Investigating Biomarkers for Primary Immunodeficiency Disorders in Patients with Autoimmune Cytopenia

Introduction: Autoimmune cytopenias (AICs) occur in the setting of immune dysregulation and autoimmunity. AICs are known to precede the classical presentation and subsequent diagnosis of primary immunodeficiency disorders (PIDs) in some cases, with a PID spectrum involving B, T, and Treg dysregulation. Conversely, PID patients are also at a markedly increased risk of AICs and patients co-diagnosed with AICs and PIDs face higher mortality rates likely due to delayed diagnosis or improper/delayed therapy. Further, co-diagnosed patients are often refractory to standard treatment. Early screening and diagnosis of PIDs in the setting of AICs could lead to more targeted therapies and potentially better patient outcomes. We intend to investigate biomarkers that differentiate immune dysregulation among those with or without PID in the setting of AIC. Methods: A retrospective analysis comparing clinical history, including symptoms, presence of PID, genetic testing and diagnosis and treatment outcomes was performed on AIC patients in our hospital from 2016 to 2019. We also performed extensive immune phenotyping by flow cytometric analysis. Patients with expansion of CD19hi CD21lo B cells were further evaluated for Tbet+ CD11chi features for potential autoreactivity. Results: Our original cohort of 103 AIC patients included patients who were diagnosed with AIC only (n=52) or co-diagnosed with AIC-PID (n=51). From this cohort, a total of 43 were evaluated for further immune phenotyping, including 35 AIC-PID patients and 8 AIC only patients. Of the AIC-PID patients, 23 (65.7%) had a genetically coded PID, including pathogenic variants in CTLA4, NFKB1, PI3K or partial DiGeorge, while 12 (34.3%) were diagnosed with CVID/CID. Flow cytometry analysis revealed significantly increased T-follicular helper cell (Tfh) population in the AIC-PID group compared to AIC only (p=0.03) and healthy donor (HD) (p=0.004). We saw a similar expansion of CD19hi CD21lo B cells within the AIC-PID group. Conversely, patients with AIC-PID had decreased percentage of T-regulatory cells (Tregs) when compared to AIC only and HD groups. These differences were amplified within the genetically coded AIC-PID patients in Treg and Tfh populations and maintained in CD19hi CD21lo B cells. Further investigation of clinical history within the symptomatic, genetically coded AIC-PID group yielded 3 (33.3%) patients responsive to treatment and 6 (66.7%) patients unresponsive or partially responsive to therapy. Interestingly, nonresponsive patients showed an increased percentage of CD19hi CD21lo B cells. Conclusions: Our analysis demonstrates the potential utility of biomarkers, notably CD19hi CD21lo B cells and their immunological function for detecting underlying PIDs within the setting of AIC patients. Though this B cell population needs further assessment regarding Tbet and CD11c expression, with studies ongoing, these biomarkers may lead to a screening methodology to increase clinical visibility and detection as well as hasten diagnosis of underlying PID and potentially predict response to therapy, therefore increasing effective treatment outcomes for patients co-diagnosed with AIC PID.