LAUSR

We read with interest the published manuscript by Sagiv et al1 highlighting the place of interleukin 22 (IL22) in the diagnosis of axial spondyloarthritis (SpA). The authors found that IL22 levels were elevated in patients with axial SpA (defined SpA: 9.7 pg/mL [0.145] vs healthy: 1.5 pg/ mL [0.16.7]; P < .001). During reading the manuscript, we found several limitations. First, the authors did not specify noninclusion criteria for all participants. Indeed, several studies demonstrated that IL22 was implicated in cancerpromoting2 and hepatitis.3 Consequently, these diseases can be responsible for an elevation of IL22. Moreover, the role of this cytokine in asthma and allergic airway diseases is still debated.4,5 We believe these diseases can influence IL22 values and should be added to the noninclusion criteria of the study to accurately evaluate the IL22 level in SpA. Extraarticular manifestations have not been collected in the study. Several studies demonstrated that IL22 was higher in patients with uveitis, psoriasis, and inflammatory bowel disease.68 Therefore, the assessment of these manifestations would have been interesting for the evaluation of IL22 in SpA. In their manuscript, the authors included 50 patients divided into 3 groups: (a) 29 patients with wellestablished axial SpA diagnosed according to 1984 modified New York criteria for ankylosing spondylitis or the Assessment in Spondyloarthritis (ASAS) classification criteria; (b) 7 patients with possible axial SpA; and (c) 14 patients judged as having an alternative diagnosis. The type of these alternative diagnoses was not mentioned in the result section. Further, the authors did not indicate who made the diagnosis of SpA in the methodology: a rheumatologist or medical staff. The criteria used to classify the patients into the second or third group were unclear. Inclusion criteria for healthy controls were also not specified. Further, the authors found that the cutoff IL22 value of 5 pg/ mL was able to discriminate patients with definite axial SpA from those with alternative diagnoses to axial SpA. The sensitivity and specificity were 0.68 and 0.86, respectively. This finding is considerable. Nevertheless, the authors did not specify the statistical test used to determine this cutoff in the method section. We suppose this cutoff was determined using the receiver operating characteristic curve and modified New York criteria for ankylosing spondylitis or the ASAS classification criteria as the gold standard for axial SpA diagnosis. The degree of accuracy of a diagnostic test depends on the area under the curve (AUC). It is classified as poor (AUC between 0.50 and 0.70), useful for some purposes (AUC between 0.70 and 0.90), and high (AUC > 0.90).9 The authors highlighted the lack of useful criteria for axial SpA diagnosis. The diagnosis of SpA is usually based on the opinion of an experienced rheumatologist. They suggested that IL22 can serve as an independent biomarker for discriminating axial SpA from its noninflammatory mimickers. They concluded that IL22 could be of practical utility. All these limitations make the generalizability of this study's results difficult. Other studies assessing IL22 in patients with axial SpA are necessary.