The Treatment of Giant Cell Arteritis in Different Clinical Settings

This paper aims to raise awareness of the different disease courses, comorbidities, and therapy situations in patients with giant cell arteritis (GCA), which require a differentiated approach and often a deviation from current treatment guidelines. With the approval of tocilizumab (TOC), which specifically binds to both soluble and membrane-bound IL-6 receptor and inhibits IL-6 receptor-mediated signaling, the spectrum of available effective treatment options has been significantly broadened. TOC yields an extensive range of possible applications that go beyond a glucocorticoid-saving effect. In this context, the treatment of GCA is dependent on the disease course as well as the associated comorbidities. The different stages of GCA in association to co-morbidities require a detailed treatment strategy.


INTRODUCTION
Giant cell arteritis (GCA) is an inflammatory disorder of medium-and large-size arteries affecting people older than 50 years. Classically involved vascular sites include the external carotid branches, the ophthalmic, vertebral, distal subclavian, and axillary arteries as well as the aorta. Segmentary inflammation leads to the occlusion of the vessel and to ischemic complications (1). On the immunological level, a complex interaction between the innate and the adaptive immune system as well as stromal and endothelial cells can be observed (2).

THE PATHOPHYSIOLOGY OF GCA
Histologically, GCA is characterized by an infiltration of the media with lymphocytes, macrophages, and giant cells (2). Inflammation may show a segmental infestation pattern in which inflammatory and non-inflammatory vascular segments are located side by side (3). The genesis of the disease is unknown. An association between infectious diseases (e.g., parvovirus B19, varicella zoster virus) and the occurrence of GCA is discussed (4)(5)(6). With regard to genetic causes, inhomogeneous data are available, whereby HLA-DRB1 * 04 is to be evaluated as a genetic risk factor for the manifestation of GCA (7). On the immunological level, there is a complex interaction between the innate and adaptive immune systems as well as stromal cells and endothelial cells (2). A special role is played by the interleukin-12-T-helper cell 1-interferonγ-axis and the interleukin-6-T-helper cell 17-interleukin-12 or interleukin-21 axis (8). Interleukin 6-triggered T-cell differentiation to T-helper cell 17 releases various cytokines that control local and systemic inflammatory processes (9). The activation of T-helper cells 1 by interleukin 12 leads to increased secretion of interferon γ, which leads to macrophage activation (9).
Currently, GCA pathophysiology can be diagrammed in two axes which explain the clinical symptoms, the systemic inflammatory response and the vascular occlusion (10). The systemic inflammatory response is associated with the innate immune system. Innate immune systems cells (vascular dendritic cells and monocytes) draw proinflammatory cytokines like Interleukin (IL) 6 which are associated with the production of acute phase proteins in the liver (mainly C-reactive protein) (11,12). The systemic inflammatory response is glucocorticoid and anti-IL-6 sensitive resulting in reduced clinical symptoms in GCA (11). Vascular occlusion is the ischaemic complication based on vascular remodeling. Activated macrophages or injured vascular smooth muscle cells produced growth factors that trigger vascular remodeling and a myofibroblast differentiation of vascular smooth muscle cells. The myofibroblast migrate into the intimal layer and deposit extracellular matrix proteins resulting in intimal hyperplasia and vascular occlusion in GCA (12). These vascular remodeling is not affected by glucocorticoids and anti-IL-6 therapy (12).
Despite improvements in the understanding of the GCA pathogenesis, glucocorticoids (GC) remain the mainstay treatment of this disease. Unfortunately, relapses are common when the GC dose is tapered, leading to prolonged treatment duration and increased incidence of adverse events (13). Methotrexate (MTX), azathioprine, TNF-alpha blockers, and cyclophosphamide have been proposed as GC-sparing agents or second-line therapy but with conflicting results (14,15).
Interleukin (IL) 6 plays a central role in the pathogenesis of GCA, and IL-6 serum levels correlate with disease activity and the likelihood of recurrence (16).
Tocilizumab (TOC) is a humanized monoclonal antibody that blocks IL-6 signaling by binding to the alpha chain of the human IL-6 receptor (17). The first results with TOC for treating GCA were published as early as 2011 (18). A first randomized phase II trial followed (19), and finally the randomized phase III study (GiACTA) led to the approval of TOC for the treatment of GCA in 2017 (20).
The initial treatment objective of GCA is rapid disease control by reducing the concentrations of serum acutephase reactants and freedom from symptoms as well as the prevention of ischemic organ damage. Treatment guidelines have been published by the European League Against Rheumatism (EULAR) (21), the British Society for Rheumatology (BSR) (22), and the French Study Group for Large Vessel Vasculitis (GEFA) (23).

TOPIC OF THIS ARTICLE
Within the scope of GCA treatment, different disease courses, and therapy situations can be observed, which require a differentiated approach. In addition, existing co-morbidities or GC-induced side effects may demand a deviation from therapy recommendations.
The present article will discuss these aspects and present possible treatment options, especially regarding the approval extension of TOC.
TOC does not only hold potential as a GC-sparing therapy but has already been used in numerous GCA therapy situations. While many case reports primarily focus on the use of the IL-6-receptor blocker in patients with refractory disease, GC dependence or intolerance (18), the randomized TOC studies included both newly diagnosed and recurrent patients (19,20).
In the registration study GiACTA, 119 patients with newly diagnosed GCA and 132 patients with relapsed GCA were enrolled (20). However, the study design allowed a different initial therapy. Thus, two groups of patients could be included: GC (20-60 mg) at baseline or an existing GC therapy with a duration up to six weeks (20). Therefore, it was possible to start with the combination of GC and TOC as conducted in the smaller phase II trial of Villiger et al. (19) as well as a later add-on of TOC to an initial GC monotherapy.
Various therapy settings can be divided into three main categories: 1. Treatment at new-onset 2. GC-tapering 3. Recurrence after therapy break.
In addition, further therapy situations such as GC-resistant or refractory disease (lack of response to GC-induction therapy), flare (relapse; a new relapse under therapy), GC-dependency (flare during GC reduction), or patients with co-morbidities and GC-induced side effects can be differentiated (Figure 1). Table 1 summarizes the substances investigated so far with the patient cohorts and the categorization into therapy situations as shown in Figure 1. Furthermore, the study objectives and the achievement of a GC-free therapy at the end of the trials are shown.
GCA is not a uniform disease. The latest GEFA guidelines differentiate two patient groups at the time of initial diagnosis: (23).
Uncomplicated GCA without ophthalmic involvement and without arteritis of the aorta or its branches GCA with ophthalmic involvement and GCA with aortoarteritis (non-complicated and asymptomatic involvement of the aorta or its branches) Accordingly, the GC dosage is adjusted to the risk profile; GCA patients with aortitis require a longer GC therapy due to increased flare rates (38).

Uncomplicated GCA
Although comparative studies on the most effective GC dosage are lacking, the guidelines recommend a GC dosage of 0.7 mg/kg body weight to 1 mg/kg body weight (maximum 60 mg/day) for this patient group (21,39). Within 24 to 48 h, GC therapy usually leads to a complete resolvement of acute GCA symptoms (40). Regarding the duration of first-line therapy, periods of two to four weeks are discussed (21,23).
Fifty percent of patients show a good response to GC therapy (41). Both the EULAR (21) and the BSR guidelines (22) recommend the use of GC-sparing immunosuppressants such as MTX to be considered early in therapy planning.
In the GiACTA study, 14% of patients with a GC taper over 26 weeks and 18% of patients with a GC taper over 52 weeks achieved a sustained remission of week 12 to week 52 (20). An increase in the GC dose due to a flare was not necessary.
However, in some cases permanent blindness cannot be avoided, because it takes up to five days after the start of a highdose GC therapy to control the inflammation of the Aa. ciliares posteriores (39).
Furthermore, a recent published study presented on a small study cohort the usefulness of TOC in the treatment of visual symptoms in GCA. The abstract clearly highlighted the noneffectiveness of TOC in blinded patients (42). In this context additionally, comparative studies are required to objectify this point, whereas permanent vision loss occur with an incidence up to 15% (43).
Patients with aortitis often require GC therapy with a duration of over two years and have higher flare rates, resulting in higher cumulative GC dosages (38). Therefore, GC-sparing therapy seems to be useful for this patient group. Interestingly, the subgroup evaluation of the GiACTA study showed that these patients benefited particularly from a weekly dose of TOC (18). Initial studies with a follow-up via positron emission tomography/computed tomography (PET/CT) scanning in a patient with aortitis showed a reduction in the uptake of 18 Ffluorodeoxyglucose under TOC therapy (44).

GCA With Comorbidities
None of the guidelines cited discuss a therapeutic approach in patients with a relative contraindication to GC (e.g., severe osteoporosis, diabetes mellitus type 2 difficult to control).
In patients showing GC contraindications at the time of diagnosis, an initial combination of GC and an immunosuppressive agent should be considered. For these patients etanercept and TOC was examined ( Table 1). In the GiACTA study, a significant proportion of the patients included had co-morbidities, which significantly restricted the long-term use of GC in particular (45). Patients with relapses had even higher incidences of typical GC-induced (long-term) side effects such as osteoporosis or arterial hypertension. However, a subgroup analysis on the effectiveness and safety of TOC is not yet available.

GC-Resistant GCA (GC-Refractory)
First-line therapy with GC does not always lead to disease control (which is not defined in GCA). Numerous articles have been published concerning GC-refractory disease (46), and some substances such as cyclophosphamide, anakinra, leflunomide, ustekinumab, and TOC have been investigated under study conditions ( Table 1). A rationale for the use of TOC is provided by a current work showing that GC and TOC differ significantly in their effect on regulatory T cells (T reg cells) (47). Consequently, the immunosuppressive effects of GC and TOC are not congruent and the use of TOC in GCrefractory patients is a promising option. Seventeen percent of patients enrolled in the GiACTA study showed refractory disease (45).

GC-Induced Side Effects
If intolerance or GC-typical side effects occur during GC therapy, the question of GC-sparing therapy arises. In particular, substances and studies that had the goal of completely renouncing GC after a certain treatment period should be taken into account. Publications that included patients in this therapy situation were found for TOC (36) and etanercept (30) ( Table 1). MTX is a possible option, but data from a metaanalysis indicate that the onset of action was observed not before 24-36 weeks (14).

GC-Tapering With No Flare
After an initial high-dose GC therapy, GC is slowly reduced (GC tapering), depending on the response. Once again, a uniformly accepted and in clinical studies preferred tapering schedule is not existing. Ideally, a planned GC tapering succeeds as intended and there is no flare-up of the disease. These patients also have the best chance of achieving a very low daily GC dose (< 5 mg) or finishing GC completely after 1 year. In the GiACTA study, 49% of patients with GC monotherapy had a flare within 52 weeks (20). These results support the already published data on GCmonotherapy over 52 weeks (13,48). In general, a significant number of patients requires permanent GC therapy, and the disease becomes GC-dependent (13).

Flare (Relapse) During GC-Tapering
A flare (relapse) under existing therapy should be distinguished from a recurrence (flare in the therapy free interval) (48). In some studies and publications, relapse and recurrence are not properly differentiated (23).
A flare during GC-tapering in the GiACTA study is an important criterion as in this case the GC dose could be increased at the physician's discretion. The GC monotherapy arms showed higher flare rates (GC-tapering over 26 weeks 68%; GC-tapering over 52 weeks: 49%) than the TOC study arms including a GC taper over 26 weeks (TOC q1w: 23%; TOC q2w: 26%). Consequently, patients under GC monotherapy had almost twice the cumulative GC doses than patients under combination therapy with TOC (20). Of special importance is also the disease control by a TOC monotherapy starting from the 26th study week (end of the GC taper), exhibiting a continuously low flare rate.

GC-Dependent
In 40.8% of patients, a certain GC dose cannot be undercut without a new increase in disease activity (46). Especially in this group of patients the question of an additive or intensified immunosuppressive therapy arises. Numerous studies have investigated whether an additional immunosuppressive therapy allows GC-sparing. However, a distinction should be made between studies involving patients in the GC-tapering phase (GC-dependent or side effects) and those initially starting with a combination therapy ( Table 1).
The current guidelines of EULAR (21), BSR (22), and GEFA (23) recommend MTX as a GC-sparing agent. It should be noted, however, that the available MTX studies did not include patients with GC-tapering issues but only newly diagnosed patients. In addition, the delayed onset of action under MTX must be considered (2). Furthermore, all three guidelines were published before the approval of TOC and the publication of the GiACTA results. Based on these findings, additionally studies were required which is comparing the effect of the GC-sparing drugs such as TOC and MTX in GC dependent patients.

QUALITY OF LIFE (QOL) AND PATIENT REPORTED OUTCOME (PRO)
QoL and PRO should be a central component of RCT in view of the frequent GC-induced side effects and its long-term application in GCA. In fact, only two studies (MTX, TOC) assessed SF-36 ( Table 1). MTX did not show any improvement of this parameter vs. GC monotherapy (35). In contrast, the TOC study arm revealed a significant improvement in QoL and a significant reduction in fatigue compared to GC monotherapy in the GiACTA study (20).
Since fatigue is a common side effect of MTX, its general use as GC-sparing therapy might be problematic in GCA patients (49).

RECURRENCE
A recurrence after a therapy break occurs in about one third of patients (23).

Acceptable Cumulative GC Dose (Recurrence With No Problems to Take GC)
Patients with recurrent disease, no contraindications to GC as well as no high cumulative CG-dosages (cut-off value is yet not defined) can again be treated with a GC monotherapy according to the published guidelines (21)(22)(23).

Recurrence With High Cumulative GC Dose
Patients suffering a recurrence after 1 or 2 years of GC therapy usually have a correspondingly high cumulative GC dosage and therefore require GC-sparing therapy. In this context, the EULAR guidelines recommend the same "initial" therapy with GC as for new-onset patients (21). According to the EULAR guidelines, an appropriate GC-sparing therapy should be initiated as well. A similar approach is recommended by the BSR (22). No differentiation is made between flare and recurrence (22). Additional immunosuppressive therapy (e.g., MTX) is also recommended. The GEFA guidelines could not identify any study that primarily included patients with relapsing or recurrent disease (23). Therefore, the data situation is weak, and in particular the administration of further immunosuppressants such as MTX (studies only in new-onset GCA patients) should be considered worthy of discussion (23).

Recurrence With GC-Induced Comorbidities
In everyday clinical practice, however, patients with recurrent disease are a particular challenge as they can develop typical co-morbidities within the first 2-4 years (50). This is shown in the GiACTA study because this group of patients exhibit a higher body weight and body mass index as well as more frequent depression and osteoporosis. In addition, this patient group was more frequently pre-treated with the combination therapy MTX and GC (17 vs. 2% in new-onset patients) (45).
This subgroup evaluation of the GiACTA study suggests that patients with recurrent disease benefit from therapy intensification, as sustained remissions were more frequently achieved in the two TOC study arms compared to the GC monotherapy arms (TOC-QW: 52.8%; TOC-Q2W: 47.8%; GC monotherapy 26 weeks tapering arm: 7.4%; GC monotherapy 52 weeks tapering arm: 14.3%) (45).
RCTs with this patient population are only available for abatacept, leflunomid, and TOC (20,24,33). A GC-sparing effect has been shown for TOC in a RCT and for leflunomide in a retrospective evaluation (Table 1). Further, the follow-up evaluation of the Villiger study (19) presented with a median time of 5 months after the last TOC application an GCA-flare in 55% of the patients (51).

DISEASE MONITORING
To monitor GCA under the treatment with TOC, traditional acute phase reactants cannot be used to control the disease activity, whereas GCA patients under the treatment with GC present also normal acute phase reactants at the time of disease flare (52). In this context, new biomarkers were required to evaluate disease activity and disease flare. Van Sleen et al. reported in an initial study the usefulness of serum calprotectin as biomarker for the detection of vascular inflammation (53) and Gonzalez et al. reported Osteopontin as a potential predictor for relapse in GCA (54). The value of the different imaging techniques like magnetic resonance imaging and Fluorine-18-fluorodeoxyglucose (FDG) PET/CT to detect disease activity of the Aorta and its branches, especially under GC-treatment is unclear (54,55). Consequently, new serological and immunological parameters should be evaluated to verify activity and flare of GCA.

SAFETY
The IL-6-receptor inhibitor has been approved for the treatment of rheumatoid arthritis (RA) in Europe in 2009 and its safety profile has been published several times (56). The GiACTA study revealed no new safety signals and no gastrointestinal perforations (20).

COSTS
The TOC-therapy in GCA is significantly increased compared to GC-therapy, where at the TOC-therapy is comparable to other biologic drugs in the treatment of rheumatoid arthritis. With ending of the TOC patent protection in 2020 and the implementation of TOC biosimilars the price of TOC will be marked down.

FUTURE PERSPECTIVES
In the course of GCA treatment, numerous situations arise in the clinical routine in which a deviation from therapy guidelines seems to be necessary. However, treatment options as well as co-morbidities are not sufficiently addressed in the recommendations. The approval extension of TOC results in an extensive range of possible applications that go beyond a GC-saving effect. Future studies are necessary to obtain better validated results in efficacy and safety for these treatment situations. New guidelines should also take into account the specific patient profile with regard to the therapy situation and co-morbidities.

AUTHOR CONTRIBUTIONS
AP, PO, and PH performed the literature research. AP and PH wrote the first draft of the manuscript, which was then reviewed and edited by PO.