297-OR: Systematic Review of Immune Interventions Trials in Recent Onset Type 1 and Effect on HbA1c

Abstract

There has been no meta-analysis that addressed the essential issue of efficacy of immunomodulatory agents on HbA1c in patients with recent onset type 1 diabetes (T1D). The aim of this study was to conduct a systematic review and meta-analysis to estimate the treatment effect of different immunotherapeutic strategies vs. placebo on HbA1c in subjects with T1D over the first 2 years of disease. We searched PubMed and Clinicaltrials.gov from 1990 until November 2022 to identify randomized clinical trials (RCTs) and observational investigations that compared the efficacy of immunomodulatory agents in adult and children with recent onset T1D. The study included 35 trials with a total of 2987 participants, 2086 in the treatment groups and 901 in the control groups. We extracted information about year of publication, study and exposure duration, immunotherapy type, no. patients, age, sex, weight, and BMI. HbA1c was evaluated at baseline, and at 3, 6, 9, 12 and 24 months. Funnel plots with confidence regions were used to explore small-study and publication biases. A random-effects model to pool effect sizes Knapp-Hartung adjustments was used to calculate confidence intervals. The evaluation outlined a significantly higher HbA1c in the treatment group at baseline (MD = 0.26 %, 95% confidence interval (CI) 0.0025 to 0.51). Heterogeneity was detected (I2 = 67%; p<0.01; prediction interval -0.71 to 1.23). Small trials with higher HbA1c in the treatment group seem to be underrepresented, however, most of the studies provided non-significant results. In analysis at 12 months (12 trials) results showed a significantly higher HbA1c in the treatment group (MD = 0.13 %, 95% confidence interval (CI) 0.09 to 0.16). This result is almost driven by one study. Results of the metanalysis indicate that overall, in the first 2 years of the disease and irrespective of the immunointervention agent tested, no effect on HbA1c levels could be detected vs. placebo. A few exceptions were observed in the data analysis with some agents. Disclosure P.Pozzilli: None. S.Fallucca: Consultant; Dompé. S.Pieralice: None. M.Marelli: Employee; Dompé. L.Mancini: None.