Aim To evaluate the long-term effect of intra-lymphatic administration of GAD-alum and a booster dose 2.5 years after the first intervention (DIAGNODE Extension study) in patients with recent-onset type 1 diabetes. Methods DIAGNODE-1: Samples were collected from 12 patients after 30 months who had received 3 injections of 4 mu g GAD-alum into a lymph node with one-month interval. DIAGNODE Extension study: First in human, a fourth booster dose of autoantigen (GAD-alum) was given to 3 patients at 31.5 months, who were followed for another 12 months. C-peptide was measured during mixed meal tolerance tests (MMTTs). GADA, IA-2A, GADA subclasses, GAD(65)-induced cytokines, PBMCs proliferation and T cells markers were analyzed. Results After 30-month treatment, efficacy was still seen in 8/12 patients (good responders, GR). Partial remission (IDAA1c < 9) had decreased compared to 15 months, but did not differ from baseline, and HbA1c remained stable. GAD(65)-specific immune responses induced by the treatment started to wane after 30 months, and most changes observed at 15 months were undetectable. GADA subclasses IgG2, IgG3 and IgG4 were predominant in the GR along with IgG1. A fourth intra-lymphatic GAD-alum dose to three patients after 31.5 months gave no adverse events. In all three patients, C-peptide seemed to increase the first 6 months, and thereafter, C-peptide, HbA1c, insulin requirement and IDAA1c remained stable. Conclusion The effect of intra-lymphatic injections of GAD-alum had decreased after 30 months. Good responders showed a specific immune response. Administration of a fourth booster dose after 31.5 months was safe, and there was no decline in C-peptide observed during the 12-month follow-up.