Temporal arteritis is a granulomatous vasculitis with preferential involvement of the aorta and its large primary branches, including the carotid and vertebral arteries that supply the brain and eyes (Ostberg 1973). In a population-based study, it was found that patients with temporal arteritis were 17.3 times more likely to develop thoracic aortic aneurysm and 2.4 times more likely to develop abdominal aortic aneurysm than age- and sex-matched controls (Evans et al 1995). Lie found that of 72 cases of aortic and extracranial giant cell arteritis, 25% did not have the clinical manifestations of temporal arteritis. The ascending aorta and arch were most frequently involved (39%), followed by subclavian and axillary (26%) as well as femoropopliteal arteries (18%), paralleling the distribution found by Ostberg in his much earlier and smaller study (Ostberg 1973; Lie 1995).
Further investigation of the variable distribution of giant cell arteritis in a cohort of patients in the Mayo Clinic series supports the concept that there is a fundamental biological difference between patients with cranial giant cell arteritis and those with similar histologic presentation primarily in the aortic arch and great vessels (Brack et al 1999). These authors compared 74 patients with subclavian and axillary giant cell arteritis, with 74 patients having cranial giant cell arteritis. The incidence of arteritis in the temporal artery biopsy was lower in the subclavian and axial group (negative biopsy in 42%) compared with the cranial arteritis group (100% positive biopsy). Subclavian and axial arteritis was biologically different from cranial arteritis cases in having higher concentrations of interleukin-2 gene transcripts in the artery tissue and over-representation of the HLA-DR beta 1*0404 allele.
The cause and pathogenesis are unknown, but considerable attention has been focused on the immunologic characteristics of the cellular infiltrate, which is dominated by mononuclear cells and multinucleated giant cells (Lie 1990). Immunohistochemical studies have shown most of the lymphocytes in the infiltrate are of the CD4+ "helper" variety. Activation of T cells requires that the inciting antigen be presented to the lymphocyte by a specific autologous human leukocyte antigen molecule. The responsible antigen is unknown, but it has been thought likely to derive from some component of the internal elastic lamina of the involved arteries, as the granulomatous infiltrate is most intense adjacent to this part of the artery wall, and the lamina is usually fragmented and otherwise damaged where the inflammation is intense (Kimmelstiel et al 1952). Also, the relative frequency of inflammation in different portions of the arteries correlates positively with the amount of elastic tissue in the media and adventitia. Thus, although the carotid and vertebral arteries are often involved, the inflammation ceases at the point where these arteries penetrate the dura to become intracranial vessels, which is also the anatomic point at which the internal elastic lamina disappears from the vessel wall (Wilkinson and Russell 1972). Nordborg and Nordborg provided evidence that the inflammatory cells and other inflammatory markers, such as interleukin-2, are concentrated at the border between the arterial adventitia and the media using monoclonal antibody against inflammatory markers. They suggest that the cellular infiltrate enters the media from the adventitial vasa vasora and migrate to the internal elastic lamina and intima (Nordborg and Nordborg 1998).
Weyand and colleagues studied allelic human leukocyte antigen polymorphisms important in determining immunoresponsiveness to exogenous antigens in the cellular infiltrate from temporal arteries of 42 patients with biopsy-proven temporal arteritis (Weyand et al 1992). The authors summarize their results as follows: "The giant cell arteritis patients shared a sequence motif spanning amino acid positions 28 to 31 of the HLA-DR beta 1 chain. In the structural model for HLA-DR molecules, this sequence motif can be mapped to the antigen-binding site of the HLA complex, suggesting a crucial role of antigen selection and presentation in giant cell arteritis."
It is interesting to note that although 60% of the temporal arteritis patients carried one of the closely related allelic variants at HLA-DR beta 1-*0401 and *0401/8, all of the patients who experienced ischemic optic neuropathy and visual loss carried the *0404 or *0401/8 allele, suggesting that these may be markers for increased disease severity.
The same authors later found that the distribution of HLA-DR beta 1 alleles among patients with polymyalgia rheumatica resembles that of patients with temporal arteritis and that the profile of patients with rheumatoid arthritis is different at that locus. This finding means that the HLA-DR beta 1 allele profile does not distinguish which of the patients with polymyalgia are destined to develop concomitant temporal arteritis (Weyand et al 1994). Conversely, in a population of patients with polymyalgia rheumatica in a Mediterranean country, Salvarani and colleagues found no significant differences in the frequencies of HLA-DRB1 types and in the expression of HLA-DRB 70-74 shared motif between polymyalgia rheumatica patients and controls (Salvarani et al 1999). Also, a series of 86 patients with polymyalgia rheumatica in Northwest Spain were found to be associated with HLA-DRB1*13/14 with only marginal increase in the frequency of HLA-DRB1*0401 and *0404 compared with controls. Sixty-two cases of biopsy-proven temporal arteritis from the same series demonstrated an association with HLA-DRB1*04 alleles, particularly *0401 and *0404 (Dababneh et al 1998). Thus, there may be population specific differences in HLA types in patients with both temporal arteritis and polymyalgia rheumatica.
The specific predilection of temporal arteritis for the elderly prompted Martinez-Taboada and others to study age-related changes in the T cell receptor repertoire in peripheral blood CD8+ T cells. They found a distinct Jbeta2.7+ CD8+ clonotype only among patients and suggested that age-related emergence of such clonotypes predisposes elderly patients to develop both polymyalgia and temporal arteritis (Martinez-Taboada et al 1996).
In a striking departure from previous ideas, Petursdottir and colleagues suggested that, based on study of aortic tissue from 3 autopsy cases and 4 surgical specimens from patients with temporal arteritis, the primary event in giant cell arteritis is loss of smooth muscle cells. They used alpha-smooth muscle actin immunoreactivity to identify smooth muscle cells in the media of these specimens. There was subtotal loss of smooth muscle in regions most remote from inflammation, but these areas gave way to zones with total loss of muscle cells next to calcified, acellular areas that were surrounded by variable granulomatous inflammation (Petursdottir et al 1996).
There has been speculation about the role a virus might play in the pathogenesis of giant cell arteritis, but there has been little proof until more recently. In a multicenter study Duhaut and colleagues investigated 305 new cases of temporal arteritis, of which 159 were biopsy proven and 76 cases of polymyalgia rheumatica, using serologic tests for IgG and IgM, directed against a number of viruses. There was a strong predominance of antibodies against parainfluenza type 1 virus in patients versus controls, the difference being larger in the biopsy positive cases (43% of patients vs. 20.9% of controls; odds ratio with controls 2.89; p+0.000006). There was no significant difference between patients and controls in the incidence of antibodies against the other viruses screened. The authors concluded that parainfluenza virus type 1 is associated with the onset of giant cell arteritis in a subset of patients (Duhaut et al 1999a). On the other hand, Gabriel, studying temporal artery biopsy tissue from 30 patients with temporal arteritis, found parvovirus B19 DNA more likely to be present in the temporal arteries of patients with temporal arteritis (29 out of 30) than in the temporal arteries of control subjects (Gabriel et al 1999).
Angiotensin I receptors were found to be increased in the vascular smooth muscle within the media of temporal arteries from patients with temporal arteritis as compared with control arteries (Dimitrijevic et al 2009). The authors suggest that treatment of temporal arteritis with angiotensin receptor blockers might be an alternative or adjunct to corticosteroids in this condition.
Recent studies indicate that 2 distinct pathogenetic mechanisms can be distinguished in temporal arteritis (Weyand and Goronzy 2003; Deng et al 2010). One, mediated by Th17 T-cells that produce interleukin 17, serves systemic inflammation with manifestations such as fever, malaise, anorexia, and headache and is corticosteroid responsive. The other, mediated by Th1 T-cells that produce interferon-gamma, mediates vascular wall inflammation and is corticosteroid resistant. It is suggested that new therapeutic approaches to the steroid-resistant inflammatory process in the large vessels, possibly involving manipulation of specific inflammatory cytokines and other inflammatory mediators, is needed.