Background Diffuse idiopathic skeletal hyperostosis (DISH, also known as Forestier's disease or ankylosing hyperostosis) is considered to be a noninflammatory skeletal disorder of unknown, but probably “degenerative” origin. NSAIDs can be helpful in those with appreciable pain and stiffness, but the benefits are often modest and may be limited by toxicity. Hitherto, no more effective pharmacological interventions have been described. This deficiency represents an important area of unmet clinical need.
Objectives To test the efficacy of TNFi in DISH, a small observational study was undertaken.
Methods Eight participants who met clinical and radiological criteria for DISH and had a BASDAI of 6 or higher at outset were treated with a TNFi for 0.33 to 8 yrs., median 1 yr. (ETA 2, ADA 1, GOL 1 CERT 4). They were compared with 12 unmatched “controls” who did not meet treatment criteria at the outset or declined TNFi therapy. The two groups were comparable in age (Median age =58 yrs. for controls and 64 yrs. for the TNFi group, P=0.16, unpaired T-test). Only one of 10 participants who had undergone HLA typing was B27 positive. None had plain X-ray evidence of sacro-iliitis. Responses to treatment and progress were determined by use of serial BASDAI scores (patients blinded to all preceding scores).
Results One of the 8 treated patients was unable to tolerate the TNFi (CERT) due to a rash which resolved promptly on cessation. Of the remaining 7, five responded to TNF inhibition with sustained falls in the BASDAI as can be seen in the graph shown below.
One of the non-responders gained no benefit and the other was only slightly better at 4 months. A sharp reduction in BASDAI similar to that observed in SpA patients was seen during the first few years of treatment in the 5 responders. Unequivocal regression was observed after 4 years in 2 of these 5 participants. In contrast, controls deteriorated progressively from a lower starting BASDAI and slowly approached the baseline in treated patients with scores mostly greater than 6 after several years. In none of the 12 control patients was there a spontaneous decline resembling that observed in TNFi treated patients.
Conclusions The results of this medium-term observational study support the efficacy of TNF inhibitors in the majority of patients with clinically active DISH. Responses were not agent specific, but rather applied to the TNFi class. The extent and duration of the response was similar to that observed with TNFi in SpA. TNFi therapy was well tolerated. It is possible that the TNFi effect was due to pain modulation. Nevertheless, the findings call into question the dogma that DISH is uniformly non-inflammatory. We consider that a RDBPCT can be justified.
Acknowledgements The authors thank UCB Australia Pty Ltd for their generous provision of Cimzia (CERT) for compassionate treatment. We also thank the Fremantle Hospital Drug and Therapeutics Committee for individual patient approval of TNFi for selected patients.
Disclosure of Interest None declared