RESUMO
PURPOSE: To evaluate the positivity rate of temporal artery biopsies (TAB) performed in suspects of giant cell arteritis (GCA) and to study the epidemiological and clinical factors associated to the biopsy result. METHODS: A retrospective, multicenter, case-control study was performed, including three hundred and thirty-five patients who underwent TAB for a suspicion of GCA from 2001 to 2010. Clinical, epidemiological and pathology data were recovered from the patients' clinical records. Histologic diagnosis of GCA was made when active inflammation or giant cells were found in the arterial wall. RESULTS: Eighty-one biopsies (24.2%) were considered positive for GCA. Clinical factors independently associated to TAB result in a logistic regression analysis were temporal cutaneous hyperalgesia (OR = 10.8; p < 0.001), jaw claudication (OR = 4.6; p = 0.001), recent-onset headache (OR = 4.4; p = 0.001), decreased temporal pulse (OR = 2.8; p = 0.02), pain and stiffness in neck and shoulders (OR = 2.3; p = 0.05), unintentional weight loss (OR = 1.33; p = 0.003) and age (OR = 1.085; p = 0.004). Other factors such as length of the surgical specimen (OR = 1.079; p = 0.028) and erythrocyte sedimentation rate (OR = 1.042; p < 0.001) were also statistically significant. The model was accurate (C-index = 0.921), reliable (pHosmer-Lemeshow = 0.733) and consistent in the bootstrap sensitivity analysis. No significant association was detected between TAB result and number of days of previous systemic corticosteroid treatment (p = 0.146). However, an association was observed between TAB result and the total accumulated dose of previous systemic corticotherapy (p = 0.043). CONCLUSIONS: Exhaustive anamnesis and clinical examination remain of paramount importance in the diagnosis of GCA. To improve the yield of TAB, it should be performed specially in older patients with GCA-compatible clinic. TAB could be avoided in patients with an isolated elevation of acute phase reactants, without GCA-compatible clinic.