It is currently unknown how long GCA should be treated with tocilizumab. In the first randomized controlled trial, the biologic agent was stopped after 52 weeks. We therefore studied what proportion of patients relapsed, when relapses occurred and whether factors might predict relapse after tocilizumab treatment discontinuation.

All patients in the tocilizumab arm who had received a 52-week treatment were evaluated. In case of lasting remission, magnetic resonance angiography (MRA) was performed and sera were taken to search for biomarkers associated with subclinical disease activity.

Seventeen of 20 patients randomized to the tocilizumab treatment arm were in lasting remission without any co-medication at week 52. Mean follow-up after study end was 28.1 months (range 17-44). Eight patients relapsed after a mean of 6.3 months (range 2-14) (six within the first 5 months, two patients at months 13 and 14, respectively). Relapsing patients were younger and showed more signs of mural enhancement in MRA compared with non-relapsing patients. MRA documented low-intensity vessel wall signals in all subjects. No morphological changes such as formation of aneurysm of aorta occurred. Biomarkers in sera did not indicate subclinical disease activity: levels of IL-6, MMP-3, soluble TNF receptor 2, soluble CD163, soluble intercellular adhesion molecule-1 and Pentraxin-3 did not differ from matched healthy controls.

The data show that a 52-week treatment with tocilizumab induces a lasting remission that persists in half of the patients after treatment stop. None of the clinical, serological or MRA findings qualify to predict relapse. Remarkably, MRA revealed a persisting wall enhancement of the descending aorta.