Giant cell arteritis (GCA) rapidly responds to high-dose corticosteroids. However, smouldering arterial inflammation can persist despite the absence of symptoms and altered acute phase reactants. In patients that are refractory, symptoms relapse during steroid tapering and vascular complications may develop. The erythrocyte sedimentation rate (ESR) and C-reactive protein (CRP) level are not sensitive enough markers to detect refractory disease.1 The neuroendocrine system regulates innate and acquired immune responses, influencing cytokine synthesis and limiting tissue damage via release of neurotransmitters and peptides in peripheral tissues. Chromogranin A in particular is a candidate marker linking neurogenic inflammation and vascular inflammation.2 We investigated by ELISA, as described previously,3 the levels of circulating chromogranin A in 40 consecutive patients with GCA (table 1), classified according to the 1990 American College of Rheumatology (ACR) criteria,4 and 40 healthy control participants. All patients had been initially treated with oral prednisone (0.75–1 mg/kg). The initial dose was gradually tapered, with the intention to reach a maintenance dose of 5–7.5 mg prednisone after 7–8 months.5 The mean follow-up time at enrolment was 29 months (range 4–118). We evaluated the response to treatment in 36 patients who had been followed for at least 8 months. We found significantly (Mann–Whitney test; p<0.001) higher levels of chromogranin A in 10 patients with refractory disease (mean (SD) 667.08 (429.16) ng/ml), defined as those whose disease relapsed before the achievement of maintenance prednisone 10 mg/day,6 as compared with responder patients (157.24 (231.96) ng/ml) and healthy controls (40.52 (21.62) ng/ml) (fig 1A). Seven patients were taking immunosuppressant drugs, all of them belonging to the refractory group, including three patients with active aortitis with aneurysm, detected by contrast-enhanced MRI or CT.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1 Elevated chromogranin A levels in giant-cell arteritis patients associate with response to treatment. Upper panel: the blood concentration of circulating chromogranin-A (CgA) was evaluated by ELISA in the serum of patients with responsive and refractory giant cell arteritis (GCA) and of an equal number of sex-matched and age-matched healthy controls. We observed significantly higher levels of circulating chromogranin A among patients with refractory disease. Data are shown as box plots, with first, second (median) and third quartiles defining box area and vertical lines extending to the 5th and 95th percentile. Lower panel: negative chromogranin A immunostain in temporal arteries samples. Upper left: histology of giant cell arteritis with intense inflammation and numerous giant cells (100× original magnification, hematoxilyn and eosin). Upper right: intense and positive staining with anti-chromogranin A 7D1 antibody in normal pancreas islet (250× original magnification). Lower left: negative immunostaining with anti-chromogranin A LK2H10 antibody in vessel with giant cell arteritis lesions (40× original magnification) Lower right: negative immunostaining with anti-chromogranin A 7D1 antibody in vessel with giant cell arteritis lesions (100× original magnification).

We found no correlation between chromogranin A and ESR and CRP levels, or with the dose of corticosteroids at the time of sample collection in responder and non-responder patients (mean prednisone dose: 9.5 (8.8) and 8.1 (4.3) mg/day respectively). Moreover we found no association between chromogranin A levels and age, sex and the presence of ischaemic optic neuropathy or polymyalgia rheumatica. We stained temporal artery samples from 10 randomly selected patients with three monoclonal anti-chromogranin A antibodies, recognising diverse epitopes of the molecule. Immunostaining was negative in all cases, irrespective of the levels of circulating chromogranin A (range 25–1189 ng/ml). Human pancreas was the positive control (fig 1B). This suggests that the vessel wall is not the source of chromogranin A in GCA. Adrenal glands are an alternative source, since stressful stimuli induce the release of chromogranin A in the blood of normal but not adrenalectomised participants.7

In conclusion, we show that high blood levels of chromogranin A identify refractory patients with GCA, reflecting persistent arterial inflammation despite treatment, even in the absence of symptoms and of altered ESR and CRP. This is in agreement with previous reports of a role for chromogranin A and its proteolytic fragments in vascular inflammation;8^–10 chromogranin A is therefore a candidate marker to assess response to treatment during follow-up and to identify patients who would benefit from higher doses of corticosteroids or immunosuppressant drugs.