The diagnosis of large-vessel vasculitis has experienced substantial improvement in recent years. While Takayasu arteritis diagnosis relies on imaging, the involvement of epicranial arteries by giant-cell arteritis facilitates histopathological confirmation. When appropriately performed temporal artery biopsy has high sensitivity and specificity. However, an optimal biopsy is not always achievable and, occasionally, the superficial temporal artery may not be involved. Imaging in its various modalities including colour-duplex ultrasonography, computed tomography angiography, magnetic resonance angiography and positron emission tomography, are emerging as important procedures for the diagnosis and assessment of disease extent in large-vessel vasculitis. Recent contributions to the better performance and interpretation of temporal artery biopsies as well as advances in imaging are the focus of the present review.
Temporal artery biopsy is the most sensitive and specific procedure for the diagnosis of giant-cell arteritis.
Imaging techniques including color-duplex ultrasonography, computed tomography angiography, magnetic resonance angiography and positron emission tomography are emerging as useful diagnostic techniques by exploring additional vascular territories both in giant-cell arteritis and in Takayasu arteritis.
Imaging may have a role in assessing disease activity and response to therapy that deserves to be further investigated.
Primary large vessel vasculitides (LVVs) are granulomatous inflammatory conditions involving large and medium-sized arteries and include 2 related but distinct entities: giant-cell arteritis (GCA) and Takayasu arteritis (TAK). Both conditions have remarkable anatomic and histopathologic similarities but also have substantial differences, mainly in demographic characteristics, disease distribution, and response to therapy. GCA predominates in elderly white people, particularly in native peoples or descendants from northern Europe, whereas TAK primarily occurs in young women and is more frequent among people of Asian or Native American ancestry. In addition, the spectrum of artery sizes targeted by GCA seems to be broader than that affected by TAK. Although TAK usually affects the aorta and its primary and secondary tributaries, the spectrum of vascular involvement by GCA ranges from the aorta to small epicranial arteries and small arteries supplying the optic nerve and retina ( Fig. 1 ). Moreover, large vessel involvement is subclinical in most patients with GCA, in whom combinations of cranial, polymyalgic, or systemic symptoms dominate the clinical picture. Development of symptomatic stenoses in large or medium-sized vessels is uncommon in GCA. In contrast, typical TAK features are related to symptomatic stenosis of the primary or secondary branches of the aorta.
The characteristic involvement of the epicranial arteries facilitates the histopathologic diagnosis of GCA, which is usually obtained through superficial temporal artery biopsy (TAB). When performed in optimal conditions, TAB has a remarkable sensitivity and provides the most definitive diagnosis of GCA. However, a TAB is not always feasible and, occasionally, the superficial temporal artery may not be involved. Imaging in its various modalities, including color duplex ultrasonography (CDUS), contrast-enhanced computed tomography (CT) with or without CT angiography (CTA), magnetic resonance imaging (MRI) with or without angiography (MRA), and positron emission tomography (PET) with or without CT (PET/CT), are emerging as seminal procedures for the diagnosis and evaluation of disease extent in patients with GCA. In TAK, diagnosis almost invariably relies on imaging techniques because of the inaccessibility of involved vessels.
Temporal Artery Biopsy
TAB is the gold standard for GCA diagnosis by showing the typical histopathologic findings, namely mononuclear cell infiltration of the artery wall (see Fig. 1 ). Giant cells are present in more than 50% of biopsies but their presence is not mandatory to establish the histopathologic diagnosis of GCA. Additional features, including fragmentation of the internal elastic lamina and intimal hyperplasia, are typical but not sufficient to confirm the diagnosis in the absence of inflammatory infiltrates. To improve the diagnostic accuracy of the TAB, several issues need to be considered.
Reducing unnecessary biopsies
About 70% of TABs performed in referral centers are negative. In most patients an alternative diagnosis is obtained after additional work-up. As a result, in a substantial proportion of patients TAB could be avoided. Several investigators have analyzed clinical data associated with TAB results. The absence of manifest or subtle cranial symptoms or abnormalities at the careful temporal artery examination is associated with low probability of a positive TAB even in patients who really have GCA. In these cases other diagnostic approaches may achieve higher yield. However, a negative biopsy may be clinically helpful in driving attention toward other potential diagnoses. Moreover, a negative TAB may have an important clinical value by reasonably ruling out a treatable disorder in pluripathologic geriatric patients with constitutional symptoms.
Improving temporal artery biopsy sensitivity
In about 10% to 15 % of patients with a negative biopsy GCA is still suspected and patients are treated accordingly in spite of the TAB results. TAB may be mistakenly negative when it is not performed or examined in optimal conditions and its sensitivity may be enhanced by improving those conditions. Because lesions may not be homogeneously distributed along the temporal artery, a lengthy fragment must be excised to achieve the maximal diagnostic performance. It has been suggested that a 0.5-cm length after formalin fixation may be sufficient but better sensitivity is obtained with fragments 2 cm or longer. Note that specimens contract around 20% or more after excision. Multiple sectioning and careful examination of multiple sections is crucial to increase the sensitivity of the TAB because lesions may be segmental.
Awareness by the examining pathologists of subtle abnormalities is also crucial. Early inflammatory changes including infiltrates surrounding vasa vasorum or inflammation of periadventitial vessels surrounding a spared temporal artery must be considered as relevant findings supporting GCA diagnosis in an appropriate clinical context ( Fig. 2 ). Sometimes, these incomplete findings are the only abnormality observed, but in other instances further sectioning reveals more characteristic features. However, incomplete findings may be equivocal: involvement of periadventitial vessels can be seen in other vasculitides affecting medium or small vessels (polyarteritis nodosa or anti-neutrophil cytoplasmic antibody (ANCA) associated vasculitis). Furthermore, slight inflammatory changes in small vessels or vasa vasorum can be seen in severe chronic infections such as endocarditis or in malignancies, possibly as a result of pattern recognition–mediated activation of endothelial cells or systemic release of inflammatory mediators. There is a gray diagnostic zone in specimens with slight inflammatory changes and, when these are observed, additional diseases must be considered and additional diagnostic work-up must be performed.
Several investigators have raised the question of whether performing bilateral TABs would increase the diagnostic sensitivity of TAB. Discordant results between TABs performed on both sides are obtained in 0% to 12.7% of cases according to various studies. Therefore, routine performance of simultaneous bilateral TAB to increase diagnostic sensitivity is not advised, although performance of a second TAB when the first one is negative can be considered in particular patients with strong suspicion of GCA or in whom obtaining a definitive diagnostic confirmation is considered crucial.
When GCA is strongly suspected a prompt initiation of glucocorticoid treatment is advised, particularly in patients with visual symptoms. There has been some concern regarding how previous glucocorticoid treatment may influence the diagnostic accuracy of TAB by clearing inflammatory lesions. This relevant issue has been addressed in several studies. It seems clear that glucocorticoid treatment for a period of 8 weeks or less does not reduce the diagnostic yield of TAB. However, over the months, lesions evolve to a healing or obsolescent stage with prominent fibrosis, vascular remodeling, and remaining scattered small foci of inflammatory cells. In some patients this pattern is observed before the initiation of glucocorticoid treatment, perhaps suggesting that the disease may have started earlier than when clinically apparent. Failure to recognize this pattern may also decrease the diagnostic performance of TAB.
Because TAB performance in optimal conditions and examination by expert and sensitized pathologists may not be invariably available, some investigators have emphasized the clinical diagnosis of GCA as the gold standard. Many investigators have used the 1990 American College of Rheumatology (ACR) classification criteria or modified versions as diagnostic criteria, even in clinical trials. However, the ACR criteria were not intended for this purpose and do not perform satisfactorily as diagnostic criteria. Many conditions in elderly people convey headache, (i.e. anemia of any origin, depression) and moderate increase of acute-phase reactants and 3 ACR criteria can easily be fulfilled. In addition, about 10% to 20% of patients lack cranial symptoms and present with atypical features, such as fever of unknown origin, unexplained weight loss, or anemia of chronic disease type, and could never be diagnosed with GCA based exclusively on clinical grounds. Moreover, clinical diagnosis relies heavily on the experience of evaluating physicians and may have an important interobserver variability. Reported series of patients with GCA considered refractory to high-dose glucocorticoid treatment contain an exceedingly high rate of patients with negative/not performed biopsies, raising the question of whether some of these patients could have a mimicking disease.
Diagnostic contribution of molecular pathology and serum biomarkers
Investigators have attempted to increase the sensitivity of routine examination of hematoxylin-eosin–stained slides by immunohistochemistry methods.
Immunostaining for T lymphocyte (CD3) and macrophage (CD68) markers may improve visualization and identification of inflammatory cells. However, small numbers of immunostained cells, particularly macrophages, may be observed in patients who are eventually diagnosed with other conditions. The meaning of immunohistochemical detection of scattered inflammatory cells is still unclear.
Several inflammatory biomarkers of GCA have been identified in tissue but their sensitivity and specificity needs to be evaluated. Circulating antiferritin antibodies have been detected in patients with both GCA and TAK. The potential role of their detection in diagnosis needs to be specifically tested and validated.
The value of temporal artery biopsy in diagnosing other conditions
As mentioned earlier, TAB occasionally shows vascular inflammation in the temporal arteries or collateral branches in vasculitides other than GCA that may present with similar symptoms. Involvement of the temporal artery or its branches by ANCA-associated vasculitis or polyarteritis nodosa has repeatedly been reported. TAB sometimes discloses histopathologic features typical of other diseases that may present with clinical signs or symptoms similar to GCA. Amyloidosis, thromboangiitis obliterans, and Kimura disease have been described in temporal artery specimens.
Timely performance and examination of TAB in optimal conditions is not universally feasible. Moreover, superficial temporal arteries may not invariably be involved and, as discussed earlier, interpretation of TAB results has some limitations. Along with the development, improvement in resolution, and widespread use of imaging techniques, these limitations have led to an active investigation of the role of imaging techniques in the diagnosis and assessment of patients with LVV.
Imaging may have various relevant roles in the evaluation of patients with LVV ( Box 1 ). It may contribute to diagnosis of vascular inflammation by detecting thickening of the involved vessels or radiotracer uptake but it is also useful in assessing disease extent by exploring multiple vascular territories. Moreover, imaging may contribute to the evaluation of treatment efficacy and may constitute a valuable outcome measure, although this aspect has not been explored in depth. However, except for color duplex ultrasonography (CDUS), most existing studies consist of case reports, small series, or post hoc or retrospective analysis including heterogeneous patients in different activity states. Prospective, controlled studies using MRA, CTA, or PET are still scarce and distinct aspects such as diagnosis, extent, evaluation of disease activity, response to treatment, and damage/scarring have not been clearly defined or specifically addressed.