Large vessel vasculitis in giant cell arteritis and polymyalgia rheumatica – prevalence, outcome and associated factors

Background: Giant cell arteritis (GCA) is the most common primary vasculitis of the elderly and is closely related to polymyalgia rheumatica (PMR). Both conditions may occur separately, simultaneously, or sequentially over time. GCA classically manifests as cranial arteritis, but large vessel vasculitis (LVV) has been recognized as part of the disease spectrum. Diagnosing LVV remains challenging as symptoms may be non-specific and imaging is necessary to establish the diagnosis. Furthermore, PMR may be the only clinical manifestation of GCA. However, the understanding of subclinical GCA in patients with PMR, including its prevalence, risk factors and prognostic significance, is limited.
GCA is associated with severe vascular and ischemic complications such as stroke, arterial stenosis, and the development of aortic aneurysms. The most feared complication of GCA is permanent vision loss, and timely diagnosis and treatment of GCA are crucial to prevent acute and chronic complications. Although fast-track clinics for GCA have reduced the delay in diagnosis, permanent vision loss is still reported in up to 13% of cases.
Established imaging modalities for the diagnosis of LVV include ultrasound, [18F]fluorodeoxyglucose positron emission tomography/computed tomography (PET/CT), CT, or magnetic resonance imaging (MRI). However, standardized criteria defining vasculitis on MRI are lacking. While imaging is well established in the diagnostic work-up of patients with suspected GCA, its role in monitoring disease activity during follow-up and predicting the disease course in GCA after treatment discontinuation remains unclear. As relapses are common in patients with GCA after treatment withdrawal, imaging biomarkers for guiding the timing of treatment stop would be helpful.
Objectives: We first aimed to investigate the prevalence of subclinical GCA in patients with newly diagnosed PMR in the literature and to identify potential patient characteristics associated with subclinical GCA in published cohorts. In a next step, we aimed to explore the hypothesis that undiagnosed subclinical GCA in patients with PMR may lead to GCA-associated vascular damage. In our second manuscript, we aimed to study the risk factors and incidence of permanent vision loss in patients with GCA and to identify obstacles which caused a delay in diagnosis. The third manuscript aimed to identify which parameters on MRI correspond to vasculitis in patients with newly diagnosed large vessel (LV-) GCA. Finally, we addressed the role of PET/CT and MRI to predict relapses after treatment stop in patients with large vessel LV-GCA.
Methods: We systematically searched PubMed, Embase, and Web of Science Core Collection for consecutively recruited cohort studies reporting the prevalence of GCA in steroid-naïve patients with PMR. Potential predictors of subclinical GCA were identified using individual patient data from seven cohorts.
In the frame of our retrospective cohort of patients with GCA, we investigated the hypothesis that a proportion of patients with newly diagnosed GCA and a history of PMR may have already had subclinical GCA at the time of PMR manifestation and compared vascular ultrasound findings (extent of vessel involvement and stenoses) between GCA patients with and without prior PMR. Furthermore, we used our retrospective cohort to examine patient and referral characteristics and trends in the incidence of permanent vision loss over the past 15 years at the University Hospital Basel.
To identify which parameters on MRI correspond to vasculitis in patients with newly diagnosed LV-GCA, we compared MRI findings to PET/CT and/or ultrasound findings on a segment level (axillary segment per side, and the thoracic aorta).
Lastly, in an exploratory cohort study, patients with LV-GCA underwent imaging at the time of treatment discontinuation. Imaging findings of patients who relapsed within 4 months after treatment discontinuation were compared to those who remained in remission.
Results: The pooled prevalence of subclinical GCA across all identified studies was 23%, and 29% in the studies using PET/CT. Inflammatory back pain and absence of lower limb pain remained weak predictors of subclinical GCA after multivariable analysis.
Newly diagnosed patients with GCA and a prior history of PMR had significantly more often LVV (51.0% vs. 25.0%, p<0.001) and stenosis within the vasculitic segments (18.4% vs. 3.1%, p<0.001) on ultrasound compared to patients without prior PMR in our retrospective cohort.
The incidence of permanent vision loss was 17.4% in our institution and did not decline over 15 years. More than half of the patients who suffered from vision loss had experienced non-ocular symptoms related to GCA for a median of 21 days but did not seek medical help until the onset of visual impairment. In multivariable analysis, patients with vision loss were older and reported more frequently jaw claudication.
Vessel wall oedema on diffusion-weighted sequences on MRI corresponded to vasculitic PET/CT findings while pathological vessel segments on MRI had a low agreement with vasculitic ultrasound findings.
None of the examined imaging parameters predicted subsequent relapse after treatment withdrawal in patients with LV-GCA. The number of segments with vasculitic findings on PET/CT and the sum of all maximum standardized uptake value (SUVmax) artery/liver ratios showed a slight tendency to be higher in patients who relapsed; however, this did not reach statistical significance.
Conclusion: The high prevalence of subclinical GCA and the accumulating evidence of the potential impact of subclinical GCA on disease outcome advocates a paradigm shift in the assessment of patients with PMR and supports the implementation of screening strategies for large vessel involvement. Our findings underscore the need to increase public and physician awareness of the potentially devastating consequences of GCA and the importance of early detection and timely medical treatment to further reduce the incidence of ischemic and vascular complications. Lastly, we did not find any parameter on imaging performed at the time of treatment discontinuation which predicted future relapse in patients with LV-GCA. The relevance of vasculitic imaging findings in patients in clinical remission of GCA for the development of aortic aneurysms should be further studied.