HLA-B52 allele in giant cell arteritis may indicate diffuse large-vessel vasculitis formation: a retrospective study.

BACKGROUND

This study aimed to identify new characteristics of elderly onset large-vessel vasculitis (EOLVV) by focusing on human leucocyte antigen (HLA) genotype, polymyalgia rheumatica (PMR), and affected vascular lesions observed on positron emission tomography/computed tomography (PET/CT) imaging.

METHODS

We retrospectively studied 65 consecutive Japanese patients with large-vessel vasculitis (LVV) who had extracranial vasculitis lesions and underwent PET/CT imaging. PET/CT images were assessed using the semi-quantitative PET visual score of each affected vessel, and the PET vascular activity score (PETVAS) and number of affected vessels were calculated. Subjects were subsequently grouped based on age at onset, superficial temporal artery (STA) involvement, and presence of PMR and compared each group according to HLA genotype. Unsupervised hierarchical cluster analysis was used to identify the patients with similar characteristics in terms of affected vascular lesions detected through PET/CT imaging. The clinical characteristics and PET/CT findings of the population newly identified in this study were examined.

RESULTS

Twenty-seven patients with EOLVV did not meet the American College of Rheumatology 1990 criteria for giant cell arteritis (GCA) and Takayasu arteritis and were considered as unclassified EOLVV (UEOLVV). The unsupervised hierarchical cluster analysis revealed that UEOLVV with PMR and large-vessel GCA (LV-GCA) formed a cluster of LVV with GCA features (i.e., PMR and/or STA involvement) when restricted to patients who were HLA-B52-positive. Patients who were HLA-B52-positive with LVV and GCA features had similar clinical characteristics and patterns of affected vessels and presented with diffuse LVV lesions. HLA-B52-positive patients who had LVV with GCA features also presented with higher PETVAS, more affected vessels, and lower rates of biologics usage and relapse compared to HLA-B52-positive patients with TAK.

CONCLUSIONS

Patients who had UEOLVV with PMR had similar characteristics to patients with LV-GCA. Patients who were HLA-B52-positive and had LVV with GCA features presented with diffuse vascular lesions and may comprise a core population of Japanese patients with EOLVV. The findings of HLA-B52 positivity and diffusely affected vessels in patients with EOLVV can be considered as suspicious findings of LV-GCA.