Augmenting the Jeffrey Modell Criteria for Optimising Screening for Inborn Errors of Immunity

Jeffrey Modell warning signs have been used to guide screening for inborn errors of immunity (IEIs) for over three decades now, but may lack sensitivity for noninfectious presentations. Our study explored JMF signs along with additional signs and some laboratory parameters and assessed their effectiveness for guiding molecular screening for IEIs. We retrospectively analysed 100 patients suspected of IEI referred to ICMR-National Institute of Immunohaematology from immunology clinics who underwent whole-exome sequencing (WES), with pathogenic (n = 50) and no variants (n = 50). We evaluated clinical presentations other than JMF-autoimmunity, autoinflammation, vaccine complications, fever with bi-cytopenia, consanguinity and laboratory parameters-absolute lymphocyte count, low naïve T cell (%) and hypogammaglobulinaemia at initial presentation. Thirty-two models with different combinations of additional signs along with JMF score were evaluated based on the AIC score to determine the best model. The model was validated on WES results of 219 patients. The JMF scores of the pathogenic/likely pathogenic group (3.54 ± 1.92) were significantly higher than those of the no variants group (1.34 ± 1.28) (p < 0.05). The best model, inclusive of consanguinity, presence of autoimmunity, vaccine complications, hypogammaglobulinaemia and low naïve T cell (%) with JMF, had a sensitivity and specificity of 82% and 67%, respectively, compared to sensitivity and specificity of 84% and 61% for JMF alone. Our study shows that the presence of any two among parental consanguinity, presence of autoimmunity, vaccine complications and low naïve T cell (%) improves the specificity of JMF criteria in guiding molecular screening for IEIs.