An open-label study of zoledronic acid (Aclasta 5 mg iv) in the treatment
of ankylosing spondylitis
Spinal lesions in ankylosing spondylitis (AS) include osteitis and spondylodiscitis—bone lesions characterised by high signal on T2w fat-suppressed or STIR magnetic resonance (MR) imaging.1 Pamidronate treats osteitis and improves spinal symptoms in non- steroidal anti-inflammatory drug (NSAID)-refractory AS,2 but treatment gains are modest. Zoledronic acid (Aclasta/Reclast; ZA) is more ‘potent’ than pamidronate,3 but it is unknown whether ZA improves osteitis in AS. We evaluated whether ZA reduces osteitis lesions in AS using an MR index of spinal oste- itis4 previously shown to correlate with, and be responsive to treatment of, AS disease activity measures.5
Patients with AS (modified New York criteria,6 >18 years old, Bath AS Disease Activity Index (BASDAI) ≥4.15, anti-tumour necrosis factor α (anti-TNFα)-naive) were recruited based on having osteitis in two or more discovertebral units (DVUs; SPARCC definition on fat-suppressed MR images),4 as determined by a local radiologist (SJG or PWB). Eligible patients received intravenous ZA 5 mg following local site guidelines. Ethics and MHRA (2007-000087-25) approval was obtained. MR images from pre-treatment and 3 and 6 months after treatment were anonymised, coded and then scored by two radiologists (PO’C and AG) experienced in SPARCC scoring, blind to the timing of scans and each others’ scores. Scoring was done on the most abnormal six DVUs on three contiguous sagittal slices of STIR images. An average of the two scorer’s total scores for each study was taken.
Clinical data were obtained pre-treatment and 3 and 6 months after ZA (NSAID use chart, spinal pain index (VAS), BASDAI, Bath AS Functional Index (BASFI) and Maastricht AS Enthesitis Score (MASES)).7
Of 11 patients recruited, 7 (6 male/1 female; age range 35– 54 years) qualified for ZA treatment (four patients not having ≥2 scorable DVUs on initial MR). No patients were taking immunosuppressants (table 1).
Baseline SPARCC scores ranged between 16–66/108. SPARCC sores improved in 4/7 patients by 3 months after treat- ment (3/4 by at least five points—the minimally important change (MIC),8 all sustained improvement through 6 months). At 6 months, 6/7 patients had improved SPARCC scores com- pared with baseline (by 19–76% with 4/7 patients by at least the MIC with another two improving by 4 and 4.5 points). The
MR observers’ scores were identical for 79% of DVUs, differing in 21% by a maximum of one only (table 2).
BASDAI improved from 6.3 to 5.1 at 6 months (mean BASFI from 6.0 improving to 5.5 at 3 months and to 4.1 at 6 months). Mean pre-treatment MASES was 5.1, improving to 3.9 at 6months. Median spinal pain (VAS) improved by 55% at 3 months (range 28–63%) sustained in 4/7 patients at 6 months.
To our knowledge this is the first report of the effect of ZA on osteitis in AS. ZA may reduce spinal osteitis in AS and may be associated with improvement in spinal symptoms for up to 6 months. Further controlled studies are required to evaluate clinical improvement in milder AS and axial spondylarthropa- thy; whether ZA-associated osteitis treatment can affect syndes- mophyte formation and whether ZA has a role as adjunctive therapy in patients treated with anti-TNFα.