How far has immunotherapy for type 1 diabetes progressed?

medwireNews Hot Topic Reviews provide up-to-date overviews of fast-moving areas of research to help healthcare providers keep abreast of the latest developments that may influence patient care.

Given the huge and unrelenting glucose management demands of type 1 diabetes, it is not surprising that prevention, or at the very least early stabilization, of the condition has long been a holy grail for researchers.

In this Hot Topic Review, we provide a whistle-stop tour of the many immunotherapies that have been tested to date, recount the successes and failures, and outline what’s on the horizon.

What is the rationale for immunotherapy in type 1 diabetes?

Type 1 diabetes is believed to result from the autoimmune destruction of beta cells, a process seemingly driven largely by T cells in its earliest stages, but with evidence for the involvement of B cells, cytokines, and other important elements of the immune response.

Thus, interventions that dampen the immune response by directly targeting effector cells or boosting regulatory cells will in theory prevent or at least slow the destruction of beta cells. A multitude of immune-modulating therapies have been developed for other autoimmune diseases and cancers, and these have well-established safety profiles, facilitating rapid testing in people with type 1 diabetes; the authors of a 2019 review identified nearly 70 studies in people with diabetes [1].

A second major strand of enquiry is whether antigen-based interventions could restore immune tolerance and halt the attack on beta cells.

In newly diagnosed people this could minimize insulin doses and reduce hypoglycemia risk; for high-risk people with autoantibodies it could delay the time to diabetes diagnosis or even prevent them from ever needing to inject insulin.

How have immunotherapies fared in clinical trials?

General immunosuppressants

Some of the earliest agents to be tested were the glucocorticoid prednisone [2]; azathioprine alone [3, 4] and in combination with prednisone [5]; mycophenolate mofetil [6]; and cyclosporine [7,8]. All of these treatments had positive effects on C-peptide decline, indicating that they slowed the decline in beta-cell function, but this did not necessarily come with improved metabolic control. Three of 20 study participants given azathioprine plus prednisone for at least a year had little or no need for insulin after 2 years of continued therapy, but this was also the case for one of the 20 people in the placebo group.

T- and B-cell targeted immunosuppressants

Most medications in this category tested thus far target T cells, with one of the earliest efforts involving antithymocyte globulin (ATG), which binds to multiple T-cell antigens. In the START trial a dose of 6.5 mg/kg had no significant effect on beta-cell function [9], yet in a later study a low dose of 2.5 mg/kg significantly slowed C-peptide decline [10], with the effect persisting 2 years after the single dose was administered [11].

More recent candidate drugs with specific mechanisms include the fusion protein abatacept, which prevents full T-cell activation by binding to two receptors expressed by antigen-presenting cells. People with new-onset diabetes given monthly abatacept infusions in a phase 2 trial had significantly improved C-peptide responses at the 2-year follow-up, but this was due to a slower rate of decline during the first 6 months only, leading the researchers to suggest that the role of the T cells in type 1 diabetes pathogenesis becomes less critical over time [12]. A phase 2 study in autoantibody-positive relatives of people with type 1 diabetes is ongoing.

Another fusion protein, alefacept, has a portion that binds a receptor expressed by T cells, particularly memory T cells, preventing their full activation, and a portion that binds a receptor on natural killer cells, resulting in the selective depletion of memory T cells. In the phase 2 T1DAL trial, two 12-week courses of alefacept did not significantly improve 2-hour C-peptide in participants with new-onset type 1 diabetes, but there was a trend toward an improvement and several secondary outcomes were significantly improved [13].

×

CD3 is a cell-surface receptor involved in the activation of cytotoxic T cells and T-helper cells, and the anti-CD3 antibodies otelixizumab and teplizumab produced promising results in people with recent-onset diabetes in phase 2 trials [14, 15]. But at phase 3 neither otelixizumab, in the DEFEND trial [16], nor teplizumab, in the Protégé study [17], significantly slowed C-peptide decline. However, exploratory analyses by the Protégé investigators suggested that success could be more likely very early after diagnosis and in children. This led directly to a trial in high-risk relatives of people with diabetes, which reported positive results at phase 2 [18], and a phase 3 trial of teplizumab in recently diagnosed children and adolescents – the PROTECT trial  – is ongoing.

Depletion of B cells with rituximab was tested in a phase 2 trial of new-onset type 1 diabetes published in 2009 [19]; four infusions over 22 days led to significantly slower C-peptide decline over 1 year, confirming an important role for B cells in type 1 diabetes pathogenesis.

Researchers are now looking at suppressing both B and T cells, by giving 4 weeks of rituximab followed by 2 years of abatacept to people at high risk for developing type 1 diabetes. Another ongoing study of note involves the anti-malarial drug hydroxychloroquine, which reduces T-cell activation, also in high-risk individuals.

Other targeted immune-modulating agents

Beyond T and B cells, many of the efforts to modulate the immune response in people with type 1 diabetes have focused on agents targeting specific pro-inflammatory molecules. Two notable trials attempted to promote anti-inflammatory molecules, by giving interferon-α (to boost interleukin [IL]-4 and IL-10) and the endogenous anti-inflammatory protein alpha‐1 antitrypsin, both of which appeared promising in preclinical research. The former study showed positive effects on beta-cell function [20] at the highest interferon-α dose used; the latter – the RETAIN trial – failed to identify a dose likely to be efficacious [21].

Direct antagonism of pro-inflammatory molecules has also produced mixed results. Two trials (published in one paper) used canakinumab and anakinra to block IL-1, but without seeing an effect on beta-cell function [22].

On the other hand, use of etanercept, which blocks tumor necrosis factor (TNF) and is widely used for the treatment of a variety of inflammatory diseases, significantly improved beta-cell function relative to placebo in a 24-week pilot randomized trial of newly diagnosed children [23]. In a similar vein, the TNF inhibitor golimumab is being tested in children and adults with a recent diabetes diagnosis, as is tocilizumab, which targets the pro-inflammatory mediator IL-6, in the EXTEND study.

Other studies showed preserved beta-cell function in adults with recent-onset diabetes given the tyrosine kinase inhibitor imatinib  [24], and those given an anti–IL-21 antibody in combination with liraglutide [25].

Vitamin D

Several trials have investigated the effects of boosting vitamin D in people with or at risk for diabetes, based on its ability to protect beta cells and help promote an anti-inflammatory immune response in preclinical studies.

Some of these clinical trials reported positive outcomes, such as enhanced T-regulatory cell activity [26], the disappearance of autoantibodies [27], and a slowed decline in beta-cell function [28]. But others showed no effect on C-peptide decline [29, 30], and all studies were small with the largest having 59 participants.

×

Agents to support beta-cell survival and regeneration

Another line of scientific inquiry is focused on medications that help beta cells to remain viable in the face of an autoimmune response. Again, trials of these agents in people with recent-onset diabetes have met with mixed success.

The REPAIR-T1D trial tested sitagliptin alongside the proton-pump inhibitor lansoprazole, aiming to increase levels of glucagon-like peptide-1 and gastrin, which are thought to help beta cells regenerate [31]. However, the combination treatment had no effect on C-peptide decline. Another trial involved gamma aminobutyric acid (GABA), a neurotransmitter with positive effects on beta-cell function and survival in preclinical studies [32, 33]. It again had no significant effect on diabetes progression, although a question mark remains over the study limitations, including an FDA-mandated limitation on the dose used. At least one ongoing study is testing a higher dose.

A notable success, albeit in a preliminary study of a small number of participants, is the antihypertensive calcium-channel blocker verapamil, which inhibited beta-cell apoptosis and promoted survival in preclinical studies. In a phase 2 trial it significantly slowed C-peptide decline during 12 months of treatment [34] and follow-up of the participants for an additional year showed stable beta-cell function in those who continued to take verapamil versus a marked decline in those who stopped [35].

Transplantation and related therapies

In 2007, researchers published a paper in JAMA describing a phase 1/2 study in which they isolated peripheral hematopoietic stem cells from 15 people with type 1 diabetes diagnosed within the preceding 6 weeks, and reinfused them after a period of immunoablation [36]. In other words, they aimed to destroy the patients’ self-destructing immune cells and replace them with naïve cells that could mature into self-tolerant cells.

This resulted in 14 of the participants becoming independent of insulin for at least 1 month, and in a later study expanded to 23 people, 20 experienced at least some time free of insulin, and one had remained off treatment for more than 4 years at the time of the publication [37].

Several groups have now replicated these original findings [38–40], and there have been many variations on the theme, most commonly infusions of bone marrow stem cells with or without mesenchymal stromal cells derived from sources such as adipose tissue or umbilical cord blood, but without the preceding immune ablation [41–46]. The majority of these studies produced positive results, usually in the form of a reduced decline in C-peptide; notably, this was the case in people with long-established diabetes, as well as in the newly diagnosed.

Other approaches involved isolating, expanding, and reinfusing patients’ own T-regulatory cells, which showed promising results [47, 48]; giving granulocyte colony-stimulating factor (GCSF) alone to mobilize stem cells, which failed to influence diabetes progression [49]; and “re-educating” patients’ immune cells by extracting them and briefly incubating them with umbilical cord blood-derived multipotent stem cells [50]. This latter procedure allows interaction with the autoimmune regulator protein, promoting the elimination of self-reactive immune cells, and during 40 weeks of follow-up resulted in significant improvements in C-peptide in 12 patients with long-standing type 1 diabetes, including six who had no residual beta-cell function at baseline.

None of these alternative methods resulted in patients achieving sustained insulin independence, however, although all the studies were small, including those testing stem cell transplantation with immunoablation.

Antigen-based therapies

The presence of autoantibodies prior to the onset of clinical type 1 diabetes offers both obvious targets for intervention with antigen-based therapies, and a clearly defined group of people who could benefit from treatment.

Prophylactic subcutaneous insulin was an early promising example [51], but the studies following it produced more mixed results. In the DPT-1 trial neither subcutaneous [52] nor oral insulin [53] were able to prevent or delay the onset of type 1 diabetes in people at high risk, and a trial of nasal insulin produced similar results [54].

A hypothesis-generating subgroup analysis of the DPT-1 oral insulin arm suggested that people with confirmed insulin autoantibodies might benefit from the intervention, but a follow-up study in this subgroup was also negative, with the possible exception of a response in a subgroup with the poorest beta-cell function [55].

However, the oral insulin dose used in this study was 7.5 mg/day. The Pre-POINT study investigators conducted a pilot trial of oral insulin in children at high genetic risk but without islet autoantibodies, the results of which suggested that a dose of 67.5 mg/day would be necessary to provoke an immune response in the majority of children [56].

The subsequent phase 1/2 Pre-POINT-Early study, however, found that 67.5 mg/day oral insulin did not in fact provoke an immune response in antibody-negative children at high risk for diabetes [57]. A very similar TrialNet study in children and adults is in progress and the much larger GPPAD-POInT study, not scheduled to complete until 2025, is examining the effects of the higher dose on autoantibody status and progression to type 1 diabetes from infanthood. Another trial of nasal insulin, using a much higher dose than the previous negative study, has reportedly not met its primary endpoint of a reduced risk for type 1 diabetes diagnosis.

An additional notable in-progress study is the intervention part of the Fr1da population screening study, in which children with islet autoantibodies will receive oral insulin at 7.5 or 67.5 mg/day or placebo.

A few recent studies have introduced variations on the prophylactic insulin theme. The largest of these, looking at the insulin B peptide, showed no effect on beta-cell function [58], but a much smaller study of this peptide from the Immune Tolerance Network showed a “lasting, robust” effect on T-regulatory cells, leading the researchers to recommend further studies of the optimal dose, timing, and other relevant parameters that might result in a successful intervention [59]. Another study tested a DNA plasmid encoding proinsulin, and found this significantly improved C-peptide levels and reduced the numbers of proinsulin-reactive cytotoxic (CD8+) T cells [60].

Researchers have also looked at treatment with an alternative autoantigen, in the shape of glutamic acid decarboxylase (GAD) in a standard vaccine formulation with alum (GAD-alum). There were initial promising results when administered to people with new-onset diabetes [61], but a phase 3 trial failed to confirm this benefit [62].

Nevertheless, several ongoing trials are continuing to test GAD-alum in people with new-onset type 1 diabetes in combination with other interventions including GABA, etanercept plus vitamin D, and vitamin D alone, the latter being a phase 3 trial specifically in people with the HLA DR3-DQ2 haplotype, following phase 2 results suggesting benefits in this subgroup [63].

In a nutshell

Multiple therapies have been tested, spilt broadly into increasingly refined attempts to dampen the immune response and efforts to induce immune tolerance. A respectable proportion of interventions have met with success, but so far only partially or transiently, or in a small proportion of participants.

This is tending to push research into specific subgroups, to an earlier phase of the disease, or to multiple therapies, with one study for example offering combination treatment with ATG, GCSF, IL-2, etanercept, and exenatide.

However, the field is at a relatively early stage, with the vast majority of published studies being phase 2 or earlier. Not helping is the still poorly defined and seemingly highly heterogeneous pathogenesis of type 1 diabetes, with recent findings as likely to reveal more complexity as to clarify known processes.

That said, the results so far support the notion that immunotherapy – with the right interventions given to the right people at the right time – could one day stop type 1 diabetes in its tracks.

medwireNews is an independent medical news service provided by Springer Healthcare. © 2022 Springer Healthcare part of the Springer Nature group