Targeting of memory T cells with alefacept in new-onset type 1 diabetes (T1DAL study): 12 month results of a randomised, double-blind, placebo-controlled phase 2 trial

Abstract

Background: Type 1 diabetes results from autoimmune targeting of the pancreatic β cells, likely mediated by effector memory T (Tem) cells. CD2, a T cell surface protein highly expressed on Tem cells, is targeted by the fusion protein alefacept, depleting Tem cells and central memory T (Tcm) cells. We postulated that alefacept would arrest autoimmunity and preserve residual β cells in patients newly diagnosed with type 1 diabetes.

Methods: The T1DAL study is a phase 2, double-blind, placebo-controlled trial in patients with type 1 diabetes, aged 12-35 years who, within 100 days of diagnosis, were enrolled at 14 US sites. Patients were randomly assigned (2:1) to receive alefacept (two 12-week courses of 15 mg intramuscularly per week, separated by a 12-week pause) or a placebo. Randomisation was stratified by site, and was computer-generated with permuted blocks of three patients per block. All participants and site personnel were masked to treatment assignment. The primary endpoint was the change from baseline in mean 2 h C-peptide area under the curve (AUC) at 12 months. Secondary endpoints at 12 months were the change from baseline in the 4 h C-peptide AUC, insulin use, major hypoglycaemic events, and HbA1c concentrations. This trial is registered with ClinicalTrials.gov, number NCT00965458.

Findings: Of 73 patients assessed for eligibility, 33 were randomly assigned to receive alefacept and 16 to receive placebo. The mean 2 h C-peptide AUC at 12 months increased by 0.015 nmol/L (95% CI -0.080 to 0.110) in the alefacept group and decreased by 0.115 nmol/L (-0.278 to 0.047) in the placebo group, and the difference between groups was not significant (p=0.065). However, key secondary endpoints were met: the mean 4 h C-peptide AUC was significantly higher (mean increase of 0.015 nmol/L [95% CI -0.076 to 0.106] vs decrease of -0.156 nmol/L [-0.305 to -0.006]; p=0.019), and daily insulin use (0.48 units per kg per day for placebo vs 0.36 units per kg per day for alefacept; p=0.02) and the rate of hypoglycaemic events (mean of 10.9 events per person per year for alefacept vs 17.3 events for placebo; p<0.0001) was significantly lower at 12 months in the alefacept group than in the placebo group. Mean HbA1c concentrations at week 52 were not different between treatment groups (p=0.75). So far, no serious adverse events were reported and all patients had at least one adverse event. In the alefacept group, 29 (88%) participants had an adverse event related to study drug versus 15 (94%) participants in the placebo group. In the alefacept group, 14 (42%) participants had grade 3 or 4 adverse events compared with nine (56%) participants in the placebo group; no deaths occurred.

Interpretation: Although the primary outcome was not met, at 12 months, alefacept preserved the 4 h C-peptide AUC, lowered insulin use, and reduced hypoglycaemic events, suggesting efficacy. Safety and tolerability were similar in the alefacept and placebo groups. Alefacept could be useful to preserve β-cell function in patients with new-onset type 1 diabetes.

Figure 1. Alefacept trial profile

Figure 2. Population means of change in stimulated C-peptide AUC mean from baseline to 12 months for alefacept and placebo treated subjects

(A) 2-hour AUCs (primary endpoint). (B) 4-hour AUCs (secondary endpoint). Bars represent 95% confidence intervals. P values were calculated using an analysis of covariance with baseline ln(AUC+1) value as a covariate.

Figure 3. HbA1c levels and exogenous insulin use in the alefacept and placebo groups

(A) HbA1c levels (%). (B) Exogenous insulin use (units/kg/day). Bars represent the 95% confidence intervals. Lines connect the mean values across visits for each treatment arm. P values for the change from baseline for both HbA1c and insulin use at week 52 were calculated using an analysis of covariance with baseline level as a covariate.

Figure 4. CD2 expression levels on lymphocyte subpopulations

Frozen PBCMs collected at baseline were analyzed for the mean fluorescence intensity (MFI) of CD2 by flow cytometry. Lymphocyte subpopulations were defined as follows: CD4 naïve (Tn): CD3⁺CD4⁺FoxP3⁻CD127^hiCCR7⁺CD45RA⁺; CD4 central memory (Tcm): CD3⁺CD4⁺FoxP3⁻CD127^hiCCR7⁺CD45RA⁻; CD4 effector memory (Tem): CD3⁺CD4⁺FoxP3⁻CD127^hiCCR7⁻CD45RA⁻; regulatory T cells (Treg): CD3⁺CD4⁺FoxP3⁺CD127^lo; CD8 Tn: CD3⁺CD8⁺CCR7⁺CD45RA⁺; CD8 Tcm: CD3⁺CD8⁺CCR7⁺CD45RA⁻; CD8 Tem: CD3⁺CD8⁺CCR7⁻CD45RA⁻; Naïve B cells: CD19⁺CD27⁻; Memory B cells: CD19⁺CD27⁺. Values are mean±SD.

Figure 5. Absolute cell counts

(A) White blood cells (WBC). (B) Total lymphocytes. (C) CD4 T cells. (D) CD8 T cells. Whole blood was analyzed real-time by flow cytometry in a central clinical laboratory. Data (cells/μL) are mean±SD.

Figure 6. Change in lymphocyte populations over time

Frozen PBMCs collected at baseline and weeks 11, 24, 35, and 52 were analyzed by flow cytometry. Percents of subpopulations (defined in figure 4) from parent populations were determined and standardized to baseline values. (A1-3) CD4⁺ naïve (Tn), central memory (Tcm), and effector memory (Tem) cells. (B1-3) CD8⁺ Tn, Tcm, and Tem cells. (C) CD4⁺ Treg. Antibody panels and gating strategies are detailed in supplementary tables 1 and 2. Values are mean±SD.

Figure 7. Ratios of Treg to naïve and memory T cells

Relative numbers of Treg and CD4⁺ and CD8⁺ Tn, Tcm, and Tem cells (subpopulations as defined in figure 4) were determined by flow cytometry (using the gating strategies described in supplementary table 2), and the ratios of Treg to the indicated T cell subpopulations calculated. (A1-3) Treg/CD4 Tn, Treg/CD4 Tcm, and Treg/CD4 Tem. (B1-3) Treg/CD8 Tn, Treg/CD8 Tcm, and Treg/CD8 Tem. Values are mean±SD.