Page 127 - Recognizing axial spondyloarthritis - Janneke de Winter

feasible in this cohort. However, a potential drawback of using SpA classification criteria is that the a priori probability of fulfilling the ASAS axial SpA, ESSG, and/ or Amor criteria is higher in these first-degree relatives than in populations used to develop these criteria (34,40,42), because the study design implies that all participants have a family history of SpA and that half of them are HLA–B27 positive. Nevertheless, the fact that first-degree relatives fulfilling the ASAS axial SpA and/or ESSG classification criteria had significantly higher scores for disease activity than those not fulfilling these criteria adds to the rightness of using these criteria for this particular study design.
Importantly, some key features of SpA, including the presence of peripheral disease or extra-articular manifestations and increased parameters of inflammation, were only rarely observed in first-degree relatives and also did not differ between those who did and those who did not fulfill the ASAS axial SpA and/ or ESSG classification criteria. Again, we cannot entirely exclude the possibility of a selection bias, since we recruited seemingly healthy first-degree relatives and since axial symptoms may be less frequently recognized by first-degree relatives and physicians than are those of peripheral or extra-articular disease. Alternatively, this observation may suggest that axial disease precedes other SpA manifestations during disease development.
Strikingly, 4 of 34 first-degree relatives (12%) not fulfilling the ASAS axial SpA and/or ESSG classification criteria had imaging abnormalities suggestive of SpA, 1 with syndesmophytes on cervical spine radiography and 3 with bone marrow edema on MRI of the SI joint. Moreover, we did observe abnormalities on MRI of the SI joint, but we did not observe spine abnormalities on MRI. These individuals did not have any other clinical or biologic signs of SpA. These imaging abnormalities might represent the subclinical phase. Future follow-up will determine whether and which of these first-degree relatives will develop clinically established SpA. We will be able to address the relevance of these observations in the near future by 1) increasing the cohort size, which will allow us to replicate the current findings, and 2) following up these first-degree relatives prospectively, focusing especially on those with imaging abnormalities only. A key question is whether these imaging abnormalities predict clinical signs and symptoms of SpA. One possibility is that a positive MRI, even when “highly suggestive of axial SpA,” can be observed more frequently in “normal individuals” than previously thought. Alternatively, some of the MRI abnormalities may indeed be the first subclinical signs of axial SpA.
More female than male first-degree relatives fulfilled the ASAS axial SpA and ESSG classification criteria. This observation could be a numerical issue which would be clarified by increasing the sample size. If this sex imbalance is confirmed, however, it might relate to this study’s classification of first-degree relatives as having SpA (including different subtypes) rather than as having AS. Indeed, the male:female ratio in AS is 2–3:1, but in other subtypes, including non-radiographic axial SpA, the number of females was higher (43,44). Alternatively, male sex may be associated with progression to full-blown disease, which is one of the research
THE PRE-SPA COHORT
125
SEVEN