor in patients with ‘poor prognosis’ CUP. This simple score, derived from routine blood tests, could be used alongside clinical assessment, standardised investigation and performance status to provide additional objective prognostic information. This may facilitate discussions with patients and assist decisions regarding further investigations and treatment. In patients with mGPS of 2 it may support early conversations about advanced care planning. We would advocate future work validating this finding in other cohorts of CUP patients, and, if supported, incorporating this biomarker into CUP clinical pathways and trials.