To explore T1D heterogeneity, the existence of endotypes, and their possible differential responses to immunotherapy administered at clinical onset, we studied 566 TrialNet clinical trial participants (MMF-DZB, anti-CD20; GAD-alum; CTLA4-Ig; Canakinumab; ATG-GCSF). A bioassay, where participant serum is used to induce transcription in healthy mononuclear cells, and unsupervised cluster analyses were used to identify 2 major groups (Gp1 and Gp2).Proportional analyses of subjects <18 years found Gp1 enriched with participants <14 years old (p=0.03); further Gp1 placebo arm participants exhibited a more rapid C-peptide rate of decline (<-0.15, p=0.02). Pathway analysis of transcriptomic data predicted inflammatory bias in Gp1, that was confirmed by detection of elevated plasma IFNγ, TNFα, and IL-6 levels (p<0.01), and higher memory T-cell abundances. When assessing response to therapy, Gp1 anti-CD20-treated participants exhibited a higher mean C-peptide compared to placebos at 6, 12 and 24 months post-onset (meta-p=0.002 vs 0.27 in Gp2) and a greater delay in the post-onset decline of baseline stimulated C-peptide (11.2 vs 5.4 months). In contrast, Gp2 CTLA4-Ig-treated participants exhibited a higher mean C-peptide compared to placebos at 6, 12, 18 and 24 months (meta p=2e-4 vs. 0.24 in Gp1) and a greater delay in the decline of baseline stimulated C-peptide (8.7 vs 5.8 months). Both Gp1 and Gp2 benefited from ATG treatment, respectively showing 9.5 and 6.7 month delays in the decline in baseline stimulated C-peptide compared to their corresponding placebos. Weighted correlation network analyses of transcriptomic data suggested that varying degrees of immunomodulation occurred in all 6 trials, including those not showing overall efficacy. In line with that observation, Gp2 canakinumab-treated participants showed a 3.4-month delay in the decline in baseline stimulated C-peptide compared to placebos. These data support the existence of two T1D endotypes that exhibit distinct immune phenotypes and therapeutic responsiveness.

Disclosure

A. Bedrat: None. T. Pant: None. S. Jia: None. M. Roethle: None. N.A. Truchan: None. Y. Chen: None. C. Lin: None. M. Hessner: None.

Funding

R01DK121528

Readers may use this article as long as the work is properly cited, the use is educational and not for profit, and the work is not altered. More information is available at http://www.diabetesjournals.org/content/license.