Page 100 - Recognizing axial spondyloarthritis - Janneke de Winter

CHAPTER SIX
(nr-axSpA), we additionally tested HLA-C*07 in 83 nr-axSpA patients of GESPIC: HLA-C*07 was negative in 59.0% (p=0.09). HLA-B*27 positivity was equal in HLA-C*07 positive and negative axSpA patients: 49 of the 65 HLA-C*07 positive axSpA patients were HLA-B*27 positive, compared to 109 of the 131 HLA-C*07 negative ax-SpA patients (75.4% vs. 83.8%, p=0.16).
We next used the SPACE cohort to test if HLA-C*07 prevalence is lower in early axSpA patients than in CBP controls. Sixty-six of the 94 axSpA patients were HLA-C*07 negative compared to 96 of the 216 CBP patients (70.2% vs. 44.4%, p<0.0001, Figure 2). Of the 23 early axSpA patients fulfilling the mNY criteria, 78.3% was HLA-C*07 negative. Of the 71 axSpA patients with nr-axSpA, 65.8% was HLA-C*07 negative. Also here, absence of HLA-C*07 was similar in HLA*B27 positive and negative patients: 22 of the 28 HLA-C*07 positive axSpA patients were HLA-B*27 positive compared to 59 of the 66 HLA-C*07 negative axSpA patients (78.6% vs. 89.4%, p=0.17). Next, we performed multivariate logistic regression in the complete SPACE population to further test whether absence of HLA-C*07 is associated with an axSpA diagnosis independently from HLA-B*27. The final model, having patients with an axSpA diagnosis and fulfilling the ASAS criteria as outcome, included inflammatory back pain, CRP level, HLA-C*07, HLA-B*27, psoriasis, uveitis and arthritis occurrence ever. This analysis showed that HLA-C*07 negativity was associated with a diagnosis of axSpA (OR 2.2, p=0.038) independently from HLA-B*27 (OR 32.8, p<0.0001). Collectively, these data in two independent cohorts demonstrate that HLA-C*07 prevalence is lower in axSpA patients than in controls, independently from HLA-B*27 prevalence.
Association between HLA-C*07 negativity and disease phenotype in axSpA
We explored if HLA-C*07 negativity was associated with a specific sub-population of axSpA patients. Table 1 shows characteristics of HLA-C*07 positive and negative axSpA patients in GESPIC. Of the 65 HLA-C*07-positive axSpA patients, 26 were male compared to 74 of the HLA-C*07 negative axSpA patients (40.0% vs. 56.5%, p=0.030). The HLA-C*07 positive axSpA patients had a mean back pain duration of 42.4±27.5 months compared to a mean back pain duration of 53.7±24.4 months in the HLA-C*07-negative axSpA patients (p=0.022). Of the 65 HLA-C*07-positive axSpA patients, 11 had psoriasis compared to 9 of 131 HLA- C*07-negative axSpA patients (16.9% vs. 6.9%, p=0.029). The mean CRP level of HLA-C*07-positive axSpA patients was 7.8±13.9 compared to 12.6±20.5 in HLA- C*07-negative axSpA patients (p=0.02). Thirty-one of the 65 HLA-C*07-positive axSpA patients fulfilled the mNY criteria compared to 82 of the 131 HLA-C*07- negative axSpA patients (47.7% vs. 62.6%, p=0.047). There were no differences in mean age, the prevalence of a positive family history, peripheral arthritis, enthesitis, IBD, uveitis, mean BASDAI or mean baseline mSASSS scores between HLA-C*07-positive and negative axSpA patients (Table 1). Both a univariate and multivariate logistic regression analysis within the axSpA patients showed that psoriasis was conversely correlated with HLA-C*07 negativity (OR 0.21, p=0.004),
98