Background/Purpose: The diagnosis of giant cell arteritis (GCA) by temporal artery (TA) biopsy requires pathologic identification of arterial inflammation, usually with giant cells. However, some biopsies may show lymphocytic infiltrates without arterial involvement (nonGCA-i). The prognostic significance of these findings is uncertain and virtually no long term follow-up studies have been done.

Methods: We identified 218 TA biopsies performed from 2010-2012 and 123 were available for blind slide review by a pathologist (RLH). Of these cases, 88 had medical records reviewed for long-term outcome. Chart review included: demographics, comorbidities, clinical presentation (constitutional symptoms, craniofacial symptoms, PMR symptoms, change in vision), and laboratory parameters. Using SPSS software, the associations between two categorical variables were analyzed with Chi-Square or FisherÕs Exact test. Mann-Whitney U test was used to analyze continuous variables.

Results: Based on assessment of symptoms at presentation and during long-term follow-up, patients with nonGCA-i were significantly less likely to have constitutional symptoms, craniofacial symptoms, CRP elevation, and vision loss (P<0.05).  The frequency of temporary or permanent vision loss of any etiology among the 3 groups was as follows: 8/18 (44.4%) patients with TA biopsy positive for GCA, 1/34 patients with nonGCA-i, (2.9%) and 5/35 (14.3%) patients with negative TA biopsy.  There was a trend towards less PMR among patients with nonGCA-i (P=0.088). The frequencies of the 3 groups meeting the 1990 ACR classification criteria (excluding biopsy result) were as follows: 16/18 (88.8%) patients with TA biopsy positive for GCA, 25/35 (71.4%) patients with nonGCA-i, and 19/35 (54.3%) patients with negative TA biopsy.  Patients with nonGCA-i received a shorter duration of corticosteroids compared to those with positive TA biopsy (P = 0.001). While 13/33 (39.4%) patients with non-GCA-i and 13/35 (37.1%) patients with negative biopsy were treated with steroids for at least 4 months, all 14 patients with TA biopsy positive for GCA were treated for at least 4 months (excluding patients lost to follow-up or deceased).

Conclusion: In this cohort, patients with nonGCA-i rarely developed clinically evident GCA-like symptoms requiring corticosteroid therapy >= 4 months.  Most importantly, patients with nonGCA-i did not develop complications of visual loss acutely or in long-term follow-up.  Based on our findings, nonGCA-i in a temporal artery biopsy is an incidental finding and is not a harbinger of GCA clinically, even with long-term followup. Treatment on the basis of nonGCA-i alone may expose the patient to unnecessary corticosteroids.  No new cases of GCA were identified in the pathological review.  However, dose and duration of corticosteroid therapy for patients with nonGCA-i should always be guided by clinical judgement.