Extra-Cranial Involvement in Giant Cell Arteritis

*João Fernandes Serôdio, Miguel Trindade, Catarina Favas and José Delgado Alves*

#### **Abstract**

Recent advances in imaging studies and treatment approaches have greatly improved our knowledge about Giant Cell Arteritis (GCA). Previously thought of as a predominantly cranial disease, we now know that GCA is a systemic disease that may involve other medium and large vessel territories. Several imaging studies have shown that between 30 and 70% of patients with GCA present with largevessel vasculitis. Moreover, a significant proportion of patients present large-vessel disease in the absence of cranial involvement. Extra-cranial disease also poses management challenges as these patients may have a more refractory-relapsing disease course and need additional therapies. Aortic dilation and aneurysms are well-described late complications of GCA involving the large artery territories. In this chapter, we discuss the clinical picture of extra-cranial involvement in GCA, focusing on improved diagnostic protocols and suitable treatment strategies.

**Keywords:** giant cell arteritis, large-vessel vasculitis, polymyalgia rheumatica, vasculitis, diagnostic imaging

### **1. Introduction**

Giant cell arteritis (GCA) is a systemic vasculitis that predominantly involves large and medium-size arteries [1]. It occurs almost exclusively in subjects aged 50 years or older, and is the most common form of systemic vasculitis among the elderly [2]. GCA is more common among caucasian female patients, with a female– male ratio of about 2–3:1. The GCA annual incidence varies with geographical location and ranges from 1.6 to 32.8 cases/100000 persons ≥50 years of age [3].

GCA is commonly defined as Large-Vessel (LV) GCA if the aorta and its branches are involved. The systemic nature of the disease was noted as early as the first cases described by Horton and colleagues in 1932 [4]. Later on, Gilmour suggested that the disease should be called "giant-cell chronic arteritis" as the temporal arteritis appeared to be only part of a more widespread vascular disease [5]. Despite this early reports, physicians have mainly focused on typical cranial symptoms and visual disturbances and have relied mostly on temporal biopsy for diagnosis. This focus is well reflected in the 1990 ACR classification criteria that emphasised the importance of headache as a cardinal symptom and temporal biopsy as its primary diagnostic tool [6]. Unfortunately, the concept of GCA as a limited cranial disease is inaccurate and obscures essential clinical features. Furthermore, the misuse of classification criteria for diagnostic purposes, may lead to underdiagnose LV-GCA [7].

In recent years there has been an increased awareness of the systemic largeartery nature of GCA. Necropsy studies have shown histologic evidence of systemic large-artery vasculitis in approximately 80% of patients [8, 9]. Recent advances in diagnostic imaging techniques have confirmed these figures, suggesting that imaging will have an increasing impact in the diagnosis and management of GCA [10–14]. Furthermore, patients with GCA are at increased risk of developing aortic dilation and aneurysms among other complications [15–17].

Altogether, these issues highlight the importance of the extra-cranial involvement of GCA which has been under-recognised and poorly managed.

### **2. Pathophysiology**

GCA is an idiopathic inflammatory granulomatous vasculitis. The aetiology is unknown, and most probably, genetic, environmental, vascular, and age-related factors concur to the development of the disease [2, 18]. In GCA, a lymphocyte and plasma cell infiltrate originates at the *vasa vasorum* in the adventitia of large vessels, which then penetrates the vessel wall leading to an intimal and media hyperplasia and vessel wall thickening [19]. Multinucleated giant cells form a complex near the intima-media complex, but they are not a requisite for diagnosis. Inflammation can be segmental, circumferential, or transmural [9, 20]. The predominance of GCA by some vessel territories and the mechanisms behind the different phenotypes like LV-GCA are still unsolved questions. In fact, most studies have been performed in temporal artery biopsies, as large arteries are not as readily accessible for histologic examination. Animal models also present limitations regarding the expression of the disease in different vascular territories. The interaction of immunopathogenic mechanisms with the different functional and anatomic characteristics of the vessel walls in different parts of the body may explain the distinct aspects of LV-GCA pathophysiology.

#### **2.1 Immunologic mechanisms in large vessel giant cell arteritis**

The critical event in initiating and sustaining the inflammatory response is thought to be the abnormal maturation and loss of tolerance of vascular Ddendritic cells (DCs), which is triggered by toll-like receptors (TLRs) [21, 22]. Differentiated DCs drive T cell and macrophage recruitment [21]. Upon the maturation of DCs, CD4+ T cells are also stimulated by local cytokines, such as IL12, to polarise into T-helper 1 (Th1) and IL6 and IL23 to polarise into Th17 cells [23].

TH17 cells are responsible for implementing a strong acute IL17 mediated inflammatory response, which leads to the overproduction of a cluster of cytokines, namely IL1β, IL6, IL23 and TNF-α [23]. Type II cytokine receptors (mainly IL6 and IL1β) signal through JAK1 homo-dimers [24] promoting further cellular activation and inflammatory response. The IL17 pathway is therefore responsible for most of the inflammatory response in the acute phase and explains the systemic nature of the disease [25, 26].

Th1 cells differentiation induces an immune response where IFN-γ is the central cytokine [27]. IFN-γ receptor signals through JAK1–JAK2 heterodimers [28]. The INF-γ signature further enhances the inflammatory response (through IL1β, IL6, and TNF-α), leading to macrophage differentiation and activation. Upon the stimulation by the granulocyte-macrophage colony-stimulating factor (GM-CSF) produced by T cells, macrophages act in sustaining inflammation and are key players in the interaction with the stromal and extracellular matrix [29, 30]. This interaction is mediated by matrix metalloproteinases (MMP) and several growth factors.

*Extra-Cranial Involvement in Giant Cell Arteritis DOI: http://dx.doi.org/10.5772/intechopen.97715*

MMP are proteases with elastolytic activity, released and activated by inflammatory cells. Smooth muscle loss and proteolytic imbalance may contribute to elastic fibre rupture, weakening of the artery wall, and cell migration [29, 31]. The IFN-γ signature is responsible for the histiocytic reaction, myofibroblast differentiation, intimal hyperplasia, neoangiogenesis, vascular remodelling, damage, and fibrosis [32]. These aspects explain the vascular manifestations and the LV complications of GCA. Current treatments efficiently inhibit the Th17-mediated response, but not the Th1 mediated expression of IFN-γ [27, 33]. Thus, the current management of vascular manifestations like artery stenosis and aneurysms is suboptimal, as vascular remodelling processes may subsist even in the absence of raised inflammatory markers [34].

Patients with polymyalgia rheumatica (PMR) present activated DCs in focal vessel infiltrates with the expression of inflammatory cytokine production (IL1β and IL6), but IFN-y is absent [35]. Therefore, it is thought that it is the IFN-γ pathway, and not IL17 activation that marks the progression to overt vascular inflammation and remodelling.

It is not yet clear why some patients have only PMR while others progress to periadventitial or transmural vasculitis. Different TLR expression on DCs may partly explain such patterns as TLR4 activation induces transmural panarteritis, while TLR5 ligands promote adventitial perivasculitis [36]. Moreover, DCs exhibit distinct combinations of TLRs in different vascular beds [37]. Thus, the phenotype of the vasculitis may depend upon the profile of the TLR driven T cell activation, which is specific of each vascular territory.

The interaction between T cells and B cells might also be implicated in the expression of LV-GCA. Recent findings in aorta tissue samples from 9 LV-GCA patients who underwent aortic aneurysm surgery, showed massive infiltration of B-cells, which outnumbered T-cells. B-cells were mainly found in the adventitia and were organised into tertiary lymphoid organs [38]. This is an uncommon observation in temporal artery biopsies.

The interaction of immune mechanisms and the vascular matrix is also demonstrated by the MMP expression in singular vascular fields. MMP2 tissue expression was observed in active temporal artery lesions and in aortic aneurysm samples obtained in 2 GCA patients. However, MMP9 was present only in temporal artery lesions and faintly detectable in normal temporal arteries and GCA-related aneurysms [17]. While MMP9 is mainly produced by inflammatory cells, MMP2 may also be expressed in smooth muscle cells and be involved in reparative mechanisms. Therefore, the expression of MMPs on different vascular beds may also impact on the clinical features of GCA.

#### **2.2 Atherosclerosis, ageing and large vessel vasculitis**

Atherosclerosis is highly prevalent among GCA patients as it is most present at an advanced age. The coexistence of these two diseases and the underlying immune mechanisms of both may tailor the phenotype of the vasculitis. It is known that patients with cardiovascular risk factors have a higher risk of developing severe ischaemic manifestations of GCA [39]. In fact, patients with ischemic complications have lower expression of IL6 suggesting that IL6 may play a protective angiogenic role to compensate for ischemia [40]. Furthermore, at the supra-aortic level, atherosclerosis most commonly affects the carotids, while LV-GCA predominantly affects the axillary arteries. Regardless of the immune profile, age and genetic factors also influence the development of atherosclerosis. In caucasians, atherosclerosis occurs later and less extensively in intracranial arteries compared to extracranial arteries. Interestingly, Asian and African populations are more affected by

intracranial atherosclerosis and also show a low prevalence of cranial GCA [3, 41]. Thus, atherosclerosis may alter vessel vulnerability or expression of GCA.

Age is an important factor that affects vascular and immune processes with a possible impact on disease vulnerability and manifestations [42]. Ageing induces significant changes in the expression of vascular MMP2 and MMP9 and reduces arterial smooth muscle proliferative capacity [43–45]. One of the main distinctions between LV-GCA and Takayasu arteritis (TAK) has been attributed to an age cutoff. Interestingly, TAK shows similar immunologic mechanisms with dysregulated activation of Th1 and Th17 pathways [46] and therefore age-related factors may be the key to explain the distinct manifestations between LV-GCA and TAK [20, 42].

## **3. Clinical features of large vessel giant cell arteritis**

#### **3.1 Clinical manifestations**

LV-GCA usually presents with prominent constitutional symptoms and a marked increase in inflammatory markers. Systemic constitutional symptoms include fever, malaise, weight loss and night sweats. Symptoms are usually nonspecific and, in up to 20% of the patients, systemic constitutional symptoms are the only clinical features of the disease with some cases being diagnosed following an investigation for fever of unknown origin [10, 18]. Aortitis is a common feature in LV-GCA. Aortitis is often pauci-symptomatic, but some patients may refer chest or back pain [18]. LV-GCA also affects the main arteries of the limbs, presenting most commonly as limb claudication. Limb claudication reflects intimal and muscular hyperplasia secondary to vascular inflammation, which leads to vessel wall thickening with lumen occlusion. Limb claudication involves the arms more frequently than the legs and may be present in up to 50% of LV-GCA patients. It can be intermittent and asymmetric despite vascular involvement being bilateral in around 80% of the patients [7, 10, 47].

The preferred vascular territories involved are the supra-aortic branches, particularly the axillary and subclavian arteries, which are involved in almost all patients with LV-GCA. Carotid and vertebral artery involvement are less frequent. Aortitis is present in around 50–65% of the patients with documented LV-GCA. Most commonly, it involves the aortic arch and the thoracic descending aorta. When the abdominal aorta is affected, there is usually involvement of the thoracic segment as well. Femoral arteries and inferior limb arteries are involved in only around 10–15% of the patients. Sometimes differential diagnosis with atherosclerosis, very commonly found in these arteries, may be difficult. Visceral arteries are rarely affected [7, 10, 12, 47–49].

#### **3.2 Clinical overlap between large vessel vasculitis, cranial giant cell arteritis and polymyalgia rheumatica**

There is a considerable clinical and epidemiologic overlap between GCA and PMR. PMR is a clinical syndrome characterised by bilateral shoulder pain, morning stiffness, shoulder or pelvic girdle weakness, and peripheral arthralgia/ arthritis [2]. Approximately 20% of PMR patients have GCA, whereas PMR is present in up to 60% of GCA patients [2, 50, 51]. PMR is also the main form of relapse in up to 50% of GCA patients, while cranial symptoms are relatively uncommon at relapse [52]. Interestingly, Positron Emission Tomography (18FDG-PET) LV fluorodeoxyglucose increased uptake was noted in 30% of patients with isolated polymyalgia rheumatica at diagnosis [53]. Therefore, PMR patients with

#### *Extra-Cranial Involvement in Giant Cell Arteritis DOI: http://dx.doi.org/10.5772/intechopen.97715*

incomplete response to corticosteroid treatment or a relapsing disease should be re-evaluated for LV involvement.

Patients with LV-GCA are more frequently women and present at a younger age, whereas patients with cranial GCA are usually men and older [7, 10, 54]. When compared with cranial GCA, patients with LV-GCA present less frequently with headache (35% in LV-GCA vs. 60% in cranial GCA), jaw claudication (22% in LV-GCA vs. 50% in cranial GCA) and also with fewer cranial ischemic symptoms (permanent visual loss in 4% in LV-GCA vs. 20% in cranial GCA) [10, 55–57]. Although there may be specificities concerning the presentation of cranial GCA and LV-GCA, they are not distinct entities (**Table 1**). More likely, we are facing a different spectrum of the same disease (**Figure 1**). Depending on the different imaging techniques used, 32–83% of the patients with confirmed cranial GCA also have LV vasculitis [10–12, 14] and 10–30% of the patients with GCA have only LV vasculitis, with no clinical, histologic or Doppler evidence of temporal artery vasculitis [10, 48, 58, 59].


*GCA, Giant Cell Arteritis; LV, Large Vessel; PMR, Polymyalgia Rheumatica;* −*, uncommon symptom or sign; +, common symptom or sign; ++, very common symptom or sign.*

#### **Table 1.**

*Clinical symptoms and signs in different subtypes of GCA and in PMR.*

**Figure 1.** *The clinical spectrum of cranial GCA, LV-GCA and PMR.*

Due to the more unspecific nature of the clinical presentation of LV-GCA, the diagnosis is often delayed or even missed. In general, patients with isolated LV-GCA have a delay in the diagnosis greater than one year compared with patients with cranial GCA [7]. It is still unknown whether this delay in diagnosis and treatment may impact the clinical course of the disease. However, LV-GCA patients relapse more frequently and earlier than those with cranial GCA and have higher corticosteroid cumulative doses and more frequently require additional immunosuppressive treatments [7, 54]. These facts suggest that patients with LV-GCA should be considered for a different management and treatment strategy, with a more tailored, eventually more aggressive approach.

## **4. Differential diagnosis**

The systemic LV involvement in GCA may resemble the presentation of Takayasu arteritis (TAK). Patients with Takayasu's disease may present with raised inflammatory markers, vascular bruits, asymmetric blood pressure measurements and limb claudication, much like patients with LV-GCA. The recent widening of the concept of vascular involvement in GCA shows that there can be an overlap between these two conditions. However, some have proposed clear distinctions. Most importantly, the epidemiology is quite different. GCA is recurrent among northern European patients, whereas TAK is more prevalent among the Asian population [60]. Another difference is the age of disease onset. GCA is almost exclusively present in patients 50 years or older, whereas TAK is common under 40 [2, 61]. However, some argue age restriction to be arbitrary and without etiologic or pathophysiologic basis [20]. In a study of 96 Japanese patients with TAK, 22% were outside the proposed age cut-off [62]. Likewise, in the study that defined the 1990 ACR Classification Criteria for GCA, 23% of the patients had less than 50 years old at diagnosis [6]. Moreover, these definitions are elusive for patients with LV vasculitis aged between 40 and 50 years. So, distinguishing GCA and TAK based only on age and epidemiology may be difficult suggesting that we might, in fact, be looking at two forms of the same disease [63].

The histopathologic findings in both GCA and TAK show a lymphohistiocytic infiltrate in the vascular wall that may be indistinguishable [20]. However, this observation may be biased due to the small number of patients undergoing vascular biopsy in TAK. Pathophysiologic mechanisms also show common features between both diseases [42, 46]. Clinically, TAK presents with a more widespread vascular involvement. The carotid and mesenteric arteries are more frequently affected in patients with TAK than GCA, while subclavian and axillary artery involvement is more prevalent in LV-GCA [63, 64]. The aortic involvement is also distinct since stenotic/occlusive lesions are predominant in TAK, whereas aneurysmal disease is more common in GCA [64]. It is unclear, however, if the differences in imaging findings represent cumulative damage due to delay in TAK diagnosis or whether other age-related immunologic and vascular factors may explain these differences. Lastly, inflammatory markers seem to be higher in patients with GCA than in TAK. Around 44% of the patients with TAK may have active vascular inflammation despite normal inflammatory marker values [65].

Several cases of small and medium vessel vasculitis have been described with temporal artery involvement, particularly granulomatosis with polyangiitis and eosinophilic granulomatosis with polyangiitis [66]. However, the presentation of ANCA-associated vasculitis with aortitis is extremely rare [67] and even more so in other forms of primary vasculitis.


#### **Table 2.**

*Differential diagnosis of large-vessel giant cell arteritis.*

Other systemic diseases present with aortitis and may also be mistaken with LV-GCA (**Table 2**). Some infections like syphilis or sub-acute endocarditis may evolve with aortitis [68, 69]. In these cases, serologic and microbiologic studies often guide the diagnosis. Other immune-mediated diseases also have aortitis as a prominent clinical feature such as Behçets disease, IgG4-related disease, and Erdheim-Chester disease. These entities often have other distinctive organ involvement and typical histologic findings pointing to a different diagnosis [69–71]. Also, in IgG4-related and Erdheim-Chester diseases, aortic involvement occurs as periaortitis and retroperitoneal fibrosis which is different from vascular inflammation. Aortitis may also be a late complication of ankylosing spondylitis. It often involves de aortic root or the iliac periaortic peritoneum. It presents late in the disease, and articular symptoms often precede it by years. With recent advances in treatment, it is expected that it will become a less common manifestation of the disease [72].
