24. Predictors of Cancer in Patients Diagnosed with Giant Cell Arteritis in Western Norway 1972-2012

Background: Giant cell arteritis (GCA) is the most common systemic vasculitis in adults, and the number of incident cases worldwide is projected to increase. Evidence as to whether or not GCA confers an added risk of cancer is conflicting.

Objective: Our aim was to investigate possible predictors of cancer in a large and well-characterized Norwegian cohort of GCA-patients. 

Methods: This is a hospital-based retrospective cohort study including patients diagnosed with GCA during 1972-2012. Patients were identified through computerized hospital records using the International Classification of Diseases coding system. Clinical information was extracted from patients’ medical journals and national registries. Further details about the inclusion process have been published previously. Data on the occurrence of cancer was obtained from the Cancer Registry of Norway. We investigated predicting factors using Cox proportional hazards regression. Selected variables were first analyzed in univariate and block regression models (block 1: clinical features including histology, block 2: laboratory and treatment factors, block 3: demographic and traditional risk factors). Variables included in the final multivariate model were selected on the following basis: P-value <0.1 in univariate or block regression or otherwise deemed clinically significant.

Results: A total of 881 patients were included following validation of the GCA diagnosis. Among these, 767 cases (71.8% women, mean age 72.5 (SD 9)) had no registered cancer prior to GCA diagnosis and were included in a time-to-event analysis, with first cancer as the event. During the observation period, 120 patients (15.6%) were diagnosed with a first cancer. We found no significant difference in the risk of malignancy compared to population controls matched according to age, sex and county of residence, but for the controls we lacked further data on potential cancer risk factors. Within the GCA-cohort we found no significant associations with the risk of cancer for the variables we could adjust for (Figure 1). Neither typical clinical findings of GCA, initial laboratory parameters nor initial or maximal (before first tapering) Prednisolone dose were predictive of cancer risk.

Conclusions: In our large cohort of GCA-patients the risk of cancer was found to be like that of population controls and no specific GCA-symptom, finding, laboratory-parameter or treatment factor were found to be predictive of cancer risk. However, our data are limited by incomplete or missing data on some cancer risk factors such as use of hormones, body mass index and family history of malignancy. Furthermore, our analyses were not able to confirm the known association of cancer with smoking and advancing age. This may be due to small numbers.

Disclosures: None