Validation of a diagnostic score for the diagnosis of autoinflammatory diseases in adults

Most autoinflammatory disorders typically come out in the pediatric population, although a limited number of patients may experience disease onset during adulthood. To date, a late disease onset has been described only in familial Mediterranean fever, caused by mutations in the MEFV gene, and in tumor necrosis factor receptor-associated periodic syndrome, caused by mutations in the TNFRSF1A gene. The relative rarity and lack of information on adult-onset autoinflammatory diseases make it likely that mutations will be found in an even smaller percentage of cases. With the aim of improving the genetic diagnosis in adults with suspected autoinflammatory disorders, we recently identified a set of variables related to the probability of detecting gene mutations in MEFV and TNFRSF1A and, in addition, we have also proposed a diagnostic score for identifying those patients at high risk of carrying mutations in these genes. In the present study we evaluated the preliminary score sensitivity and specificity on a wider number of patients in order to validate the goodness of fit of the model. Two hundred and nineteen consecutive patients with a clinical history of periodic fever attacks were screened for mutations in MEFV and TNFRSF1A genes; detailed information about family/personal history and clinical manifestations were also collected. For the validation of the score we considered data both from the 110 patients used to build the preliminary diagnostic score and from the additional 219 patients enrolled in the present study, for a total number of 329 patients. Early age at disease onset, positive family history for recurrent fever episodes, thoracic pain, abdominal pain and skin rash, which are the variables that had previously been shown to be significantly associated with a positive genetic test result (12), were used for validation. On univariate analysis the associations with a positive genetic test were: age at onset (odds ratio [OR] 0.43, p=0.003), positive family history for recurrent fever episodes (OR 5.81, p < 0.001), thoracic pain (OR 3.17, p < 0.001), abdominal pain (OR 3.80,p < 0.001) and skin rash (OR 1.58,p=0.103). The diagnostic score was calculated using the linear combination of the estimated coefficients of the logistic multivariate model (cut-off equals to 0.24) revealing good sensitivity (0.778) and good specificity (0.718). In conclusion, our score may serve in the diagnostic evaluation of adult patients presenting with recurrent fever episodes suspected of having an autoinflammatory disorder, helping identify the few subjects among them who may be carriers of mutations in IIIEFV and TNFRSF1A genes