Large vessel vasculitis: the search for response biomarkers

KEYWORDS:

• Large vessel vasculitis
• Takayasu arteritis
• giant cell arteritis
• T cells
• biomarkers
• imaging

1. Background

The most common large vessel vasculitides (LVV) in adults are giant cell arteritis (GCA) and Takayasu arteritis (TA) [1]. The former is seen in those >50 years, while the latter typically affects those under 40 years. GCA and TA share histologic features, including focal granulomatous pan-arteritis characterized by lymphomonocytic infiltration and occasional multinucleate giant cells. The resulting intimal hyperplasia predisposes to luminal obstruction, arterial wall remodeling with disruption of elastic laminae, excessive deposition of extracellular matrix, and fibrosis. Understanding of LVV pathogenesis remains limited and all aspects of clinical management are suboptimal. The few available clinical trials are limited to GCA, and TA remains an orphan disease. Diagnosis is typically delayed and understanding of pathogenesis is poor. Treatment options are limited, with undue reliance upon corticosteroids resulting in significant side effects [2–4]. Disease complications can be severe and life-threatening. The lack of disease-specific biomarkers is a critical contributor to this rather dismal picture. However, recent renewed efforts raise the hope of significant advances in disease management. The focus of this editorial is to briefly present our view of biomarker research in GCA and TA, presenting recent advances and the main unmet needs and pitfalls.

2. Biomarkers may improve diagnosis and identify homogenous LVV subsets

The heterogeneous nature of GCA and TA in terms of presentation, disease course, and pattern of arterial involvement may delay diagnosis. Noninvasive imaging biomarkers can help address this problem [5]. ¹⁸F-fluorodeoxyglucose positron emission tomography (FDG-PET) identifies metabolically active cells to reveal active arterial wall inflammation in active GCA and TA [2,4], proving particularly useful for demonstrating aortitis in older patients with systemic symptoms and the absence of classical GCA symptoms including headache and jaw claudication. Likewise, high-resolution Doppler ultrasound can reveal the hypoechoic halo sign and circumferential wall thickening characteristic of temporal arteritis [4].

Imaging data may also inform a current area of debate, namely whether GCA and TA represent a single spectrum of disease or distinct entities [6]. GCA was classically considered to affect extracranial large and medium-sized cephalic arteries, while TA involves the aorta, the pulmonary arteries, and their main branches. However, wider use of imaging has highlighted a fact previously recognized in post-mortem studies, that 60% of patients with GCA have evidence of disease affecting the aorta and its major branches. A significant proportion of these patients do not suffer symptoms associated with temporal arteritis and hence a new poorly characterized disease subtype ‘large-vessel GCA’ has evolved. The identification of novel biomarkers is now required to improve the nosography of LVV and guide accurate clinical diagnosis. Improved classification is essential for understanding potential subtype differences in pathogenesis, clinical course, and response to therapy, and for the design of clinical trials involving homogeneous populations [7].

3. Genetic biomarkers

Further genomic analysis represents an important component of the future work and is likely to identify more than the three predominant phenotypes mentioned above. This approach has the potential to stratify patients in terms of predisposition to arterial stenoses, aneurysmal dilatation, subdiaphragmatic arterial involvement, blindness, and response to therapy. Currently available data have revealed differences between GCA and TA both at MHC and at non-MHC loci [8,9] and further results are awaited with interest.

4. Disease activity biomarkers

Definition of the optimal assessment of LVV disease activity and response to treatment remains an unresolved conundrum [10]. Clinicians are unduly reliant upon erythrocyte sedimentation rate (ESR) and C-reactive protein (CRP) levels. Recent research has sought methods that distinguish active and inactive diseases according to multi-item activity indices. For TA, these include the National Institute of Health (NIH) criteria and the Indian Takayasu Activity Score (ITAS) [10]. Although these measures have proved useful, further work is required for both TA and GCA. However, the search for novel plasma biomarkers has proved frustrating. Plasma levels of cellular adhesion molecules and coagulation-related proteins, such as von Willebrand factor and tissue factor, were unrelated to activity status [11]. Other potential biomarkers, including matrix metalloproteinases, interleukin (IL)-6, IL-8, and IL-18, have been investigated in relatively small studies, predominantly yielding data concerning measurement of disease activity. Similarly, N-terminal pro-brain natriuretic peptide (NT-proBNP) levels associate with active disease, while in stable disease, NT-proBNP may act as an indicator of disease severity and damage and merits further prospective study [12].

We believe that future biomarker discovery is dependent upon improved understanding of LVV pathogenesis. Novel biomarkers for different phases of disease need to be identified and used in combination with clinical examination and noninvasive imaging data, to generate multi-dimensional disease activity and damage scores. To date, most biomarker studies have used disease activity as a reference. However, disease activity per se may not reliably reflect the evolution of arterial involvement. Indeed, patients thought to have inactive disease with normal acute-phase reactants may experience progressive vascular involvement. This occurrence may be particularly frequent in patients treated with IL-6 receptor antagonist [13]. Thus, on occasion, arterial progression may reflect local inflammatory responses and/or tissue remodeling, rather than systemic inflammation.

Regardless of whether inflammatory or noninflammatory progression has occurred, the resultant intimal hyperplasia in both GCA and TA represents a stereotyped arterial response to injury driven predominantly by myofibroblasts, with the subsequent fibrosis extending into the media and adventitia of involved arteries [14]. Adventitial involvement is more prominent in TA and intimal changes in GCA. Importantly, these mechanisms are likely to yield novel biomarkers and may ultimately be identified with high-resolution imaging.

5. Future approaches – disease activity versus arterial progression

We propose that arterial progression rather than disease activity should be used as the reference when designing LVV candidate biomarker studies. Assessment of arterial morphology with ultrasonography (US), computed tomography (CT), or magnetic resonance (MR) would be needed for this purpose. An unmet need in this regard is standardization of the definition of arterial progression and more than a single definition might be required. Thus, for clinical studies requiring hard clinical outcomes, focusing on a luminal definition of arterial progression might be ideal, as it is directly linked to adequacy of blood flow in large arteries. In contrast, pathogenic studies focusing on the mechanisms of inflammation and remodeling require arterial wall data including assessment of wall thickness, enhancement, and lesion length.

Novel biomarkers should be challenged with providing nonredundant information beyond disease activity, ideally reflecting the activation of pathways at least partially independent of pro-inflammatory cytokines such as IL-6. One example is pentraxin-3 (PTX3), a long pentraxin generated locally at sites of inflammation where it modulates tissue remodeling and repair. PTX3 levels do not appear to correlate with ESR, CRP, NIH activity criteria, or ITAS [15,16], but with arterial progression and local inflammation, as assessed by imaging in TA [16] and with ischemic events in GCA [17]. In further longitudinal studies, PTX3 should be assessed as a potential indicator of local progressive arterial injury or remodeling.

Studies of GCA have revealed pathogenic roles for T-cell and neutrophil populations and may point the way to new biomarkers. Recent data suggest that Th9 and Th17 cells characterize distinct lesion patterns in GCA [18]. Comparison of Th1 and Th17 cells shows that the former are less steroid-sensitive and associate with arterial remodeling and relapsing disease, while Th17 cells are steroid-sensitive and responsible for IL-6 secretion and the systemic inflammatory response [5]. In contrast, Th1 cells were more steroid responsive than Th17 cells in TA [19]. Thus, disease-specific interplay between these two T-cell populations and their utility as biomarkers remains to be elucidated. Neutrophils exert a suppressive effect on T-cell proliferation in GCA which is lost during steroid tapering, indicating a potential pathogenic role for neutrophils and a novel avenue for biomarker research [20].

Large vessel involvement in GCA remains poorly understood. A major unresolved issue is the identification of those patients at risk of progressive dilatation of the ascending aorta despite apparent clinical remission. It is unknown whether hemodynamic factors drive progressive dilatation of an arterial wall previously weakened by inflammation or whether smoldering local disease also contributes. Although no biomarker has been identified, enhanced aortic uptake during the inflammatory phase detected by FDG-PET scanning may predict late aortic progression, and so identify a subpopulation of patients suitable for biomarker studies [21,22].

6. Novel imaging approaches

Given the limitations of an approach confined to clinical and laboratory assessment of LVV, an ideal future scenario might involve assessment based upon imaging biomarkers reflecting local pathogenic events, alongside matched clinical and plasma biomarker data. State-of-the-art imaging techniques allow identification of pathology associated with tissue inflammation, including neovascularization revealed by contrast enhancement during CT, MR, and contrast-enhanced US with micro-bubbles. The latter is relatively inexpensive, widely available, and reproducible [23]. However, analysis is somewhat limited by being confined to the carotid, temporal, and subclavian/axillary arteries. MR may reveal additional inflammation-related features in the arterial wall such as edema. Co-registration of PET with CT improves anatomical signal localization and differentiation of arterial and periarterial tracer uptake. Improving PET resolution and the emergence of PET-MR imaging offers exciting possibilities in the vasculitides. Precise identification of active arterial wall inflammation, combined with sensitive demonstration of treatment responses, is the ‘holy grail’. However, consensus has yet to be reached on how closely imaging findings reflect arterial wall inflammation, remodeling, and LVV disease activity, and so prospective analyses are awaited with interest.

Given the limitations of using clinical disease activity as a reference, it remains to be determined whether functional characterization of arteritis using novel PET ligands will provide data independent of clinical activity and reflective of key pathogenic events including progression. In this regard, [¹¹C]-PK11195 binds a receptor expressed by activated macrophages and sensitively detects active LVV in GCA and TA [24].

7. Summary

In conclusion, there are no biomarkers universally accepted for detection of LVV activity or response to therapy. This in part reflects limited understanding of disease pathogenesis and inability to define homogenous patient groups. However, we believe that prospects are good and changes in the reference points used for clinical studies, alongside a combined genomic, cell and molecular biology and imaging approach, will ultimately result in accurate stratification of patients and to the effective use of targeted biologic therapies.

Declaration of interest

J Mason has received travel support or participated in medical board meetings with Pfizer and Roche. The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed.

Additional information

Funding

This paper has been supported by the National Institute for Health Research (NIHR) Biomedical Research Centre Funding Scheme at Imperial College.