A regression predictive model for QuantiFERON-TB Gold Plus® indeterminate results in immunosuppressed patients

Abstract

Background: Screening for latent tuberculosis infection using Interferon-Gamma Release Assays is a routine procedure prior to the initiation of anti-tumor necrosis factor (TNF) biotherapy or immunosuppressive therapy. However, indeterminate results are relatively frequent and are an obstacle to treatment initiation. Aim: The aim of this cross-sectional study was to estimate the frequency of indeterminate QuantiFERON-TB Gold Plus® test results in Tunisian patients, and to analyze the potential clinico-biological risk factors associated with these indeterminate results. Methods: Whole blood samples from 712 patients being monitored for autoimmune diseases and candidates for anti-TNF biotherapy or switch of immunosuppressive therapy were used to screen for latent tuberculosis infection using the QuantiFERON-TB Gold Plus® test. Based on literature background, the following variables were tested for the association with indeterminate results: gender, age, diabetes, immunosuppressive drugs, lymphocyte count, Neutrophil-to-lymphocyte ratio, serum albumin, and estimated glomerular filtration rate. Results: The QuantiFERON-TB Gold Plus® test was negative in 572 (80.3%) patients, positive in 106 (14.9%), and indeterminate in 34 (4.8%) cases. Positive results were significantly associated with a family history of confirmed and treated tuberculosis, OR (95% CI) = 52 (20.2–134.3). The use of immunosuppressive drugs and duration of treatment were significantly associated with the occurrence of indeterminate results: OR (95% CI) = 24.5 (5.8–103) and OR (95% CI) = 1.004 (1.002–1.007), respectively. Biologically, lymphopenia, hypoalbuminemia, and decreased estimated glomerular filtration rate were significant risk factors for indeterminate results: p = 5 E-6, p = 4.3 E-4, and p = 0.002, respectively. Thus, a multiple logistic regression model based on these three biological parameters enabled us to develop a predictive score for indeterminate results with a sensitivity of 91.2% and a specificity of 99.9%, AUC = 0.9964 (0.9917–1), p = 2.8 E-52. Conclusion: Immunosuppressive therapy, lymphopenia, hypoalbuminemia, and kidney failure appeared to be risk factors for indeterminate QuantiFERON-TB Gold Plus® results.