Anatomic Pathology / PERIVASCULAR INFLAMMATION IN TEMPORAL ARTERY BIOPSY SPECIMENS

We retrospectively compared 81 temporal artery biopsy specimens demonstrating perivascular inflammation without evidence of temporal arteritis and 76 specimens demonstrating no inflammation. Patients with perivascular inflammation included 43 women (mean age, 71.2 years). Nineteen patients met the 1990 American College of Rheumatology (ACR) criteria for the diagnosis of temporal arteritis. All patients demonstrated chronic perivascular inflammation consisting primarily of lymphocytes. Granulomas were noted in 4 specimens. Internal elastic lamina disruption, intimal fibroplasia, and dystrophic calcification were noted in 86 arteries examined. Fibrosis or scarring of the vessel walls was observed in 10 specimens. Corticosteroid therapy was beneficial to 33 of 56 patients. In patients with no evidence of inflammation (50 women; mean age, 66.6 years), 21 met ACR criteria for temporal arteritis. Histologically, disruption of the elastic lamina was noted in 75 of 81 arteries biopsied, intimal fibroplasia in 66, microcalcifications in 5, and fibrosis or scarring in 5. In this group, 47 patients received corticosteroid therapy; clinical improvement was noted in 28. Patients with chronic perivascular inflammation but no arteritis seem no more likely to have temporal arteritis on clinical grounds than similar patients without inflammation on biopsy. Temporal arteritis has been long recognized as a distinct entity since its initial description in the literature in 1890 by Hutchinson.1 Epidemiologic studies have shown an incidence rate of up to 30 cases per year per 100,000 persons older than 50 years in some populations.2-4 Temporal artery biopsy has long been advocated for patients in whom a diagnosis of temporal arteritis is suspected. Because of the potential complications of corticosteroid therapy, biopsy confirmation of the diagnosis is reassuring before initiating treatment. The histopathologic diagnosis is predicated on the recognition of a lymphohistiocytic infiltration of the temporal artery wall, frequently accompanied by multinucleated giant cells. Temporal arteritis is notoriously patchy, with the false-negative biopsy rate, as determined by correlation with clinical findings, reported as 42% to 61% in published series.5-7 Histopathologic parameters that would permit reclassification of these false-negative biopsies as “positive” or “probable” temporal arteritis have yet to be identified, if in fact they exist. The significance, in particular, of perivascular inflammation adjacent to the artery has not been determined. We attempted to study this issue by retrospectively reviewing a series of such temporal artery biopsy specimens that were marked by chronic perivascular inflammation but demonstrated no evidence of vasculitis. Materials and Methods All available temporal artery biopsy specimens from January 1990 to December 1993 were reviewed retrospectively. After histologic review, biopsy specimens were placed into 1 of 3 categories: (1) arteries with definite evidence of Anatomic Pathology / ORIGINAL ARTICLE Am J Clin Pathol 2001;115:342-347 343 © American Society of Clinical Pathologists active arteritis, (2) arteries with chronic perivascular inflammation but no definite evidence of arteritis, or (3) biopsy specimens that showed no evidence of inflammation. The presence of inflammation, generally lymphocytic or granulomatous in nature, within the artery wall was required to make a diagnosis of temporal arteritis ❚Image 1❚. In each case, H&E-stained sections and Movat pentachrome–stained sections of artery were reviewed. Arteries were sectioned into 2 to 14 pieces (mean, 5.0 pieces). The number of levels generated from the paraffin block ranged from 2 to 58 (mean, 13.5 levels). Discrepancies between the original diagnosis and diagnosis on review was resolved by majority consensus of the authors. Since the focus of this study was on the cases with perivascular inflammation and no active arteritis, all 81 cases with this histologic finding were included. A comparison group of 76 cases without evidence of inflammation on biopsy also was evaluated. Histopathologic features evaluated included the presence of perivascular inflammation and the types of inflammatory cells observed in the infiltrate, disruption of the internal elastic lamina, fibrosis or scarring of the vessel wall, presence of granulomas, evidence of intimal fibroplasia, microcalcifications, and thrombosis. In each case, the medical records were reviewed for clinical information, including the age and sex of the patient at the time of biopsy, site of biopsy, erythrocyte sedimentation rates (ESRs) before biopsy (within 1 week and before treatment) and at the most recent follow-up after the biopsy, symptomatology that resulted in the patient seeking treatment, clinical impression at the time of biopsy, and information about corticosteroid therapy and responsiveness to therapy. In each case, particular note was made about whether the patient fulfilled the American College of Rheumatology (ACR) criteria for the diagnosis of giant cell arteritis.2 The criteria include the following: (1) age at disease onset, 50 years or older; (2) onset of new headaches; (3) temporal artery tenderness to palpation or decreased pulsation unrelated to arteriosclerosis of cervical arteries; (4) ESR, 50 mm/h or more by the Westergren method; and (5) a biopsy specimen with artery showing vasculitis characterized by a predominance of mononuclear cell infiltration or granulomatous inflammation, usually with multinucleated giant cells. The presence of at least 3 of these 5 criteria has been associated with a sensitivity of 93.5% and a specificity of 91.2% for the diagnosis of giant cell (temporal) arteritis.2 The control and perivascular inflammation groups were compared on the proportion meeting the individual ACR criteria (any 3 of the 5 criteria) with chi-square tests.